Design Strategy for Art Copper Alloys’ Colors Through Machine Learning and Oxidation Treatment

Abstract

Copper alloys are often used as a substitute for gold because of their bright gold color and low production cost. Seventeen kinds of Cu-Al alloys, Cu-Zn alloys, and Cu-Sn alloys with different compositions were fabricated to investigate the effect of chemical composition and oxidation treatment on their colors. In addition, machine learning was used to establish the relationship between all alloys’ composition and color. The best model for each alloy also had been sought out. More importantly, through the analysis of the calculated results, the tree-structure-based models, gradient boosting regressor, random forest regressor, and decision tree regressor, exhibited a better predictive power on the small data set. Subsequently, Cu-4.17 wt.% Al, Cu-24.3 wt.% Zn, and Cu-10.1 wt.% Sn alloys were selected for oxidation at 100°C, which showed that later oxidation could also change sample color.