Testing Two Types of Centrifugal Pumps

Are you trying to do a direct comparison of these two pumps at identical conditions of head and flow? Are you looking at abrasive particle wear or wear due to erosion/corrosion or cavitation etc? Do you need to test the pumps at a specific point on the curve? All these questions and many others will govern your test set up.

Running them in parallel will mean that the suction pressure and the discharge pressure for each pump will be fundamentally the same if you get the piping layout correct. There will be slight differences of differential head across the pumps due to nozzle sizes, though depending on flow these may not be significant. To know whether this is possible you would need to show us the pump curves and the system curve.

Putting the pumps in series gives very different suction conditions for the two pumps.

Yes, I want to do a direct comparison of these two pumps at identical condition of head and flow. I am looking at wear due to both erosion/corrosion and cavitation. I am looking into spinning the smaller pump at 2400 rpm and the bigger one at 1200 rpm.

You cannot directly compare machines at two different speeds, for all possible parameters. Mechanical wear & tear goes up according to the square of the speed, in round numbers. If you are comparing two pumps, selecting one or the other for the same application, then maybe that is valid, but you are going to be years getting reasonable or useful data

Well I have had some experience with pump testing and using pumps as motor loads so (for what it's worth), if I were building a rig to test two pumps I would have a large central container holding the test liquid (water) and the two pumps (one on either side of the container) sucking water out of the bottom of the container and then pumping it into the top of the container.

I would make the container deep enough so that the outflow water being pumped back into the tank did not interfere with the pump inflow (same with the tank width and spacing between the two pumps). Valves located in the outflow pipes allow for setting of the water flow and therefore pump motor load.

Due to the closed-loop nature of the test rig design the two pumps don't have to run at the same flow rate. One thing to be aware of however is that the water heats up due to pump mechanical friction so evaporation needs to be controlled.

For pumps of this size using a large residential water storage tank may be the cheapest option.

Hope this gives you a few ideas.

looks promising, cheap and simple, just will take more space than I allotted to it.

I'm just not sure where to start with the dimentions of the tank, this is my first crack at testing drives under load on a bench, so I would greatly appreciate if you elaborate.

The rotary converter's max load is 10hp and I doubt if I ever go there, these are going to be a functional tests of the VFD in range from 0.5 to 7hp, so I'm assuming water cooling shouldn't be a problem, so 50 liters, 100l?

I'm also not sure how to select the motor to cover such a range of loads.., two different motors with pumps, for lower and higher range?, doesn't sound good to me.

May be resistive load for smaller drives?, so as you can see I'm groping.

So far what I fix goes into the original application so I have not had to worry about it.

The first two answers you got were on the mark. You also did not indicate the impeller diameter. Generally pumps have 3 dimensions given, Suction, discharge and max diameter which would be indicative of the Total Discharge Head possible with the pump which combined with the specific gravity of the pumped fluid would give you a differential pressure that the pump could produce given enough HP input.

If you are making a comparison you would want to maintain the same hydraulic conditions on each, same inlet pressure and same discharge pressure so you will need valves on both sides of the pumps so you can throttle back to maintain the same pressures I would think.

Like the man said you haven't really told us What you're comparing. Obviously the capacities will be quite different between a 6x5 pump and a 8x6 pump unless they're running at different RPM.

You could be comparing a 6x5x12 running @3600RPM to a 8x6x21 running @ 1800 RPM and checking for inlet vane erosion for all we know.

As far as giving good answers, remember, as in computers, the answers can often be no better than the input of the question.(Garbage in = Garbage out)

Why is there a need to do this test?

I used horse troughs instead of tanks. Suction above the trough and discharge metered back in the trough. I was testing hydraulic motors connected to centrifugal pumps.

This doesn't make any sense to me, any pumping system should be designed for purpose....If you want to compare two systems that seems to me could be done much cheaper and faster using a software program designed for that....

...the rest is just quality and compatibility of materials...

....and proper installation configuration

As I understand it, you are wanting to determine efficiency and durability for the two pumps and compare these for identical conditions.

Durability will be very hard to determine unless you are willing to spend a few years. (Assuming you don't have test conditions that cause suction cavitation.)

The rest will require instrumentation including:

• suction & discharge pressure gauges
• flow meter
• metering of input electrical energy (amps, volts, power factor, frequency, etc.) If using a VSD this is usually available on the display.
• a way to determine slip between input frequency and shaft speed
• acoustic measurement of any cavitation.
• thermometer

I strongly recommend against any valves to reduce the inlet flow/pressure. You will need a discharge throttling valve (butterfly may work well). I suspect you will need a tank of at least 10,000 liters (or size sufficient for at least 10 minutes of pump flow). You may need a way to control the water temperature. The design or inlet piping will determine if the flow is laminar or turbulent, as these will affect wear, etc.

Then you can begin. If you are controlling conditions for same flow rates and same inlet pressure, the discharge pressure will vary. If controlling the pressures, the flow rates will vary.

Are you trying to create pump curves? or working with existing ones to look at efficiency?

You can estimate pump durability from a comparison of pump speeds.

Sounds to me like you need a carefully written protocol for the testing, approved by whomever is interested in the results.

--JMM

Helllo it is important the flow and tdh requieres to rub the test. For example if you want both at the same flow and same tdh you can run the test in parallel. Give me more data of your required test to understand your goals