[martinw]

Dear All,

I apologise if this in the wrong forum.

I need to measure the pressure drop though a water pipe line at different flow rates, and I am scared. My original idea was to put a couple of static pressure gauges at each end of the pipeline, each on a Tee at right angles to the main flow, but then, through the dim mists of time, I remembered Bernoulli's principle, which suggests that the flow through the pipe line will create a suction force that will be read by each dial gauge.

My guess is that if it is a perfectly straight pipeline, with identical entry and exit points near the Tees where the dial gauges are, there will be no need to worry because the suction forces will be equal at both gauges.

In my case, there may be different turbulent conditions at each end that might louse things up. My knowledge of fluid dynamics is about zero, but would it help to severely restrict the pipe diameter leading into the dead end Tees that have the static pressure gauges????

Any input, or other suggestions would be hugely appreciated.

Best wishes,

Martin

[Geof]

I would say your guess is correct; both gauges can be expected to have the same bias.

I seem to recall, vaguely, about having a pitot tube facing downstream to avoid the effect you are concerned about but that is probably gilding the lily.

What you can expect to find, I think with this experiment is that your pressure drop remains fairly linear with flow rate up to a certain point and then drops dramatically. The drop is when the flow is fast enough that turbulent flow not laminar flow occurs; although I think this is dependent on the pipe diameter and the viscosity so you may not see any change.

[martinw]

Dear Geof,

Thanks.

I guess what I'm asking is that if I have laminar flow at one end of the pipeline and a somewhat more turbulent one at the other end, is there a way of reducing the effect of any bogus differential reading between the two static gauges?
Do you think it would help to severely throttle the pipe diameters leading into the Tees that lead to the static gauges? As I said, my knowelge of Mr Bernoulli's work is strictly limited.

Best wishes,

Martin

[Geof]

I think rather than 'throttling' the diameter in the tees you need to have a very tiny hole in the wall of your main flow pipe. Or maybe an insert in the tee that has a radiused end to match the bore of the pipe.

I wonder if you are maybe being too picky. What type of precision are you aiming for? Plus/minus 20% or 0.2%.

Try different setups and the same initial driving pressure and flow rate to see if the pressure difference varies. If you find that it does I think then you will be faced with taking measurements at several different orifices for the side tap and then extrapolate to zero to get the best value.

[martinw]

I think rather than 'throttling' the diameter in the tees you need to have a very tiny hole in the wall of your main flow pipe. .

Dear Geof,

This is absolutely not aerospace stuff, and we can re-jig things very easily. We can live with an accuracy of a few percent. I absolutely agree that to do this "aerospace" we would need profiled tappings. Many thanks for the advice.

Best wishes,

Martin

[groomden]

would this help?
http://www.efunda.com/formulae/fluid..._flowmeter.cfm

looks easy enough to set up and there seems to be some of that 'science' stuff behind it too

edit: to clarify....
well, that's more to show the principle, if you do a search there's lots of software to calculate pressure drop based on orifice plates in a system....

[Bubba]

I think you will see more variation based on the "quality" of the gage. I have run tests similar to this for years on underground mains. Also, don't forget to take into account any elevation changes. For fresh water, it is .432 lb per ft of head. IF you can obtain static conditions, you can measure the static pressure on each end and the difference will indicate the elevation change (if your gages have been properly calibrated).

[martinw]

Dear groomden, Bubba, and Geof,

I am enormously grateful for your input.

The "pipeline" is only about 15 metres long, with many twists and turns, and the ends are only about 1.2 metres apart vertically. Thanks for the static head stuff Bubba, I completely forgot about that!

The pipeline is very small bore, and there will probably be a pressure drop of about 1 bar
(say 15 psi) at the flow rates we are going to be using.

Groomden, I am familiar with orifice plates and have used them for measuring air flows, but you and Geof got me thinking about the shape of the tappings into the Tee for the static gauges. The pipeline has an equal flow rate at each end, (eg no leaking branches) , so my current guess is that as long as I put the pressure tappings at equally turbulent or laminar sections at each end, I should be OK. My suspicion is that, because turbulent flow is extremely unpredictable, the Tees will have to go into a straight section of laminar flow pipe that conforms to the the general standard: about 10 upstream and 5 downstream diameters, or something (I forget the figures) . If I stick to those rules, maybe the inlet to the Tees will not matter too much as long as they are about the same.



