Carbon dioxide emissions must be reduced as a countermeasure against global warming. A gasoline direct injection (GDI) engine is known for its high fuel economy, but there are more particulate emissions. To trap these particulates, a gasoline particulate filter (GPF) is needed. In this study, numerical simulations of the flow field in the GPF have been performed using the lattice Boltzmann method (LBM). Three types of filters were tested by our proposed approach, named as a digital material technique, where virtual materials are created by the simulation. Focusing on the pore characteristics of the filters, we used filters with the same pore size but different porosity. For comparison, the real internal structures of the actual GPF were detected using an X-ray CT method. The simulated flow field and the pressure drop across the filter wall were investigated, comparing with those of the actual GPF. As a result, it was found that the larger the porosity, the smaller the pressure drop. By setting the pore size and porosity to the same values as the actual GPF, a similar flow field was obtained, and the value of the pressure drop was close to that of the actual GPF.