• Save
Demi 2011 proceedings
Upcoming SlideShare
Loading in...5
×
 

Demi 2011 proceedings

on

  • 1,939 views

 

Statistics

Views

Total Views
1,939
Views on SlideShare
1,939
Embed Views
0

Actions

Likes
0
Downloads
0
Comments
0

0 Embeds 0

No embeds

Accessibility

Categories

Upload Details

Uploaded via as Adobe PDF

Usage Rights

© All Rights Reserved

Report content

Flagged as inappropriate Flag as inappropriate
Flag as inappropriate

Select your reason for flagging this presentation as inappropriate.

Cancel
  • Full Name Full Name Comment goes here.
    Are you sure you want to
    Your message goes here
    Processing…
Post Comment
Edit your comment

Demi 2011 proceedings Demi 2011 proceedings Document Transcript

  • University of Banja LukaFaculty of Mechanical Engineering26th - 28th May 2011 DEMI 2011 10th Anniversary International Conference on Accomplishments in Electrical and Mechanical Engineering and Information TechnologyPROCEEDINGS ZBORNIK RADOVA BANJA LUKA, May 2011.
  • PROCEEDINGS ZBORNIK RADOVA University of Banja LukaFaculty of Mechanical Engineering BANJA LUKA, May 2011.
  • ii
  • PROCEEDINGS OF THE 10th ANNIVERSARY INTERNATIONAL CONFERENCE ON ACCOMPLISMENTS IN ELECTRICAL AND MECHANICAL ENGINEERING AND INFORMATION TECHNOLOGY Under patronage of: Ministry of Science and Technology of the Republic of Srpska, Ministry of Industry, Energy and Mining of the Republic of Srpska and City of Banjaluka Publisher: Faculty of Mechanical Engineering Banja Luka For publisher: PhD. Miroslav Rogi , Full Professor Editor in Chief: PhD. Gordana Globo ki-Laki , Associate Professor Organizational board: Gordana Globo ki-Laki , PhD. Associate Professor, Chairman Miroslav Rogi , PhD. Professor Snežana Petkovi , PhD. Associate Professor Zdravko Milovanovi , PhD. Associate Professor Petar Gvero, PhD. Associate Professor Strain Posavljak, PhD. Assistant Professor Darko Kneževi , PhD. Assistant Professor Tihomir Latinovi , PhD. Assistant Professor Valentina Golubovi -Bugarski, PhD. Assistant Professor Milan Tica, MSc. Mechanical Engineering Stevo Borojevi , MSc. Mechanical Engineering Bojan Kneževi , MSc. Electrical Engineering Branislav Sredanovi , BSc. Mechanical Engineering Branislav Jovkovi , BSc. Mechanical Engineering Milivoj Stipanovi and Ljubo Glamo i , PhD. Ministry of Industry, Energy and Mining of the Republic of Srpska Technical processing: Milivoj Stipanovi Circulation: 190 iii
  • iv
  • SCIENTIFIC BOARDBlagojevi Aleksa, Faculty of Mechanical Engineering Banjaluka, B&HBlagojevi Drago, Faculty of Mechanical Engineering Banjaluka, B&HBobrek Miroslav, Faculty of Mechanical Engineering Banjaluka, B&HBojani Pavao, Faculty of Mechanical Engineering Beograd, SerbiaBulatovi Miodrag, Faculty of Mechanical Engineering Podgorica, Montenegro osi Ilija, FTN Novi Sad, SerbiaDaki Pantelija, Faculty of Mechanical Engineering Banjaluka, B&HDolo ek Vlatko, University of Sarajevo, B&H uri kovi Veljko, Faculty of Mechanical Engineering Banjaluka, B&HFilipovi Ivan, Faculty of Mechanical Engineering Sarajevo, B&HGerasim uk G. Vasilj, “KPI”, UkrainiaGruden Dušan, TU Wien, AustriaIvkovi Branko, Faculty of Mechanical Engineering Kragujevac, SerbiaJokanovi Simo, Faculty of Mechanical Engineering Banjaluka, B&HJoviševi Vid, Faculty of Mechanical Engineering Banjaluka, B&HKoji Miloš, Harvard University, USAKostolansky Eduard, University of Cyril and Metodius Trnava, SlovakiaKozi or e, Faculty of Mechanical Engineering Beograd, SerbiaLuka Duško, University of Applied Science, GermanyMaksimovi Stevan, Aeronautical Institute, SerbiaMileti Ostoja, Faculty of Mechanical Engineering Banjaluka, B&HMili i Dragomir, Faculty of Mechanical Engineering Banjaluka, B&HMilutinovi Dragan, Faculty of Mechanical Engineering Beograd, SerbiaNedi Bogdan, Faculty of Mechanical Engineering Kragujevac, SerbiaNinkovi Dobrivoje, ABB Turbo-Systems AG, SwitzerlandOgnjanovi Milosav, Faculty of Mechanical Engineering Beograd, SerbiaPeši Radivoje, Faculty of Mechanical Engineering Kragujevac, SerbiaPlan ak Miroslav, FTN Novi Sad, SerbiaPop Nicolae, North University of Baia Mare, RomaniaRadovanovi Milan, Faculty of Mechanical Engineering Beograd, SerbiaRavano Giambattista, University SUPSI, SwitzerlandSavi Vladimir, FTN Novi Sad, SerbiaSchmied Joachim, Delta JS, SwitzerlandSeok Park Hong, University of Ulsan, KoreaŠljivi Milan, Faculty of Mechanical Engineering Banjaluka, B&HSokovi Mirko, Faculty of Mechanical Engineering Ljubljana, SloveniaStefanovi Milentije, Faculty of Mechanical Engineering Kragujevac, SerbiaStegi Milenko, FSB Zagreb, CroatiaThomeensen Trygve, Norwegian University of Science and Technology, NorwayTodi Velimir, FTN Novi Sad, SerbiaTufek i emo, Faculty of Mechanical Engineering Tuzla, B&HVeinovi Stevan, Faculty of Mechanical Engineering Kragujevac, SerbiaVereš Miroslav, Faculty of Mechanical Engineering, Bratislava, SlovakiaZeljkovi Milan, FTN Novi Sad, SerbiaZrili Ranko, Faculty of Mechanical Engineering Banjaluka, B&HZrni Nenad, Faculty of Mechanical Engineering Beograd, Serbia v
  • vi
  • CONTENTKEYNOTE LECTURES ................................................................................................... 11. Claudio R. Boër SUSTAINABLE INNOVATION AND INNOVATION FOR SUSTAINABILITY: THE EVOLUTION AND INVOLUTION OF PRODUCTION PARADIGMS ........... 32. Giambattista Ravano COOPERATION MODELS BETWEEN UNIVERSITIES AND INDUSTRY IN APPLIED RESEARCH, SWITZERLAND CASE STUDY AND SOME PRACTICAL EXAMPLES ...................................................................................... 53. Radivoje Peši , Stevan Veinovi TRANSPORT ECOLOGY AND GLOBAL CLIMATE CHANGE ............................ 74. Milosav Ognjanovi DESIGN CONSTRAINTS AND ROBUST DESIGN AS THE MODERN APPROACH TO MECHANICAL STRUCTURE DEVELOPMENT DESIGN CONSTRAINTS AND ROBUST DESIGN AS THE MODERN APPROACH TO MECHANICAL STRUCTURE DEVELOPMENT ........................................... 215. Neven Dui RENEWABLE ENERGY AS A DRIVER OF ECONOMIC GROWTH ................. 356. Stojan Petrovi , Božidar Nikoli , Emil Hnatko, Jovo Mr a, Stevan Veinovi MALICIOUS ECOLOGY ON VEHICLES AND TRAFFIC EXAMPLE ................. 55A. MECHANICS AND DESIGN ................................................................................... 577. Michail Leparov, Georgi Dinev, Marieta Jancheva ABOUT RECEIPTE OF VARIANTS OF TECHNICAL OBJECTS ....................... 598. Leparov M., Yancheva M. ABOUT THE INTEGRATION OF ASSEMBLY UNITS ........................................ 659. Darko Kneževi , Aleksandar Milašinovi , Zdravko Milovanovi ANALYSIS OF INFLUENCE OF LENGTH OF DEVELOPMENT OF BOUNDARY LAYER ON FLOW RATE THROUGH RADIAL CLEARANCE WITHIN HYDRAULIC CONTROL COMPONENTS ........................................... 7110. Nebojša Radi , Goran Sekuli , Dejan Jeremi ANALYTICAL-NUMERICAL STRESS ANALYSIS OF SPUR GEARS WITH STRAIGHT TEETH ............................................................................................. 7711. Georgy Dinev, Marieta Yancheva CAD DESIGN OF FLEXIBLE FRICTION COUPLING ....................................... 8312. Dragan Lišanin, Marinko Petrovi , Nenad Grujovi , Jelena Borota COMPARATIVE ANALYSIS OF THE FORMATION OF SMALL GRAIN GUIDANCE ......................................................................................................... 8713. Pavle Stepani , Željko urovi , Aleksa Krošnjar, Aleksandra Pavasovi COMPARISON OF TECHNIQUES FOR DETECTION OF FAILURE ROLLING ELEMENT BEARINGS ....................................................................... 9314. Strain Posavljak, Miodrag Jankovic, Katarina Maksimovic CRACK INITIATION LIFE OF NOTCHED METALLIC PARTS EXPOSED TO LOW CYCLE FATIGUE ....................................................................................... 99 vii
  • 15 Sr an Bošnjak, Zoran Petkovi , Miloš or evi , Nebojša Gnjatovi , Nenad Zrni DESIGN IMPROVEMENTS OF THE BUCKET WHEEL WITH DRIVE ............11116. Aleksandar Marinkovi , Aleksandar o i , Bratislav Stojiljkovi , Milan Vuli evi DESIGN OF TESLA-TIFFANY CASCADE FOUNTAIN AS A SAMPLE OF TESLA`S RESEARCH CREATIVITY IN FIELD OF MECHANICAL ENGINEERING .................................................................................................11717. Svetislav Lj. Markovi DESIGN SEALS FOR REAL CONNECTIONS .................................................12318. Valentina Golubovi -Bugarski, Drago Blagojevi , or e i a, Branislav Sredanovi DETECTION OF STRUCTURAL DAMAGE LOCATION USING FREQUENCY RESPONSE FUNCTION DATA .................................................12919. Dragi Stamenkovi , Mato Peri DETERMINATION OF RESIDUAL STRESSES IN TUBULAR WELDED STRUCTURAL COMPONENTS ........................................................................13520. Živko Pejašinovi , Zorana Tanasi , Goran Janji EFFECT OF MATERIAL PROPERTIES OF MEASURING FORCE TRANSDUCER ELASTIC ELEMENTS TO METROLOGIY CHARACTERISTICS .........................................................................................14521. Siniša Kuzmanovi , Milan Rackov EVALUATION OF CONCEPTUAL SOLUTIONS OF UNIVERSAL HELICAL TWO STAGE GEAR UNITS ..............................................................................15122. Ivica amagi , Nemanja Vasi , Zijah Burzi FATIGUE ANALYSIS FROM FRACTURE MECHANICS ANGLE ....................15923. Slobodanka Boljanovi , Stevan Maksimovi , Strain Posavljak FATIGUE LIFE ESTIMATION OF CRACKED STRUCTURAL COMPONENTS .................................................................................................16524. Ibrahim Badžak, Remzo Dedic, Mersida Manjgo HYDRAULIC INSTALLATION OF EKO CONTAINER ......................................17325. Vesna Rankovi , Nenad Grujovi , Goran Milovanovi , Dejan Divac, Nikola Milivojevi PREDICTION OF DAM BEHAVIOUR USING MULTIPLE LINEAR REGRESSION AND RADIAL BASIS FUNCTION NEURAL NETWORK .........17926. Nenad Zrni , Sr an Bošnjak, Vlada Gaši , Miodrag Arsi SOME ASPECTS IN FAILURE ANALYSIS OF CRANES ................................18527. Stevan Maksimovi , Ivana Vasovi , Mirko Maksimovi SOME ASPECTS TO DESIGN OF AIRCRAFT STRUCTURES WITH RESPECTS TO FATIGUE AND FRACTURE MECHANICS ............................19128. Andrija Vuji i , Nenad Zrni STATE-OF-THE-ART IN LIFE CYCLE ASSESSEMENT AS A CORE OF LIFE CYCLE DESIGN .......................................................................................20329. Mersida Manjgo, Ljubica Milovi , Zijah Burzi STRESS INTENSITY FACTOR AND ITS EFFECT OF STRUCTURAL INTEGRITY ........................................................................................................20930. Vukojevi Nedeljko, Hadžikaduni Fuad, Pavi Mate VIBRATORY STRESS RELIEVING OF TANK FLANGS ..................................215viii
  • 31. Ranko Antunovic VIBRODIAGNOSTICS OF ROTATION MACHINES ........................................ 221B. PRODUCTION TECHNOLOGIES AND ENGINEERING ...................................... 22932. Kramar D., Sokovi M., Kopa J. ADVANCED CUTTING TECHNOLOGY – HIGH-PRESSURE JET ASSISTED MACHINING ....................................................................................................... 23133. Milan Milovanovi , Milentije Stefanovi ANALYSIS OF THE EFFECTS OF APPLYING NEW MATERIALS .................. 24134. Tomasz Kudasik, Tadeusz Markowski, Olimpia Markowska, S awomir Miechowicz APPLICATION OF RAPID PROTOTYPING RESINS FOR PHOTOELASTIC TESTING ............................................................................................................ 24735. Slavica Cvetkovi AUDITING PROCESS DESIGN COMPANY LOGISTICS SYSTEM ................. 25336. Andonovic Vladan, Vrtanoski Gligorce CAD/CAM TECHNOLOGY IN DENTAL MEDICINE .......................................... 25937. Zoran Janjuš, Aleksandar Petrovi , Aleksandar Jovovi , Radica Proki -Cvetkovi , Predrag Ili CHANGES VOLTAGE COMPACTION POLYPROPYLENE FILLED WITH GLASS POWDER .............................................................................................. 26538. Miletic Ostoja, Todic Mladen CHANGING THE WALL THICKNESS PROFILE IN THE PROCESS OF PROFILING ........................................................................................................ 27139. Plavka Skakun, Miroslav Plan ak, Dragiša Viloti , Mladomir Milutinovi , Dejan Movrin, Ognjan Lužanin COMPARATIVE INVESTIGATION OF DIFFERENT LUBRICANTS FOR BULK METAL FORMING OPERATIONS ..................................................................... 27540. Borislav Kovljeni COMPLEXITY ANALYSIS OF CAD/CAM SYSTEMS INTEGRATION IN ERP BUSINESS ENVIRONMENT ............................................................................. 28141. S. Aleksandrovi , T. Vujinovi , M. Stefanovi , V. Lazi , D. Adamovi COMPUTER CONTROLLED EXPERIMENTAL DEVICE FOR INVESTIGATIONS OF TRIBOLOGICAL INFLUENCES IN SHEET METAL FORMING .......................................................................................................... 28542. Radu Alexandru Ro u, Viorel Aurel erban, Mihaela Popescu, U u Drago Cosmin Locovei DEPOSITION OF TITANIUM NITRIDE LAYERS BY REACTIVE PLASMA SPRAYING ......................................................................................................... 29143. S awomir Miechowicz, Tadeusz Markowski, Tomasz Kudasik, Olimpia Markowska DESIGN AND FABRICATION OF MEDICAL MODELS WITH RAPID PROTOTYPING TECHNIQUES AND VACUUM CASTING .............................. 29744. Djordje Vukelic, Branko Tadic, Janko Hodolic, Igor Budak, Milovan Lazarevic DEVELOPMENT AN EXPERT SYSTEM FOR MACHINING FIXTURE DESIGN .............................................................................................................. 30345. Bogdan Nedi , Gordana Globo ki-Laki DEVELOPMENT MODEL FOR CONTROL METAL CUTTING PROCESS .......... 309 ix
  • 46. Aurel Prsti , Zagorka Acimovi -Pavlovi , Zvonko Gulišija, Mirjana Stojanovi DEVELOPMENT OF EPC PROCESS FOR MANUFACTURING PARTS IN AUTOMOTIVE INDUSTRY ................................................................................31547. Obu ina Mur o, Škalji Nedim, Smaji Selver EFFECT OF SURFACE ROUGHNESS ON WOOD ADHESION ......................32148. M. Stefanovi , D. Viloti , M. Plan ak, S. Aleksandrovi , Z.Gulisija, D. Adamovi FORMING LIMIT INDICATORS IN METAL FORMING .....................................32749. Run ev Dobre, Gligor e Vrtanoski, Ljup o Trpkovski HEATED TOOL BUTT WELDING OF POLYETHYLEN PIPES .........................33750. Marija Mihailovi , Aleksandra Patari , Zvonko Gulišija, Miroslav Soki INCREASING PRODUCTION EFFICIENCY THROUGH CASTING QUALITY IMPROVING BY ELECTROMAGNETIC FIELD APPLYING ..............................34351. Zorana Tanasi , Goran Janji , Bobrek Miroslav, Živko Pejašinovi INFLUENCE OF ORGANIZATIONAL CULTURE ON BUSINESS PERFORMANCE ................................................................................................34952. Robert Molnar, Drago Soldat INNOVATION-THE KEY FACTOR IN ENTREPRENEURIAL CYCLES ............35553. Vid Joviševi , Stevo Borojevi , Gordana Globo ki-Laki , Branislav Sredanovi LABORATORIES UNDER REQUIREMENTS OF DIRECTIVES AND STANDARDS OF EUROPEAN UNION .............................................................36154. Sanja Petronic, Andjelka Milosavljevic, Biljana Grujic, Radovan Radovanovic Radmila Pljakic LASER SHOCK PEENING OF N-155 SUPERALLOY EXPOSED TO AGGRESSIVE MEDIUM ....................................................................................36755. Bogdan Mari , Ranko Boži kovi LEAN CONCEPT TOOLS IN PROCESS OF TECHNICAL SYSTEMS OVERHAUL ........................................................................................................37356. Ranko Radonji , Milan Šljivi , Živko Babi , Milentije Stefanovi NUMERICAL SIMULATION OF HOLE FLANGING OF CIRCULAR SHEETS ..37957. Dejan Luki , Velimir Todi , Mijodrag Miloševi , Goran Jovi i ONE APPROACH TO THE DEVELOPMENT AND IMPLEMENTATION OF FLEXIBLE MANUFACTURING SYSTEMS ........................................................38558. Milentije Stefanovi , Srbislav Aleksandrovi , Dragan Adamovi PAPER ABOUT PAPERS IN THE AREA OF METAL FORMING PRESENTED AT DEMI CONFERENCES HELD SO FAR ................................39159. Todic Mladen, Miletic Ostoja POSITION OF THE NEUTRAL SURFACE DEFORMATION AT BENDING TWO LAYER COMPOSITES .............................................................................39960. Milena Cosi , Zagorka Acimovi -Pavlovi , Zvonko Gulišija, Mirjana Stojanovi , Zoran Janjuševi POSSIBILITY TO USE RHEOCASTING PROCESS FOR MANUFACTURING PARTS IN AUTOMOTIVE INDUSTRY ...............................................................40561. Zvonko Gulišija, Marija Mihailovi , Aleksandra Patari , Zoran Janjuševi PROPERTIES OF THE WROUGHT Al ALLOY 7075 OBTAINED BY ELECTROMAGNETIC CASTING PROCESS ....................................................40962. Vrtanoski Gligorce, Andonovic Vladan RAPID TECHNOLOGY IN DENTAL BIOMECHANICS ......................................413x
  • 63. Stevo Borojevi , Vid Joviševi , Gordana Globo ki-Laki , or e i a, Branislav Sredanovi , Marko Radisavljevi SELECTION OF VARIANT FOR MATERIAL FLOW TYPE IN CONDITIONS OF GROUP APPROACH USING THE SOFTWARE SYSTEM TECNOMATIX PLANT SIMULATION ......................................................................................... 41964. Velimir Todi , Dejan Luki , Mijodrag Miloševi , Jovan Vukman TECHNOLOGICAL BASIS FOR THE DEVELOPMENT AND IMPLEMENTATION OF FLEXIBLE MANUFACTURING SYSTEMS ................ 42765. Goran Janji , Predrag Nagraisalovi , Zorana Tanasi , Miroslav Bobrek, Živko Pejašinovi THE PROCESS OF MEASURING EQUIPMENT MANAGEMENT AND ITS AUTOMATIZATION ............................................................................................ 43366. S. Mazzola, P. Pedrazzoli, G. Dal Maso, C. R. Boër VIRTUAL ENVIRONMENT PLATFORM FOR INDUSTRIAL OPERATION AND MAINTENANCE ......................................................................................... 441C. THERMOTECHNIQUE AND ENERGETICS ......................................................... 44767. Nataša Soldat, Mirjana Radiši BASIC ASPECTS OF DEFINING MECHANICAL-TECHNOLOGICAL SOLUTIONS FOR THE PRODUCTION OF BIOGAS FROM LIQUID MANURE 44968. Crnojevi C., Le i M. DETERMINATION OF PRESSURE DROP TWO-PHASE FLOW OIL AND GAS FOR ISOTHERMAL FLOW IN HORIZONTAL PIPELINE ......................... 45369. Igor Andreevski, Gligor Kanev e, Ljubica Kanev e, Aleksandar Markoski, Sevde Stavreva DEVELOPMENT AND APPLICATION OF REGULATORY DISPERSION MODEL FOR AIR POLLUTION ASSESSMENT ................................................ 45970. Gordana Tica, Veljko uri kovi , Petar Gvero DIMENSIONING OBJECTS COOLING SYSTEM FOR PREDETERMINED KNOWN RELIABILITY ....................................................................................... 46571. Mi a Vuki , Velimir Stefanovi , Predrag Živkovi , Mirko Dobrnjac EXPERIMENTAL INVESTIGATION OF THERMAL AND FLOW PROCESSES IN SHELL AND TUBE HEAT EHCHANGERS ........................... 47572. Popov G., Klimentov Kl., Kostov B. INVESTIGATION OF THE ENERGY CONSUMPTION IN REGULATING THE FLOW RATE OF PUMP SYSTEMS ................................................................... 48173. or e S. antrak, Slavica S. Risti , Novica Z. Jankovi LDA, CLASSICAL PROBES AND FLOW VISUALIZATION IN EXPERIMENTAL INVESTIGATION OF TURBULENT SWIRL FLOW ........................................... 48974. Popov G., Klimentov Kl., Kostov B. METHODS TO ESTIMATE THE ENERGY CONSUMPTION IN REGULATING THE FLOW RATE OF PUMP SYSTEMS .......................................................... 49575. Majid Soleimaninia NUMERICAL INVESTIGATION OF HEAT TRANSFER ENHANCEMENT IN NATURAL CONVECTION AND FORCE CONVECTION IN A FLUID SATURATED VARIABLE POROSITY MEDIUM ............................................ 501 xi
  • 76. Milica Grahovac OBJECTIVE FUNCTION DEFINITION FOR PRIMARY HVAC SYSTEM TOTAL COSTS MINIMIZATION .........................................................................51577. Diana Alina Bistrian, Manuela P noiu, Tihomir Latinovi , Marcel Topor PARALLEL SOLUTIONS TO ACCELERATE MATHEMATICAL ALGORITHMS IN HYDRODYNAMIC STABILITY PROBLEMS ........................52378. Sevde Stavreva, Marko Serafimov, Igor Andreevski REDUCING CONCUMPTION OF ENERGY OF DATA CENTERS ...................53379. N. Mani , V. Jovanovi , D. Stojiljkovi RESULTS OF EXPERIMENTAL INVESTIGATION OF PELLET STOVE ACCORDING TO EN 14875 ...............................................................................53980. Mirjana Radiši , Nataša Soldat SOME EXPERIENCES IN THE PRODUCTION OF BIOGAS FROM LIQUID MANURE ............................................................................................................54981. Aleksandar Stjepanovi , Sla ana Stjepanovi , Ferid Softi , Zlatko Bundalo TEMPERATURE CHARACTERISTICS OF PHOTOVOLTAIC MODULE .........55582. Predrag Živkovi , Gradimir Ili , Mirko Dobrnjac, Mladen Tomi , Žana Stevanovi , WIND POTENTIALS ASSESMENT IN COMPLEX TERRAIN ...........................56183. Ljubo Glamo i WIND POWER RESOURCES IN THE REPUBLIC OF SRPSKA ......................56784. Milovan Kotur, Gostimir Radi COVENANT OF MAYORS FORESEEN ACTIVITIES AT DISTRICT HEATING COMPANY – TOPLANA A.D. BANJA LUKA ....................................57585. Milovan Kotur, Branko Usorac, Petar Gvero, Gordana Tica PARTIAL REPLACING HEAVY FUEL OIL WITH BIOMASS IN THE DISTRICT HEATING COMPANY IN GRADIŠKA ..............................................58186. Milovan Kotur, Zoran Kneževi , Petar Gvero, Gordana Tica BIOMASS PROJECT IN DISTRICT HEATING COMPANY (DHC) IN PRIJEDOR, BIH ..................................................................................................587D. TRAFFIC MEANS ..................................................................................................59387. Stojan Petrovi , Božidar Nikoli , Emil Hnatko, Jovo Mr a, Stevan Veinovi MALICIOUS ECOLOGY ON VEHICLES AND TRAFFIC EXAMPLE ................59588. Zlatomir Živanovi , Zoran Jovanovi , Željko Šakota A COMPARATIVE ANALYSIS OF CNG AND HYBRID BUSES VS DIESEL BUSES ...............................................................................................................60789. Milan Milovanovi , Dragoljub Radonji , Saša Jovanovi ADJUSTMENTS OF VEHICLES WITH GAS DRIVE ........................................61390. Dalibor Jajcevic, Raimund Almbauer APPLICATION OF A CYCLIC BOUNDARY CONDITION FOR CFD SIMULATIONS OF A 2-CYLINDER IC-ENGINE ...............................................61991. Melisa Velic, Semir Mulalic, Adnan Pecar CALCULATING THERMODYNAMIC PROPERTIES BY CREATING AND USING MODEL OF A DIESEL ENGINE WITH SIX CYLINDERS ....................627xii
  • 92. Mile Rai evi , Miroslav Demi , Nebojša Rako, Predrag Milenkovi DETERMINING THE DURATION OF VIBRATION MEASUREMENTS OF HUMAN BODY IN LABORATORY CONDITIONS BY SUBJECTIVE METHODS ........................................................................................................ 63393. Miroljub Tomi , Stojan Petrovi , Slobodan Popovi , Nenad Milji DUAL PORT INDUCTION SYSTEM FOR DMB 1.4 MPI ENGINE ................... 65194. Blaževi A., Bibi Dž., Filipovi I. FUNCTION AND ADOPTION OF IC ENGINES DUAL MASS FLYWHEEL ..... 66195. Jelena Eric Obucina, Jovanka Lukic HYDRAULIC PUMP IN THE OF VEHICLE STEERING SYSTEM ................... 66796. Aleksandar Davini , Radivoje Peši , Dragan Taranovi , Miroslav Ravli IGNITION SYSTEM OF MULTIPROCESSING OTTO/DIESEL ENGINE ......... 67397. Filipovi I., Milašinovi A., Blaževi A., Pecar A. IMPACT OF THE SPECIFIC ABSORBERS ON THE DYNAMIC LOAD OF THE IC ENGINE’S CRANKSHAFT ................................................................... 68198. Jasna Glišovi , Jovanka Luki , Danijela Miloradovi IMPROVEMENTS OF GROUND VEHICLES FUEL ECONOMY USING REGENERATIVE BRAKING ............................................................................. 68799. Predrag Živkovi , Mladen Tomi , Gradimir Ili , Mirko Dobrnjac, Vladimir Lazovi , INFLUENCE OF TRAFFIC ON AIR QUALITY IN NIŠ ...................................... 693100. Boran Pikula, Ivan Filipovi , Mirsad Trobradovi INVESTIGATION OF DYNAMICS CHARACTERISTICS OF HYBRID VEHICLES ......................................................................................................... 699101. Vladan Ivanovic, Decan Ivanovic, Vladimir Pajkovic LANDFILL GAS AS A FUEL FOR A VEHICLE FLEET FOR THE CITY LANDFILL .......................................................................................................... 705102. Jovanka Luki , Radivoje Peši , Dragan Taranovi NVH INVESTIGATION OF POWER STEERING SYSTEM HYDRAULIC PUMP ................................................................................................................ 711103. Vojislav B. Krsti , Božidar V. Krsti , Vuki N. Lazi POSIBILITY OF DETERMINATION THE ROUTES FOR TRANSPORTATION OF HAZARDOUS GOODS ON THE BASIS OF THE RISK LEVEL ................. 717104. Božidar V. Krsti , Vojislav B. Krsti , Ivan B. Krsti POSSIBILITIES DETERMINATION OF THE OPTIMAL STRATEGY FOR PREVENTIVE MAINTENANCE OF THE CARDAN SHIFT VEHICLE USING POLYCRITERION OPTIMIZATION .................................................................. 723105. Pikula Boran, Filipovic Ivan, Kepnik Goran RESEARCH OF THE EXTERNAL AERODYNAMICS OF THE VEHICLE MODEL .............................................................................................................. 731106. Vladimir R. Pajkovi ROAD TRAFFIC SAFETY PERFORMANCE IN MONTENEGRO .................... 739107. Dobrivoje Ninkovic SURVEY OF METHODS FOR CALCULATING THE WAVE ACTION IN THE MANIFOLDS OF INTERNAL COMBUSTION ENGINES .................................. 745108. Dragan Taranovic, Radivoje Pesic, Jovanka Lukic, Aleksandar Davinic TEST BENCH FOR NON-STANDARD MEASUREMENT CHARACTERISTICS OF RECIPROCATING COMPRESSOR ........................................................... 759 xiii
  • 109. Branislav Aleksandrovi , Rajko Radonji , Marko apan, Aleksandra Jankovi THE RESEARCH OF CORRELATIONS BETWEEN MOTORCYCLE OSCILLATORY PROCESSES DURING THE NONSTEADY MODES OF MOTION ............................................................................................................765110. Izudin Deli , Izet Ali TRENDS IN DEVELOPMENT OF CATALYTIC CONVERTER OF INTERNAL COMBUSTION ENGINES (ICE) .....................................................771111. Pantelija Daki , Sreten Peri MONITORING OIL FOR LUBRICATION OF TRIBOMECHANICAL ENGINE ASSEMBLIES ....................................................................................................777E. MECHATRONICS ..................................................................................................793112. Milan Paripovi ANALYSES THE DAMAGE CAUSED BY ATMOSPERIC DISCHARGE AND OVERVOLTAGE PROTECTION ........................................................................795113. Slaviša Todorovi , Miroslav Rogi AUTOMATION AND OPTIMIZATION OF PROJECT OPERATIONS IN THE BRIDGE CRANE DESIGN PROCESS ...............................................................801114. Corina Daniela Cun an, Ioan Baciu, Loredana Ghiorghioni DC STABILIZER WITH DIGITAL CONTROL .....................................................807115. Marija Mili evi , Vladimir Kaplarevic, Zoran Dimi , Vojkan Cvijanovi , Mirko Bu an DEVELOPMENT OF DISTRIBUTED CONTROL SYSTEM FOR ROBOTS CONTROL BASED ON REAL-TIME LINUX PLATFORM ..................................813116. Miroslav Rogi , Bojan Kneževi , Branislav Risti DEVELOPMENT OF THE CONCEPT OF INTERACTIVE EDUCATION IN MECHATRONICS ..............................................................................................819117. Ivan B. Krsti , Božidar V. Krsti , Dragan I. Milosavljevi EFFECTIVENESS DETERMINATION OF ELECTRONIC DEVICES PREVENTIVE MAINTENANCE ..........................................................................825118. Mihailo P. Lazarevi , Vasilije Vasi , Aleš Hace, Karel Jezernik FURTHER RESULTS ON MODELING, INTEGRATED DESIGN AND SIMULATION OF A MECHATRONIC SYSTEM WITH FPGA ...........................831119. Miroslav Grubiši , Snježana Rezi IMPACT OF SENSOR FAILURE ON WORK OF ELECTRONICALLY CONTROLLED DIESEL ENGINES ....................................................................837120. Miroslav Kostadinovi , Zlatko Bundalo, Dušanka Bundalo IMPLEMENTATION OF PLANTWEB ALERTS IN A DELTAV SYSTEM ...........843121. Vahid Bagher Poor, Majid Hashemipour IMPLEMENTATION OF RFID TECHNOLOGY AND SMART PARTS IN WIRELESS MANUFACTURING SYSTEMS ......................................................849122. Tihomir Latinovic, Sorin I Deaconu, Remiquez Labudski, Marcel Topor INTELLIGENT APPROACH FOR MOBILE ROBOT SIMULATOR WITH ROBOSIM SOFTWARE .....................................................................................857xiv
  • 123. Milutinovic D., Glavonjic M., Slavkovic N., Kokotovic B., Milutinovic M., Zivanovic S., Dimic Z. MACHINING ROBOT CONTROLED AND PROGRAMMED AS A MACHINE TOOL .................................................................................................................. 863124. Vladimir Kaplarevi , Marija Mili evi , Jelena Vidakovi , Vladimir Kvrgi NEW APPROACH FOR DESSIGNING ROBOT PROGRAMING SYSTEM BASED ON L-IRL PROGRAMING LANGUAGE ................................................ 873125. Kostic Aleksandra, Velic Melisa, Bektesevic Jasmin PRACTICAL STRATEGIES FOR STABILISATION OF ALGORITHMS BASED ON SECULAR EQUATIONS OF RSPDTM .......................................... 877126. Platon Sovilj, Nenad abrilo, Vladimir Vuji i , Ivan Župunski REMOTE MEASUREMENTS BY ZIGBIT WIRELESS MODULE ...................... 885127. K. Abhary, D. Djukic, H-Y. Hsu, Z. Kovacic, D. Mulcahy, S. Spuzic, F. Uzunovic SOME ASPECTS OF KNOWLEDGE ENGINEERING ...................................... 893128. Nenad Miloradovi , Rodoljub Vujanac, Blaža Stojanovi STACKING AISLE WIDTH FOR FORKLIFT TRUCKS IN PALLETIZED STORAGE AND HANDLING SYSTEMS ........................................................... 899129. Mihajlo J. Stoj i , Bojan Kneževi THE CONTROLLER DESIGN FOR TRACKING TRAJECTORY WITH CONTROLLED JERK ......................................................................................... 905130. Deaconu, S. I., Opri a N, Popa, G. N., Latinovic T. ULTRASONIC WELDING SYSTEM FOR AUTOMOTIVE WIRINGS INDUSTRY ......................................................................................................... 911131. Dražen Pašali , Zlatko Bundalo, Dušanka Bundalo, Miroslav Kostadinovi WIRELESS SENSOR NETWORKS IN HOME AUTOMATION ......................... 917132. Mihailo Lazarevi , Petar Mandi , Vasilije Vasi SOME APPLICATIONS OF NEUROARM INTERACTIVE ROBOT AND WEBOTS ROBOT SIMULATION TOOL ............................................................ 923F. MAINTENANCE OF TECHNICAL SYSTEMS ....................................................... 929133. Rusmir Bajri , Enver Omazi , Fehmo Mrkaljevi AVAILABILITY ANALYSIS OF IRREDUNDANT TECHNICAL SYSTEMS ........ 931134. Aleksandar Živkovi , Milan Zeljkovi , Milorad Rodi , Milivoje Mijuškovi COMPUTER AND EXPERIMENTAL DETERMINATION OF THE HUB UNIT LIFE .................................................................................................................... 937135. Danijela Nikolic, Vanja Sustersic, Jasmina Skerlic DECENTRALIZED WASTEWATER TREATMENT SYSTEMS IN LARGE SETTLEMENTS ................................................................................................. 943136. Ivan B. Krsti , Dragan I. Milosavljevi , Božidar V. Krsti DETERMINATION THE PERIODICITY OF MANAGING OF PREVENTIVE MAINTENANCE OF TECHNICAL SYSTEMS ................................................... 949137. Mihaela Popescu, Radu Alexandru Ro u, Carmen Opri , Ibolyka Bran ENVIRONMENT PROTECTION FOR WELDING AND ALLIED TECHNIQUES . 955138. Milomir upovi , Desimir Jovanovi , Bogdan Nedi FTA AND FMEA IN PREDICTING INCIDENTAL CONDITIONS IN CABLE CARS AND SKI LIFTS ....................................................................................... 961 xv
  • 139. Milorad Panteli , Sr an Bošnjak MAINTENANCE AND LIFECYCLE OF THE EXCAVATION UNITS ..................967140. Dusan Jovanic, Drago Soldat MODELING MAINTENANCE ACTIVITIES ON A WELDED CONSTRUCTION USING IDEF0 METHODOLOGY ........................................................................973141. Jasmina Skerlic, Vanja Sustersic, Danijela Nikolic NATURAL SYSTEMS OF WASTEWATER TREATMENT IN SMALL SETTLEMENTS ................................................................................................. 979142. Miodrag Milutinovi , Vladimir Popovi PROCEDURES FOR RISK BASED MANAGEMENT AND MAINTENANCE RIMAP PRINCIPLE ............................................................................................985143. Dragoslav Dobraš, Sead Avdi PURPOSE AND METHOD EDUCATION OF INTERNATIONAL RECOGNIZED PERSONNEL FOR WELDING ..................................................997144. Aleksandar Majstorovi SAFETY PARTS OF BODY TO RESPIRATION WITH BREATHING APPARATUS IN DANGEROUS ZONE ............................................................1003145. Stojan Simi STATUS OF MAINTENANCE SERVICE IN COMPANIES IN REGION IN TIMES OF RECESSION ..................................................................................1009146. Remigiusz LABUDZKI IDENTIFY CHARACTERISTICS OF OBJECTS IN MACHINE VISION ............1015xvi
  • PREFACEThis year, the 10th Anniversary International Conference on Accomplishments inElectrical, Mechanical Engineering and Information Technology DEMI is going to beheld. The number of countries taking part in DEMI is increasing so this year scientistsand researchers from 17 countries are going to take part. This modest jubilee, but witha considerable number of participants, points to the fact that that DEMI Conference isbecoming an internationally recognized conference with respectable participants fromtechnologically developed countries.The first DEMI Conference was held in 1998 and was organized by the Faculty ofMechanical Engineering in Banja Luka having aimed to draw together university andinstitute scientists and professionals as well as experts from a very weakened, post-war economy. Such mission and role of the DEMI Conference has remained one of itspriorities to this day. From 1998-2003, the DEMI Conference was being held everyyear, and after that time, it started to be held every two years. The DEMI became atraditional conference and it took over a significant but demanding role of gatheringresearchers and scientists as well as economy experts with the aim to foster fasterimplementation of contemporary research and new technologies in productionprocesses in order to ensure better competitiveness of our industry. Definitely, this taskhas not been an easy one to achieve. thThis 10 anniversary DEMI Conference has seen a significant progress. For the firsttime, the official language of the Conference is English. The number of papers to bepresented at the Conference is 145 from 17 countries (former Yugoslav countries,Austria, Germany, Switzerland, Italy, Norway, Poland, Romania, Bulgaria, Turkey,Australia, New Zealand). The previous DEMI 2009 Conference brought togetherresearchers and scientists from 11 countries which indicates a greater interest in thisconference. The Conference activities will be realized in six sections which encompassthe following thematic fields: Production Technologies and Engineering, Mechanicsand Design, Traffic Means, Thermotechnique and Energetics, Maintenance ofTechnical Systems and Mechatronics. Key note lectures will be delivered by scientistsfrom universities from Switzerland, Norway, Serbia and Croatia who are involved inresearches of the highest scientific level in the fields of: mechatronics, energetics,modern approaches to machine structure and product design, the ecology oftransportation and global warming, paradigms of contemporary production and acompany’s adjustment to the current market requirements, etc.Therefore, we hope that the papers to be presented will contribute in considering andreflection on the present situation of research in mechanical and electrical engineeringas well as information and communications technologies in the regional conditions,enable comparison with developed European countries and offer new models for fasterimplementation of contemporary research as to encourage production in the country.As it is the most important scientific and professional conference in the fields ofmechanical engineering, information and communications technologies in our country,this is expected from the DEMI Conference with good reason. xvii
  • On behalf of the Organizational Committee of the DEMI 2011Conference, I would liketo thank all authors, members of the review team as well as institutions, companiesand individuals who contributed to the realization of the Conference program.The International Conference on Accomplishments in Electrical and MechanicalEngineering and Information Technology DEMI 2011 will be held at the Faculty ofMechanical Engineering, University of Banja Luka. We are looking forward towelcoming you as our dear guests. Welcome to the 10th anniversary DEMI 2011Conference.Banja Luka, 12 May 2011 President of the Organizational Committee of the 10th International DEMI 2011 Conference Gordana Globo ki-Laki , PhD. Associate Professorxviii
  • KEYNOTE LECTURES
  • SUSTAINABLE INNOVATION AND INNOVATION FOR SUSTAINABILITY:THE EVOLUTION AND INVOLUTION OF PRODUCTION PARADIGMS Claudio R. Boër1Summary: The consumer market is continuosly evolving requiring new products withmore functions, better design, more personalized, and lately, more sustainable. Theevolution of products has been followed closely by the evolution of the means tomanufacturing them, to assemble them and to delivering them to the market. Studyingthe evolution of production, it is possible to see that different paradigms have appearedfrom pre-industrialization, to industrialization, to mass production and then masscustomization. Innovation has always been the driver of the evolution of products andrelated process and the present keynote will present an overview of the past evolutionand the future trends. In particular it will be shown how important is that innovation issustainable in time for a company to be successfull and how innovation is a key for thefuture sustainable production paradigm.Key words: Production paradigm, sustainble production, mass customization,innovation.1 Prof. Dr. Ing.Claudio R. Boër, Director ICIMSI-SUPSI, Manno-Lugano, Switzerland 3
  • COOPERATION MODELS BETWEEN UNIVERSITIES AND INDUSTRY IN APPLIED RESEARCH, SWITZERLAND CASE STUDY AND SOME PRACTICAL EXAMPLES Giambattista Ravano1Summary: Switzerland has a very well established economic environment of smalland medium size industries (SMI). Thanks to a positive financial situation and a long-term policy of development an interesting system of cooperation between Universities(particularly Universities of Applied Sciences) and SMI could be established.Some key successful factors are described: Direct financing of industrial projects with proved business plan Participation of main key player in strategic decision (industrial organizations, research partners, Federal and Cantonal government) Balanced mix of cooperation and concurrence between research partners.Thanks to this approach a positive trend to innovation could be reached.Three examples of applied research projects in the fields of “materials for precisionmechanics, traffic systems, biotech precision systems” which turned into innovationand new business are presented.1 Prof. Giambattista Ravano, Director Department of Innovative Technology, University of Applied Sciencesand Arts of Southern Switzerland (SUPSI), CH -- 6928 Manno 5
  • TRANSPORT ECOLOGY AND GLOBAL CLIMATE CHANGE Radivoje Peši 1, Stevan Veinovi 2Summary: Heavy critics claim that a group of Western scientists has put together theprepositions for Kyoto protocol. The Kyoto protocol limits the production of thefollowing six components with anthropogenic activities: carbon dioxide, methane,nitrous oxide, per fluorocarbons, sulphur hexafluoride and hydro fluorocarbons. Themost dangerous role is assigned to carbon dioxide, so, on that basis, the elimination ofcarbon dioxide production is proclaimed as ecological success. The escalation iscontinued in such a pace, that clean technologies (?), clean fuels (?) and cleanvehicles (?) and similar are defined. The controversy over global warming gets evenmore complicated when you include politics, economics, greed, and the self interestsof the various governments, NGOs and companies. Astronomer Milutin Milankovicstudied changes in the orbital eccentricity, obliquity, and precession of Earthsmovements. Such changes in movement and orientation change the amount andlocation of solar radiation reaching the Earth. He hypothesized that when some parts ofthe cyclic variations are combined and occur at the same time, they are responsible formajor changes to the earths climate (even ice ages). The next stage of power trainand fuel strategy involves using new high economy combustion engines that can berun with partially renewable fuels and used worldwide. Researchers are hard at workexploring new fuels, engines and vehicle technologies- but there are not clean cars,clean energy or full renewable fuels. Contrary to popular belief electric vehicles don’thold all the emission answers, until the electric power comes from coal-fired powerstations or part renewable sources (hydro, wind and sun).Key words: global warming, IC engine, renewable fuels, transport ecology1. INTRODUCTION Rational and controlled vehicle use in transport and traffic is an obligation forthe future. Any activation of energy bears a burden to the environment in full amount. Awarm-up lifts light gases like oxygen and nitrogen, while steam and carbon-dioxidechoke the environment. Such are the signs of nature: there are no pure energy forms,no clean fuels, and no clean engine, neither clean vehicle. Actual reserves and natural1 PhD Radivoje Peši , professor, Kragujevac, Faculty of Mechanical Engineering in Kragujevac, Serbia, E-mail: pesicr@kg.ac.rs2 PhD Stevan Veinovi , retired professor, Kragujevac, Faculty of Mechanical Engineering in Kragujevac,Serbia, E-mail: vpst@kg.ac.rs 7
  • Peši R., Veinovi S.gifts – coal, petroleum and gas – should be used less as power sources and more asraw materials. The greatest contribution to ecological prolongation of life on our planetis given by rational, economical technologies and products that carefully engage thegifts of nature. Political leaders gathered in Kyoto, Japan, in December 1997 to consider aworld treaty restricting anthropogenic production of greenhouse gases, chiefly carbondioxide (CO2). They feared that CO2 would result in anthropogenic - caused globalwarming – hypothetical severe increases in Earth’s temperatures, with disastrousenvironmental consequences. During the past 10 years, many political efforts havebeen made to force worldwide agreement to the Kyoto treaty [1]. There is no convincing scientific evidence that anthropogenic release of carbondioxide, methane, or other greenhouse gasses is causing or will, in the foreseeablefuture, cause catastrophic heating of the Earths atmosphere and disruption of theEarths climate. Moreover, there is substantial scientific evidence that increases inatmospheric carbon dioxide produce many beneficial effects upon the natural plant andanimal environments of the Earth. Oregon Petition, from the Oregon Institute ofScience and Medicine, signed by over 17,000 international scientists including morethan 2000 of the worlds leading climatologists, meteorologists and planetary/atmospheric scientists [2]. Carbon Dioxide currently at 370 ppm, for it to be dangerous it would have to beat 15,000 ppm. This could not be reached even if every fossil fuel was burned.Thousands and thousands of studies show that higher levels of CO2 are good forplants. Many scientists believe plants still are not getting enough CO2. Tomato farmersusing exhaust from electricity to grow their tomatoes. Vegetation looses less waterunder higher CO2 levels, meaning vegetation in drought prone areas will live longerand produce more. Rice (the most eaten food in the world) was shown to increasemass and use less water with higher CO2 levels. Meaning the most important food inthe world highly benefits from CO2 increase. Fig. 1 Global temperatures8
  • Ecology of transport and global change earth’s climate The eruption of volcano Tambora in Indonesia in 1815 caused the Year withoutsummer because of global cooling. Due to the destruction of crops, disease,contamination of water, etc., tens of thousands more died in the next few followingyears. In 1991 Volcano Pinatubo (Philippines) caused the entire earth to cool by 0.5 °Ffor over one year, fig. 1. The misjudgment of Intergovernmental Panel on Climate Change (IPCC) andPotsdam Institute for Climate Impact Research (PIK) are not overlooked. The 50,000-member American Physical Society criticized the CO2 hypothesis as well as manyGerman scientists, physicists and meteorologists. Even former herald of climatedisasters are now well established that the climate is strongly influenced by nature (sun+ ocean current) than the CO2 in the atmosphere. But it is also annoying: For about tenyears, directed the global temperature is not more to the global warming prophets, ithas ceased to rise, although more and more diligently CO2 leaves the smokestacksand exhaust pipes of the people. Figure 2 shows the petition of 18,000 American geologists against climatepolicy USA Government. Fig. 2 Climate disaster: Cause is the sun! Fig. 3 Annual mean surface 3 December 1973 9 April 2001 temperatures in the contiguous The slope decreasing The slope increasing United States between 1880 and Intermediate Trends 2006. The slope of the least-squares Fig. 4 Is it Global Cooling or Global trend line for this 127-year re cord is Warming? 0.5ºC per century [1]. 9
  • Peši R., Veinovi S. Surface temperatures in the United States during the past century reflect thisnatural warming trend and its correlation with solar activity, as shown in fig. 3.Compiled U.S. surface temperatures have increased about 0.5°C per century, which isconsistent with other historical values of 0.4 to 0.5°C per century during the recoveryfrom the Little Ice Age. This temperature change is slight as compared with other natural variations.Three intermediate trends are evident, including the decreasing trend used to justifyfears of global cooling in the 1970s. Between 1900 and 2000, on absolute scales ofsolar irradiance and degrees Kelvin, solar activity and temperature increased fig. 4. Secular Variations of the Planetary Orbits, (French: Variations Séculaires des Orbites Planétaires, abbreviated as VSOP) allows prediction of past and future orbital parameters with great accuracy. is obliquity (axial tilt). e is eccentricity. is longitude of perihelion. e·sin( ) is the precession index, which together with obliquity, controls the seasonal cycle of insolation. is the calculated daily-averaged insolation at the top of the atmosphere, on the day of the summer solstice at 65 N latitude. Benthic forams and Vostok ice core show two distinct proxies for past global sea level and temperature, from ocean sediment and Antarctic ice respectively. Vertical gray line is current conditions, at 2 kilo years A.D. Fig. 5 Past and future Milankovic cycles [3,4]10
  • Ecology of transport and global change earth’s climate The Serbian scientist Milutin Milankovic is best known for developing one of themost significant theories relating Earth motions and long-term climate change.Milankovic dedicated his career to developing a mathematical theory of climate basedon the seasonal and latitudinal variations of solar radiation received by the Earth. Nowknown as the Milankovic Theory, it states that as the Earth travels through spacearound the sun, cyclical variations in three elements of Earth-sun geometry combine toproduce variations in the amount of solar energy that reaches Earth: 1. Variations in the Earths orbital eccentricity (e) – the shape of the orbit around the sun. 2. Changes in obliquity ( ) – changes in the angle that Earths axis makes with the plane of Earths orbit. 3. Precession – the change in the direction of the Earths axis of rotation, i.e., the axis of rotation behaves like the spin axis of a top that is winding down; hence it traces a circle on the celestial sphere over a period of time. Together, the periods of these orbital motions have become known as Milankovic cycles, fig. 5. Changes in orbital eccentricity affect the Earth-sun distance. Currently, adifference of only 3 percent (5 million kilometers) exists between closest approach(perihelion), which occurs on or about January 3, and furthest departure (aphelion),which occurs on or about July 4. This difference in distance amounts to about a 6percent increase in incoming solar radiation (insolation) from July to January. Theshape of the Earth’s orbit changes from being elliptical (high eccentricity) to beingnearly circular (low eccentricity) in a cycle that takes between 90,000 and 100,000years. When the orbit is highly elliptical, the amount of insolation received at perihelionwould be on the order of 20 to 30 percent greater than at aphelion, resulting in asubstantially different climate from what we experience today. As the axial tilt (obliquity) increases, the seasonal contrast increases so thatwinters are colder and summers are warmer in both hemispheres. Today, the Earthsaxis is tilted 23.5 degrees from the plane of its orbit around the sun. But this tiltchanges. During a cycle that averages about 40,000 years, the tilt of the axis variesbetween 22.1 and 24.5 degrees. Because this tilt changes, the seasons as we knowthem can become exaggerated. More tilt means more severe seasons–warmersummers and colder winters; less tilt means less severe seasons–cooler summers andmilder winters. Its the cool summers that are thought to allow snow and ice to last fromyear-to-year in high latitudes, eventually building up into massive ice sheets. There arepositive feedbacks in the climate system as well, because an Earth covered with moresnow reflects more of the suns energy into space, causing additional cooling. Changes in axial precession alter the dates of perihelion and aphelion, andtherefore increase the seasonal contrast in one hemisphere and decrease theseasonal contrast in the other hemisphere. Using these three orbital variations, Milankovic was able to formulate acomprehensive mathematical model that calculated latitudinal differences in insolationand the corresponding surface temperature for 600,000 years prior to the year 1800.He then attempted to correlate these changes with the growth and retreat of the IceAges. To do this, Milankovic assumed that radiation changes in some latitudes andseasons are more important to ice sheet growth and decay than those in others. Then,at the suggestion of German Climatologist Vladimir Koppen, he chose summerinsolation at 65 degrees North as the most important latitude and season to model, 11
