CFD-DEM simulations of the Small-Scale Challenge Problem 1 (SSCP-I) fluidized bed are performed to assess the effect of a microstructure-based particle-gas force model on bed dynamics and particle velocity statistics. Three force-modelling approaches are compared: the conventional Di Felice drag model, the microstructure model with van Wachem mean drag, and the microstructure model with Di Felice mean drag, in which the Di Felice model provides the mean drag while the microstructure-based model accounts for drag deviations and lateral forces. The simulations are evaluated against experimental measurements of the pressure-drop spectrum and Eulerian particle velocity statistics, and bubble statistics are additionally compared with literature correlations.

The results show that the combined microstructure with the Di Felice approach improves the prediction of the dominant pressure-drop frequencies, indicating a better representation of the characteristic slugging dynamics. Mean particle velocities are captured with comparable accuracy by the different approaches, while simulations, including microstructure-dependent force contributions, show improved agreement for particle velocity fluctuations. In particular, the granular temperature is substantially better predicted than with the conventional Di Felice model. Bubble diameters and bubble rise velocities are only weakly affected by the choice of force model. Overall, the results demonstrate that including local microstructure information in particle-gas force modelling can improve fluctuation-dominated quantities in unresolved CFD-DEM simulations of fluidized beds.