PTS Flowbench Forum

It has been quite a while since I have spent appreciable time on this forum, so I have spent a fair amount of time reading, trying to catch up. A short explanation to preface a couple of questions.

I have had a simple pitot bench for many years, eventually collecting many flow elements to cover a wide range, plus a few orifice plates from Bruce. Even though simple, it seems reasonably powerful. I have tested the 2.68" plate from Bruce up to 29.75" w.c., or in the mid 500cfm range. The bench can be relatively accurate, but certainly not as stable as I would like. For example, testing the same part using a range of flow element diameters and finding significant differences in the flow reading. It certainly can work for comparative testing, but overall (reasonable) accuracy, not so much. This actually led to a question I asked 2 1/2 years ago regarding preferred flow velocity through a pitot flow element. Since there has never been a reply, I pretty much walked away and have considered an orifice based flow element, off and on, for a while now. I found some humor in Bruce's comment of feeling guilty for leading members down the wrong path. I believe all knowledge to be useful, even if it is to find out you were wrong. Yes, there may have been time and money misdirected, but not wasted. My guess is a pitot bench can have less friction loss internally, but there certainly must be some pressure drop though the flow tube or no movement would occur. I guess only time will tell if the loss through a flow tube and an orifice plate are comparable or not, but not really relevant to what I need to know:

Since my flowbench has a separate motor box, on wheels, the bench was connected to the motor box with the flow elements, both the box and flow elements easily reversible. I believe the pitot style elements can be reasonably replaced with a container for an orifice plate. In keeping with the redneck nature of my bench, I am thinking of using two five gallon plastic pails, end to end, with the plate in the center and an appropriate baffle on the inlet side. A couple of toilet flanges on the ends and I have a pregnant version of the pitot elements. Hopefully good enough for proof of concept. I have two inclined gauges on the wall, approx. 8.5" rise by 40", and a 17" rise by 44". I hope the two different scales will allow two separate flow ranges for any orifice plate. This of course begs the question, how low can the Delta p be, using water manometers and expect reasonable accuracy? If I use the 2.68", 515cfm @ 28", plate from Bruce, are the standard pressure drop formulas usable for this application, such as 389cfn @ 16" and 275cfm @ 8". The exact numbers aren't important, only the two different range question.

2seater wrote: This of course begs the question, how low can the Delta p be, using water manometers and expect reasonable accuracy?

There is no simple answer to that, the higher the pressure drop across the measurement orifice, the more accurate and repeatable the results are very likely to be.

The reason being, that at a very low delta p, the low velocity air entering and flowing through the orifice will be much more easily disturbed and influenced by any up stream turbulence, air buffeting, or any flow bias. The design of the settling chamber and it's size and configuration, and indeed the entire bench, becomes a much more critical factor in achieving dependable results.

A relatively high design orifice pressure drop is going to have a much higher flow velocity, much more violence and fury right at the measurement orifice, and any slight imperfection of flow up stream of the orifice plate will have a correspondingly much smaller effect on the final measured differential pressure.

Use the highest measurement orifice pressure drop that you possibly can afford to run, and you will definitely see better accuracy and repeatability all else being equal. It will greatly reduce the effects of any design imperfections in the rest of the bench.

That covers most of it, but a sloping manometer with a very small rise also becomes much more critical, especially levelling and setting the zero, as well as keeping the tube dead straight.
The closer to vertical the manometer, the less critical all these possible sources of error become.

Thank you for the input. I can certainly see the benefits of the orifice system but I have always imagined the power sapping caused by the second restriction. I could be completely wrong. I guess it would be super simple to just add a probe to the motor box to see what the full depression looks like. I guess the simple things get overlooked.

I shall set about making an orifice assembly up and tee both inclined and the velocity probe verticle gauges together to see what's what.

A dark side bench is still going to have significant pressure drop through the long length of measurement ducting required to achieve the required reasonably high velocity laminar (?) flow.

The one supreme advantage of a ratiometric orifice bench is that the readings require no correction.
What you see on your manometer IS the flow measurement.
You could even calibrate your water manometer scale directly in CFM if you wished to do so.

