The separator, a microporous membrane in lithium-ion batteries (LIBs), serves as a crucial component by providing the pathway for lithium-ion transport as well as preventing internal short circuits through physical isolation of electrodes. Its structural properties, such as porosity, pore size, and tortuosity, significantly influence not only its ionic conductivity but also the battery performance and cycle life. Given their importance, various structural characterization techniques have been introduced to elucidate separator microstructures. However, conventional 3D imaging methods, including X-ray tomography and ion beam slicing, face critical limitations in analyzing nanoscale pore structures due to insufficient resolution or beam-induced damage. To address these issues, we propose a stochastic methodology for forming virtual 3D separator structures from 2D surface images of the separator for obtaining reliable structural parameters without severe thermal distortions. The generated virtual 3D models are quantitatively validated against experimental data. Furthermore, by systematically varying structural factors, such as porosity and pore size, we analyze their influence on ion transport properties, providing new insights into dominant design parameters. This study proposes a versatile simulation framework that enables an in-depth understanding of internal microstructure and guides the rational design of advanced separators for high-performance LIBs.