  • Peši R., Veinovi S.reasoning that great ice sheets grew near this latitude and that cooler summers mightreduce summer snowmelt, leading to a positive annual snow budget and ice sheetgrowth [3,4]. But, for about 50 years, Milankovics theory was largely ignored. Then, in 1976,a study published in the journal Science examined deep-sea sediment cores and foundthat Milankovics theory did in fact correspond to periods of climate change (Hays et al.1976 [5]). Specifically, the authors were able to extract the record of temperaturechange going back 450,000 years and found that major variations in climate wereclosely associated with changes in the geometry (eccentricity, obliquity, andprecession) of Earths orbit. Indeed, ice ages had occurred when the Earth was goingthrough different stages of orbital variation. Since this study, the National Research Council of the U.S. National Academyof Sciences has embraced the Milankovic Cycle model. ...orbital variations remain themost thoroughly examined mechanism of climatic change on time scales of tens ofthousands of years and are by far the clearest case of a direct effect of changinginsolation on the lower atmosphere of Earth (National Research Council, 1982 [6]).2. RENEWABLE ENERGY Renewable Energy – Energy derived from sources that are not depleted whenused, therefore their use causes little environmental impact. Examples are wind power,hydroelectric energy and solar energy. There are some definitions of Renewable energy resource: Renewable energy resource is: an energy resource that is replaced rapidly bynatural processes. Some examples of renewable energy resources are sunlight,hydropower (water falling through a dam), and wood; When you use some sunlight towarm your back, more is made almost immediately available; Water above the dam iscontinually replaced by rainfall. If you chop down a tree and burn its wood in yourcampfire, it takes awhile for the forest to grow enough to replace that wood, but it willhappen within your lifetime [7]; and Any energy resource that is naturally regenerated over a short time scale andderived directly from the sun (such as thermal, photochemical, and photoelectric),indirectly from the sun (such as wind, hydropower, and photosynthetic energy stored inbiomass), or from other natural movements and mechanisms of the environment (suchas geothermal and tidal energy); Renewable energy does not include energy resourcesderived from fossil fuels, waste products from fossil sources, or waste products frominorganic sources [8]. To our knowledge, there is no legal definition as to what renewable means -and the meanings proffered vary quite a bit. Technically, all sources of power arerenewable, just at different rates - so the primary difference between a renewable andnon-renewable is the rate of replenishment. That, in itself, should be a flag that this(in science terms) is a rather arbitrary and subjective definition. Who is to say whatreplenishment rate is good or bad, and on what basis? Let we consider this definition: Renewable is an energy resource that isreplaced rapidly by natural processes… Non-renewable is any resource that is notreplaced in a reasonable amount of time (our lifetime, our children’s lifetime …) and isthus considered ‘used up’ and not available to us again. Such words as rapidly and12
  • Ecology of transport and global change earth’s climatereasonable are subjective, relative terms - not scientific. Another pivotal aspect ignoredin these definitions is the fact that although a source (e.g., wind) may be quicklyreplenishable, it uses up other resources (e.g., land) that are non-replenishable. Wewill run out of suitable land for wind power sooner than we will run out of fossil fuels.Shouldn’t the entire package be assessed as a whole? Considering the variability, inadequacy, and political nature of its currentiteration, there is some merit to just exterminate renewable from our vocabulary. Butnature abhors a vacuum, so for sound bite reasons, if we refuse to use renewable,then we would be well-advised to come up with a good replacement. Then, our choicesare to redefine renewable so that it makes more scientific sense, or to come up with asubstitute [8]. They shared their conundrum with a group of energy experts.Interestingly, they were unanimous in their consensus that there was no hope ofsalvaging renewable. One environmentalist said: “Several years ago, I came to theconclusion that the word renewable, applied as a source of energy, was a pejorative -and I treat it as such today (much as I do terms like windmills and windfarms). Theseare all words bowdlerized of any positive meaning, designed by the craven to casuallyseparate people from the contents of their wallets. And so, in my public comments, Ialways connect renewables with fraud. Rather than refine the definition, I move that weridicule the very concept. Instead I recommend promoting the principle of our makingdecisions based on energy density, or something in that vein.” The other scientist said: When questioned on renewable, it is relatively simpleto explain the First Law of Thermodynamics concerning conservation. Energy cannotbe new, thus cannot be renewed. All we are doing is transforming one manifestation ofenergy into another - and we should be doing it in a clean, non-polluting, preferablynon-carbon-based manner. This avoids (most times) the controversial subject ofpotential, unquantified global warming versus thermal equilibrium - which is too longand too complex for most listeners. The authors suggest that we should normally usethe phrases clean energy, and clean sources, because they are more accurate thanrenewable [9].3. PROS AND CONS OF BIOFULES The production of biofuels for transport faces several challenges.3.1 Energy balance There is controversy over the energy balance of biofuels production. Theenergy balance is the amount of energy needed over the life-cycle to produce biofuels(input) versus the amount of energy produced (output). According to studies byPimentel and Patzek, it takes more energy to make ethanol than is contained in theethanol itself. Other studies (e.g. by the US Department of Agriculture) indicate that theenergy balance is positive. The balance also varies largely according to the crops usedand the transformation process.3.2 Climate change reduction potential In principle, biofuels are "carbon neutral": when they are used, no more carbondioxide is released than has been absorbed during the growth of the plants used to 13
  • Peši R., Veinovi S.make these biofuels. Therefore replacing fossil fuels with biofuels for transport couldhelp in the fight against climate change, fig. 6. Fig. 6 Biofuels are carbon neutral But other studies, including a May 2007 report [10] by the United NationsEnergy division, contest this conclusion, saying that the use of biofuels could actuallyincrease greenhouse gas emissions as land would be converted from forests, wetlandand reserves for conservation to grow more corn or soya beans. The report notes thatwith respect to reducing greenhouse-gas emissions, biofuels would be moreappropriately used for combined heat and power production rather than for transport.3.3 Land Use Using agricultural land to grow bio-energy crops would compete with the use ofland for food and animal feed production, driving up the prices of commodities likecereals. According to the European Environment Agency (EEA), reaching theinitial 5.75% target of the biofuels directive would already require biofuel crops totake over between 4% and 13% of the total agriculture area of the EU-25. Nevertheless a July 2007 study by the Commissions DG Agriculture foreseesthat reaching the new 10% target for biofuels in transport by 2020 would not overlystretch the (EUs) land availability, requiring a relatively modest 15% of arable land,which it claims could be largely covered by set aside land, currently reserved under theCommon Agricultural Policy in order to limit excessive production by farmers [11].European NGOs say palm oil producers in Indonesia and Malaysia are damaging theenvironment by planting crops in natural forests and in the middle of protected animalhabitats, and accuse the industry of illegally logging rainforests and violating the rightsof local communities.14
  • Ecology of transport and global change earth’s climate3.4 Environment and sustainability Energy crops generally require more fertilizers and pesticides to grow. Theyalso require more water, draining the earth’s already scarce resources. What’s more,biodiversity loss - especially in developing countries seeking to enter this growingmarket - is an important risk as forests and grasslands are cleared to plant the vastquantities of crops needed to make a significant dent in the use of oil in transport. Calls for binding sustainability criteria to be introduced in laws promotingincreased biofuel use therefore emerged from all sides. In its proposed RenewablesDirective of 23 January 2008 [12], the Commission proposed to introduce certainstandards, including an obligation for biofuels counting towards the 10% target todeliver life-cycle CO2 savings of at least 35% compared to fossil fuels and a ban onbiofuels planted in protected areas, forests, wetlands and highly biodiversegrasslands. But Members of the European Parliament (MEPs) insisted on tougherconditions. In September 2008 [13], the Industry and Energy Committee backed areport demanding that biofuels offer at least 45% carbon emission savings comparedto fossil fuels – a figure that would rise to 60% in 2015. They also insisted thatadditional social and environmental criteria be included to protect natural resourcesfrom both direct and indirect land use changes and to guarantee respect for humanrights and adequate working conditions in biofuel plants, especially in developingnations. In an attempt to reach a compromise between the 27 member states on theissue, a special ad-hoc working group was set up at the end of February 2008, with theaim of drafting "core criteria" for biofuels (EurActiv 01/04/08). After months of in-fighting, EU ambassadors appeared to have reached a consensus in September. The EUs Renewables Directive, adopted on 26 March 2009, initially requires a35% CO2 saving, which will then be scaled up to "at least 50% in 2017 and 60% in newinstallations thereafter. It stipulates that biofuels and bioliquids taken into account inthe 10% target must not be produced from raw materials from land with "highbiodiversity value", land that has a high carbon stock, or peat lands. The final life-cycle CO2 reduction requirement will be crucial for the biofuelindustry. Indeed, typically, biodiesel made from European-grown rapeseed results in agreenhouse gas saving of 44% while the typical figure for ethanol made from EU sugarbeet is 48% [13]. When using fossil fuels we can assume that on the part of the land, that isplaned to grow crops for biofuel, we could plant trees. In this case the trees will be ableto absorb a larger amount of CO2 than the crops for biofuels for the equal amountenergy in the fuels. Thus, the statement that biofuels have a lower CO2 emission thanfossil fuels is questionable.3.5 Biofuel costs Biofuels are more expensive than traditional fossil fuels. Therefore taxexemptions are needed to make them competitive. The second generation biofuelspromise to be cheaper but are still under development. In some countries like Brazil,biofuels can be produced at cheaper prices. 15
  • Peši R., Veinovi S.4. TRANSPORT ECOLOGY The term transport ecology or ecology in transport implies a complex balancingof the impact of every anthropogenic activity on the environment according to the lawsof nature! Unilaterally reduction emissions only one component can not to declare asenvironmentally, especially when it leads to even greater negative consequences onthe other side. An initial environmental requirement is ordered refineries to reduce the amountof lead and sulphur in gasoline. Products of combustion of such fuels have much moretoxic components, heavy metals and Particulate Matter (PM). Table 1 presents the international drive cancerous categories of fuel, lubricantsand bitumen.Table 1 Carcinogenic category (IRAC- International Research Agency for Cancer) Degree of Group Fuel and lubricants Carcinogenic Benzene, 1 Proven lubricants, bitumen 2A Probable Gasoline 2B Possible Diesel, heating oil 3 Not classified MTBE, alcohols All of our refineries used metal additives in the engine fuel in order to increaseof the octane number, fig. 7! There is insisting on the separation of gasoline at theleaded and lead-free gasoline in our regulations! Someone has forgotten to bring inthe legislation for the prohibition of all metallic additives?! (So says the internationalstandards and recommendations!). Metal additives, from motor fuels, are leading topollution: engines, catalysts, the people, all living creatures, all plants, the land, water,and air! Fig. 7 World companies offer other metal additives The worst feature of particle emissions (largely as a result of metals in fuelsand additives) is to stimulate cancerous diseases. What is the specific of PM emissions from conventional (and perhaps modern)Diesel engines? The conventional diesel engines emit large amounts of particlesduring starting, cold engine operation at full load and when the post-injection is occur.16
  • Ecology of transport and global change earth’s climateParticles from diesel engines are dangerous because they behave like hoveringobjects with large surfaces. Nevertheless they belong the group of hygroscopicallysubstances that are beautiful and easy can to penetrate to all parts of the lungsbecause they are less (PM2) from human bronchioles (PM5 = 0.006 mm). Sources of PM emissions in road transport are listed in Table 2.Table 2 Sources of PM emissions in road transport Sources of PM PM10, mg/km Commercial vehicles with diesel engine 380 - 1 000 Passenger cars without catalyst 3 - 900 Truck tires 400 – 3 000 Road surface 375 - 11 700 It is absurd to make partial and unrealistic regulations on toxic, PM and CO2emission quantities, without thorough environmental assessment of the automobile andoil industries, transport and traffic in general. Fig. 8 shows a simultaneous decrease inemission of toxic components with advanced catalytic technologies. Also visible is the increase in the amount of carbon dioxide or in other words,the fuel consumption. In passenger cars the new ecological regulations result inincreased fuel consumption by about 10% and in the trucks up to 15%. Just such aconclusion confirmed by fig. 9. Fig. 8 Unrealized European Forecasts Fig. 9 Regulations on reducing toxic by 2012 (According to the “auto-oil" emissions is regularly followed by the program) increase in fuel consumption Just what to expect from modern models of passenger cars with air powerperformance over 200 HP and an electronic limited speed (usually 250 kilometres perhour)? They are regularly equipped with multi-degree mechanical (six to eight degrees)or automatic transmission, a lot of (useful and useless) electronics and electronicstability control for wanton entry into curve on the road? Statistical processing of roll-over vehicle after the collision shows that the SUV- Sport utility vehicle (because of the high centre of gravity) and speed of vehiclesprone to roll-over and at such time the typical security measures are insufficient.Braking power must be nearly ten times the engine power. Therefore the wheels is 17
  • Peši R., Veinovi S.scratching and scraping the surface of road plucking the pieces of asphalt and rubber.The largest emissions of PM is from wearing asphalt road surface, tires and allfrictional surface such as brakes and clutches, fig. 10a.Fig. 10a The high-speed, acceleration and Fig. 10b EU directives: the road surface - braking tires fall apart asphalt - contain carcinogenic components Launched particle were dispersing in the surrounding air which breathes allliving organisms later, and all particles contain carcinogenic ingredients, fig. 10b. 1- fuel pump, 2- pressure sensor, 3- a tube for fuel under pressure, 4- pressure regulator, 5- injectors, 6- gasoline tank, 7- LPG tank, 8- ECU for mixing and selecting of fuels Fig. 11 New motor technology enables the use of a combination of fuel as a substitute for additives The large tankers are already used the crude oil which they are transporting asfuel for their engines. All army in the world in the project tasks prescribe multi-fuelpower for combat vehicles. Aviation fuel is not shared according to value of octaneand/or cetane but they shared according to caloric value of fuel which determines the18
  • Ecology of transport and global change earth’s climateaction radius (Range!). Strategic Development Goals reveal that the future for roadtransport -and not only for him- in vehicles which use ecological fuel. The current oilrefinery with its irrational consumption of the oil and the high pollution will acceleratethe disappearance of fossil energy reserves [14,15]. Figure 11 shows a demonstration of modern engine technologies. Liquefiedpetroleum gas (LPG) has over 100 octane value and therefore LPG has a role ofadditives. An analogue picture we can apply to a combination of diesel fuel andbiodiesel (biodiesel have a Cetane number of 60 to 80) and again the biodiesel havethe role of additives to improve inflammability of diesel fuel.5. CONCLUSION A review of the research literature concerning the environmental consequencesof increased levels of atmospheric carbon dioxide leads to the conclusion thatincreases during the 20th and early 21st centuries have produced no deleterious effectsupon Earth’s weather and climate. Increased carbon dioxide has, however, markedlyincreased plant growth. Predictions of harmful climatic effects due to future increasesin hydrocarbon use and minor green house gases like CO2 do not conform to currentexperimental knowledge. Human use of coal, oil, and natural gas has not harmfully warmed the Earth,and the extrapolation of current trends shows that it will not do so in the foreseeablefuture. The CO2 produced does, however, accelerate the growth rates of plants andalso permits plants to grow in drier regions. Animal life, which depends upon plants,also flourishes, and the diversity of plant and animal life is increased. Human activities are producing part of the rise in CO2 in the atmosphere.Mankind is moving the carbon in coal, oil, and natural gas from be low ground to theatmosphere, where it is available for conversion into living things. We are living in anincreasingly lush environment of plants and animals as a result of this CO2 increase.Our children will therefore enjoy an Earth with far more plant and animal life than thatwith which we now are blessed. Such are the signs of nature: there are no pure energy form, no clean fuel, noclean engine, and neither clean vehicle. Actual reserves and natural gifts – coal,petroleum and gas – should be used less as power sources and more as rawmaterials. The greatest contribution to ecological prolongation of life on our planet isgiven by rational, economical technologies and products that carefully engage the giftsof nature. The term transport ecology or ecology in transport implies a complex balancingof the impact of every anthropogenic activity on the environment according to the lawsof nature! Unilaterally reduction emissions only one component can not to declare asenvironmentally, especially when it leads to even greater negative consequences onthe other side.ACKNOWLEDGMENTS The paper is the result of the research within the project TR 35041 financed bythe Ministry of Science and Technological Development of the Republic of Serbia. 19
  • Peši R., Veinovi S.LITERATURE[1] Robinson A., et al., (2007.) Environmental Effects of Increased Atmospheric Carbon Dioxide, Journal of American Physicians and Surgeons, vol.12, no. 3 pp.79-90.[2] Petition project, - http://www.oism.org/pproject/ , accessed on 2011-04-07[3] Milutin Milankovitch - http://earthobservatory.nasa.gov/Features/Milankovitch/, accessed on 2011-04-07.[4] Spasova D. et al, (2007). Milutin Milankovitch a traveler through distant worlds and times. Ministry of Environmental Protection of Republic Serbia, Belgrade.[5] Hays, J.D.; Imbrie, J.; Shackleton, N.J. (1976). "Variations in the Earths Orbit: Pacemaker of the Ice Ages". Science 194 (4270): 1121–1132.[6] National Research Council, (1982.) Solar Variability, Weather, and Climate, Washington, D.C.: National Academy Press, 1982, p. 7.[7] Definition: Renewable Energy Resource - http://www.cpast.org/Articles/ fetch.adp?topicnum=11, accessed on 2011-04-07[8] TREIAs definition of renewable energy has been adopted by the Texas legislature- http://www.treia.org/mc/page.do?sitePageId=49495, accessed on 2011-04-07.[9] "Renewable" Energy: In Search of Definition - http://masterresource.org/?p=1643, accessed on 2010-09-107.[10] Sustainable bioenergy: a framework for decision makers - http://esa.un.org/un- energy/pdf/susdev. Biofuels.FAO.pdf, accessed on 2010-09-107.[11] Biofuels: Impact on agriculture modest says Commission - http://www.euractiv.com/en/climate-environment/biofuels-impact-agriculture- modest-commission/article-165913, accessed on 2010-12-07.[12] COM(2008) 19, Directive of the European parliament and of the council on the promotion of the use of energy from renewable sources, Brussels, 23.1.2008, COM(2008) 19 final[13] Biofuel-makers denounce target downgrade - http://www.euractiv.com/en/ transport/biofuel-makers-denounce-target-downgrade/article-175298, accessed on 2010-12-07.[14] Peši R., Davini A., Veinovi S., (2005.) One engine for all fuels – one fuel for all engines, Proceedings ISBN 86-80941-30-1 – Paper EAEC05YU-EN01, 10th EAEC European Automotive Congress, Belgrade 2005, pp. 1-10.[15] Peši R., Davini A., Veinovi S., (2008.) New engine method for biodiesel cetane number testing, Thermal Science, vol. 12, no. 1, pp. 125-138.20
  • DESIGN CONSTRAINTS AND ROBUST DESIGN AS THE MODERN APPROACH TO MECHANICAL STRUCTURE DEVELOPMENT 1 Milosav OgnjanoviSummary: Actual approaches in technical systems design understand horizontalintegration of various scientific and technological knowledge and various market anduser needs in harmony with human environment. Robust design provides high level ofdesign results in the first attempt and design structures insensitive of service conditionsvariation. Axiomatic method in connection with robust design gives possibility fordesign parameters definition using various design constraints. The article has intentionto present and analysed relations between these methods and approaches. The casestudy of power transmission (PT) components is carried out in order to presentefficiency of the new approaches in design parameters selection and calculation.Reliability, vibrations and noise as design constraints in the stage of the Embodimentdesign of PT components make conditions for axiomatic method application androbustness provision. Reliability as the design constraint is defined and modelled in aspecific way suitable for this purpose and application. Also, the model of gearvibrations and gear units noise generation is presented in a new way suitable forapplying as the design constraint. Those design constraints provide design parametersdefinition in an efficient way, with high-level service quality indicators. The presentedmodels are based on a great volume of experimental data about service conditionsprobability, gear and bearing failure probability, gear units vibration and modalbehaviour etc. Theoretical knowledge and models are insufficient yet to provide thenecessary data. The article contains presentation of testing methods and dataprocessing oriented to provide data necessary for the application in the suggestedapproach.1. INTRODUCTION Tendency to achieve better design performance imposes necessity for aHolistic approach that also considers integration in product development. Recentresearch on integrated design includes system integration, requirements integration,knowledge integration, and method and process integration [1]. System integration hasa motivation to support the concept of “whole system” design, as opposed to itsseparate components design. Requirements integration considers „vertical“ and1 Prof. dr Milosav Ognjanovi , University of Belgrade, Faculty of Mechanical Engineering, Kraljice Marije 16,11120 Belgrade, Serbia, mognjanovic@mas.bg.ac.rs 21
  • M. Ognjanovi„horizontal“ integration. Vertical integration considers development of productrequirements through different design stages similar to simultaneous (concurrent)engineering principle. Actual trend of integration is horizontal one which considersdifferent areas of product development because products are becoming more multi-disciplinary and their boundaries are expanding especially in the sense of thenumerous new requirements. In Fig.1 some of the possible fields of productdevelopment are presented in the form of horizontal integration. Except fromrobustness, the product needs to satisfy aesthetic, ergonomic, eco and otherrequirements. Holistic approach can provide effective results. Also, separate design ofmultidisciplinary products is not acceptable, and therefore mechanical and electronicdesign has to be integrated. Biomimetric design [2] has an important role in searchingfor new principles and solutions in the area of biological systems, in order to apply intechnical systems, but integrated in a holistic design. Software design Robust Biomimetric design design Mechanical Electronic Horizontal design design integration of product development Aesthetic and design Ergonomic design design Eco - design Fig. 1 Horizontal integration of product development and design areas2. ROBUST DESIGN OF PT COMPONENTS Robust design means the technical products insensitive to variation ofoperating conditions and also products which are successfully designed in the firstattempt. Power transmission components operate in extremely random conditions.Operating regimes (loads, speeds, etc), production conditions and failure processesare random. It is known that random processes are possible to identify, present andanalyze only using the experimental approach supported by statistical indicators.22
  • Design constraints and robust design as the modern appr. to mech. structures developmentPower transmission operating regime for a certain machine system (vehicle, dredge,etc.) is possible to identify by systematic measurement in service conditions. Possibilityof power transmission components (gears, bearings, sealing sets, etc.) failure can beidentified by failure probability which needs extensive laboratory tests of the listedcomponents. The relation between service regime and failure probability leads tocomponent reliability. Using these elementary reliabilities as design limitations(constraints) for design parameters definition, the robustness of power transmissionunit is fulfilled. Similar situation occurs with the vibration and noise of powertransmission components. The level and frequency structure of vibration and noiseproduced by these components can be used as design constraints for designparameters definition and harmonization of their interaction. The relations betweenoperating conditions and component parameters and dynamic responses have to beharmonized using experimental approach. Fig. 2 Power Transmission Systems Design – PTSD model For the purpose of power transmission system design the specific procedure isestablished and presented in Fig. 2 in the form of the design model. The PTSD (PowerTransmission Systems Design) model consists of four modules which respect generaldesign procedure of technical systems and specific needs for the design of PTsystems. The first is Solution Module which is oriented to the creation of conceptualsolutions for certain service conditions. Power transmission systems in conceptualsense are a variation structure i.e. conceptual solutions are the result of variouscombinations of gear pairs, shafts, bearings etc. In interactive communication themodule offers all possible combinations and stores in the Conceptual Base. The nextLAHP-module has the task to adapt limitations and constraints to every conceptual 23
  • M. Ognjanovisolution and to every design component. Limitation Analytic Hierarchy Processing –LAHP module divides the conceptual design in sub-conceptions or function carriers,and for that structure the processing limitations and constraints in hierarchy order.General transmission ratio of the system is decomposed to the level of everytransmission stage. General value of the system reliability (chosen in advance) ishierarchically decomposed to the level of possible failure. The LAHP-module is the keypart of that approach which supports reverse calculation in order to fulfilled one of thefeatures of design robustness i.e. to design the system with a chosen general level ofreliability. The next is Design Parameters Definition – DPD module based on thecalculation of design parameters, especially dimensions, using axiomatic approach. Byobserving the axiomatic rules and by inclusion of design constrains this module fulfilsthe features of robustness. The last module is Priority Module whose task is to checkwhich design solution satisfies service limitations, such as volume, weight, efficiency,cost etc. in a better way. The calculation of priority indicators is interactive and givespossibility for additional corrections and adoptions. The main feature of design robustness contains DPD module based onaxiomatic rules. These are the two axioms, the axiom of independency and the axiomof information minimum. In Fig.3 the principle of DPD module based on axiomatic rulesis presented. Functional Requirements FR of every design component are defined byservice conditions, where the system operates. This FR is necessary to transform intoDesign Parameters – DP of design component. Transformation matrix [A] established -1by Suh [3] for the relation in Fig.3 is inverse matrix [A] . Numeric values of matrixmembers define the relations in design component, which are constrained bynumerous limitations, such as safety or reliability, stiffness, standards, rules, etc.These limitations and constraints are the result of service conditions, which is deductedby LAHP module to the level of design component. For the purpose to present thisrelation more clearly, the following example is processed. Fig. 3 Relations in DPD module In Fig. 4 the example is presented. The assembly of the gear, shaft andbearing is defined by a great collection of design parameters, especially dimensions.The calculation of dimensions is reduced to the three dimensions, gear diameter d,gear width b and shaft diameter dsh. In this way, the axiom of information minimum is24
  • Design constraints and robust design as the modern appr. to mech. structures developmentfulfilled. Other dimensions are in relation with those calculated. Matrix [G] (Fig.4) is theshape vector which defines transformation of parameters in the all shape dimensions.This is the shape parameterization where varying of the shape parameters varies thecomplete shape and dimensions. In Fig.4 are presented the two shapes of the sameassembly obtained in this way. Similar approach is incorporated in CAD tools for theshape modelling. Fig. 4 Example of DP minimization and variation The structure of the matrix [A] according to Suh can be uncoupled, coupled anddecoupled. The ideal situation is with uncoupled matrix where one DP is responsiblefor one FR. Real situation is more complex. In order to obtain the decoupled matrix of -1transformation, the matrix [A] is presented in the form of matrix [C] in the followingform. d c11 0 0 0 T1 3 d sh 0 c22 0 0 ….(1) T C 0 0 c33 0 T SE 0 0 0 c44 1 The axiom of independency is conditionally fulfilled. The members of matrix [C]and design parameters d, dsh, carrying capacity of the bearings C and seal typeindicator SE have to be calculated successively. The member c11 is in relation withelementary reliability R1 against wear failure of gear pair, c11 f R1 . After thecalculation of gear diameter d, it is possible to calculate the shaft diameter because theshaft loads depend of the gear diameter, i.e. c22 f d , R2 , and also of the shaftreliability R2. The next step is calculation of bearing carrying capacity using matrixmember c33 which is in relation with the both diameters d and dsh, the total number of 25
  • M. Ognjanovibearing revolutions along the service life n and of bearing reliability R3, i.e.c33 f d , d sh , n , R3 . At the end, the matrix member c44 is in relation to the shaftdiameter dsh, to the total number of the shaft revolutions n and to seal reliability R4, i.e.c44 f d sh , n , R4 . The main feature of robustness is covered by reliability. Thevalues of calculated parameters have to be insensitive to service conditions varying.For example, gear diameter calculation is in the form of T 13 ……. (2) d k11 3 2 c11 Hdes Tmax Hdes Design available gear teeth flank stress Hdes is in direct relation to reliabilityagainst teeth flank failure R1. Unreliability Fp=1-R1 is the complex function of serviceconditions probability p and of failure probability PF under these service conditions, i.e.Fp=p PF. If in the service life the gear pair is exposed to the flank stress H1 with n 1cycles and to stress H2 with n 2 cycles and to H3 with n 3 cycles (Fig. 5), gear wearunreliability can be calculated as 3 ni Hi i Fp pi PFi pi n PFi 1 e i i 1 ; ; ………….. (3) Teeth wear (failure) probability PFi is presented by Weibull’s functions, wherethe parameters of those functions i and i are defined for every Hi and n i (see Fig.5).For this purpose, it is necessary to have the area of failure probability for a certain gearpair, which can be obtained by extensive gear wear testing. If calculated unreliability isclose to unreliability which is defined as design constraint, the maximal stress can bechosen as design available stress, i.e. Hdes= H1. If it is not, it is necessary to changethe relations in Fig. 5 and to repeat the calculation. However, that definition of Hdesincludes in this way all randomness and variations of service conditions, the calculateddesign parameters are insensitive to service conditions varying [4]. Fig. 5 Relation between service stress varying and gear wear probability26
  • Design constraints and robust design as the modern appr. to mech. structures development3. RELIABILITY AS DESIGN CONSTRAINT The presented approach in the design of power transmission componentstowards the new approaches as axiomatic and robust design can be successful with agreat volume of experimental data only. The new methods can succeed in the designprocess [5] but remain to be carried out towards experimental data. However, the datahave to be oriented and adapted to be suitable for application in a new way. Reliabilityis the term with a very wide area of applications. For the purpose of applying the newmethods and approaches, the reliability is defined in a specific way (equ.3). The mainfeatures of reliability as design constraint are the following. Firstly, elementary reliabilityis connected to possible failure, not to the component of the system. In one possiblefailure a few components can participate or one component can be exposed to morethan one failure. Secondly, this elementary reliability is composed of causeprobabilities which produce failure in order to avoid possible failures by designactivities (Design Parameters Definition). In this regard, the elementary reliability iscomplex probability composed of operating stress probability and failure probabilityunder that operating stress. Both of these probabilities are the result of extensiveexperimental research. For the purpose of power transmission components design, theexperiments and the data processing contains a few sub-fields: failure probabilitytesting of power transmission components; then measurement and statistical analysisof service loads and load (stress) spectrum creation which can represent the wholeservice life of every component; reliability testing of gear transmission units or theentire power transmission system. PF(N) log HN PF=0.9 PF=0.1 PF( H a) b) logN Fig. 6 Gear failure probability testing: a) back-to-back testing rig, b) the range of gear failure probability distribution Testing of failure probability of gear transmission components is an extremelyextensive process of laboratory testing. It is necessary to make numerous tests andthen apply statistical data processing in order to obtain the range of failure probabilitydistribution. In Figure 6a back-to-back testing rig for gear wear testing is presented. Inorder to obtain the field of wear probability distribution (Fig.6b), it is necessary to testthe three sets of gear pairs and define the three Weibull’s distribution functions. Usingthese functions and logarithmic scale, the range of failure probability distribution isdefined. This range is an important source of failure probability data for every stresslevel or for every stress cycles number (see Fig.5). 27
  • M. Ognjanovi Testing of gear train transmission reliability can be carried out with the kind ofobjectives and tasks. Complete reliability test of this system can be extremelyextensive and it is very difficult to accelerate this test. In order to accelerate and tomake efficient experimental approach, a combination of calculation and experimentscan provide good results in a shorter time. Reliability model of gear drives based onthe results of failure probabilities and load spectrums of gear drive componentsprovides possibility to calculate reliability for any of service conditions. Reliability testsin this sense are directed to the check of calculated results for chosen conditions. InFig.7 the testing of car gearboxes reliabilities is presented. For this purpose back-to-back system is used, which provides a long-time testing under full load and speed ofrotation with a small consumption of energy. Testing is carried out according to loadspectrum obtained by service conditions analysis Fig. 7 Reliability testing of car gearboxes in back-to-back system4. VIBRATION AND NOISE AS DESIGN CONSTRAINTS In mechanical systems (especially in gear transmission systems), there existvibrations caused by fluctuating forces (forced vibrations) and natural vibrations whichare the result of relation between design parameters and disturbance processes in thesystem. Besides, in these systems there is one specific kind of vibrations caused bysuccessive teeth impacts. The vibration caused by impact becomes free damped afterthat impact, and attenuates (disappears) in a short time by internal damping. The gearmeshing process is a very good example in this sense (Fig. 8a). Elastic deformationsof previous teeth in mesh replace position of first teeth contact point from position Ainto position A’. Besides that, the position A’ is out of contact line, the teeth begin tomesh with impact. Every impact generates free vibration with the natural frequency ofthe system fn. This vibration attenuates in a short time by internal damping of thematerial and by friction. The next impacts which are repeated with the frequency f (Fig.8b) restore free natural vibrations again and again. These vibrations can be calledrestorable free vibrations. Figure 8b shows the measured results of these vibrations forthe gear speed of rotation 250 rpm, the teeth number z=41, i.e. the teeth mesh(impact) frequency f=nz/60=170Hz, and the time between the impacts 1/f=0.006seconds. The natural frequency of the gear system is fn=2200Hz and the time period ofnatural vibration is 1/fn=0.00045 seconds. The maximal amplitude level for thisexample is R1 4g (g-earth acceleration). The impact force, Fc=vc(cme)0.5, isproportional to the collision velocity vc, the teeth stiffness c and the mass me which isequivalent to the reduced rotating masses mr.28
  • Design constraints and robust design as the modern appr. to mech. structures development 1/fn R1 A’ a) b)Fig. 8 Restorable angular free vibrations: a) individual teeth impact and free vibrations, b) time function of restorable free vibrations for low rotation speeds Figure 9a presents the results of measured gear vibrations. The total level ofvibrations increases until the resonance is (9000rpm-Fig.9a). In the supercritical teethmesh frequency range, the vibration level slightly increases or fluctuates close to thesame level. In the case of forced vibrations the level of vibrations decreases (line 3 Fig.9b). It can be explained by the domination of restorable free vibrations (Fig. 8). b) c) a) d) Fig. 9 Restorable free vibrations caused by gear meshing: a) measured results, b) comparing of response curves of forced (3) and restorable (2) vibrations, c) frequency spectrum for critical mesh-frequency, d) frequency spectrum in supercritical mesh-frequency range [6] 29
  • M. Ognjanovi Evidence for the specific nature of gear vibrations is frequency spectrumsobtained by the FFT analyses (Fig 9c and d). The first one (Fig.9c) shows resonantvibrations, when the teeth mesh frequency f is equal to the natural frequency of thegear pair, fn. This natural frequency responds with a natural vibration level close to 60g.For a higher mesh frequency f>>fn (supercritical mesh frequency range) (Fig.9d), thefrequency spectrum is dominated by a set of natural frequencies fni with significantlylower responses (less than 9g). One can notice that there is no response for the testedmesh frequency f=17066Hz, but only for natural frequencies. With the variation ofmesh frequency f, the set of responding natural frequencies fni varies, and theirresponse levels change. This shows that the modal structure of the system is notstable and that the dissipation of disturbance energy in this domain is significantlyhigher. That is the reason why the total vibration level for f>>fn (Fig. 9a) varies. a) b) c) Fig. 10 Testing rigs for gear vibration measurement and analysis: a) driving sub-system with speed and load varying, b) measurement of angular vibration until to 6000 rpm, c) measurement of linear vibration until to 40000 rpm. In Figure 10 are presented testing rigs with close power circulation (back-to-back system) used for gear vibration testing. Testing rigs are prepared for variation ofgear design parameters, load (torque), speed of rotation until it is 40,000 rpm, and gearvibration (linear and angular) and noise measurement and frequency analysis. Vibrations of the system are the result of interaction between disturbance in thesystem and sensitivity (response) of the system to disturbance. The interaction is acomplex process and the results cannot be predictable with the exact value. Numericanalysis using the FEM and modal testing provides results for the purpose of thementioned relation prediction. In Fig.11 an example of gear drive housing analysis ispresented. Vibration energy and noise emitted by housing walls are a part ofdisturbance energy caused by teeth impacts, load and stiffness fluctuation etc. Energyabsorbed into the elastic structure of machine parts (for example, gears) is transmitted30
  • Design constraints and robust design as the modern appr. to mech. structures developmentthrough the structure and a small part is emitted into the surroundings in the form ofvibration energy and noise. The process of absorbing and transmitting this energythrough the elastic structure of a machine system is interesting to be treated on thepresented example of gear transmission covers (housings). The mechanisms oftransmission, attenuation, amplifying, frequency modulation etc of elastic waves andemission into the surroundings are analyzed using numerical and experimentalmethods. The cover of the system (gear transmission housing) is the most effective inthis process. This approach contains the analysis of disturbance energy transmissionthrough machine parts, attenuation in the contacts, losing inside the parts or amplifyingby excitation of natural vibrations. Modal analysis and modal testing of the geartransmission housing have enabled identification and proving of these processes.Numerical and experimental impact energy transmission through housing enabled toidentify the conditions for a certain modal shape excitation. Also, these results enabledto define mechanism of frequency modulation of emitted noise and vibrations in thesurroundings. The measurement of the noise emitted in the surroundings was alsocarried out to prove the starting hypothesis and developed mechanism of disturbancetransmission, and to define the value of energy transmission ratio. a) b) Fig. 11 Example of sensitivity analysis of gear unit housing: a) modal analysis using FEM, b) modal testing [7] Analytic hierarchy processing (LAHP module –Fig. 2) in the case of vibrationand noise limitations (constraints) is the deduction of the total vibration and noise level 31
  • M. Ognjanoviof the system to the vibration levels of the system components. The process ofdeduction can be characterised by a number of the following vibration features. Numerous disturbance sources and vibration and noise sources are in a very complex relation. The difference between the total vibration or noise level and an individual source with maximal effect can be small. The level of vibration in the sub-critical frequency range is proportional to the ratio between disturbance frequency and natural frequency. The level of vibrations in the resonant frequency range is proportional to damping characteristics of the system. The level of vibrations in the supercritical frequency range is proportional to disturbance intensity and frequency. According to the mentioned problem, the Limitation Analytic HierarchyProcessing - LAHP consists of disturbance frequency calculation and vibration levellimits definition based on the type of the system. Vibration limits of the system sub-assemblies are possible to define by varying design parameters in the DPD module. The vibration level of design solutions can be reduced by a set of designparameters corrections which have to be harmonised and lead to vibration levelminimisation. The Campbell’s diagram (Fig.12) shows the relation between thedisturbance frequencies kf (teeth mesh frequency f) and the natural frequencies of thesystem fni. The high level of vibration is equality of these frequencies and is marked inFigure 12 by signed points. With the aim to minimise the level of vibration, two groupsof design rules are possible to define. The first of them contains variations of designparameters which can vary the disturbance (service) frequencies kf and reducedisturbance forces intensity. The second group of design actions contains variations ofdesign parameters which can vary the natural frequencies fni with the aim to avoidequality with the service frequencies. For both groups there exist numerous researchresults which are necessary to apply. When this passive approach to vibration andnoise reduction is not enough, the active approach can be the next step. By generatingopposite vibration and noise, absolutely steady and silent gear transmission units arepossible to obtain. n f f 2f 3f 3f n5 2f fn5 n4 f n3 fn4 fn3 n2 n1 fn2 fn1 f n Fig. 12 Campbell’s diagram of relation between disturbance and natural frequencies in the system32
  • Design constraints and robust design as the modern appr. to mech. structures development5. NEW MATERIALS IN MODERN DESIGN Choice of material in design process is the top decision which is result ofimpact of material characteristics and necessary features of mechanical components.Designers experience and knowledge about traditional materials give the main supportfor these decisions. Contradictions and contradictory effects in choice of materials veryoften make them not fully successful. It is no material with characteristics which cancover necessary features of components. Additionally some of characteristics makenegative effects. Creation of materials with certain necessary characteristics is the taskof designers and material science. The common name for materials of this type ofmaterials is composite materials. Development of these materials was started forextreme specific applications. Now it is relative wide area of various materials whichcovering many areas of traditional materials application. For example traditionalmaterial for gears is the group of various steels. These materials have good volumeand surface strength and durability but support vibration and noise generation. Plasticgears have no enough level of strength and durability. Special way of gear productionof composite material can satisfy of all necessary features. Special gear systems inautomotive and similar areas for specific kind of motion (Fig. 13a) have to be verysilent and reliable. The only good solution is special composite material. a) b) Fig. 13 The new design and new kind of material production Technology of production of new (composite) materials is various, starting fromcasting similar to plastic until to very modern rapid prototyping (production)technologies. Materialisation of 3D CAD models using 3D printers (Fig.13b) istechnology in expansion. This technology start with usage of plastic materials, continueby using various kinds of powder including metal powder and sintering, until to usageof malty material printing in order to produce composites. Rapid production of matricesfor composite casting or similar production is good implementation of 3D printing in thetrend of technical system components development with certain characteristics ofmaterial. Biomimetrics (Bionics) is the field of science oriented to transforming ofbiological systems into technical. In the area of materials (composites) this directionand methodology is very perspective in the future. Exist wide area of various kind ofnatural composites in flora as wood, bamboos etc., at animals, insects etc. 33
