In the case of multiaxial loading, different models are used for prediction of fatigue life. Usually, they are divided into 3 subgroups: stress‐based for high‐cycle regime, strain‐based for low‐cycle regime, and energy‐based models, which are considered to be the most universal and applicable for both high‐cycle and low‐cycle regimes. The application of energy‐based models requires knowledge of the energy dissipated during the plastic strain, as well as the elastic strain energy. Calculation of this energy in turn requires specification of stress and strain tensor components. Determination of the stresses acting in real objects is a complicated task, however. In this paper, an attempt was made to answer the question of whether thermography enables the determination of plastic strain energy in low‐cycle uniaxial and multiaxial fatigue tests, including non‐proportional loadings. The second question was whether it is possible to use this methodology for fatigue life prediction. For this purpose, a history of temperature changes was recorded during fatigue tests conducted on CuZn37 brass, using a thermographic camera. On this basis, the values of plastic strain energy density, dissipated in the fatigue loading cycle, were calculated next. These values were compared with the values calculated from the hysteresis loops, determined from force and torque measurements, together with strains measured with a biaxial extensometer. The strain energy density was further used for prediction of fatigue life with the application of 2 models: the energy‐based model of Ellyin and a strain‐based model, where the energy was used as a non‐proportionality factor. The predicted fatigue lives were compared with experimental ones. The results can be considered as very satisfactory.

The paper presents a thermographic method of accelerated determination of theS-Ncurve. In the presented method, theS-Ncurve was developed based on energy-related parameter with the assumption of its dependency on the stress amplitude. The tests made on C45 steel and X5CrNi18-10 steel under reversed bending revealed that theS-Ncurve obtained by accelerated thermographic method fits inside the 95% confidence interval for theS-Ncurve obtained from the full test.

A new thermographic method that enables simultaneous accelerated determination of the fatigue limit and the S-N curve was presented in this paper. The fatigue limit determination method was based on a constant rate of temperature rise occurring in second phase of the specimen fatigue life. The S-N curve determination method was based on energetic parameter with assumption of its dependency on the stress amplitude. The tests made on X5CrNi18-10 steel under reversed bending revealed that the fatigue limit value obtained from accelerated thermographic tests as compared to the value obtained from full test differs by 5.0 %. The S-N curve obtained by accelerated thermographic method fits inside 95 % confidence interval for the S-N curve obtained from full test.