Prediction of performance of a TiAlN coating in profile turning using a grey box approach (PreProCoat)

点击翻译按钮获取中文摘要

In the PreProCoat project, the development and use of an approach based on greybox models is being pursued, with which the application behavior of a TiAlN/TiN tool coating can be predicted for the case of the dry machining process of profile grooving in quenched and tempered steel C45. With the grooving profile, the load conditions in the contact zone between the outgoing chip and the coating change and thus the locally distributed thermo-mechanical load spectrum. In experimental machining investigations, first correlations between machining parameters as well as tool properties and criteria for the evaluation of the process course, the coating condition and the machining quality are worked out. In an analogy test, individual areas of the contact zone are investigated with regard to the local distribution of temperature and cutting forces and correlated with the machining parameters. The tests are statistically validated up to the end of life of the coating system. Process intermittent wear measurements as well as downstream coating condition analyses accompany the experimental phase. The results obtained will also serve to validate an FE chip formation model to be subsequently created, which can be used to supplement and extend the experimental investigations by numerical means. In particular, the FE model enables the local and temporal derivation of the thermo-mechanical stress collectives in the coating. The required analytical submodels for friction and material behavior are taken from the literature. Tribomechanical analyses on contact pairs of AlTiN/TiN-coated test specimens and C45 material samples provide the required specific friction parameters. With the aid of damage models, the local residual stresses in the coating system are identified from the thermo-mechanical load collectives, among other things, and a possible coating failure is derived from this. The calculations with the FE stress model are carried out iteratively along the coating service life, whereby the submodels for friction and material behavior as well as the geometry model of the tool are adjusted on the basis of the analytical condition determination before each iteration step. Since a purely analytical approach is likely to result in fuzzy prediction, AI/ML approaches are pursued with which experimentally obtained analysis data can be correlated with corresponding influencing variables. The coupling of the resulting data-driven models with the already implemented analytical models is finally done on the basis of a greybox structure to be created.