The emphasis in the first half of the project was the assessment of the simulation chain, the simulation of the billet processing, the analysis of the circumferential yield strength variations, the implementation of the grain class model for the simulation of duplex grain structures as well as the design and forging of duplex discs.
During the assessment, the simulation chain was applied for the design of new direct aged forgings showing reasonable results for the calculation of the yield strength and the low cycle fatigue limit. The simulation of the billet processing was assessed with the emphasis on the material models for recrystallization during radial forging. It was shown that the used recrystallization model is insufficient to describe the occurring mechanisms during radial forging in the simulation. However, the simulated local strain, strain rate and temperature data was used in the subsequent simulation chain.
In order to analyze the circumferential variation high-resolution measurements were performed and fundamental material mechanisms were investigated during the whole project (all three report periods). Basic differences between the direct aged condition and standard aged condition were analyzed in detail. Systematic effects involving dislocation structures, precipitation kinetics and the interaction between dislocations and precipitates caused by a multi-blow forging were observed and used to develop a hypothesis on the effects causing circumferential yield strength variations. This knowledge is of scientific interest and very valuable for BSTG as a forging supplier of critical parts. The results were published and presented at scientific conferences.
Furthermore a grain class model to describe the duplex microstructure was implemented in the simulation tool. Due to an incompatibility of the model with the simulation tool caused by interpolation after remeshing the model could not be used. An alternate modelling approach was therefore developed using the meta-dynamic recrystallization to predict regions with non-homogeneous grain structures.
In the second reporting period this model was used to design the processing route of a turbine disk. In order to validate the developed model the forged disks were cut up and microstructure as well as mechanical properties were investigated. It could be shown that the model is only suited for a qualitative prediction of the microstructure. In order to be able to make quantitative predictions, the originally planned grain class model was developed as well in a parallel project.
During the third reporting period the grain class model was parametrized and implemented into the automated simulation chain. Further turbine disks were designed, forged and fully investigated to validate both developed models. By the end of the project, it could be shown that both models are suited for the prediction of mechanical properties. The first model can give an overview qualitatively of the regions that are endangered of the occurrence of non-uniform grain structures. The second model is also able to quantify the microstructure (e.g. grain size, fractions). The use of the models is a very powerful tool to improve the production route of turbine disks and to make them even safer in use. Therefore, the models will be used at BSTG for the design of turbine disks. The modeling and simulation results were published and presented at scientific conferences.
