Dataset for Designing extrusion dies on the basis of eXplainable Artificial Intelligence

Published: 4 July 2022| Version 1 | DOI: 10.17632/dy97xr6t8h.1
Contributor:
Juan Marcos Llorca

Description

This dataset contains the porthole die data used to develop a design support tool for aluminium extrusion porthole dies based on eXplainable IA. The tool is useful for dies with 4 cavities and 4 ports per cavity. Dataset includes the geometrical data related to 596 different ports from 88 first-trial 4 cavities and 4 ports per cavity dies. In terms of the R² metric and the results obtained with the application examples, the results obtained with this ML-based model are significantly better than those of a previous model based on linear regression. It also includes the results and geometries of the FEM simulations performed to validate the model.

Files

Institutions

  • Universitat d'Alacant

Categories

Experimental Design, Machine Learning, Die Design, Profiles Extrusion, Predictive Modeling

Licence