looking to set up a test for a rotary valve using a non-positive displacement pump.

[CamTheNewby]

Hello,

I am looking to set up a small test to calculate the performance capabilities of a rotary valve, and modify the part for optimal performance based on the test results. Part of the system will require a pump. I have never really looked into purchasing a pump myself, but I do know what kind of requirements there are for the system. I want it to run in a closed loop if possible, no check valve required ideally. Also, I want to test it while the valve inlet and outlets are subjected to a range of different psi pressures ranging from around 60 or 65 psi, to about 100 psi. I am hoping to just be able to leave it, so that the pressure remains at a steady psi pressure, and I can just leave it over night to run. I don't know if this would effect the choice of parts, but the diameters of the inlets and outlets are around 1.6".

Could anybody recommend what kinds of pumps/lengths of hoses or pipes would be good for this kind of experiment? I was thinking about an impeller pump of some sort, and a non-positive displacement. Any suggestions would be greatly appreciated. I'm pretty new to this stuff, so please excuse me if I sound like I don't know what I'm talking about because I don't really haha. Anyways, thanks!

 

[CamTheNewby]

as a bit of additional info, I want water to run through the pipe, there are no particulates in the water, and the flow rate is approximately 1 gallon per minute

 

[LittleInch]

This rotary valve of yours, Is it an on / off type valve that you're testing or more like a fixed speed PD meter which delivers packets of liquid at a fixed flowrate? what sort of pressure difference are you going to see across the valve?

I think you're probably better off using some sort of accumulator and dumping your output via a control valve into a sump tank and then pumping it from there back into your accumulator which would allow you to vary the pressure inlet and out let quite easily. This would need a PD pump really with a pressure relief back to the same sump to allow pressure ratios of 60 to 100 psi, or you could get a variable speed pump to vary the pressure, especially if your flow is a regular flow, not on/off.

1 gallon a minute is fairly small - you're into lab type and injection pump sizes here.

If you've got any sort of concept sketched out, post it - it all adds to the info which is in your head, but not ours.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[CamTheNewby]

supposedly (and this is kind of counter intuitive to me), I can use a non positive displacement pump and create a dead head situation (no flow rate at all, 0 gallons per minute), and then depending on the rigidity of the impeller, i am able to create a pressure depending on the amount of power delivered to the pump. Can i get verification on this? Or a suggested type of pump to use? the rotary valve is designed to run at 75 psi, and we want to mimic this. Forget the range, just fixed, unless it is possible to achieve a varying range using a non positive displacement pump.

 

[LittleInch]

The thing is that pumps like to pump, that's what they're designed for and so for your application you need something a bit different to allow the pump to operate properly. How much time is spent flowing as opposed to not flowing? The pressure or head developed by the pump is a function of speed, size of the impellor for a centrifugal pump and inlet pressure. Power has nothing to do with it so long as there is enough to turn the pump at its fixed speed. The way to change pressure is to change speed.

However if the pump is not pumping for more than about 20% of the time and for more than 20-30 seconds, it won't like it and ou need to allow for some sort of minimum flow bypass.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[CamTheNewby]

well, i think you bring up an interesting point, and I fully agreed with you, but i have someone who i am speaking to who seems absolutely sure that if you use a non positive displacement pump, then you are going to get some back from the flexible impeller, and so there will be a flow created, but it might just not be in a loop like one would expect, but.. maybe something like if the pipe was 30 feet in diameter, and there was wind flowing around in it. Some goes one way, and some goes another way. Eddys and such.. Maybe something to do with flow turbulence. I don't think it all needs to be going in the same direction. I'm pretty sure that the intention is to run the pump with a fully restricted pipe somewhere down the way, and to maintain a steady 75 psi.

 

[LittleInch]

There is a grain of truth in what is said as flow is never stable at no flow conditions, but for all intents and purposes there is no flow through the pump. At the size you are talking about there probably is no real problem, but I don't understand why you're so against a PD type pump (gear or some other pulseless pump with a full flow relief able to be adjusted to whatever pressure you want. You wouldn't really want this in a commercial situation, but for testing at low volumes and pressure you don't really care about a few watts being thrown away. This would allow you to vary your input pressure, start stop for as long as you want providing the relief line goes back t some sort of decent sized dump tank. PD pumps just want to pump a fixed volume at the speed you run it at and so long as your motor can accommodate the highest pressure you need then regulate the pressure by means of the re-circulation valve.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[btrueblood]

Regardless of the type of pump, realize that all of the power consumed by the pump will be dissipated in the fluid stream as heat. For your closed loop, expect the temperature to rise, and eventually peak at some level much warmer than ambient air. I do this type of testing "all the time" for products my company builds, and use a small cooler (HX with a fan) with a thermostat to adjust the temperature. With a decent pipe insulation, I can quite easily push water temperatures up over 200 deg. F or higher on pump power alone, for a small enough loop.

 

[CamTheNewby]

Well, I really appreciate the help guys. Thanks for all the info. We are still looking around for the type of pump that's going to meet our exact requirements, and it's likely to be a flexible impeller pump. Consideration will definitely be taken for what kind of piping we use, in order to make sure that the heat generated is not too great. LittleInch, if it were me making the final decision on this, I would probably go with your proposal, but since its not my money that we're using, I don't really have the final decision. If the proposed method of the flexible impeller pump with a dead head situation doesn't work, then we will likely go with something like what you suggest.

 

[LittleInch]

You need to satisfy my curiosity now as to why you think a flexible impeller pump is going to be your chosen type. I had to go look it up to see what they were, but from what I can see they don't offer you any advantages over a more conventional pump type. The think we've not talked about properly is what your pressure and flow is doing wile you cycle this valve. A small pump probably won't mind suddenly running to the end of the curve and back, but it won't do it any good and will introduce a level of surge and vibration if the flow goes from nothing to flat out then back again. A big accumulator and some sort of flow restriction d/s your valve would probably help.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way