Civil


Civil

Engineering AI

Coastal and offshore infrastructure experiences complex, uncertain loading from waves, currents, and storm surges. We develop uncertainty quantification frameworks and graph neural operator surrogates to understand and predict structural response under hydrodynamic loading — enabling safer and more resilient civil infrastructure.


Hydrodynamics

Wave loading on coastal and offshore structures is inherently stochastic — wave heights, periods, and directions are uncertain, and small changes can produce large variations in structural response. We develop rigorous uncertainty quantification (UQ) frameworks and graph neural operator (GNO) surrogates to characterise and predict structural response efficiently, without the need for thousands of expensive high-fidelity simulations.

  • Sensitivity analysis and statistical characterisation of structural response extremes
  • Surrogate-based UQ for wave–structure interaction
  • Graph neural operators for ill-posed inverse problems in wave loading
  • Handling irregular geometries and unstructured meshes
  • Validation against physical flume experiments
Luo et al. — Framework for uncertainty quantification of wave–structure interaction in a flume, Computational Particle Mechanics (arXiv 2025)
Luo, Fourtakas & Harish — Graph neural operators for ill-posed problems in wave–structure interaction (2026, SSRN preprint)
Wave–structure simulation

Software & Tools

We develop and maintain open-source software tools to support the research community in wave–structure interaction analysis and uncertainty quantification.

  • HydroUQ — Uncertainty quantification framework for hydrodynamic loads on coastal structures
  • CFD Notebooks — Jupyter notebooks for CFD simulations of coastal and offshore flows
  • NHERI SimCenter — Broader suite of natural hazards engineering simulation tools
Open-source tools