Computational Fluid Dynamic Predictions of Mixing in a Double-Rushton Stirred Tank
Author
Locht, Pascal
Term
4. term
Education
Publication year
2010
Submitted on
2010-06-02
Abstract
Væskeblanding er en helt central operation i den kemiske procesindustri. For at designe gode miksere er det vigtigt at forstå, hvordan væsken bevæger sig (strømningsfeltet) under hele processen. Denne afhandling bruger Computational Fluid Dynamics (CFD) til at undersøge strømningsfeltet i en Double‑Rushton omrørt tank. Mixerens geometri blev modelleret i CFD‑softwaren Gambit, og flere simuleringer blev gennemført i Fluent for at studere udvalgte parametre. For at validere modellen sammenlignes resultaterne med et nyere studie af samme mixer, som anvender en mere kompleks modelleringsmetode. Vi analyserer ligheder og forskelle mellem vores resultater, teorien og den publicerede litteratur og vurderer, hvad de betyder for relevansen af vores modelleringsvalg. Inden simuleringerne gives en teoretisk gennemgang af omrørte beholdere, den anvendte turbulensmodel (RANS, Reynolds‑Averaged Navier–Stokes), den valgte modelleringsmetode (kildetermer, dvs. ekstra termer i ligningerne) og grundlæggende CFD‑begreber. Studiet peger på, at selv om der er visse ligheder mellem litteraturen og vores simuleringer, er det vanskeligt at opnå samme resultater, når forskellige modelleringsmetoder bruges.
Fluid mixing is a core operation in the chemical process industries. To design effective mixers, it is essential to understand how the liquid moves (the flow field) throughout operation. This thesis uses Computational Fluid Dynamics (CFD) to examine the flow field in a Double‑Rushton stirred tank. The mixer geometry was built in the CFD software Gambit, and multiple simulations were run in Fluent to explore selected parameters. To validate the modelling, the results are compared with a recent study of the same mixer that uses a more complex modelling approach. We analyze agreements and discrepancies among our simulations, theory, and the published literature, and assess what they imply for the relevance of our modelling choices. Before the simulations, we review background on agitated vessels, the turbulence model used (RANS, Reynolds‑Averaged Navier–Stokes), the chosen modelling approach (source terms, i.e., adding extra terms to the equations), and core CFD concepts. The study indicates that, although our results share some similarities with the literature, achieving the same outcomes is difficult when different modelling approaches are used.
[This abstract was generated with the help of AI]
Documents