Utilization of Digital Twins and Other Numerical Relatives for Efficient Monte Carlo Simulation in Structural Analysis

Bernt Johan Leira, Arifian Agusta, Sebastian Thöns

Research output: Chapter in Book/Report/Conference proceedingBook chapterResearchpeer-review

129 Downloads (Pure)


Analysis of structures will in general involve large and complex numerical models, which require extensive computation efforts. These models are frequently referred to as digital twins. This analysis becomes particularly cumbersome for cases where a large number of response calculations are repeatedly performed, such as in the case of Monte Carlo simulation. One way of avoiding this will be to introduce simplified numerical models, which are no longer twins but some kind of more distant numerical relative. As an example of such a simplified numerical representation, a so-called response surface model can be applied in order to overcome the excessive computational efforts. Such models are also sometimes referred to as meta-models or cyber-physical models. One possible approach is to use a response surface model based on first- or second-order polynomials as approximating functions, with the function parameters being determined based on multivariate regression analysis techniques. In this chapter, various types of approximate models are first discussed in connection with a simplistic example. The application of response surface techniques is subsequently illustrated for a quite complex physicsbased structural model for an offshore jacket structure in combination with Monte Carlo simulation techniques.
Original languageEnglish
Title of host publicationTheory, Application, and Implementation of Monte Carlo Method in Science and Technology
Number of pages20
Publication date2019
Publication statusPublished - 2019


  • Digital representation
  • Structural analysis
  • Monte Carlo simulation
  • Response surface techniques
  • Structural integrity management


Dive into the research topics of 'Utilization of Digital Twins and Other Numerical Relatives for Efficient Monte Carlo Simulation in Structural Analysis'. Together they form a unique fingerprint.

Cite this