Calibration pressure - 28" or much higher?

[SWR]

[SWR]

[bruce]

[quote="SWR"][color=#000000]I thought an orifice with "square edged" hole was a sharp edge design if the corners between the flat side and the hole was exactly 90

[SWR]

[bruce]

Here is a pic off the net showing the backside bevel on a plate. I have never compared a square edge plate to a sharp edge plate. I have always just machined the sharp edge plate in alum. (might have to try compare the two)

The dia to thickness ratio on a square edge plate might make a change in the Cd? With the sharp edge plate you have a constant since the edge is dead sharp and the thickness of the plate does not come into play.

[Stef]

The Cd of the square edge orifice will be similar to that of a sharp edged one. However they should perform the same way i.e. as the flow (and Reynolds number) increases through the orifice the Cd value will drop. Once past a certain Reynolds number the Cd value will stay steady and be independent of flow.
If your plastic orifice is deflecting under high differential pressure (across the orifice) then the Cd value will increase, skewing your flow numbers. It doesn't take much deflection (less than 1/8") to increase the uncertaincy of the measurement by 1-2%.
Looking at ISO 5167 they recommend a minimum of 1/8" thick plate in stainless steel for pipes your size and single direction flow. For bi-directional flow the thickness goes down to 0.079" in stainless steel (becomes a square edged orifice) and the maximum pressure differential across the orifice is reduced to limit deflection.
I would suggest replacing the plastic orifice with a steel one to eliminate the chance of deflection then retest with the calibration orifices.
If there are still problems after that you can look at other potential areas of error, like uneven velocity profile in the pipe (due to swirl etc.) and eliminate those one by one, checking the calibration between each change to systematically find the cause of error.

[Thomas Vaught]

Any feel for when the CD would stabilize as far as Reynolds
number goes?

I have seen the CD go down as you mentioed with increased flow on a square edged orifice.

Tom V.

[SWR]

[Thomas Vaught]

I would use the spreadsheet for the orifice calcs.

It is readily available on the forum and allows more accurate
calculations vs a "set" constant that is only correct in one situation.

JMO

Tom V.

[Stef]

[Stef]

[color=#000000][quote="SWR"]It was deflecting quite a bit more than I thought. So I am changing them all to aluminium or steel.

But,I am uncertain to what I should do. My orifices sit in short sections of 5" pipe attached with V-band,nitrile-o-ringed speed-clamps,and will fit both ways. Turning the orifice around or changing to a different one takes no more than about 5 seconds with this setup. So the question is,with test accuracy in mind: Should I rather make the orifices sharp edged,1/4" alu for stiffness,and just turn them 180

[Tony]

SWR,

On factor you seem to be completely ignoring here is that the Cd of an orifice when located in a "relatively" small diameter pipe will not be anything like the Cd of the same orifice when mounted flush on an infinite flat surface facing completely undisturbed upstream air.

It may come close if the pipe were a couple of feet or more in diameter at low flow velocities. But flow along that five inch pipe will be heading straight towards that orifice, and will have considerable axial velocity, and also an unknowable and incalculable degree of turbulence.

The flow entering an identical orifice facing an undisturbed upstream air volume will mostly be radial and fall inwards towards the orifice from around the edges.

The flow conditions are going to be as different as night and day.

The whole point here is that the flow conditions upstream a flat flush mounted orifice are easily defined, known and repeatable. A sequence of stepped orifice sizes will always behave predictably. A two inch orifice will always flow EXACTLY one quarter as much as a four inch orifice.

But I can tell you for sure, a five and a half inch orifice in a six inch pipe is not going to have anything like the same flow characteristics or Cd as a five and a half inch flat, flush mounted orifice plate facing still air.

[SWR]

[bruce]

One question I would pose is why would you want to put the orifice in a pipe? Pipes are more inclined to the pitot style of bench.

There have been many orifice style benches built by members of the forum using the plenum design. The design is simple and works. Why make your project harder and more complicated than it needs to be to obtain the desired repeatability and accuracy level one needs? Just looking at that ISO standard makes me go hummmmm . . .

It is really not that hard to make a repeatable and accurate flowbench from the info on this forum. All one has to do is look at the "pass-it-around plate" thread to see how accurate the home-built bench can be compared to the pro built benches that are out there.

[Tony]

I totally agree with Bruce on this.

Pipes are for pitot (or venturi) measurment, orifice plates require the largest practical plenum or settling chamber upstream. The upstream settling volume simply cannot be made too large.

The sharp orifice in a pipe flow measurement method is very widely used in process engineering simply because it is cheap and extremely practical where there is a vast maze of tightly packed pipework in an industrial plant.

It is never intended to be used for precision laboratory grade flow measurement. It is a bit like the empty/full gauge on a fuel tank. You don't need to know the precise quantity of fuel remaining, just a rough indication to avoid sudden nasty surprises.

Likewise many orifice in a pipe process flow meters used in industry just indicate a high/low flow range without any pretense of precision. The method is just not suitable for an accurate airflow bench, and you will NEVER see an orifice in a pipe used in any commercial airflow bench.