New preprint on data-driven surrogate modeling of rough surface contact

This manuscript presents a surrogate modeling framework for predicting the effective contact area in rough surface contact problems, which plays a critical role in various multi-physics applications such as wear, sealing, and heat or electrical transfer. While highly accurate numerical techniques like the Boundary Element Method are available, their computational cost poses challenges in multi-query contexts, such as uncertainty quantification, parameter identification, and multi-scale modeling. To address this, we develop a fast and efficient data-driven approach built on a simulation-based database that captures a broad range of parameters. The proposed method significantly reduces the computational workload while maintaining predictive reliability, making it suitable for integration into multi-query and multi-scale modeling workflows.

Sahin, T., Bonari, J., Brandstaeter, S., & Popp, A. (2025). Data-Driven Surrogate Modeling Techniques to Predict the Effective Contact Area of Rough Surface Contact Problems (Version 1). Preprint, submitted for publication arXiv