  • M. Ognjanovi6. CONCLUSION Modern approach to mechanical structure development very often is thesegment of horizontal integrated multidisciplinary work. Power transmissioncomponents in the future have to be of high quality integrated with electronic controlsystems, software and intelligent systems, etc. Design methodology has to be efficientand provide high level of quality indicators. Robust and axiomatic design usingreliability, vibration and noise as design constraints are guaranty for success.1. The robust design can be provided by precise definition of constraints and by these constraints application in the design process. For this purpose, the design process is defined as a set of four modules: Solution module, LAHP module, DPD module and Priority Rating module.2. Reliability is defined in a specific way as a complex probability of service stress probability and failure probability of machine parts exposed to this stress. The procedure of design parameters definition (DPD) using elementary reliability as a constraint is established.3. The design for vibration and noise offers an approach to design parameters harmonisation aiming at vibration and noise level reduction. For the purpose to vibration and noise reduction, the mechanism of vibration and noise generation in power transmission systems is presented.4. Tendencies of the new materials development are also presented.ACKNOWLEDGEMENT This article is a contribution to the Ministry of Science and Education of Serbiafunded projects TR 035006.LITERATURE[1] Liu, S., Boyle, I.M., (2009). Engineering design: perspectives, challenges, and recent advances, Journal of Engineering Design, Vol. 20, no.1, p. 7–19[2] Lenau, T, (2009). Biomimetrics as a design methodology – Possibilities and challenges, CD Proceedings of International Conference on Engineering Design ICED’09, Stanford 2009, p. 5.121-5.132.[3] Armen Z, James K, Lusine, B. (2007). Modeling and analysis of system robustness, Journal of Engineering Design Vol.18, p. 243-263.[4] Ognjanovic M., Benur M. (2011). Experimental Research for Robust Design of Power Transmission Components, Meccanica, DOI: 10.1007/s11012-010-9331-y[5] Liang, J., Mourelatos, Z.P., Nikolaidis, E. (2007). A Single-loop Approach for System Reliability-based Design Optimisation, - Journal of Mechanical Design, Vol.129, p.1215-1224[6] Ognjanovic M., Agemi F. (2010). Gear vibrations in supercritical mesh-frequency range caused by teeth impacts, Strojniski vestnik – Journal of Mechanical Engineering, Vol. 56, no.10, p. 653-662.[7] Ciric-Kostic, S., Ognjanovic, M. (2007). The Noise of Gear Transmission Units and the Role of Gearbox Walls. FME Transactions Vol. 35, p. 105-112.34
  • RENEWABLE ENERGY AS A DRIVER OF ECONOMIC GROWTH 1 Prof. dr. sc. Neven DuiSummary: The new European energy strategy, so called energy-climate package,sets up ambitious goals of having 20% renewable energy in gross energy demand,reduce carbon dioxide emissions by 20% and improve energy efficiency by 20% by2020. One of the goals of this is to use the necessary need to reduce the influence onenvironment and increase the security of energy supply for opening new and betterquality jobs, regional development and reindustrialization of the European economy.This is also a potential for new development of Croatian, and also southeast Europeaneconomies, based on future inevitable investments that may be used to spearhead thelocal economy growth. Countries should approach this in an integral way, best for theirindividual economies. It will be necessary to lay down financial support mechanismsthat will create both, supply and demand for new technologies. On one side, financialmechanisms that open the way to increased investment in renewables and energyefficiency are being put in practice, like feed-in tariffs, but others are needed to supportinnovative development of new technologies which create jobs in industry andacademia. New financial mechanisms, often used in countries that successfully gainfrom increasing both supply and demand of renewables, are suggested. They wouldhelp finance the development of new products, by bringing together the researchcapacities of academia and industry, while in the same time avoiding extra budgetarystrains. Most of renewable energy technology, especially those that will stem fromapplication of Buildings directive, are quite simple and can be produced locally.Meanwhile, those technologies have high development costs, due to need toconstantly innovate, which can then be supported through the suggested financialmechanism.1 Prof. dr. sc. Neven Dui , Faculty of Mechanical Engineering and Naval Architecture (FSB) University ofZagreb 35
  • Neven Dui Renewable energy as a driver of economic growth DEMI 2011 Banja Luka, May 26/28, 2011 Prof.dr.sc. Neven Dui , University of Zagreb Faculty of Mechanical Engineering and Naval Architecture (UNIZAG FSB) © FSB 2011. EU energy context Security of energy supply • Import dependence from 50% to 70% by 2030 Employment and regional development policies • Deindustrialization and trade liberalization • “Boosting growth and jobs by meeting our climate change commitments” Mitigation of global warming Environmental protection Sustainable development © FSB 2011. Energy - climate package 2008 20-20-20 till 2020 20% RES 10% biofuels Energy neutral buildings from 2018 20% CO2 emission reduction 20% more energy efficiency Energy neutral buildings from 2018 © FSB 2011.36
  • Renewable Energy as a Driver of Economic Growth The new renewables Directive1. Sets mandatory national targets for renewable energy shares, including 10% biofuels share, in 2020 (Articles 3 and 5)2. Requires national action plans (Article 4)3. Standardises “guarantees of origin” (certifying the renewable origin of electricity or heat) and enables the transfer of these to provide flexibility to Member States (Articles 6, 7, 8, 9 and 10)4. Requires reduction of administrative and regulatory barriers to the growth of renewable energy (Article 12), improvements in provision of information and training (Article 13) and improves renewables’ access to the electricity grid (Article 14)5. Creates a sustainability regime for biofuels (Articles 15-18) © FSB 2011. The new renewables Directive © FSB 2011. Nova Direktiva za obnovljive izvore energije Obaveze zemalja Zapadnog Balkana u postizanju udjela obnovljivih izvora energije u ukupnoj potrošnji do 2020 © FSB 2011. 37
  • Neven Dui New RES Directive © FSB 2011. New RES Directive Plin Vjetar Neto promjena instaliranih kapaciteta za proizvodnju elektri ne energije u EU, 2000- 2010 [MW], EWEA PV © FSB 2011. RES-e in EU, 1996-2006 © FSB 2011.38
  • Renewable Energy as a Driver of Economic Growth Production of primary energy, EU EUROSTAT © FSB 2011. RES Directive - reforms of administrative and regulatory barrierssimplification and streamlined proceduresplanning authorities to consider renewableenergy and district heating and cooling systemsminimum levels of renewable energy in buildingcodes for new or refurbished buildingspromotion of energy efficient renewable energycertification regimes for installers; mutualrecognitionpriority access to the grid system for electricityfrom RESto develop grid infrastructureto review cost sharing rules © FSB 2011. RES Directive - Promotion of biofuelsEstablishes a biofuels sustainability schemeBiofuels not meeting the criteria will not beeligible for target counting, obligation schemes,tax exemptions or other supportMember States responsible for verification ofcompliance however Commission can decide that“certification schemes” give reliable proof ofcompliance10% biofuels by 2020• 2nd generation biofuels x 2• RESe + RESH2 x 2,5 © FSB 2011. 39
  • Neven Dui Energy package – ETS post 2012 EU wide ETS auctioning No free emissions for power generation sector (except cogeneration) – increasing the cost of electricity from coal © FSB 2011. Energy package – buildings New and reconstructed buildings zero energy from 2019 Define zero energy by 2010 Quota of old buildings to be reconstructed National Action Plans by 2011 Strict insulation regulation from 2006 for new and reconstructed buildings (>1000m2, 2009 - >50m2) Smart metering Feasibility of solar heating, heat pumps, biomass heating, cogeneration • Spain: obligation to supply 25% of heating • New package will include obligation Phase out incandescent light bulbs by 2011 Passive buildings ... More efficient appliances, boilers, etc. © FSB 2011. How to create RES market? How to create RES market? • Demand Supply • Feed-in tariff • Obligation © FSB 2011.40
  • Renewable Energy as a Driver of Economic GrowthRES-E support schemes in EU © FSB 2011.Financial Support Mechanisms • Feed-in tariff systems are characterized by a specific price normally set for a period of several years, which must be paid by electricity companies, usually distributors, to domestic producers of green electricity. A variant of the feed-in tariff scheme is the premium mechanism. • Green certificate system RES-E is sold at conventional power-market prices. In order to finance the additional cost, all or some consumers are obliged to purchase a certain number of green certificates from RES-E producers. • Tendering procedure the state places a series of tenders for the supply of RES-E which is then supplied on a contract basis at the price resulting from the tender. • Tax incentives are used as an additional policy tool. Tax incentives may be a tax credit or a cash payment or an exemption from tax obligations or low VAT. • Investment incentives: A common investment subsidy is a grant for the installation of capacity. © FSB 2011.Effectiveness of a promotion policy for onshore wind Source: OPTRES, 2007 © FSB 2011. 41
  • Neven Dui Evaluation of different FITs design options (Source: ISI 2008) © FSB 2011. International Feed-in Cooperation • Web-Site: Research Events Legislative documents Links http://www.feed-in-cooperation.org/ © FSB 2011. Feed-in tariffs (FITs) • Already applied in over 60 countries, states and provinces, • a highly effective tool for boosting the viability, and hence value, of the renewables industry, • FITs have been empirically proven to generate the fastest, lowest- cost deployment of renewable energy. • European Commission, SEC(2008) 57, update on renewable energy policies in the European Union (EU): ‘‘well-adapted feed in tariff regimes are generally the most efficient and effective support schemes for promoting renewable electricity’’ • 85% of all new wind capacity and nearly 100% of the new PV capacity since 1997 were installed in countries using FIT © FSB 2011.42
  • Renewable Energy as a Driver of Economic GrowthFeed-in Tariffs (FITs) • Fixed • Premium • Net meetring Progression of the remuneration level for wind onshore within the market option © FSB 2011.Feed-in tariff design • One of the most important aspects of a feed-in tariff design is the determination of the tariff level and the duration of support. • One possibility is to set the tariff level based on the electricity generation costs from renewable energy sources. Alternatively, the support level of RES-E can be based on the avoided external costs induced by RES generation. © FSB 2011.RES Electricity Cost – Supply Curves ADEG project calculation for 2010 (Croatia) © FSB 2011. 43
  • Neven Dui External cost 6 Global warming External costs, c€/kWh 5 Noise Materials 4 Crops Occup. Health 3 Public health Source DG TREN 2 1 0 Coal Oil Gas Nuclear Wind Biomass © FSB 2011. Feed-in tariff design • Should provide technology-specific tariff levels. • Investment for the plant • Other costs related to the project, such as expenses for licensing procedures • Operation and maintenance (O&M) costs • Fuel costs (in the case of biomass and biogas) • Inflation • Interest rates for the invested capital • Profit margins for investors • Rejection or curtailment © FSB 2011. FITs design • Stepped tariff design: different levels of remuneration are paid for electricity of the same RES-E technology. • flat tariff design : the opposite of a stepped tariff design e © FSB 2011.44
  • Renewable Energy as a Driver of Economic Growth 28.Guarantee of origin • ’ guarantee of origin’ means an electronic document which has the sole function of providing proof to a final customer that a given share or quantity of energy was produced from renewable sources as required by Article 3(6) of Directive 2003/54/EC; 28 © FSB 2011.Conclusions • RES-E support requires continuity and long term investment policy • Technology-specific tariff levels should be applied • Energy policy should provide mechanisms to ensure the penetration and to improve the integration of RES-E into the grid • A premium tariff option can be applied to increase market orientation • Tariff degression provides incentives for cost reductions © FSB 2011.Conclusions • Stepped tariffs may be applied to reflect different power generation costs within the same technology: May lead to higher administrative complexity and to reduced transparency If not properly designed overall efficiency of the system may be decreased • Extra premiums may help to reach policy goals • Monitoring of success of financial schemes and RES policies should be applied. • Revision and proposal of new tariffs every 4-5 years. © FSB 2011. 45
  • Neven Dui Conclusion An effective scheme is one that: • provides tariffs for all levels, from domestic to large- scale developments, • takes account of the level of development of each technology and benefits like (fuel savings, avoided emissions, local jobs, etc.) • guarantees long term investment security, • is administratively simple and easy to explain in order to ensure investors and public acceptance. © FSB 2011. © FSB 2011. Possible programmes in Croatia • 100000 solar roofs – hot water – 2000 jobs • 100000 heat pumps – heating and cooling – 1200 jobs • Refurbishing buildings to current regulation in 10 years Demand reduction 23 PJ, 10% less primary energy in Croatia Construction and building material industry – more than 25000 new jobs All can be local jobs and technologies © FSB 2011.46
  • Renewable Energy as a Driver of Economic Growth How to create RES market? How to create RES market? • Demand Supply • Project development • Design • Installing • Equipment production © FSB 2011. How to create RES market?How to create equipment supply?• Support to innovative R&D• Another 2020 goal, 3% GDP for innovations and R&D• Source? Emission tax? © FSB 2011. How to create RES market?Industry Academia partnership• Innovation Union 2020• Constant flow of innovation and human resources is necessary for competitive production• Supporting development is investing in future © FSB 2011. 47
  • Neven Dui Torwards 2050 Four pillars of Post carbon society Renewable Energy Buildings as Positive Power Plants Energy Storage Smart grids and Plug-in Vehicles © FSB 2011. Denmark SOURCE: STORIES project • Main RES: Wind, solar and biomass Main support schemes: 1. Wind power • On- and Offshore Committees • Feed-in tariffs Depend on the year the turbine was connected to the grid and its age Range 1,3-8 €cent/kWh • Tendering procedure (for offshore wind turbines) 2. PVs • Subsidies withdrawn in 01.01.2002 • Large scale demonstration projects carried out between 1998-2007 • R&D for new types of PV cells 3. Wave energy • R&D stage, in 2003 the first grid connected wave power plant went into operation © FSB 2011. SOURCE: STORIES project Finland • Main RES: hydro power and biomass Main support schemes: • Investment Subsidies: 30-40% depending on the novelty of project • Guaranteed Access to the Grid • Tax Subsidies • Feed-in Tariff or Green Certificates © FSB 2011.48
  • Renewable Energy as a Driver of Economic Growth SOURCE: STORIES projectFrance • Main RES: hydro power and biomass Main support schemes: • Feed-in tariffs Depend on the technology Range: 6,57 -40 €cent/kWh Duration: 15-20 years Special tariffs for Corsica and overseas Departments (PVs: 40 c€/kWh + premium for building integration of 15 c€/kWh Geothermal: 10 c€/kWh + premium for energy efficiency between 0 and 3 c€/kWh) • Tendering system for large renewable projects • Tax credit: 50%of the equipment cost for household installations • Reduction of VAT-Tax: 5,5% reduction in VAT for residential energy equipment © FSB 2011. SOURCE: STORIES projectGermany • Leader in wind energy, photovoltaics, solar thermal and biofuel sectors, both as RES-E producer and manufacturer • Main core of renewable Energy Act: Priority access for RES-E to the grid Priority transmission and distribution Grid operators must purchase the electricity produced from RES Equalisation of additional costs between all grid operators and electricity suppliers Feed-in tariffs Higher tariffs for geothermal, PVs and certain types of biomass Decreased tariffs for onshore wind installations at locations with very good yield Annual decrease of tariffs for all technologies except small hydro plants (1%- 6,5%) to take account of technical development © FSB 2011. SOURCE: STORIES projectGermany 2/2 • No specific support schemes for hybrid plants • Growing interest in exploring the potential of combined RES and energy storage systems particularly for hydrogen production applications First hybrid power plant (wind/biogas - 120MW and hydrogen-500kW) installed on grid in 2008 © FSB 2011. 49
  • Neven Dui SOURCE: STORIES project Spain • World’s second larger producer of wind energy • Feed-in tariffs: fixed total price+price premiums with recognising the environmental benefits Project developers should choose one of the following options: 1. Transfer electricity to the system through the transport or distribution grid 2. Sell the electricity on the wholesale electricity market Feed-in tariff and premium differentiate according to technology and the age of installation Feed-in tariffs are revised every four years Internal rate of return of 7% © FSB 2011. SOURCE: STORIES project Ireland Main support scheme: • Feed-in tariffs • Range between 5.7-7.2 €cent/kWh • Adjusted taking into account annual indexation • Tax Relief • Relief capped at 50% of all capital investments (excluding land), net of grants, on a single project • Extension of the relief until 31 December 2011 • Other incentives: • Ocean Energy Strategy, Organization of an all-island market, Establishment of Renewable Energy Development Group and Irish Energy Research Council © FSB 2011. SOURCE: STORIES project Ireland Other policy Incentives • Ocean energy strategy Launched in 2006 Increasing the capacity for research and development, both within academic institutions and commercial entities developing devices • All-Island market • Renewable Energy Development Group Focus on achievement of Ireland’s 2010 target for electricity from RES Considering policy options and support mechanisms available to Government to stimulate increased use of biomass and CHP Developing a strategy for the promotion of ocean energy • Irish Energy Research Council Coordination of energy research and technological development and innovation (RTDI) activities and deliver an integrated approach to energy RD&D © FSB 2011.50
  • Renewable Energy as a Driver of Economic Growth SOURCE: STORIES projectIreland 2/2 • No special regulation for RES storage systems • Study on: Implementation of a wind energy storage facility at Sorne Hill Wind Farm Different electricity storage technologies and their potential to address wind energy intermittency in Ireland • In operation: Desalination plant with load management on Inis Mean © FSB 2011. SOURCE: STORIES projectUK • Renewable Obligation Order: Obligation on electricity suppliers to source an increasing share of their power sales from RES Issue of Renewables Obligation Certificates Several technologies eligible under RO Penalty • Climate Change Levy: RES-E is excempted © FSB 2011. SOURCE: STORIES projectMalta • RES market still at an early stage • Substantial solar and wind potential • Support schemes: A fixed feed-in tariff of 46.6 €/MWh for PV installations below 3.7 kWp; and A reduction in value-added tax on solar systems from 15% to 5% © FSB 2011. 51
  • Neven Dui SOURCE: STORIES project Cyprus • The system of support to RES includes the following components: Definition of the price of kWh in the market. Provision of long-term contracts (15 years) Definition of financial incentives. Feed-in tariff system is implemented combined in many cases with capital subsidization. © FSB 2011. SOURCE: STORIES project Conclusions • Each country adopts different measures on promoting RES, depending on the RES type and the year of implementation • No special legislation or policy framework for hybrid power plants with the notable exception of Greece © FSB 2011. SOURCE: STORIES project Greece • Main RES: hydro power and active solar Main support scheme: • Feed-in tariffs Distinction between the interconnected system and the non- interconnected islands (particularly favourable tariffs for solar power plants) • Photovoltaic Development Plant Installation of 540MWp in the mainland Installation of 200MWp on the islands Geographical distribution Plants <150 kWp are exempted from the obligation to obtain a production license • Investment incentives Subsidies: 30-40% of the installation cost depending on RES technology, region and type of enterprise © FSB 2011.52
  • Renewable Energy as a Driver of Economic Growth SOURCE: STORIES projectGreece Hybrid power plants • Law 3468 sets the fundamentals for the promotion of hybrid power plants in the non-interconnected islands • Public consultation of a proposed Code by RAE for the operation of hybrid power plants (July 2008) • Remuneration of hybrid power plants for: Capacity availability Energy supplied to the grid from the hybrid station’s firm output units or RES units • Hybrid power plants have to pay for the electricity absorbed from the grid in order to fill their storage systems © FSB 2011.Greece SOURCE: STORIES project Hybrid power plants Remuneration of: • Capacity Availability: The price to be paid to the producer for the availability of the firm output units of the hybrid station cannot be less than the price paid for the availability of the units of a new incoming conventional plant of corresponding capacity • Energy produced from the hybrid station’s firm output units, exploiting the energy stored in the storage system, and supplied to the grid of the island: Conducted on the basis of the mean marginal cost of the conventional peak units operating on the island during the last year (annual mean variable cost) • Energy the RES units of the hybrid station supply directly to the Network of the island: Priced according to the existing feed-in tariff depending on RES technology (currently 87.42 €/MWh for wind energy in non-interconnected islands) • Energy the hybrid station absorbs from the Network for fulfilling its storage system: Conducted on the basis of the mean variable cost of the conventional base units operating on the island during the last year (annual mean variable cost) • Power produced in the RES units of the hybrid station and directly supplied to the Network of the island may be compensated to the power the hybrid station absorbs from that network for fulfilling its storage system © FSB 2011. SOURCE: STORIES projectGreece Hybrid power plants Hybrid power plants • PPC has installed two hybrid power plants Island of Kythnos: PVs+Batteries Island of Ikaria: Wind+hydro power • RAE received applications for installing hybrid power plants in Crete and Sikinos. Applications are currently being reviewed © FSB 2011. 53
  • Neven Dui THANK YOU FOR YOUR ATTENTION! ! Neven.Duic@fsb.hr © FSB 2011.54
  • MALICIOUS ECOLOGY ON VEHICLES AND TRAFFIC EXAMPLE Dr inž. Stojan Petrovi , Dr inž. Božidar Nikoli , Dr inž. Emil Hnatko, Dr inž. Jovo Mr a, dr inž. Stevan VeinoviSummary: During the epoch of human progress and ecology there is a big interestfor malware ecology as well protection against them. Identifying the entire value streamfor each product is the first step in ecological thinking. By improvement a technicalfacilities we degrease energetic and ecologic environment overload. For example,vehicles are the cases where to positive effect of complex energy balance over thepropulsion system in them. Around 60 million vehicles are produced per year and 800million vehicles are in traffic today in the world. The biggest power has been situated intransport sector: up to 15 times bigger of all energy resources in the EU and up to 10times bigger of all energy resources in Serbia.Such huge number of vehicles has, of course, a certain negative influence on natureand human environment. The overall pollution of environment –air, water and soil –enforcement genetic bases uncontrolled changes wherever the entire world of ourPlanet exists. The global conclusion is that there is no chance to change the road fleetfrom fossil fuels to electric power. 55
  • A. MECHANICS AND DESIGN
  • ABOUT RECEIPTE OF VARIANTS OF TECHNICAL OBJECTS Michail Leparov1, Georgi Dinev, Marieta Jancheva2Summary: One of the important problems of the design process of assembly units isreceiving of many variants to select the optimal variant by specific criteria. In this workare offered some ways for assembled unit. Every modification fulfil the conditions: Performs the same function as the main original technical product; Differs from the original product on characteristics of the parts (forms, number, location) and / or connections between them.The reports presented several ways for modifications of the article by heuristicalgorithms, and they are illustrated by examples.Key words: variants assembly unit, change, technical object, geometrical, design,methods for geometrical design.1. INTRODUCTION One of the important problems in the design process of the assembly unit (AU)has received a number of options to choose according to specific criteria of optimaloptions. Some methods for obtaining variants of TO are given in [2-7]. Under theversion of the assembled unit in this work means transformation of the product, wherethe new product differs from the source technical object (TO), but performs the samemain function. Through this work offers some new ways of getting the options variantsof given AU. Every variant of the option meets the conditions: Performs the same function as the main output product; Differs from the original object on characteristics of the parts (forms, number, location) and/or connections between them. The methods are obtained by logical analysis and are represented by heuristicalgorithms i.e set of guiding steps for search of solutions.1 Prof. PhD, Michail Leparov,Technical University of Sofia, Faculty of Mechanical Engineering, (e-mail:mleparov@tu-sofia.bg)2 Assoc. Prof. PhD Georgi Dinev, Technical University of Sofia, Faculty of Mechanical Engineering, (e-mail:gdinev@tu-sofia.bg)2 Asist. Prof. PhD Marieta Jancheva,Technical University of Sofia, Faculty of Mechanical Engineering, (e-mail: mmarieta@)mail.bg) 59
  • Michail Leparov, Georgi Dinev, Marieta Jancheva2. METHODS2.1 Method of “Repair"Algorithm1. Specify the problem in the source is TO.2. It is assumed TO that it must be "repaired" (correct) because it are damaged(literally or figuratively). Make a list of all possible failures. For this purpose, madelogical analysis, power analysis, polls "repair" firms, surveys among users and others.3. For each damage:3.1. Damage analyses, and taking into account what happens when it immediately,and to bring what changes related damage components.3.2. It is assumed that it TO begin to function only when executed operation of thedamage.3.3. Based on the analysis addresses the problem, to some extent resembles thesituation of damage.If necessary, carry out appropriate changes in TO.Example 11. Lets solve the problem which demands a new look and additional functions, orother realization of the TO function, depicted in Figure 1.2. a) button 1 after pressure blocked; b) spring 5break; c) the forces applied to the cover 2 screws 3 are developed and damage; d) efforts in the spring and buttons1 a cover 2 are deformed up in the middleupwards; e) efforts in the spring and screw the casing 4 is broken;3.1. a) button can not move; b) the button is not return to the starting position; c) cover 2 to move freely up and down and prevents the operation of the facility; d) the button 1 can do necessary working stroke; e) spring fell in the resulting hole;3.3. a) One possible solution is given in Figure 2; button and cover form a bayonet,pressure and rotation of 90 º projections button (fig.2b) stand outside the holes in thecover, so the button can not return to the starting position (return is realized in his newrotation of 90 º); b) fig.2v; object consists only of two-button detail and housing-related thread, inrewinding button when he moved in either direction (up, down); c) fig.2g; in rotation of the button it moves from the bottom with threaded body andthrown up from the lower spring where as sharply circuit, after a new rotation button iswrapped thread at the top of the bonnet on the development of the upper spring buttonmoves it down sharply, thus reducing the spark to turn the electric circuit; d) fig.2d; cover was replaced by an elastic membrane that also serves as a spring; e) figure 2, moving up is done manually, a return to the starting position, by usingspring.60
  • About reseipte of variants of technical objects Fig. 1 Original object "button”Fig. 2 New implementations of GF to TO "button” 61
  • Michail Leparov, Georgi Dinev, Marieta JanchevaExample 21. Lets solve the problem is demand for a new look and additional functions, or otherrealization of the function of TO “Door”.2. a) the key is broken and so far has remained in the lock; b) the spring that returns the lock is broken up; c) hob of the door is removed from one hinge; d) spring, which influence on locking plate is broken;3.1. a) can not put a key lock and unlock on the door; b) not influence lock on tongue; c) the door does not fulfil it’s main function-closing and opening a hole in the wall; d) the key is not influence on locking plate;3.3. a) door only locks closing; b) lock has two positions-the first is the traditional, which has normal function, thesecond is obtained after pressing a little lock in which the tongue is locked, and thus issuspended and open the door without a key; c) the door can be rotated to it’s base and becoming part of a window, as is hole inthe wall is partially open to air penetration, while access to space is limited; d) primary key does not affect the plate, and on the inclusion of a code devicewhich, after entering the correct code, unlatched the lock and allows the switch tomove the locking plate.2.2 Method "Paradoxes"Algorithm1. Pinpointing the problem.2. Determination of the action which is necessary to be realization to solve theproblem. Note: In the presence of TS can be detained and it’s main functions OFi. The effectof TO it’s main function, defined by a verb.3. The action (and any OFi):3.1. Searching opposite states, e.g. open-closed, include- exclude, move forward andbackward moves, etc.3.2. For each pair participating in the opposite state:3.2.1. Identify basic features. This can be done by defining these parts, e.g. throughthe dictionary.3.2.2. Incorporation is a sign of one with the sign of the second opposite state.Combinations can be of type "adjective-noun" or of another type.Combinations are formed properly grammatical terms.3.2.3. Analyze the combination. Record what is understood as meaning thecombination.3.2.4. As results of the analysis sought to decision the problem.Example1. Search a new realization of the closure of TO " Reservoir ".2. Open.3.1. Close-open (closed-open).62
  • About reseipte of variants of technical objects3.2.1. According to [1]: (Closed) - which is limited, there is no free access; (Open) - in which is free to go; - Restricted, proprietary, unavailable; - Free, passable.3.2.2. Some possible phrases are: - Limited freedom (free limitations); - Limited passable (passable limitations); - Unavailable freedom (free unavailability ); - Inaccessible passable (passable inaccessibility).3.2.3. One possible subjective interpretation is: (Free limitation) the movement of the cover is only one coordinate axis; (Walked limitation) obstacles, but it is possible to achieve "mechanism" to open; (Free inaccessibility) is insufficient space to open, it is destructive; (Walked unavailability) can be done "path" to inaccessibility;3.2.4. Some possible solutions for closing the tank are given in Fig.3 by letter symbolsin the figures correspond to letter names of the phrases from t.3.2.3. Fig.3 is a baselinefigure of fig.3b-cover slides along grooves on fig.3v-fold is somewhat cut into 4 partsreservoir can bend to fig.3g-shell is attached to the environmental technicalenvironment, and fig.3d-fitting cover to the body going through braces. Fig. 3 Some solutions to close of TO "Tank” 63
  • Michail Leparov, Georgi Dinev, Marieta Jancheva3. CONCLUSION In the work are considered the following methods: a) "Repair", and b)"Paradoxes". The methods allow a systematic search of variants. They can be used asa "manual" design, and as a basis for developing of system formalized for obtainingmodifications of random assembly unit.LITERATURE[1] Interpretative Dictionary of Bulgarian AZ, Team, S., MAG-77, 1997.[2] . , . „ ’ 2009 , ” , 2009.[3] Dinev G, . Leparov, M. Vicheva About search of variants of product with CAD, The 5 th Intern.Simp. “Shape, Mechanical and Industrial Design of Products in Mechanical Engineering 2008”, the Faculty of Technical Sciences, Novi Sad, Serbia, 2008.[4] . . , . „ ’ 2008 , ” , 2008.[5] ., . CAD , 17 . - . . . „ ” 2008, , 2008.[6] ., . , . CAD , . “ , , ”, , .4-5, 2008.[7]. . , . . . , 16 . - . . . “ -2007 “, , , 2007.64
  • ABOUT THE INTEGRATION OF ASSEMBLY UNITS Leparov M.1, Yancheva M.2Summary: One of problems in design process is integration of two already designedor purchased assembly units. In this work is propose a method called „Surfaces ofdetails” used to classified and lead the actions of searching for integration the twoexisting assembly units.The algorithm is illustrated by example. The results of the thiswork could be used in the design of mechanical and other products.Key words: integration of assembly unit, assembly unit, design1. INTRODUCTION In the process of design of technical products very often arises problem ofintegraion parts of the designed product, incl. assembly units. This process is intuitivethanks to the engineering experience of design team. In the technical books of designthis matter is not enough illustrated very well [1-4 ect.] The purpose of this work isrevealing of some opportunities to integrate the given assembly units (AU).2. METHOD OF ”SURFACE OF DETAILS” AlgorithmDefault: to AU 1. Analysis: 1.1. Clasification of requirements for each of the two outputs AU in terms oftheir functioning in order to deploy one off so as not to interfere with the functioning ofothers AU. 1.2. Clarifying the need for common links between both AU (if any), eg.Presence of generalized flows, particularly the disposal of a AU (component part)against the other one AU (component part) ect. 1.3. Choose one AU for primary, when he will join the ect. AU. Baisic is the AU,which apparently is more massive and which shall be the second AU (support AU).When difficulties of determination for such is considered any of them AU.1 Prof. PhD eng. Mihail Leparov, Bulgaria,1000 Sofia, Bvd. Kliment Ochridski-8, Technical University,Faculty of Mechanical Engineering, e-mail: mleparov@tu-sofia.bg2 Assistant Prof. PhD eng. Marieta Yancheva-Popova, Bulgaria,1000 Sofia, Bvd. Kliment Ochridski-8,Technical University, Faculty of Mechanical Engineering, e-mail: mmarieta@mail.bg 65
  • Leparov M., Yancheva M. 1.4. Classification of existing types of compounds to ensure the immobility ofboth AU. Classification may be more detailed or superficial. 1.5. Preselection ways of assembling both AU, and for this purpose to analyzethe classification of t.1.4 and reported characteristics of boht AU. 2. Systematization 2.1. Determination of potential surfaces of the support AU, which serve to join thebase AU. To this end: 2.1.1. Idetify parts (components) of the support AU. 2.1.2. For each part (components) to determine its surfaces. 2.1.3. Analyze the possibility of any surface directy or indirecty (through) othersurfaces or details) to serve for attaching the assembly support AU taking into accountthe chosen mode of assembly of t. 1.5. If necessary are make any changes in AU, soas to mainitaain its principal function.2.2. Determination of potential surfase of the base AU which will serve to join thesupport Au. To this end: 2.2.1. Determination of details (components) of the base AU. 2.2.2. t.2.1.3.2.3. Refine options (combinations) of the mutual disposition of both AU.2.4. Selection criteeria adopted by one or more options for further design work.2.5. Structural arrangement of each variant of t.2.4. If nesessary, point of view ofoperation, assembly and disassembly, are made corresponding changes in either orboth.Fig.1 Assembly drawing Elektromagnetic Fig. 2 Assembly drawing Indicator light66
  • About the Integration of Assembly Units Fig. 3 Some probabilities inregration assembly unitsExample Let be given AU by fig.1.Elektromagnetic and AU from fig.2 Indicator light.Components and operation of both are AU: el. magnet : 1-coil, 2- restrictive sleeve,3-housing,4–axis, 5- anchor,6- screw. Supplying the coil creates an electromagnetic fieldwhich is concentrated by magnetic 3.The field attracts the armature 3, which is securedto the axle 4 through tight fit. Axis through the hole in its lower end is connected to andmoved undepicted device. When off the coil under the action of gravity axis 4, alongwith five anchor occupy its original position. Indicator Lamp: 1 - building, 2 - contactplate, 3 - disc insulator (2 pcs.), 4 - washers (2 pcs.) 5 - base 6 - screw, 7 - screw 8 -lamp. One end of the lamp 8 is fed through a conductor, whose tip is placed betweenthe washers 4 and housing 1 and the second end of the lamp by-wire connected to thebase 5 ,the screw 7 and the contact plate 2. Let the indicator light to signal openingaction of the electromagnet (attracted anchor).1.1. Requirements.(el.magnet) – the space around the bottom of the axle must be free of attachmenttransfer device;- Air gap between the armature 5 and housing 3 must be free to move the anchor;- Details over 6 and 2 should be open about their position when moving the anchor;- Space in the prismatic part of det.3 must be free to join the environmental technical 67
  • Leparov M., Yancheva M.environment;(Indicator lamp) - 8 lamp must be visible to a person standing on one side of theelectromagnet; - Is there access to the locations of the wires connecting them;1.2 Require general links:- Electric wires, power lamp must be connected with the power of the coil;- The relationship between the two AU to be fixed.1.3 Electromagnets fig.1 is a massive part, and then it will be the base unit, theindicator lamp will perform an secondary role.1.4. The classification of compounds between the two are in first approximation is:1. Detachable1.1. By assembly 1.2. By bolting the two parts concatenated1.3. Through bolts (screws,tenon) compound 1.4. By elastic element 1.5. By pinconnector 1.6.By bayonet.2. Permanent: 2.1. By soldering, 2.2. By gluing, 2.3. By welding,2.4. By riveting, 2.5.By roll, 2.6 By pressed detail.Note: Each assembly component can participate actively in those compounds in whichits a constituent of the compound or passive, in which its pressed by the othercomponents, some of which are active.1.5. Let shortlisting includes the following modes of entry: 1.1-1.4.2.1.1. Detailsof the indicator light is given at the beginning of examples.2.1.2. The surfaces of each piece is visible from fig.1 and fig.2.2.1.1. Details of the indicator light is given at the beginning of examples.2.1.2. The surfaces of each piece is visible from fig. 1 and fig. 2.2.1.3. Some possibilities are: (Det.1) a) in the upper forehead is pierced with a threaded hole axis parallel to the axisof the AU, the lamp is pressed against the electromagnet by a screw that is screwedinto threaded holes (in a spur is added) b) same but on the lower forehead c) bythreaded (assembly) at the top of the outer cylindrical surface, d) by thread (assembly)at the bottom of the outer cylindrical surface (the same should be extended), e) lateralcylindrical surface is drilled threaded hole perpendicular to the axis of the AU, theattachment will be made by screw, which is wrapped in 1;(Det.2) a) the plate can be brought out through a crack in 1, thus after el.isolation canserve as an hanger for attachment of the AU, for this purpose can be used as a contactas the lower and top side,b) media from the base is pressed between parts 2and 3;hanger reaches the plase of fastening through a crack in det.1.(Det.3) a) puck can be extended outward like a case det.2 b) hanger from the base AUis pressed between parts 3 and 5 or between det. 1 and 3;(Det.4)a) as b) det.3(Det.5) a) open the plate (base) mounts the AU, eg. with screw connection, as thecontact surface between both can serve as upper and lower surfaces b) bending of thetwo horizontal parts down will change the location of the connecting screw connectiond) under the condition of b) can be used connection type assembly or hinge ( the plateitself acts as a shaft or hole);(Det.6) a) hanger of the base is pressed between the parts 6 and 4 or between det.1and 4 b) under the head of det.6 put elastic piece, which is implemented by mountingon the base;(Det.7) a) det.7 associated whit threaded hole in the base;(Det.8) a) thread the lamp 8 can be used to create a threaded coupling for attachment68
  • About the Integration of Assembly Unitsto the auxiliary base AU.Note: In some cases, the main 5 becomes unnecessary.2.2.1.Details of the electromagnet are given at the beginning of examples.2.2.2. Some possibilities are:(Det.1) a) through the slit in the hanger det.3 support AU webs pass over the spool ofthe coil (or slit in it) and bend down (groove in the spool), Option: Add this to thehanger electromagnet, and it is attached to the lamp;(Det.2) a) joint between 2 and hole in an support AU b) by bolting (added ext intodet.2);(Det.3) a) in the upper forehead is drill two screw holes with axes parallel to the axis ofthe AU, the lamp is pressed against the electromagnet by a screw that is screwed intothreaded holes (Different placements of the holes), b) by concentric to the hub 2 moreholes or a bulge, as is done by joining thread (assembly); c) on the lateral cylindricalsurface is pierced with a threaded hole perpendicular to the axis of the AU, theattachment will be made by screw (screws), which wrapped in det.3 d) by using openaccession of an electromagnet to environmental protection technical e) with additionalopenings on the upper face of the prismatic section (the same may be extended), f)with additional openings on the lateral surface of the prism division g) using the twofronts of the prismatic section for attachment of auxiliary through elastic element type"pinch" (base) or clip is fixed by an additional screw connection;(Det.4) a) by a pin axis that is perpendicular to the axis of the base is attachedsupported AU with threaded or assembly;(Det.5) a) in the cone surface is pierced threaded hole that mounts the AU, providedthat it does not interfere with the interaction of the electromagnet with environmentaltechnical environment.(Det.6) a) hanger of the support AU pressed between parts 2 and 6 or between det.2and 3 b) under the head of det.1 put elastic piece, which is implemented by theattachment of the supporting AU;The total number of combinations K of attachment of both AU is a.v = K = 18.14 = 252,where a and c is the number of possible anchorages for the AU and accordingly thebase AU. For some combination require additional changes accordingly, and anycombinations there are variations of structural conversions.Some of these combinations are given in Fig.3, where by a long dashed double- dottednarrow line shows the silhouette of the support AU.3. CONCLUSION In this work the method “Surface of details” has been reviewed. He has beenpresented and illustrated the algorithm by example. The method allows a systematicsearch of options for integration of assembly units. It could be used as a "manual"design, and as a basis for developing a formalized system desing of integration ofproducts.LITERATURE[1] Pahl G., W. Beitz. Engineering Design. A Systematic Approach. Springer - Verlag, Berlin, 2001. 69
  • Leparov M., Yancheva M.[2]. Otto K., K. Wood Product Design. Techniques in Engineering and New Product Development, NJ, Prentice,Hall, 2001.[3]. Polovinkin A.I. Fundamentals of Engineering Design, 3-e Edition, St. Petersburg, ed.Lany,2007.[4]. Leparov, M., M. Vicheva, M. Georgiev, Fundamentals of Engineering Design, Softtreyd, S., 2008.70
  • ANALYSIS OF INFLUENCE OF LENGTH OF DEVELOPMENT OF BOUNDARY LAYER ON FLOW RATE THROUGH RADIAL CLEARANCE WITHIN HYDRAULIC CONTROL COMPONENTS 1 2 3 Darko Kneževi , Aleksandar Milašinovi , Zdravko MilovanoviSummary: In hydraulic control components (directional control valves, servo valves,LS regulators, etc.) there is a problem of determining of flow rate through the radialclearances of different lengths overlap between the control piston and body ofcomponent. As lengths of overlap are relatively small, it is necessary to analyze thedevelopment of a full profile of fluid velocity through radial clearance (in sleep modeservo valve has length of overlap equal zero; directional control valves, depending onthe diameter, have lengths of overlaps of 0.6 to 0.8 mm; while in work of the LSregulator, piston is continuously located in areas of small opening, zero overlap andsmall length of overlap). Common values of radial clearance is approximate in therange from 1 m (minimum values for servo valves) to 50 m (maximum value for axialpiston pumps). [1]In this paper is analysed development of full-length profile of velocity inside radialclearance (i.e. development of boundary layer).Key words: hydraulic components, radial clearance, length of overlap, boundary layer1. INTRODUCTION Complexity of mathematical description of the flow of hydraulic fluid through theradial clearances stems from the fact that the mathematical description can not be thesame for flow of hydraulic fluid through the clearances at zero overlap between thepiston and cylinder, the small lengths of overlap and the long lengths of overlap. Flow of the hydraulic fluid through orifice at zero overlap (cylindrical ring),depending on pressure, temperature, fluid type and size of the clearance, can belaminar or turbulent. Due to small size of clearances in hydraulic components, flow ofhydraulic fluid through orifice at zero overlap is mostly in the laminar or transition zone.The discharge coefficient of flow through the orifice in the laminar flow (or transitionfrom laminar flow to turbulent) is a function of Reynolds number. Determination ofRejnolds number complicates the fact that it depends on the viscosity of fluid. Viscosity1 Assist. Professor, Darko Kneževi , Banja Luka, Faculty of Mechanical Engineering, darkokn@blic.net2 Assist. Professor, Aleksandar Milašinovi , Banja Luka, Faculty of Mechanical Engineering, acom@blic.net3 Assoc. Prof., Zdravko Milovanovi , Banja Luka, Faculty of Mechanical Engineering, mzdravko@urc.rs.ba 71
  • Darko Kneževi , Aleksandar Milašinovi , Zdravko Milovanoviof working fluid is significantly changed because of the sudden transformation ofhydraulic energy into heat (coeficient of dynamic viscosity of working fluid at the exit ofthe orifice can be 5-6 times smaller than the entrance to the orifice). [2] For small length of overlap (ie short path of flow of oil through the radialclearance), flow takes place relatively quickly, so that the generated heat, for the mostpart, is retained in the oil and the process can be treated as adiabatic. Further, thevelocity profile along the clearance is variable and it is necessary to analyze length ofdevelopment of the boundary layer between the piston and cylinder (ie, thedetermination of length needed for fully development of velocity profile along radialclearance). Fig. 1 Flow of hydraulic fluid through radial clearance For longer lengths of overlap, fluid flow is quite slow, so fluid is exchanged heatwith the environment. The change of state of hydraulic fluid in the flow can beconsidered isothermal. Such case flow through the radial clearance is analyzed in theliterature, but with one flaw: it ignores the change of oil viscosity with a change ofpressure. This neglect can be lead to the calculated flow rate is 50% higher than theactual flow rate (depending on value of of working pressure in the hydraulic system ). By determining the length of development of the boundary layer in flow of fluidthrough the radial clearance can be defined criteria for the application of appropriateformulas for determining the flow rate. [2]2. DEVELOPMENT OF BOUNDARY LAYER AT FLOW OF FLUID THROUGH RADIAL CLEARANCE This analysis is based on analysis of development of boundary layer for flowingincompressible fluid with uniform velocity profile at the entrance. The boundary layerhas grown in thickness to comletely fill the clearance. Viscous effects are ofconsiderable within the boundary layer. For fluid outside the boundary layer viscouseffects are negligible. As the radial clearance between the piston and cylinder, cr, small compared tothe radius of the piston, it is acceptable analysis of this flow as the flow between twoparallel plates at a distance cr. The symbol (x) denotes the thickness of the boundary layer on the surface ofthe piston for any x.72
