One chromatography column developer was having challenges scaling up an existing design to a larger version. Our team worked with the manufacturer using Computational Fluid Dynamics (CFD) to understand the reasons for the reduced performance. CFD modeling was used to pinpoint reduced performance when moving from existing, small scale units to upscaled columns.

In the work we peformed, we were able to validate a CFD model and calibrate porous dispersion coefficients for the existing column. Moving to the larger column, however, initial results could not be reproduced with CFD. It was found that for the upscaled column, the synthetic textile mesh that forms the boundary between the distributor and the packed bed had deformed under the packing pressure of the column, such that the distributor was pinched off, and the outer radius of the packed bed was starved of flow. Computing the deformation of this mesh around the radial rib supports—and including this effect in the CFD simulations—showed that the upscale experimental results could be reproduced. The solution was verified in subsequent experiments.

Through a recommended design change, we were able to show that the high level of performance typical of the existing column could be reproduced in the upscaled column. The improved design is also able to be packed more consistently, and operates at a lower pressure loss than the original.