Pitot Bench Operation

[roepke44]

So I am new to this forum, and i thank you for all the good info. I have a question about the pitot style bench operation. here goes...

So the flow in a pipe has a certai velocity profile across any given cross section. In fully developed, laminar, no pressure gradient, flow it is a parabola. in fully developed turbulent flow, it is almost constant across the section. But in general, this profile may or may not even be axisymetric due to the short pipe lengths and multiple pipe bends in the measurement circuits. Also, this profile could change depending on the flow conditions.

So do people with Pitot benches do a sweep of a cross section for each test condition and then find an average velocity and use that in their mass flow rate calc? do you just assume that the velocity is uniform and use flow numbers for comparison purposes and don't care about the systematic error? how do people handle this?

One way i can think of is to set up a traverse system, where i measure the velocity profile (measure te velocity at multiple points from one wall to the next) for every test point. This would take a really long time to complete a test. The other way i would think to handle this is to use a flow straightener section and long sections of pipe before the pitot probe. I would then spend one good afternoon doing miltiple sweeps at all flow rates achievable and make sure my flow was uniform & axisymmetric. then correlate the average velocity to a measured velocity of the pitot-static probe. But this would add flow losses to the system and reduce the maximum flow rate.

I am trying to decide on which kind of bench to build for small motorcycle engine parts (< 50 cfm). I'm just trying to make sure i understand what is normaly done before i build. Thanks for your input.
-Josh

[jfholm]

Josh,
First welcome to the forum!
Second, Thank you so much for the great argument "for" orifice style flow benches.

John

[Tony]

Welcome to the Forum Josh.

I am not a pitot guy, and have never built a pitot bench, and know next to nothing about them...

But I believe one common technique is to use an "averaging" pitot tube, that has several impact velocity measurement holes located right across the flow cross section.

That is supposed to give a fairly representative figure for impact velocity over a range of laminar and disturbed flow conditions.
All very dark mysterious science, which is why I much prefer orifice plate flow measurement.

Bruce can probably tell you much more about this, he was once a dedicated pitot disciple, before he came over from the dark side.

[roepke44]

First off, Thanks for the warm welcome.

I have my doubts about some aspects of orifice plate systems, but i will save those for another thread. For now, let's just leave it at i think a pitot style system can be more accurate (less sources of errors), simpler, and have a wider operating range.

-Josh

[Tony]

I think a pitot style system can be more accurate (less sources of errors), simpler, and have a wider operating range.

-Josh

Have to respectfully disagree there on all counts Josh.......

The basic idea of an orifice flow bench is that you flow the SAME AIR through two constrictions placed in series and measure the relative pressure drops.
This is a true ratiometric measurement.
One constriction is your unknown test piece, the other constriction your reference test orifice.

The "quality" of the flowing air makes no difference to that measurement.
Summer, winter, or up a very high mountain, makes no difference to the reading.
You need make no corrections to the measurement, it will always read the same.

What could be simpler, or more REPEATABLE than that ?

Realize that we are not trying to figure out the exact number of air molecules flowing through a port under some standard operating conditions. We do not need a highly accurate absolute flow measurement figure, and that is an extremely important distinction.

All we need to know is if flow has increased or decreased by some very small amount (with respect to our reference orifice) since the last time it was tested, and the ambient test conditions may have changed hugely over that time.

Anyhow, that is the basic philosophy behind an orifice style bench, and it is what most people here have come to accept as the best test and measurement technique for our needs.

We need simplicity, stability, and REPEATABILITY with fewest sources of introduced error to detect and quantify very small changes from test to test, not an accurate absolute mass flow measurement.

[roepke44]

I can see the great value in not having to make temperature or pressure or humidy corrections. that is a great feature of the orifice style system. and yes, there is some error in measuring temperature and pressure to use in these corrections, but i am somewhat confident in my ability to keep these errors small and, more importantly, to know what they are.

in a orifice system, i see a lot of unknown error sources. such as orifice plate deflection, vena contracta movement, static pressure measurement locations, and coefficient of discharge changes to name a few. I am not convinced that the Cd of a plate is always constant.

from my standpoint, the pitot system has fewer parts and the errors are known. i think both systems can be made to be read the same results, but i am more comfortable with the pitot style system. The above sources of error could be neglible, but i haven't seen any evidence either way. So for me, a pitot system is less risk.

back to the topic, I have heard of averaging pitot tubes, but never used them myself. If i remember right, they still need some sort of fudge/correction factor in there data processing. i am not a big fan of these correction factors, because it is difficult to determine what makes them changeand by how much they change. i was under the impression that they only read total pressure, so what do you do for static pressure? or do you just measure it at the edge of the duct? does anyoe know how accurate these averaging pitots are?
Thanks
-Josh

what happens if the standard orifice is damaged, or bent, or the edge is no longer sharp, the standard has changed, and you have a problem.