Thanks,

Best wishes,

Martin

[Bubba]

Martin,
I think your dead on with the pipe diameters and the straight sections, putting the gages in the Tees, could be a disaster from a data taking standpoint.

[martinw]

Martin,
I think your dead on with the pipe diameters and the straight sections, putting the gages in the Tees, could be a disaster from a data taking standpoint.

Dear Bubba,

ARRRRGH!

Sorry, I thought that I could put the pressure gauges on a Tee off a straight length of pipe that was experiencing laminar flow. There would be the same similar identical arrangement at each end of the "test section". Do you reckon that would be OK? My guess is yes.

Sorry, I am probably being stupid. Hard day...

Best wishes,

Martin

[Bubba]

I think I agree with you as long as the straight length is long enough (10 dia???)

also, you must have piping downstream of the gage or it will simply read atmospheric (0 psig) if it is right at the end of the pipe and it is dumping into the open. (Obviously, I don't know your piping system)

[Geof]

What is your high pressure at the inlet end?

[martinw]

What is your high pressure at the inlet end?

Dear Geof,

OK, I have not explained the concept very well. My apologies to all who have taken the trouble to give advice on this thread.

Here goes.

1)There is a header tank that contains water. It is an open tank.

2) There is an outlet at the bottom of the tank.

3) The pipe drops vertically about 36" (say 1 metre) and goes into the inlet of a centrifugal pump.

4) The pump can deliver a flow of about 2000 litres per hour against a head of 2 bar (say 30 psi). It may not be asked to do so.

5) The water is pumped through various "bits and bobs" before entering a test section. Calculations suggest that there will be a pressure drop of at least 1 bar through the test section

6) The water exits the "test section", goes through a few more "bits and bobs", and is then discharged though an open vent pipe over the header tank.

7) A pretty simple circuit.

All I need to do is get a handle on the pressure drop through the "test section" at different flow rates. Height between the pressure tappings is approx 1 yard/metre


8) I can put a by-pass after the pump back into the header tank, to adjust the flow into the test section.

This is not aerospace... 5% would be fine on the delta P.

I probably have not answered your question. My guess is that the pressure above atmospheric, at the entrance to the test section is about 0.2 bar max when the pressure through the "bits and bobs" at the on-coil side is added in.

Best wishes,

Martin

[philbur]

Your limit on accuracy is going to be dependent on the gauges you use. Are you planning on proper instrument gauges or going to rely on regular general purpose gauges. You can buy delta pressure gauges which will give you a much better resolution. I have used these before to make a reasonable accurate flow meter when combined with an orifice plate.

Have you done any calcs to see what Reynolds number range you have? This will tell you whether you are in the transition area between turbulent or laminar flow. Transition zone is between a Reynolds number of 2000 and 4000.

An excellent book, for practical engineers, when calculating flow behavior in piping systems is "Fluid Mechanics by Frank M White". Another but less detailed is "Flow of fluids" by Crane

Regards
Phil

Dear All,

I apologise if this in the wrong forum.

I need to measure the pressure drop though a water pipe line at different flow rates, and I am scared. My original idea was to put a couple of static pressure gauges at each end of the pipeline, each on a Tee at right angles to the main flow, but then, through the dim mists of time, I remembered Bernoulli's principle, which suggests that the flow through the pipe line will create a suction force that will be read by each dial gauge.

My guess is that if it is a perfectly straight pipeline, with identical entry and exit points near the Tees where the dial gauges are, there will be no need to worry because the suction forces will be equal at both gauges.

In my case, there may be different turbulent conditions at each end that might louse things up. My knowledge of fluid dynamics is about zero, but would it help to severely restrict the pipe diameter leading into the dead end Tees that have the static pressure gauges????

Any input, or other suggestions would be hugely appreciated.

Best wishes,

Martin

[martinw]

Dear philbur,

Thank-you for your advice.

We do not need high precision. Yes, a differential gauge might give a bit more accuracy, but the tappings into the pipeline question remains the same whether we use a differential gauge or two individual static ones.

