September 2023
First Fuel Pump Flow Tests
I knocked out a couple initial tests for the Bandit Fuel pump this weekend. The test setup has the discharge of the pump recirculating to the inlet side via an orifice, with pressure transducers on either side. I started with a 0.125” Orifice which balanced out at about 77psid. I then calculated the real EOD for my design system curve, finding it to be between 0.090 and 0.100 inches. I ended up trying a 0.038” orifice just for kicks but it built too much backpressure and stalled the motor before it could build enough speed to run continuously. I drilled out an orifice that came out to be 0.098” and ran tests 5 and 6 in that configuration.
The sensors are gauge PTs that I convert to psia in data reduction but it seems like they can read some vacuum because the suction pressure reading dropped about 3psi. I’m not sure if RPT-1 bottomed out or if the real pressure was lower but the overall DP ended up being about 127psid in steady state which corresponded to about 0.128 Kg/s of flow.
Fluid power can be computed by finding the product of volumetric flow rate and pressure rise. Also since the pump is positive-displacement, the shaft speed is known for a given flow rate. The pump outputs a known mL/R and volumetric efficiency of 91% meaning I calculated expected speed to be 1736 RPM.
With a motor of a known Kv constant it is possible to calculate the current and voltage of the motor for a given torque and speed. Factoring in resistive heat loss, it is then possible to see at which point the max power output of the motor intersects with the required shaft power for given pump conditions.
The data show test 6 slightly lower on the power curve than expected. This could be due to overlooked sources of mechanical or electrical inefficiency. Another contributor could be if RPT-1 is reading higher than reality which would mean that the real pressure rise is larger than measured corresponding to a higher fluid power. For now I will add an additional efficiency coefficient of 0.86 in my calculations to account for the difference.
repeating the analysis for test #2 shows that the coefficient is higher for this test so it could be a function of speed or some other variable and might be worth looking further in to. Either way, the next steps are to increase the battery voltage to approach the actual design conditions.