On the calculation of the strain energy release rate for thin-gage facesheet sandwich structures using the J-integral method

Abstract

Facesheet-to-core disbonding in sandwich composites is a critical failure mechanism that can occur under relatively low, out-of-plane loads. Test methods such as the single-cantilevered beam (SCB) and drum peel tests are being considered for standardization to estimate the strain energy release rate (SERR) for computationally predicting failure and structural sizing purposes. In particular, the present SCB standard uses modified beam theory (MBT) and the area method to estimate the SERR, but these have limitations: MBT assumes linear-elastic behavior, while the area method requires a complex loading/unloading testing scheme to subsequently calculate the SERR. In this study, an alternative for SERR estimation is investigated using the J-integral method. Unlike MBT, the J-integral does not rely on linear-elastic assumptions, enables instantaneous SERR calculation, and is expected to be more widely applicable. Therefore, SCB tests were conducted on sandwich composites with relatively thin-gage (1.65 mm) facesheets, and the corresponding SERR estimates from J-integral, MBT, and area method were compared. In addition, finite element analysis using a virtual crack-closure technique was employed to verify the J-integral approach. For small normalized disbond lengths (<40 mm/mm), all analytical methods showed good agreement. Deviations over 10% in SERR estimates between the J-integral and MBT methods were observed for large normalized disbond lengths (>40 mm/mm), which is attributed to geometric nonlinearities affecting the estimate of the effective crack extension. Using the predicted load and angles, the predicted J-integral is shown to be in good agreement with corresponding experimental values. Overall, the J-integral method appears to be a robust, efficient alternative to estimate SERRs and appears to account for geometric nonlinearities.