Structural Robustness against Accidents

STRUCTURAL ROBUSTNESS
AGAINST ACCIDENTS
Franco Bontempi*, Marco Lucidi, Pier Luigi Olmati
*PhD, PE, Professor of Structural Analysis and Design
School of Civil and Industrial Engineering
University of Rome La Sapienza
Rome - ITALY
1

LINEAR interactions NONLINEAR
LOOSEcouplingsTIGHT
3

Design Complexity
(Optimization)
Loosely – Tightly Couplings (Interactions)
Nonlinear–LinearBehavior
9

SYSTEM CONTINGENCY
Nature
Characteristics
Weakness
…
Nature
Characteristics
Strengths
…
COUPLINGS / INTERACTIONS
NONLINEARITY
10

Structural Robustness
=
Structural Survivability
13

Structural Sistems Performance
15

Levels of Structural Crisis
UsualULS&SLS
VerificationFormat
Assessment
1st level:
Material
Point
2nd level:
Element
Section
3rd level:
Structural
Element
4th level:
Structural
System
17

Structural Robustness (1)
ATTRIBUTES
RELIABILITY
AVAILABILITY
SAFETY
MAINTAINABILITY
INTEGRITY
SECURITY
FAILURE
ERROR
FAULT
permanent interruption of a system ability
to perform a required function
under specified operating conditions
the system is in an incorrect state:
it may or may not cause failure
it is a defect and represents a
potential cause of error, active or dormant
THREATS
NEGATIVE CAUSE
STRUCTURALQUALITY
more robust
less robust
18

•Capacity of a construction to show regular
decrease of its structural quality due to
negative causes.
•It implies:
a) some smoothness of the decrease of
structural performance due to
negative events (intensive feature);
b) some limited spatial spread of the
rupture (extensive feature).
19

Robustness comparison
5
8 6 97
12 10 1311
4 2 31
l
VIERENDEEL STRUCTURE ROBUSTNESS
00,51
0 1 2 3 4 5 6 7 8 9 10
Damage Level
PU [ad] MAX MIN
6
6
1
2
3
7 8 9 4 5 1010
TRUSS STRUCTURE ROBUSTNESS
00,250,50,751
0 1 2 3 4 5 6 7 8 9
Damage Level
PU [ad] MAX MIN
High element connectionHigh element number
14
5 3 1 2 4 6 8
l
7 5 3 1 2 4 6 8
14
11 9 1210
18 20 1921
13 15 1617
l
STATIC
INDETERMINANCY
i = 4 i = 12
9
3
12
1 2
11 5 6 13 14
17
22

•Capacity of a construction to show regular
decrease of its structural quality due to
negative causes.
•It implies:
a) some smoothness of the decrease of
structural performance due to
negative events (intensive feature);
b) some limited spatial spread of the
rupture (extensive feature).
26

Bad vs Good Collapse
STRUCTURE
& LOADS
Collapse
Mechanism
NO SWAY
“IMPLOSION”
OF THE
STRUCTURE
“EXPLOSION”
OF THE
STRUCTURE
is a process in which
objects are destroyed by
collapsing on themselves
is a process
NOT CONFINED
SWAY
27

Cascade Effect / Domino Effect
• A cascade effect is an inevitable and sometimes
unforeseen chain of events due to an act affecting a
system.
• In biology, the term cascade refers to a process that,
once started, proceeds stepwise to its full, seemingly
inevitable, conclusion.
• A domino effect or chain reaction is the cumulative
effect produced when one event sets off a chain of
similar events.
• It typically refers to a linked sequence of events where
the time between successive events is relatively small.
28

High Probability Low Consequences
HPLC
events
31

LowProbabilityHighConsequences LPHC
events
32

HPLC
High Probability
Low Consequences
LPHC
Low Probability
High Consequences
release of energy SMALL LARGE
numbers of breakdown SMALL LARGE
people involved FEW MANY
nonlinearity WEAK STRONG
interactions WEAK STRONG
uncertainty WEAK STRONG
decomposability HIGH LOW
course predictability HIGH LOW
HPLC – LPHC EVENTS
35

