Multi-objective parameter calibration and filling simulation verification of fireworks composite powder based on DEM-RSM and desirability function approach

Fireworks rely on accurate calibration of mixed powder parameters for quality, ignition effects, environmental pollution, and safety. This study combines experiments and simulations to measure basic powder parameters, including density, moisture content, and repose angle. Based on particle scaling theory, an enlargement model is established. The discrete element method coupled with response surface methodology (DEM-RSM) is used to calibrate eight parameters, including inter-particle static friction and powder-acrylic static friction. By minimizing the relative error between experimental and simulated repose angles, the optimal parameters are obtained. DEM-RSM is also employed to determine the critical filling speed and critical residence time at a target filling ratio of 100%. The Desirability Function Approach is introduced for multi-objective optimization and validation using repose angle and filling ratio. Results show repose angle errors below 1% and filling ratios exceeding 99%, confirming reliability. The calibrated parameters support future studies on combustion characteristics, dispersal behavior, and environmental pollution potential.