Real‐Time Hybrid Fire Simulation on Steel Frames: Coupling Strategy and Parameter Effects

Abstract

Understanding and modelling the complex interactions and load redistribution mechanisms of steel structures under fire conditions remains difficult even today. In fact, structural fire safety research and application still lack concrete methods that accurately capture phenomena in structural systems. While large‐scale experiments are expensive and pure numerical simulations suffer from validation and uncertainty issues, Hybrid Fire Simulation (HFS) offers a promising solution. However, the sensitivity and stability of HFS methods for stability problems of structural members and frames remain unexplored. This paper presents a large‐scale benchmark setup and develops an HFS method for members in uniaxial compression, evaluating its sensitivity and stability. Key parameters governing coupling stability and model fidelity are quantified, including the stiffness definition and update strategy of the user‐defined element representing the physical substructure (PS), the displacement‐switch threshold and ABAQUS convergence settings. Two thermomechanical configurations are examined: (i) thermo‐elastic behaviour with temperature‐dependent Young's modulus, and (ii) thermo‐elasto‐plastic behaviour including thermal expansion. In both cases, the PS was tested at ambient temperature using a disc‐spring stack. The study identifies baseline parameter settings that maintained stable coupling and consistent global response, providing initial settings for future tests with a heated PS.