Vacuum motors in parallel

[wesman07]

Hello all,

My question is about vacuum motors in parallel. I am using two of the same Ametek motors to power my home built flow bench that I am trying to use to measure airflow through a cylinder head. These motors come with a performance chart that I thought I could use to measure air flow in cfm.

The research that I have done lead me to believe that two motors in parallel will double the air flow for a given vacuum depression. My calibration efforts are indicating that this is false. I would like to understand why as well as how to estimate airflow if possible.

What I know is that the depression readings at zero to 1/2” orifice are accurate. A 2” orifice gives me a depression reading around 10” w.c. Instead of 4.4 as indicated on the performance chart.

Any help is appreciated. Please see photo attached.

Thank you,
Wes

 

[LittleInch]

You've got less flow.

Where is the orifice?

Can you sketch this system please so we can see what you can see in front of you

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[1503-44]

Yes please post a picture.

I think that your cylinder head has too much resistance to allow the higher flow of the 2"orifice to pass through it, but it's no problem for the smaller capacities of the smaller orifices to get through. You are starving suction flow with high flow resistance of the cylinder head, which increases the vacuum to 10"wc. It will allow 98 x 2 = 196 to pass, but not 224.

What reading do you get when the cylinder head is not attached?

A black swan to a turkey is a white swan to the butcher.

 

[wesman07]

The calibration tests are being performed through a piece of lexan that has seven holes in it ranging from 0.25” - 2”. The test bore or largest orifice is 4” which will produce .84 w.c. I can confirm that there are no leaks in the system.

Is there a “system curve” that I need to calculate?

Please see photo attached of the bench without the back and test bore installed.

 

[wesman07]

If I run only one pump and block off the other, the manometer reading matches the depression on the graph.

 

[1503-44]

The box and your inlets comprise "the system".
Friction of the air flow through the inlets and the box is proportional to the flow squared. Double the flow and Friction is 4x the previous value. Square edge openings have friction losses too. The coefficients are around 0.6, then x v^2/2g to get the friction loss. And you have inlets at the holes into the box and another into the pump inlet itself to account for. I suspect most of the friction losses are occurring at the inlets, as the box looks relatively large. The cut sheet probably accounts for the pump inlet itself, as maybe they think a pipe of equal diameter connects there, not sure.

I think if you double the area of the holes in the box, or rather the lexan, maybe you'll get pressures closer to 2x the cut sheet values. I'm too lazy to do the math. Can you try that, just with the 1.5", and please let me know what happened? Or this is easier. Block all the holes but one. Get the pressure value and then see if it corresponds at least roughly to the pressure for a hole with half the area. If it does, problem solved.

A black swan to a turkey is a white swan to the butcher.

 

[wesman07]

I’ll plot the depression for the seven test orifices on the chart. I assume you mean double the area of the orifice, and not double the diameter?

 

[1503-44]

Right. The Area.

A black swan to a turkey is a white swan to the butcher.

 

[wesman07]

You might be on to something here. Two of the test orifices I used are almost exactly have the area of other plotted points.

 

[wesman07]

So if the vacuum correlates to half the orifice area given on the chart, where would that put the CFM?

 

[1503-44]

I was thinking about that. As the lake rose, I didn't notice the alligators. Rereading your first post, you say you want to calibrate air flow from the pump's data sheet. That's a problem, because as we know the air flow will be the result of the combination of the pump and the duct system's characteristics, not just what the pump's data sheet says. We really need an independent means of measuring the flow, some kind of flow meter.

Without a mass flow metering device per say, one way to measure flow is with a manometer, but that requires an orifice plate installed in a pipe, where both have well defined diameters for which we could calculate the flow rate through the orifice from well known relationships of flow vs pressure drop across the orifice plate. Or maybe we could estimate it by other means, such as measuring the velocity of smoke moving through the duct and trying to determine the average velocity.

I'll think some more about that.
Anybody else have some ideas?

A black swan to a turkey is a white swan to the butcher.

 

[1503-44]

I was thinking about that. As the lake rose, I didn't notice the alligators. Rereading your first post, you say you want to calibrate air flow from the pump's data sheet. That's a problem, because as we know the air flow will be the result of the combination of the pump and the duct system's characteristics, not just what the pump's data sheet says. We really need an independent means of measuring the flow, some kind of flow meter.

Without a mass flow metering device per say, one way to measure flow is with a manometer, but that requires an orifice plate installed in a pipe, where both have well defined diameters for which we could calculate the flow rate through the orifice from well known relationships of flow vs pressure drop across the orifice plate. Or maybe we could estimate it by other means, such as measuring the velocity of smoke moving through the duct and trying to determine the average velocity.

I'll think some more about that. First, I'll need some more coffee.
Anybody else have some ideas?

A black swan to a turkey is a white swan to the butcher.

 

[1503-44]

It seems that the charts are accurate with one pump.
I'll try calculating a Cv coefficient (relates flow to pressure drop) using that pump data and we'll try to see if we can make some inference to the flow with two pumps based on that Cv.

A black swan to a turkey is a white swan to the butcher.

 

[wesman07]

Thank you for taking the time to confirm that there is a issue with the system. This was a design I came up with but I am not capable of proving or disproving it mathematically.

What I did do was test the flow range on a cylinder head that I knew where the flow rates should fall. I believe it’s reading 10-20 cfm lower than it should at 20” depression and fall further off the chart as depression decreases… but that’s hard to confirm.

There may be a way to correct it mathematically, and if there is I would be more than happy to try it. The two other options I can think of is using a Pitot tube to measure velocity, or some sort of hot wire mass air flow sensor.

 

[wesman07]

I found a calculator that figures cfm from orifice size and vacuum depression. It turns out vacuum and cfm times two is accurate. Vacuum per orifice size is not, which is fine. Thanks for the help.