PhD thesis: Numerical Modeling and Simulation of Polymerization Reactions in Coiled Flow Inverters

PhD defense : Dhiraj GARG

Team : MecaFlu

Title : Numerical Modeling and Simulation of Polymerization Reactions in Coiled Flow Inverters

Abstract : This thesis aimed at improving the modeling and simulation of free radical polymerization (FRP) in batch as well as in flow reactors. A generalized explicit analytical solution (AS) was obtained in case of variable volume, bulk/solution polymerization, homogeneous and isothermal batch reactor. The reaction steps included initiation, propagation, transfer to monomer, transfer to solvent, transfer to chain transfer agent (CTA), termination by combination and disproportionation. Different models of gel, glass and cage effects were also implemented explicitly. AS was validated against numerical solutions as well as published experimental data and was found in good agreement. Furthermore, its applicability was extended to conditions for which AS was not derived, i.e. non-isothermal conditions. The versatility and flexibility of AS over the complete range of monomer conversion were thus demonstrated. Then, to broaden even more its applications range, AS was used in CFD simulations. A new and simple transformation was proposed to make kinetic rate coefficients dimensionless in terms of concentration. This enabled chemical data to be fed in molar form to CFD modeling. It also enabled easy coding and debugging by keeping the original form of generation terms intact. The results were found to be improved after validation against experimental data. This transformation was then used for evaluating three tubular microreactor geometries, namely straight tube reactor (STR), coiled tube reactor (CTR) and coil flow inverter reactor (CFIR), under different feed conditions (unmixed or perfectly mixed) at very low Reynolds numbers (<1). The modeling for FRP was performed with constant or variable fluid physical parameters (density, viscosity and thermal conductivity) along with discrete variation of diffusion coefficients. Their effects on simulation results were observed and compared with published experimental data for 4 different monomers and were found to match perfectly. Results for mixed feed condition were found to be independent of microreactor geometry. CFIR seems to be the most promising reactor design under microreaction investigated conditions as it allowed the best control over polymer characteristics.

The presentation will take place on Tuesday March 14th at 2.00pm in the seminar room, ICube (4 rue Boussingault, Strasbourg).