  • Analysis of Influence of Length of Development of Boundary Layer on Flow Rate Through Radial Clearance Within Hydraulic Control Components Fig. 2 Development of boundary layer inside radial clearance Conditions to the lack of velocity gradient at the boundary of the boundary layerand sliding on the surface of the piston, are given folowing expressions uz 0 0, (1) u 0. (2) z z Assuming change of velocity profile by parabolic law 2 u z z 2 , (3) ucand applying the continuity equation 2 cr 2 z z u0 cr 2 uc 2 dz uc dz , (4) 0we obtain u0 uc . (5) 2 1 3cr Karman has developed an approximate method of solving the boundary layer,which is based on the application of momentum equations in integral form, applied toan element of boundary layer, restricted areas 1, 2, 3 and 4 (Figure 3). [3]Isolated element of the boundary layer in Figure 2 is shown in Figure 3. 73
  • Darko Kneževi , Aleksandar Milašinovi , Zdravko Milovanovi Fig. 3 Isolated element of boundary layer For steady two-dimensional flow, neglecting the mass forces and applying abasic property of the boundary layer p 0, zx component of momentum equation, applied to the CV, is x d dp d p p dx p dx dx w dx u u dz dx dx dx 0 x x (6) d uc uc n ds u 2 dz u 2 dzdx 0. 2 0 dx 0 By applying the continuity equation for incompressible fluid flow v n dS 0, (7) KPon the element of boundary layer of unit width dS ds 1 or dS dz 1 , the continuityequation (7), applied to the control volume CV, becomes dQ Q x u c n ds Q x dx 0 , (8) 2 dxwhere:74
  • Analysis of Influence of Length of Development of Boundary Layer on Flow Rate Through Radial Clearance Within Hydraulic Control Components x x Q x v n dz udz . (9) 0 0 We combine eguations (8) i (9) to get x dQ d uc n ds dx udz . (10) 2 dx dx 0 If for two-dimensional steady flow, velocity and pressure are change in the law uc uc x i p p x ,applying Eulers equations of motion, gives duc 1 dp uc . (11) dx dx Integral over the surface 2 is x dQ d uc uc n ds uc dx uc udzdx . (12) 2 dx dx 0 By applying Newtons law of viscosity, shear stress on the surface of the pistoncan be written as du w . (13) dz z 0 By incorporating equations (10), (11), (12) and (13) into equation (6) anddividing by dx , is obtained x x x du d d du uc c dz uc udz u 2 dz . (14) dx 0 dx 0 dx 0 dz z 0 By integrating equation (14), we get 20 cr . (15) x 3Re x Here is 2u0 cr 2Q 2Q Re , (16) d wReynolds number. By putting 75
  • Darko Kneževi , Aleksandar Milašinovi , Zdravko Milovanovi cr , 2we get the length of the full development of the boundary layer, xtl, xtl 0.0375 Re . (17) cr From equation (17) can be seen that for very low Reynolds number Re 30 ,which is often the case in the hydraulic control valves, the length of the development ofthe boundary layer is less than the size of radial clearance cr.3. CONCLUSION It is obvious that in determining the flow rate through the radial clearance withinthe hydraulic component must be used various mathematical expressions in thefunction of the length of the overlap. The criterion for the application of the appropriateexpressions (for zero overlap, short or long overlap) can not be determined withoutexperimental investigations. The reason for this is that the energy of hidraulic fluid flowing through radialclearance is transformed into heat energy which can lead to increase of fluidtemperature. As working fluids in hydraulic systems are commonly used mineral oils,which significantly changes the viscosity with changes of temperature and pressure, itis the determination of flow rate of fluid through the radial clearance within thehydraulic components is very complex. While the mathematical description of fluid flow through long clearances isrelatively simple, real difficulty is a mathematical description of the flow of hydraulicfluid through the radial clearances of small length of overlap (in the limiting case atzero overlap), which are of great importance in hydraulic control components. For shortlengths of overlap can not be ignored impact of changes of viscosity of hydraulic fluidwith changing temperatures, and must be taken into account the influence ofdevelopment of full profile of velocity within the radial clearance. It can be concluded that the formula for determining the flow through the radialclearance at short length of overlap had to be covered: clearance geometry, operatingconditions (pressure and temperature of fluid), the type and physical properties of fluidand changes of these properties with change of temperature and pressure. Expression(17) can be used as orientation for defining criteria for dividing radial clearances onshort and long, and applying the appropriate formula for calculating of flow rate.LITERATURE[1] Savi V., Zirojevi LJ. (2003). Uljna hidraulika 3. IKOS, Novi Sad.[2] Kneževic D. (2007). Influence Clearances Geometry in Hydraulic Components of Automation Control on Efficiency of Hydraulic Systems, PhD Thesis, Novi Sad.[3] Munson B.R., Young D.F., Okiishi T.H. (2002). Fundamentals of Fluid Mechanics, John Wiley & Sons, New York76
  • ANALYTICAL-NUMERICAL STRESS ANALYSIS OF SPUR GEARS WITH STRAIGHT TEETH 3 Nebojša Radi 1, Goran Sekuli 2, Dejan JeremiSummary: The main task of the study is to determine the stress state of spur gearswhere the teeth run straight across the wheel, using analytical-numerical analysis. Forthe purpose of stress analysis of spur gears, two methods are used: analytical method,performed in accordance with ISO 6336 standards and numerical method which isbased on ANSYS v12 software package.Of these two analysis, only the numerical one is used in two cases, that is, whenfriction between two gears as their teeth mesh together, appears even when it isneglected. After this, derived results are compared.Key words: stress in the flank of the tooth, stress in the root of the tooth, numericalanalysis, analytical analysisINTRODUCTION The first step in the procedure is the stress state determination in the flank ofthe tooth and in the root of the tooth of gears. This is one of the main tasks and mostimportant things when checking calculation of load capacity. For the purpose of thistask, a pair of the same spur gears is used. Stress state of these particular gears isdefined through analytical and numerical analysis. Analytical method is proposed bythe standard ISO 6336 while the numerical one is based on ANSYS v12 softwarepackage. The paper briefly describes the process of developing the module, simulationand analysis of conditions. Derived results are compared and analyzed.1. DETERMINATION OF THE STRESS STATE USING ANALYTICAL ANALYSIS In order perform the analytical analysis following data are required: the moduleofa gear m=3 mm, number teeth z=20, pitch diameter d1=60 mm, vertex of a circlediameter da=66 mm, shaft opening B=15 mm, gear width b=30 mm, roating momentT=9.954 Nm, dynamic durability in the flank of the tooth Hlim=480 N/mm2, and dynamicdurability in the rooth of a tooth Flim=262 N/mm2.1 Doc. dr Nebojša Radi , Isto no Sarajevo, Mašinski fakultet I. Sarajevo, (nesor67@yahoo.com)2 Dipl. inž. maš, Goran Sekuli , Višegrad, (sekulic.goran@gmail.com)3 Dipl. Inž. maš, Dejan Jeremi , Isto no Sarajvo, Mašinski fakultet I. Sarajevo, (dejan.jeremic@yahoo.com) 77
  • Nebojša Radi , Goran Sekuli , Dejan JeremiOperating stress in the flank of the tooth Durability of tooth flank implies their resistance to pitting formation. Consideringthat pitting is primarily a consequence of material fatigue due to high surfacepressures, the operating stress is expressed by following formulations: Ft u 1 H ZH ZE Z Z ZB K A KV K H KH d1 b u (1)Where is: Ft - circumferential force d 1 - pich diameter u – kinematic transfer capacity b – gear width K A is the working condition factor and includes increasing of load due to variations ofinput and output moments. K V is a dynamic factor, keeping in mind load increase dueto internal dynamic effects. K H is load distribution factor in the flank of tooth. K H isthe influence of unequal load distribution on the pairs of gears in a constant meshDrive gear circumferential force is expressed by the next formula: T Ft r1 (2) 9.954 Ft 331.8 N 0.03 In this case, load factors are: K A 1 , K V 1 , K H 1.23 , K H 1.45 ,KF 1.37 , K F 1.304 . Other important factors that fulfill above requirements for a proper opration ofoperating stress of a gear, are: shape factor of a gear Z H 2.49 , material flexibility Nfactor Z E 189.8 ( ) , single mesh factor Z B 1 , constrain angle factor Z 0. 9 , mm 2conjugate gear factor Z 1. Inserting the previously calculated values into the operating stress in the flankof tooth expression, we get next results: 331.8 1 1 N H 2.49 189.8 0.9 1 1 1 1 1.23 1.45 344.9 60 30 1 mm 2Operating stress in the root of the tooth According to the ISO standards, as the working stress in the root of the tooth,bending stress S is used. Therein, we neglect two things: tangential stress whichappears in a consequence of shearing action and normal stress due to pressure.Adequate correction factors were introduced in the calculation. Taking intoconsideration load factors too, the expression of the working stress becomes:78
  • Analytical-Numerical Stress Analysis of Spur Gears with Straight Teeth Ft F Y Fa YSa Y Y K A KV K F KF b m mn (3)Thus, shape factor of the tooth flank is Y Fa 2.86 , stress concentration Y Sa 1.56 ,constrain angle factor Y 0.73 and chamfered teeth factor Y 1.Putting all these parameters in the formulation of the operating stress in the root of thetooth, we get: 331.8 N F 2.86 1.56 0.73 1 1 1 1.37 1.304 21.45 30 3 mm 22. DETERMINATION OF THE STRESS STATE USING NUMERICAL ANALYSIS A numerical method is carried out through ANSYS v12 software package. It isperformed for two cases: in the first case, when there is no friction between meshedgears and in the second case when the friction coefficient is taken into account. Inorder to obtain the results of the stress state using numerical analysis, it was bothnecessary to create a module of a gear and modify it into gear set, and then to definethe conditions of analysis. The model is formed by extrusion from geometry. Evolute of a curve of toothprofile is got through involute function, calculating coordinates of the points used for itsdrawing in ANSYS v12 software package. After that, drawing of a vertex circle, rootcircle and shaft opening is done. Removing redundant lines, we formed geometry as inFigure 1. Fig. 1 Geometry and model of a gear Fig 2 shows a pair of gears derived from touching meshed gears only at onepoint in no-load. 79
  • Nebojša Radi , Goran Sekuli , Dejan Jeremi Fig. 2 A pair of gears Defining terms that simulate normal conditions and limitations, we began bydefining a contact region between the flank bearing surfaces of tooth of gears.Furthermore, we selected a type of contact which was the contact without friction(choosing frictionless option). The following step is the formation of finite elementmeshes. Optimal mesh is formed by choosing a typical region with coupling of gearsand decreasing size of final element. In this way, we reduced the time of networkforming and with comminuting of mesh in the mentioned region, we increased theaccuracy. Fig. 3 Optimal mesh of final elements We complete defining condition analysis introducing two things: the load, thatis, torque of 9954 Nmm in the upper (drive) gear and a fixed support at the top smoothsupport at the root fo a gear. After that, derived results are graphically presented byfigures 4 and 5. Figure 4 illustrates the values of the stress state when friction betweenmeshed gears is neglected, while Figure 5 shows the stress state between meshedgears where friction is defined through type of contact along with friction coefficient 0.15 .80
  • Analytical-Numerical Stress Analysis of Spur Gears with Straight Teeth Fig. 4 The result of the stress state when friction is neglected Fig. 5 The result of the stress state with friction at the contact3. COMPARISON OF THE RESULTS DERIVED THROUGH ANALYTICAL AND NUMERICAL ANALYSIS The results of the stress state derived, in one case, through analytical methodand in two cases, using numerical method, are given in the following diagrams.Diagram a) illustrates the comparative results from which we can conclude that valuesof the stress in the flank of tooth in all three cases have minor variations. Maximumoperating stress in the root of the tooth in both cases of numerical method has minorderivations, while the analytical analysis calculated value is much lower. Radni napon na boku 500 500,00 zupca 450 450,00 Radni napon u podnožju 400 400,00 350 350,00 Radni napon na 300,00 300 boku zupca 250 250,00 Radni napon u 200,00 200 podnožju 150 150,00 100 100,00 50 50,00 0 0,00 Numeri ka Numeri ka Analiti ka Numeri ka Numeri ka Analiti ka metoda metoda (sa metoda metoda (bez metoda (sa metoda (bez trenja) trenjem) trenja) trenjem) a) b) Fig. 6 Comparing results 81
  • Nebojša Radi , Goran Sekuli , Dejan Jeremi The reasons for the higher values of the stress state in the root of the toothusing analytical method may be due to stress concentration of the radius. Accordinganalytical recommendations, it should be taken into account that the gear is treated asa cantilever and that the stress distribution is linear. For the results derived bynumerical method, we cannot say that we have a linear distribution as in the case ofbending, where the root of the tooth is loaded according ISO standards. If we read theresults of numerical analysis from the areas without stress concentration, than theimage of the compared results is as in the diagram b). From these results, we are ableto say that deviation for analytical method is less, compared to numerical one.4. CONCLUSION Determination of stress state of spur gears with straight teeth is made in orderto compare the results derived using numerical analysis with the results from analyticalanalysis based on ISO standards. Analyzing the results, we can conclude that stressstates in the flank of tooth both for numerical and analytical method, are approximate.Also, we deduce that in relation to the position of maximum stress in the root of thetooth, critical section of analytical method does not correspond with the position ofmaximum stress gained by numerical method. Numerical analysis shows that thedeveloped model can be used in a simple way for different cases of analysis. It allowsus to change the parameters of condition analysis, as we have in this case altered thetype of contact of meshed gears.LITERATURE[1] Miltenovi , V.: Mašinski elementi. Mašinski fakultet, Niš, 2006.[2] Miltenovi , V.: Mašinski elementi – tablice i dijagrami. Mašinski fakultet, Niš, 2006.[3] Lee, H.: Finte Element Simulations with ANSYS Workbench 12.Schroff Developments Corporation, 2010.[4] *** www.khkgears.com[5] *** www.ansys.com82
  • CAD DESIGN OF FLEXIBLE FRICTION COUPLING Georgy Dinev1, Marieta Yancheva2Summary: In this work are presented approach for optimal design and constructivedocumentation of flexible-friction coupling in CAD medium. On base of advance designcalculation are prepared constructive documentation of coupling.The first design process step is creating 2D geometrical modeling of assembly unit. Onbase of specify functional parameters and dimensional chains are create 3Dgeometrical model of assembly unit. Finally the design mechanical products arechecked by animation model for collision between component parts. The result can beuse in education the students “Bachelor and Master degree” in science field“Mechanical Engineering”.Key words: CAD, design, flexibly, function, coupling, dimensional, analysis, education,control, documentation, modeling.1 INTRODUCTION Dimensional analysis is important activity in the process of designing productsand development of design documents. So far considered some approaches to the design of mechanical products interms of dimensional-precision analysis [2],[6],[7]. Modern systems CAD/CAM/CAE enables to improve the quality in thedevelopment of design documentation [8]. As a result, in promote precision byautomated assembly of mechanical product. Furthermore, the software allows toinspected, by simulation of the created geometric patters and demand for theircollisions in their assembly [5]. Despite the great possibilities of software in use, thecredit lies in the knowledge and skills of the working team. In this case the basic stones are following: Dimensional-structure analysis of a mechanical product-elastic frictional coupling from a methodological viewpoint; Simulation analysis of a mechanical product and check the parametric models of the component parts.1 Assoc. Prof. PhD Georgi Dimitrov Dinev, Bulgaria,1000 Sofia, Bvd. Kliment Ochridski-8, TechnicalUniversity, Faculty of mechanical Engineering, e-mail: gdinev@tu-sofia.bg2 Assistant Prof. PhD eng. Marieta Konstantinova Yancheva-Popova, Bulgaria,1000 Sofia, Bvd. KlimentOchridski-8, Technical University, Faculty of mechanical Engineering, e-mail: mmarieta@mail.bg 83
  • Dinev G., Yancheva M. These stages are applicable mainly for the purpose of designing course onPrinciples of Mechanical Design. The main tasks are solved in this work are: To create a geometric model of elastic frictional coupling; To make dimensional structural analysis of choose variant constructive decision; Analyzing the results for optimization and of axial parameters-clearances, in the design of mechanical product; To propose an algorithm to developed constructive documentation of assembly unit.2 INVESTIGATION, RESULTS AND DISCUSSION2.1 Geometrical modeling of elastic friction coupling-mechanical product On base the specified output data base calculations is developed 3Dgeometrical model of the elastic coupling through the graphical system Solid Works,given in Fig.1. [7] Fig. 1 3D geometrical model of the elastic coupling2.2 Dimensional structual analysis of the assembly unit For the functional dimensioning mehanical product of design through modernCAD/CAM/CAE systems are based the following algorithm: output data;84
  • CAD design of flexible friction coupling Modeling of component parts; 3D/2D models; Construction of the graph model; Calculation of axial chains; On the basie of the developed 3D geometrical model of the assembly unit to study the collision between the relevant it parts; From an inspection of 3D geometrical models of component parts to make corrections of their sizes by removal of material. Fig. 2 2D drawing with coded contact surfaces On base of the developed constructive solution are choosed a system of twodimensionale chains. Values of the axial parameters are selected on basestandardized catalog information. The functional elements of the assembly unit aregived in Fig. 3. 85
  • Dinev G., Yancheva M. Fig. 3 Scheme dimensional chains2.3 Analysis of results Since the elastic friction coupling has a flexible element, assembly parameterP1f, and is selected by catalog material are change from 0,5 to 5 mm during operationprocess due to deviation of the shaft [3]. From the examination of the 3D geometricmodel despite the changing parametres, there was not collision and a strong influenceon the value of the axial parameter P2f. 3D geometrical model can be use for simulation analysis and verification toparametric models of the component parts. That the, students will get knowledge andskills for use creative design methods of mechanical products. Also motivation to worktogether in particular and especially to analyze and discussion analytical solutions [4].1. CONCLUSION The assembly unit of mechanical product – flexibly coupling in CAD medium isdeveloped. On the basis of 2D geometrical model is made a dimensional analysis ofthe product. A methods for constructive documentation of mechanical product isproposed. 3D geometrical model of this assembly unit can be investigated by graphicalsystem NX-CAD/CAM The results can be used in design process on “Principles ofmechanical design”.LITERATURE[1] Dinev G.D., N. Gizdov (2009) About teaching in fundamentals construction and CAD – Bulgarian journal of Engineering design, Vol. 2, p.74-79.[2] Dinev G.D, (2009) Fundamentals of constructive documentation of gear transmission with cylindrical wheels, SOFTTREJD, Sofia, p.190, (in Bulgarian).[3] Dinev G.D. (2011) Course on Principles of Mechanical Design, AVANGARD PRIMA, p. 98.[4] Dinev G., Z. Vitlarov (2011) Handbook for laboratory exercises on Principles of Mechanical Design, AVANGARD PRIMA, p. 40.[5] Dinev G.D. (2011) An approach for simulation design of mechanical product, 6-th International conference “ OPTIROB” 2011, Sinaja, Romania, May 2011 (printed).[6] Sandalski B. P., P. V. Goranov, G. D. Dinev, I. Z. Nikolova (2008) Fundamentals of design and CAD, SOFTTREJD, Sofia, p.338 (in Bulgarian).[7] Goranov P.V, G.D. Dinev, V.M. Stancheva, M.K. Jancheva (2005) Acseptual model of internet oriented system for tolerance analyses of mechanical assemblies. Proc. of Inter. Conference “Challenges on higher Education”, Sozopol, vol.3, p.64-67.[8] User guide on Solidworks (2010).86
  • COMPARATIVE ANALYSIS OF THE FORMATION OF SMALL GRAIN GUIDANCE Dragan Lišanin1, Marinko Petrovi 2, Nenad Grujovi 3, Jelena Borota4Summary: Classical theory of internal ballistics for small-arms represents a very roughapproximation. The assumption is that interior profile of the barrel of the small caliberscan be defined on the basis of theoretical and practical tests primarily (experiment) fora family of weapons, i.e. all weapons of small caliber (including 12.7 mm). The aim isto determine the feasibility of standardization of one or the other barrel profile (12" and7"), based on the advantages of one over the other profile, verification of the modifiedinternal ballistics model and the confirmation of such a hypotheses.Key words: small grain, ballistics, small caliber, weapon1. INTRODUCTION The barrel is an essential part of weapons. It represents the part in which thereis twofold energy transformation: chemical energy of combustion of gunpowder isconverted into heat and the heat into mechanical energy. The barrel provides thenecessary initial projectile speed and direction of the boom, and the projectiles thathave no wings to stabilize them, through the helical grooves, it provides high rotationneeded for stability when the projectile is moving towards its aim. The classical theory based on geometrical law of combustion and adiabaticprocesses in small arms barrel, gives a fairly satisfactory solutions for large-caliberweapons, because the assumptions on which these methods are based do not differfrom the actual firing processes. There are several two-phase models developed in thelast decade that represent a significant step forward in studying the process of firing insmall-arms [1]. To solve the internal ballistics problem the most realistic strategy is togo from the works by P.S. Gough and that the solution to the model should be soughtby using the finite differences model [2], [3]. Fundamentals of numerical treatment PDJ were given by Richtmayer andMorton, Roach, Patankar and Smith, and for the numerical solution of two-phase flowin small arms we used three methods: MacCormack, Lax-Wendroff and Spalding. PSGough [3] and others, and Gokhale and Krier used MacCormack two-stage, explicit1 dipl.maš.inž. Dragan Lišanin, Kragujevac, Mašinski fakultet (lisanin@sbb.rs)2 prof.dr Marinko Petrovi , Kragujevac, Zastava namenski proizvodi3 prof.dr Nenad Grujovi , Kragujevac, Mašinski fakultet (gruja@kg.ac.rs)4 dipl.maš.inž. Jelena Borota, Kragujevac, šinski fakultet (jborota@gmail.com) 87
  • Dragan Lišanin, Marinko Petrovi , Nenad Grujovi , Jelena Borotamethod with finite differences. Later for numerical modelling of flame propagationthrough the granular environment Spalding implicit method is used [4]. The entire survey was realized as a contribution to the study of flamepropagation through a granular environment [4]. The experimental determination,calculation of pressures at firing weapons, and nonstationary gas dynamics is appliedto the internal ballistics problem of small-arm weapons. In the US armament the 5.56 mm machine gun M16 with a bullet M193 wasintroduced, but other members of NATO did not accept it. Parallel study of several typesof bullets, adopt the same bullet for the all members. For the survey Belgium prepared abullet under the designation SS109, which uses the same cartridges and has the samedimensions as the U.S. M193, but the projectile has significantly enhanced features(which required pitching twisting grooves of the barrel from 12" to 7"). Inside profile of the barrel for small-arms can be defined in terms of theoreticaland above all practical tests [experiment] to a family of weapons that meet thefollowing requirements: safety of use and reliability and simplicity of construction andmaintenance. For analysis was taken the internal profile of the barrel of the small firingarms. As the basis was taken 5.56 mm caliber, and discussed it in two variants ofexisting models internal ballistics two-phase flow [5].2. COMPARATIVE ANALYSIS Belgium conducted a parallel investigation with three types of bullets (7.62 mmx 51 SS77; 5.56 x45 mm M193 and 5.56 mm x 45 TW7" SS109, valid gunpowderL110, P112 armour). Advantage of the system 5.56 mm was showed and confirmedthe accuracy in operational steps in twisting grooves of grain SS109 and M193.Thereby, they set the initial velocity of the projectile SS109, 4g weight V0 = 940 m/s ofthe tube length of 508 mm and V0 = 925 m/s from the barrel length 465 mm. Swedenalso conducted tests of ballistic projectile stability of 5.56 mm projectiles fired fromweapons of different length and steps twisting grooves (guns: U.S. M16A1 - 12", theBelgian FNC - 7" and the Swedish FFV 890C - 9"). Based on these findings, for the SS109 projectile the step folding grooves 7"(177.8 mm) is confirmed. Because the bullet 5.56 mm SS109 has the same capsule asthe 5.56 mm M193 bullet (SS92), two basic parameters for design of internal alignmenttube can be defined: a step twist and propellant chambers. The increased velocity of the projectile is a function of increased pressurepropellant gases p, which in cal. 5.56 mm exceeds the value of 3600 bar. It also meansthat it increases the specific pressure on the working side gutters of the barrel, i.e.pressure force of projectile shell is also greater. When moving through the channel of the barrel, the projectile receives a largeradial velocity along the axial rotation axis. The shell is faced with significant centrifugalforces of inertia that are transmitted to the surface of the guidance since the projectileis in contact with it. At the exit from the barrel, the centrifugal force cause shear stressin the shell and that may even harm the strength of shell and proper route of theprojectile. Between the shell and core there is usually soft lead lining - layer that, liningthe engraved projectile and gutters protect the channel of barrel from intense damage. The most frequently used method of increasing ballistic lifetime of the barrel ischrome. The reason for this is an excellent reliability of this production in terms of the88
  • Comparative analysis of the formation of small grain guidanceachieved level of the barrel performance compared to other technologies. Classically applied cutting technology did not give many opportunities tochange a rectangular profile of the rifling and its substitute. Rifling forging is a relativelynew method, which is currently brought to a satisfactory level of quality. Its use, unlikeother technologies has two main advantages: projectile deposits are made assimultaneous continuation of rifling operation and with this technology a more flexiblevariation of rifling profile is possible. Fig. 1 Shapes of the grooves The shape of the cross-section or the profile of the projectile guidance was fora long time considered to have very significant effect on weapons ballistics, Fig 1.Studies of the practical results have shown that it has the largest influence on the wearand lifetime of the barrel. For selection of a profile, the valid criteria is determined todefine the restrictions such as: a minimum distribution of hits on the target, themaximum lifetime of barrel and a minimum of cleaning (the low-maintenance) andminimum cost pipe. When shooting caliber weapons below 12.7 mm, there are 4 or 6grooves. Adoption of the four-grooves for the profile of the barrel 5.56 mm SS109 isaccompanied by the satisfaction of demands that the specific pressure on the workingside of the guidance in this new profile does not exceed the value that occurs in theprofile with 6 grooves. Three profiles were tested: profile 1 (with 6 grooves), profile 2and profile 3 (with 4 grooves), for the purpose of comparative analysis and the generalresults, Tab. 1.Table 1 Profile dimensions Profile 1 Profile 2 Profile 3 Calibar d [mm] 5.50 – 5.59 5.44 – 5.53 5.38 – 5.47 Groove do [mm] 5.63 – 5.72 5.64 – 5.73 5.61 – 5.70 Change interval d and do [mm] 0.09 0.09 0.09 Twofold depth of the 0.10 – 0.16 0.17 – 0.23 0.20 – 0.26 groove 2 1 [mm] Surface Sc [mm2] 24.51 – 25.31 24.36 – 25.16 24.23 – 25.03 89
  • Dragan Lišanin, Marinko Petrovi , Nenad Grujovi , Jelena Borota3. MATHEMATICAL MODEL FOR THE TWO-PHASE FLOW WITH QUASISTATIC ENGRAVING THEORY The two-phase flow model takes the pressure of forcing p0 as developedpropellant gas pressure when the projectile was launched and the engraving happend. The main drawback of this model is the empirical defining the pressure offorcing, and then the Drozd method defines the initial values of parameters and theduration of combustion till the moment of initiation of the projectile. Basic assumptions of the model would be implemented as follows: Launching of the projectile at the pressure generated in the cartridge for overcoming a strong link between the projectile and cartridge (plucking the projectile force). The projectile travels to the contact of front grooved barrels and transition cone. Speed that the projectile reaches on the road is relatively small and does not affect the process of quasistatic engravement. Engravement is a time process and is not instantaneous, but a gradual transition in which the projectile passes the transition cone with its length. The force between the shell and surface of the transition cone remains constant and is provided through the yield strength of material under static conditions. Engraving force increases with the moving of the projectile, caused by increasing of the contact surface. The system of equations of two-phase flow at combustion of gunpowder can berepresented by partial differential equations (PDE) in matrix form: Y Y As b (1) t s matrix of partial derivatives depending on the time t and generalised coordinate s T T Y u ub e Y u ub e and (2) t t t t t t s s s s s s matrix 5x5 of the coefficients with the partial derivatives from dependent variables by generalised coordinate s 5,5 As akl k 1,l 1 (3) matrix of the right of the equation 5 b bk k 1 (4) or the general form Y m Y k (akl l) b (5) k t l 1 s where, u - gunpowder gas velocity, ub - gunpowder grain velocity, - porosity,as a measure of space that can be fulfil by gunpowder gases, - density ofgunpowder gases, e - internal energy per unit volume of gunpowder gases; m 590
  • Comparative analysis of the formation of small grain guidanceduring the combustion of gunpowder, and m 3 after the combustion, and some ofthe coefficients akl are equal to 0. In addition to these factors, the mathematical model also depends ongunpowder pressure, velocity of grain combustion, mass of grain, parameter of surfaceforces, energy per unit volume used for grain heating etc. Quasistatic engravement model [7], was developed for artillery and applied to asmall-arms, does not take into account the change in cross section due to groovedbarrels [2]. This simplification is possible for the artillery, where the diameter of therotating band is larger than the diameter of grooves, which is very often not the casewith the projectile of firing weapons.4. RESULTS AND PRECISION Real experiment was carried out with 59 samples of barrels cal. 5.56 mm TW7[6]. The aim was to make atmospheric conditions [temperature, humidity, atmosphericpressure ...) and factors related to the techniques involved in experiment (assembly,handling, loading, firing ...) as similar as possible for all samples. The average time ofmaximum pressure values were measured at four measuring points by piezoelectrictransducers (p1msr, p2msr, p3msr, p4msr). The average values with standard deviation aregiven in Table 2.Table 2 Statistic analysis of pressure Profile p1msr st. dev. p2msr st. dev. p3msr st. dev. p4msr st. dev. Profile 1 3502 100 2608 110 1042 68 700 22 Profile 2 3470 322 2403 534 1075 80 708 18 Profile 3 3628 150 2748 128 1088 42 705 21 Projectile velocity was measured at 10 m from the mouth tube. The averagevalues of calculated and measured initial projectiles velocities (Vocal, Vomsr) reduced tothe mouth tube are given Table 3.Table 3 Statistic analysis of initial velocities Profile Vocal Vomsr st.dev. Profile 1 863 873.2 11.0 Profile 2 865 867.2 11.6 Profile 3 868 868.2 7.3 It is obvious that experimental results deviate very little from the calculated,only by +1.2% for Profile 1, +0.27% for profile 2 and +0.02% for the third profile Thismeans complete correspondence of theoretical and experimental models. The averagevalues of Rs and D with the profiles are given in table 3, with measurements incentimetres. As with velocities, statistic analysis shows that the accuracy obeys theGaussian distribution with standard deviations, and that the results of the measured 91
  • Dragan Lišanin, Marinko Petrovi , Nenad Grujovi , Jelena Borotainitial velocities correspond to the results of precision.Table 4 Statistical analysis of precision Profile D40 RS40 st.dev. Profile 1 4.18 1.16 0.28 Profile 2 3.81 1.03 0.14 Profile 3 3.62 1.01 0.165. CONCLUSION The most stable results of the initial velocity and precision provides the profile 3with the lowest standard deviation of 7.3 m / s and the smallest average radial drift ofthe hits RS40 = 1.01 cm, and diameter of the distribution of hits D40 = 3.62 cm. This work has no pretensions to study the mathematical models analysed inliterature and there is absolutely reason to doubt that information. That leaves roomsfor discussion about quality, dimensional accuracy and uniformity of the designedbarrels, the quality and uniformity of ammunition and quality of measurementequipment. To completely define the most appropriate form of the projectile guidance it isnecessary to repeat the experimental research several times with barrels andammunition of different quality and make measurements with quality measuringequipment in various weather conditions. The aim of this paper is to point to one possible way of defining a standardizedprojectile guidance, as for calibre 5.56 mm, and for all calibres up to 12.7 mm, whichmust meet the following criteria: security Service - the pressures, initial velocity andaccuracy must be in pre-defined borderline, simplicity of construction - different levelsof quality of tools and human resources, ease of maintenance, low cost of makingbarrels and certitude.LITERATURE[1] Cvetkovi , M. (1984). The use of non-stationary gas dynamics for the interior ballistics problem of weapons of small caliber, Ph.D. thesis, Zagreb.[2] Tan i , LJ. (1997). Numerical solution of unsteady internal ballistics model of small caliber weapons, Ph.D. thesis, Belgrade.[3] Gough, P. S., Zwarts, F. J. (1977). Modeling Heterogeneous Two-Phase Reacting Flow, AIAA Journal, AIAA / SAE 13th Propulsion conference, Orlando, Florida, July 11-13, Vol. 17, No. 1, pp. 17-25.[4] Jaramaz, S. (1991). Contribution to the study of flame propagation through a granular environment. Ph.D. thesis, Faculty of Mechanical Engineering, Belgrade.[5] Tan i , LJ. (1987). Determination of the pressure forcing the firing of weapons. Master thesis, Zagreb.[6] Petrovic, M. (1999). Analysis of a new form of projectile guidance for barrel cal. 5.56 mm. Ph.D. thesis, Mechanical Engineering, Belgrade.[7] Stiffler, K. (1982). Projectile sliding forces in a rifled barrel. Mississippi State University.92
  • COMPARISON OF TECHNIQUES FOR DETECTION OF FAILURE ROLLING ELEMENT BEARINGS 1 2 Pavle Stepani , Željko urovi , Aleksa Krošnjar 3, Aleksandra Pavasovi 4Summary: This paper presents advanced techniques for the detection of bearingfailures based on vibration signals. Capability of detection and diagnosis of someeffective techniques are discussed and compared on the basis experimental results. Inparticular, we analyzed the ID3 data mining technique, detection using statisticalpattern recognition and application of hidden Markov model (HMM). Comparing theexperimental results showed that the highest accuracy of detection techniques hasbeen realized using HMM, and for determining the type of bearing damage easier touse a statistical approach to pattern recognition. It is shown that the presence of faultscan be detected on-line monitoring of the probability of trained HMM for the correctbearing, which is determined by feature extraction from vibration signals. In addition,the application of HMM can be extended to consider problems of prognosis andprediction of bearing condition, which gives as an output time to failure.Key words: failure detection, data mining, statistical pattern recognition, hiddenMarkov model, condition based maintenance1. INTRODUCTION Most modern techniques for the detection of bearing failures is based on theanalysis of vibration signals collected from the bearing housings. The common goal ofall techniques is to detect the presence and type of failure at an early stage and tomonitor its development in order to assess the remaining life of a system and choosethe appropriate maintenance plan. Monitoring and predicting of the condition industrialsystems and processes at running are activities whose importance in recent yearsgreat, as based upon the concept of preventive maintenance. Timely knowledge of adefect allows the user to correct the problem occurred during a regular repair in orderto avoid additional, unplanned and often very costly delay because of sudden failure.The basic elements necessary for successful diagnosis, prognosis, and conditionbased maintenance (CBM) are shown in figure 1.1 Dipl. eng. el., Pavle Stepani , Belgrade, Lola Institute, pavle.stepanic@li.rs2 Prof. dr, Željko urovi , Belgrade, Faculty of Electrical Engineering, zdjurovic@etf.rs3 Dipl. eng. el., Aleksa Krošnjar, Belgrade, Lola Institute, aleksa.krosnjar@li.rs4 Prof. dr, Aleksandra Pavasovi , Belgrade, Faculty of Electrical Engineering, apavasovic@etf.rs 93
  • Pavle Stepani , Željko urovi , Aleksa Krošnjar, Aleksandra Pavasovi Fig. 1 The basic elements condition based maintenance The sensors are mounted on bearings that are monitored that gives thevibration signals. Then the feature extraction is to convert the signal of vibration insome type of parametric representation suitable for further analysis. Diagnostics isengaged in fault detection, isolation and identification of them when it happens.Detection of fault has a task to indicate that if something is wrong in the observedsystem, task of fault isolation is that find the defective component in the system whileidentification should determine the nature of errors which is detected. In the block forprediction is based on prognostic models and algorithms estimated the remaininglifetime of the observed bearings. Condition based maintenance is generally limited to the failure diagnosisperformance than the remaining life assessment. Diagnosis of failure is one of themost important phases in the CBM system. As a first step in diagnosis is to determinewhether the error is present or not, and therefore this paper will present and comparethree different methods for detection of bearing failures. Detection of rolling elementbearing faults actually represents the problem of pattern recognition and classification.The following section will briefly discuss these methods and then focus more onimplementation techniques and the results obtained in this study. For detaileddescriptions of the techniques recommended listed references.2. METHODS2.1 Decision Tree Decision tree is one of data mining techniques, and basically represents aclassifier that shows all the possible outcomes of a process. The paths that lead tothese outcomes are in the form of tree structure. The nodes that separate the differentclasses of branching based on if-then conditions. Algorithms to build trees arerecursively embedded in the decision tree partitioning the training set of datasuccessively more complete subsets. Partitioning criteria determine characteristicdifferences between decision trees, which are created different algorithms. ID3algorithm uses entropy as an indicator and from initial training set partitioning it madeless subset [1]. Features with the highest information gain are selected for the branchnodes based on information theory. Construction of decision tree very much dependson how the features of the training sample selected.94
  • Comparison of Techniques for Detection of Failure Rolling Element Bearings2.2 Statistical Pattern Recognition The main goal of pattern recognition is to decide which category the observedpattern belongs. One of the ways in which the pattern recognition can be performed isthat based on the series of patterns estimated conditional probability density functions,but such a procedure involves a very large amount of data and is associated with anumber of numerical problems. Therefore, it is often applied classification methods arenot optimal in any sense, but they are very simple. Commonly used parametricmethods are linear and quadratic classifier [2]. In this paper we design a quadraticclassifier based on the desired output.2.3 Hidden Markov Models HMM consists of two stochastic processes. The first process is a Markov chainwith finite number state, while another process for the given state generates a series ofobservations. The word "hidden" means that the HMM states are not directlyopservable. Since the sequence of states, as the basic stochastic process is notobserved, it can only be estimate through another set of stochastic processes thatgenerate sequences of observations. Thus defined model is the basis for thedescription of a number of different problems. From the point of view of modeling thevibration signals from rolling element bearings, HMM can be viewed as a combinationof two stochastic processes. These are the hidden Markov chains which are temporaryvariability of vibration signals in the hidden state sequence and observable processthat explains the spectral characteristics of vibration signals. So, despite the fact that inevery discrete time instant t system is in state qt , it every step emits a visible symbol,which in general can be a continuous function (e.g. spectrum). HMM can be describedby the following equations: X k 1 AX k Vk 1, (1) Yk CX k Wk ,where X k denote hidden process, Yk denote observation process, Vk and Wk areterms for noise, and A and C are parameters. In general HMM is represented by triplet A, B and , i.e. in compact notation A, B, , where the matrix A is the probability distribution of transition states,matrix B is the probability distribution of occurrence of certain symbols in certain states,and is the initial distribution of state [3]. The problem of training HMM is a subject ofadaptation (or re-estimation) model parameters A, B, in an effort to learn andcoding the characteristics of observational sequences that enable the model cancontinue to identify similar observation sequences. Optimization of parameters can bedone Baum-Welchs iterative algorithm [3], which in this paper and used.3. EXPERIMENTAL RESULTS Experimental data were collected from the station for testing rolling elementbearings. For the purposes of research and comparison of the above techiques, in thispaper we used vibration signals obtained from accelerometer mounted horizontally onthe bearing housing. Data were collected for the two states of bearings: normal(without defects) and defective (with some type defect). The activity preceding 95
  • Pavle Stepani , Željko urovi , Aleksa Krošnjar, Aleksandra Pavasoviclassification is the determination of which measuring quantities will in the bestpossible way characterize the pattern that needs to be classified. For the first twomethods (ID3 and quadratic classifier), extracted by 9 parameters in the time domainand 9 parameters in the frequency domain for a total of 18-dimensional feature vector,characterizing each of the tested bearings [2,4]. For HMM failure detection featureextraction from vibration signals was performed using Linear Predictive (LP) modelingfor the model of 10th order.3.1 ID3 Decision Tree ID3 uses information gain as a measure of selection between features. In orderto decrease computational complexity of algorithm for classification, dimensionreduction was performed in 18-dimensional feature vectors in the new 6-dimensionalspace of features. The reduction of dimensions was based on KL transformation [4].The process of discretization of continuous features in the ID3 algorithm is actually aprocess of selecting optimal threshold [1]. The classification was performed in twoclasses: defective and healthy (i.e. functional) rolling element bearings. Figure 2 showsthe decision tree for the detection of bearing condition using ID3 algorithm. To test theclassifier used is the "leaving-one-out" method which is estimated classificationaccuracy of 95.41%. y3 54.04 y3 54.04 y4 0.69 y4 0.69 y2 47.53 y2 47.53 y5 2.59 y5 2.59 y1 1333 y1 1333 Fig. 2 ID3 decision tree3.2 Quadratic Classifier Quadratic classifier is designed using the desired output method. For theapplication and control of quadratic classifiers previously performed dimensionreduction of the 18-dimensional feature vectors into two-dimensional space of features.Representation of patterns in two-dimensional space allows visualization of randomvectors and easily calculate the parameters of quadratic classifiers. Dimensionreduction was carried by using the criterion based on scatter measure [2]. Figure 3shows the position of two-dimensional vectors that characterize the analyzed bearingsand projected quadratic classifier. Testing the classifier, "leaving-one-out" method wasobtained classification accuracy of 98.98%. In Figure 3 shows the obviouslyseparability between defective and functional bearings. In addition, within the class ofdefective bearings noticeable clustering around the characteristic centers which mightindicate to a particular type of bearing damaged. This is primarily result of intelligentdimension reduction based on scattering matrices which takes account of classseparability.96
  • Comparison of Techniques for Detection of Failure Rolling Element Bearings y2 y1 Fig. 3 Quadratic clasifier ( - defective bearings, × - functional bearings)3.3 HMM Fault Detection For detecting presence of faults is enough to train one HMM for functionalrolling element bearings. Sequences of observations to train HMM is obtained bydividing the vibration signal of length one second division in ten equal segments. Foreach of these segments (length 0.1 s) the autocorrelation method of LP coefficients arecalculated for a model of order 10th. Accordingly, observational sequence consists of10 observations, i.e., O O1O2O3 O10 where each of the observations Ot a 10-dimensional vector of features extracted from the LP model. Based on trainingobservations for defective and healthy bearings codebook is generated using K-meansalgorithm [3]. Codebook size is M = 128 and on its basis will be carried out vectorquantization (i.e. discretization) of continuous feature vectors. A structure model ofHMM was chosen left-to-right model with three states and discrete observationsymbols. Calculate the probability of resulting in a estimate to what extent a sequencecorresponding to the adopted model in forward-backward procedure [3]. If theprobability above a predefined threshold, then in bearing is not present defect,otherwise the bearing is faulty. One way to select an appropriate threshold is theminimum value of probability calculated for training observation sequences fromnormal bearings: thresh min P O (2) normal In Figure 4 a) shows the iterative step in the training of HMM and thecorresponding logarithm of probability at each step. Figure 4 b) shows the logarithmsof probability of vibration signals for functional bearings and bearings with defects,obtained for the HMM of functional bearings. As the figure shows bearing faults aredetected with accuracy 100%. Probability of functional bearing is separated by aclearly defined threshold out of probability bearings with defect. On-line monitoring ofprobability HMM for functional bearings indicates ability to forecasting the condition ofbearing. Developing fault in the bearing will be reflected as an unexpected change inprobability, i.e. a decrease in the probability. Over time, as failure progresses, theprobability of HMM for a functional bearings has a downward trend. When the bearingscloser to end of useful life, the probability decreases rapidly, which indicates seriousdamage. 97
