FUZZY LOGIC-BASED MODELING OF A CENTRIFUGAL BLOOD PUMP PERFORMANCE VIA EXPERIMENTAL DATA OF NEWTONIAN AND NON-NEWTONIAN FLUIDS

Abstract

Using fuzzy logic methods, some complex experiments that are not possible due to critical limitations can be simulated in a short time. In this study, experimental data of Newtonian 40% aqueous glycerin solution (GS) and non-Newtonian 600ppm aqueous xanthan gum solution (XGS) working fluids were used to model the hydraulic performance of a centrifugal blood pump. A novel fuzzy logic-based model (FLM) for modeling the hydraulic performance of the pump model is proposed. In the proposed model, there are two inputs which are flow rate and impeller rotational speed and one output which is head pressure. In FLM, the range for flow rate is 1-7.8L/min in GS and 1-8L/min in XGS, and for head pressure 50-245mmHg in GS and 50-215mmHg in XGS. In addition, impeller rotational speed range is 2700-3600rpm for both fluids. The estimated results with FLM were validated with the experimental results and it was seen that the FLM was compatible with the experimental results with an accuracy of 96.25%. These results imply that the developed FLM is acceptable and can be used to assist in determining the performance of blood pumps.