RUNAWAY (1)
effect
time
decomposability
course predictability
36

EFFECT
RUNAWAY(2)
decomposability
course predictability
37

Framework of Analysis
HPLC
Eventi Frequenti con
Conseguenze Limitate
LPHC
Eventi Rari con
Conseguenze Elevate
Complessità:
Non linearita’,
Interazioni,
Incertezze
Impostazione
del problema:
Deterministica
Probabilistica
ANALISI
QUALITATIVA
DETERMINISTICA
ANALISI
QUANTITATIVA
PROBABILISTICA
ANALISI
PRAGMATICA
CON SCENARI
38

A Black Swan is an event with the following three attributes.
1. First, it is an outlier, as it lies outside the realm of regular expectations,
because nothing in the past can convincingly point to its possibility.
Rarity -The event is a surprise (to the observer).
2. Second, it carries an extreme 'impact'.
Extreme impact - the event has a major effect.
3. Third, in spite of its outlier status, human nature makes us concoct
explanations for its occurrence after the fact, making it explainable and
predictable.
Retrospective (though not prospective) predictability -
After the first recorded instance of the event, it is rationalized by hindsight,
as if it could have been expected; that is, the relevant data were available
but unaccounted for in risk mitigation programs.
References: Taleb, Nassim Nicholas (April 2007). The Black Swan: The Impact of the Highly Improbable (1st ed.).
London: Penguin. p. 400. ISBN 1-84614045-5.
Black Swan Events
39

HAZARD
IN-DEPTH
DEFENCE
HOLES DUE TO
ACTIVE ERRORS
HOLES DUE TO
HIDDEN ERRORS
FAILURE PATH
44

=
45

48
• The cladding system is a crucial component of the
building for protecting the inside against external
explosions.
• In this experimental program three specimens are
tested.
• The first specimen (A) is conventionally designed with
a minimum amount of required reinforcement (0.15
%),
• the second specimen (B) is designed to achieve a
specific maximum deflection if subjected to a specific
blast demand,
• and the third specimen (C) is equal to the specimen
(B), subjected to a larger explosive charge.
Introduction

49
• All the specimens are horizontally simply supported and
the explosive charge is orthogonally suspended at 1500
mm from the center of the exposed blast side of the
concrete panel.
• The experimental texts were conducted at the facility of
the R.W.M. ITALIA s.p.a. (www.rwm-italia.com) at
Domusnovas (Sardinia - Italy).
• Finite Element Analyses (FEAs) are carried out with the
explicit Finite Element (FE) code LS-Dyna® for predicting
the deflection of the precast panels. Solid elements are
utilized for modeling the concrete instead beam
elements are adopted for modeling the reinforcement.
Contact algorithm for modeling the boundary conditions
is utilized.
Introduction

55 Test matrix
The specimens are simply horizontally supported, and the supports are made by concrete blocks.
The explosive charge is suspended at 1500 mm from the panel surface and it is orthogonal with the
center of panel surface. The supports are 400 mm high and the lateral open space between the
panels and the ground is closed by sandbags (see Fig.1). In this way the shock wave would be not
able to diffract on the back face of the panels.
Fig. 2 - Longitudinal section of the testing site
The explosive, provided by the R.W.M. ITALIA
s.p.a., is the PBXN-109 (composed by the 64.12 %
of RDX, the 19.84 % of Aluminum, and the 16.04 % of
Binder)
Panel
t a b c
[mm] [mm] [mm] [mm]
A 150 1550
B 200 1160 1550
C 200 880 1550 2030
Thickness of the panels and position of the meter devices
Two kinds of displacement meter are used and
provided by the R.W.M. ITALIA s.p.a.: the comb
device and the coaxial tube device, as well as
shown in figure.

58 Experimental results
Specimen A
The specimen A is designed with the minimum reinforcement for a concrete cladding wall panel.
The deflection of the specimen A reached the full
scale value of the coaxial tubes device. The panel
during the deflection impacted the external tube
of the coaxial tubes device and the panel stopped
its deformation.
The maximum and residual
deflection of the panel is so
108 mm.

