Residual stresses from Friction Stir Processing
Description
Dataset Description This dataset provides 2D residual stress distributions at specific locations on a grid. The stresses are recorded in two principal directions: the longitudinal/processing (x-direction) and the transverse (y-direction). The dataset contains two types of data: 1) Experimental data • The experimental stress values were obtained from a friction stir processed 3mm thick cast aluminum alloy 380 plate. The photograph of the plate with marked grid points used for stress characterization is provided in the accompanying Excel sheet. • All stress data corresponds to the grid points shown in the photo, and symmetry across the processing line has been assumed for analysis purposes. • The characterization method used was the hole-drilling electronic speckle pattern interferometry (ESPI) (https://doi.org/10.1007/s00170-024-14769-0) 2) Simulated stresses • The simulated stress values represent through-thickness averaged stresses in the plane of the sheet. These simulations were performed using the Thermal-Pseudo Mechanical (TPM) model (https://doi.org/10.1007/s00170-023-11199-2) in FEM. • The simulated dataset is divided into two subsets: one for training and validation of the machine learning model (400 sets) and the second for testing its generalization capabilities (150 sets). • Each CSV file corresponds to a single simulation stress distribution. The first column contains stresses in the longitudinal/processing (x-direction), while the second column contains stresses in the transverse (y-direction). • The coordinates associated with the stress data points are provided in a separate file. This file specifies the x-coordinate in the first column and the y-coordinate in the second column.
Files
Steps to reproduce
1) Experimental data was obtained using the hole-drilling electronic speckle pattern interferometry (ESPI) (https://doi.org/10.1007/s00170-024-14769-0) 2) These simulated data was obtained using the Thermal-Pseudo Mechanical (TPM) model (https://doi.org/10.1007/s00170-023-11199-2) in FEM.
Institutions
- Pacific Northwest National Laboratory