Documente Academic
Documente Profesional
Documente Cultură
Structural damage accumulation and control for life cycle optimum seismic
performance of buildings
Luis Esteva, Orlando Javier Díaz-López, Alberto Vásquez and Jesus Alberto León
Institute of Engineering, National University of Mexico, Ciudad Universitaria, 04510 México, DF, Mexico
in the degradation of the mechanical properties of the system and Deierlein (2013) examine the effectiveness of alternative
components. These criteria are formulated within a life cycle retrofit strategies for the control of seismic risk in old reinforced
optimisation framework, taking into account uncertainties concrete frames, considering different damage levels, including
associated with concepts such as the initial mechanical prop- collapse. The importance of a proper combination of seismic
erties of the system, the characteristics of the processes lead- and environmental hazards in the evaluation of the lifetime
ing to their degradation and the accuracy of the tools used to seismic performance was observed in all the systems consid-
assess the damage levels or the actual values of the mechanical ered.
properties at given times. They focus attention on systems with This article starts with a brief examination of the problems
wide space-and-time variability of live loads, but high intensity faced in the selection of local and global damage indicators that
accidental excitations, such as earthquakes or hurricanes, are can be used for the development of practically applicable crite-
not considered. ria for the assessment of the influence of damage accumulation
Moan (2005, 2011) presents similar ideas for specific on the seismic vulnerability functions of building structures.
applications to offshore platforms; inspection, maintenance and Some quantitative information is then shown about this influ-
repair decisions depend on specified upper bounds to the ence for a family of typical systems of this kind. This is followed
cumulative failure probability for given time intervals. Stewart by the mathematical models adopted here for the life cycle
(2006) presents some illustrative examples about the potential optimisation process and the results of several applications are
of using spatial variability analysis to obtain improved esti- examined, with an emphasis on their implications for the opti-
mates of the expected maintenance costs based on pre-specified mum engineering decisions. Finally, some practical problems
values of local damage indicators. Sanchez-Silva, Klutke and are discussed about the assessment of the damage indicators
Rosowsky (2011) propose a life cycle expected performance for specific systems and desirable lines of future research are
analysis, assuming that repair and maintenance decisions are identified.
taken with information about the global capacity of the system Seismic vulnerability functions for the systems studied in the
that results from its initial value, corresponding to the undam- following are determined with the aid of previously developed
aged condition, and the reductions produced by the damage criteria and models (Esteva et al., 2011; Esteva & Díaz-López,
accumulation process. 2006; Esteva, Díaz-López, & Ismael-Hernández, 2010). Atten-
Probabilistic models are proposed for this process, but tion is focused on the analysis of the influence of some varia-
no information is provided about the detailed degradation bles on those functions and on the resulting life cycle present
processes in the individual members that lead to the global values of expected damage costs. The main independent varia-
capacity reductions considered. Chiu, Hsiao, and Jean (2013) bles considered are: (a) the nominal value of the lateral strength
proposed a procedure for assessing expected costs and benefits reduction factor specified to allow different levels of non-linear
of seismic retrofitting in multi-storey buildings. They adopt the behaviour, (b) initial damage accumulated from the response
assumption that a building is restored to its original state after to previous seismic excitations, (c) use of energy dissipating
each seismic event, regardless of the damage level reached, thus devices, (d) threshold damage values for repair of the structural
eliminating the possibility of damage accumulation. In a later system or replacement of energy dissipating devices. Expected
study, Chiu (2014) presents a criterion for the reliability-based performance is expressed in terms of the following indicators:
service life assessment of deteriorating reinforced concrete (a) global and local fatigue damage indexes, (b) vulnerability
buildings, including the influence of cumulative damage; dam- functions, including final expected damage and failure proba-
age states are determined by visual inspection, cumulative dam- bility in terms of intensity and initial damage, (c) expected fail-
age evolution is expressed in terms of a Markov chain model, ure rates for given values of initial damage, (d) optimum life
taking into account degradation of structural properties due to cycle repair and replacement strategies for structural frames
corrosion. The study is focused on a life cycle analysis based with or without energy dissipating devices.
