ANÁLISE DE MODELO MATEMÁTICO MONODISPERSO PARA A REMOÇÃO DE PARTÍCULAS NA FILTRAÇÃO UTILIZANDO MÉTODO DE LEVENBERG-MARQUARDT

Abstract

Mathematical modeling for models monodisperse suspensions removal constitutes a challenge in terms of both complexity algorithmic and computational performance depending on the complexity of the interrelationships between the particles of various sizes and spatial configurations. The present work analyzes the efficiency of particle removal by comparing the direct up and down filtration, a mathematical model of monodisperse particle removal in filtration, using the method of Levenberg-Marquardt. Double filtration tests were used on a pilot scale, with data filtering up and quick filter descendant to the sensitivity analysis and calibration of models. Disparate results were obtained with the regression coefficient for multivariate equations (R2) of the order of 50% up to 99% of adherence to experimental data. These results indicate the feasibility of models and the need for further studies in the subject.

Index Terms – flocculation, mathematical modelling, re-flocculation, aggregation, ruptures.