Predicting the performance of pressure filters

As a step towards removing heuristics from the design and specification of filters, it is shown how accurate experimentation and modelling can be utilised to predict the performance of a pressure leaf filter over a range of experimental conditions. An experimental apparatus combining the principles of mechatronics with a dead-end pressure leaf filter is outlined, along with sample data which outline its versatility. Experimental data obtained from the apparatus are compared with predictions given by classical filtration theories incorporated within flexible computer simulations. Batch experimental results obtained at constant pressure with aqueous suspensions of calcite, talc and zinc sulphide are shown to match well with theoretical predictions of the important design parameters cake height and cumulative volume of filtrate, particularly when cake compressibility is lower. Constant pressure data are further used to predict constant flow filtrations where the applied pressure is manipulated via computer software.