Deep Learning-based FEA Surrogate for Sub-sea Pressure Vessel

Overview

Created deep learning surrogates for finite element analysis (FEA) of sub-sea pressure vessels. The DNN model approximates full FEA simulation results at a fraction of the computational cost, enabling rapid structural design evaluation for underwater autonomous systems.

Approach

• Train a deep neural network on FEA simulation data (stress, displacement fields)
• Use the surrogate as a fast evaluator during design optimization loops
• Validated on sub-sea pressure vessel geometries relevant to UUV design