Atmospheric conditions - Do they affect flow on my orifice bench?

[Niklas]

[color=#000000]Will my flow reading need to be corrected in some way for barometer pressure, humidity and temp?

Lets assume that I have very accurate digital manometers (for barometer, orifice delta, test pressure, pitot) coupled to a modern daq module that are precision calibrated to give exact pressure readings in [force/area] units such as [Pa] (N/m^2) in my software.

Say I did a test at 29" Hg (~98205 Pa), humidity was 29% RH and temp was 25

[Thomas Vaught]

According to Super Flow (Harold Betties) the orifice plate type flow benchs like the SF300, SF600, and SF1020, being "ratiometric" benchs will not need corrections for Barometer, Temp, and Humidity. The SF110 and 120 benches will need corrections for temperature as they are "motor centered" benches vs "orifice centered" benches.

The "orifice centered" benches "see" the same Humidity, Barometric pressure, and Temperature at both the Test Part and the Orifice so no corrections required.

If you had a Laminar bench you MUST do corrections for the Humidity, Temp, B.P. as the device works differently from an orifice.

A last point. All of the info which states that the perfect orifice cd will be .62 is based on water testing which uses incompressible fluids calculations. Air is compressible (even though only a small amount of compressibility occurs). This will lower the CD from the "perfect" .62 value.

[Niklas]

Thanks Thomas,
I'm building a orifice bench with 7 motors, upgradeable to 9, it will be a unidirectional flow type bench with one blow and one suck-hole on the top. I'm building my own data acquisition with motorola MPX sensors and the Labjack U12 usb daq, I'm making my own software for this in DAQFactory. I also included a weather station in the build since I need to know humidity, temp and pressure of the ambient air to be able to convert indicated water column pressure from the u-tube manometer used for sensor calibration to "real" pressure in [Pa].

Weather station:
MPXHZ6115AC6U motorola mpx 15-115kPa pressure sensor
EI-1050 temp&humidity digital chip
Flowbench:
MPXV5004DP 15.74" H2O, only 12" used for 100% orifice dp together with the +/-4V input of the U12 to get best possible resolution with only 12 bits A/D. Would have been better with a lower pressure sensor but the only suitable mpx is the 7002 which is a positive/negative pressure differential sensor with +/-8" H2O range so that one wouldnt help accuracy...
MPXV5010DP 40" H2O for test pressure
MPXV5010DP 40" H2O for pitot pressure
EI-1050 temp&humidity digital chip for correction since the bench is a "motor center" type bench when run in exhaust mode. The humidity part here has no use but as I already used all 8 analog inputs (=4 differential channels) of the U12 I decided to use a EI-1050 here also. It may detect if I spill a beer in the bench though.

[DaveMcLain]

The orfice plate benches are supposed to be ratiometric because the same atmospheric conditions that effect the test piece also effect the measurment orfice plate in the same way.

The SF110 and 120 or any "blower center" flowbench require a correction for the temperture difference from one side of the motor to the other because in that case the temperture of the test side is different than the measurment side because when the motor pumps air it also heats it. The correction factor is easy to apply to correct the results, no big deal. Computer software like the stuff from Performance Trends will apply this correction factor automatically if you configure it for a motor centered bench configuration.

[Thomas Vaught]

Are you planning on drawing through the "Suck" hole when you do your "Blow" testing?

You might find you are restricting the flow with the hole.

I would recomment that you follow a design like 86 rocco and I have where you open a port hole
(like a 6" one) when you are doing the "blow" testing. Same deal for a second port hole when
you are doing "suck" testing.

The motors like to see a nice opening on either type of testing.

The SF 300 and SF600 benches have a special rotary valve that does the same job but many
hear have used extra port holes to do the same thing.

Tom V.

[Niklas]

[color=#000000]OK, the design allows additional portholes to be added later without problems, both the "blow" and the "suck" hole will be 5"diameter with a 1/2" blend radius on the inside of the bench, I have a "stub stack" (bass-reflex tubing with a large radius bell) that i plan on using on the "suck" hole when blowtesting and a similar thing on the "blow" hole for "suck" testing, only it has a 90

[Tom]

[color=#000000]Hi folks,

I would like to bring this subject back up.

Perhaps I dont get it,
BUT if you want to get the speed of the air with a Pitot style, orifice style or venturi style you
actually need the density of the air that is moving or not ? If not why ?
With the speed you calculate the flow cfm, m

[86rocco]

[Tom]

[86rocco]

[color=#000000]The "standard" air density is 0.0764 lb/cu ft, this is the density of dry air at 15

[Tony]

One easy way to prove the "theory" is to set up two identical orifice plates in series flow, and measure the two pressure differentials. They will be the same, even though the first orifice works in higher density air than the second. Swap the orifice plates around and repeat the test. That should prove to you that the ratiometric method DOES work.

The only real way to screw it up is to add (or remove) significant amounts of external heat to the air between the two orifice plates, as will occur in a blower centered flowbench.

That is the single main attraction of an orifice style bench, ambient air changes make absolutely no difference to the results. Both restrictions in the flow path will have exactly proportional pressure drops.

As soon as you begin adding corrections and fiddle actors to any type of measurement, you can start getting into real trouble fairly fast.

[Mouse]

Can someone show me why a Pitot tube would not work in the same way as an orifice in this situation.

John

[Tony]

With an orifice bench you are measuring and comparing the pressure drops across two flow restrictions arranged in series. It is a very simple ratiometric measurement, with no complications.

With a pitot measurement, you are taking an indirect measurement. Measuring one thing and using correction factors to convert that measurement to infer something quite different.

I suppose an analogy would be using astronomic tables and an accurate clock to calculate the height of a building from its shadow on sloping ground, compared to just lowering a measurement tape from the roof to the ground. Both systems work, one is simple the other is far more complicated.

An experienced and mathematically competent surveyor or navigator might wonder at the clumsy stupidity of some fool hanging off a ladder.

The fool on the ladder that can barely count, knows for sure the measurement is correct.

[Tom]

[color=#000000]Hello,

I am 100% with you and also agree 100%.
May be I can not bring my point right up to you guys.

Please try to follow me with following.
Lets take the other thread as an example (passing around the orifice made by Bruce).

- Bruce made a orifice diameter x inch
- He flowed it on his bench 250 cfm (atmosphere condition #1)
This measuring is, lets say 100% accurate and so is our reference
- He passes it around the US to the next one who will flow it on his bench
with conditions #2 (different to #1). If he has no information about the conditions #1
from Bruce he will have a different flow number than 250 cfm or am I here wrong ?
If so, why no difference ?
May be 86rocco is correct and the changes within are small, but if you will have weather brake down you will have a change in all your numbers.

IF the air conditions are that small why do we have the mass air sensor ?
From flow to mass is only the density between.

If it is unimportant what kind of conditions you have for the reference why do we have a

[bruce]

Humm we could do a more scientific approach to the pass it around plates by machining up another set with specific rules of testing to prove or disprove these questions?

The plates are a "blind test" its being done more or less to see how close everyone is not as a calibration of your own personal flowbench (which is not implied with any previous post, I'm just stating it) Nobody but myself knows what the plates are machined to so when you flow them you have no idea what they are suppose to flow or have flowed for others.

You do raise some questions in my mind now which I will try to do some testing on my bench in the very near future. I flow the plates on a pitot style flowbench that is corrected for shop conditions.

The reference point is the size of the hole and calculated flow with a .62 cd at standard conditions. If you setup your initial flowbench calibration at anything different than those conditions then that should be taken into consideration I would think? Least that is what I have always done.