Prediction of load threshold of fibre-reinforced laminated composite panels subjected to low velocity drop-weight impact using efficient data filtering techniques

This work is concerned with physical testing of carbon fibrous laminated composite panels with low velocity drop-weight impacts from flat and round nose impactors. Eight, sixteen, and twenty-four ply panels were considered. Non-destructive damage inspections of tested specimens were conducted to approximate impact-induced damage. Recorded data were correlated to load–time, load–deflection, and energy–time history plots to interpret impact induced damage. Data filtering techniques were also applied to the noisy data that unavoidably generate due to limitations of testing and logging systems. Built-in, statistical, and numerical filters effectively predicted load thresholds for eight and sixteen ply laminates. However, flat nose impact of twenty-four ply laminates produced clipped data that can only be de-noised involving oscillatory algorithms. Data filtering and extrapolation of such data have received rare attention in the literature that needs to be investigated. The present work demonstrated filtering and extrapolation of the clipped data using Fast Fourier Convolution algorithm to predict load thresholds. Selected results were compared to the damage zones identified with C-scan and acceptable agreements have been observed. Based on the results it is proposed that use of advanced data filtering and analysis methods to data collected by the available resources has effectively enhanced data interpretations without resorting to additional resources. The methodology could be useful for efficient and reliable data analysis and impact-induced damage prediction of similar cases’ data.