Ein Simulationsmodell zur Identifikation von Diffusionskoeffizienten anhand charakteristischer Kenngrößen aus Sinterexperimenten
New modeling approaches open up ever more possibilities for the simulation of microstructure de-velopment in solid phase sintering. In order to take advantage of these, the validation and use of sin-tering models requires precise knowledge of the material properties, especially the diffusion coeffi-cients. However, hardly any reliable data are currently available for the surface and grain boundary diffusion of most sinter materials. Against this background, a sinter model is developed in the present work and, in order to determine the required but difficult to access material properties on the basis of experimentally determined characteristic parameters, it is used in a method for parameter identifica-tion. As a simulation model, a two-particle model is developed, in which the material transport is realized using predetermined diffusion paths. The local differences in chemical potential along the diffusion paths are calculated by the derivatives of the interfaces according to the associated volumes. This novel approach allows, for the first time for this class of sinter models, to treat all diffusion mecha-nisms in a uniform way. The behavior of the model is first validated using experimental data from the literature. It is observed that the diffusion coefficients, for which simulation results are in good agreement with the literature data, can be excellently approximated by an Arrhenius equation. This observation is seen as an im-portant confirmation of the plausibility of the model behavior. The suitability of the model for parameter identification is demonstrated by determining the interfa-cial diffusion coefficients of titanium. The basis is formed by the characteristic quantities neck growth and shrinkage from experiments designed close to application. The parameter identification, where it is necessary to find the combination of diffusion coefficients that leads to the best agreement of the parameters in simulation and experiment, is formulated as a scalar optimization task and solved with a particle swarm algorithm. Potential for improvement is recognized in the experimental design, since the experiments for the -phase of titanium provide an ambiguous result. The sintering model itself, however, is again confirmed by the predictive behavior shown in comparative experiments. The final comparison of the own model compared to a particle pair modelled with the phase field method reveals differences in the modelling and in the model behavior, which rather suggests the validation of phase field models based on experiments than to question the model developed here. For sintering models using the phase field method, the modelling of shrinkage in particular is per-ceived as insufficient. The need for precise knowledge of the diffusion properties for their validation once again illustrates the need for methods to determine diffusion coefficients, as developed in this thesis.
Use and reproduction:
No license. The provisions of the German Copyright Act (UrhG) apply.
Please note that individual components of the publication may be subject to other licensing or copyright conditions.