INFORMATION ENTROPY DISTRIBUTION FOR SEQUENTIAL FRAGMENTATION IN SINGLE FIBERS

It is possible to demand that in estimating probabilities of events in the real world, one should include physical measures by taking into a ccount the phase space belonging to each event as priors (initial weig hts) of the probabilities. Data from single-fibre fragmentation experi ments on aluminium-based composites are analysed by combining phase sp ace and informational inputs to predict fragment length distributions in a macroscopic system. A single constraint (arising from the physics of the experiments) suffices to yield qualitative agreements; a furth er constraint brings quantitative fit.