What question did this study set out to answer?

This research aims to develop a rapid method for digital characterization of cement-based materials by modeling electron energy deposition fields.

Rapid electron energy deposition modelling for digital characterization of cement-based microstructures using a physics-informed cGAN

View Full Paper

Ask AI

Bookmark

Authors

Randomized trial demonstrates rapid digital characterization of cement-based microstructures, suggesting a new approach for high-throughput analysis.

This research aims to develop a rapid method for digital characterization of cement-based materials by modeling electron energy deposition fields.
Developed a physics-informed conditional generative adversarial network (cGAN) as a surrogate for Monte Carlo simulations.
Utilized density, effective atomic number, and beam energy as physical conditions for training the model.
Implemented PatchGAN-based adversarial learning to enhance local morphology and response preservation.
Achieved a reduction in depth-profile MAE from 0.046 to 0.024 with the cGAN model.
Reduced peak error from 61.0% to 20.6% compared to a CNN baseline.
Predicted electron energy deposition fields in 4.42 ms, achieving 834- to 2687-fold acceleration over traditional Monte Carlo simulations.

Li et al. (2026) studied this question.

View Full Paper

Ask AI

Bookmark

Randomized trial demonstrates rapid digital characterization of cement-based microstructures, suggesting a new approach for high-throughput analysis.