Statistics of acoustic emission in paper fracture: precursors and criticality

We present statistical analysis of acoustic emission (AE) data from tensile experiments on paper sheets, loading mode I, with samples broken under strain control. The results are based on 100 experiments on unnotched samples and 70 samples with a long initial edge notch. First, AE energy release and AE event rates are considered for both cases, to test for the presence of "critical points" in fracture. For AE energy, no clear signatures are found, whereas the main finding is that the event rate diverges when a sample-dependent "critical time" of the maximum event rate is approached. This takes place after the maximum stress is reached. The results are compared with statistical fracture models of heterogenous materials. We also discuss the dependence of the AE energy and event interval distributions on average event rates.