Fraunhofer-Gesellschaft

Publica

Hier finden Sie wissenschaftliche Publikationen aus den Fraunhofer-Instituten.

Investigations on the Application of the Downhill-Simplex-Algorithm to the Inverse Determination of Material Model Parameters for FE-Machining Simulations

 
: Hardt, M.; Schraknepper, D.; Bergs, T.

:

Simulation modelling practice and theory 107 (2021), Art. 102214
ISSN: 1569-190X
English
Journal Article
Fraunhofer IPT ()

Abstract
Modeling the machining process by means of simulation techniques became more and more popular within the last decades. The increasing use of simulation techniques can be attributed to the enhancing computational performance. To accurately model the machining process, different input models are required. Among them, the material model has a crucial impact on the quality of the simulated results. To determine the underlying material model parameters, an inverse procedure has been established, where the material model parameters are adjusted within a chip formation simulation until a reasonable match between simulations and experiments is achieved. However, the procedure of the inverse determination requires high computational efforts and is not robust. This paper presents a novel approach to determine material model parameters inversely from FE-machining simulations by means of the Downhill-Simplex-Algorithm. An effect study was conducted to investigate the influence of the initial simplex, the boundary conditions, and the underlying parameters of the Downhill-Simplex-Algorithm. Further, the validated procedure has been extended to consider multiple cutting conditions and an adaptive step-size has been implemented into the algorithm. The results of the procedure were validated by re-identification of process observables from the target simulations and it was aimed to re-identify the material model parameters of the target simulations. Based on the algorithm, a robust and systematical method has been developed to determine material model parameters inversely from the machining process.

: http://publica.fraunhofer.de/documents/N-615045.html