A. Candeo, F. Dughiero, University of Padova, Padova, Italy
With the increasing demand for high-quality products, both in terms of reliability and accuracy, surface heat treatments have become a mandatory step in the production chain of many mechanical components, especially gear wheels. They allow to increase the superficial hardness of the treated parts, while leaving the core properties unaffected, thus preserving ductility and extending fatigue life. In fact, wear and pitting resistance as well as bending and impact strength are decisive factors which influence gear performances.
Delivering uniformly contoured profiles, case hardening remains the most common process in industry, despite some intrinsic disadvantages such as high distortions and long process times. On the other hand, induction surface hardening still remains a challenging task to carry out, due to the slightly different achievable contour and radial hardness distribution, aspects which must be taken into consideration when deciding on case depth requirements. The main benefits of switching to induction are related to a significant reduction in process steps and fabrication costs, ease of compressive residual stress achievement, and lower distortions. However, the tuning of the process may be quite tedious and requiring many hours of laboratory trials, technical expertise, and metallurgical analyses.
Moreover, an accurate and realistic 3-D modeling of all the concurrent physical phenomena occurring within the work-piece during the heat treatment represents a tricky task, mainly due to the inter-dependence and cross-coupling of electrical, magnetic, thermal, metallurgical, and mechanical properties of the material.
This paper will show the perspectives and issues of 3-D multiphysics gear hardening simulation. This will cover the electromagnetic and thermal problems related to eddy current heating, together with the metallurgical and mechanical aspects induced by phase transitions and stress/strain evolution. The simulated profiles will be shown and compared with experimental results from real test-cases.