Induction hardening simulation of crankshaft

Induction hardening is extensively employed in industries such as automotive, aerospace and heavy machinery, where components require superior mechanical properties and long-term durability. Despite its advantages, the process induces tensile residual stresses that may compromise component performance by initiating micro-cracks. Therefore, optimizing the process design is essential to achieve a balanced outcome in terms of surface hardness, residual stress distribution and fatigue resistance. Traditionally, process optimization has relied heavily on experimental trials, which are both time-consuming and costly. This highlights the need for more efficient and reliable alternative methods. However, induction hardening is inherently a complex multiphysics process, involving strong couplings between thermal, electromagnetic, phase transformation and mechanical phenomena. Additionally, several material properties exhibit significant temperature dependency, introducing non-linearities that make accurate stress prediction highly challenging. Consequently, no analytical methods currently exist that can precisely predict the outcomes of the process. In response to the challenges faced by the Swedish heavy-duty truck manufacturer Scania, this thesis presents a comprehensive simulation model of the induction heating and cooling processes applied to the crankshaft of the 6 cylinder, 13 liter Scania Super engine. The model, developed with COMSOL Multiphysics version 6.3 and JMATPRO version 12.4, couples electromagnetic heating, transient thermal behavior, phase transformations and stress evolution to capture the full complexity of the process. The simulation provides detailed predictions of key process outcomes, including the temperature distribution at the end of induction heating, the resulting hardness profile and the residual stress field throughout the material. These results are validated against available experimental data to evaluate the model’s accuracy and limitations. Finally, the thesis outlines future directions aimed at further improving the precision and predictive capability of the numerical simulation that has been developed.