CFD modeling for the prediction of molecular weight distribution in the LDPE autoclave reactor: Effects of non-ideal mixing

Sunkyu Shin, Solji Choi, Jonggeol Na, Ikhwan Jung, Min Kyu Kim, Myung June Park, Won Bo Lee

Research output: Contribution to journalArticlepeer-review

2 Scopus citations

Abstract

A commercial Low-density polyethylene (LDPE) reactor was modeled to investigate the effects of highly nonlinear hydrodynamics in an industrial-scale autoclave reactor on the polymer Molecular weight distribution (MWD). To reduce the extremely high computational burden of Computational fluid dynamics (CFD) runs and detailed MWD calculations, a combination of the CFD multicompartment model and probability generating function transformation is proposed. The validity of the proposed model was verified by a comparison with experimental data on industrial equipment. The computation time was shown to be only 37% of the actual operation time. The proposed model showed that strong downflow and weak reverse-flow coexisted to form circulation flow between the disks in the autoclave. Polymerization mainly occurred in the strong downflow region of the tubular regime. The circulation flow increased the accumulation of dead polymers to enhance the chain transfer to polymer, which produces long-branched living polymer chains and termination by recombination for a broad MWD with bimodality. Further analysis based on the simulation without the circulation flow showed a narrow MWD and no bimodality. These results indicate that the flow pattern must be properly controlled to produce polymers with desired properties in large-scale polymerization reactors.

Original languageEnglish
Article number131829
JournalChemical Engineering Journal
Volume427
DOIs
StatePublished - 1 Jan 2022

Keywords

  • Autoclave
  • Bimodality
  • CFD
  • MWD
  • Non-ideal mixing

Fingerprint

Dive into the research topics of 'CFD modeling for the prediction of molecular weight distribution in the LDPE autoclave reactor: Effects of non-ideal mixing'. Together they form a unique fingerprint.

Cite this