CAD Modeling method of the electrospun membrane under multifractal dimension optimization control

Abstract

Based on an analysis of large sample data, this paper improves the calculation method of the fractal dimension in an electrospun membrane and proposes a method to generate a computer-aided design (CAD) model of an electrospun membrane under the control of fractal dimension. Fifteen electrospun membrane samples of PMMA and PMMA/poly(vinylidene fluoride) (PVDF) materials were prepared under similar concentrations and voltage parameters, and 525 SEM images of the surface morphology with a resolution of 2560 × 1920 were taken as a dataset. The feature parameters, such as fiber diameter and direction, are extracted from the image. Second, based on the minimum value of the power law behavior, the pore perimeter data were preprocessed to calculate the fractal dimensions. A 2D model was reconstructed randomly based on the inverse transformation of the characteristic parameters. The genetic optimization algorithm adjusts the fiber arrangement to realize the control of characteristic parameters, such as the fractal dimension. Based on the 2D model, a long fiber network layer with a thickness consistent with the depth of the SEM shooting is generated in ABAQUS software. Finally, a solid CAD model of the electrospun membrane with realistic thickness was constructed by combining multiple fiber layers. The result shows that the improved fractal dimension exhibits multifractal characteristics and distinct sample differences, which are more similar to the experimental results. The proposed 2D modeling method of the long fiber network can allow the control of various characteristic parameters, including the fractal dimension, and can generate the required model quickly.