Figure 9 – Final solid fat content for all 5 cases: comparison between experiment (blue, left) and simulation (purple, right).
Figure 9 illustrates that the (average) eventual solid fat content found in the simulations is in very good agreement with the experimental data, certainly given the uncertainty (represented by the error bars) in both experiments and simulations. The error bars of the simulations are based on differences found in simulations with different sets of calibration coefficients, such as in Figures 5 and 6. The experimental uncertainty is once more clear from the different values of the solid fat content of the similar tests #3 and #4: test #4 was stopped earlier than #3 and therefore should contain more oil indeed (as in the simulations), while this was not observed in the tests. In test #5, the average solid fat content was predicted too high while the outflow simulation (see Figure 7) was very well predicted.
Finally, we checked whether the simulation reproduces the solid fat profile in the eventual filter cake as found in the pilot-plant tests. To this end, the filter cake produced in test #2 was removed from the filter chamber and sliced with the help of an egg slicer into five layers of approximately equal thickness. This is a tedious operation, as oil dripped on the cake during removal and cutting was difficult, with the cake easily crumbling and melting during handling. The solid fat content of ten samples out of each slice was measured and averaged to construct