Computers & Chemical Engineering, Vol.106, 322-338, 2017
A novel factorial design search to determine realizable constant sets for a multi-mechanism model of mixing sensitive precipitation
A mechanistic model for the reactive precipitation of amine mono- and di-hydrochloride (AHC) salts was developed in the first stage of this work (Komrakova and Maluta, 2016). In a companion paper, an extensive experimental investigation of the reaction products is reported (Ershad, 2013). The experimental work shows a sudden shift from mono- to di-production as the blend strength of 4, 4'-methylene dianiline (MDA) decreases, and a strong dependence of di-concentration on excess HCl in the feed for the high blend strength experiments. Due to the complexity of the reaction-precipitation system, eleven constants related to physical properties and rates were identified in the model. An extensive search of the literature returned only realistic ranges for most of these constants. In this work, a novel method for reducing the uncertainty in the space defined by these eleven constants is presented. The trends in the experimental data were successfully reproduced and the model results provide a number of important insights into the two complex reaction outcomes. A factorial design search for similar loosely bounded parameter spaces is proposed. (C) 2017 Elsevier Ltd. All rights reserved.
Keywords:Amine hydrochloride salts;Reactive precipitation;Full factorial design;Mechanistic model;Mixing;Optimization