Diesel trains have a wide range of applications, especially in nonelectrified mountainous regions with many tunnels, where the ventilation performance of the power packs is crucial to ensure the smooth operation of the trains in tunnels. The smoke emitted from the power packs affects indoor air quality through air conditioning units (ACUs) on the top of the train. In this study, the interest is to understand the flow field around the power pack at the bottom of the train and diffusion of smoke on the top of the train. Numerical simulations were conducted using the incompressible unsteady Reynolds-averaged Navier–Stokes and shear stress transport k-ω two-equation turbulence model along with the slip-mesh technique. The simulation method and parameter settings were verified based on experimental data. The results show that the upstream fan flow is greater than the downstream in the same power package. The downstream fan flow fluctuates to a greater extent. The increase in train speed leads to a decrease in the fan flow and has a greater impact on the upstream fan flow. The downstream smoke concentration in the same ACU intake in the head car tail car is greater than that in the upstream. Compared to the train speed, the blockage ratio has a limited effect on fan flow and smoke diffusion. Therefore, to optimize the ventilation performance, a priority should be given to the speed of the train.