Development of a universal impeller test rig for scaled testing of high performance impellers.

Abstract

This dissertation presents the validation of a universal impeller test rig, designed by the author and constructed at the University of KwaZulu-Natal (UKZN). The research was conducted as part of UKZN’s Aerospace Systems Research Group’s (ASReG) work into liquid rocket propulsion. The rig will be used to evaluate the performance of an impeller, developed as part of ASReG’s research, for use in a hypothetical launch vehicle’s fuel turbopump. Head rise versus flow rate characteristics, as well as cavitation performance will be assessed by the rig. The power requirements of the impeller necessitated the reduction in rotational speed and geometric size of the test case. Scaling laws and dimensionless numbers were used to predict the test case performance based on the design performance. This predicted performance was then used to determine specific parameters used in the rig design. Validation of the rig and testing procedures was performed using a standard industrial KSB ETA 125 – 200 centrifugal pump, by comparing the experimental results with those of the supplier. Head rise characteristics were determined by measuring the change in pressure between the inlet and discharge of the pump and then plotted against the flow rate for varying system heads. Cavitation performance was assessed by decreasing the inlet pressure while maintaining a constant flow rate. This was performed at various flow rates within the range of operation. Head breakdown, vibration and noise levels, both in the time and frequency domains, were used to assess the cavitation performance. The head rise versus flow characteristics of the pump, determined on the rig, showed good agreement with the supplier’s data. Cavitation performance, determined by head breakdown, was also in accordance with the supplier. It was found that both the vibration and general noise levels increased, indicating the presence of cavitation, before any head breakdown was detected. By monitoring the level of the high frequency noise in particular, > 10 kHz, the presence of cavitation was detected at a significantly higher inlet pressure than would be suggested by the head breakdown approach.