Predicting crackling noise in compressional deformation

Abstract

Compressional deformation of wood as a porous material is a test system for the prediction of avalanches. We analyze the predictability of ‘woodquakes’ or acoustic emission (AE) events from small wood samples, based on a geophysics analogy, the Omori production law. The main idea is that an enhanced aftershock rate leads to an increased event probability after a ‘main shock’ or an event that exceeds a pre-chosen energy threshold. We present an analysis of the resulting relation, and show that combined with an AE timeseries analysis procedure implemented on a fast signal processing setup this allows to confirm the ensuing prediction scheme. By maximizing the prediction efficiency by tuning the parameters of prediction trials like ‘when’, ‘for how long time’ and ‘for how big events’ aftershock probabilities up to close to 50% are realized and confirmed in on-line experiments. We also repeat the procedure for the main shock prediction utilizing a version of the Omori law for fore shocks. The implications of these trials are discussed in terms of the efficiency of avalanche prediction.