Understanding earthquake swarm behaviours is imperative for assessing whether they may serve as precursors to larger seismic events, with direct implications for seismic hazard assessments. Many studies suggest that swarms are primarily driven by fluid-fault interactions, while others report that aseismic slip can also generate swarms. However, the physics underlying these triggering mechanisms and the role of intrinsic fault properties remain elusive. Here we employ dynamic fault rupture simulations to model swarm sequences in the absence of fluid. We test a suite of fault settings with different rate-and-state frictional parameters by systematically varying the parameters a and b, which represent the direct effect and state evolution terms, respectively. Specifically, we vary the ratio (a-b)_vw/(a-b)_vs between velocity-strengthening (VS) and velocity-weakening (VW) regions to identify fault conditions favorable for generating swarm-like sequences. We then analyze the simulated sequences using observational techniques, including spatiotemporal migration, hydraulic diffusivity fitting, and Gutenberg-Richter magnitude-frequency analysis. Our results indicate that swarm generation is prevalent when (a-b)_vw/(a-b)_vs lies between approximately 0.65 to 1. At a ratio of 1, 57% of the data display swarm sequences, increasing to 100% as the ratio decreases. The resulting b-values range from 0.77 to 4.30, consistent with natural swarm observations. Interestingly, even in the absence of fluids, some simulated sequences exhibit a diffusive migration behavior, with inferred diffusivities of 0.03-9.2 m²/s, also within the observed range reported for natural swarms. These findings suggest that swarms can emerge from aseismic processes under complex fault conditions while presenting key features typically associated with fluid-driven swarms.