Prediction tool for plasma nitriding results
In a joint project of the Fraunhofer Institute for Surface Engineering and Thin Films (IST) and the WOT Chair of the Chemnitz University of Technology (IWW), a prediction tool for plasma nitriding results has been developed for twelve steel grades so far, which in addition to a broad database of nitriding results also offers a software-based prediction trained via an artificial neural network.
The IGF project IGF 18741 BG entitled „Prognosis tool for plasma nitriding processes for the surface layer treatment of tools and components“ of the German Society for Electroplating and Surface Technology was funded by the Federal Ministry for Economic Affairs and Energy through the German Federation of Industrial Research Associations „Otto von Guericke“ e.V. as part of the programme for the promotion of joint industrial research on the basis of a resolution of the German Bundestag.
In plasma nitriding, there are complex relationships between the nitriding parameters (such as treatment time, process temperature and gas composition), the alloying elements contained in the base material, the component geometry and the nitriding result. From the user's point of view, the most important parameters for evaluating the nitriding result are nitriding hardness depth and the hardness gradient in the component edge zone. In the past, there was a lack of reliable knowledge about the above-mentioned relationships, which was mostly due to a lack of systematically recorded data.
As part of the IGF project „ProgPlas - Prognosis tool for plasma nitriding processes for the surface treatment of tools and components“ more than 500 combinations of twelve different materials with different sets of process parameters were therefore produced and then evaluated with regard to the nitriding result. All the data was then processed in the form of user-friendly result maps, which provide a quick overview of the materials, nitriding parameters and results. On the other hand, the compiled database was used to create an artificial neural network that identifies the effective relationships of all process- and material-related input parameters with the nitriding hardness curve. The extensive data also enables the generation of nitriding hardness curves from unlearned input data. For a user-friendly usability of the trained network, a graphical user interface was set up.
The prediction tool created will also be enlarged in the future with further experimental data and thus experience an increasingly general applicability. The data available so far already contribute to better exploiting the potential of nitriding in practice by enabling the user to fall back on optimal, i.e. effective and at the same time efficient process parameters. The optimised approaches can, for example, increase the tool life of tools or the service life of components.
Further links:
Online forecasting tool:
https://mala.txm.de/webapps/home/session.html?app=NitridingTool
PDF download of the article:
https://www.tu-chemnitz.de/mb/WOT/forschung/forschungsprojekte/aif_bmbf_sab/IGF18741.pdf
The final report on the project can be requested by e-mail at the following address:
thomas.grund@mb.tu-chemnitz.de
Image:
left: Exemplary result map, overview of material, nitriding parameters and result; right: User interface of the prognosis tool based on an artificial neural network.
27.1.2020 – Projects of the professorship ( thomas.grund@mb… )
