Generated Data and Trained Checkpoints for a PC-FNO Thermal-Management Heat-Conduction Surrogate Study

This dataset archive contains generated finite-difference heat-conduction datasets, trained neural-surrogate checkpoints, and numerical experiment outputs used in the manuscript entitled "A Physics-Constrained Fourier Neural Operator Surrogate for Thermal-Management Heat Conduction and Sparse-Sensor Heat-Source Inversion". The archive supports reproduction of the reported U-Net, FNO, PC-FNO, lightweight DeepONet, out-of-distribution, grid-verification, sensitivity, and sparse-sensor inverse heat-source identification studies. The companion code and reproducibility material are provided separately as Supplementary_Code_Reproducibility.zip.