  • Pavle Stepani , Željko urovi , Aleksa Krošnjar, Aleksandra Pavasovi Fig. 4 a) HMM training curve b) Probabilities of normal and defective bearings4. CONCLUSIONS This paper compares some of the modern techniques for the detection ofbearing failures from vibration signals based on experimental results. Classifier basedon the ID3 decision tree is clear and easy to understand. Good characteristic of thisclassifier is a short period of training. Estimated accuracy of detection bearing faults is95.41%. One of the disadvantages of training phase, which can lead to less accuracyof detection is over-fitting. Quadratic classifier is easy to use and has a high accuracy of fault detection98.98%. This is primarily result of intelligent dimension reduction based on scatteringmatrices which takes account of class separability. As a result of this reduction ofdimension is a clustering of damaged bearings pattern around specific centers, whichcan be used to determine the fault type. Detection of bearing failures using discrete HMM of vibration signals showedclassification accuracy of 100%. Time training of HMM is longer than the previous twotechniques and algorithm used in training do not guarantee reaching a globalmaximum criterion function. Monitoring probability of HMM for functional bearings canbe predicted his condition in the future and estimate the remaining useful life.Acknowledgments This work was created within the research project "Development of the devicesfor pilots training and dynamic flight simulation of modern combat aircraft: 3 DoFcentrifuge and 4 DoF spatial disorientation trainer" that is supported by the Ministry ofScience and Technological Development, Republic of Serbia.REFERENCES[1] Quinlan, J.R. (1986). Induction of decision trees. Machine Learning 1, 81-106.[2] Stepanic, P., Latinovic I.V., Djurovic, Z. (2009). A new approach to detection of defects in rolling element bearings based on statistical pattern recognition. Int. J. Adv. Manuf. Technol., Vol. 45, No. 1-2, pp. 91-100.[3] Rabiner, L.R. (1989). A Tutorial on Hidden Markov Models and Selected Applications in Speech Recognition. Proceedings of the IEEE 77(2), pp. 257-286.[4] Stepani , P., Krošnjar, A. (2009). Detekcija ošte enja kotrljaju ih ležajeva primenom ID3 stabla odlu ivanja. Zbornik radova DEMI 2009, pp. 323-328.98
  • CRACK INITIATION LIFE OF NOTCHED METALLIC PARTS EXPOSED TO LOW CYCLE FATIGUE Strain Posavljak1, Miodrag Jankovic2, Katarina Maksimovic3Summary: The flat specimens with a central hole and one turbojet engine compressordisk have been studied in this paper as representatives of notched metallic parts. Inconditions of loading by blocks of positive variable force, the crack initiation life ofspecimens was determined experimentally. Using different approaches based on lowcycle fatigue criteria, crack initiation life of mentioned specimens was estimated also.Original Neubr’s rule and Topper’s and Sonsino-Birger’s modification of this rule, wereused for determination of stress-strain response at critical point. Estimation of crackinitiation life was carried out using Palmgren-Miner’s rule of linear damageaccumulation supported by Morrow’s, Manson-Halford’s and Smith-Watson-Topper’scurves of low cycle fatigue. Experimentally obtained results and results of estimation ofcrack initiation life were compared and analyzed. It was shown that approach whichinclude Sonsino-Birger’s modification of original Neuber’s rule and Palmgren-Miner’srule of linear damage accumulation supported by Morrow’s curves of low cycle fatigue,was made the best result of estimation.This approach was used for crack initiation lifeestimation of the studied turbojet engine compressor disk. For the purpose ofestimation, two blocks of rotation frequency were taken into account. One block thatpresents different engine ground controls and one block that presents different types offlights. Estimated crack initiation life of disk, expressed in flight hours, was comparedwith practical crack initiation life.Key words: flat specimens with central hole, turbojet engine compressor disk,damages, crack initiation life estimation1. INTRODUCTION Crack initiation life estimation of real notched metallic part exposed to low cyclefatigue understands knowing of data about blocks of loads, cyclic events in that blocks,cyclic properties of material used or nominated for workmanship, stress-strainresponse at critical point or point of expected crack initiation and damages provoked byall cyclic events.1 Ph.D., Strain Posavljak, Banja Luka, Faculty of Mechanical Engineering, (s.posavljak@urc.rs.ba)2 Ph.D., Miodrag Jankovic, Belgrade, Faculty of Mechanical Engineering, (mjankovic@mas.bg.ac.rs)3 Ph.D., Katarina Maksimovic, Belgrade, Office of Water Mangement, (tince@net.rs) 99
  • Strain Posavljak, Miodrag Jankovic, Katarina Maksimovic Stress-strain response at critical point (local stress-strain response) for allcyclic events and method of identification and counting of those events, have specialimportance. Local stress-strain response may be determined by measurement, by thefinite element method and by the methods that relate local stresses and strains withtheir nominal values. Flat specimen with central hole and one turbojet engine compressor disk, asnotched metallic parts, are dicussed in this paper. Methodology of crack initiation lifeestimation of these parts, based on low cycle fatigue criteria, is presented.1. CASE OF FLAT SPECIMENS WITH CENTRAL HOLE2.1 Some information about specimens Three flat specimen with central hole was made of stell 13H11N2V2MF in heattreatment state (Heating at 1000°C, Oil quenching, Tempering at 640°C, Air cooling).Experimentally obtained cyclic properties of this steel that marked as quenched andtempered (QT) steel, are contained in Table 1.Table 1 Cyclic properties of QT steel 13H11N2V2MF [1] No Propertiy Value 1 Modulus of elasticity, E [MPa] 229184.6 2 Cyclic strength coefficient, K’ [MPa] 1140.0 3 Cyclic strain hardening coefficient, n 0.0579 4 Fatigue strength coefficient, ’f [MPa] 1557.3 5 Fatigue strength exponent, b -0.0851 6 Fatigue ductility coefficient, ’f 0.3175 7 Fatigue ductility exponent, c -0.7214 Crack initiation life (CIL) of mentioned specimens marked with SPC1, SPC2and SPC3, was determined experimentally. Using the universal MTS system, the samewere loaded by blocks of positive variable force. Results of experimentally obtainedCIL are included in Table 2.Table 2 Results of experimentally obtained CIL of specimens [1] Specimen Crack initiation life [Blocks] SPC1 4000 SPC2 3600 SPC3 4200 Geometry of flat specimens with central hole, original block of positive variableforce and the same block with identified cyclic events, are shown in Fig. 1. 100
  • Crack Initiation Life of Notched Metallic Parts Exposed to Low Cycle Fatigue Fig. 1 Geometry of flat specimens with central hole, original block of positive variable force and the same block with identified cyclic events Cyclic events within block of positive variable force, defined as X-Y-X forcecycles, were identified using method of »reservoir« [1]. X-Y-X force cycles, sortedaccording to level /i/ and number of appearing /Ni/ within block, are contained in Table 3.Table 3 Identified X-Y-X force cycles within block of positive variable force Level Xi-Yi-Xi force cycle Nuber of force cycles within block, i [kN] Ni 1 6-120-6 1 2 48-120-48 3 3 72-120-72 2 4 48-96-48 12.2 Stress-strain response at critical specimen point At the beginning, the flat specimen with central hole was observed as idealelastic body. Its linear stress response, provoked by maximal force Fmax = 120 kN, wasobtained using the Finite Element Method (FEM) implemented in I-DEAS MasterSeries software [2]. Discretization was carried out by plane stress parabolic finiteelements. FEM model with trajectories of principal stresses 1 is shown in Fig. 2. Principal stress 1 has maximum value at critical specimen point P, 1max = 1,P= 1789.13 MPa and it with belonging strain is not real. Stress for point which is satisfying far from stress concentration area, taken asnominal stress n = 480 MPa, is equal to the ratio of force Fmax and surface of fullspecimen section (S = 5 50 = 250 mm2).Stress concentration factor Kt = 3.727, equal to the ratio of stress 1,P and nominalstress n, was served for transformation of linear stress-strain response at criticalspecimen point into nonlinear. Respecting memory of metals, nonlinear stress-strain response, at criticalspecimen point, was described by stabilized hysteresis loops assigned to all ith forcecycles contained in Table 3. Neuber’s rule [3], Topper’s modification [3,4] and Sonsino-Birger’s modification of this rule [5,6], were used for determining of nonlinear stressstrain response. The next two systems of equations 101
  • Strain Posavljak, Miodrag Jankovic, Katarina Maksimovic (1) 2 K2 t ni E 1 (2) n 2 i 1,2,3,4 E 2K belonge to Neuber’s rule. The first equations in these systems are two forms ofNeuber’s hyperbola [3]. The second equation in (1) is equation of cyclic stress-straincurve and the second equation in (2) is equation of Masing’s curve [3] for QT steel13H11N2V2MF. The values of nominal stresses ni and their ranges ni that used in (1) and(2), were calculated using expressions Yi 103 ni S (3) Yi Xi 103 ni i 1,2,3,4 S Values of cyclic properties, used in (1) and (2), with known stress concentrationfactor Kt, were taken from Table 1. Systems (1) and (2) in wich instead of Kt used effective stress concentrationfactor Kf=3.582, have been related to Topper’s modification of Neuber’s rule. Sonsino-Birgers modification of Neubers rule was based on the application ofthe next systems of equations 1 Kt ni Kt ni 1 2 E 1 (4) n i 1 E K 1 Kt ni Kt ni 1 2 E 1 (5) n 2 i 1,2,3,4 E 2K The first equations, in systems (4) and (5), present two forms of Sonsino-Birger’s curve. The second equations correspond to the second equations in systems(1) and (2). The values of Kt , ni , ni , E, K’ and n’ used in solving of systems (1) and 102
  • Crack Initiation Life of Notched Metallic Parts Exposed to Low Cycle Fatigue(2), used and for solving of systems (4) and (5) also. Numerical results of nonlinearstress-strain response at critical specimen point, based on the application of Neubersrule and Toppers and Sonsino-Birgers modification od this rule, are contained in Table4, Table 5 and Table 6.Table 4 Numerical results of nonlinear stress-strain respose at critical specimen point –Neubers rule i Xi-Yi-Xi force cycle mi i i [kN] [MPa] [MPa] 1 6-120-6 125.013 1514.669 0.00832044 2 48-120-48 345.916 1072.864 0.00468568 3 72-120-72 524.554 715.588 0.00312232 4 48-96-48 167.278 715.588 0.00312232Table 5 Numerical results of nonlinear stress-strain respose at critical specimen point –Toppers modification of Neubers rule Level Xi-Yi-Xi force cycle mi i i i [kN] [MPa] [MPa] 1 6-120-6 131.960 1490.921 0.00780806 2 48-120-48 361.741 1031.359 0.00450238 3 72-120-72 533.549 687.744 0.00300084 4 48-96-48 189.934 687.744 0.00300084Table 6 Numerical results of nonlinear stress-strain respose at critical specimen point -Sonsino-Birgers modification of Neubers rule Level Xi-Yi-Xi force cycle mi i i i [kN] [MPa] [MPa] 1 6-120-6 115.114 1496.466 0.00791852 2 48-120-48 326.998 1072.699 0.00468495 3 72-120-72 505.556 715.582 0.00312229 4 48-96-48 148.439 715.582 0.00312229 Nonlinear stress-strain response at critical specimen point, determined bySonsino-Birger’s modification of Neuber’s rule, is presented in Fig. 2. 103
  • Strain Posavljak, Miodrag Jankovic, Katarina Maksimovic Fig. 2 Fflat specimen with central hole: FEM model with linear stress response described by principal stress 1 trajectories (a) and nonlineal stress-strain response at critical specimen point determined by Sonsino-Birger’s modification of Neuber’s rule2.3 Specimen damages and crack initiatioin life estimation Estimation of specimen damage D, as damege at critical point, provoked byblock of positive variable force in Fig.1, was carriad out using Palmgrem-Miners rule oflinear damage accumulation [3,7,8,9] defined on the way 4 4 Ni D Di (6) i 1 i N 1 fi In above expression the damage provoked by ith force cycle is marked with Di.This damage presents the ratio of number of appearing Ni of ith cycle within block ofpositive variable force and number Nfi of the same cycle that flat specimen made of QTsteel 13H11N2V2MF can endure up to crack initiation. Numbers Ni are given in Table 1and numbers Nfi were determined using Morrow’s, Manson-Halford’s and Smith-Watson-Topper’s curve of low cycle fatigue (LCF). Morrow’s (M) curve [3,10] is the firstequation of the next system f mi Nb f c f Nf 2 E (7) i i 1,2,3,4 2 2 The first equation of system 104
  • Crack Initiation Life of Notched Metallic Parts Exposed to Low Cycle Fatigue c b f mi Nb f f mi f Nc f 2 E f (8) i i 1,2,3,4 2 2is Manson-Halford’s (MH) curve [3] and the first equation of system 2 2b b c PSWT max E f Nf E f f Nf 2 (9) i PSWT PSWT, i max,i E i 1,2,3,4 2is Smith-Watson-Topper’s (SWT) curve [3] of low cycle fatigue, where PSWT presentsSmith-Watson-Topper’s perimeter in which max,i = mi + i/2 . Stresses mi in systems(7) and (8) are mean stresses. Combining applied approaches of determining of nonlinear stress-strainresponse at critical specimen point, with different curves of LCF, using Palmgren-Miner’s rule, we obtained estimated values of damages D and crack initiation life (CIL =1/D) expressed in blocks of positive variable force. These estimated values arecontained in Table 7.Table 7 Estimated values of damages D and CIL for the flat specimen with central hole Sonsino- Toppers Birgers modification Neubers rule modification of Neubers of Neubers rule rule M curve of MH curve of SWT curve M curve of M curve of LCF LCF of LCF LCF LCF D 0.000431586 0.000858519 0.000535187 0.000335017 0.000349002 CIL 2317 1164 1868 2984 2865 The best result of estimated CIL of the flat specimen with central hole wasobtained using Topper’s modification of Neuber’s rule in combination with Morrow’scurve of LCF. This result is rejected because the application of effective stressconcentration factor Kf is debatable. Result of estimated CIL obtained using Sonsino-Birger’s modification of Neuber’s rule in combination with Morrow’s curve of LCF (CIL =2865 blocks) is accepted as the best estimated result. The same amountsapproximately 80 % of the smollest experimentaly obtained CIL that amounts 3600blocks (Table 2). 105
  • Strain Posavljak, Miodrag Jankovic, Katarina Maksimovic2. CASE OF ONE TURBOJET ENGINE COMPRESSOR DISK3.1 Some information about disk Here observed disk is the first stage low pressure compressor rotor disk ofR25-300 turbojet engine. It is dominantly loaded by centrifugal forces of blades andown centrifugal forces. For disks of this kind, engine ground controls and flights,described by blocks of rotation frequency /n/ in time /t/, have important role. In the caseof observed disk, two blocks of rotation frequency, taken into account. Blocks A and Bthat present different engine ground controls and different flights (Fig. 3). Fig. 3 Blocks of rotation frequency (block A presents different ground controls and block B presents different flights) Identified X-Y-X cycles of rotation frequency in blocks A and B are given inTable 8. Identification was carried out similar to identification of force cycles in block ofpositive variable force in Fig. 1.Table 8 Identified X-Y-X cycles of rotation frequency Block A Block B i Xi - Yi - Xi Ni i Xi - Yi - Xi Ni 1 0-100-0 1 1 0-100-0 1 2 35-100-35 3 2 70-100-70 3 3 50-100-50 1 3 75-100-75 2 4 80-100-80 2 4 80-100-80 1 5 85-100-85 1 5 85-100-85 3 6 35-85-35 1 6 70-95-70 1 7 75-90-75 1 8 70-85-70 1 Prescribed service life of observed disk, which because of premature crackscan not be reached, amounts 1200 flight hours. Practical service life as crack initiationlife expressed in flight hours (CILh), for Weibull’s probabilities 0.001 and 0.999,amounts 28 and 676 flight hours [1]. Material for workmanship of disk is steel13H11N2V2MF. Achieved quality of this steel is questionable. Because of that, thequestion was asked. Whether the disk made of QT steel 13H1N2V2MF with cyclicproperties in Table 1, can have a longer CIL? 106
  • Crack Initiation Life of Notched Metallic Parts Exposed to Low Cycle Fatigue3.3 Stress-strain response at critical disk point For determining of stress-strain response at critical disk point, it was enough toobserve one blade and critical region of our disk (at the first as ideal elastic bodies).Linear stress response and nodal reactions at blade root contact surfaces, inconditions of maximal rotation frequency n=186 s-1, were obtained using the finiteelement method (FEM) [2]. Using FEM and the same rotation frequency, mentionedreactions in transformed form used as nodal forces for obtaining of linear stressresponse of critical region of disk. Axysimmetric linear stress response of disk, when itobserved as blisk (bladed disk), was obtained also (Fig. 4). Fig. 4 Linear stress response of critical region of disk (left) and linear axisymmtrical stress response of disk when it observed as blisk (right) According to Fig. 4, maximal equivalent stress eq,max at point of expected crackinitiation (critical point P) and belonging strain are unreal. It can see that critical point Pcorresponds to real crack initiation. Equivalent stress at point P’ of blisk, whichcorresponds to critical disk point P, taken as nominal stress n, was served forcalculation of so-called equivalent stress concentration factor Keq = 7.45 which isdefined as ratio of eq,max and n. Similar, as in case of the flat specimen with central hole, nonlinear stress-strainresponse at critical disk point was determined using Sonsino-Birger’s modification ofNeuber’s rule based on application of the system of equations (4) and (5) in whichinstead of Kt used Keq. The values of corresponding cyclic properties of QT steel13H11N2V2MF were the same. Nominal stresses ni and their ranges ni which areused in mentioned systems, were obtained using expressions 2 Yi ni 223 100 2 2 Yi Xi ni 223 (10) 100 100 i iA 1,2,...,6/ i iB 1,2,...,8 Numerical results of nonlinear stress-strain response at critical disk point,provoked by blocks of rotation frequency A and B, are included in Table 9 and Table 10. 107
  • Strain Posavljak, Miodrag Jankovic, Katarina MaksimovicTable 9 Numerical results of nonlinear stress-strain response at critical disk point,provoked by block A Xi-Yi-Xi mi i i i [%] [MPa] [MPa] 1 0-100-0 112.778 1484.231 0.00768147 2 35-100-35 157.173 1395.440 0.00650403 3 50-100-50 235.722 1238.343 0.00545603 4 80-100-80 555.780 598.227 0.00261026 5 85-100-85 624.315 461.157 0.00201214 6 35-85-35 -42.239 996.615 0.00434976Table 10 Numerical results of nonlinear stress-strain response at critical disk point,provoked by block B Xi-Yi-Xi mi i i i [%] [MPa] [MPa] 1 0-100-0 112.778 1484.231 0.00768147 2 70-100-70 431.368 847.050 0.00369601 3 75-100-75 491.333 727.121 0.00317266 4 80-100-80 555.780 598.227 0.00261026 5 85-100-85 624.315 461.157 0.00201214 6 70-95-70 350.542 685.396 0.00299058 7 75-90-75 410.124 564.704 0.00246396 8 70-85-70 200.794 385.902 0.001683843.4 Disk damages and crack initiation life estimation Estimation of damages DA and DB provoked by blocks of rotation frequency Aand B, of observed disk, was carried out using Palmgren-Miner’s rule in the form 6 6 Ni DA Di A i 1 i 1 Nfi A (11) 8 8 Ni DB Di B i 1 i 1 Nfi B Numbers Ni in (11) we taken from Table 8 and numbers Nfi determined usingsystem (7), in which index i taken values i=iA=1,...,6 and i=iB=1,...,8. Resource between the two overhauls of R25-300 turbojet engine amounts 400flight hours [1]. The different engine ground controls in that resource is covered with438 blocks A and the different flights with 685 bocks B of rotation frequency. On thebase of this fact, simple expressions for estimation of damages D400h and D1h weredeveloped. 108
  • Crack Initiation Life of Notched Metallic Parts Exposed to Low Cycle Fatigue (12)Damage D400h is damage accumulated between two overhauls and D1h is the one flighthour damage. From (12), using values for DA and DB, we obtained values for damages D400h,D1h and estimated crack initiation life expressed in flight hours (CILh). The data aboutDA, DB, D400h, D1h and CILh = 1/D1h are contained in Table 11.Table 11 The data about DA, DB, D400h, D1h and CILh DA DB 0,00060924 0,00025664 438DA 685DB D400h 0,26684543 0,175796791 0,44264222 D1h = 0,00110661 CILh = 903 In Fig. 5, estimated CILh = 903 flight hours is compared with practical CIL. Fig. 5 Practical crack initiation life (CIL) and estimated crack initiation life (CILh) of observed disk Estimated CILh = 903 flight hours is greater than practical CIL = 676 flight hoursthat corresponds to Weibulls probability P(t) = 0.999 [1]. 109
  • Strain Posavljak, Miodrag Jankovic, Katarina Maksimovic3. CONCLUSION Methodology of crack initiation life estimation, applied for specimens withcentral hole and for observed disk, can be applied in the similar form, for all notchedmetallic part expossed to low cucle fatigue. Because of the use of efective stress concentration factor Kf in the endurancelimit area and because that the same, in the finite endurance are, depends on thenumber of load cycles N, application of Toppers modification of Neubers rule, fordetermination of nonlinear stress-strain response, should be rejected. On the example of the first stage low pressure compressor rotor disk of R25-300 turbojet engine it is shown that with cyclic material properties we can influence oncrack initiation life. Research in relate to crack initiation life of notched metallic parts are notfinshed. The question of accuracy of crack initiation life estimation and today is open.There are many unsolved problems. For example, application of linear damageaccumulation is questionable, because it does not take into account interactionbetween cycles and order of their appearance. On the estimation accuracy of crackinitiation life, accuracy of nonlinear stress-strain response, at critical points,significantly influence. Here is accuracy increased by introducing of Sonsino-Birger’scurve that is more accurate than Neuber’s hyperbola.LITERATURE[1] Posavljak, S. (2008). Fatigue Life Investigation of Aero Engine Rotating Disks, Doctoral dissertation (in Serbian), University of Belgrade, Faculty of Mechanical Engineering.[2] Lawry M. H. (1998). I-DEAS Master Series, Mechanical CAE/CAD/CAM Software, Student Guide, Structural Dynamics Research Corporation, SDRC Part Number P-60002[3] Bannantine, J.A., Comer, J., Handrock, J. (1990). Fundamentals of Material Fatigue Analysis, Prentice-Hall, Englewood Clifs, New Jersy.[4] Topper T. H., Wetzel R. M., Morroe JoDean (1969). Neuber’s Rule Applied to Fatigue of Notched Specimens, Journal of Materials, JMLSA, Vol 4, No. 1, pp. 200-209.[5] Sonsino, C.M. (1993). Zur Bewertung des Schwingfestigkeitsverhaltens von Bauteilen mit Hilife örtlicher Beanspruchungen, Konstruktion 45, 25-33.[6] Birger, I. A. (1985). Prognozirovanie resursa pri malociklovoj ustalosti, Problemy prochnosti, No 10, 39-44.[7] Jankovic, M. (2001). Low Cucle Fatigue (in Serbian), University of Belgrade, Faculty of Mechanical Engineering.[8] Palmgren, A. (1924). Die Lebensdauer von Kugellagern, Verfahenstechnik, Berlin, 68, 339-341.[9] Miner, M. A. (1945). Cumulative Damage in Fatigue, Journal of Applied Mechanics, 76, A159-164.[10] Morrow, J. (1968). Fatigue Design Handbook, Advances in Fatigue, Vol. 4. Society of Automotive Engineers, Warrendale, Pa., Sec 3.2, pp. 21-29. 110
  • DESIGN IMPROVEMENTS OF THE BUCKET WHEEL WITH DRIVE Sr an Bošnjak1, Zoran Petkovi 2, Miloš or evi 3, Nebojša Gnjatovi 4, Nenad Zrni 5Summary: Generaly, there are three basic reasons for the bucket wheel excavatorssubsystems redesign: (1) better customization of the machine versus operatingconditions; (2) easier maintenance and (3) failure of the substructure underconsideration. The improved design solutions, outlined in this paper, were developedto satisfy the first and second reason. Namely, an obsolete conception as well asconstant perrenial exploitation in heavy duty conditions made it necessary to redesignthe excavating device of the bucket wheel excavator SchRs 350. Several characteristicproblems which occur during exploitation and maintenance are herein discussed/listed.This paper presents a concept of redesign which includes substitution of the existigspur - geared bucket wheel gearbox with a planetary one and the redesign of the BWbody and its connection with the output gearbox shaft. It also shows the results of thecomparative load analysis for the original and the redesigned bucket wheel body.Protection of the original bucket wheel drive of the excavator SchRs 1760 was realisedby magnetic powder clutches. Because of the relatively high price and also theproblems noticed during exploitation (e. g. overheating resulting in fusing of thepowder) this study developed a solution design of the bucket wheel drive with fluidcouplings. Thanks to the mentioned redesign, partial revitalizations of the bucket wheelexcavators are realised, leading to considerable increase of the machines reliability aswell as the reliability of the complete surface mining system, and, additionally, makingthe maintenance process easier.Key words: bucket wheel, gearbox, overload clutch, external load identification,redesign, FEM1. INTRODUCTION Bucket wheel excavators (BWE) present the backbone of the open pit coalmining system. Speaking freely, the above class of machines is built up andaround thebucket wheel (BW), which in turn realizes the basic function of the machine – soil1 Professor Sr an Bošnjak, Belgrade, Faculty of Mechanical Engineering, sbosnjak@mas.bg.ac.rs2 Professor Zoran Petkovi , Belgrade, Faculty of Mechanical Engineering, zpetkovic@mas.bg.ac.rs3 Researcher Miloš or evi , Belgrade, Faculty of Mechanical Engineering , mddjordjevic@mas.bg.ac.rs4 Teaching Assistant Nebojša Gnjatovi , Faculty of Mechanical Engineering, ngnjatovic@mas.bg.ac.rs5 Associate Professor Nenad Zrni , Faculty of Mechanical Engineering, nzrnic@mas.bg.ac.rs 111
  • Sr an Bošnjak, Zoran Petkovi , Miloš or evi , Nebojša Gnjatovi , Nenad Zrniexcavation. The design of the BW and its drive determines basic BWE performances –above all, capacity. As stated in [1] „design errors in these areas would be difficult andexpensive to correct, if at all, after the machine has been constructed”. From a historical point of view, BWs have been developed from the cell-type(which was. up untill the mid-1950s, the only BW design type in use), through the half-cell type (introduced in the mid-1960s), to the present cell-less type which became theprefered BW design. During almost a century of BW improvement, the BW gearboxes wereimproved as well.. Thanks to their compactness and relatively small weight, planetarygears pushed out the spur – geared gearboxes, which have been used predominantlyuntil then. Besides BWs and gearboxes, under permanent improvement are also theconcepts of their bearings as well as the design of supporting elements and powertransmission parts [2]. The natural tendency to constantly improve the performance of the BWE,especially its capacities, has not been followed by adequate calculation methods andproduction technology. Good proof of this is the relatively frequent damage of BWEsubsystems [3-7]. Regardless of the cause, failures of high performance machines, ofwhich BWE is one, always lead to very high financial losses [8,9]. Generally, there are three basic reasons for the BWE redesign: (1) bettercustomization of the machine versus operating conditions; (2) easier maintenance and(3) failure of the studied substructure. This paper will present several BW with drive redesign solutions developed bythe University of Belgrade - Faculty of Mechanical Engineering.2. CASE 1 – BWE SchRs 350 An obsolete conception and constant perrenial exploitation in heavy dutyconditions made it necessary to redesign the excavating device of the BWE SchRs350, Fig. 1(a), which was put into service in 1961 [10]. One bedding of the BW shaft (bearing mark 23076K) is placed between BWand gearbox, Fig. 1(b). In order to enable replacement of the mentioned bearing it isnecessary to dismantle the complete gearbox, including the output shaft gear,Fig. 1(c), which is a relatively long and complex procedure. First of all, the machineshould be disburdened by reclining the BW boom, counterweight and dischargingboom, Fig. 2(a). The dismantling of the BW gearbox should be done only uponcompleting the previous. Because of the structure complexity, it is necessary to openthe gearbox, Fig. 1(c), and do its succesive dissasembling. Only then is it possible todismantle the gear from the output shaft, Fig. 1(c). After replacing the bearing , thegear is to be heated before installation, Fig. 2(b). A minimum of 7 days is needed tocarry out the above procedure. Low reliability, followed by frequently occurring breakdowns, leads toconsiderable decrease of the BWE exploitation effects. Besides that, it was necessaryto do the following: (a) Build in a new BW gearbox of the planetary type, providing axialdissasembling of the BW and its shaft connection; the existing BW drive enables a -1 -1 -1choice of BW shaft speed among the following values: 6 min , 7.5 min and 9 min ,while the speed of the appropriate planetary gearbox output shaft is 8.2 min-1;112
  • Design improvements of the bucket wheel with drive(b) Redesign and manufacture the new BW, Figs 2(c) and 2(d), with a larger number ofbuckets: instead of nB = 8 for the original , nB = 10 for the redesigned BW. Changeability of number of buckets to catch the soil is the main cause ofchangeability of external load caused by the resistance-to-excavation. The in-housedeveloped software RADBAG [6] was applied for analyzing the external load, Fig. 3,and determining its dynamic characteristics caused by the change in the total numberof buckets. Based on the research results, it is conclusive that the redesignedexcavating device gives more favorable external loading. In order to make the substitution of the BW shaft bearing easier and thedowntime of the machine and complete system substantially shorter, the connectionbetween the BW gearbox and the BW shaft is redesigned, Fig. 4. Finally, the general view of the redesigned BW boom head is shown in Fig. 5. Fig. 1 (a) BWE SchRs 350; (b) BW shaft bedding; (c) BW drive dismantling Fig. 2 (a) BWE temporary supporting; (b) heating the output gear; (c) 3D model of the redesigned BW body; (d) BW body FEA Fig. 3 External load coordinates 113
  • Sr an Bošnjak, Zoran Petkovi , Miloš or evi , Nebojša Gnjatovi , Nenad Zrni Fig. 4 Redesigned connection of the BW gearbox and BW shaft Fig. 5 Redesigned BW boom head3. CASE 2 – BWE SchRs 1760 The original solution of the BW drive is shown in Fig. 6. The BW gearbox (5) isdriven by two AC motors (1) of 550 kW each, which are connected to the input stagesby cardan shafts (2) and overload clutches (3) of the magnetic powder type. Becauseof the relatively high price of overload clutches and the problems noticed duringperennial exploitation (e.g. overheating resulting in fusing of the powder), it wasnecessary to substitute the above mentioned clutches with fluid couplings. The type offluid couplings (TE) is determined by the external load characteristics, Fig. 7, which areidentified by applying the procedure described in [6]. The nominal fluid coupling size(866) is defined by the motor power and motor shaft nominal number of speed. Thedrum of the existing double shoe brake is mounted on the output shaft of one of thefluid couplings, Fig. 8. Installation of the fluid couplings in the BW drive increasesreliability and enables full machine overload protection.114
  • Design improvements of the bucket wheel with drive Fig. 6 BWE SchRs 1760 : 1 – AC drive motor; 2 - cardan shaft; 3 - magnetic powder clutch; 4 - brake; 5 – BW gearbox; 6 - auxiliary drive motor; 7 - auxiliary drive brake-2800 Mzr [kNm]-2600-2400-2200-2000-1800-1600-1400-1200 o [ ] -70 -61 -53 -46 -40 -35 -30 -25 -20 -15 -11 -7 -2 2 7 11 15 20 25 30 35 40 46 53 61 70 Fig. 7 Dependence of the BW torque on the slewing platform position (angle ) Fig. 8 Fluid coupling: (a) without brake; (b) with brake 115
  • Sr an Bošnjak, Zoran Petkovi , Miloš or evi , Nebojša Gnjatovi , Nenad Zrni4. CONCLUSION The presented concepts of the BWE working devices redesign eliminate theoriginal design drawbacks which deal with the accessibility and possibility of repair orsubstitution of the damaged subassemblies as well as with safety against overloading.Besides, they are in full agreement with contemporary BWE excavating deviceconceptions. Finally, thanks to the presented redesign, partial revitalizations of thebucket wheel excavators were realised, leading to a considerable increase of themachines reliability as well as the reliability of the complete surface mining systemwhile at the same time making the maintenance process easier.LITERATURE[1] Durst, W., Vogt, W. (1989). Bucket Wheel Excavator. Trans Tech Publications, Clausthal-Zellerfeld.[2] Gnilke, M. Intelligent retrofit solutions for bucket wheel excavators - Review of recent projects in central - and eastern Europe, from http://www.baumaschine.de, accessed on 2009-10-11.[3] Rusi ski, E., Czmochowski, J., Iluk, A., Kowalczyk, M. (2010). An analysis of the causes of a BWE counterweight boom support fracture. Engineering Failure Analysis, vol. 17, no. 1, p. 179-191.[4] Rusi ski, E.; Harnatkiewicz, P.; Kowalczyk, M., Moczko, P. (2010). Examination of the causes of a bucket wheel fracture in a bucket wheel excavator, Engineering Failure Analysis, vol. 17, no. 6, p. 1300-1312.[5] Bošnjak, S., Zrni , N., Simonovi , A., Mom ilovi , D. (2009). Failure analysis of the end eye connection of the bucket wheel excavator portal tie-rod support. Engineering Failure Analysis, vol. 16, no. 3, p. 740-750.[6] Bošnjak, S., Petkovi , Z., Zrni , N., Simi , G., Simonovi , A. (2009). Cracks, repair and reconstruction of bucket wheel excavator slewing platform. Engineering Failure Analysis, vol. 16, no. 5, p. 1631-1642.[7] Bošnjak, S.; Petkovi , Z.; Zrni , N.; Panteli , M., Obradovi , A. (2010). Failure analysis and redesign of the bucket wheel excavator two-wheel bogie, Engineering Failure Analysis, vol. 17, no. 2, p. 473-485.[8] Dreyer, E. (1995). Cost-effective prevention of equipment failure in the mining industry. International Journal of Pressure Vessels and Piping, vol. 61, no. 2 - 3, p. 329-347.[9] de Castro, P.M.S.T., Fernandes, A. A. (2004). Methodologies for failure analysis: a critical survey. Material and Design, vol. 25, no. 2, p. 117-123.[10] Bošnjak, S., Petkovi , Z., Gnjatovi , N., or evi , M. (2009). Redesign of the Bucket Wheel Excavating Device. MHCL’09 Conference Proceedings, p. 123-128.AcknowledgementsA part of this work is a contribution to the Ministry of Science and TechnologicalDevelopment of Serbia funded project TR 35006.116
  • DESIGN OF TESLA-TIFFANY CASCADE FOUNTAIN AS A SAMPLE OF TESLA`S RESEARCH CREATIVITY IN FIELD OF MECHANICAL ENGINEERING 1 2 3 Aleksandar Marinkovi , Aleksandar o i , Bratislav Stojiljkovi , Milan Vuli evi 4Summary: Nikola Tesla (1856-1943) made numerous inventions and discoveries invarious fields, from transmission of electrical energy, fluorescent lighting, wirelesstransmission and remote control to numerous applications of high-frequency currentsin industry and medicine and peculiar inventions in the field of mechanical engineeringand aeronautics. One of his inventions which differs considerably from everythingelse this great scientist and inventor dedicated the largest part of his creativity to – wasa fountain. The abundant documents in the archives preserved in the Nikola TeslaMuseum in Belgrade demonstrate Tesla’s great engagement in this at first glanceless significant segment of his creative activity. This paper presents the originaldesign of Tesla-Tiffany cascade fountain, as a result of Tesla’s creativity in field ofmechanical engineering. Using the original sketches, technical drawings and somecalculations related to the dimensions of fountain elements, the authors analyze anddescribe his designs of fluid propulsion and flow. In addition, the paper gives somerecent calculations as a comment and contribution to Tesla’s original design, in orderto present some modifications and improvements in his construction. CAD modelsbased on original variants of fountain and its components are also presented.Key words: Nikola Tesla, Mechanical Engineering, Design, Cascade Fountain1. INTRODUCTION Tesla’s abundant and interesting creative activity in the field of mechanicalengineering has only sporadically been quoted in the greatest number of biographicaldepictions of his life and opus. Accordingly, such terms as e.g. turbines, pumps,fountains, speedometers or aeroplanes, often act confusing even on a connoisseur,1 Assistant Professor, Aleksandar Marinkovi , Belgrade, Mechanical Engineering Faculty University ofBelgrade, Serbia, amarinkovic@mas.bg.ac.rs2 Teaching and Research Assistant, Aleksandar o i , Belgrade, Mechanical Engineering Faculty Universityof Belgrade, Serbia, acocic@mas.bg.ac.rs3 Museum Curator, Bratislav Stojiljkovi , Belgrade, Nikola Tesla Museum, Serbia, bratislav.stojiljkovic@tesla-museum.org4 MSc student, Milan Vuli evi , Belgrade, Mechanical Engineering Faculty University of Belgrade, Serbia,v_milance@hotmail.com 117
  • Aleksandar Marinkovi , Aleksandar o i , Bratislav Stojiljkovi , Milan Vuli evideemed to be at least averagely informed about Tesla’s inventions and discoveries.Nikola Tesla devoted himself to machine engineering at the beginning of the 20thcentury, at the time when his grandiose Long Island project failed, due to the lack offunds. He desperately needed a new invention at the time, similar to the inductionmotor, required by the industry, of a simple construction and cheap to produce, in theorder to fund his further research. Next twenty years of his life Tesla dedicatedexclusively to researches in the field of mechanical engineering, making also use of hisinventions in the field of alternating currents (motors and generators) in order to securethe necessary funds [1]. Thanks to his specialist knowledge and exceptional creativity he was able tosolve some technical problems in mechanical engineering and to realize a wide rangeof new inventions and constructive solutions based explicitly on physical laws. One ofTeslas solutions in the field of mechanical engineering, less known until now, is hispatent for a fountain. This invention differs from customary solutions applied tofountains and aquariums of the time, where the fluid was sprayed or jetted withappropriate devices for decorative purposes [2].2. TESLAS FOUNTAIN AND COOPERATION WITH TIFFANY STUDIOS The Tesla fountain, a completely new type for that time, had to introduce a newtype of beauty, which relied principally "on a fascinating spectacle of the large mass offluid in motion and the display of seemingly great power". Water or other fluid would belifted from the fountain receptacle to the horizontal flared out top of conduit, to overflowit in the form of a unique circular cascade, making a splendid sight for an observer’seye. By applying only 1/25 of a horsepower (~30W) to the shaft and assuming a lift ofeighteen inches (~45.7cm), more than one hundred gallons (~378,5l) per minute couldbe propelled to flow over the flared top of conduit, one foot in diameter with the depthof the overflowing fluid of approximately one-half inch (~1.27cm). As the circulationwould be extremely rapid the total quantity of liquid required would be comparativelysmall [3]. Fig. 1 The circular water cascade at the top of the overflow The new type of fountain would allow for «realization of beautiful and strikingviews through additional colourful illumination and the disposition of voluminouscascades», with relatively low consumption of energy. Tesla’s application for his modelof fountain was filed at the United States Patent Office on October 28, 1913, and theletters patent No.1.113.716 was issued one year later, on October 13, 1914. Over118
  • Design of Tesla-Tiffany Cascade Fountain as a Sample of Tesla`s Research Creativity in Field of Mechanical Engineeringmore than eight years, with shorter or longer breaks, Tesla developed numerousmodels of fountains, but without significant commercial success. The preservedarchive documentation, e.g. general designs, technical drawings, calculations,correspondence with companies and individuals, magazine articles, various brochuresand the like, provide an insight into the scope of his research in this field of mechanicalengineering. After submitting his patent application for "certain new and usefulimprovements in fountains" at the end of October 1913, Tesla started searching for theways and options to realize his invention. He offered the model of his cascadefountain, based on the new principle, to Louis Tiffany, inventor of a special stainedglass technique, famous designer and the owner of Tiffany Studios. The review of thepreserved archive documents (letters from the correspondence between these twoimportant men and between Tesla and Tiffanys associates, as well as calculations,drawings of the fountain and other documents), shows that this cooperation lasted fromDecember 4, 1913, to January 9, 1916 [4]. At the end of 1913 and the beginning of 1914 Tesla intensively cooperated withJoseph Briggs, Manager of Tiffany Studios, on the development of particular structuralparts of the cascade fountain and the manner of its illumination. Joint venture wasformalized by signing a contract on business cooperation between Nikola Tesla andStudy Tiffany on February 17, 1914. For the Tesla – Tiffany cascade fountain Teslaelaborated a complete design of the plant and its components. He also specifiedcharacteristics of the propelling motor and proposed a very interesting solution forcomplete electric illumination. Fig. 2 Vertical cross section of Tesla-Tiffany Cascade Fountain, New York, December 22, 1913 For circulation of water and creation of “the fascinating spectacle of a largemass of fluid in motion”, the classic propeller was planned, with blades directly fixed tothe driving shaft of electromotor. The driving assembly was placed vertically along thecentral axis of fountain in the interior of the elevated conical overflow. In the design of 119
  • Aleksandar Marinkovi , Aleksandar o i , Bratislav Stojiljkovi , Milan Vuli evithis fountain Tesla did not make use of his new principle of energy transmission fromthe fluid and to the fluid i.e. of his invention of a turbine with bladeless discs, patentedon May 6, 1913. For the illumination of the fountain and realization of a more beautiful and morefascinating spectacle Tesla used different automatically controlled coloured lights.Electric lighting of Tesla – Tiffany cascade fountain consisted of two lighting systemsindependent from one another, differing in number and power of electric bulbs used.One system, installed on the outer side of the fountain, was composed of four sets, twobulbs of different colours in each, distributed at an angle of 90º and placed verticallyunder the top of the elevated conical overflow. Two sets had blue bulbs while theremaining two were equipped with violet and yellow-green bulbs, respectively. Theother lighting system was placed in the interior of the fountain, under the glass bell of aparticular design. This system consisted of three bulbs of greater power, distributed atan angle of 120º and installed horizontally above the driving assembly of the fountainin the middle of the elevated conical overflow. A circular glass plate, composed of nineequal segments, was placed above the horizontally built-in bulbs. Each one of thesesegments was of stained glass: yellow-green, blue or violet. The glass segments of thesame colour were fixed at an angle of 120º, thus forming a uniformly coloured set. Alternation of colours of electrical lighting during the work of the fountain, asrequested by Louis Tiffany, had to be realized according to the following sequence:beginning from the yellow-green the colour had to be changed into blue, from blue toviolet, from violet into blue and finally from blue into yellow-green. The time required forthese alternations had to be about three minutes. Though the cooperation between NikolaTesla and Louis Tiffany ended withoutthe realization of planned projects and expected commercial success, it represents apart of Tesla’s opus in the field of fountains and makes evident his broad interests,extreme creativity and ease in solving complex engineering problems in his own,original way.3. CALCULATIONS OVERVIEW WITH 3D MODELS Starting point for hydraulic calculation of fountain will be the weir height,denoted as h in Fig 3. Flow over weirs is very complicated in nature because variouscomplex phenomena are present: streamlines curvature, recirculation flows, dominanteffect of inertial forces, etc. In this paper simplified hydraulic calculation is considered,based on engineering principles which are simplified theory corrected by experimentaldata [5]. Based on original drawings and fontain dimensions, we may assume that weirheight is h = 1 cm . Now the volume flow rate over weir is determined by followingequation: Q Cd D g h3 / 2 , where Cd is discharge coefficient and g is gravitational acceleration. Dischargecoefficient is complex function and it depends on Reynolds and Webber number andalso on weir shape. It is determined experimentally and in literature values are knownfor other types of weirs, like crested weir, V shaped weir, trapezoidal weir, etc. For typeof the weir on the fountain there are no data in literature, but based on other similartypes it can be estimated that it’s value is Cd = 0.85.120
  • Design of Tesla-Tiffany Cascade Fountain as a Sample of Tesla`s Research Creativity in Field of Mechanical Engineering Fig. 3 Sketch of the weir of Fountain Now it possible to calculate volume flow rate using previous equation and it’svalue is Q = 4.182 lit/s . With this flow rate average fluid velocity over weir can becalculated using the equation: Q uav,1 u1 , D h and it’s value is u1 = 0.532m/s . Now the the pump head can be calculatedusing famous Bernoulli equation, written for water level in lower reservoir of thefountain and the weir. The equation is: 2 u1 u2 Hp H Hl H K* 1 , 2g 2g where is kinetic energy correction factor (Coriolis coefficient) and Hl is sumof all head losses. Head losses are usually expressed in terms of fluid kinetic energy,like Ku12/2g, where K is head loss coefficient. Based on geometry of fountain and it’sdimensions, and velocity distribution in the weir cross section it can be estimated thatK= + K 5, and it follows that value of pump head is Hp = 0.53m. With somestandard value of efficiency factor of pump p = 0.8 it can estimated that power neededfor work of the pump is P = g Hp Q / p = 30W. It can be concluded that powerneeded for pump drive is relatively low, which is expected if we have in mind that lift ofthe fountain is only H = 0.51m. Based on data for flow rate and pump head we canchoose appropriate pump for this fountain from catalogs of pump manufacturers. Concerning the calculation of transmission mechanism for lightening system,we can say that main goal was to reduce nominal rotational speed of electric motorfrom nem = 1150min-1 to requested nex = 0,28min-1 for driving of lighting system. For thispurpose Tesla had constructed a transmission system that consists of four stage gearreduction, with 96 and 12 tooth cylindrical gears and gear ratio i = 8 for each stage.Taking into account a friction torque loss during the rotation of lightening mechanism,we could calculate electric power for driving this mechanism [6]. Preliminary calculationof electric power needed for the transmission system gives us insignificant energy loss(less than a 1 mW) compared with hydraulic part, due to a very small dimensions andlight gears material such as negligible external load. At the end of this paper, the Fountain CAD model was made using CATIA V5software tool [7], where complete Assembly such as driving mechanism were designedas shown in Fig. 4. 121
  • Aleksandar Marinkovi , Aleksandar o i , Bratislav Stojiljkovi , Milan Vuli evi Fig. 4 CAD Model of the Fountain and driving system with electric motor 4. CONCLUSION This paper is an attempt aimed to present the Tesla-Tiffany Cascade Fountainas an example of Tesla’s reseach creativity in field of Mechanical Engineering. Basedon materials from Nikola Tesla Museum with original sketches, technical drawings andsome calculations related to the dimensions of fountain elements, authors analyse thelast variant of fountain and its components focused on calculation and energy lossaspects of its operating. Even it was not the main objective of the paper, here ispresented the CAD model of the Fountain where its assembly with driving andtransmission mechanism could be observed. For further investigation of this Tesla’sinventions in field of Mechanical Engineering, authors attend to continue analysis offew other transmission and shape variations in order to present some modificationsand improvements in his construction and also to made CAD models based on originalvariants of the Fountain.LITERATURE[1] B. Jovanovic (2001). Tesla – Spirit, Work, Vision, Freemental, Belgrade.[2] Tesla III Millennium, The Fifth International Conference, Proceedings, 15- 18.10.1996. Belgrade,[3] Nikola Tesla. (1996). Patent IV, Institution for Textbooks and Teaching instruments, Belgrade.[4] Teslas Fountain. (2007). Nikola Tesla Museum, Belgrade.[5] White, F.M. (2003), Fluid Mechanics, McGraw Hill.[6] Decker, K.H. (1995), Maschinen elemente. Hanser Verlag, 677 S.[7] Hoffmann, M., Hack, O., Eickenberg, S. (2005). CAD/CAM mit CATIA V5. Carl Hanser Verlag. München, Wien.122
  • DESIGN SEALS FOR REAL CONNECTIONS 1 Svetislav Lj. MarkoviSummary: In order to ensure non-leakage of surfaces, the joints are sealed with flatseals of elastic material. Increase of the reliability of sealing is achieved by using theseal coated firming. Soft material seals are used for joints made with rivets, hooks ...,when the seal on connection suffers the additional burden. Setting the seal is based onfixing the seals in parallel with surfaces that are to be sealed, so that a seal covers theentire surface.Key words: gaskets, joints, design1. INTRODUCTION Sealing gaskets of soft material is always associated with relatively bigger orsmaller changes in distance between the parts which are fixed. In mechanicalengineering it is usually needed to seal compounds based on “metal on metal”principle, while preserving the correct mutual arrangement of parts circuit. When installing gaskets it is very important to take into account the propersetting, centering gaskets. To increase the reliability of sealing, gaskets are lubricatedand sealants are used.2. SEALING OF SOLID COMPOUNDS The task of sealing the solid compounds can be solved in several ways. Nondemountable and rarely separable joints are sealed with mixtures with sealingcapabilities: bakelite, liquid glass (sodium silicate). Surfaces previously treated by ironing or by scratching, when connecting arecovered with a thin layer of sealant - lubricant. Oilers are usually produced by the colorof molten red hot oil (lead white, lead paint), iron powder or colloidal graphite in oil. Italso applies to suspension of colloidal graphite, covered with dry silver-plated graphite. Another way of sealing the solid compounds consists of setting up asubmerged elastic seals of rectangular or circular cross-section of joint surfaces.Gaskets are placed in the channels created along the entire rim of the joint. In the freestate, the gasket gets above the joint appearances of strictly fixed size (Fig. 1, left),which depends on the material and the desired force of sealing. In the tensile state of1 Dr Svetislav Lj. Markovi , prof., Visoka škola tehni kih strukovnih studija a ak, Srbija,e-mail: svetom@open.telekom.rs 123
