The document does not provide enough context to determine what "Adv Serv Ltd" stands out. It appears in a citation for a conference proceedings, but the full abbreviation is not defined.

1. Creep-fatigue Interaction Testing
Fatigue lives in metals are nominally time independent
below 0.5 To.,,. A: higher temperatures, fatigue lives
are altered due to rime-dependent, thermally activated
creep. Conversely, creep rates are altered by super.
imposed fatigue loading. Creep and fatigue generally
interact synergistically to reduce material lifetime.
Their interaction, therefore, L_ of £mpbrtanc_) to
structural durability of high-temperature structures
such as nuclear reactors, reusable rocket engines, gas
turbine engines , terrestrial steam turbines, pressure
vessel and piping components, casting dies, molds for
plastics, and pollution control devices. Safety and life-
cycle costs force designers to quantify these inter-
actions. Analytical and experimental approaches to
creep-fatigue began in the era following World War 1I.
In this article experimental and life prediction
approaches ate reviewed for assessing creep-fatigue
interactions of metallic materials. Mechanistic models
are also discussed briefly.
1. I'est Facilities
Modern creep-fatigue testing facilities are sophisti-
cated closed-loop,servo-con_olled, electro-hydraulic
fatiguetestingmachines('I-Ialfordetal._ adapted
to high-temperature operation (Fig.---l_. Typical
machinesconsistoftwo orfourposts,a rigidloading
frame, and a hydraulic cylinder mounted in the center
ofthebase.The cylindertransmitsh_'draulic pressure
from a pump throughan electromcscrvovalve.A
specimengripismounted on theupperend of the
cylinderram, and anotherisattachedto a loadcell
mounted to the upper cross member to measure axial
loadtransmittedthroughthespecimen.Modern grips
utilizehydraulicsto alignand securethespecimen.
Alignment of the specimen and loadingaxes is
importantto minimizebending loads and avoid
buckling.Specimenshavea uniformlyreducedcross-
sectionthatisas uniformlyheatedas possibleand
experiencesthe cyclicstressesand strains.Cyclic
strains are measured and controlled with commercially
availableextensometers with resolutionson the order
of I0microstrain,
A specimenheatingsystemisthe core of high-
temperature testing. Test machine components such as
the extensometer sensing element, specimen grips,
load cell, and all hydraulic components must be kept
cool.The mostversatile heating systemusesinduction
coilssurroundingthespecimen.High-frequencycur.
rentpassingthroughthecoilsinduceseddycurrentsat
the surfaceof the specimen. Specimen heating results.
The high thermal conductivity of most alloys results in
a specimen with minimal thermal gradients in the
radial direction. Coil spacing controls axial thermal
gradients.Threeindependentlycontrolledcoilscanbe
sx0015
sx0020
This report is a preprint of an article submitted to a journal for
publication. Because of changes that may be made before formal
publication, this preprint is made available with the understanding
that it will not be cited or reproduced without the permission of the
author.

2. /
C:/jan/emr409037 Apr04-pro p lc- lm 3 _ 2)
i C
I
i/
Error Specimen_
Gain I
f I
[ .. ve ,o.v, I [
,(Servovalve
Figure 1
Modern creep-fatigue testing machine.
employed to minimize the axial thermal gradient. Coil
spacing will allow visual observation of the specimen
and permits attachment of an extensonjgter sporting
ceramic extension rods that transmiq_gage section
displane.mmz_ to the more distant sensini[ element. A
thermocouple held in direct contact with the specimen
surface supplies feedback for temperature control.
Commercial noncontacfing optical pyrometers are
also available. Alternate heating techniques, such as a
combination of induction and radiation, are required
for low conductivity materials such as ceramics. A
small metallic susceptor surrounds the specimen and is
heated by induction. The susceptor then radiates heat
to the specimen. The low thermal mass of this system
permits reasonably fast temperature changes. Radi-
ation furnaces are quite common. They use coiled
Nichrome wire or siliconcarbideelements or even
banks of quartz lamps. Furnaces offerhigh thermal
stability because of the large thermal mass, but are
slow to respond to intentional temperature changes
during thermomechanical testing. Direct resistance
heating also has been used successfully, but remains
uncommon. High current (low voltage) passed
sxO025
i
|$

