Author(s)

S. K. Hashemi & M. A. Bradford

Abstract

Currently, there is significant interest amongst computational mechanics researchers in the area of investigating structures under blast loads. The explosive effect can impart damage, ranging from minor to full structural failure. Recent advancements in computer technology have enabled the ability to implement software in an efficient and cost-effective way to model complicated blast scenarios. To achieve better agreement between numerical and experimental models, the behaviour of the materials must be defined precisely and correctly. Dynamic loads are much more complicated than static loads, and so all parameters which could possibly affect the results and their further interpretation should be assigned carefully. Therefore, different values for C and P – as Cowper-Symonds strain-rate coefficients in LS DYNA – are considered in a simulation of W150 × 24 steel beams under two different blast shots, and are compared with experimental results to determine strain-rate effects. To distinguish between different models, an error analysis is used based on scaling the absolute differences between the exact and derived results with the application of exponential and linear utility functions for four different performance criteria. The results show that the strain-rate effect must be taken into account in models containing blast loads, even when the strain rates experienced are relatively small. Moreover, the maximum deflections along the beam length show less dependency, while the residual deflections and maximum strain depend significantly, on the strain rate. The best model with less average error is derived when C and P equal 20 s-1 and 7, respectively. Keywords: numerical, blast, steel beam, strain rate, LS-DYNA.

Keywords

numerical, blast, steel beam, strain rate, LS-DYNA.