ChromaTech: A discontinuous Galerkin spectral element simulator for preparative liquid chromatography
•ChromaTech is presented to solve a liquid chromatography general rate model.•The discontinuous Galerkin spectral element method is shown to be stable and fast.•The performance of ChromaTech is tested by comparing against CADET.•Superior efficiency of spectral methods over second order spatially acc...
Gespeichert in:
Veröffentlicht in: | Computers & chemical engineering 2020-10, Vol.141, p.107012, Article 107012 |
---|---|
Hauptverfasser: | , , , , |
Format: | Artikel |
Sprache: | eng |
Schlagworte: | |
Online-Zugang: | Volltext |
Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
Zusammenfassung: | •ChromaTech is presented to solve a liquid chromatography general rate model.•The discontinuous Galerkin spectral element method is shown to be stable and fast.•The performance of ChromaTech is tested by comparing against CADET.•Superior efficiency of spectral methods over second order spatially accurate methods is shown.•ChromaTech is suitable as a building block in advanced model-based digitalization strategies.
ChromaTech is a simulator for preparative liquid chromatography processes with mass transport described by the pore and surface diffusion general rate model. A discontinuous Galerkin spectral element method is used for spatial discretization with exponential decay of approximation errors within elements. The code is validated by numerically reproducing a high-precision reference obtained with CADET-semi-analytic. The performance of ChromaTech is tested by comparing against CADET, a dedicated code based on a weighted essentially non-oscillatory finite volume method with second (low) order spatial accuracy. Reassuringly, ChromaTech provides exactly the same chromatograms as CADET for multicomponent protein purification cases with linear and non-linear adsorption dynamics. However, the numerical results show, that ChromaTech has superior efficiency in terms of computational cost and discrete problem size without compromising stability. The spatial discretization is the major difference between the two codes for solution of the pore and surface diffusion general rate model. Thus, it demonstrates, that spectral methods are not just competitive with second (low) order accurate methods often used by default, but simply a superior approach for spatial discretization of liquid chromatography flow problems in terms of computational efficiency. |
---|---|
ISSN: | 0098-1354 1873-4375 |
DOI: | 10.1016/j.compchemeng.2020.107012 |