A mean stress-based method for processing data in SHPB tests

The split Hopkinson pressure bar (SHPB) technique is widely used to test the dynamic mechanical properties of materials. However, conventional three-wave and two-wave methods for processing SHPB data introduce reconstruction errors in the stress–strain curve due to the incomplete satisfaction of the uniform stress assumption, affecting accuracy, especially for materials like rock and concrete. This paper proposes a mean stress-based SHPB data processing method (MS method) that avoids these reconstruction errors. First, by analyzing the stress-strain formulations in conventional models using one-dimensional elastic wave theory, a novel stress-strain reconstruction model (MS model) is developed. Then, the theoretical self-consistency and reconstruction errors of the three models are evaluated, revealing the causes of reconstruction errors in the conventional models and the advantages of the MS model. Finally, a data processing method is then proposed by combining the MS model with an algorithm to estimate the specimen’s mean stress. Applying the MS method to the concrete specimen SHPB data processing not only reveals the reconstruction errors inherent in the conventional methods but also demonstrates the significant value of replacing them with the MS method in improving the accuracy and applicability of SHPB technique.