on the probabilistic estimation of future events, but decisions
to be made in the future do not take into account any updated 2. Seismic vulnerability functions
information about the observed seismic demand and damage In practical engineering applications related to perfor-
processes. However, it is recognised that the proposed method mance-based earthquake-resistant design, the estimation of the
provides useful information related to maintenance based on seismic reliability of non-linear multi-storey buildings for the
both serviceability and safety. given values of the ground motion intensity is ordinarily based
The joint influence of seismic and environmental hazards in on a measure of the probability that the lateral deformation
the process of damage accumulation on the life cycle expected capacity of the system, determined by conventional pushover
performance of bridges was studied by Akiyama, Frangopol, analysis, is greater than the peak value of the corresponding
and Matsuzaki (2011), who estimated seismic demands and non-linear response demand for an ensemble of earthquake
performance of a reinforced concrete bridge pier, taking excitations with the specified intensity. However, probabilistic
into account the deterioration of its mechanical properties estimates of the deformation capacities of multi-storey build-
resulting from the hazard associated with airborne chlorides. ings obtained by means of pushover analysis are tied to severe
Similar studies are presented by Ghosh and Padgett (2011) for limitations because according to this approach it is not possible
ageing bridges, by Biondini, Camnasio, and Palermo (2014) to account for (a) the influence of cumulative damage associ-
for a four-span continuous bridge, and by Celarec, Vamvat- ated with the cyclic response, and (b) the dependence of the
sikos, and Dolšek (2011) for multi-storey reinforced concrete lateral deformation capacity on the response configuration of
buildings, considering the influence of corrosion over time. Liel the system when it approaches failure. This has fostered the
850 L. ESTEVA ET AL.
into an updated mechanical model of the whole system and 3.3. Spatial distribution of damage: its implication on
used to assess the expected influence of the residual damage seismic vulnerability functions and long-term expected
on the potential increase of its vulnerability function under performance
the action of a future external excitation, in this case an earth- Spatial distribution of damage throughout a structural
quake ground motion of unknown intensity. Decisions about system is often characterised by significant irregularities.
repair and maintenance actions can then be made in terms of Damage concentrations tend to occur at those elements
risk-acceptance and cost-benefit criteria, formulated within a or sub-assemblages with the highest ratios of the magni-
life cycle framework. tudes of the internal forces associated with the accidental
In many cases, visual inspection may not provide enough excitations (earthquake ground motions, in this example)
quantitative information about the local damage levels and, as to those corresponding to the permanent loads. Because the
a consequence, about their potential impact on the increase mentioned irregularities determine the dominant potential
of the seismic vulnerability function of the system. This may failure mechanisms of the system under the action of future
occur, for instance, if main structural members are hidden by earthquakes, the functions relating the expected increments
some architectural finishing elements the removal of which in the seismic vulnerability functions to the global indica-
might be considered too expensive; it may also happen in hys- tors of initial damage are affected by very large uncertain-
teretic energy dissipating devices, which may not show any ties. It is reasonable to expect that significant reductions in
evidence of damage before the failure condition by low cycle these uncertainties can be achieved through the use of vec-
fatigue is reached. In many of these cases it would be justified tor damage indicators, integrated by components describ-
to make preliminary estimates of local and global damage after ing simultaneous values of local damage levels at individual
the occurrence of moderate and high intensity events. These structural members or segments. This is another concept
estimates can be derived from instrumental response records deserving attention in research programs in the near future.
obtained during those events or from simulated time histories The influence of damage accumulation on the expected
of local strains or distortions at critical sections, generated by seismic performance of the systems can be expressed in terms
means of mathematical models of the system; they can also be of the increasing values of the expected failure rates νF, deter-
based on structural health monitoring studies performed after mined in accordance with the following equation:
the occurrence of excitations deemed to be significant.
If these damage estimates lead to significant increments in
( )
∞ d𝜈Y y
∫0
the values of the seismic vulnerability function, further actions 𝜈F = − PF (y)dy (2)
dy
should be taken to improve the knowledge of the decision-
maker about the local damage conditions and their possible where νY(y) is the seismic hazard function at the site, expressed
impact on the updating of the latter function. The first step in terms of the mean annual rate of occurrence of intensities
might be to perform a detailed visual inspection, covering all greater than y, and PF(y) is the probability of failure given that
the critical sections that form part of each relevant global fail- the intensity is equal to y. This probability grows with time, as
ure mechanism. Final estimates of the vulnerability function the accumulated damage grows.
should include the information coming from the visual inspec-
tion as well as that derived from the previous steps. For this 4. Life cycle analysis including damage accumulation
purpose, probabilistic models of the epistemic uncertainties Esteva et al. (2011) present the basic concepts of a life cycle
associated with the local damage levels and the corresponding framework (Rackwitz, 2000; Rosenblueth, 1976) normally
mechanical properties should be developed in accordance with adopted for the formulation of optimum engineering deci-
Bayesian probability concepts. sions applicable to the design, construction and maintenance
The information needed for the establishment of these mod- of structural systems. Within this framework, a utility function
els might be partially derived from observations about physical as given by Equation (3) is defined as the sum of the present
evidences of local damage and values of fatigue-damage indica- values of the expected benefits and costs to be generated since
tors resulting either from recorded structural response histories the construction of the system to its eventual demolition or
or from those determined by means of mathematical models. abandonment:
852 L. ESTEVA ET AL.
[ ∞
]
−𝛾T
∑
U = B − C0 − E Di i (3)
i=1
Figure 6. Values of damage and normalised damage indexes for system B10-4
Figure 13. Values of damage and normalised damage indexes for system BD2. Figure 14. Values of damage and normalised damage indexes for system BD4.
