Settling Experiments With Iso-Optical Systems for Coalescence-Model Validation

Design of settlers can be challenging, since coalescence behavior varies widely between systems. For settler design, a numerical tool has been developed, which is based on considering the behavior if individual representative drops (ReDrop concept). In these simulations, the coalescence modeling is a major challenge due to the complex interactions of drops upon approach and coalescence. A consistent coalescence model has been proposed in the past and a standardized settling cell developed, which allows quantitative analysis of settling behavior for a specific material system. The settling cell is lighted with an LED panel from behind to ensure proper lighting and a video taken of the batch-settling experiment, which is then quantitatively evaluated. For validation of the coalescence model, quantitative data on the time-dependent holdup profile in this standardized settling cell is desired. This can be obtained by utilizing an iso-optical system, to which a dye is added. In an iso-optical system the refractive index of both liquid phases is identical so that refraction at the interface vanishes. Thus, in an iso-optical system, a dispersion appears fully transparent. Adding a dye, which partitions preferentially into one phase, then allows to quantify how much of that phase is present at any height position in the cell. To determine the drop-size distribution with a SOPAT-Probe, which is inserted into the settling cell, slight concentration changes are introduced so that the interfaces become visible and can be detected with the help of this optical probe and thus accessible to evaluation. The time-dependent holdup profile is automatically evaluated from the video taken. This evaluation also includes some corrections e.g. for inhomogeneities of LED panel used for lighting the cell as well as the camera. These quantitative data on drop sizes and time-dependent holdup are then used as a basis to validate some fundamental aspects of the coalescence modelling. The experiments e.g. clearly show the presence of a densely-packed zone in addition to the close-packed zone, which is usually assumed. In this densely-packed zone the drops are not in direct contact but settle freely at high values of holdup. This has to be appropriately modelled in the settling model. The densely-packed zone previously had been suspected from the ReDrop modelling but is now quantified experimentally (Leleu, Pfennig, 2019). The results of the ReDrop simulation are in good agreement with the experimental data