PTS Flowbench Forum

Grey,

Do not in the least worry about the edge of the disk a .125 or 3MM aluminum disk with precision bored orifices will not let you down. Some of us here use a square edge plate in our benches and they calibrate spot on.

The SF600 issue is the angle of the plate and the approach of the air from the test fixture along with orifice holes that are not machine cut and are out of round.

Rick

1960FL wrote:Grey,

Do not in the least worry about the edge of the disk a .125 or 3MM aluminum disk with precision bored orifices will not let you down. Some of us here use a square edge plate in our benches and they calibrate spot on.

The SF600 issue is the angle of the plate and the approach of the air from the test fixture along with orifice holes that are not machine cut and are out of round.

Rick

Lucky I didn't go with iteration #13 where I had everything inside on a 45º angle.

I read somewhere on here that someone did all the sharp edge Vs square edge tests, but unfortunately didn't post any technical specifics, so it looks like I will have to plug away and finish the rest of the bench and test it all before final sealing.

Grey i will assure you the only differance between a Sharp edge and a square edge if made properly is the CD value used in calibration and your calibration plates should be sharp edge From Bruce so you are working against a known standard.

Rick

Brucepts wrote:What does FP recommend for their DM for a range of plates? I would venture to say they have no clue

You were right, they don't know.

Now I have to take the FP-1 apart.

Motorola MPXV5010DP is the sensor they use which won't help you that much as all the programming is on a chip.

Brucepts wrote:Motorola MPXV5010DP is the sensor they use which won't help you that much as all the programming is on a chip.

At least the sensor looks pretty linear until 10kpa, with error increasing with pressure diff. Temp pretty stable from 0º to 90ºC

http://www.datasheetcatalog.com/datashe ... 10DP.shtml

The internals of the FP1 are pretty underwhelming, just breadboard, no soldered printed circuit board - that's a corrosion problem. where's the desiccant -.

So the sensor is most accurate at low pressure Δ, the orifice is more accurate a high pressure Δ, and the DC affects the reading exponentially with increasing pressure Δ.

It seams a square edge internal orifice would be more stable.

Based on this :-

It looks like orifice plates get more accurate with pressure Δ, eg, the 3.490" has less than 1% error from 200cfm onwards but before that it has substantial error.

Is there a table that shows accuracy over pressure Δ for sharp edge and square edge orifice plates. I was wondering where the sweet spot is for orifice plates.

Also, I'm assuming that the CD of the plate is comparing the plate flow against the 147cfm@28" per square inch number, so do sharp edge orifice plates have to have different bevel angles on the back for different sized orifices so that the CD is the same for all the plates.

If you are using a water manometer there seems to be a "sweet spot" I personally always flow no less than 50% of what the orifice is capable. So on a 400 cfm orifice I would not flow less than 200 cfm with it. Some people have said 40% or more on an orifice.

Now a caveat here, with a good digital manometer you seem to be able to get away with just one orifice. I have done this comparison back to back. To stay accurate with my water manometer I have to keep changing the orifice in the bench. With my digital manometer I can flow the whole test all lifts with just the largest orifice I need.

John

jfholm wrote:Now a caveat here, with a good digital manometer you seem to be able to get away with just one orifice. I have done this comparison back to back. To stay accurate with my water manometer I have to keep changing the orifice in the bench. With my digital manometer I can flow the whole test all lifts with just the largest orifice I need.

John

Unfortunately that's my problem, I don't have a good DM, I have a FP1 .
Bruce recommended multiple orifices, so I'm trying to work around some limitations.
The FP1 is limited to 6 CD inputs(3 int and 3 exh), which I could get around in post on Excel.
The FP1 is more accurate at low pressure Δ
Orifices are more accurate at high pressure Δ

I need to extend the flow ranges to overcome the FP1 limitations, so I'll use the top 60% of an orifices flow.