Diesel particulate filter (DPF) made from cordierite with a porous filter media structure has become an essential component in diesel engines to meet the increasingly the strict emission standards. Hence, a more accurate and universal DPF porous filter media model is necessary to analyze the complex mechanism and optimize its performance. In this work, a new optimization method was proposed for the initial DPF model constructed by the quartet structure generation set (QSGS), and the fractal and spectral dimension were used to quantify the actual effects of the optimized DPF model. Results showed that the method could effectively repair dead zones and eliminate burr boundaries, thus decreasing the complexity of the pore structure and increasing the effective flow area of the DPF model. The increase in porosity decreased the fractal dimension and increased the spectral dimension in porous media. Additionally, simulation results showed that the proposed method was more effective on burr boundaries. This allows the optimization method to be adopted more effectively in porous media modeling and provides improved guidance for optimization. Not only DPF, the proposed method offers a new idea for the structure optimization of porous filter media.