Specimen B
This panel is designed to achieve a specific performance under a blast load, so the specimen B is
designed for blast (the amount of explosive is the same of the specimen A, 3.5 Kg TNTeq).
The maximum and the residual deflection achieved by the
specimen B is of 70 mm and 35 mm respectively.
The specimen B shows a ductile failure with a diffuse crack patterns on the central one third of
the panel span (the major cracks are 3 mm width). However, some radial crack patterns are present,
this is due to the short stand-off distance, and develops a flexural mechanism as designed.

Specimen C
The specimen C is equal to the specimen B but the blast demand is greater for leading significant
damages to the panel without reach a failure. The specimen C would test the blast resisting range
of the panel over the limit of his specific design (the amount of explosive is increased at 5.5 Kg TNTeq).
The maximum and the residual deflection are of
123 mm and 82 mm respectively.
Heavy crack patterns are assessed. Along the mid-span of the panel diffuse cracks are present
with significant width until 10 mm. Moreover some cracks at the mid-span pass through the
panel cross section thickness (maximum width of the crack passing is the 5 mm)

61 Numerical investigation
In order to reproduce the experimental tests numerically the explicit Finite Elements (FE) code LS-
Dyna® is adopted.
To simulate physic phenomena, in this study a “Lagrangian” method is adopted and the uncoupled
approach is preferred, thus the blast load is computed and applied independently from the
structural response of the concrete wall panels.
The FE models have constant solid stress elements for the concrete, and beams elements for the
reinforcement. To bond the beams and solid elements, the LS-Dyna® keyword Constrained
Lagrange in Solid is used.
For reducing the computational effort the model of the specimens are only a square part of the
panel, so opportune boundary conditions are provided.
Support
Blast load BC
Panel
Detail view of the finite element model
The concrete supports of the panels are explicitly
modeled and the contact between the panel and the
support is provided by the LS-Dyna® keyword
Contact Automatic Surface to Surface. Furthermore,
in order to take into account correctly the clearing
effect the boundary conditions for the blast load are
provided; a rigid surface modeling the other three
quarter of the panel is added.

Numerical investigation
The material constitutive law of the reinforcement is the kinematic hardening plasticity model
and the strain rate effects is accounted for by the Cowper and Symonds strain-rate model.
The parameters selected for this model are:
• D=500 s-1;
• q=6;
• steel Young’s modulus=200 Gpa;
• Poisson coefficient=0.3;
• yielding stress=543 MPa 0
2
4
6
8
0.001 0.1 10 1000
DIF[-]
Strain-rate [1/sec]
Compressive
Tensile
A
B
C
Reflecting surface
Reflecting surface
Dynamic Increase Factor relation
1
Density
2.248 lbf/in4
s2
2.4*103
kg/m3
fcm
4060 psi
28 N/mm2
Cap
retraction
active
Rate
effect
active
Erosion none
Input data for the concrete model
Concrete model input data
Due to the walls delimiting the testing site, multiple reflections of the original shock wave
occurred. Consequently the blast load on the specimens is greater than the blast load on a
specimen tested in an open space.
Using the uncoupled approach the
image charge method (instead of the
ALE method) provides acceptable
results without increasing the
computational effort.
Elementary scenario of reverberating
Shock waves
Image charge
side
Stand-off α
[m] [degrees]
West 6009 27
North 4705 35
South 4705 35
East 13505 13
Table 1: Image charge positions1
62

0
40
80
120
160
200
240
280
0 0.05 0.1 0.15 0.2 0.25
δ[mm]
time [sec]
Experimental
Numerical
Specimen A
δmaxδres
0
10
20
30
40
50
60
70
80
0 0.05 0.1 0.15
δ[mm]
time [sec]
Numerical
Experimental
Specimen B
δmax
δres
(a) (b)
0
20
40
60
80
100
120
140
0 0.05 0.1 0.15
δ[mm]
time [sec]
Numerical
Experimental
Specimen C
δmax
δres
0
40
80
120
160
200
240
280
0 0.05 0.1 0.15 0.2 0.25
δ[mm]
time [sec]
Specimen A
Specimen B
Specimen C
(c) (d)
Figure 1: Experimental and numerical mid-span displacement

