Research on Microscopic CFRP Stochastic Pore Model Based on Pore Morphology

Abstract

The microscopic pore morphology of the samples were observed, and their statistical data were analyzed. Specifically, the distribution, shape and size of the pores were characterized, and the related histogram about the pore length distribution was also drew. According to CFRP pore statistical data, MATLAB was used to build a random pore model based on random process theory. The random medium was highly improved by using the extreme value search method. Thus, a random void model was deduced, well matched with our experiment results. During the establishment of the model, to make it closer to the actual pores, such important parameters as roughness factor, partition length and autocorrelation function were analyzed, and the statistical data of actual pores were used to optimize the parameters. Finally, a strict one-to-one correspondence was built between the stochastic model and the actual situation about pores. The model can provide theoretical support for our subsequent studies about explaining the effect of pores on ultrasonic testing and the scattering of ultrasonic waves in the pores during the testing process.