Computer-based methods for constructing parametric statistically equivalent models of the microstructure for the analysis of its elastic characteristics
Computer-based methods for constructing parametric statistically equivalent models of the microstructure for the analysis of its elastic characteristics
Babak Haghighi1
1) 1. Department of Physics, Ma.C., Islamic Azad University, Mashhad, Iran,
IOI Code
:
ACEF-ACAAAK
Publication :
6th. International Congress On Engineering, Technology and Applied Sciences(etas2027.com)
Abstract :
The analysis of microstructure material is one of the important parts of the manufactured products quality control. The microstructural study allows obtaining information about the state of the material by the structure image and does not require additional equipment for testing the sample. The paper analyzes cast iron with the inclusion of spheroidal graphite. This material is widely used in mechanical engineering. Suspension axles, connecting rods, gears, shafts, braking elements are produced from ductile iron.
The paper proposes computer methods for constructing parametric statistically equivalent models of high-strength cast iron microstructure to analyze its elastic characteristics. The metallographic images of the material are studied using the OpenCV computer vision library. Contours of graphite inclusions are determined. Graphite is assumed to have a circular shape. The mathematical expectation and variance of the inclusions size are approximated. The dependence of the inclusions size and their concentration is found. According to this dependence, a statistically equivalent artificial microstructure of cast iron is generated. A finite element model is built to find the elastic characteristics of the material. It is accepted that ferrite is an isotropic material, and graphite has a hexagonal crystal lattice structure. The stress state is considered based on flat models.
Two hundred numerical experiments are carried out for 17 values of the inclusions concentration. The results obtained for elastic constants are statistically averaged. The dependences of the Poisson coefficient, elastic modulus, and displacement on the concentration of inclusions are established. To assess the reliability of decisions, the mixture rule is applied. The results of the rule applying confirm the correctness of the constructed models. However, the upper boundary of the confidence interval exceeds the upper estimate of the elastic modulus. This result arises from the complexity of the real properties of the graphite material, and also because the orientation of the main axes of graphite crystal cannot be determined. Therefore, a comparison is possible only on average and value.
Keywords : image processing finite element method material properties ferrite image processing finite element method material properties elastic modulus