Monte carlo simulation of a probability screen

The interaction of a probability screen with spherical particles was simulated in a computer model. Repeated runs were made using randomly selected, slightly differing initial conditions, from which the penetration probability was obtained at a particular particle size under given screen operating conditions. The calculations were then repeated with differing particle sizes and operating conditions, and the results were found to compare favourably with experimental data obtained in a previous study. It was shown that the stochastic model correctly predicts the effects of the principal operating variables (screen angle of inclination and aperture size, frequency and amplitude of vibration). Because of excessive computer time requirements of this model a second, semi-empirical non-stochastic model was developed whose results closely match those of the earlier model. The second model is suitable for incorporation into a plant simulation computer package. On a simule par ordinateur l'interaction d'un ecran de probabilite avec des particules spheriques. Des essais successifs ont ete effectues avec des conditions initiales selectionnees de maniere aleatoire mais quasi semblables, a partir desquels la probabilite de penetration fut obtenue a une certaines taille de particules pour des conditions de fonctionnement de l'ecran donnees. Les calculs ont ete ensuite repetes avec differentes tallies de particules et conditions operatoires, et on a trouve que les resultats se comparent favorablement aux donnees experimentales obtenues lors d'une etude anterieure. On a montre que le modele stochastique predit correctement les effets des variables operatoires principales (angle d'inclinaison de l'ecran et dimension d'ouverture, frequence et amplitude de vibration). Un second modele non stochastique, semi-empirique, a ete developpe en raison des besoins excessifs en temps de calcul, dont les resultats se rapprochent etroitement du premier modele. Le deuxieme modele peut ětre incorpore dans un modele de calcul de simulation d'installation.