A simple wave separation method for Split Hopkinson Bar experiments using linear encoders

Abstract

The Split Hopkinson bar is a well-established instrument for testing material properties at high strain rates. Despite its popularity, the method has limitations due to its measurement principle, which involves the propagation of the strain wave in elastic slender bars. A key limitation is the superposition of strain waves, which primarily limits the maximum duration of experiments. To address this, wave separation (or wave deconvolution) techniques have been developed to separate overlapping strain waves. However, existing methods often involve complex algorithms, wave dispersion issues requiring an analytical model of the bar's material, or expensive experimental equipment. This paper introduces a simple wave separation technique using linear magnetic encoders as velocity sensors in a Split Hopkinson bar. The approach relies on solving wave propagation equations at a single point, using velocity signals from the linear encoder and strain data from a conventional strain-gauge. The method offers several advantages, including simplicity in instrumentation and calculation, suppression of wave dispersion effects, easy implementation, and cost-effective sensors. We validate the method through numerical simulations with a custom finite element code, analyzing error sources and their impact. Experimentally, we demonstrate the principle using void tests (experiments without a specimen) and compare the results with the conventional strain-gauge method. The technique is further applied to various Split Hopkinson bar systems and materials, including compression, tension, and cellular materials. The results are promising, with good performance within the application range of the method. The paper concludes with a discussion of its advantages and limitations.