Non-spherical 2-dimensional particle size analysis from chord measurementsusing Bayes' theorem

A method of deconvoluting 2-dimensional particle size distributions from ch ord size data is presented and evaluated. This is the Probability Apportion ing Method (PAM3). It assumes that the particles (or droplets) can be repre sented by super quadrics and are cut randomly by a sensor to give a chord m easurement. Starting from an assumed uniform particle distribution, Bayes' theorem is used to calculate hit probabilities for each particle type and t he population is then recalculated. The process is then repeated until ther e is no significant further change in the calculated distribution. Using nu merical simulations PAM3 is shown to be quite accurate and robust for a num ber of different types of particle shapes provided there is a sufficient nu mber of accurate measurements.