Application of Parameter Selection and Estimation Techniques in a Thermal Styrene Polymerization Model

Abstract

A model is developed to describe thermally‐initiated polymerization of styrene between 100 and 170 °C. The model accounts for generation and consumption of styrene adduct. Chain transfer to adduct is the only transfer reaction used. Autoacceleration is modeled using the break‐point method of Hui and Hamielec. Using formal ranking and parameter selection techniques that account for parameter sensitivity, correlation and uncertainty, 4 of the 40 model parameters are selected for estimation to improve fit between model predictions and data. After estimation, the model predicts conversion data with a standard error of 5%, and provides excellent fit to a MWD curve obtained at 100 °C. Simulation results confirm that high‐temperature degradation reactions are not important in the temperature range of interest.

A complete understanding of the reaction kinetics for the industrial production process of expandable polystyrene

“…The first model is the w p ‐model (entry 1 in Table ), which is a purely empirical model with an exponential decay for the (average) apparent termination rate coefficient depending only on the polymer mass fraction w p . The second model also relates to an average apparent termination coefficient and is the one previously used by Woloszyn et al for the simulation of the same EPS process . It is a free volume‐based diffusion model (entry 2 in Table ), following the Vrentas and Duda free volume theory and making a differentiation between translational, segmental, and reaction diffusion .…”

“…For completeness, it is mentioned here that the other FRP data, that is, those in the absence of blowing agent, were taken from the work of Woloszyn and McAuley. 65…”

“…The second model also relates to an average apparent termination coefficient and is the one previously used by Woloszyn et al for the simulation of the same EPS process. 24,65 It is a free volume-based diffusion model (entry 2 in Table 2), following the Vrentas and Duda free volume theory and making a differentiation between translational, segmental, and reaction diffusion. [93][94][95][96] It should be noted that several variants of this model 2 exist, including those in which a chain length dependency is explicitly included in the diffusional contribution.…”

“…[93][94][95][96] It should be noted that several variants of this model 2 exist, including those in which a chain length dependency is explicitly included in the diffusional contribution. [97][98][99] In the present work, it was decided to only use the most extensive and recent model by Woloszyn and McAuley 24,65 that can be seen as an extension of the diffusion model applied by Kotoulas et al 23 Note that also a pure "series" model can be used (only use of the most outer equation of entry 2 in Table 2). 22,30,64 This diffusion model has been particularly proposed to be formally representative for the description of radical polymerization processes with monomers leading to less pronounced impact of diffusional limitations.…”

Exploring the Full Potential of Reversible Deactivation Radical Polymerization Using Pareto-Optimal Fronts

“…A higher Pterm leads to an increase in Ð but also to a lowering of EGF. In the case of the NMP of styrene, the reaction probability for chain transfer to dimer (PtrD; Figure 7f) needs also to be considered, specifically at elevated temperature [62,74].…”