My guess is that if we put straight lengths of pipe before the Tee, and after, things might work, as long as the straight lengths are the same at both ends of the test section. Say 10 upstream internal pipe diameters, and 5 downstream.

Quite how I test this idea is , at present, unknown.

Thank-you.

Best wishes,

Martin

[NC Cams]

I second the motion for the delta P type gage. They are quite adept at sorting out and/or averaging some of the complicating bits and bobs the crop up in trying to find flow losses in already piped systems.

Regarding the differential capability, thsi depends on a bunch of other things.

Let's say your one test point is 12-15 inches above the other (mind isn't capable of working in metric, sorry, had dialysis about 4-5 hours ago and I'm just not well yet), aand you have a Delta P capability on the gage, well you used up your measurement range via the height differential 'tween your two test points.

If there is a throttle of some/any sort 'tween the test points, you run the risk of going from essentially laminar, non cavitated flwo to anything but. Difference potential exists but how much, how bad, I dunno.

The common wisdom to use a "straghtening section" of 10 x D is well founded and should be utilized, YOu can also eliminate/tune each measurement adaptor/section, by doing linearity flow evuations on each just to see/make srue they are both at least comparable under the same flow conditions.

I had a similar flow sort of deal that we had to do to simulate oil flow potential in an engine - we used air as the flow agent and simple delta P monometers. We were looking for CHANGE, noe absolute flow change. It was a long story, after the fact, CYA deal that my manager wanted proof of. He knew the idea would work, he just wanted some paperwork to CHA when teh VP's cam down squacking - Reason: we were in the process of invalidation the VP's buddy as a vendor and VP's buddies carry more juice when it comes to requiring PROOF that their parts are a POS.

Bottom line, simple gaging works if you THINK it thru. Coming here was a good idea. Surely some of the older farts here have BTDT.

[martinw]

Dear NC Cams,

Thank-you for that, I hope this finds you well.

We need a measurement of absolute pressure as will as the diff. between each end of the test section, so two gauges make sense. This job is rough science, and I am enormously grateful for the advice from everybody who took the time to post.

I will let you know how things go.

Best wishes,

Martin

[ckelloug]

I'm late to the party but I'd think that first you would want to figure out Reynolds number for the system and determine if the flow is laminar anywhere in the system.

If it's not, and I suspect that it won't be given your flow rate without a big pipe, then you aren't going to have to worry about whether the flow is laminar on one side of the system and turbulent on the other.

The actual diameter of the pipe, the material being transported and the flow rate are critical to getting any meaningful answer. The answers to virtually all of your questions change dependent on the exact values.

After that, it's a matter of looking in a chart of fluids equations and finding the equation for pipe pressure drop and plugging into it optionally looking up the loss data if available from your pipe, valve and fitting manufacturers.

Once you know what size value you are trying to measure, then you know what to measure it with. See http://en.wikipedia.org/wiki/Manometer

Good luck. It's been a while since I've talked to you. It looks like the last person interested in E/G is me

[martinw]

Dear ckelloug,

I very much doubt that the EC thread does not gain from your expertise. Respect.


My take on the delta P measurement is pretty simple for this very small pipeline.

1) the flow rate at each end is identical

2) if the tappings to the gauges are identical , and the flow in the main pipeline is the same (laminar or whatever) , my guess is that the "whole thing comes out in the wash".

Same conditions at both ends of the pipeline, no problem.

I will report back.

Thanks for the input.

Best wishes,

Martin

[Geof]

I'm late to the party but I'd think that first you would want to figure out Reynolds number for the system and determine if the flow is laminar anywhere in the system.....
Once you know what size value you are trying to measure, then you know what to measure it with. See http://en.wikipedia.org/wiki/Manometer

Good luck. It's been a while since I've talked to you. It looks like the last person interested in E/G is me

I am puzzled.

Surely if you know the Reynolds number for the system you just calculate the pressure drop for varous flow rates; measurement is not essential. Although you are assuming laminar flow.

By taking measurements of the pressure drop at different flow rates you are obtaining the data to calculate and effective Reynolds number. Probably to be rigorous the pressure drop versus flow should be graphed to get an indication whether you stayed within a laminar flow region.