# Centrifugal Pump

This experiment has four objectives to it of which are measuring the standard performance characteristics of the pump, comparing the performance characteristics for three different Reynolds numbers and explaining the effects observed and to also make predictions about the performance for a pump with an impeller diameter of 228mm using the data obtained in the experiments. This experiment was conducted under room temperature. The motor was first adjusted to 3000 rpm and the input valve which was used to control the input pressure was adjusted in 0.1bar steps. Pump Inlet and delivery static pressure and pressure difference across the Venturi meter was recorded. The procedure was stopped when the Venturi meter pressure difference became less than 0.05bar. The process was then repeated for motor speeds of 2500 rpm and 2000 rpm. It was observed that the power input of the pump increased linearly as flow rate increased. Apart from that, the total head decreased as the flow rate was increased and this was governed by the Bernoulli’s Principle. It was calculated that the Reynolds number for speeds of 3000rpm, 2500rpm and 2000rpm are 4.1707x107, 3.4756x107 and 2.7805x107 respectively. From these numbers, it was observed that at 2000rpm, efficiency was greatest, recording 50.3% efficiency and this was possibly due to the smaller turbulent flow and thus having a smaller resistance due to the friction coefficient of the pipe. The predicted hydraulic power and head for the 228mm diameter impeller was found to be 2897.04W and 252.7kPa respectively. However, the prediction is only theoretical and might not have the same result if carried out experimentally. This is because, with increasing impeller size comes increasing flow rate which would tend give rise to unbalanced radial forces, increased shaft deflection, wears out the bearing and seal life faster and also reduces efficiency. Thus, this would result in the hydraulic power being lower than predicted. The confidence limit of the flow coefficient was calculated to be 0.02195 ± 0.00075. This flow coefficient corresponds to a head coefficient of 2.8 which has a tolerance of ± 0.1. In order to reduce the uncertainty, the experiment could be repeated several times to obtain several graphs and obtain an average value of flow coefficient. In addition to that, bends and pipe fittings can be lessened to reduce the formation of eddy currents in the fluid. Impellers, bearings and seals could also be replaced prior to the experiment to reduce fluid friction. This would include regularly checking for vibration to predict bearing damage or any misalignments.

EXPERIMENTAL SETUP

The procedure and apparatus as outlined in the lab sheets were followed to obtain the data for the following conditions. The chosen rotational speeds were 3000rpm, 2500rpm and 2000rpm. Please refer appendix for values obtained during the experiment and values for necessary working condition.

RESULTS

The figures below show the graphs of efficiency, power input (brake horse power) in kW and total head (m) against flow rate, Q (m3/s) respectively.

Figure [ 1 ]: Graph of Efficiency against Flow Rate, Q

Figure [ 2 ]: Graph of Power Input against Flow Rate, Q

Figure [ 3 ]: Graph of Total Head against Flow Rate, Q

The figures below show pump performance graph (which includes pump total head, power input and efficiency) at 3000rpm, 2500rpm and 2000rpm against flow rate. Best Efficiency Point (BEP)

Best Efficiency Point (BEP)

Figure [ 4 ]: Centrifugal Pump Performance Curves at speed of 3000rpm

Best Efficiency Point (BEP)

Best Efficiency Point (BEP)

Figure [ 5 ]: Centrifugal Pump Performance Curves at speed of 2500rpm

Best Efficiency Point (BEP)

Best Efficiency Point (BEP)

Figure [ 6 ]: Centrifugal Pump Performance Curves at speed of 2000rpm

The figures below show the graphs of head coefficient and power coefficient against flow coefficient for rotational speed of...

Please join StudyMode to read the full document