A thermo-mechanical fatigue (TMF) occurs when a mechanical as well as a thermal, periodically changing load acts on a component. In typical technical applications where thermomechanical fatigue has a significant influence, the number of load cycles is in the range of low cycle fatigue strength, which means at a number of cycles of 104 to 105, and the resulting elongations are usually above the elastic limit, especially because this decreases with increasing temperatures. A usual description of the connections takes place via the Coffin-Manson relationship. Analogous to the Woehler curve, its representation within a double logarithmic diagram results in a falling line. Instead of the load changes, however, the plastic strain is displayed on the abscissa.
- "In-Phase test" - Phase angle 0°
- "Out-of-phase test" - 180° phase angle
- "Clockwise Diamond test" - 90° phase angle
- "Counter-Clockwise Diamond test" - phase angle -90°
- or according to the application with any angle in between
The cyclical thermal and mechanical loads are not necessarily synchronous, but in many cases offset by a phase angle to one another. Depending on the phase angle, this results in significantly different material loads and achievable number of load cycles. Since the time sequence of the individual events has a significant influence on the damage processes, TMF tests distinguish between:
The processes in a material exposed to a thermomechanical load are considerably more complex than if they are triggered only by the pure mechanical stresses. In addition, thermomechanical fatigue is generally not isolated, instead it is superimposed by other loads, especially in high-temperature applications, where creep tendency and corrosion processes are more frequent. Both make it difficult to evaluate. An absolute statement is associated with a correspondingly greater uncertainty, especially when comparing different materials whose material data were obtained under different test conditions.
The numerical evaluation of thermomechanical fatigue and the prediction of the maximum allowable load changes are basically carried out using the same methods as for low cycle fatigue. Here, too, the local plastic strains must be determined according to their history of origin. Since the temperature curve is an important influencing factor, a coupled multiphysics simulation is necessary. The numerical effort increases accordingly. Since it is difficult to estimate which combinations of thermal and mechanical load conditions are decisive, a reduction of the calculation effort by quasistatic equivalent load cases is only possible to a limited extent.
Application-specific, alternative approaches exist to predict thermomechanical fatigue with the aim of keeping calculation times short. However, these methods often lose their absolute accuracy, but are particularly suitable for the localisation of the weak points and their comparative evaluation. This can be of particular interest in the concept phase, especially if no specific material properties are yet available.