N icolas Graux

In recent years, rising raw material and transport costs have increased the constraints on the manufacturing industry. In order to limit the impacts, diversification of material supplier is one of the keys. However, it is not easy to get exactly the same material characteristics from two different suppliers. Even if the material is selected according to standards, the process and treatment of the material can change the mechanical properties, especially the fatigue properties. Unfortunately, static tests are not reliable for estimating fatigue properties and standard fatigue testing campaigns are too long and too expensive to have a large number of suppliers.

Measuring self-heating under cyclic loading is one of the solutions to get a quick evaluation of fatigue properties at a reasonable cost. With dedicated post-processing, it is possible to estimate the endurance limit and dispersion with a high level of confidence for reliable comparison. However, the self-heating measurement requires expertise at all stages of the test.

This study will focus on proposing the first steps of an automated test and post-processing to compare a self-heating measurement to a reference self-heating curve. To achieve this objective, steel specimens will be tested under cyclic loading. The reference of self-heating curves will be carried out for different configurations (tension, torsion, mean stress…). Then, an automatic protocol will be proposed considering issues, such as measurement errors or the minimum number of measurement points to be performed. Finally, the protocol will be tested on other grades of ferrous materials.

Session