  • Svetislav Lj. Markovithe surface of the joint a contact occurrs, and the gasket material gets plastic or elasticseal deformation and allows the sealing of the surface of the joint (Fig. 1, right). Fig. 1 Sealing compound, “metal on metal” principle, the positions before and after execution To increase the tightness of the seal surfaces that are sealed, shallowchannels are carved up (Fig. 2) in which deformable gasket materiasl are set. For thisreason, a seal cam should be ridged (Figure 3). With tightening, the reefs getcompressed and form a red channel that acts like labyrinth seal. Fig. 2 Sealing compound “metal on metal” principle, by use of the channels, the positions before and after execution124
  • Design seals for real connections Fig. 3 Sealing compound “metal on metal” principle, with the cam seals, the positions before and after execution In Figure 4 it is presented the elastic sealing gasket set in an enclosed space,which consist of the channel on the one side and protrusion on another surface. Thisway of presentation is an advantage over the round edges, which consist of channelsand drains that must be made with the necessary degree of accuracy. Fig. 4 Sealing compound “metal on metal” with round flanges Round (circular) rims are sealed with the elastic metal rings (Fig. 5), usually Z -type (so-called Gopher rings). Fig. 5 Sealing compound by Gopher rings, the positions before and after execution 125
  • Svetislav Lj. Markovi3. TIN GASKETS To ensure the tightness of the surface, the joints are usually sealed by tin matmade of elastic material. Parts of the mechanical transmissions are assembled withsoft seals (in those cases where there is no necessity of exact mutual position ofparts). Gaskets of soft material are used for joints realized with the helping hand ofscrews, bolt, hooks ... when seal suffers tensile loads. There is a big application of steel-reinforced seals that include elastic materials(rubber, plastics, asbestos ...), closed by a coating of soft metal (copper, brass). Thediversity of such seals is shown in Figure 6. Fig. 6 Reinforced-reinforced seals Figure 7 shows an example of application of reinforced gasket for sealingthreaded connections. Gaskets of this type can be used multiple times. Fig. 7 Sealing of threaded continuation by reinforced seals4. SEALING COMPOUNDS WHICH CONTAIN LIQUIDS Figures 8-15 show the sealing possibilities of cylindrical joints which are subjectto the pressure of fluid (a case of “wet” tubes of an internal combustion engine cooledwith liquid).126
  • Design seals for real connections Fig. 8 Fig. 9 Fig. 10 Fig. 11 Fig. 12 Fig. 13 Fig. 14 Fig. 15 The simplest form of sealing is achieved by placing a rubber ring round sectionin the paper tube channel (Fig. 8). In the free state the ring withdraws above thesurface of the paper tube. In introducing paper tube in the water jacket, the ring iscompressed and reinforces the paper tube in the casing (lining). To ensure that sealing 127
  • Svetislav Lj. Markovicompound is placed well, a few rings one after the other are placed (Fig. 9). Theimproved construction of such a seal is shown in Figure 10. Here is a channel createdwith a chamfered part directed to the side opposite of the of fluid pressure effect.Under the pressure of fluid the rings are constantly pushing the narrowed part of thechannal and adhere onto the sealing surfaces with the force proportional to thepressure. Among the rings there is a furrow which has a drainage hole connected tothe atmosphere (air). In case of breach of the first ring, the fluid leaks out through thedrain hole. The second ring which is, in a given case, relieved of the pressure,prevents further penetration of liquids. To increase the reliability of the working party of the sealing, several sealingrings are set (Fig. 11). Different types of channels and rings are shown in Figure 12. Inthese cases it is necessary to assure that the intersection channel is higher than theintersection of the rings, otherwise the rubber as incompressible material can cause asignificant radial force and the occurrence of deformation at the seal. To provide acompact grip of the rings to the wall “jackets”, on the inside of the rings wavy two-sidering springs are set (Figure 13). In order to avoid overloading of the sealing rings and to maintain a constantpressure during the exploitation, in the compounds are introduced elastic elements inthe form of annular conical spring (Figure 14) or the cam ring springs of round crosssection (Figure 15). In mechanical engineering it is often necessary to seal compounds whichcontain liquids, and which are connected to each other by round windows andopenings. They are to be sealed by tin seals of elastic material. Compounds that areexposed to high pressure and high temperatures are tightened by asbestos reinforcedseals (Figure 16). Fig. 16 Copper-asbestos gasket5. CONCLUSION Rubber is widely used for sealing joints. Mostly used is synthetic rubber whichis has many advantages over natural rubber. Some of the advantages of syntheticrubber are acceptable resistance to oil, petrol, kerosene, high temperature, highelasticity. Disadvantages of rubber can be repaired by its reinforcement (by pressing)or vulcanization.REFERENCES[1] P. I. Orlov, Osnovi konstruirovanija 1, „Mašinostrojenije“, Moskva, 1988.[2] Markovi S.: Technological suitability of machinery parts, Proceedings 12th International Research/Expert Conference „Trends in the Development of Machinery and Associated Technology“ TMT 2008, Istanbul, Turkey, 2008.[3] Markovi S.: Razvoj oblika mašinskih proizvoda u zavisnosti od tehnologije izrade, Zbornik radova sa 32. savetovanja proizvodnog mašinstva Srbije sa me unarodnim u eš em, Novi Sad, 2008.128
  • DETECTION OF STRUCTURAL DAMAGE LOCATION USING FREQUENCY RESPONSE FUNCTION DATA Valentina Golubovi -Bugarski1, Drago Blagojevi 2, or e i a3, Branislav Sredanovi 4Summary: The method presented in this paper is designed to locate the damage inthe structure, using data from the frequency response function. The method is basedon the Generalized Damage Index. Functioning of method is demonstrated bynumerical example, where freely supported beam was analyzed.The experimentalexamples were performed for two beam like structures: freely supported beam andcantilever beam.The main advantages of the method are that it is not necessary toundertake a modal identification, there is no need for any analytical or numerical modelof the structure, and it uses all measured data, in form of frequency responsefunctions, without further treatment.Keywords: mechanical structure, damage location, modal analysis, frequencyresponse functions.1 INTRODUCTION Structural damage can be defined as changes introduced into a system thatadversely affect the current or future performance of that system. If one is focused onthe study of damage identification in structural and mechanical systems, than damagemay also be defined as any deviation in the structure’s original geometric or materialproperties that may cause undesirable stresses, displacements or vibrations on thestructure. These weaknesses and deviations may be due to: cracks, loose bolts,broken welds, corrosion, fatigue, etc. All of them should cause a decrease in thestructure’s stiffness, and some will also affect its mass and damping properties.Therefore, structural damages should always, at a sufficient level of severity, cause achange in a structure’s vibration behavior, described by modal properties: naturalfrequencies, damping loss factor and mode shapes. Since the changes on the dynamiccharacteristics can be measured and studied, it is possible to trace what structuralchanges have caused the dynamic characteristic to change, thus identifying damage. Itis obvious that concept of damage detection is based on comparison between two1 Doc. dr Valentina Golubovi -Bugarski, Banja Luka, Mašinski fakultet Banja Luka, (valentina@urc.rs.ba )2 Prof. dr Drago Blagojevi , Mašinski fakultet Banja Luka, (dbl@urc.rs.ba )3 Doc. dr or e i a, Mašinski fakultet Banja Luka, (djordje@urc.rs.ba )4 Branislav Sredanovi , Mašinski fakultet Banja Luka 129
  • Valentina Golubovi -Bugarski, Drago Blagojevi , or e i a, Branislav Sredanovidifferent states of the system, one of which is assumed to represent the initial andoften undamaged state. The structural damage detection can be divided into fourlevels, based on the amount of information provided regarding the damage state.According to Rytter [1] these four levels are: 1) identification of damage existence in astructure; 2) identification of damage existence in a structure and location of damage;3) identification of damage existence in a structure, location of damage andquantification of damage severity; 4) identification of damage existence in a structure,location of damage and quantification of damage severity and prediction of theremaining service life of the structure.2 DETECTION OF DAMAGE LOCATION In practical damage evaluation it is difficult to excite the structural modes(frequency) of a higher order, because the excitation of higher modes needs entering ahigher level of excitation energy, which is technically hardly feasible. Therefore, it is ofinterest to study the techniques of evaluation of structural damage that require only afew basic mode shapes and/or modal frequency. Such methods usually do not requirean analytical model of the structure, but only a few modal frequencies and modalshapes that meet state structures before and after the occurrence of damage, whichcan be obtained by dynamic tests. Sampaio and Maia [2] present such method basedon the Detection and Relative Damage Quantification indicator concerning thedetection and the relative severity of damage, as well as the Generalized DamageIndex for localization of damage. This method belongs to the class of methods that usethe change in the frequency response functions to detect, locate and relatively quantifythe damage. The main advantages of the method are: 1) it is not necessary to performmodal identification; 2) there is no need for any analytical or numerical model of thestructure; 3) it uses all measured data in the form of frequency response functions,without further treatment. Theoretical basis of the method is derived from the equation of motion of amultiple-degree of freedom system with hysteretical damping, which is often used indescribing of complex structure’s dynamics [3]. For such system, the systemreceptance matrix, containing all inforamtion about the dynamic characteristic of thesystem is given by: 2 1 K iD M (1) Each element jk ( ) of the matrix corresponds to an individual frequencyresponse function, FRF, describing the relation between the response at a particularcoordinate j and a single force excitation applied at coordinate k: Xj (2) jk ( ) Fi 0, i 1...N ; i k Fk The column vector, k, of the receptance matrix, k , describes the shape(in space) exhibited by the structure at each excitation frequency , given by the130
  • Detection of Structural Damage Location Using Frequency Response Function Dataresponses normalized by the applied forces. When a structure is damaged its stiffnessand damping change and, in consequence, so does the receptance matrix: d d 1 (3) ( ) K idD 2 Mwhere the superscript d stands for damaged. It is reasonable to assume that thesmaller the degree of correlation between the column vectors, k ( ) and d k ( ) , the larger the damage. The Generalized Damage Index, GDI, can be applied to the localization stagein damage detection, that is level 2 according to Rytter. To localize a damage of thestructure, a comparison between shape vectors should be done. The GeneralizedDamage index, based on the coefficient i(f) calculated from the curvature of thefrequency response function for each location p, the frequency fi and the level ofdamage d, is defined as : N N d d d d i f i f i f i f i f i f i 1 i 1 d p, f N (4) N d d i f i f i f i f i f i f i 1 i 1where represents the conjugate operator and N is the total number of co-ordinatesor measuring points. The second derivate of the FRFs can be obtained by a centraldifference approximation: i f i 1 f 2 i f i 1 f , i=p=(2... N-1) (5) As mentioned above, it is necessary to measure the FRF functions for twodifferent states of structure, one of which is assumed to represent the initial and oftenundamaged state, while the other state corresponds to a level of damage “d”.3 NUMERICAL EXAMPLE To show the effectiveness of the proposed method, the model of a free beamis used, made by the finite element software package ANSYS [4]. The model is madeof 160 beam elements, with 161 nodes and two degrees of freedom per each node, (y, x). Beam dimensions are 400 10 10 mm (L b h), Young s modulus is 2.1x1010N/m2, the density is 7820 kg/m3, Poissons ratio 0.3. The first four modal frequenciesand corresponding mode shapes of beam are obtained by modal analysis in thefrequency range of 1-3000 Hz, with frequency resolution of 1 Hz. The harmonicanalysis is done applying a random excitation of unit force at the node 30, while theresponse to the given excitation is measured in every fourth node, that is in a total for40 nodes the response is determined to a given excitation of structure, figure 1a. Thedamage is simulated by reduction of the cross-section of 64-68th beam element, figure 131
  • Valentina Golubovi -Bugarski, Drago Blagojevi , or e i a, Branislav Sredanovi1b. Six levels of damage are simulated: do is the undamaged beam, while the levels ofd1-d5 are damaged beam by 10%, 20%, 30%, 40% and 50% reduction of cross-section, respectively. The harmonic analysis procedure is repeated for each level ofdamage. Using routines of Matlab software, the Generalized Damage Index iscalculated from the equation (4), for each of the six levels of damage d = (0,1,2,3,4,5),four modal frequencies f = (f1, f2, f3, f4) and location p = (1,..,N-1). The calculatedvalues are graphically interpreted at figure 2, overlapping curves of GDI index for all sixlevel of damage. Force Damaged elements d0 d1 d2 d3 d4 d5 30 64 Fig. 1 a) harmonic analysis of the beam, b) six levels of damage From the graphic display of GDI it is obvious that the Generalized DamageIndex reliably indicates the location of 16 and 17, which correspond to the damagedelements 64-68. Also may be noted that value of the GDI increases with increasing thelevel of damage for the location of damaged elements, so that the GDI givesinformation about the relative increase of damage. Indikator ošte enja GDI za pet nivoa ošte enja GDI 1.03 1.02 GDI_1 1.01 GDI_2 GDI 64 68 GDI_3 1 GDI_4 0.99 GDI_5 0.98 1 4 7 10 13 16 19 22 25 28 31 34 37 40 LokacijaLOCATION vora) (broj Fig. 2 The GDI indicates the location of damage for each level of damage4 EXPERIMENTAL EXAMPLES For the experimental analysis we used two steel beams of dimensions 40010 10 mm [4]. One beam was suspended by common strings to simulate free-freeconditions; the other was clamped to simulate the cantilever beam, figure 3. On thefreely supported beam there were 19 uniformly arranged measuring points, while onthe cantilever beam there were 15 measuring points. The damage was simulatedintroducing a cut at a certain location of the beam, figure 4. The propagation ofdamage was simulated by deepening the cut for 1 mm at each of 7 (5) levels ofdamage for each beam.132
  • Detection of Structural Damage Location Using Frequency Response Function Data Fig. 3 Measuring equipment and beam boundary conditions: a) the freely supported beam, b) the cantilever beam The test beams were excited by modal hammer (type Endevco 2302-10) ineach of 19 (15) nodes. Accelerometer (Bruel & Kjaer type 4507), which caught thesignal response, was set as showed at figure 4. Excitation and response signals weremeasured by the measuring system PULSE (Bruel & Kjaer type 3560 C) in thefrequency range 0 3200 Hz, and processed in Pulse Labshop 9.0 software. Modaltest was repeated for each level of damage. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 14 13 12 11 10 9 8 7 6 5 4 3 2 1Fig. 4 Models of test beams with position of cut, measuring points and accelerometers: a) freely supported beam, b) cantilever beam After 10 modal tests of the freely supported beam had been done, all FRFsmeasured at location 5 were overlapped, figure 5a. The cantilever beam was modaltested for 7 times, and the overlapped FRFs measured for location 11 are showed atfigure 5b. It can be noted that there is some frequency shift due to increasing of thedamage, that is frequencies move to the left (decreasing) due to of decreasing of thestiffness of the beam (when level of damage increasing). For the purpose to locate the damage, the Cumulative GDI was calculated bysuccessive adding the values of GDI (eq. 4), for the each level of damage. Figure 6ashows cumulative GDI indicating the location of damage between measurementlocation 14 and 15 of free-free beam. Figure 6b shows cumulative GDI indicating thelocation of damage between measurement location 4 and 5 of cantilever beam. 133
  • Valentina Golubovi -Bugarski, Drago Blagojevi , or e i a, Branislav Sredanovi [(m/s²)/N] Frequency Response H1(odgovor,pobuda) - File (Magnitude) [(m/s²)/N] Frequency Response H1(odgovor,pobuda) -50% (Magnitude) D:ValentinaDOKTORATG1G1 bo1Frequency Response H1(odgovor,pobuda) - Mark OVERLOAD ValentinaDOKTORATU k3k3 50%Frequency Response H1(odgovor,pobuda) - OVERLOAD 10k 1k 1k 100 100 10 10 1 1 100m 0 400 800 1.2k 1.6k 2k 2.4k 2.8k 3.2k 0 400 800 1.2k 1.6k 2k 2.4k 2.8k 3.2k [Hz] [Hz] Fig. 5 Overlapped FRFs: a) free-free beam, b) cantilever beam G2 K3 suma sukcesivnih nivoa ošte enja Indikator GDI-suma sukcesivnih nivoa ošte enja 1.08 4 1.06 1.06 1.04 1.04 15 1.02 14 1 5 GD I 1.02 GDI 0.98 1 0.96 0.94 0.98 0.92 0.96 0.9 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 1 2 3 4 5 6 7 8 9 10 11 12 Lokacija (broj vora) Lokacija Fig. 6 The GDI indicates location of damage: a) free-free beam, b) cantilever beam5 CONCLUSION This paper presents one method for detection of damage location usingfrequency response functions data. The results of numerical example and experimentsshow that the Generalized Damage Index, GDI, is able to detect damage location onthe structure. As certain measurement inaccuracies happened on the certain locationsduring testing it could be averaged by successive adding the values of GDI for theeach level of damage. However, it is supposed that GDI should increase continuouslyon the location on damage. So, some improvement of the GDI indicator is proposedthat is the Cumulative GDI. Proposed damage detection method showed goodperformance even for the hammer excitation and one response transducer available.LITERATURE[1] Rytter, A. (1993). Vibration based inspection of civil engineering structures, Ph.D. Dissertation, Department of Building Technology and Structural Engineering, AalborgUniversity, Denamrk.[2] Sampaio, R.P.C., Maia, N.M.M. (2008). Strategies for an efficient indicator of structural damage, Mechanical system and Signal processing, Vol.22.[3] Maia, N.M.M, Silva, J.M.M. et al. (1997). Theoretical and Experimental modal Analysis, Research studies Press Ltd.[4] Golubovi -Bugarski, V. (2010). Models of correlation between structural damage and dynamics response of mechanical structure, Ph.D. Dissertation, University of Banja Luka, B&H.134
  • DETERMINATION OF RESIDUAL STRESSES IN TUBULAR WELDED STRUCTURAL COMPONENTS Dragi Stamenkovi 1, Mato Peri 2Summary: Numerical simulation of welding process with an example of two tubes ofboth same and different steel materials properties have been carried out in this paper.It is shown that the residual stress field and temperature field are symmetric in case ofwelding same materials, while in case of welding materials with different thermo-mechanical properties they are asymmetric. So, it can be concluded that besidethermo-mechanical properties of material, pipe diameter and pipe wall thickness canalso influence the asymmetry of residual stress and temperature field.Key words: Welding process, Residual stress, Dissimilar butt-weld, Finite elementmethods, Abaqus1. INTRODUCTION Residual stresses and distortion are common unavoidable adverseconsequences of welding process that appear due to localized heating and subsequentrapid cooling material. High residual stresses in regions near the weld can causefatigue, stress or fracture corrosion cracking. Therefore, estimating the magnitude anddistribution of welding residual stresses is of great importance. In the early days of exploration of residual stresses and deformations in thewelded structures, we relied on experimental methods: X-ray diffraction [1, 2], holedrilling [3], ultrasonic [4] and cracking [5]. In order to estimate residual stresses, in thelast 30 years, finite element method has been used, by which is tried to substitute arather expensive experiment [6-15]. Analytical methods remained limited to simplerproblems. On the example of welding two tubes with the same material properties,Teng [6] proved that pipe diameter and wall thickness influence the size anddistribution of residual stresses in welding of steel pipe. The welding of pipes and plates of steel with different thermo-mechanicalproperties is used in the case of constructing power plants and ships, which can bevery well simulated using numerical methods. In the example of welding two tubes ofsize ø219 x 8.18 mm, and with the ‘’V’’ shape of weld groove, made from carbon andalloy steel, Akbari [7] showed that the temperature fields and residual stress fields areasymmetric.1 MSc, Dragi Stamenkovi , Belgrade, Termoelektro d.o.o., dragi33@gmail.com2 MSc, Mato Peri , Zagreb, Bureau of Energetics and Mechanical Engineering, mato.peric@bestprojekt.hr 135
  • Dragi Stamenkovi , Mato Peri This paper presents two numerical simulations of welding two steel pipes ofdimensions ø324.0 x 3.96 mm with the constant width of the weld of 3.84 mmthroughout the whole tube wall, model A and model B. In the model A the welding oftwo pipes of the same carbonate steel was simulated, whilst in the model B the weldingof tubes of carbon and stainless steel is stimulated.2. FINITE ELEMENT ANALYSIS APPROACH The numerical simulations performed in this work use sequential thermal-stress solution procedure in which the transient heat transfer analysis is followed bythe thermal stress analysis. Temperatures predicted by the heat transfer analysis areused as the loading for thermal stress analysis.2.1.1. Thermal model The governing equation for transient non-linear heat transfer analysis is: T T T T kx ky kz Q C (1) x x y y z z twhere, k x , k y , k z thermal conductivity in the x , y and z respectively, T is thecurrent temperature, Q is the heat generation, is the density, C is the specificheat capacity and t is the time, respectively. General solution of equation (1) isobtained introducing the initial and boundary conditions, as follows: Initial condition: T ( x, y, z ,0) T0 ( x, y, z ) (2) Boundary conditions: T T T kx Nx ky N y kz qs hc (T T ) hr (T Tr ) 0 (3) x y zwhere, N x , N y , N z are the direction cosine normal to the boundary, hc , hr are theconvection and radiation heat transfer coefficients respectively, qs is the boundaryheat flux and Tr is the temperature of radiation heat source and T are thesurrounding temperature. Radiation heat losses are dominant near the weld and canbe expressed by equations: 2hr F (T 2 Tr )(T Tr ) (4) 8 2 4where, 5.67 10 J/(m K ) is the Steffan-Boltzman constant, is the effectiveemissivity and F is the configuration factor. Moving away from the weld zone radiationheat losses decrease and increase the share of convection heat loses hc . The totalheat input is given by: UI q (5) VH136
  • Determination of residual stresses in tubular welded structural componentswhere, the arc efficiency, V is volume of the weld bead, I is current, U is the arcvoltage and VH is volume of the activated weld bead element.2.2. Mechanical model The equilibrium equation can be expressed as: ij bi 0 (6) ij ji (7) In equations (6) and (7) ij is the stress tensor, is density and bi is thebody force. In numerical simulation, it was used thermal-plastic constitutive equations: d D ep d C th dT (8) d D ep d C th dT (9) e p thwhere, D is the elastic stiffness matrix, D is the plastic stiffnes matrix, C is thethermal stiffness matrix, d is the strain increment and dT is the temperatureincrement.3. BUTT-WELDED OF TWO PIPES Two numerical simulations of welding were performed, model A and model B.In the model A the welding of two pipes of the same carbon steel, A 106-B, issimulated. In the model B the welding of tubes of carbon steel, A 106-B and stainlesssteel, A240 TP304, is stimulated. Diameter of tubes are ø324 mm, thickness of wall3.96 mm and lenght of each tube 200 mm. Mechanical and thermal properties aregiven in Figure 1, 2, 3 and 4. The material is modelled as an elastic – ideally plastic.Welding parameters chosen for this analysis are as follows: tungsten inert gas welding,welding current I=110 A, welding voltage U=20 V and welding speed v=5 mm/s. Thefollowing values are assumed: the convective heat transfer hc=15 W/m2K, the arcefficiency =70% and the emissivity = 0.8. Welding of pipes is modelled in singlepass. The specific heat flux input is q 2.076 1010 J/m3s. Width of welding bead is3.84 mm and constant throughout the whole tube wall. The analyses are made withstandard method for modelling of weld. Three-dimensional mesh consists of 14400elements. The same mesh is used both for thermal and mechanical analysis. Inthermal model DC3D8 elements are used, while for mechanical model C3D8 elementsare used, according to ABAQUS code [15]. The smallest elements whose size is 5 x1.92 x 1.32 mm are used at the root of the welded joint. Geometry of welded pipes andmechanical boundary conditions are shown in Figure 5. 137
  • Dragi Stamenkovi , Mato Peri 1,8 1,6 Thermal conductivity 1,4 Density Specific heat Thermal Properties 1,2 C (J/g/°C) 1 -2 3 (10 g/mm ) 0,8 0,6 0,4 (10-1J/mm/s/°C) 0,2 0 0 200 400 600 800 1000 1200 1400 1600 Temperature (°C) Fig. 1 Thermal properties of steel A106-B 300 Youngs Modulus (GPa) 250 Yield Stress (MPa) Thermal Expansion Coefficient Mechanical Properties Poisson ratio 200 (10-7m/m/K) 150 100 50 10-3 0 0 200 400 600 800 1000 1200 1400 1600 Temperature (°C) Fig. 2 Mechanical properties of steel A106-B 1,2 Thermal conductivity 1 Specific heat Density (10-2g/mm3) Thermal Properties 0,8 C (J/g/°C) 0,6 0,4 (10-1J/mm/s/°C) 0,2 0 0 200 400 600 800 1000 1200 1400 1600 Temperature (°C) Fig. 3 Thermal properties of steel A240 TP304138
  • Determination of residual stresses in tubular welded structural components 300 Youngs Modulus (GPa) Yield Stress (MPa) 250 Thermal Expansion Coefficient Poisson ratio Mechanical Properties 200 (10-7m/m/K) 150 100 50 10-3 0 0 200 400 600 800 1000 1200 1400 1600 Temperature (°C) Fig. 4 Mechanical properties of steel A240 TP304 400 200 200 y O324 x 3.96 x weld z A240 TP304 A106-B 0° start of welding 90° 270° 180° mechanical restraint in x, y and z direction mechanical restraint in y and z direction mechanical restraint in y directionFig. 5 Geometry of welded pipes and mechanical boundary conditions 139
  • Dragi Stamenkovi , Mato Peri4. RESULTS Temperature field 50 s after welding for model A and model B are shown inFigures 6 and 7. Comparison of temperature profiles in z direction for model A andmodel B, 50 s after welding is shown in Figures 8. Temperature profiles, shown inFigures 8, are the same at inner and outer surface of tubes for Model A and Model B.In Figures 9 and 10 are shown residual stress field , z , in z direction for Model A iModel B. In Figures 11 and 12 are shown comparison of residual stresses in zdirection at inner and outer surfaces for =180°, Model A and Model B. Fig. 6 Temperature field 50 s after welding, Model A Fig. 7 Temperature field 50 s after welding, Model B140
  • Determination of residual stresses in tubular welded structural components 120 100 Model A 80 Model B Temperature (°C) 60 40 A240 A106-B 20 0 0 50 100 150 200 250 300 350 400 -20 Z - Coordinate (mm)Fig. 8 Comparison of temperature profiles in z direction, =180°, Model A and Model B Fig. 9 Residual stress in z in z direction, Model A Fig. 10 Residual stress in z in z direction, Model B 141
  • Dragi Stamenkovi , Mato Peri 250 200 Model A 150 Model B Temperature (°C) 100 50 A240 A106-B 0 0 50 100 150 200 250 300 350 400 -50 -100 Z - Coordinate (mm)Fig. 11 Comparison of residual stresses z in z direction, at inner surface for =180°, Model A and Model B 100 Model A 50 Model B 0 Temperature (°C) 0 50 100 150 200 250 300 350 400 -50 A240 A106-B -100 -150 -200 Z - Coordinate (mm)Fig. 12 Comparison of residual stresses z in z direction, at outer surface for =180°, Model A and Model B5. DISCUSSION AND CONCLUSION Numerical simulation of welding two tubes was conducted, carbon steel A-106B, A model, and after this the numerical simulation of welding two tubes made fromtwo different materials carbon steel A106-B and stainless steel A-240 TP304, model B.The weld of the same width across the whole thickness of the tube wall was assumed.Weld of temperature field of such a form becomes uniform along the thickness of thetube at any point of time. Model A: the symmetry of the temperature field was confirmed (Figures 6 and 8),after this stress fields (Figures 9, 11 and 12) which coincides with the results in Ref [7]. Model B: the asymmetry of the temperature field was confirmed (Figures 7 i 8),after this stress fields (Figures 10, 11 and 12). Maximum tensile and compressive strain in z direction in welded tubes madeof carbon steel and stainless steel, at outer and inner surface of the tube are morebalanced than in Ref [7], which could be attributed to the uniformity of temperature fieldalong the thickness of the wall because of the constant width of the weld tube wallthickness.142
  • Determination of residual stresses in tubular welded structural componentsLITERATURE[1] Norton JH, Rosenthal D. Stress measurement by X-ray diffraction. Proceedings of the Society for Experimental Stress Analysis 1944;1(2):73-76.[2] Norton JH, Rosental D. Recent contribitions to the X-ray method in the field of stress analysis. Proceedings of the Society for Experimental Stress Analysis 1947:5(1):71-77.[3] Pang HL, Pukas SR.Residual Stress Measurements in a crusi-form welded joint using hole drilling and strain gauges. Strain 1989; February:7-14.[4] Chu SL, Peukhrt H, Schnieder E. Residual stress in a welded steeel plate and their measurements using ultrasonic techniques. MRL Bull. Res. Dev., 1987;1(2):45-50.[5] Masubuchi K, Martin DC. Invesigation of residual stresses by use of hydrogen cracking. Welding J., 1961;40:553s-563s.[6] Teng TL, Chang PH. Three-dimensional thermomechanical analysis of circumferentiallly welded thin-waled pipes. International Journal of Pressure Vessels and Piping 75 (1998) 237-247.[7] Akbari D., Sattari-Far I. Effect of welding heat input on residual stresses in butt- weld of dissimilar pipe joints. International Journal of Pressure Vessels and Piping 86 (2009) 769-776.[8] Deng D. FEM prediction of welding residual stress and distortion in carbon steel considering phase transformation effects. Materials and Design 30 (2009) 359- 366.[9] Deng D, Liang W, Murakawa H. Determination of welding deformation in fillet – welded joint by menas of numerical simulation and comparison with experimental measurements. Journal of Materials Processing Technology 183 (2007) 219-225.[10] Deng D, Murakawa H., Liang W., Numerical simulation of welding distortion in large structures. Computer methods in applied mechanics and engineering 196 (2007) 4613-4627.[11] Lee CH, Chang KH, Three-dimensional finite element simulation of residual stresses in circumferential welds of steel pipe diameter effects, Material Science and Engineering A (2008) 210-218[12] Peric M., Tonkovic Z., Karsaj I.: Numerical analysis ofresidual stresses using a shell/3D modeling technique, Proceedings of the International Conference on Advances in Welding Science and Technology for Construction, Energy and Transportation AWST-2010, pp.75-80[13] Stamenkovic D., Vasovic I.: Finite Element Analysis of Residual Stresses in Butt Welding Two Similar Plates, Scientific Technical Review, ISSN 1820-0206, 2009,Vol. LIX, No.1, pp.57-60.[14] Peric M., Stamenkovic D., Milkovic V.: Comparison of residual stresses in butt- welded plates using software packages abaqus and ansys, Scientific Technical Review , ISSN 1820-0206, 2010,Vol.60,No.3-4, pp.22-26[15] ABAQUS/Standard, Users guide and theoretical manual, Version 6.8, Hibbit, Karlsson & Serensen, Inc. (2008). 143
  • EFFECT OF MATERIAL PROPERTIES OF MEASURING FORCE TRANSDUCER ELASTIC ELEMENTS TO METROLOGIY CHARACTERISTICS Živko Pejašinovi 1, Zorana Tanasi 2, Goran Janji 3Abstract: An influence of caracteristics of the material of elastic element on metrologiccaracteristics of measuring force transducer is analysed in this paper. The mostimportant caracteristics of the materials and these elements are included and theirsignificance in fulfillment of the general functions and metrologic caracteristics ofmeasuring transducer are pointed out. Results of the analysis are the basis for choiceof materials and for design of the optimal shape of these elements, and consequentlymake this process much faster and more efficient.Keywords: force tranducer, elastic element, material of elastic element.1. INTRODUCTION Measuring force transducers designed on the principle of strain gauge are theserial connection of three transducer elements: elastic element, gauge and measuringbridge. F Elastic Strain R Measuring Um element gauge bridge Fig. 1 Structural scheme of measuring force transducer Elastic elements are first-primar transducing elements and the most importantmechanical parts of measuring force transducer. Input values of elastic elements areforces (or torque, pressure etc.), and output values are displacements (linear orangular) or strains. Elastic elements are suitably shaped rigid bodies (rod, beam, ring etc.) on whichsurfaces gauges are applied and connected to the respective bridging configuration.Their function is to act as a reaction to measured force. In such activities elastic elementsare deformed along with applied gauges, electrical resistance of gauges changes, and in1 Živko Pejašinovi Ph.D. Assistant Prof., Banja Luka, Faculty of Mechanical Engineering, (zivkop@teol.net)2 Zorana Tanasi Ph.D., Banja Luka, Faculty of Mechanical Engineering, (tami9997@teol.net)3 Goran Janji MSc., Banja Luka, Faculty of Mechanical Engineering, (gjanjic@urc.rs.ba) 145
  • Živko Pejašinovi , Zorana Tanasi , Goran Janjimeasuring diagonal of the bridge occur the changes equivalent to the measuring force.Tensions, i.e. strains of elastic element must remain purely elastic, that is within theHooke’s law, and strong enaugh to provide an appropriate output signal. Their main caracteristics are: linearity between forces and strains, smallhisteresis, small creep, small tension relaxations etc. Achieving these characteristicsmainly depend on the structural design of elastic elements and properties of thematerial from which it is made. These caracteristics of elastic element which have the greatest influence onmetrologic caracteristics of measuring transducer are analysed in this paper. Analysisincludes mechanical and thermal characteristics and features related to the workabilityof the material of elastic element.2. MECHANICAL CHARACTERISTICS OF MATERIAL The most important mechanical caracteristic which predominantly affects theaccuracy or measuring force transducer is linear relationship between tension andstrain in area of material elasticity of elastic element. According to Hookes lawrelationship between tension and strain in this area is determined with expression: E (1)where is: - tension in material, E - module of material elasticity, - relative strain. Linearity of expression 1 depends on material elasticity module of elasticelement. For the stress range corresponding to the rated load of elastic element,material should have perfect linearity between stress and deformation. In addition tomaterial of elastic element, also other influences can cause the occurrence ofnonlinearity at the output of the transducer. In general, such materials that have stablecharacteristics in a wide range of factors influencing changes are needed for elasticelements construction. Such materials are rare in practice. Their characteristics to agreater or lesser extent depend on the rate of change and the size of workload, time ofactivity, vibration, temperature, etc. Under the load, elastic element is deformed instantly, and then thedeformations continue to increase under the law of damping (Fig. 2a). After unloading,the deformations of elastic element disappear, also with a time delay. Thisphenomenon is called the material creep (with delayed elastic effect) and depends onthe speed of change of measured values. When loads are changing rapidly creeping isnot so prominent as in a short period of time fails to develop. The highest creep occurswith medium frequency loads 20-200 Hz [6]. It is usually not manifested individually,but together with mechanical hysteresis, resulting increase of hysteresis loops. It isespecially expressed at elevated temperatures. Discrepancy between the curves depending on the stress and strain, obtainedby the loading and unloading elastic elements, creates a hysteresis loop (Fig. 2b).Width of hysteresis loop of elastic elements for stresses in the range of rated load islow. It increases with increasing stress so behaviour of material of elastic elementwhen it is loaded above the rated load must be considered.146
  • Effect of material properties of measuring force transducer elastic elements to metrologiy characteristics b a b 0 t 0 a. a b. Fig. 2 Diagrams: a) creep, b) elastic hysteresis The mechanical and thermal treatment of elastic elements should be strictlyadhere to the prescribed treatment regime. Long and over-heating of the elasticelement leads to local changes in mechanical properties as reflected by the increase inhysteresis. In some cases, because of the importance of hysteresis, size of theallowable stresses of elastic elements are not determined by the factor of safetycompared to the material yield strength, but based on the size of allowable hysteresis.This applies to the elastic elements for high-class accuracy, and allowed stresses donot exceed 0.1 02 [6]. Mechanical characteristics of materials such as the limit of proportionality,elastic limit, yield strength and tensile strength is not related directly to thecharacteristics of elastic elements. As the limits of proportionality and flexibility areconsidered as conditional borders, allowed level of work load (and thus the maximumsize of mechanical signals) is determined by the limit of the material flow. Safetyfactors for this limit are chosen in the range 0.3 to 0.5 (for elastic elements of thehigher class of accuracy to 0.1) [6]. The nominal load of elastic element that representsthe upper limit of measurement range is defined in this way. Elastic elements for forcemeasuring are required to withstand much heavier loads than the nominal. Theseloads are called safe or secured load limits. The material must not happen any change(plastic deformation, cracks) to the boundary of the safe load. The value of the safeload limit depends on the type of elastic element and is 150-200% of rated load, or300-500% compared to the limits of destruction of materials [2, 3, 4]. In order toprevent breakage (especially in the areas of stress concentration) the material musthave sufficient flexibility. Therefore, the elastic elements rarely require higher hardnessof 50 HRC, because the materials with higher hardness prone to brittle fracture. If it iscertain that the safe load limit to be exceeded it is necessary to install some protectionagainst overload. For elastic elements dynamicaly loaded with a large number of cycles, it isnecessary to take into consideration fatigue strength and regarding it to select theappropriate safety factor. For most of elastic elements, fatigue strength is of secondaryimportance, since the sensitivity of electro-strain gauges to fatigue is limiting constraint. In some cases of choise of material of elastic element may be limited byspecific requirements for material density, temperature coefficient of linear expansion,the corresponding relationship between elastic modulus and bulk density, etc. 147
  • Živko Pejašinovi , Zorana Tanasi , Goran Janji3. THERMAL PROPERTIES OF MATERIALS Previously analyzed mechanic properties of materials used for making elasticelements are related to standard operating conditions (room temperature withoutaggressive and other influences). As the electro transducers with strain gauges can beused in a wide temperature range and under different external influences, specialattention should be paid to the thermal properties of materials, since at extremetemperatures, range of materials to create elastic elements is substantially limited. Experimental studies indicate that mechanic properties of some metals (yieldstrength and tensile strength) significantly depend on temperature. Figure 3 showsseveral diagrams obtained during the investigation of carbon steel in tension atdifferent temperatures [5]. The diagram shows that the yield strength at highertemperatures is less obvious and above 300ºC it can not be seen in the diagram.According to the same diagrams, steel strength rises to 250ºC and then rapidlydecreases. Napon Tension 400 0C 500 0C 200 0C 250 0C 100 0C 300 0C 20 C 0 Specifi no izduženje specific strain Fig. 3 Diagrams of tensile of carbon steel at different temperatures As these sizes do not relate directly to the elastic elements, more interestingare the first parts of these diagrams. According to these diagrams limit ofproportionality of steel decreases with increasing temperature. At the same time, theslope of the diagram reduces and thus also the modulus of elasticity for 1-3% at 60 ºC[4]. This information is very important because it directly influences to the linearity ofthe materials in the field of elasticity according to the expression 1. Keeping this inmind it is necessary to choose a material for which the elastic modulus in a certaintemperature range is linear and stable enough. In this case, the errors that occur dueto changes in elastic modulus can be taken into account by introducing corrections orcompensations, which is done only for elastic elements of higher class accuracy. Another important requirement for the material of elastic element is the stabilityof modulus in time. Research shows that the instability arising in the material after itsmechanic and thermal treatment is associated with the subsequent action andrelaxation. The lowest time-dependent instability modules of elasticity have materialswith no significant residual stress.148
  • Effect of material properties of measuring force transducer elastic elements to metrologiy characteristics By analyzing the thermal properties of the material of elastic elements, oneshould bear in mind that the electro-strain gauge is a source of heat that is transferredto the elastic element. Temperature gradients in the asymmetrical elements, especiallyif they are unbalanced in relation to applied gauge, can cause interference at theoutput. To minimize these temperature effects during the reaction of elastic elements,the elements should be of symmetrical forms with respect to applied gauge. The transducers in use are exposed to a larger or smaller temperaturechanges. The impacts of these changes can be reduced by appropriate constructionalsolutions. In this case, one must take into consideration heat flow path between thecasing of the transducer and the elastic element, and within the element. For anytechnical solution, the change of temperature gradients within the element is inverselyproportional to the thermal conductivity of the material element. Therefore, the thermalconductivity of the material of elastic element is important feature that must be takeninto consideration when selecting materials. The effects of thermal expansion ofmaterials of the first order at exit of elastic element are eliminated with self-compensating gauges and full-bridge configuration. In addition, there are higher-ordereffects arising from changes in the cross section dimensions, torque arm and the otherparameters, and is about 0.1 to 0.2% at 60ºC [4]. The analysis of thermal effects on the material of elastic element and its behaviorin such conditions shows that they are numerous and significantly affect the final results.Thermal effects are successfully solved by applying these recommendations and usingsome existing practical and very efficient methods of compensation of these effects.4. WORKABILITY OF MATERIAL OF ELASTIC ELEMENT For selecting the material of elastic elements, in addition to mechanical andthermal properties, an important properties are related to their workability. In theproduction of elements the principle of monolithism should be respected. In this waythe equality of all influential factors on the characteristics of elastic elements isensured. All connections of any kind applied to the elastic elements, to a greater orlesser degree cause a slight displacement and friction that can cause certain non-linearity. Welded joints require special attention because they create residual stresses,and metallurgical effects contribute to the appearance of micro-plasticity and limitingendurance to fatigue. Fig. 4 Elastic element of the two-component force measuring transducer 149
  • Živko Pejašinovi , Zorana Tanasi , Goran Janji By adjusting the basic, relatively simple shapes of elastic elements to the taskof measuring, very complex configurations of elastic elements are formed (Fig. 4).Keeping this in mind and respecting the previously mentioned principle, a complicatedand precise machining is a common requirement in the manufacture of elasticelements. In this case, the workability of the material is an important feature to betaken into account when selecting materials because it significantly affects theperformance and cost. More materials suitable for elastic elements have good workability in the statebefore heat treatment. The predicted mechanical properties are achieved bysubsequent heat treatment at appropriate temperatures and cooling in oil or water. Inelastic elements of complex shape with segments of various thicknesses, theseprocesses are accompanied by the occurrence of subsequent deformation of thedesigned shape. Materials prone to such changes should be avoided in the selection.5. CONCLUSION In the design of force transducers based on the principle of electro-resistantstrain gauges central place belongs to the design of their elastic elements. Forsuccessful realization of this task, it is necessary to know the material properties ofelastic elements. The choice of material of elastic elements must be compatible withcommercial by type, quantity and usable form, and sometimes price. None of thematerials to create elastic element does not possess all necessary features so that thechoice represents a compromise between the existing properties. In the particularcase, the optimal material should be selected according to the specific priority offeatures. Summarizing the previously performed analysis it should be noted that any kindof imperfection in the chosen material of elastic element has negative effect on themetrological characteristics of the transducer in which an elastic element is embedded.LITERATURE[1] Pejašinovi Ž., Prilog optimalnom oblikovanju elasti nih elemenata mjernih pretvara a sile u cilju poboljšanja metroloških karakteristika, Doktorska disertacija, Mašinski fakultet Banja Luka, 2005.[2] Gommola G., The Application and Installation of Load Cells, Spectris Messtechnik GmbH. – Frankfurt/Main, Zarbock, 2000.[3] Hoffmann K., An Introduction to Measurements using Strain Gages, Hottinger- Baldwin Messtechnik GmbH, Darmstadt,1989.[4] Measurements Group, Strain Gage Based Transducers – Their Design and Construction, The Technical Staff of Measurements Group, Inc – North Carolina, Releigh, 1988.[5] Timošenko S., Otpornost materijala II dio, Gra evinska knjiga, Beograd, 1966.[6] Makarov R. A., Tenzometrija v mašinostroenii, Mašinostroenie, Moskva, 1975.150
