Sustainable And Circular SAND Recycling (SANDRA)

Purpose and goal:

This project stands as a pioneering endeavour to digitally transform the casting process in foundries, by integrating empirical measurements, numerical modelling, and machine learning within sand reclamation processes. The purpose of this project is to reduce the environmental impact of the sand recycling processes. The project aims to introduce a machine learning model for sand recycling to help foundries optimize process parameters, better manage their sand recycling process, optimize their material consumption and reduce the amount of casting defects.

Expected results and effects:

The project will develop innovative models for more efficient sand reclamation and overall, a more stable casting process, which will contribute to a sustainable, resource efficient and circular production system. Expected results are to reduce the environmental impact of foundries by reducing the amount of sand waste and demand for new sand using machine learning model to optimize the sand recycling.

The approach is to increase sand reclamation to its maximum while maintaining and/or improving the quality of castings. Approach and implementation:

A machine learning project begins with gathering data from pertinent sources, which involves extracting vital manufacturing details, and formatting them appropriately. The project starts with gathering data on characteristics of raw materials like sand, binders, and additives incorporated into sand and core mixtures using unique tools and methodologies developed by project partners.

After data collection, the data undergoes pre-processing to be able to produce and train a machine learning model tailored to enhance the sand recycling procedure and elevate casting quality.