Water Brake Dyno

[ABXLiaison]

I am trying to design a small engine dyno using a water pump. The problem I have getting started is pump selection. The motor I will be testing is a 50 to 60 hp maximum and 10,000 rpm maximum. My first thought is to reduce the driven speed of the pump by possibly 10:1. How do I determine the gpm and maximum pressure pump required to adequately load this engine. My plan is to increase the engine load by slowly closing a valve to increase the pump pressure. I believe that by comparing the output pressure to the rpm and torque measured off the free to rotate pump housing. The torque data would be collected using a load cell attached to the pump housing.

 

[smckennz]

This doesnt really answer your question, but may help.
I designed and built a power absorber for 400bhp 7000 rpm in my final year at uni about 100 years ago. My supervisor had worked for Heenan & Froude, so the ideas were his. The absorbing element was a fluid flywheel out of a Daimler preselector clutch/gearbox assembly; suitably modified. The rotor and stator had radial vanes inside semi toroidal housings.
Water was admitted into the "PCD" of the stator which was also fitted with vent lines. The centrifugal force from the rotor slung the water outwards through the clearance gap between the rotor and stator periphery, and into a casing bolted to the stator. A valve on the casing could be adjusted to control the water holdup in the casing and the rotor/stator. Varying the holdup varied the water shear area between the rotor and stator and hence the torque.

Using a centrifugal pump will not work that well as the range will be quite narrow at fixed speed and impeller diameter. The torque will also increase with flow rate. You could use a PD pump. To select, look at the pressure/flow/speed/power curves, and pick one that requires something in excess of 60hp (say 80hp) at a suitable speed.Stay with moderate (say around 150 psi) pressures as this will keep components at a reasonable price.

The power absorbed is a function of head and flow.
P = 60HP = 45kW
P = Flow(l/s) X S.G. (1 for water) X Head (m) X 9.81/(1000 (w/kW) X efficiency(fractional))

1m head = 10kPa
for efficiency, work on about 0.85 for a PD pump.

Cheers

Steve

 

[Artisi]

First up you need to decide what accuracy you need from this "dyno", if you want to split hairs over the input measurements, I think you should look at a different method.
Anyway, I would suggest a centrifugal pump unit that requires around 70 - 75 HP at BEP when running at 3600 RPM, couple this with a 3:1 reducing gear drive. You will then need to have this unit accurately hydraulically tested (calibrated)with the flow, pressure and power input carefully measured and plotted onto a "curve" at say 3600 rpm and 2,3 or 4 other speeds.
This will then establish the power required to either reproduce a pressure or a flow which could be read from the "curve" at a given speed. You will then need to set-up your dyno with a water loop, this will need to exactly duplicate the suction and discharge configuration of the test unit where you had it calibrated, any difference to this especially the suction side could well change the "performance" from the calibrated result.
If you think all this is worth while and you have any questions let me know and will see if I can assist you further.

Ian


International College
Naresuan University
Phitsanulok
Thailand

 

[scalleke]

What you want to do is posssible and even at the operating speeds that you are mentioning.

I would shy away form a gearbox to reduce from 10000 RPM as this will be a pack of troubles.

Centrifugal pumps can operate at 10000 RPM and easilly so. Go for an open impelkler with generous clearances the impeller can be straight vaned. You do not need a volute You can just used a concentric design with a tangental diffuser style outlet.

Only two things that will be an issue will be shaft bearings and making sure that your pump is always filled with liquid.

It will be hard to find a bearing that can operate at 10000 RPM and the type of loads that you mention.

At 10000 RPM you will get high pressures but only at the immediate outlet of the pump and immediately after the pressure control valve you will be able to work at ambient pressure.

Best Regards.
Scalleke

 

[hydrae]

My plan is to increase the engine load by slowly closing a valve to increase the pump pressure. I believe that by comparing the output pressure to the rpm and torque measured off the free to rotate pump housing. The torque data would be collected using a load cell attached to the pump housing.

Torque and power requirements are not linear in a pump, most pumps require more power wide open than closed down, so you may get the opposite effect as you close the valve the power requirement decreases. The efficiency is where the energy goes.

The waterpump type dyno that I briefly worked with at university, controlled power by the amount of water allowed to enter the pump, so under low power, the pump pushed a water/air mixture thus with the lower specifice gravity equals lower power, power was measured by a load cell preventing rotation of the pump and RPM. Operation of this unit was not clean, the water/air mixture was not stable so the torque varied and vibrated.

suggestions
I would look at a magnetic coupling for control which will allow the pump to spin at a slower speed for less loading although I do not know if they are rated for 10,0000 rpm
or better yet an electric motor (generator) and variable resitor bank?

Hydrae