Design and Control of Chemical Process Systems

Abstract

The MATLAB codes were modified for nonlinear recursive least square (NRLS) algorithm, nonlinear recursive orthogonal least square (NROLS) algorithm, PI and PID controllers and used to study performance of various chemical processes. The NRLS method was typically adopted to maintain cell concentration in the fermentor, outlet concentration in isothermal three CSTRs in series and temperature trajectory of free radical polymerization of methyl methyacrylate (MMA) in batch reactor. The number average molecular weight of polymethyl methyacrylate (PMMA) was controlled by using NROLS algorithm, PI and PID controllers. newlineBy using the modified MATLAB code for simulated annealing (SA) and differential evolution (DE) optimization techniques, bulk polymerization of MMA was studied to optimize conversion. The results obtained conclude that DE algorithms are very robust and efficient to find out the global optimum of a function with ease and accuracy. newlineThe ring opening polymerization (ROP) kinetics of L-lactide was studied in bulk at 150-190oC with stannous octoate [Sn(Oct)2] and 1-pyrene butanol as catalyst and co-catalyst respectively. The kinetics of L-lactide ROP follows the co-ordination insertion mechanism. The objective of present experimental work was to achieve high monomer conversion and high molecular weight polylactide. The effect of different parameters; namely, time (0-1h), temperature (150-190oC) and co-catalyst to catalyst ratio (10-30) on the properties of polylactide was studied. The rate of propagation obtained increases with increase in temperature. The depolymerization reactions dominate at the high temperature and polymerization time. For the reversible ROP of L-lactide the pre-exponential factor (A) and the activation energy (E) were calculated and the values obtained are E= 24 KJ/mol and ln A= 16.2. The experimental results obtained were studied to account for the reversible activation, propagation, termination and validated by modeling using MATLAB. The model developed successfully.