STRUCTURE AND INFRASTRUCTURE ENGINEERING 857
Figure 15. Variation of expected damage in the frame along the building height
for system BD4.
(2) For the cases studied, the variability of the local/global damage Chiu, C.K., Hsiao, F.P., & Jean, W.Y. (2013). A novel lifetime cost-benefit
analysis method for seismic retrofitting of low-rise reinforced concrete
indexes is very similar, regardless of the lateral-force-reduction buildings. Structure and Infrastructure Engineering, 9, 891–902.
index Q assumed in the structural design. Cornell, C.A. (1969). A probability-based structural code. Journal of the
(3) Expected failure rates for the systems studied range from 10−5 American Concrete Institute, 66, 974–985.
to 10−4, for the condition of no initial damage. For initial values Corotis, R.B., Hugh Ellis, J., & Jiang, M. (2005). Modeling of risk-based
of IDFN equal to .5, these rates may increase by a factor as high inspection, maintenance and life-cycle cost with partially observable
Markov decision processes. Structure and Infrastructure Engineering, 1,
as 100. 75–84.
(4) Systems with energy dissipating devices have lower vulnerabil- Cosenza, E., Manfredi, G., & Ramasco, R. (1993). The use of damage
ity functions than conventional structural systems with equal functionals in earthquake engineering: A comparison between different
values of the lateral strength and stiffness. methods. Earthquake Engineering and Structural Dynamics, 22,
(5) Present values of life cycle expected damage cost functions 855–868.
Ellingwood, B.R. (2005). Risk-informed condition assessment of civil
are not very sensitive to the damage threshold repair index for infrastructure: State of practice and research issues. Structure and
systems designed for Q = 2, while pronounced variations are Infrastructure Engineering, 1, 17–18.
observed for systems designed for Q = 4, both for conventional Esteva, L., Campos, D., & Díaz-López, O. (2011). Life-cycle optimisation
frames and for systems with energy dissipating devices. in earthquake engineering. Structure and Infrastructure Engineering, 7,
33–49.
Esteva, L., & Díaz-López, O. (2006). Seismic reliability functions for complex
6.2. Recommendations for future studies systems based on a secant-stiffness reduction index. Proceeding of
Near-future research activities should be aimed at improving International Federation for Information Processing, WG7.5 Working
engineering tools for the assessment of the influence of initial Conference, Kobe, Japan.
structural damage on the seismic reliability and expected per- Esteva, L., Díaz-López, O., & Ismael-Hernández, E. (2010). Seismic
formance of multi-storey buildings and for the selection of effi- vulnerability functions of multi-storey buildings: Estimation and
applications. Structure and Infrastructure Engineering, 6, 3–16.
cient repair and maintenance strategies. Some specific concepts Esteva, L., & Díaz, O. (1993). Influence of cumulative damage on the seismic
requiring attention are: reliability of multistory frames. Proceeding of Fifth IFIP WG7.5 Working
(1) Efficient criteria and methods to estimate and identify local Conference, Takamatsu, Japan.
damage and its relation with global damage, including both Esteva, L., & Ismael, E. (2003). A maximum likelihood approach to system
instrumental records and analytical models. reliability with respect to seismic collapse. Proceeding of International
Federation for Information Processing, WG7.5 Working Conference,
(2) Selection of ‘efficient’ damage distribution patterns; i.e. loca- Banff, Canada.
tions where damage should be made to concentrate, so that Frangopol, D.M. (2011). Life-cycle performance, management and
they lead to easy, reliable and efficient retrofit actions. optimization of structural systems under uncertainty: Accomplishments
(3) Application of the foregoing concepts to three-dimensional and challenges. Structure and Infrastructure Engineering, 7, 389–413.
Ghosh, J., & Padgett, J.E. (2011). Probabilistic seismic loss assessment
systems subjected to two simultaneous horizontal components.
of aging bridges using a component-level cost estimation approach.
Earthquake Engineering & Structural Dynamics, 40, 1743–1761.
Disclosure statement Ismael, E., & Esteva, L. (2006). A hybrid method for the generation of
No potential conflict of interest was reported by the authors. artificial strong ground motion records. Proceedings of First European
Conference on Earthquake Engineering and Seismology, Geneva,
References Switzerland.
Akiyama, M., Frangopol, D.M., & Matsuzaki, H. (2011). Life-cycle Jeong, G.D., & Iwan, W.D. (1988). The effect of earthquake duration on the
reliability of RC bridge piers under seismic and airborne chloride damage of structures. Earthquake Engineering and Structural Dynamics,
hazards. Earthquake Engineering & Structural Dynamics, 40, 1671–1687. 16, 1201–1211.