Change the orifice, change the manometer scale, and you have a direct reading water manometer bench.

It is certainly possible to design your orifice bench to have a low working pressure drop across the measurement orifice, but it definitely will make the orifice much more sensitive to turbulence and flow instabilities within the bench.

Personally I would much rather go the other way, accept a high pressure drop and reduced maximum flow bench capacity, but know for sure my bench was stable and repeatable.

I certainly agree there must be pressure drop in the pitot flowtube or nothing would move. I am questioning myself as to why I never actually checked it? I'm sure it could be calculated too, but I have pretty much determined I will change over to the orifice. I am sure some of the previous mystery readings I have seen are due to correction factors. I do like the idea of getting rid of the corrections.

I think I have come to terms with the possibilty I will lose total capacity, but a new air mover is still in the long range plan, so not a big deal. For reference, my bench is simply a piece of countertop with a second piece of 3/4" plywood laminated to the underside for stiffness, which is bolted to the wall, a small settling chamber underneath with a 4" elbow exiting the bottom, and a vacuum motor box on wheels, which is connected to the 4" elbow under the bench with the various flow tubes. The two connecting heights are designed to be straight shot, so little loss from direction changes. I know an optimal design for the orifice will be different, but a quick and simple test will do for now. It should prove interesting to recheck some previously cataloged items.

I have never used a pitot bench, so have no idea what typical pressure losses through the measurement system may be.

Adding corrections to each and every reading is not only slow and error prone, but there will always be some inevitable reading and rounding errors.

Those long pipes required in a dark side bench look fairly restrictive.
I know from flow testing exhaust systems and intercooler pipework the cumulative pressure drops can add up to become surprisingly high in terms of inches, but in psi they don't look all that bad.

I would not be too difficult to directly compare the pressure drop of a pitot flow tube to a comparative orifice.
The difference in the total required blower pressure differential, taking into account the unavoidable test pressure drop, may end up being quite small between the two.

Yes, I know what you mean about pressure drops in the items mentioned. As a matter of fact I have never tested a head, but lots of mass air flow, throttle bodies, mufflers, cat's. intercoolers, homemade turbo piping etc... Those small psi differences sure do add up to a tall column of water By the way, I liked the suggestion of a cfm scale. Something to think about.

2seater wrote: I liked the suggestion of a cfm scale. Something to think about.

The way I had once considered doing this (a very long time ago), was to have a whole series of identical sloping manometers screwed to a wall, one mounted directly above the other.

Each would have it's own direct reading CFM scale to suit a particular orifice measurement range, of 2:1 flow range
(from 25% to 100% of each manometer scale length).
Identical scales could be drawn up, but with different CFM numbers on each manometer.

The zero pressure point could conveniently be tweeked over a small range by simply sliding the CFM scale back and forth.

High end calibration of each CFM range on each manometer, could be corrected by adjusting the incline slope against one of Bruce's calibration plates.
The reason for using multiple manometers, instead of just swapping scales on one manometer, would be that the manometer slopes may need to be set slightly different for each measurement orifice to get the high end CFM calibration correct.

Hopefully as you switch between flow ranges the CFM readings should stay the same. It should be possible to get this exactly right by adjusting each manometer incline individually.

I had thought of fitting a single cam lobe, and a bunch of microswitches around the outside of my orifice selection turret.



As I selected each different measurement flow range on my turret, the air lines going to the manometers would be switched automatically via small 12 volt solenoid valves. The type of automotive solenoid valves commonly used in emissions vehicles.

I have never actually done any of this, but it all seems quite possible and practical to do, could be made to look quite impressive, and with a bit of scrounging should cost almost nothing to put together.

I can see that it would be possible, but a bit more than I have in mind at this time If reasonably priced digital instrumentation didn't exist, it might be considered. As always, suggestions are appreciated.

Yes, this was all from a very long time ago.
Years before Bruce, Rick, and I started working on the Forum digital manometer project.

The only reason I mentioned water manometers at all, is that for many people starting out with their very first flow bench, it is still the lowest cost way to get a bench up and running.