Condition monitoring of hot forging tools using acoustic signals

The SMASH research project addresses a key challenge in hot die forging, particularly for small and medium-sized manufacturers. High thermal and mechanical loads lead to severe tool wear and crack formation, causing premature tool replacement and production defects. Existing monitoring methods, such as optical inspection and force measurement, are often too costly, difficult to integrate, or unreliable in industrial environments.

SMASH develops a cost-effective inline monitoring system using acoustic emission analysis, intelligent signal processing, and machine learning. The system detects and predicts tool wear as well as thermal and mechanical crack initiation during forging. Machine learning algorithms identify damage-related signals and separate them from industrial background noise. The approach is validated in an automated test environment that represents realistic production conditions.

The project aims to deliver a modular software demonstrator for real-time monitoring of tool condition and part quality. By enabling reliable prediction of remaining tool life, SMASH reduces unexpected failures, minimizes scrap, extends tool lifetime, and improves resource efficiency in industrial forging.