Banon, H., Biggs, I.M., & Irvine, H.M. (1981). Seismic damage in León, J.A. (2010). Funciones de daño acumulado para diseño de edificios
reinforced concrete frames. ASCE Journal of the Structural Division, con disipadores de energía [Cumulative damage functions for design
107, 1713–1729. of buildings with energy dissipating devices] (MsC thesis). Graduate
Banon, H., & Veneziano, D. (1982). Seismic safety of reinforced concrete Program in Engineering, National University of Mexico, Mexico
members and structures. Earthquake Engineering and Structural (in Spanish).
Dynamics, 10, 179–193. Liel, A.B., & Deierlein, G.G. (2013). Cost-benefit evaluation of seismic risk
Biondini, F., Camnasio, E., & Palermo, A. (2014). Lifetime seismic mitigation alternatives for older concrete frame buildings. Earthquake
performance of concrete bridges exposed to corrosion. Structure and Spectra, 29, 1391–1411.
Infrastructure Engineering, 10, 880–900. Moan, T. (2005). Reliability-based management of inspection, maintenance
Bradley, B.A., Dhakal, R.P., Cubrinovski, M., MacRae, G.A., & Lee, D.S. and repair of structures. Structure and Infrastructure Engineering, 1,
(2009). Seismic loss estimation for efficient decision making. Bulletin of 3–62.
the New Zealand Society of Earthquake Engineering, 42, 96–110. Moan, T. (2011). Life-cycle assessment of marine civil engineering
Campos, D., & Esteva, L. (1997). Modelo de comportamiento histerético structures. Structure and Infrastructure Engineering, 7, 11–32.
y de daño para vigas de concreto reforzado [Hysteretic behavior and Park, Y.J., & Ang, A.H.S. (1984). Mechanistic seismic damage model
damage model for reinforced concrete beams]. Proceedings of XI for reinforced concrete. ASCE Journal of the Structural Division, 110,
National Conference on Earthquake Engineering, 567–573. Veracruz, 722–739.
Mexico: Sociedad Mexicana de Ingeniería Sísmica (in Spanish). Picazo, Y., Diaz Lopez, O., & Esteva, L. (2015). Seismic reliability analysis
Celarec, D., Vamvatsikos, D., & Dolšek, M. (2011). Simplified estimation of buildings with torsional eccentricities. Earthquake Engineering and
of seismic risk for reinforced concrete buildings with consideration Structural Dynamics, 44, 1219–1234.
of corrosion over time. Bulletin of Earthquake Engineering, 9, Powell, G.H., & Allahabadi, R. (1988). Seismic damage prediction
1137–1155. by deterministic methods: Concepts and procedures. Earthquake
Chiu, C.K. (2014). Reliability-based service life assessment for deteriorating Engineering and Structural Dynamics, 16, 719–734.
reinforced concrete buildings considering the effect of cumulative Rackwitz, R. (2000). Optimization — The basis of code-making and
damage. Structure and Infrastructure Engineering, 10, 1101–1118. reliability verification. Structural Safety, 22, 27–60.
860 L. ESTEVA ET AL.
Rosenblueth, E. (1976). Optimum design for infrequent disturbances. Vamvatsikos, D., & Cornell, C.A. (2002). Incremental dynamic analysis.
ASCE Journal of the Structural Division, 102, 1807–1825. Earthquake Engineering and Structural Dynamics, 31, 491–514.
Roufaiel, M.S.L., & Meyer, C. (1987). Analytical modeling of hysteretic Vásquez, A. (2010). Funciones de daño acumulado para edificios de concreto
behavior of R/C frames. Journal of Structural Engineering, 113, 429–444. reforzado [Cumulative damage functions in reinforced concrete
Ruiz, V.M. (2000). Confiabilidad de marcos simples con disipadores buildings] (MsC thesis). Graduate Program in Engineering, National
[Reliability of simple frames with energy dissipators] (MsC thesis). University of Mexico, Mexico (in Spanish).
Graduate Program in Engineering, National University of Mexico, Wang, M.L., & Shah, S.P. (1987). Reinforced concrete hysteresis model
Mexico (in Spanish). based on the damage concept. Earthquake Engineering and Structural
Sanchez-Silva, M., Klutke, G.-A., & Rosowsky, D.V. (2011). Life-cycle Dynamics, 15, 993–1003.
performance of structures subject to multiple deterioration mechanisms. Williams, M.S., & Sexsmith, R.G. (1995). Seismic damage indices for
Structural Safety, 33, 206–217. concrete structures: A state‐of‐the‐art review. Earthquake Spectra, 11,
Stewart, M.G. (2006). Spatial variability of damage and expected 319–349.
maintenance costsfor deteriorating RC structures. Structure and
Infrastructure Engineering, 2, 79–90.