CRE’s contributions to ESREL SRA-E 2025 conference
From June 15–19, 2025, two PhD researchers from the Chair for Reliability Engineering (CRE), Mauricio Misraji and Ali Osman Mert Kilicsoy , participated in the ESREL SRA-E 2025 conference, held at the University of Stavanger in Norway. The event brought together experts from the fields of safety, reliability, and risk analysis, providing an excellent platform for sharing research and engaging in insightful discussions. Both researchers had the opportunity to present their work and represent the Chair at this conference.
Mauricio Misraji Research:
First Excursion Probability Estimation of a Bilinear Conservative Oscillator Subject to Gaussian Loading
Abstract:
The study of oscillators under stochastic loading is essential in fields such as mechanical, ocean, wind, and earthquake engineering. The uncertainty arising from the intrinsic nature of the loading can be quantified using the first excursion probability, which measures the likelihood that an oscillator’s response will exceed a specified threshold during stochastic excitation. However, estimating this probability is challenging due to the need to manage a large number of random variables to represent the load, as well as potential nonlinearities in the restoring force and the non-stationary nature of both the loading and the oscillator’s response.
This work presents an efficient method for estimating first excursion probabilities for nonlinear oscillators subjected to Gaussian loading, particularly for systems with a bilinear, conservative restoring force. The technique focuses on exceedance probabilities within the nonlinear response range by dividing the calculation into two components: first, estimating the probability of failure in the elastic range, and then assessing the probability of failure in the inelastic range.
To estimate the probability within the elastic range, an advanced simulation-based variance reduction method is employed. This method effectively addresses scenarios where the oscillator's response remains linear. This technique also generates samples of the oscillator's maximum response in the inelastic range, which are subsequently used to construct an extreme value distribution for the inelastic maximum response. By synthesizing the response data, this distribution facilitates a more accurate and efficient estimation of first excursion probabilities across both response ranges. In this context, the variance reduction sampling method is optimally utilized to explore both elastic and inelastic ranges, while the extreme value distribution leverages this information to enhance the estimation of the first excursion probability. A numerical example is provided to illustrate the application of the proposed approach.
Ali Osman Mert Kilicsoy Research:
A Sobolev Trained Neural Network Surrogate with Residual Weighting Scheme for Computational Mechanics
Abstract:
Repeated evaluation of system responses through models become necessary when quantifying uncertainty or optimizing such system. This task can accurately be done through use of complex numerical models such as finite elements. However these models bring with them high computational cost which scales with the complexity of the observed system. Therefore, the use of surrogate models is very practical as they can provide a feasible accuracy for less computational cost. Neural networks represent one type of such surrogate models, whereby a set of data is used to train the neural network model on. The incorporation of sensitivity data, called Sobolev training, can elevate the model performance in accuracy and training time by expanding the loss with additional terms. Each term is pondered with a coefficient weight, which are optimized in parallel training through an adaptive scheme. We use this neural network model in a case study of computational mechanics with regards to its performance.
As CRE continues its mission to bridge theoretical innovation with practical impact, we are proud of our PhD candidates’ contributions at ESREL SRA-E 2025. The conference offered a valuable platform for presenting ongoing research, engaging with leading experts in the field, and fostering interdisciplinary dialogue to advance safety, reliability, and risk-informed engineering practices across Europe and beyond.
Conferences details sourced from: ESREL2025