3. /
C:/jan/emr409037 Apr04-pro p Ic-lm 4 (X 31
"_'.rough the specimen results in seE-heating due to the
specimen's electrical resistivity. This technique is the
=ost di_cult to conuol t_causeoftbe _l"enfly high
£_e:mal response. Heating of the test section does not
reiy on the thermal conductivity of the specimen;
cooling, however, does. Maintaining an accurate
constanttemperatureisdilr_-ult.
Modern testfacilities arc typicallyinappropriatefor
long-term(_ 1 month)testing.Lessexpensive,more
reliabl¢, screw-driven machines are sul:_riorfor
purpose. If long-termcreep-fatigue resistanceis a
designrequirement,long-termtestresultsarenecess-
ary to verifylifepredictionmodels for accurate
extrapolation.
_0030
2. MeehanDma and Life Prediction Modela sx0035
Fatigue crack initiation and early growth mechanisms
differ substantially from those for creep cracking.
Fatigue damage is time-lndependent to-and-fro crys-
tallographic slip, most evident at free surfaces. Creep
damage occurs throughout the volume of stressed
material and requires the thermally activated diffusion
of atomic vacancies. Creep damage due to grain
boundary sliding, however, intersects free surfaces in
much thesameway as fatigueslip.
The additionofcreeptoan otherwisepuretension/ sx0040
compressionfatiguecyclecan be accomplishedin
threedistinctlydifferentways as shown below.A
baselineconditionof purefatiguecyclingwillhave
negligiblecreep presentif a high enough cyclic
frequency(_>IHz) isimposed.
- Constantstrainingratelow enoughto introduce
creepinadditiontocrystallographicslip.
- Constantstressascreepoccurswithtime.
- Constanttotalstrainwhilestressrelaxes,con-
vertingelastictocreepstrain(leftpanelsof Fig.2).
Sinceany one of thesethreeways of introducingsx0045
creepintoa cyclecouldbe appliedto tensiononly,
compressiononly,or bothtenslonand compression,
thereemerge,in theextreme,onlyfourbasiccom-
binationsofcompletelyreversedcreep-fatiguestress-
straincycles:
- Tensileplasticstrain(p)reversedby compressive
plasticstrain(p),i.e.,(pp)
- Tensileplasticstrain(p)reversedby compressive
creepstrain(c),i.e.,(pc)
- Tensilecreepstrain(e)reversedby compressive
creepstrain(c),i.e.,(cp)
- Tensilecreepstrain(c)reversedby compressive
plasticstrain(p),i.e.,(cc)
The rightpanelsof Fig.2 illustratetypicalstress- sx0050
strainhysteresisloopsfor strainhold typecycles.
Invariablycreep-fatiguecyclesinvolve(p.v)straining
in additionto (cp),(pc),or (cc)components.Such
cyclesdemand use ofservo-controlledsystemswith
extensometersand state-of-the-artfunctiongener-

4. /
C:/jan/emr409037 Apr04-pro p It- lm 5 (X 4)
,',Slnm c, $1_u
, , .'
_., / ", / ', , _
84
B2
81
,:%/ :Y___.a'
/'qtarc 2
Creep.fatigue test cycles.
ators. The cycles are simplifications of actual complex
service cycles. Experimental replication of service
histories would require complex programming achiev-
able through a computer-controlled servo system.
Usually, life prediction models are relied upon to link
single lah_,to_ ¢_e.s _th mlggi¢ se_,d¢.¢t'_,_.s.
The deformation, crack initiation, and early growth
mechanisms, in general, will be different for each of
these discrete cycles. Schematic models developed
to illustrate these differences have been compared to
experimental observations for a variety of alloys
(Manson and Halford [I]_. Of the more than 100
creep-fatigue life predicuon models proposed since the
1950s,onlyabout10% havebeenappliedextensively
enoughtowarranttheirreview(Miller[_ Halford
et al. _. The three most widely used-aTffthe time-
and cyc--'ffd-fractionrule (ASME _), strai_ange
partitioning (SLIP), and damage--_-echanics. Few
specitically address inevitable oxidation interactions.
The ASME Code Case is extremeIy conservative and
does not directlyaddresswhen hilurewould occur.
Thatconditionisacceptablebecausepowerplantsdo
not havethesame stringentweightand performance
requirementsas,forexample,aerospacepropulsion
sx0055
. .-_: .?.:_;.;_.:,.. , -" 7, .

5. v
I
C:[_an/emr409037 Apr04-pro p lc- lm 6 (X 5)
systems. The Inner require mtr.h g_mm:r life .1_
accuracy to conserve weight while m2,_._, z a
combination of durability and performance.
3. Co_ Re_ks
Creep-fa6gue interactionis reasonably well uader-
stood at the phenomeuological level.A sisni_:ant
short-termexperimentaldatabase existsand several
satisfactory analytic models are in use for estir_fing
cyclic lives. Modeling of oxldatioo interactlom with
creep-fatigue, lack of long-term databases, and verified
long-term extrapolation procedures remain as im-
portant areas of research.
See also: Creep-Fatigue: Oxidation Interactions
Ix0060
Creep-Fatigue: Oxidation Interactions
Biblioftaph¥
_'_ricaa Societyof Mechaaical Engineers1986 Code Case/¢-
47-2$. ASME, New York
l:I'_ord G R, Lerch B A, McC_w M A 2000 Fatigue, ¢_-eep-
fatigue, and thermomechanical fatigue life testingof alloys In:
(ed.}ASM Han_ook. Vol.8, MechanlcalTeJgln$and
Evaluation. ASM International, Materials Park, OH,Sect. 8B
a_l'_'nson S S, Halford G R 1984 Relation of cyclic loading
pattern to mlcrostmctural fracture in creep-fatigue In: C J
Beevets (ed.) Proc. 2nd Int. Conf. Fatlfw and FaHfue
' Threzho!d_ (Fatigue 84). Engineering Materials Adv Serv Ltd,
Wartey, UK, Vol. 3, pp. 1237-55
l_[il]'erD A, Priest R H, EUison E G 1984 A review of material
response and life prediction techniques under fatigue-creep
loading conditions. High Temp. Mater. Process. 6, 155-94
G. R. Halford
!QA: What is_,dv Serv! short for?