Computational & Technology Resources
an online resource for computational,
engineering & technology publications
Computational Science, Engineering & Technology Series
COMPUTATIONAL TECHNIQUES FOR CIVIL AND STRUCTURAL ENGINEERING
Edited by: J. Kruis, Y. Tsompanakis and B.H.V. Topping
Application of Soft Computing Techniques in Life-cycle Optimization of Civil and Marine Structures
D.M. Frangopol and M. Soliman
Department of Civil and Environmental Engineering, ATLSS Engineering Research Center, Lehigh University, Bethlehem, United States of America
D.M. Frangopol, M. Soliman, "Application of Soft Computing Techniques in Life-cycle Optimization of Civil and Marine Structures", in J. Kruis, Y. Tsompanakis and B.H.V. Topping, (Editors), "Computational Techniques for Civil and Structural Engineering", Saxe-Coburg Publications, Stirlingshire, UK, Chapter 2, pp 43-58, 2015. doi:10.4203/csets.38.2
Keywords: reliability, importance sampling, Bayesian network, optimization, Markov Chain, Monte Carlo simulation.
Performance assessment and life-cycle optimization of large scale structural systems require advanced computational techniques to enable more accurate modelling of the structural performance and better understanding of their lifetime behaviour. These techniques should be capable of considering complex component-system interactions while accounting for aleatoric and epistemic uncertainties associated with resistance and load estimations, deterioration phenomena, and the effects of life-cycle interventions on the performance of structural systems. The quantification and reduction of these uncertainties are essential for more accurate life-cycle management processes. This can be performed with the aid of structural health monitoring and non-destructive inspection data, along with updating and inverse methodologies. This paper discusses the application of soft computing techniques in life-cycle performance prediction, service-life estimation, uncertainty quantification, and multi-objective optimization. The applications of incremental nonlinear finite element analyses, response surface modeling using design of experiments concepts, and advanced sampling methods for identifying probabilistic performance indicators of deteriorating structures are discussed. Updating techniques based on Markov Chain Monte Carlo simulation and Bayesian Networks, and their implementation in evaluating the structural performance are also presented. Moreover, this paper also discusses the integration of these performance prediction methodologies for determining optimal life-cycle management strategies using multi-objective genetic algorithms. Applications of these concepts to bridges and naval vessels are presented.
purchase the full-text of this chapter (price £20)