  • EVALUATION OF CONCEPTUAL SOLUTIONS OF UNIVERSAL HELICAL TWO STAGE GEAR UNITS Siniša Kuzmanovi 1, Milan Rackov2Summary: Universal gear reducers are very simple and quite improved mechanisms,but despite of that their further development is constantly made in order to find bettersolution. On the basis of the present solutions it can be note that almost everymanufacturer has its own conceptual solution of gear reducer. Different conceptualsolutions of universal gear reducers do not make a problem for customers because it isnot the point of their interest. They are only important to have quality gear unit,primarily a strong, reliable, durable, with low-vibration and noise and that is cheap, alsothey want short delivery time, low maintaining costs and the repair time is short.However, manufacturers of gear are very important what kind of concept solution isadopted and therefore it is necessary to evaluate some conceptual solutions during thetime of their development in order to adopt the best. This paper gives the evaluation ofparticular solutions in order to realize the degree of goodness of some solutions.Key words: evaluation, conceptual solution, universal gear unit, reducer1. INTRODUCTION Universal helical gear reducers are the mechanisms which have very wideapplication in mechanical engineering. They are regularly bought as finalized productand none of their costumers and users do not think about their manufacturing.However, manufacturers of gear reducers are continuously working to improve thetechnical characteristics of their products in order to achieve certain advantages at themarket. This paper deals with the affect of arrangement of pinions, gear wheels andbearings inside of gear unit on the technical characteristics of universal helical gearreducers. Universal gear reducers are usually made as single, two, three stage andmultistage gear units. Some manufacturers do not produce single stage gear unitsbecause they are not often required, and the same case is with multistage gearreducers. Multistage gear units are usually made by combining of two and three stageunits, although there are manufacturers which make only multistage gear units inspecial large housing with several gear pairs, but it is rare case. Universal reducers can be made as geared motors with special reducer electric1 Prof. DSc., Siniša Kuzmanovi , Novi Sad, Univ. of Novi Sad, Faculty of Technical Sciences, (kuzman@uns.ac.rs)2 Ass. MSc., Milan Rackov, Novi Sad, Univ. of Novi Sad, Faculty of Technical Sciences, (racmil@uns.ac.rs) 151
  • Siniša Kuzmanovi , Milan Rackovmotor or with standard IEC motor and they can be made without installed motor withfree input shaft or with IEC motors interface. Gear reducer without installed motor is now very little used because of morecomplex structures and difficult installation. It usually need another coupling (forconnecting motor and reducer), larger space for installation (because of those couplingapplication), and there are additional difficulties with aligning input shaft of reducer andmotor shaft. However, manufacturers of gear reducers (due to market competition)must also have those reducers in their offer.2. CONCEPTUAL SOLUTIONS OF GEARS AND BEARINGS ARRANGEMENT There are different practical conceptual solutions of gear unit design despitethe small possibilities of variation, arrangement and position of gears and bearingswithin the gear reducer housing. These possibilities are the smallest for the singlestage gear reducers and the following conceptual solutions are possible here, which isshown on Fig.1 and Fig.2. Two stage gear reducers are available as geared motors,Fig.2-1 (with special reducer electric motor or with standard IEC motor), with anadapter for IEC motors, Fig.2-2 (with one or two bearings) and without installed motor,Fig.2-3 (with free input shaft). Geared motors with standard IEC motors are notconsidered here, because most of the manufacturers do not offer this solution, and itwould only slightly expand the number of possible conceptual solutions. In this paperonly two stage gear reducer are analyzed. Fig. 1 Possible conceptual solutions of universal single stage gear reducer (1 – geared motor, 2 –adapter for IEC motor, 3 – adapter with free input shaft)3. PROBLEM INTERPRETATIONS Based on the theory of possible conceptual solutions, it should adopt the bestsolution, which is not so simple. In fact, almost every manufacturer of gear reducer hasa different conceptual solution of gear units, so that it is very difficult to determinewhich solution is best. The output shaft is usually designed that only one bearing accepts axial forcein order to make cheaper solution and make possible easier installation and removal ofthe output shaft assembly. Minimum consumption of material must be considered during design, the lowervolume of machining as possible and sufficient strength and stiffness of the housing,with a simple assembly and removal of gear unit. Additionally, it is necessary to takeinto account the possibility of using the same set of gears in the frame of particular sizeof gear reducer. For example, gear pair from single stage gear unit is the first gear pairof two and three stages gear reducer.152
  • Evaluation of Conceptual Solutions of Universal Helical Two Stage Gear Units 1. Universal two stage gear reducer with special reducer motor installed 2. Universal two stage gear reducer with adapter for IEC motor or with standard IEC motor 3. Universal two stage gear reducer with free input shaft (so-called B interface) Fig. 2 Characteristic conceptual solutions of universal two stage gear reducer Today there is a tendency to achieve a high gear ratio, so it means to use largedriven gear, which resulted in a need to make the frontal opening of the housing (and/orthe top opening for two and three stages gear reducer). It is necessary because thestandard frontal opening (which size depends on the electric motor flange) cannot be usedfor installing so large gears. This concept housing, so called monoblock construction, issomewhat weaker, but their possibility of simple installation gives them a certainadvantage. Large back cover of single stage gear reducer provides completely installationof large gears, so there is no need to make opening on the top of gear unit. Easy assembly and removal of gears must be provided when defining theirarrangement and position with smaller as possible overall dimensions of gear reducer.Also, minimum use of special tools must be enabled because today fast delivery isrequired (usually within 72 hours) and fast repair of gear reducer, which is usually donein service centers near the major markets since fast delivery, service, maintenance andrepair are often a crucial factors for decision to purchase gear reducer. Generally, fastand quality service is an essential support for gear reducer sales, and to ensure thisconstructive gear reducer must be prepared for it.4. THE PROPOSAL OF EVALUATION Evaluation of conceptual solution can be done in many ways by a simplecomparison of so-called partial indicators of quality: the weight of gear reducer (m),gear ratio (u), output torque (T2), etc. Also, complex indicators can be used: the ratio 153
  • Siniša Kuzmanovi , Milan Rackovbetween gear ratio and weight (u/m), ratio between output torque and weight (T2/m),etc. In order to obtain more complete assessment of gear reducer, it is necessary tointroduce more complex indicators, such as uT2/m. It should be noted that this methodof assessment can evaluate only gear units with the same shaft height. In order toobtain a more complete assessment, it is necessary to introduce the shaft height value(h) and to considerate complex indicator (uT2/h) or more complex indicator (uT2/mh),where the value of these indicators should be as higher. Of course, evaluation can be done for gear reducer housings made of samematerial (cast iron). If the materials are different, obtained values will not becomparable, although obtained assessment can refer a lot about the benefits ofindividual solutions. Also, it should note that some manufacturers are oriented to highload capacity (high output torque) and small gear ratio, while others are oriented tosmaller output torque and high gear ratios. Additionally, some manufacturers offer twogear sets in the frame of one housing (or the same shaft height): with small transmittedtorque and high gear ratio and the other with high transmitted torque and small gearratio, so that they can successfully compete with other manufacturers. With thisapproach manufacturers slightly complicate their product range, but certainly achievegreat advantage at the market. In a rough assessment of these solutions it cannotobtain enough accurate evaluation of the goodness of particular solutions, andtherefore it is necessary evaluate assessment for each gear ratio for the same shaftheight and compare it with other manufacturers’ solutions.5. ASSESSMENT OF CHARACTERISTIC SOLUTIONS This paper deals with assessment of technical characteristics of two stageuniversal gear reducers (foot mounted type) of several leading manufacturers of gearboxeswhich names will not be mentioned, so they will be marked as Manufacturer 1 to 6 (Fig.3).Geared motors are not observed, but variant without installed motor with the classical solidinput shaft, in order to make accurately comparison and evaluation of gear units. Man. 1 Man. 2 Man. 3 Man. 4 Man. 5 Man. 6 Fig. 3 Frontal view of universal single stage gear reducer On the basis of analyzed solutions of these six manufacturers, it can beconcluded that the Manufacturers 1, 3 and 6 offer ten sizes of a two stage gearbox (tenshaft heights), but for some shaft heights they offer two gear sets in the frame of onehousing: with small transmitted torque and high gear ratio and the other with hightransmitted torque and small gear ratio. It can be concluded the most required shaftheights are between 90 and 160 mm, so they offer two gear sets and can successfullycompete with other manufacturers. Manufacturer 2 offers eleven sizes of gear units,Manufacturer 4 offers eight and Manufacturer 5 offers seven shaft heights. The values154
  • Evaluation of Conceptual Solutions of Universal Helical Two Stage Gear Unitsof output torque (T2), gear ratio (u), weight (m) and housing volume (V) for each shaftheight are shown in Fig.4. T2 max, u max Nm h, mm h, mm 3m, kg V, m h, mm h, mm Fig. 4 Diagrams of maximal output torque (T2), gear ratio (u), weight of gear reducer (m) and housing volume (V) of universal two stage gear reducers with free input shaft for different axis heights (h), Manufacturers 1 - 6 Analyzing obtained curves, it can be noticed that the increase of output torqueis fairly consistent for all manufacturers and it is especially higher for the Manufacturers1 and 6 for greater shaft heights. Manufacturer 2 and 6 offer the highest gear ratio.Furthermore, it is obvious that reducers of all manufacturers have almost constantvalue of gear ratio, except the Manufacturer 2 which offers high gear ratio for shaftheight until 200 mm and after that it is reduced. Comparing the reducer weight, it canbe observed that the weight is evenly increased for all manufacturers and the reducersof Manufacturer 2 have the highest weight for largest shaft heights. Increasing thevolume is also the same, but for higher shaft heights reducers of Manufacturers 1 and6 have the largest overall dimensions. However, Manufacturer 6 reducers arenonsymmetrical and they have lower volume than it’s calculated. Since gearboxes are not standardized, manufacturers offer a variety of reducershaft heights. Shaft heights of these six manufacturers are not same, except for theheight of 250 mm, which suggests this is the most required shaft height for two stage 155
  • Siniša Kuzmanovi , Milan Rackovgear units. Also, similar shaft height is for 130 mm (125, 130 and 132 mm) and 315mm (300 and 315 mm). T2 max, Nm umax m, kg V, m3Fig. 5 Comparison of output torque (T2), gear ratio (u), weight of gear reducer (m) and housing volume (V) of universal two stage gear reducers with free input shaft for the shaft height h = 250 mm, Manufacturers 1 - 6 Therefore, the initial assessment is taken for shaft height of 250 mm which allsix manufacturers can provide. Comparison of output torques, gear ratios, weights andvolumes for shaft height of 250 mm is given in Fig.5. Comparing the most important technical characteristics of gear reducers withshaft height 250 mm, it is observed that the biggest output torque is provided by gearreducer of Manufacturer 5 (T2max = 5000 Nm). The highest gear ratio is achieved byreducer of Manufacturer 6 (umax = 62.04, three times higher than Manufacturer 3). Gearunit of Manufacturer 5 has the smallest weight (m = 117.1 kg). Gear reducers ofManufacturer 1 and 6 have the largest mass and volume. Gear unit of Manufacturer 2have large weight and the smallest volume, which means the Manufacturer 2 hasbetter arranged and more compact design, but unfortunately with the lowest outputtorque. Therefore, every reducer of these manufacturers has some typicalcharacteristic, so that according to these basic indicators it is impossible to adopt thebest solution. Therefore, complex indicators are used for evaluation, such as the ratio ofoutput torque and weight (T2 max/m) or the ratio of gear ratio and weight (umax/m), Fig.6. Comparing complex indicators of two stage gear units (Fig.6-1,2) of those sixmanufacturers, it can be concluded that Manufacturer 5 has an advantage, because ithas the highest torque per weight unit and gear ratio value per weight unit. After themreducers of Manuf. 3 and 6 have high torque per weight unit, but reducers of Manuf. 6and 2 have high gear ratio per weight unit. However, after this comparison it is notclear which manufacturer has better reducers after Manufacturer 5.156
  • Evaluation of Conceptual Solutions of Universal Helical Two Stage Gear Units It should considerate more complex indicators such as umax T2 max / m to obtaina more complete assessment of the quality of gear reducers, Fig.6-3. Based onanalysis for this indicator for shaft height 250 mm and consideration of achieving ashighest as possible output torque and gear ratio for the minimum mass of gearbox, it isconfirmed that gear units of Manufacturer 5 are the best, and after them there are gearunits of Manufacturer 6 which have the most rational utilization of material, much betterthan other gear reducers. Therefore, it should adopt such kind of gear units withmaximum output torque and gear ratio per mass unit.T2 max / m u max / m u max T2 max / m 1 2 3Fig. 6 Comparison of complex characteristics of universal two stage gear reducers with free input shaft for the shaft height h = 250 mm, Manufacturers 1 - 6 This evaluation has been performed from the viewpoint of design optimization,which means obtained amount of technical parameters based on the designed weightor volume. If considerate the price of gear unit, it would obtain more complexassessment of gear reducer, but since the price is a market category, it will not bediscussed here.u max T2 max / u max T2 max / mh mV h, mm h, mm Fig. 7 Complex indicators of technical characteristics quality of gear reducers for different shaft heights of universal two stage gear reducers with free input shaft 157
  • Siniša Kuzmanovi , Milan Rackov This method of assessment can be used only for gear units with the same shaftheight, so the evaluation for the shaft height of 250 mm cannot guarantee the sameconclusion for other shaft heights. Introducing the shaft height values is necessary toobtain more complex evaluation that will allow comparison of gear units for differentsizes of gearbox. Therefore, it should considerate an indicator u max T2 max /h or evenmore complex u max T2max /mh or u maxT2 max/Vh (Fig. 7). Analyzing obtained indicators, the gear ratio and output torque are in the mostfavorable comparison with the weight and volume for smaller shaft heights (100 - 200mm). Therefore, the Manufacturers 2, 5 and 6 have the best relations of complexindicators until the shaft height 315 mm. Manufacturer 3, which provides the lowestgear ratio, has the lowest values of these complex indicators. After shaft height 355mm, reducers of Manufacturer 1 are leading and the best solution.6. CONCLUSION Based on the implemented evaluation, it can be seen that the weight ofgearbox of certain manufacturers are not uniform. Also, gear ratios and output torqueare different for particular manufacturer (even within the certain sizes of the samemanufacturer). Based on complex indicators, such as the ratio of output torque andweight (T2 /m), or the gear ratio and weight (u/m), we can see a big difference betweendifferent solutions, but reducers of Manufacturer 5 should be highlighted because theyhave the best relations of technical characteristics. In order to obtain a more completeassessment, a complex indicator (u max T2max /m or u maxT2 max/V) is introduced intoconsideration on the basis of which was also confirmed the advantage of Manufacturer5. It should consider that this method of assessment can evaluate only gearboxes withthe same shaft height. Shaft height is introduced into consideration in order to obtainmore complex assessment, so indicators (u max T2max /mh) or (u maxT2 max/Vh) are used.Based on these evaluations, it is concluded that Manufacturers 5, but also 2 and 6have reducers with the best relations of complex indicators, so that this concept shouldbe adopted in the development of new two stage series of gear units. Therefore, itshould adopt such kind of gear units with maximum output torque and gear ratio permass unit. Of course, the price as market characteristic is an additional argument for thepurchase of gear unit, but also the speed of delivery, provided service, spare parts,which is all not discussed in this assessment of most favorable and optimal design.LITERATURE[1] Kuzmanovi , S. (2009) Universal Helical Gear Reducers, University of Novi Sad, Faculty of Technical Sciences, Novi Sad[2] Kuzmanovi , S., Ianici, S., Rackov, M. (2010). Analysis of Typical Method of Connection of Electric Motor and Gear Unit in the Frame of Universal Motor Gear Reducers, Machine Design 2010, Faculty of Technical Sciences, Novi Sad, p. 141-146.[3] Rackov, M., Kuzmanovi , S. (2011). Proposal of Assessment Method for the th Conceptual Design of Universal Helical Gear Reducers, Proceedinga of 7 International Scientific Conference IRMES 2011, University of Niš, Mechanical Engineering Faculty, 27-28 April 2011, Zlatibor, Serbia, p. 469-474.[4] Catalogues: SEW, Nord, Rossi, Bonfiglioli, Motovario, Radicon158
  • FATIGUE ANALYSIS FROM FRACTURE MECHANICS ANGLE Ivica amagi 1, Nemanja Vasi 2, Zijah Burzi 3Summary: Crack propagation influenced by variable loads that are lower than theloading level of quasistatic fracture is called material fatigue. Fatigue is causing thelargest number of damages and failures of parts and structures during exploitation. In anumber of cases damages can be assigned to the material state, but a lot of damagesare consequence of a poor constructive design and more often bad welded joints [1].Accordingly, fatigue tendency of a part of machine, structure or welded joint, does notdepend merely on fatigue strength of the material used for a part fabrication but itsgeometry as well. Therefore, when increasing a machine or structure fatigueendurance is set as a task, it should be taken into consideration that mere selection ofmore durable material for parts fabrication is not sufficient and often is inefficient, andnew constructive solutions presents better alternative. Defects in metals and variousforms of stress concentration (notches, grooves, holes welded and mechanical joints)in elements can’t be avoided. Therefore, period until fatigue crack initiates, Ni, can beneglected, from angle of total fatigue life until crack appearance, Nu. Considering thatmacroscopic crack growth rate can’t be affected by structural alternation, it remains todetermine the crack growth rate in laboratory conditions for the given material ofspecimen or element, and after the crack size determination during occasionalinspection to estimate the remaining life of machine part or structure. Theses of linearelastic fracture mechanics provide the possibility for this estimation.Key words: fatigue, crack, welded joint, fracture mechanics parameters, fracture1. INTRODUCTION The basic progress that fracture mechanics made in the material fatigue scopeis in analytical parse of the fatigue fracture phenomenon on the initiation period, inwhich fatigue crack appears, and the propagation or the growth period that follows andin which crack enlarges to the critical size at which sudden fracture occurs. Thus, totalnumber of cycles, Nt, after which crack occurs, is divided to the number of cyclesneeded for fatigue crack initiation, Ni, and the number of cycles need for itspropagation up to the critical size, Np:1 M.Sci, Ivica amagi , Blace, Faculty of Technical Sciences K. Mitrovica, ivica.camagic@pr.ac.rs2 M.Sc, Nemanja Vasi , Kragujevac, Faculty of Technical Sciences K. Mitrovica, nemanja.vasic@pr.ac.rs3 PhD, Zijah Burzi , Belgrade, Military Institute of Techniques, zijah_burzic@rvkds.net 159
  • Ivica amagi , Nemanja Vasi , Zijah Burzi Nt = Ni + Np . (1) Parallel introduction of experimental and theoretical approach alloweddevelopment in the study of a material behaviour under variable load influence, sincejust theoretical approach cannot fully explain initiation and propagation of the fatiguecrack. Today, all factors that influence the da/dN = f(DK) dependence, in so called low-cycle fatigue when plastic deformation is established in the loop of a single hysteresis,are intensively researched. Stress and strain state analysis, on a tip of a growingfatigue crack, by linear-elastic fracture mechanics (LEFM) procedure have lead to theParis equation formulation for all metals and alloys, connecting the fatigue crackgrowth rate with the stress intensity factor range on a crack tip.2. LINEAR-ELASTIC FRACTURE MECHANICS CONCEPT Crack growth during fatigue is a very complex process that depends on aseries of variables such as: Effective stress field intensity on the crack tip defined by K-factor; Loading type and form; Working environment (aggressiveness, temperature, humidity); Mechanical and metallurgical properties of the metal. Introduction of the fracture mechanics into fatigue behaviour researchoriginated from the crack growth analysis during the cyclic loading. Change of thecrack length, a, is shown in Fig. 1 [4], in terms of cycles number, N, on a three levels ofthe upper stress u ( 1> 2> 3) with the lower stress l=0, where each specimenhad an equal starting crack length a0. It is noticeable that with the incensement of cycles number, N, and cracklength, a, crack growth rate, defined with tangent slope, continuously rises. Also,increase of the stress range , is followed by more rapid velocity gradient increase. Inother words, crack with length a1 shown in Fig. 1 grows more rapidly with stressamplitude 1 compared to 2 and 3 respectively. A number of theoretically and empirically defined dependences in followingform da/dN = f(F, a), can be found in literature, and they emphasize the importance ofloading and crack length. First to define the range of stress intensity factor K = f( , a)as a basic parameter that controls the fatigue crack growth rate as: K = Kmax - Kmin = Y( max - min) ( a)1/2 = Y ( a)1/2, (2)were Paris and associates [2, 3,]. If at the same time min < 0 it is accepted that Kmin =0 because the K-factor does not exist for pressure stresses. Crack growth rates da/dN in terms of the K are determined by graphical ornumerical procedure from an appropriate a-N, t curve. Experimental results presentedon a double-logarithm scale most commonly have specific S-shape, schematicallyrepresented in Fig. 2 [2]. Three regions can be noticed on this curve from the crack growth mechanismand different influential factors intensity points of view. In region I decreasing of the Kis followed by sudden reduction of the crack growth rate. K value at which the ratesare at the order of magnitude 10-10 m/cycle or lower defines the sensitivity threshold ofthe stress intensity factor range – fatigue threshold, Kth [5]. Below Kth fatigue cracks160
  • Fatigue Analysis From Fracture Mechanics Anglebehaves as cracks without growth tendency. In region II the dependence log(da/dN) tolog K is basically linear and it is represented with straight line described by Paris andassociates [2] with the degree function as: da C ( K )m (3) dN 1 > 2 > 3 3 2 Crack length, a Fracture 1 a1 Number of cycles, N Fig. 1 Fatigue crack length dependence on the number of cycles Sudden crack growth, before final fracture, appears in region III. This apparentinstability is connected with the K-factor maximum value, Kmax, approaching the criticalvalue of the fracture toughness, KIc, for given material, which is connected to the earlyphase of brittle fracture. This possibility is expressed within high strength materials,with low fracture toughness, where tensile specimens’ dimensions allow linear-elasticbehaviour at the K-factor levels close to the fracture toughness at plane deformation,KIc, as well. Paris equation application proved to be especially fertile in the domain of weldmetal fatigue and welded structures. Unlike homogenous materials and parts madefrom them, where the number of cycles necessary for fatigue crack initiationdominantly participates in the total number of cycles until fracture, it was noticed soonafter the Paris equation was formulated that with welds the total number of cycles untilfracture is mainly determined by the fatigue crack growth. Following checks confirmedthat the number of cycles needed for the crack initiation within different kinds of weldsand materials does not exceed 25% of the total number of cycles until fracture [5]. Thereason for this lies in geometric unhomogenousity of welds and in a presence of smallenough surface roughness, especially on crossing between raised and base metal, inform of intrusions or inclusions of slag with depth not higher than 0,02-0,04 mm, onwhich initial fatigue crack rapidly forms caused by the stress concentration [5]. Based on the previous analysis of the influence factors on fatigue crack growthrate it can be seen that beside the K range, as the most influential, the crack growth 161
  • Ivica amagi , Nemanja Vasi , Zijah Burzirate is influenced by many mechanical, geometrical and metallurgical factors andenvironment properties. However, till now, a series of empiric, semi empiric or purelytheoretical models for prediction of the crack growth rate was developed, but it isspecific that each of them is applicable in certain K region and for certain metals andtesting conditions. da log dN R>0 I II III R< 0 da m = C ( K) dN K th K c log K Fig. 2 Principle form of the growth rate change da/dN = f( K) for R = 0 and S-curve shift directions for R 0 Paris equation, given in expression 3, where C and m are material constants, isvalid only in the crack growth region II (Fig. 2), and only for the single value of R =Kmax/Kmin ratio. Since crack growth rate lines are nearly parallel for different values of Rin region II [1], coefficient m values will be the same as well, but coefficients C willdiffer, since they depend on the R ratio. That dependence can be simply presented as: da Cv ( K ) mR , (4) dN (1 R ) nvwhere coefficients Cv and nv are determined for the ratio R = 0. Expression 4, with substitution: nv n v K K max , (5) (1 R )was modified by Walker in the following form: da nv n Cv K mR nv K max Cv K mv K max , v (6) dNwhere mv = mR - nv. Forman and associates suggested expression:162
  • Fatigue Analysis From Fracture Mechanics Angle da CF ( K ) mF CF ( K ) mF 1 K max , (7) dN (1 R ) K Ic K K Ic K maxthat beside the R ratio, includes influence of rapid fracture as well, when the Kmax isconverging to the fracture toughness KIc. This kind of behaviour can be found withinhigh strength metals, with low fracture toughness, where sizes of specimens for fatiguecrack growth rate testing allow linear elastic behaviour at the K-factor levels close tothe KIc, as well. Klesnil and Lukas included the R ratio influence on the crack growth behaviourclose to the threshold Kth (region I). That empiric model, valid for the region II as well,has a following form: da ( K ) mK m CK K thK , (8) dN (1 R )where CK and mK are material constants, and is a constant that is dependent on thematerial-environment system. The influence of heterogeneity of structure and mechanical properties can beoverviewed more clearly on the example of butt welded joint (2/3 X-weld) of highstrength low alloyed steel Nionikral-70, welded with Tenacito-75 basic low-hydrogenelectrode in diameters of 3,25 and 4 mm, supplied by Acroni Jesenice. The fatiguecrack growth rate diagrams, da/dN, the stress intensity range change, K, for thesamples tested at the room temperature are given in Fig. 4. Values of the coefficient Cand m, and the stress factor intensity range on a fatigue threshold Kth, are given inTable 1. Fig. 4 da/dN- K dependency diagram for: a) BM, b) WM and c) HAZ specimens 163
  • Ivica amagi , Nemanja Vasi , Zijah BurziTable 1 Paris equation coefficients Growth region Coefficient C Coefficient m Kth, MPa m1/2 -14 BM-1 3,98 10 4,139 10,22 WM-1 3,30 10-19 8,462 8,45 HAZ-1 2,90 10-16 6,403 8,713. CONCLUSION Without a doubt fracture mechanics approaches have large potential forsuccessful application in all cases of welded structures safety evaluation. Brittlefracture tendency in presence of a crack and in variable load conditions should beconnected with slope change of the part of the curve in Paris law validity area. Inparticular example, slower crack growth is confirmed at specimens with a crack in basemetal and heat affected zone, because for the same crack growth rate higher stressintensity factor is requested. Maximal crack growth rate can be expected on a level ofstress intensity factor range witch approaches the fracture toughness at plane strainKIc, since brittle fracture is achieved at such level.LITERATURE[1] Miller, K.J., O Donnell, W.J. (1999). The fatigue limit and its elimination, Fatigue Fracture Engineering Materials Structures, Vol. 22, p. 545-557.[2] Paris, P.C., Erdogan, F. A Critical Analysis of Crack Propagation Laws, Trans. ASME, Journal Basic Eng., Vol. 85, No. 4, p. 528.[3] Paris, P.C., Sih, G.C., Stress Analysis of Cracks, Fracture Toughness Testing and Application, STP 381, p. 30-83.[4] Burzi , Z. (1997). Ispitivanje promenljivim optere enjem glatkih i zarezanih epruveta, 7. Tematski zbornik radova, Eksperimentalne i numeri ke metode u oceni integriteta konstrukcije, V. Plana, p. 75-92.[5] Radon, J.C. (1982). Determination of Threshold Stress Intensties, Fatigue of Low Alloy Steel BS4360-50D, Int. J. Fatigue, p. 225.164
  • FATIGUE LIFE ESTIMATION OF CRACKED STRUCTURAL COMPONENTS Slobodanka Boljanovi 1, Stevan Maksimovi 2, Strain Posavljak3Summary: This paper presents a computational model/method for calculating theresidual fatigue life for surface-cracked structures. The analysis considers the majorthree-dimensional aspects for the semi-elliptic crack problem including crack shape(planar crack with crack front curvature) and local stress variation in a plate. The plateis subjected to a tensile load. Stress intensity factor, as an important parameter forfatigue life estimation is determined by applying analytical and numerical methods. Thefatigue life of surface-cracked problems has been estimated using the Paris equationby incrementing the crack depth. Calculated residual life correlates well withexperimentally obtained data.Key words: fatigue; semi-elliptical surface crack, crack growth; life evaluation1. INTRODUCTION Structural components are subject to environment during service and incombination with possible cracks initiated at structural discontinuities such as holes ormaterial defects this could lead even to failure. Reliable fatigue life prediction is veryimportant for safe design and maintenance of structural components subjected tocyclic loading [1,2]. Under cyclic loading or under static loading and due to effects ofdifferrent environment factors any surface flaw, if it exists, has the potential ofsubcritacally growing into a surface crack. Analysis of the structure containing suchflaws is needed for modeling and estimation of the corresponding crack propagationrate as well as residual fatigue life. Experimental investigations show that crack shape of propagating surfacecracks in a plate under cyclic loading is approximately semi-elliptical. Due tocomplexity of geometry, for fatigue crack growth analysis of semi-elliptical problems, itis neccessary to examine both the depth and surface directions. Actually, theoreticalmodel for fatigue crack growth is based on two coupled Paris fatigue crack growthrelations. For reliable prediction of crack growth rates and fracture strengths ofcomponents accurate stress analysis of surface crack problems is needed. Over the1 Ph.D, Slobodanka Boljanovi , Belgrade, VTI, Ratka Resanovi a 1, (slobodanka.boljanovic@gmail.com)2 Ph.D, Stevan Maksimovi , Belgrade, VTI, Ratka Resanovi a 1, (s.maksimovic@open.telekom.rs)3 Ph.D, Strain Posavljak, Banja Luka, Faculty of Mech. Eng., Stepe Stepanovi a 75, (s.posavljak@urc.rs.ba) 165
  • S. Boljanovi , S. Maksimovi , S. Posavljakyears, several methods have evolved to compute stress intensity factor for structuralcomponents containing semi-elliptical cracks. In general, for fatigue crack growthanalysis of surface crack problems, two approaches may theoretically be employed tomodel three-dimensional cracking problems. The first approach in which surface crackis analysed as a full three-dimensional problem, and the second, which consists ofreplacing the surface crack by an equivalent two-dimensional or line crack createdmathematically by combining suitable analytical models with correction functions.According to that, some of numerical methods which can be used to calculate stressintensity factor for fatigue crack growth analysis are: the finite element method [3], theboundary-integral equation method [4,5], the finite element alternating method [6] andthe 3-D line method [7-9]. The objective of this paper is to formulate a computational model for fatigue lifeestimation of structural components with surface semi-elliptical cracks. In fatigueanalysis, both, crack depth direction and surface direction are investigated. Moreover,the stress intensity factor is calculated by applying analytical method and numericalmethod (FEM).2. FATIGUE CRACK GROWTH An important aspect in the crack growth analysis of cracked structuralcomponents is the evaluation of residual life under service conditions. This means thecalculation of the time i.e. number of loading cycles required to grow a crack from atolerable size to a critical length. The fatigue life analysis of a complex surface crack problem requiresconsidering two crack growth directions. Actually, Newman and Raju [3] intoduced thatthe aspect ratio change of surface cracks should be calculated by assuming that asemi-elliptical profile is always maintained and for fatigue life estimation it is adequateto use two coupled Paris fatigue laws known as two-point plus semi-ellipse method: da mA db mB CA KA , CB KB , (1) dN dNwhere KA and KB are the ranges of stress intensity factor at the depth and surfacepoints of a surface crack and they can be calculated by applying analytical and/ornumerical methods, CA, CB, mA and mB are material constants experimentally obtained.In crack growth analysis very important aspect is to evaluate fatigue life up to failure.Final number of loading cycles for surface cracked problems can be calculated for bothdirections if equations for crack growth rate are integrated: af bf da db N depth mA , N surface mB . (2) a0 CA KA b0 CB KB Eq. (2) can be used to estimate the number of loading cycles required for thesurface crack to grow through the thickness for the assumed incremental increase incrack depth. The corresponding incremental increase in the surface direction iscalculated using Eq. (3). Every subsequent crack length is determined as: ai 1 ai a and bi 1 bi b. (3) This process is repeated until the crack depth reaches the through-thickness.166
  • Fatigue life estimation of cracked structural componentsThe incremental life is estimated corresponding to every incremental growth and thecumulative total life is evaluated.3. STRESS INTENSITY FACTOR FOR SEMI-ELLIPTIC CRACK PROBLEM In the study of fatigue crack growth and failure behaviour of surface cracks, theessential element is an accurate calculation of the stress intensity factor. Stressintensity factor solutions are required for the assessment of fracture strength and,residual fatigue life, or for a damage tolerance analysis to aid structural design. In fracture mechanics the stress intensity factor can be calculated withfollowing equation: KI Y a , (4)where is the level of external loading, a presents the crack length and Y is thegeometry correction factor. Due to the fact that surface cracked problem is examined inthis paper, Eq. (4) must be adjusted to considered geometry. The complex geometryof surface cracked problems could be introduced by formulating the adequategeometry correction factor. First, Irwin [10] proposed an expression for the stressintensity factor, around an elliptical crack in an infinite elastic solid, subjected touniaxial tension and based on an analysis by Green and Sneddon [11]. As a result ofIrwins considerations, the geometry correction factor along the elliptical crack in aninfinite solid subjected to uniaxial stress acting normal to the plain of the crack canbe expressed as: 2 0.25 2 a 2 sin cos b Y (5) Ekwhere a is the depth and 2b is the surface length of a semi-elliptical flaw. Theparameter is the angle in the parametric equation of ellipse and E(k) is a completeintegral of the second kind [12], i.e.: 1.64 0.5 a Ek 1 1.47 . (6) b Later, the geometry correction factor for a surface crack in an infinite elasticsolid subjected tensile was modified in order to define the stress intensity factor forsurface crack in a finite elastic solid [12-14] (Fig.1). Actually, in Eq. (5), themagnification factor Mf is included, which clearly shows the effects of a front and backfree surfaces of the plate: 2 0.25 2 a 2 sin cos b Y Mf . (7) EkThe factor Mf for finite width w and finite thickness t is formulated as: 167
  • S. Boljanovi , S. Maksimovi , S. Posavljak 2 4 a a Mf M1 M 2 M3 g fw , (8) t twhere M1, M2, M3 present factors depending of depth crack length a and surface cracklength b, i.e.: a M 1 1.13 0.09 , (9) b 0.89 M2 0.54 , (10) a 0.2 b 24 1 a M3 0 .5 14 1 , (11) a b 0.65 bg is factor depending of depth crack length a, thickness t, as well as angle : 2 a 2 g 1 0.1 0.35 1 sin , (12) tand factor fw can be expressed like: b a f w sec . (13) 2w tWith defining the geometry correction factor (Eqs. (7)-(13)), it is possible to calculatethe stress intensity factors at any point along a semi-elliptical crack in a finite platesubjected to tensile load. Additionally, in this paper, numerical approach is built for thestress intensity factors calculation by applying a finite element method. All calculatedresults using analytical and numerical approaches for determination of the stressintensity factors are presented in the section that follows. Fig. 1 Geometry of a semi-elliptical surface crack in a finite plate.168
  • Fatigue life estimation of cracked structural components4. NUMERICAL RESULTS To illustrate computation model for fatigue life of structural components with asemi-elliptical crack, a few numerical examples are presented in this section. Theseexamples examine stress intensity factor calculation as well as fatigue life estimation.In order to verify the validation of presented model for surface crack growth simulationobtained results are compared with experimental data.Example 4.1 Crack growth estimation of semi-elliptical surface crack This example deals with crack life calculation of structural element with semi-elliptical surface crack. Analysed external loading is axial with constant amplitude(Pmax=35 kN, R = 0.1). Geometry characteristics are: a0=2.42 mm, af=9.43 mm,b0=2.54 mm, bf=12.95 mm, 2w=28 mm, t=11 mm, L=102 mm. Material characteristicsof AISI 4130 alloy steel are: y=415 MPa, u=559 MPa, E=210 GPa and Parisparameters CA=6.35*10-12, mA=3.99, CB=5.61*10-12, mB=3.89. The first parameter that has to be examined in fatigue crack growth analysis ofsurface-cracked structure is the stress intensity factor. This important parameter iscalculated by applying Eq. (4) with defined geometry characteristics as well as externalloading. Due to the fact that surface crack is analysed, it was neccessary to examinethe calculation of stress intensity factors for both surface and depth directions. Forcalculation of the stress intensity factors Eq. (4) with Eqs. (6)-(13) were used.Computed values of stress intensity factors for adequate crack increments arepresented in Fig.2. AISI 4130 (R=0.1) 50 40 KIA , KIB [MPa m ] 0.5 30 20 10 0 0 2 4 6 8 -3 10 12 14 a, b [m] (x 10 ) KIa-a (Calculated curve) KIb-b (Calculated curve) Fig. 2 Stress intensity factor versus crack length (a- Depth growth, b-Surface growth). Using the fatigue performance data, according to the structural geometry anddefined fatigue model in previous section, it is possible to calculate crack growth rateand fatigue life up to failure. Obtained results for crack length versus number of loadingcycles up to failure are presented in Fig.3 for both, depth and surface directions. At thesame Figures, all computed results for number of loading cycles up to failure arecompared with experimental data [15]. 169
  • S. Boljanovi , S. Maksimovi , S. Posavljak AISI 4130 (R=0.1) AISI 4130 (R=0.1) 10 14 12 8 10 a [m] (x10 ) a [m] (x10 ) -3 -3 6 8 4 6 4 2 2 0 0 0 2 4 6 8 10 12 14 16 18 0 2 4 6 8 10 12 14 16 18 4 4 N [cycles] (x10 ) N [cycles] (x10 ) a - N (Calculated curve) b - N (Calculated curve) a - N (Experiment) b - N (Experiment) Fig. 3 Crack length versus number of loading cycles for semi-elliptical crack (a- Depth growth, b-Surface growth). It is indicated in Fig.3 that the estimated values of number of loading cycles upto failure are conservative when compared to experimental data. In engineeringpractice existance of conservativity in fatigue crack growth analysis is always benefitialsince in this way safe residual service life of structural elements could be determined.Additionally, conservativity of computed results is often result of defined criteria (oftentoo strict) which are used to formulate adequate analytical relations for crack growthanalysis.Example 4.2 Stress analysis of a surface-cracked plate In this example, stress intensity factor calculation was carried out. Theconsidered plate with surface crack (a=b) is subjected to tensile load. External loadingis with constant amplitude (P = 50 kN). The plate is made of AISI 4130 steel 0.5( y=415MPa, u=559MPa, E=210 GPa, KIC=80 MPam ). Geometry characteristics ofthe surface cracked plate are: a0=b0=3 mm, 2w=50 mm, t=10 mm. In addition to analytical approach for stress intensity factor evaluation used inthe previous example, numerical approach based on finite element method is built inthis paper. For this purpose singular six-node finite elements [16,17] are used.Actually, step-by-step, for each increment of crack length different meshes aremodeled by using super-elements around crack tip [18], Fig. 4. Fig. 4 The part of finite element mesh of the surface-cracked plate.170
  • Fatigue life estimation of cracked structural componentsThe step-by-step procedure is repeated until the computed crack growth is very closeto the final failure of the plate. A representation of the finite element analysis for thesurface-cracked plate (a=b) made of AISI 4130 steel is shown in Fig.4 and Fig.5. V1 L1 23.39 C1 21.97 Surface 11 20.55 19.14 17.72 16.3 14.89 13.47 12.06 10.64 9.223 7.807 6.391 4.975 Z 3.558 Y X 2.142 Output Set: MSC/NASTRAN Case 1 Deformed(0.0366): Total Translation 0.726 Contour: Solid VonMises Stress Fig. 5 Stress distribution of the surface-cracked plate subjected to tensile load.Computed results are listed in Table 1 for stress intensity factor for different values ofcrack length in both directions (surface and depth) by using proposed analyticalmethod and a finite element method.Table 1 Comparison of the computed stress intensity factors using analytical andnumerical method (Positions: A - =900, B - =00) a=b KAFEM KAAnal. KBFEM KBAnal. -3 0.5 0.5 0.5 Step 10 [m] [MPam ] [MPam ] [MPam ] [MPam0.5] 1 3 6.74 6.55 7.92 7.41 2 4 7.88 7.68 9.24 8.87 3 5 8.89 8.41 11.18 9.99 4 6 9.99 9.81 12.48 11.41 It can be observed in Table 1 that the analytical method gives almost the samesolutions as finite element method, so both methods for stress intensity factorcalculation can be used in crack growth analysis of surface-cracked plate subjected totensile load.5. CONCLUSION The fatigue growth model to analyse the semi-elliptical crack in a platesubjected to tensile load is developed. For the stress intensity factor calculations, bothanalytical and numerical approaches are applied. Fatigue life up to failure is estimatedby using two-point plus semi-ellipse method. The fatigue lives up to failure arecompared with experimental results available in the literature. Comparisons ofcomputed number of loading cycles up to failure, point out the fact that the analyticalapproach for calculation of stress intensity factor is in a good agreement with thoseobtained by applying finite element method. Presented fatigue growth model isapplicable in engineering practice for fatigue life estimation of semi-elliptical crackproblems in a plate subjected to a tensile. 171
  • S. Boljanovi , S. Maksimovi , S. PosavljakACKNOWLEDGMENT This research work is financially supported within the project No. OI 174001(SANU-Mathematical Institute, Belgrade) as well as the project No. TR 35052 (VTI-Aeronautical Department, Belgrade) by the Ministry of Science and TechnologicalDevelopment, Serbia.LITERATURE[1] Boljanovi , S, Maksimovi , S., Djuri , M. (2009). Analysis of Crack Propagation Using the Strain Energy Density Method. Scientific Technical Review, vol. LIX, No.2, p.12-17.[2] Boljanovi S., Maksimovi , S. (2009). Fatigue Life Analysis of Cracked Structural Components using Crack Closure Effects, SEECCM 2009 Conference Proceedings, Rhodes, Greece.[3] Newman, J.C.Jr., Raju, I.S. (1981). Stress Intensity Factor Equations for Cracks in Three- Dimensional Finite Bodies Subject to Tension and Bending loads. NASA TM 83200.[4] Cruse, T.A. (1972). Numerical evaluation of Elastic Stress Intensity Factor by the Boundary-Integral Equation Method. In: Swedlow, J.L. (editor). The Surface Crack: Phisical Problems and Computational Solutions. American Society of Mechanical Engineers, New York, p.153-170.[5] Cruse, T.A. (1974). An Improved Boundary-Integral Equation Method for Three Dimensional Elastic Stress Analysis. Computers and Structures, vol. 4, p.741-754.[6] Smith, F.W., Sorensen, D.R. (1976). The Semi-Elliptical Surface Crack – A Solution by the Alternating Method. International Journal of Fracture, vol. 12, no. 1, p.47-57.[7] Rice, J.R. (1972). The Line-Spring Model for Surface Flaws. In: Swedlow, J.L. (editor). The Surface Crack: Physical Problems and Computational Solutions. American Society of Mechanical Engineers, New York.[8] Parks, D.M., Lockett, R.R. (1980). Development and Calibration of the Non-Linear Line- Spring Model for Analysis of Surface Cracked Plates and Shells. In: EPRI Ductile Fracture Research Review Document, EPRI Rep. No-80-10-LD WS 80-912, p.9.1-9.12.[9] Delale, F., Erdogan, F. (1981). Line-Spring model for Surface Cracks in a Reissner Plate. International Journal Engng. Sci. vol.19. p.1331-1340.[10] Irwin, G.R. (1962). The Crack Extension Force for a Part Through Crack in a Plate. Trans. ASME. Journal of Applied Mechanics, vol. 84, p.651-654.[11] Green, A.E. , Sneddon, I.N. (1950). The Distribution of Stress in the Neighborhood of a Flat Elliptical Crack in a Elastic Solid. Proc. Cambridge phil. Soc., vol. 46, p.159-164.[12] Jones, R. et al. (2004). Weight Functions, CTOD, and Related Solutions for Cracks at Notches. Engineering Failure Analysis, vol. 11, no. 1, p.79-114.[13] Anderson, T.L. (1995). Fracture Mechanics: Fundamentals and Applications. CRC Press.[14] Newman, J.C.R.I., Raju, I.S. (1986). Stress-intensity Factor Equations for Cracks in Three- dimensional Finite Bodies Subjected to Tension and Bending Loads. In: Atluri Satya N. (editor). Computational Methods, Elsevier Science, p.312-334.[15] Song, P.S., Shieh, Y.L. (2004). Crack Growth and Closure Behaviour of Surface Cracks. International Journal of Fatigue, vol.26, no.4, p.429-436.[16] Sukumar, N., Kumusa, M. (1992). Application of the Singular Finite Element to Crack and Sharp Notches in Orthotropic Media. International Journal of Fracture, vol. 58, p.177-192.[17] Maksimovic, S., Burzic, Z. and Maksimovic, K. (2006). Fatigue Life Estimation of Notched Structural Components: Computation and Experimental Investigations. ECF16 Conference Proceedings, E.E. Gdoutos (editor), Alexandroupolis, Greece, Springer.[18] Msc/NASTRAN, Theoretical Manuels.172
  • HYDRAULIC INSTALLATION OF EKO CONTAINER Ibrahim Badžak1, Ph.D., Remzo Dedic2, Ph.D., Mersida Manjgo3, Ph.D.Summary: This paper describes the hydraulic installation that is embedded in theprototype design of ecological containers. Identify the problems that were created tosolve this task, and the ways in which they are resolved.Keywords: hydraulic installation, ecological containers1 INTRODUCTION Ecological containers are a novelty in the way of collecting waste in the narrowcity centers, parks and pedestrian zones, where the rule resides a large number ofpeople on a relatively small area. In this way, there is a large amount of waste that isdifficult to take. The proposed solution provides a significant contribution to solving thisproblem.2 WORKING PRINCIPLE Ecological container is a steel frame in which the store an adequate number ofstandard containers, which are then lowered into a special mechanism suitable holesdug in the sidewalk or the street. (Fig. 1)1 Univerzitet „Džemal Bijedi “ Mašinski fakultet Mostar, Sjeverni logor bb, BiH, ibrobadzak@hotmail.com2 Sveu ilište u Mostaru, Fakultet strojarstva i ra unarstva , Matice hrvatske bb, BiH, rdedic2001@yahoo.com3 Univerzitet „Džemal Bijedi “ Mašinski fakultet Mostar, Sjeverni logor bb, BiH, manjgom@yahoo.com 173