The following table shows the summary of the results for each specimen reporting both the
maximum and the residual deflections of the experimental and numerical investigations. Moreover
the support rotation θ is shown for both the experimental and numerical investigations.
Specime
n
Experimental Numerical Experimental Numerical
δmax
[mm]
δres
[mm]
δmax
[mm]
δres
[mm]
θmax
[deg]
θres
[deg]
θmax
[deg]
θres
[deg]
A 108* 108* 244 240 4.0* 4.0* 8.9 8.8
B 70 35 58 50 2.6 1.3 2.1 1.8
C 123 82 114 106 4.5 3.0 4.2 3.9
* Full scale value
Looking at the maximum support rotations
experimentally assessed:
• specimen B goes over the Moderate
Damage CDL but does not exceed the
Heavy Damage CDL;
• specimen C does not exceed the Heavy
Damage CDL
Component damage levels θ [degree] μ [-]
Blowout >10° none
Hazardous Failure ≤10° none
Heavy Damage ≤5° none
Moderate Damage ≤2° none
Superficial Damage none 1
1
<<<<<<<<<<<<<
Component damage levels (CDLS) for
U.S. antiterrorism performance-based
blast design approach

The below figure shows the simulated crack patterns of the three specimens; in view is the brittle
damage parameter in the range from 0.95 to 1
Specimen CSpecimen BSpecimen A
(a)
Specimen C
Specimen B
Specimen A
(b)
Figure 1: Crack patterns of the specimens: (a) back view, (b) longitudinal view

=
71

Keywords
•Complexity
• Predictability
•Dependability
•Structural Robustness
•Accident Scenarios
• Back-analysis
• Learning
72

Next Dating
SPRING 2017
KICK-OFF MEETING ON
EXPLOSION GROUP IN ITALY
UNIVERSITY OF ROME LA SAPIENZA
SCHOOL OF CIVIL AND INDUSTRIAL ENGINEERING
analisi-strutturale@uniroma1.it
Scientific Coordination by Franco Bontempi
Technical Coordination by Dario Porfidia
73

ATTRIBUTES
THREATS
MEANS
RELIABILITY
FAILURE
ERROR
FAULT
FAULT TOLERANT
DESIGN
FAULT DETECTION
FAULT DIAGNOSIS
FAULT MANAGING
DEPENDABILITY
of
STRUCTURAL
SYSTEMS
AVAILABILITY
SAFETY
MAINTAINABILITY
permanent interruption of a system ability
to perform a required function
under specified operating conditions
the system is in an incorrect state:
it may or may not cause failure
it is a defect and represents a
potential cause of error, active or dormant
INTEGRITY
ways to increase
the dependability of a system
An understanding of the things
that can affect the dependability
of a system
A way to assess
the dependability of a system
the trustworthiness
of a system which allows
reliance to be justifiably placed
on the service it delivers
SECURITY
High level / active
performance
Low level / passive
performance
88

Prevention
Prorection
Risk = Probability · Magnitudo
89

Risk=Probability·Magnitudodo
discretizationinlog-logplane
90

Option 1 – Risk avoidance, which usually
means not proceeding to
continue with the system;
this is not always a feasible
option, but may be the only
course of action if the
hazard or their probability of
occurrence or both are
particularly serious;
Risktreatment
92

Option 2 – Risk reduction, either through
(a) reducing the probability
of occurrence of some
events, or (b) through
reduction in the severity of
the consequences, such as
downsizing the system, or
(c) putting in place control
measures;
Risktreatment
93

Option 3 – Risk transfer, where insurance
or other financial
mechanisms can be put in
place to share or completely
transfer the financial risk to
other parties; this is not a
feasible option where the
primary consequences are
not financial;
Risktreatment
94

Risktreatment
Option 4 – Risk acceptance, even when it
exceeds the criteria, but
perhaps only for a limited
time until other measures
can be taken.
95

Structural Robustness against Accidents

Recommended

Recommended

More Related Content

What's hot

What's hot (20)

Viewers also liked

Viewers also liked (20)

Similar to Structural Robustness against Accidents

Similar to Structural Robustness against Accidents (20)

More from Franco Bontempi Org Didattica

More from Franco Bontempi Org Didattica (20)

Recently uploaded

Recently uploaded (20)

Structural Robustness against Accidents