Uncertainty Quantification of Soil-Structure Interaction in Tunnel Linings by Polynomial Chaos Expansion

Uncertainty Quantification of Soil-Structure Interaction in Tunnel Linings by Polynomial Chaos Expansion

Stať ve sborníku v databázi WoS či Scopus

The paper is focused on uncertainty quantification of soil-structure interaction in tunnel linings using a surrogate model in form of Polynomial Chaos Expansion (PCE). Tunnel design is a complex and complicated task since it is strongly associated with a great number of load and material uncertainties. Moreover, modelling the soil-structure interaction multiplies the complexity and non-linearity of a tunnel engineering problems. In order to handle such uncertainties, finite element method with random input variables has proven to be a very accurate tool. The probabilistic analysis is typically performed by Monte Carlo simulation (MC), simulating uncertainties according to their complete probability distributions and statistical correlations. The computational burden of MC represents the main obstacle to its use in complex numerical models and it is therefore not practical for industrial applications. The solution can be an efficient approximation of the original mathematical model by computationally efficient analytical function – a surrogate model. In this study, the surrogate model in form of PCE is utilized, allowing for analytical post-processing (statistical and sensitivity analysis). Uncertainty quantification is focused on estimation of spatial variability of internal forces caused by the soil-structure interaction.

The paper is focused on uncertainty quantification of soil-structure interaction in tunnel linings using a surrogate model in form of Polynomial Chaos Expansion (PCE). Tunnel design is a complex and complicated task since it is strongly associated with a great number of load and material uncertainties. Moreover, modelling the soil-structure interaction multiplies the complexity and non-linearity of a tunnel engineering problems. In order to handle such uncertainties, finite element method with random input variables has proven to be a very accurate tool. The probabilistic analysis is typically performed by Monte Carlo simulation (MC), simulating uncertainties according to their complete probability distributions and statistical correlations. The computational burden of MC represents the main obstacle to its use in complex numerical models and it is therefore not practical for industrial applications. The solution can be an efficient approximation of the original mathematical model by computationally efficient analytical function – a surrogate model. In this study, the surrogate model in form of PCE is utilized, allowing for analytical post-processing (statistical and sensitivity analysis). Uncertainty quantification is focused on estimation of spatial variability of internal forces caused by the soil-structure interaction.

Concrete Structures; Polynomial Chaos Expansion; Spatial Variability; Tunnel Linings; Uncertainty Quantification

Concrete Structures; Polynomial Chaos Expansion; Spatial Variability; Tunnel Linings; Uncertainty Quantification

NOVÁK, L.; GAKIS, A.; KŘÍŽEK, M.; NOVÁK, D.; SPYRIDIS, P.

2025

08.05.2024

Springer Science and Business Media Deutschland GmbH

9783031602702

20th International Probabilistic Workshop

512

519

8

https://link.springer.com/book/10.1007/978-3-031-60271-9