PTS Flowbench Forum

I thought I would do a followup to the orifice conversion for my pitot bench. As stated before, I was pleased with the surprising accuracy, and of course, the repeatability. I needed a higher capacity and I received a 550 cfm @ 16" orifice from Bruce as the measuring orifice. As it turns out, I could use a little larger still, but it is very close. Prior to testing the new orifice plate, I went through the accuracy of the inclined manometer and verified it against two different verticle well type manometers as well as a u-tube.

When I ran the sequence of orifice plates, the original four used in the first mockup tracked very close, but, indicated 5%-6% low. The 2.68" plate rated @ 515cfm, which in the first test was the measuring orifice, indicated 510 cfm, or within 1%. I couldn't come up with a logical test adjustment to get a consistant result. I thought perhaps I had reached a bad combination of orifice size vs pipe diameter. I did not use a baffle in the airstream in the original construction so taking a cue from the forum posts I decided to try adding one. It is 7" in diameter, located 3" from the air entry, which is approx. 10" from the test orifice. To my surprise, and relief, all five test plates now line up almost perfectly, even down to the smallest @ 160 cfm. I did not make any adjustments to the formulas used for orifice diameter or the orifice in a pipe calculator. As used in a pipe, the 550 cfm orifice comes out as 542.4cfm, which was used as the base flow @ 16". I used the inclined scale calculator, without adjustment, and found the numbers come out almost perfectly, over and over.

Originally I speculated that an orifice system may be more restrictive than a pitot, but I now doubt that is the case. For reference, with all motors running, I get 33.5"wc at the 2.68" orifice (564 cfm) and 17.4" across the 3.185" orifice in the pipe, or 565.6cfm. Pretty good agreement and capacity from the six original cheap Surplus Center motors. I read through the entire five page orifice flow in a pipe thread from last spring with some humor. Speculation about the accuracy and imagined problems with the PTS bench design, were filed away where they belong. It needs no defending. The easiest thing to do, is test your ideas and speculation. What I have put together works better than I could have hoped, and if everything written were to be believed, it probably wouldn't have happened at all.

The baffle plate is a very important feature, as it shields the orifice from being blasted with direct axial airflow, which is an unnatural act as far as the orifice is concerned.
Design and dimensions of the settling chamber and baffle are not extremely critical, as long as the basic concepts are understood and followed.

If you can get a calibrated test orifice and a calibrated measurement orifice to agree up around full maximum airflow, the smaller sizes should all readily fall into line.
It is up around max airflow that velocities and turbulence within the bench are at their highest and any problems or limitations with the bench design will begin to show up.

It can be an extremely rewarding feeling to have all the numbers agree with expectations, and repeat again and again.
You now have a stable and repeatable flow bench that you have confidence in.
Well done !

Thanks. I think a one cfm variance from calculated and/or tested values from 515 down to 160 is way better than I expected. In everything I have seen regarding an orifice in a pipe doesn't mention a baffle, but the expectation is a long straight run of xx pipe diameters. The funny thing is, a round pipe to a round orifice does seem to make a bit of sense. Not to dredge up the old problem finding with a square box )

Orifice in a pipe is very common for liquid flow measurement especially in the petrochemical industry, simply because it is cheap and very convenient to just bolt something between two pipe flanges.

But for serious accuracy, orifice in a pipe is a crime against nature.

Fire up your bench and try lowering a long piece of wool or string into the test hole.
It will follow the direction of airflow.
You will quickly see that most of the air rushes towards the test hole radially, from across the flat surface of your bench top. It then falls over the sharp edge of the hole.
Very little air rushes down vertically straight from above the test hole.

The measurement orifice flows the same way from what "should" be a large stagnant air volume contained within the settling chamber.
The baffle lets the orifice draw air in from around the baffle in a natural way.
Any fierce air jet aimed right at the measurement orifice upsets the natural flow process, and the flow coefficient.
Orifice in a pipe is the worst possible way to go about it.

I am with you on the air flow path. I have preached that to anyone that would listen. I also understand the desirability of a still air volume. What baffles me is, why does what I have work so well? Perhaps just lucky in this case? Years ago I was involved with an astronomy club, and there were many that ground their own mirrors and essentially made their stuff from the ground up, but sometimes were so involved in the details they forgot to look at the sky. This is meant in jest but a true story nonetheless. I truly appreciate the advice and input.

The air is definitely the Boss.....

Minimising errors, and letting the air just do what it wants to naturally do is probably the key.

The other thing is keeping the measurement orifice pressure drop high enough so that the effects of any slight up stream turbulence or instability have an absolutely minimal effect on the flow coefficient.

The bench works very well now and couldn't be happier. Even with standard water manometers it is so much faster. I am now testing my first head and now have another question. I have no issue with intake testing, but I "think" I saw a reference about a possible coefficient difference when blow testing. I tried searching but came up empty. Since my vacuum motor box is separate, It is a simple matter to turn it around for exhaust tests, as well as reversing the flow path. There is a substantial temperature rise over time with vacuum motors, so, is there a correction required where the temperature is variable and the test piece is at the terminal end of the setup? I toyed with the idea of adapting the exhaust port to the intake of the bench and attaching the cylinder adapter to the chamber, so the flow path is correct but that has plenty of issues as well.

Provided the air passing through both the measurement orifice and the test piece are at the same temperatures, no correction will be required.

As the blower discharge temperature inevitably slowly changes in blow mode, the exact same air still flows through both, and so the relative ratiometric pressure drop across both still remains exactly the same.

The blower can be located up stream (blowing) or down stream (sucking) it does not matter, as long as the same air at the same temperature flows through both the test piece and measurement orifice.

The test pressure will sometimes drift slightly as the motors heat up and the mains supply voltage varies, but the accuracy of the flow measurement will not if you can hold the test pressure constant.

If you fitted the Forum automatic test pressure controller to your bench, even that small irritation goes away completely, and measurements become even faster and easier !

Tony, I appreciate the quick reply. That will make life much easier. I know of and appreciate all of the technology available but will have to wait for now. What I have now is certainly more time consuming, but I have more of that to invest right now.