New here and to flow benches

[Wesman07]

Hello my name is Wes. I’m new here as well to flow benches in general. I built a cost effective pressure drop bench to test if an experimental concept was viable. Basically I am using two of the same ametek motors, reading vacuum with a digital manometer and using the performance chart to figure out flow. It is very accurate at low flow but falls in the upper range. I believe the issue is turbulence. Does anyone have a suggestion to try. Should I shield the pick up tube or a flow straightener and if so, what is the best location?

[Brucepts]

Welcome Wes,

My first thought is motor performance chart figures do not add up CFM, they are based on one motor, so using 2 motors would not be double the CFM at the given rating.

1+1 does not equal 2 when it comes to vacuum motors there is a decreasing efficiency as more motors are added I have found.

[Wesman07]

Bruce,

Thank you for the response. I should clarify the inaccuracies. At zero flow the motors should pull 63in/h20 and it pulls about 61. With a 2” orfice it’s should pull 4.4in/h20 but it get a reading near 10. The test offices are not perfect, but they are not that inaccurate.

After reading around here I understand this is not the accepted flow bench design, thank you for helping anyway. This project took off as a quick and dirty way to test if an experimental concept was worth pursuing.

I have found that if I bring the signal tube through a small aerosol can cap I can get the reading significantly more accurate through the entire range. The height relation between cap and tube is very sensitive. I have no idea why this works. I thought the issue was turbulence, but I’m starting to dismiss that idea.

Going back to your theory… if the motors are not adding up to exactly double the rating, wouldn’t It be reading a lower vacuum depression?

[Brucepts]

Bruce,


After reading around here I understand this is not the accepted flow bench design, thank you for helping anyway. This project took off as a quick and dirty way to test if an experimental concept was worth pursuing.

Going back to your theory… if the motors are not adding up to exactly double the rating, wouldn’t It be reading a lower vacuum depression?

I'm not sure there is any "accepted" flowbench design hence the forum we have here to discuss flowbenches

Looking at your design does present some thought on my end about my theory based on your bench design, looking at it you have the blowers in 2 different plenums so they are not working against each other as would be in a single plenum design so my theory might just be that a theory based on single plenum design flowbenches . . . something for me to ponder

[Tony]

What seems to happen if you have several motors operating very close together in a cluster, they seem to try to rob air from each other around the air intakes. There is no fixed allowance for this, it depends on the physical layout and air turbulence, but as more motors in close proximity are added, the contribution of each usually becomes slightly less.
As has been pointed out, your design does not suffer that problem.
What you will find though, is that actual airflow can change a lot with the ac supply voltage and motor temperature. As the motor winding temperature increases, the electrical resistance of the copper increases, reducing power and motor rpm.
If you take measurements at different times of the day and different days the readings will never duplicate exactly, and that is a problem.

You can never be quite sure if that 4% flow increase or decrease is from your most recent port modification, or due to random changes in your flow bench.

[1960FL]

Welcome Wes, I think I might take a step backwards in learning and solving your problem, start by blocking off one motor, then cut your test orifices out of thinner material so the CD is closer to a standard. Run your test against the motor chart and see how you compare, you may also want to build a simple plastic tube pickle jar manometer to test against your electronic one. Get a baseline on your setup, then test the other motor, Does the data jive? if so then your airflow path and or your motors fighting against each other could be contributing to your problem.

Rick

[Wesman07]

With one motor running and one motor blocked off the the depression per orifice size is on scale. With both motors running the depression per orifice scale starts to run high starting after zero and is significantly higher at a 2” orifice. After spending to time with the charted data I was able to determine that two motors in parallel does not necessarily double the cfm per orifice diameter, it doubles the AREA of an orifice for a given depression. So 4.4” w.c. falls inline with a 2.82” test orifice rather than the 2.00” indicated on the chart.

I haven’t quite put this all together to see if any of this is remotely useable. I’m in the process of installing a pitot tube for a third party testing but given the size constraints that won’t be very accurate either. I’m also looking into making a new bench. Im just trying to understand the general design and Taylor it to my needs and budget.

[Wesman07]

Quick update. Yesterday I drilled eight holes in a piece of plexiglass and measured the vacuum for each orifice. I then used the orifice calculator I found on this forum. Using excel I created a scatter plot and overlaid the two curves. There was a 10cfm discrepancy at two middle points. The other six points were within 2cfm.

Now my test orifices were made with wood bits so they themselves have a degree of inaccuracy. However, I think it can be concluded that this $300 bench is accurate enough to guide a beginner in the correct direction and deliver ball park cfm numbers.

[Brucepts]

Now my test orifices were made with wood bits so they themselves have a degree of inaccuracy. However, I think it can be concluded that this $300 bench is accurate enough to guide a beginner in the correct direction and deliver ball park cfm numbers.

Repeatability is what you are after, then you know your numbers are giving you results you can use for comparing from one test to another. This is more of a concern over accurate CFM numbers, you are headed in the right direction with your understanding