Investigating the predictive capacity of Eulerian CFD to model long-duration blast loads on finite cross-section geometries

Typically defined by positive pressure durations over 100ms, long-duration blasts can generate dynamic pressures (blast winds) capable of exerting damaging drag loads on slender structural elements such as columns. With limited availability of appropriate drag coefficients for specific structural geometries or different section orientations, Computational Fluid Dynamics (CFD) can provide a valuable tool for calculating blast interaction and loading on user-specified geometries. Commercially available CFD programs or ‘hydrocodes’ with shock wave modelling capabilities remain based on solving the inviscid Euler equations. the ability to analyse long-duration blasts is still not confidently offered however, with no prior studies examining the accuracy of modelling interaction with relatively much smaller, finite geometries. this remains particularly challenging due to large wavelengths and time durations inherent to long-duration blasts, usually limited by impractical solution domains and computing resource. This paper presents a comparative investigation between numerical simulations and experimental results to assess the predictive capability of Eulerian CFD as a tool for calculating long-duration blast drag loading on an intricate I-section geometry from different angles of incidence. Calculated pressure-time histories on exposed geometry surfaces demonstrated good agreement although reduced accuracy and under-prediction occurred for shielded surfaces manifesting as overestimated net translational loading. Numerical discrepancies were attributed to the inviscid Euler equations underpinning the CFD solver, limiting accuracy when resolving complex aerodynamic flows at bluff I-section orientations. Results of this study provide new understanding and awareness of the numerical capability and limitations of using CFD to calculate long-duration blast loads on intricate geometries