Adapting U-Net for linear elastic stress estimation in polycrystal Zr microstructures
This work provides a faster alternative for stress estimation in materials science, specifically for zirconium microstructures, but it is incremental as it adapts an existing neural network architecture.
The authors tackled the problem of estimating linear elastic compatibility stresses in zirconium polycrystalline microstructures by adapting a U-Net convolutional neural network, achieving speedups of 200x to 6000x with memory savings compared to finite element analysis, albeit with up to a 10% reduction in accuracy.
A variant of the U-Net convolutional neural network architecture is proposed to estimate linear elastic compatibility stresses in a-Zr (hcp) polycrystalline grain structures. Training data was generated using VGrain software with a regularity alpha of 0.73 and uniform random orientation for the grain structures and ABAQUS to evaluate the stress welds using the finite element method. The initial dataset contains 200 samples with 20 held from training for validation. The network gives speedups of around 200x to 6000x using a CPU or GPU, with signifcant memory savings, compared to finite element analysis with a modest reduction in accuracy of up to 10%. Network performance is not correlated with grain structure regularity or texture, showing generalisation of the network beyond the training set to arbitrary Zr crystal structures. Performance when trained with 200 and 400 samples was measured, finding an improvement in accuracy of approximately 10% when the size of the dataset was doubled.