High-Throughput Experiments for Rare-Event Rupture of Materials

Abstract

The conditions for rupture of a material commonly vary from sample to sample, and depend on the types of loads, such as monotonic or cyclic loads. Of great importance to applications are the conditions for rare-event rupture (e.g., rupture of 1% of samples), but their measurements require a large number of samples and consume much time. Here we measure the conditions for rare-event rupture by developing a high-throughput experiment. For each run of the experiment, we print 1,000 samples under the same nominal conditions, and pull them simultaneously using a kinematic mechanism of one degree of freedom, so that all samples are subject to the same stretch. We automate the process of identifying the rupture of individual samples by processing the video of the experiment. We load the samples monotonically, and record the stretch for each ruptured sample. We also load samples cyclically to a prescribed amplitude of stretch, and record the number of cycles for each ruptured sample. We study rare-event rupture using the Weibull distribution and the peak-over-threshold method from extreme value statistics. The high-throughput experiments enable the prediction of rare events with high accuracy and confidence.