Optimizing the Aluminum Recycling Upgrade Process Using Multi-Sigma®

This case study showcases how Multi-Sigma’s AI-driven analysis platform optimizes the aluminum recycling upgrade process by enhancing mechanical properties while minimizing greenhouse gas (GHG) emissions and costs. This project was conducted as a part of a NEDO-funded project, this study leverages AI to drive significant advancements in sustainability and performance.

1. AI analysis

A dataset containing 18 process samples was analyzed. The AI model was trained using key input parameters from the aluminum recycling process, including impurity concentration, solution treatment duration, high-pressure sliding process, and aging treatment conditions. The output parameters included the mechanical properties of the recycled aluminum, as well as associated GHG emissions and cost estimates.

AI analysis 
Multi-Sigma Neural network analysis 
Parity plot

2. Factor Analysis

The factor analysis identified the most influential parameters affecting the aluminum recycling process:

  1. Length of the high-pressure sliding process (~ 41% influence).
  2. impurity concentration (~ 19% influence).
  3. solution treatment time (~ 12% influence).

3. Multi-objective Optimization of the Recycle Process

Multi-Sigma’s optimization module was utilized to determine the ideal recycling process parameters. The objective was to maximize key mechanical properties—such as tensile strength, 0.2% proof stress, and elongation—while minimizing GHG emissions and cost.

Multi-objective Optimization of the Recycle Process

Source: NEDO project (Development of advanced circulation technology for aluminum materials)
https://www.nedo.go.jp/english/activities/activities_ZZJP_100195.html
https://aizoth.com/research-project/nedo/