[Tony]

I

in a orifice system, i see a lot of unknown error sources. such as orifice plate deflection, vena contracta movement, static pressure measurement locations, and coefficient of discharge changes to name a few. I am not convinced that the Cd of a plate is always constant.

True, but if the entering air upstream of an orifice is made reasonably stable due to having a suitable settling chamber design, and with a reasonably high pressure differential maintained across the orifice, the results can be shown to be amazingly stable and repeatable.

Anyhow Josh, a great many people here on the forum have switched over from originally using a pitot bench to using an orifice bench.

Nobody that I am aware of, has ever gone the other way from an orifice bench to a pitot bench.
There are practical as well as theoretical reasons why that is so.

[Brucepts]

I for one have made the switch from pitot to orifice style

I am so happy I made the switch that I no longer offer pitot flow elements for sale anymore! Nor can I even understand why anyone would go that route in my opinion?

A pitot bench needs to many flow elements to cover a range of CFM compared to an orifice style flowbench.

Tony has already pointed out quite a few more differences and positives. The design of an orifice bench is also simpler, my PTS orifice style flowbench has one moving part (slider plate for direction change) and it has no effect on the bench accuracy as it's after the testing orifice plate.

I have sold in the past more than a couple of flow elements and averaging pitot tubes and feel bad about it now knowing what I do now using my orifice style flowbench. I have helped a few of those customers over to an orifice style flowbench out of "guilt" and so I can sleep better at night

I do welcome the discussion though . . . and welcome to the forum

[roepke44]

OK, so how do you calibrate the calibrated plates? maybe use another calibrated plate? at some point i would assume you use some sort of other method of measuring the mass flow rate than an orifice plate. there are many. LFE, vane anemometers, paddle wheel anemometers, vortex shedding meters, ones that use sound waves, etc, etc. all are valid if used correctly.

while we are on the subject, has anyone ever tried to quantify the error in their measurements? It would be an interesting comparison between people with different types of benches.

I don't believe any measurement unless it is accumpanied by at least an estimate of the error in that measurement. the fact that a portjob increased my airflow by 10 cfm across the board doesn't mean anything. the equipment could be measuring +/- 10 cfm or more. If that's the case, that means that the shop really has no idea if it helped or hurt.

the only method for truely knowing the error in a system is to measure the system with a method that is at least an order of magniude more accurate than the original system.

I don't want anyone to get the impression that i am against orifice plates. i think they are great. for me, i just think that i could build a small flowbench with a pitot static tube, some pipe, some wood and some tubing. and my bench would be just fine. what i'm asking is how is the velocity measurement normally done to account for flow none-uniformity.

-Josh

[Tony]

OK, so how do you calibrate the calibrated plates?

-Josh

At the most basic level, anyone can just bore a hole into a metal plate themselves, and call that their flow standard to which all flow measurements are then referenced.

You can then begin making very fine and consistent comparative measurements with the greatest of ease, and use that to do your own airflow development work at home, quite independent of anyone else.
The beauty of this is it works !!!!
And it is a very sensitive and repeatable test method.

The question then arises, how does "Joe's" 300 CFM orifice flow standard compare to "Fred's 300 CFM orifice flow standard ?

Long ago, Bruce made some test orifices in various sizes and they have traveled all over the Nation, being measured on a large number of home built and high dollar commercial air flow benches.
The conclusion being that the home built flow benches performed as well for accuracy and consistency as the very best commercial benches.
With care you can build something at home that can read within a very few percent of what might be in the middle of the error spread for the very best equipment available.

Bruce now has fabricating these reference orifice plates down to a fine art, and can supply anything you need at minimal cost.

All this is the result of a lot of hard work over many years by dedicated people right here on this Forum.
The Forum orifice bench plans (available from Bruce) are the result of many years of testing and development and the latest combined thinking of many Forum members.