  • Ibrahim Badžak, Ph.D., Remzo Dedic, Ph.D., Mersida Manjgo, Ph.D. Fig. 1 Ecological container In this way the surface of the pavement or in the plane of the square remainsthe only series basket wastes. Each of them indicated that the wastes are injected(plastic bottles, glass bottles, aluminum bottles, paper, etc.) Inserted items to cart, the body of the basket, sinking into a container locatedunder the basket, resulting in a higher capacity, thereby reducing the number ofrequired discharge. It also prevents animals to approach through garbage and rubberinsulation significantly reduces the intensity of odors. Emptying the container is achieved in the same way as the classical discharge: After the arrival of a vehicle to transport waste joins hydraulic installation by quick-detachable coupling, a shifting lever valve wiring is accomplished raising support structure with a container to the level of the pavement (Figure 2).174
  • Hydraulic installation of eko container Fig. 2 Container Lifting Containers are then extracted from the supporting structure one by one, empty the classical way, and then returned to supporting construction (Figure 3). After completion of discharge, the action of the valve handle wiring, the bearing structure of the container is again lowered into the hole Workers separate hydraulic installation and go. Fig. 3 Empty containers ecological3 HYDRAULIC INSTALLATIONS Hydraulic scheme of hydraulic installation of ecological container is shown inFigure 4. 175
  • Ibrahim Badžak, Ph.D., Remzo Dedic, Ph.D., Mersida Manjgo, Ph.D. Executive elements of hydraulic systems are two hydraulic cylinders (1)connected at the ends of supporting structure. As the friction in the guides and the possible uneven distribution of loads canbe expected for some delay in one cylinder to the other, so in hydraulic system installintegrated flow regulator (3) which is substantially prevented. The cylinders are one way to have a smaller number of lines. Fig. 4 Hydraulic installation ecological container The cylinders are one way to have a smaller number of lines. The control element (6) is a distributor 3 / 3, and the inability to purchase theprototype installation is embedded distributor 4 / 3. Manage distributor is done by hand,a return to the neutral position is achieved with springs. On suporting construction container flow between the regulator and cylinderare made of metal tubing, the flow regulator valve and a hose with a coupler finallyenabling you to quickly assembling and disassembling hydraulic installations, withoutleaking of hydraulic oil.176
  • Hydraulic installation of eko container Driving part of the hydraulic installation is achieved with a vehicle for waste.This reduces the price of hydraulic installation of eco-container, and avoids potentialproblems mounting the drive under the earth (because of the possibility of waterpenetration), and bringing power lines to container. The drive unit is axial - piston pump. Reservoir contained 100 gallons ofhydraulic oil.4 EXPERIMENTAL TESTING Experimental testing of ecological containers carried on the polygon in thefactory that produced it. (Figure 5). Fig. 5 Experimental testing of a prototype Experimental data have found certain irregularities in raising the supportstructure, and the cause it was too big gap between the sliding surfaces and the air stillpresent in a hydraulic installation. After irradiation, reducing the gap and hydraulic installations, these problemshave disappeared.5 CONCLUSION Ecological containers represent a new way of waste disposal, which veryefficiently solve several problems (transport, environment, health) Experiments with the prototype confirmed the hypothesis and the possibility ofusing hydraulic installation vehicles for transport of waste. The following steps shouldmake certain improvements in design, a larger series would reduce the price of theseproducts and thus make them more accessible to the market, as well as theirapplication in practice. 177
  • Ibrahim Badžak, Ph.D., Remzo Dedic, Ph.D., Mersida Manjgo, Ph.D.LITERATURE[1] Basta, T., (1972). Mašinska hidraulika, Nau na knjiga, Beograd[2] Jovanovic, P.,Ciric, M., (1985) Cilindri u hidraulici i pneumatici, Održavanje mašina i opreme, Beograd[3] Jovanovic, S., (1981) Uljna hidraulika I, Tehni ka knjiga, Beograd,[4] Jovanovic, S., (1983) Uljna hidraulika II, Tehni ka knjiga, Beograd,[5] Kelic, V., (1985) Hidroprenosnici, Nau na knjiga, Beograd,[6] Savic, V., (1987) Hidrauli na ulja i održavanje hidrauli nih sistema, Mašinski fakultet Zenica,[7] Savic, V., (1987) Mašinska hidraulika, Mašinski fakultet, Zenica,[8] Savic, V., (1988) Uljna hidraulika, Dom štampe, Zenica,[9] Momirski, M., (1981) Elementi teorije skeletnih konstrukcija, Fakultet tehni kih nauka, Novi Sad178
  • PREDICTION OF DAM BEHAVIOUR USING MULTIPLE LINEARREGRESSION AND RADIAL BASIS FUNCTION NEURAL NETWORK Vesna Rankovi 1, Nenad Grujovi 2, Goran Milovanovi 3, Dejan Divac4, Nikola Milivojevi 5Summary: The safety control of dam is supported by monitoring activities and is basedon mathematical models. The variations of hydrostatic pressure and temperature arethe main variables that should be taken into account when analyzing the results of theconcrete dam observations. Deterministic models based on mechanical principles areoften difficult to construct. Statistical procedures, such as multiple linear regression(MLR), have been applied to dam safety to determine the influence of external loadson the structure deformation. The relations between the loads and dam behavior arenonlinear. Radial basis function (RBF) neural network can be successfully applied tofunction approximation, forecast, and dynamic systems identification. Neural networkmodeling from measured data is effective tool for the approximation of uncertainnonlinear systems. This paper presents novel approach based on the use of RBFnetwork to estimate dam behaviour. Mathematical models based on experimental dataare developed. MLR and RBF neural network models for prediction of dam behaviourhave been compared with the measured data on the basis of correlation coefficient.Key words: Dam Safety, Multiple Linear Regression, Radial Basis Function NeuralNetworks, Dam Behaviour, Modeling1. INTRODUCTION The variations of hydrostatic pressure, temperature and other unexpectedunknown causes, such as the time effects, are the main variables that should be takeninto account when analyzing the behaviour of the dam. Different models have beendeveloped and used to analyze dam behaviour. These are grouped into two generalcategories: deterministic and statistical models, [1]. The deterministic modeling1 Ph.D., Vesna Rankovi , Kragujevac, Faculty of Mechanical Engineering, University of Kragujevac,(vesnar@kg.ac.rs)2 Ph.D., Nenad Grujovi , Kragujevac, Faculty of Mechanical Engineering, University of Kragujevac,(gruja@kg.ac.rs)3 MSc., Goran Milanovi , Mrkonji Grad, HES Vrbas, (g.milanovic@hesvrbas.com)4 Ph.D., Dejan Divac, Belgrade, Jaroslav erni Institute for the Development of Water Resources,(ddivac@eunet.rs)5 Ph.D., Nikola Milivojevi , Belgrade, Jaroslav erni Institute for the Development of Water Resources,(nikola.milivojevic@gmail.com) 179
  • Vesna Rankovi , Nenad Grujovi , Goran Milanovi , Dejan Divac, Nikola Milivojevirequires solving differential equations for which closed form solutions may be difficultor impossible to obtain, [2]. The advantages of the statistical method, such as multiplelinear regression, consist in simplicity of formulation and speed of execution. The radial displacement of one or several points of the dam is is an importantbehaviour indicator. The radial displacement in any point of the dam is nonlinearfunction of hydrostatic pressure and temperature and other unexpected unknowncauses. One of successful neural network applications is to model complex nonlinearbehaviour. Mata [3] compared multiple linear regression and multilayer perceptron (MLP)neural network models for the prediction of the upstream–downstream displacement ofarch dam recorded by a pendulum. The models are generated on the basis ofexperimental data of time histories of reservoir level and external temperature and ofstructural responses. An improved neuro-wavelet modeling methodology to model andforecast the horizontal displacement of the dam was applied by Cao et al., [4]. The aim of this paper is to construct a MLR and RBF neural network models topredict the radial displacement of arch dam and demonstrate its application toidentifying complex non-linear relationships between input and output variables. The RBF neural network has the capability of universal approximation [5-7] asmultilayer perceptron. In most of the literature, RBF neural networks are considered asa smooth transition between fuzzy logic and neural networks. RBF network isespecially interesting because of their simple but adaptive structure and effectivelearning ability. Although RBF networks usually need more neurons in hidden layerthan MLP, they are easier to initialize and train than MLP networks [5].2. CASE STUDYThe Bocac dam, on river Vrbas, is a medium size dam (Fig. 1). Fig. 1 Upstream face of dam and cross-section through block 8 Dam is located in Republic Srpska about of 25 km from the city of Banja Luka.The dam was constructed between 1976 and 1981. It is a double curvature arch damand height is 66 m, crest length 221.4 m. The crest thickness is 3m-13m while themaximum thickness of the dam base is 14.4m. The minimum, normal and maximumoperating levels are 254, 282 and 283 m above sea level (asl), respectively. The totalcapacity of reservoir is 52.7 106 m3 . The dam is equipped with a monitoring system180
  • Prediction of dam behaviour using multiple linear regression and radial basis function neural networkto measure a particular parameters, such as: concrete, water and air temperatures,reservoir water level, horizontal and vertical displacements, rotations, foundationdisplacements, movements of joints, strain, stress, uplift pressure, foundationdisplacements and seepage. The three pendulums were installed to measure radial and tangentialdeformations. In this paper it is analyzed the radial displacement of point V1 at block 8with the proposed methods. A data set include 392 data samples. The available set ofdata was divided into two sections as training and test set. Data from January 2000 toDecember 2008 are used to train, and data from January 2009 to December 2010 areused to test, (Fig.2). Fig. 2 The radial displacement of the point V1 at block 83. OVERVIEW OF THE MLR, RBF NEURAL NETWORK3.1 Multiple linear regression N Consider a training data set x1, y1 , x2 , y 2 ,..., x p , y p , where Txi x1i x2 i ...xNi is a vector of input variables and y i is the corresponding outputvalue, p is the number of training data points. The multiple linear regression model is given by: MLR ym 0 1 x1 2 x2 ... N xN (1)where i , i 0,1,..., N represent unknown parameters.The i can be estimated by minimizing the sum of the squares of the errors: 2 2 2 y1 y m1 y2 y m2 ... yp y mp (2)where y mi denote the MLR output value from the i-th input element: y mi 0 1 x1i 2 x2i ... N xNi (3) 181
  • Vesna Rankovi , Nenad Grujovi , Goran Milanovi , Dejan Divac, Nikola MilivojeviThe matrix form Eq. 2 is: T y X y X (4) 1 x11 x21 xN 1 1 x12 x22 xN 2 T T where: X , 0 1 ... N , y y1 y 2 ... y p 1 x1p x2 p xNpThe least squares estimator of is given by: 1 XT X XT y (5)3.2 RBF network RBF network consists of one hidden layer and takes Gaussian functions as itsbasis functions. The typical structure of an RBF network with N inputs, s hiddenneurons and one output is shown in Fig. 3. Fig. 3 Structure of RBF neural network The inputs x1 , x2 , …, xN are applied to all neurons in the hidden layer. Theoutput ai 1 of each hidden unit i is then computed by: - ni21 ai 1 e , i 1,2,...s (6)where: 0.5 N 2 ni i, j 1 - xj bi 1 (7) j 1 ln 0.5 i, j 1 is center of the Gaussian functions for i-th hidden neuron and bi 1 , spwhere sp is the radius or spread of all radial basis functions.182
  • Prediction of dam behaviour using multiple linear regression and radial basis function neural network The output of the RBF neural network is given by: s RBF ym 1,i 2 ai 1 b1 2 (8) i 1where 1,i 2 is the weight from the i-th hidden layer neuron to the output neuron andb1 2 is the bias at output neuron. There are different learning strategies which can be used to design of an RBFnetwork, depending on how the centers of the radial basis functions of the network arespecified. The most popular existing learning strategies include: fixed centers selectedat random, self-organized selection of centers, and supervised selection of centers. Inthis paper, the first strategy is used to determine the parameters of the RBF network.The learning method requires training data set. In this method, the centers of theGaussian functions i , j 1 are randomly selected from the training data set. Spread ofradial basis functions sp is important parameter for constructing an RBF neuralnetwork. If spread is small, the RBF is very steep so that the neuron with the weightvector closest to the input will have a much larger output than other neurons. A largerspread leads to a large area around the input vector where hidden neurons willrespond with significant outputs. In this paper the Matlab Neural Network Toolbox isused for the implementation of the RBF network. The Matlab function newrb is applied.After obtaining i , j 1 values for all hidden neurons, the weights 1,i 2 and bias b1 2 aredetermined using multiple linear regression techniques.4. SIMULATION RESULTS Statistical model gives the displacement as the sum of three terms: the first isrelated to the hydrostatic pressure, the second term takes into account function of timeand the third term is due to air temperature change. The general statistical model usedfor the prediction of the displacement of a point in a dam is: MLR y m h, t , s 0 1h 2 h2 3 h3 4 h4 5 e t 6 t 2 (9) 7 cos d 8 sin d 9 sin d 10 sin d cos d 2 jwhere h is water level, t is the elapsed time expressed in years, d is the 365season varying between 0 and 2 , j represents the number of days since January 1st. The input variables ( x1, x2 ,..., x10 ) of the MLR model are h, h 2 , h3 ,h 4 , e t , t , cos d , sin d , sin2 d , sin d cos d .The MLR model for the prediction of the radial displacement of point V1 is: y m V 2 h, t , s 1.7591 10 5 h 2 5.8059 10 9 h3 2.2973 10 11 h 4 MLR (10) 6.6749cos d 4.6632sin d It can be seen from Eq. (10) that the elapsed time did not have a significanteffect on the radial displacement of point V1. Correlation coefficients of 0.9595 for the 183
  • Vesna Rankovi , Nenad Grujovi , Goran Milanovi , Dejan Divac, Nikola Milivojevitraining and 0.9748 for the test set are obtained between the measured and MLRmodeled values of the radial displacement. Based on the MLR analysis results, the inputs of the RBF model are the h 2 , 3 4 h , h , cos d and sin d . The output of the model is the radial displacement ofpoint V1. Different models were constructed and tested in order to determine thevalues of sp and goal. The spread and goal are varied from 1 to 2 and from 0.05 to 0.1,respectively. It is found that RBF with spread equal to 1.7 and goal equal to 0.07produced the correlation coefficients of 0.9637 and 0.9766 between the predicted andthe measured values of the radial displacement of point V1 for the training and testsets, respectively.5. CONCLUSION In this paper, RBF neural network and MLR model were developed to predictthe radial displacement of the arch-dam. The performance of the RBF network andMLR model were tested using correlation coefficients. Results of simulation, presentedin this paper, show that the application of the neural network to prediction of radialdisplacement gives a slightly higher coefficient of correlation values for training andtest sets. Proposed approach based on RBF network can be a very efficient tool anduseful alternative for the computation of seepage, stress, or crack opening of dam.ACKNOWLEDGMENT: The part of this research is supported by Ministry of Science in Serbia,Grants III41007 and TR37013.LITERATURE[1] ICOLD. (2003). Methods of analysis for the prediction and the verification of dam behaviour. Tech. rep. Swiss Committee on Dams.[2] Szostak-Chrzanowski, A., Chrzanowski, A., Massiéra, M. (2005). Use of deformation monitoring results in solving geomechanical problems—case studies. Engineering Geology, vol. 79 , no.1-2, p. 3 –12.[3] Mata, J. (2011). Interpretation of concrete dam behaviour with artificial neural network and multiple linear regression models. Engineering Structures, vol. 33, no. 3, p. 903–910.[4] Cao, M., Qiao, P., Ren, Q. (2009). Improved hybrid wavelet neural network methodology for time-varying behavior prediction of engineering structures. Neural Computing & Applications, vol. 18, no. 7, p. 821–832.[5] Park, J., Sandberg, I. W. (1991). Universal approximation using radial basis function networks. Neural Computation, vol. 3, no. 2, p. 246–257.[6] Leonard, J.A., Kramer, M.A., Ungar, L.H. (1992). Using Radial Basis Functions to Approximate a Function and Its Error Bounds. IEEE Transactions on Neural Networks, vol. 3, no. 4, p. 622–630.[7] Schilling, R.J., Carroll, J. J., Al-Ajlouni, A. F. (2001). Approximation of nonlinear systems using radial basis function neural networks. IEEE Transactions on Neural Networks, vol. 12, no. 1, p. 1–14.184
  • SOME ASPECTS IN FAILURE ANALYSIS OF CRANES 1 Nenad Zrni , Sr an Bošnjak 2, Vlada Gaši 3 , Miodrag Arsi 4Summary: Failure of a structural or mechanical component of cranes usually can beassociated with materials-related problems and/or design-related, as well as thefabrication-related problems or inadequate structural maintenance. Also, cranecomponents and structure experience a spectrum of stresses while operating.Therefore, about ten percent of material handling high-performance machines failurescan be attributed to fatigue failure. In most cases these failures were unexpected andlead to catastrophic consequences. This paper discusses some aspects in failureanalysis of cranes, particularly high-performance ones, gives the background for failureanalysis and presents some typical examples of failure. The aim of this article is toencourage practitioners in the failure investigation process to look beyond themetallurgical issues and to also examine the loads and stresses.Key words: cranes, failure analysis, finite elements method1. INTRODUCTION Exploitation of heavy duty and high-capacity lifting/conveying and earthmovingmachines such as cranes (tower cranes, container cranes, ship unloaders, gantrycranes, elevators, conveyors, etc.), bucket wheel excavators and stacker/reclaimersunder the action of highly pronounced dynamic loads, may lead to failures of theirstructural parts, substructures and subassemblies - plastic deformations, cracks andfractures. Failures of the cranes’ structural parts unavoidably lead to serious damagesor total collapses; these accidents are often followed by very high financial losses(millions of €) and possibly serious injuries or crane-related fatalities. A lot of datareveals on the serious consequences of cranes’ accidents. For instance, more than500 US construction workers died in crane accidents between 1984 and 1994,according to a study of Occupational Safety and Health Administration (OSHA), whilethe International Union of Operating Engineers revealed that 502 workers died in 480separate accidents - major causes included assembly/dismantling (12 percent), boombuckling (8 percent), rigging failure (7 percent) and overturn (7 percent). [1]. The USDepartment of Labor OSHA in May 2010 estimated there are as many as 82 fatalities1 Associate Professor, University of Belgrade, Faculty of Mechanical Engineering, nzrnic@mas.bg.ac.rs2 Professor, University of Belgrade, Faculty of Mechanical Engineering, sbosnjak@mas.bg.ac.rs3 Assistant, University of Belgrade, Faculty of Mechanical Engineering, vgasic@mas.bg.ac.rs4 Senior Research Associate, Institute for Testing Materials, Belgrade, miodrag.arsic@institutims.rs 185
  • Nenad Zrni , Sr an Bošnjak, Vlada Gaši , Miodrag Arsiannually associated with cranes in construction, and said a more up-to-date standardwould help prevent them. The increase of cranes numbers was followed by theincrease of accidents, so in 2009, according to the source [1], have been reported 303accidents with 197 deaths (in the same period, according to Aviation Safety Network -http://aviation-safety.net/, have been reported 30 aircraft accidents with 757 fatalities).Statistics reveal that crane accidents are the most common cause for construction sitedeath. There are over 125,000 cranes used in operation in the construction injury andanother 100,000 in general and maritime industries. The causes of crane-relateddeaths in US construction industry in the period 1992-1996 are given in Fig. 1 [2]. Fig. 1 Causes of crane-related fatalities in US construction (1992-2006)[2] * Included 64 struck by falling booms/jibs, ** Included 21 falls from cranes, 9 falls from crane baskets, 8 from crane loads, ***Other causes included 9 highway incidents. Types of cranes usually involved in statistical data are mobile cranes, towercranes, floating or barge cranes and overhead cranes. At least 71% of all crane-relatedaccidents involved mobile cranes, Fig. 2a. Mobile cranes are involved e.g. in 63% of allcrane collapses, tower cranes are involved in 5% of all crane-related incidents, whileother and unspecified cranes were involved in 24% of all crane related incidents [2].The reasons of the crane collapses are shown in Fig. 2b [2]. According to [3], there is30 plus major accidents annually worldwide with around 50 deaths. Also, since theyear 2000 there have been over 1112 tower crane accidents which have resulted inover 778 deaths and countless injuries [3]. Of course, many accidents are neverreported on, so the data can perhaps be doubled. Worldwide tower crane accidentstatistics in the year 2009 gives 176 accidents resulting in 75 deaths, while in the year2010 we have 112 accidents with 62 deaths [3]! The incidents analyses statistics showthat a non-negligible number of reasons leading to the accidents remain unknown. Mechanical and civil engineering structures are always designed to carry theirown dead weight, superimposed loads and environmental loads such as e.g. wind orwaves. These loads are usually treated as maximum loads not varying with time andhence as static loads. In some cases, the applied load involves not only staticcomponents but also contains a component varying with time which is a dynamic load.In the past, the effects of dynamic loading have often been evaluated by use of anequivalent static load, or by an impact factor, or by a modification of the factor of safety186
  • Some Aspects in Failure Analysis of Cranes[4]. Many developments have been carried out in order to try to quantify the effectsproduced by dynamic loading. Examples of structures where it is particularly importantto consider dynamic loading effects are the construction of tall buildings, long bridgesunder wind-loading conditions, buildings in earthquake zones, high-performancecranes etc. Exploitation of high-performance cranes in insensitive working conditionsprovides fertile ground for the occurrence of fatigue cracks, while extremeenvironmental conditions may also cause disastrous consequences, even whenmachines are out of operation. Brittle fracture occurrence of vital parts of the structureis also possible in the installation stage of the machine [5]. The above examplesconfirm that a failure with disastrous consequences is possible at any stage of theproduct life cycle [5]. In addition, they confirm the factual existence of four mainreasons for the collapse of high-capacity lifting/conveying and earthmoving machines[6]: (1) design faults, the so-called ‘designing-in’ defects, [6,7]; (2) manufacture faultscausing the so-called ‘manufacturing-in’ defects, [6,7]; (3) exploitation faults –according to [7], these causes can be named ‘operating-in’ defects; and (4) extremeenvironmental impacts – unusual occurrences (extreme storm, earthquake, fire) –according to [7], these causes can be named ‘environment-in’ defects. Of course,machine failures are often the result of a combination of several different causes [6]. Fig. 2 a) Overturning of the mobile crane, b) Statistics of crane collapses2. SOME EXAMPLES IN CRANES FAILURE Several different examples of crane failures and accidents, reported in the lastdecade, can be found in the following references and explained shortly in this paper.Failure analysis of mobile harbor crane wheel hub is presented in [8], while the failureanalysis demonstrated that the mechanism of failure was fatigue due to the service-induced growth of a fatigue crack and stress-concentrating effects of that crack onmotioninduced stresses. Failure analysis of a large ball bearing of a dockside crane isshown in [9]. Analysis of the bearing showed that faulty hardening of the bearingcaused pitting of the raceways, which led to failure of the bearing. It should bementioned that in both papers [8,9] the authors didnt carry out a thorough structuralanalysis and/or e.g. employ FEM to evaluate the failure. Practically they used amettalurgical approach to analyze failure. On the other hand some failure casestudies encompassed a more detailed structural analysis including FEM [3, 10, 11, 12,13]. For instance fatigue damage analysis and repair procedure at the large 250 kNportal crane placed in the shipyard is shown in [10]. The fatigue cracks occurred atseveral critical points, i.e., on the bottom of the tower and on both legs of the portal, 187
  • Nenad Zrni , Sr an Bošnjak, Vlada Gaši , Miodrag ArsiFig. 3a). Previous attempts of repair by simple welding of cracks were not successful,because new cracks were detected soon after the repair, Fig. 3b).When cracksreached the critical length (300–500 mm), the exploitation of the crane was stoppedand detailed analysis was carried out. First cracks were detected soon after the cranewas placed in the shipyard. The crane was still in usage with the reduced allowedcarrying capacity (50 kN instead of nominal 250 kN), but the cracks continued to grow.In order to find the source of crack initiation and growth, the complete documentationand calculations were checked. It was found that the calculations were performed byusing the finite element method and simple beam elements, without taking into accountthe stress concentrations, the influence of inertial forces and the wind. Those facts ledto perform the complete static and fatigue analysis, what involved measurements ofreal stresses during the typical maneuvers. Fig. 3 a) Location of fatigue cracks, b) Cracks after repair The catastrophic collapse of the crane on the Milwaukee Brewers baseballstadium retractable roof project (Miller Park) that could be the most awesome liftaccident of all time is shown in [11], while the investigation of the cause and origin ofthe collapse of an overhead tower crane at an office building construction site inBellevue, USA, is given in [12]. Failure analysis of container crane boom failureoccurred while the boom were being lowered is shown in [13]. The most recent failureanalysis of a tower crane that was used during the construction of a hydroelectricpower plant is given by the authors of this paper in [3]. Due to the breakaway of thegusset plate of the counterjib truss structure, Fig. 4a) and 4b), plastification occurredand the counterjib ballast fell from the crane. This caused the crane’s loss of staticstability whereupon the jib leaned (at a distance of approximately 29 m from the axis ofrotation) onto the structure of the adjacent portal crane. Because of this there was nototal overturning of the crane. Fig. 4 a) Fractured gusset plate, b) 3D model of the joint fracture188
  • Some Aspects in Failure Analysis of Cranes In order to diagnose the cause of the gusset plate fracture the authors in [3]carried out FEA to calculate the working stress in the gusset plate accompanied bychemical composition and mechanical properties testing and microstructure testing. Itwas concluded that the fracture of the gusset plate originated from the cumulativeinfluence of the following factors, such as influence of the parent metal hardening inthe fracture zone.3. CONCLUSION Failure of a structural or mechanical component usually can be associated withmaterials-related problems and/or design-related problems (which may include,depending on the definition of design, unexpected service environment). Materialsfailure analysis investigations are usually carried out first, and the objective is todetermine what the material can reveal about the cause of failure [14]. This type offailure analysis consists of first evaluating the failed material for evidence of the failuremechanism such as fatigue, overload, corrosion and environmentally assisted cracking. But, at the same time an engineering review of the component design andservice application to determine the loads, displacements, temperatures, vibrations,and other service-related factors is required. It may also include numerous analytical,classical, and computer techniques that are available to assist in a structural designfailure analysis. However, traditional analytic techniques have their limitations in thatonly relatively simple and idealized structures can be analyzed using simplified loadingand materials property assumptions. When complex designs, such as high-performance material handling and mining machines, transient loadings, and nonlinearmaterial behavior need to be evaluated, computer-based techniques are used. This iswhere finite element analysis (FEA) is most applicable and provides considerableassistance in design analysis as well as failure analysis [14]. Finite element analysis isone of the most common tools used by design engineers in failure analysis[3,10,11,12,13]. It has application in the structural/mechanical fields to determine thestress, strain, and displacements of structures subjected to different types of loading.Therefore, in the failure investigation process it is necessary to look beyond themetallurgical issues and to also examine the loads and stresses. The majority of large structures such as cranes have significant welding andhence small defects (cracks) are already present and in [15] is assumed that smallcracks are present and that crack propagation is the only part of the fatigue process.Because of non-uniform temperature gradient and local elastic/plastic deformationsduring the welding and cooling process, residual stresses are found in all weldedstructures. The residual stresses profiles and their magnitudes are difficult to quantify.Therefore it is usual for welded structures to ignore the effects of mean stress andwhether the stress is tensile or compressive. An investigation given in [15] afteranalyzing failures of over sixty materials handling machines found that about tenpercent of shiploader, stacker, reclaimer and stacker/reclaimer failures can beattributed to fatigue failure. Similar can be concluded for other type of material handlingmachines. In most cases these failures were unexpected and lead to catastrophicconsequences. The process of fatigue means that steel strength deteriorates under theaction of cyclic loads and this may ultimately lead to cracking and the unexpectedfailure of structures. 189
  • Nenad Zrni , Sr an Bošnjak, Vlada Gaši , Miodrag ArsiACKNOWLEDGMENT A part of this work is a contribution to the Ministry of Science and TechnologicalDevelopment of Serbia funded project TR 35006.LITERATURE[1] CraneAccidents.com, official site, from http://craneaccidents.com/stats.htm, accessed on 2011-03-28.[2] McCann, M. (2010). Understanding crane accident failure: A report on causes of deaths in crane-related accidents, Crane and Rigging Conference. Presented at CRC 2010, May 27, Houston, Texas, US.[3] Zrni , N., Bošnjak, S., Gaši , V., Arsi , M. (2011). Failure analysis of the tower crane counterjib. Procedia Engineering, Elsevier, accepted for publication.[4] Kuntiyawichai, K., Burdekin, FM. (2003) Engineering assessment of cracked structures subjected to dynamic loads using fracture mechanics assessment. Engineering Fracture Mechanics, Vol. 70, p. 1991-2014.[5] Hadianfard, MJ, Hadianfard, MA. (2007) Structural failure of a telescopic shiploader during installation. Journal of Failure Analysis and Prevention, Vol. 7: p. 282–91.[6] Bošnjak S., Arsi , M., Zrni , N., Rakin, M., Panteli M. (2011) Bucket wheel excavator: Integrity assessment of the bucket wheel boom tie-rod welded joint. Engineering Failure Analysis. Vol. 18. p. 212–222.[7] Gagg, CR. (2005) Failure of components and products by ‘engineered-in’ defects. Case studies. Engineering Failure Analysis. Vol. 12.1000–1026.[8] Shirokoff, J. Mobile harbor crane wheel hub fatigue failure. (2003) Practical Failure Aanalysis.; Vol. 3(1). p. 81-84.[9] Ost, W., De Baets, P., De Waele, W. (2004) Failure of a large ball bearing of a dockside crane. Engineering Failure Analysis. Vol. 11. p. 335–53.[10] Domazet, Ž., Krstulovi -Opara, L., Stupalo, S. (2005) Fatigue cracks and failures in cement industry, shipbuilding and power plant facilities. Engineering Failure Analysis. Vol. 12. p. 819–833.[11] Ross B., McDonald, B., Vijay Saraf, SE. (2007) Big blue goes down. The Miller Park crane accident. Engineering Failure Analysis. Vol. 14. p. 942–961.[12] McDonald, B., Ross, B., Carnahan, RA. The Bellevue Crane Disaster. (2011) Engineering Failure Analysis, doi: 10.1016/j.engfailanal.2010.09.003.[13] Soderberg, E. (2008) Container crane boom collapse: Cause and prevention. Port Technology International. Vol. 39. p. 60-65.[14] Service, T. (2002) ASM Handbook Vol. 11, Failure Analysis and Prevention (ASM International). Finite Element Modeling in Failure Analysis. p. 380-389.[15] Morgan, R. Gatto, F., Ford, P. (2005). Fatigue of Complex Structures for the Mining Industry. Proceedings Australian Structural Engineering Conference. Sydney, N.S.W.: Engineers Australia. p. 753-759.190
  • SOME ASPECTS TO DESIGN OF AIRCRAFT STRUCTURES WITH RESPECTS TO FATIGUE AND FRACTURE MECHANICS Stevan Maksimovi 1, Ivana Vasovi 2, Mirko Maksimovi 3Summary: Attention in this work is focused on developing computation methods ofdamaged aircraft structural components with respect fatigue and fracture mechanics.Computation method is based on combining singular finite elements to determinestress intensity factors for cracked structural components with corresponding crackgrowth lows that include effect of load spectra on number of cycles or blocks up tofailure. To demonstrate efficient computation procedure in fatigue life estimation herenumerical examples are included. Attention in this work is focused on design of aircraftwing-fuselage joints. Computation procedure to strength analyze with respects fatigueand fracture mechanic is applied to cracked aircraft lug type structure. Computationresults are compared with correspond experiments. Good agreement betweencomputation and experimental results is obtainedKey words: Fatigue, Fracture mechanics, Aircraft structures, Damaged lugs, FiniteElements,Crack growth, Strain energy density method1. INTRODUCTION Damage tolerance application to the light training aircraft structural componentsis limited to critical parts. A part, that if it fails, alone may cause the loss of an aircraft isclassified as a critical part. This definition means, that aircraft wing-fuselageattachments must be comply with the damage tolerance requirements [1,2]. The maingoal is a safe life design, i.e. a slow crack growth structure not requiring any insectionduring its full life. The Damage Tolerance approach assumes the components have apreexisting flaw from which a crack will grow under dynamic loads. This assumptionmakes it possible to account for in-service or manufacturing defects in determining thedynamic life. The Damage Tolerance Methodology uses fracture mechanics to predictthe fatigue crack growth in a structure. In the design analysis of a slow crack growthstructure it is most important to make correct estimates for the early portion of thecrack growth process, because it is there the life is. In most cases this implies thatmaximum accuracy is needed for small corner cracks.1 Ph.D, Stevan Maksimovi , Belgrade, VTI, Ratka Resanovi a 1, (s.maksimovic@open.telekom.rs)2 Ph.D student, Ivana Vasovi , GOŠA institute Belgrade, (ivana.vasovic@gmail.com)3 Ph.D student, Mirko Maksimovi , Water supply and Sewerage, Belgrade, (mirkom@open.telekom.rs) 191
  • S. Maksimovi , I. Vasovi , M. Maksimovi Attachment lugs are particularly critical components in crack initiation andgrowth because of their inherently high stress concentration levels near the lug hole.For these reasons, it is important to develop analytical/numerical as well asexperimental procedures for assessing and designing damage tolerant attachment lugsto ensure the operational safety of aircraft. Over the years, several extensive studies[3-5] have been made on lug fatigue performance, involving both experimental andnumerical means. In the work of fatigue crack growth and fracture behavior of attachment lugs[6,7], an accurate calculation of the stress intensity factor is essential. Over the yearsseveral methods have evolved to compute the stress intensity factors for structuralcomponents containing cracks. These methods include analytical as well asexperimental approach. The experimental backtracking approach was used to deriveempirically the stress intensity factors for structural components using the growth ratedata of through-the-thickness cracks for simple geometry subjected to constant-amplitude loading. The finite element method can be applied easily to the propagation of the crackunder constant amplitude loading condition, but it is difficult to use under randomloading condition due to the complexity of modeling the retardation behavior of thecrack (crack closure effect suggested by Elber), which can be introduced byoverloadings.Accurate stress-intensity factor (SIF) solutions are required to conduct thoroughdamage tolerance analyses of structures containing cracks. Exact closed form SIFsolutions for cracks in three-dimensional solids are often lacking for complexconfigurations; therefore, approximate solutions must be used. Over the past twodecades, considerable effort has been placed on developing computationally efficientmethods which provide highly accurate SIF solutions for cracks in three-dimensionalbodies. The purpose of this investigation was to test the accuracy of the crack growthmodels. All necessary parameters, such as material property data, stress intensitysolutions, and the load spectrum, were defined. To determine residual life of damagedstructural components here are used two crack growth methods: conventionalForman`s crack growth method and crack growth model based on the strain energydensity method. The last method uses the low cycle fatigue properties in the crackgrowth model.2. COMPUTATION PROCEDURES TO LIFE ESTIMATIONS An analysis to show complience with a crack growth criterion, as required by adamage tolerance specification, involves following steps: Calculation of the stress intensity factors Calculation of the critical crack size Compilation of crack growth data Adoption of a model for prediction of fatigue crack growth In the design analysis of a slow crack growth structure it is most important tomake correct estimates for the early portion of the crack growth process, because it isthere the life is. In most cases this implies that maximum accuracy is needed for smallcorner cracks. Stress intensity Factor (SIF) is an important fracture parameter to192
  • Some aspects to design of aircraft structures with respects fatigue and fracture mechanicspredict the behaviour of the joint in the presence of cracks using Linear ElasticFracture Mechanics (LEFM). For a successful implementation of the damage tolerancephilosophy to the design and in-service operation of structures subjected to fatigueloading it is crucial to have reliable crack growth prediction tools3. STRESS INTENSITY FACTOR OF CRACKED LUGS In general geometry of notched structural components and loading is toocomplex for the stress intensity factor (SIF) to be solved analytically. The SIFcalculation is further complicated because it is a function of the position along the crackfront, crack size and shape, type loading and geometry of the structure. In this workanalytic and FEM were used to perform linear fracture mechanics analysis of the pin-lug assembly. Analytic results are obtained using relations derived in this paper. Goodagreement between finite element and analytic results is obtained. It is very importantbecause we can to use analytic derived expresions in crack growth analyses. Lugs areessential components of an aircraft for which proof of damage tolerance has to beundertaken. Since the literature does not contain the stress intensity solution for lugswhich are required for proof of damage tolerance, the problem posed in the followinginvestigation are: selection of a suitable method of determining othe SIF, determinationof SIF as a function of crack length for various form of lug and setting up a completeformula for calculation of the SIF for lug, allowing essential parameters. The stressintensity factors are the key parameters to estimate the characteristic of the crackedstructure. Based on the stress intensity factors, fatigue crack growth and structural lifepredictions have been investigated. The lug dimensions are defined in Fig. 1. =t W Output Set: Case 12Step 0.500000 Contour: Plate Top Equivalent Stre 29.01 27.28 25.55 23.82 22.09 b 2R b 20.36 18.63 16.9 L 15.17 a 13.45 11.72 9.987 8.258 6.529 Y 4.8 3.071 H Z X 1.342 r F Contact pin/load FE model Fig 1 Geometry and loading of lugsTo obtain stress intensity factor for the lugs it is possible to start with generalexpression for the SIF in the next form K YSUM a (3.1) 193
  • S. Maksimovi , I. Vasovi , M. Maksimoviwhere: Y – correction function, a- the crack length. This function is essential indetermining of the the stress intensity factor. Primary, this function depends on stressconcentration factor, kt and geometric ratio a/b. The correction function is defined usingexperimental and numerical investigations. This correction function can be defined inthe next form [13]: 1.12 kt A (3.2) YSUM k Q a A b k er a b (3.3) w 2 R b (3.4) 2 2 2 R 2 R r 3.22 10.39 7.67 (3.5) w w a 3 U 10 b (3.6) Q a 3 10 b 2 2 R 2 R (3.7) U 0.72 0.52 0.23 H H a 1.895 1 b A 0.026 e (3.8) The stress concentration factor kt is very important in calculation of correctionfunction, eq. 3.2. In this investigation. A contact finite element stress analysis was usedto analyze the load transfer between the pin and lug.4. CRACK GROWTH ANALYSIS USING CONVENTIONAL APPROACH For crack growth analysis and fatigue life estimations have been used variousconventional crack growth models. Many of these models achieve correct solutions ofcrack growth analyses for cracked structural elements under cyclic loads of constantamplitude. Hovever, for construction under cyclic loads of variable amplitude in form ofload spectrum such as in aircraft cases it is necessary tu include the effects of shapeof load spectra and its effects of estimation life of structural elements [9]. Forman, Newman and others [8] developed the NASGRO equation, which is anequation often used to describe crack growth. This equation describes the crackgrowth curve in terms of the crack length a , the number of cycles N, the stress ratio R,the stress intensity factor range DK, and material constants, C, n, p, q through best fitsof the da/dN - DK data.194
  • Some aspects to design of aircraft structures with respects fatigue and fracture mechanics p K th n 1 da 1 f K C K q (4.1) dN 1 R K max 1 Kcwhere: a - crack length, N-number of cycles, C, n, p, q – are experimentally derivedmaterial parameters , K is the stress intensity factor (SIF), K th is the threshold stressintensity factor, R is the stress ratio, K c - is the critical stress intensity factor. TheNewman closure function isone of these terms and is define as f : K op max R , A0 A1 R A2 R 2 A3 R 3 ; R 0 (4.2) f K max A0 A1 R ; 2 R 0and the coefficients are given by: 1 2 Smax A0 (0.825 34 0.05 ) cos( 2 0 A1 (0.415 0.071 ) Smsx / 0 A2 1 A0 A1 A3 A3 2 A0 A1 1where: – is the plane stress/strain constraint factor, ( max/ 0) is the ratio of maximumstress to the flow stress. The threshold stress intensity factor range is calculated by thefollowing empirical equation: 1 (1 Cth R ) a 2 1 f K th K th 0 / (4.3) a 1 (1 Ao )(1 R) Relation (4.1) represents one general crack growth models based onconventional approach. This relation can be transformed to conventional Forman`scrack growth model [5]. In region III rapid and unstable crack growth occurs, soForman at al. Proposed equation for region III as well as for region II, [9]. n da C K (4.4) dN 1 R KC Kwhere KC is the fracture toughness. Forman`s equation has been developed to modelof unstable crack growth domain (III).5. CRACK GROWTH MODEL BASED ON STRAIN ENERGY DENSITY METHOD In this work fatigue crack growth method based on energy concept isconsidered and then it is necessary to determine the energy absorbed till failure. Thisenergy can be calculated by using cyclic stress-strain curve. Function between stress 195
  • S. Maksimovi , I. Vasovi , M. Maksimoviand strain, as recommended by Ramberg-Osgood provides good description of elastic-plastic behavior of material, and may be expressed as 1 n/ 2 . (5.1) E 2k /where E is the modulus of elasticity, /2 is strain amplitude and /2 is stressamplitude. Equation (5.1) enables the calculation of the stress-strain distribution byknowing low cyclic fatigue properties. As a result the energy absorbed till failurebecome [10,11] 4 / / Wc f f (5.2) 1 n/where f/ is cyclic yield strength and f/ - fatigue ductility coefficient. Given the fact thatstrain energy density method is considered, the energy absorbed till failure must bedetermined after the energy concept is based on the following fact: The energyabsorbed per unit growth of crack is equal to the plastic energy dissipated within theprocess zone per cycle. This energy concept is expressed by Wc a = p , (5.3)where Wc is energy absorbed till failure, p- the plastic energy and a - the crack length.In equation (5.3) it is necessary just to determine the plastic energy dissipated in theprocess zone p. By integration of equation for the cyclic plastic strain energy densityin the units of Joule per cycle per unit volume 10 from zero to the length of theprocess zone ahead of crack tip d* it is possible to determine the plastic energydissipated in the process zone p. After integration relation of the plastic energydissipated in the process zone becomes 1 n/ K I2 p (5.4) 1 n/ E I n/where KI is the range of stress intensity factor, - constant depending on the strainhardening exponent n/, In/ - the non-dimensional parameter depending on n/.Fatigue crack growth rate can be obtained by substituting Eq. (5.2) and Eq. (5.4) in Eq.(5.3) da 1 n/ 2 / / KI K th , (5.5) dN 4E I n/ f fwhere Kth is the range of threshold stress intensity factor and is function of stress ratioi.e. Kth= Kth0(1-R) , (5.6) Kth0 is the range of threshold stress intensity factor for the stress ratio R = 0 and iscoefficient (usually, = 0.71). Finally number of cycles till failure can be determined byintegration of relation for fatigue crack growth rate196
  • Some aspects to design of aircraft structures with respects fatigue and fracture mechanics ac / / da 4 E I n/ f f N B , B (5.7) a0 KI K th 2 1 n/and KI YS a, (5.8) Equation (5.7) enables us to determine crack growth life of different structuralcomponent. Very important fact is that equation (5.7) is easy for application since lowcyclic material properties (n/, f/, f) available in literature are used as parameters. Theonly important point is stress intensity factor which, depending on the geometrycomplexity and the type of loading, could be determined by using analytical and/ornumerical approaches.6. NUMERICAL VALIDATION To illustrate computation procedures in damage tolerance analysis and residuallife estimations of damged structural components here are numerical examplesincluded.6.1 Life estimation of damaged structural elements Subject of this analyses are cracked aircraft lugs under cyclic load of conctantamplitude end spectra. For that purpose conventional Forman crack growth model andcrack growth model based on strain energy density method are used. Material of lugsis Aluminum alloy 7075 T7351 with the next material properties: 2 m=432 N/mm Tensile strength of material 2 02=334 N/mm KIC=2225 [N/mm3/2] Dynamic material properties (Forman`s constants): CF=3* 10-7, nF=2.39. / / / Cyclic material properties: f =613 P , f =0.35, n =0.121. The stress intensity factors (SIF`s) of cracked lugs are determined for nominalstress levels: g = max=98.1 N/mm2 and 2 min=9.81 N/mm . These stresses aredetermined in net cross-section of lug. The corresponding forces of lugs are definedas, Fmax= g (w-2R) t = 63716 N and Fmin= 6371.16 N, that are loaded of lugs. Forstress analyses contact pin/lug finite element model is used.For cracked lugs defined in Table 6.1, with initial cracks a0, SIF`s are determined usingfinite elements, Table 6.2. To obtain high-quality results of SIF`s cracked lugs aremodeled by singular finite elements around crack tip. 197
  • S. Maksimovi , I. Vasovi , M. Maksimovi 83.3 t = 15 mm 160 R20 a 44.4 R41.6 F = 6371.6 daN Fig 6.1: Geometry of cracked Fig. 6.2. Finite Element Model of cracked lug lug 2 with stress distributionTable 6.1 Geometric parameters of lugs [13] Lug Dimensions [mm] No. 2R W H L t 2 40 83.3 44.4 160 15 6 40 83.3 57.1 160 15 7 40 83.3 33.3 160 15 The stress intensity factors of cracked lugs are calculated under stress level: g= max=98.1 N/mm2, or corresponding axial force, Fmax= g (w-2R) t = 63716 N. Inpresent finite element analysis of cracked lug is modeled with special singular quarter-point six-node finite elements around crack tip, Fig. 6.2. The load the model, aconcentrated force, Fmax, was applied at the center of the pin and reacted at the otherand of the lug. Spring elements were used to connect the pin and lug at each pairs ofnodes having identical nodal coordinates all around the periphery. The area of contactwas determined iteratively by assigning a very high st