Abstract: To address the challenges of removing light gases ( \mathrmH_2 ) in ultra-high vacuum systems, the significant decay in pumping efficiency of turbomolecular pumps (TMPs) across different flow regimes, and the lack of theoretical guidance for structural design, a two-dimensional numerical model for a TMP blade row was established using the Direct simulation Monte Carlo (DSMC) method in OpenFOAM. Based on the validation of grid independence and comparison with Kruger’s theoretical values and Sawada’s experimental data, the effects of inlet pressure, speed ratio, and structural parameters, including the pitch-to-chord ratio and blade-row inclination angle, on pumping performance under different flow regimes were systematically investigated, and contour maps were used to reveal the coupling effects between operating pressure and major geometric parameters. The results indicate that as the flow regime shifts from free molecular flow to transition flow, the increase in intermolecular collision frequency causes the pumping performance to decay logarithmically; moreover, the performance gain brought by increased rotational speed gradually saturates at higher pressures. It is found that H_\max peaks when the pitch-to-chord ratio is approximately 1.2, while K_\max decreases monotonically with the increase of the pitch-to-chord ratio. Furthermore, the optimal blade angle for maximizing pumping speed shifts towards smaller angles as the pressure increases. In addition, considering the extremely low compression ratio of hydrogen, this study explains the physical rationale for variable-parameter turbine blade-row designs, such as segmented pitch-to-chord ratios and variable blade-angle profiles, from the perspective of microscopic molecular transport and collision mechanisms, thereby providing a theoretical basis for the structural optimization of high-performance TMPs.