Tractorsport Flowbench Forum Archive

Posted: **Mon Oct 18, 2004 11:36 am**

I recieved some orifices from Bruce for calibrating my flow bench. The look great and will allow me to do good cal.

Question for all,
Where are you people locating the inclined manometer tap points?
I have mine about one inch on either side of the orifice plate.
My readings are not accurate. There are down by about 20+ CFM. I have no leaks.
Based on what I read last night form a perfromance trends article/manual both tap points should be "as close as practically possible to the orifice".
Keeping that in mind I'm going to try moving the taps right next to the orifice. Has anyone had similar findings?
or are people finding that 1" is close enough?

Some background: The orifice plat is located in a straight pipe, ie no plenum upstream. I don't intend on changing to a plenum unless a last resort.

Posted: **Mon Oct 18, 2004 6:20 pm**

There are some difficulties with the orifice in the pipe approach, the air is being forced towards the orifice, and it is going to be turbulent downstream, and probably upstream as well.

Turbulent air up stream, will give the same orifice a different flow coefficient than if still or laminar air enters the orifice. It could be either higher or lower.

Attempting to measure static pressure in a moving air-stream can only be done reasonably accurately with a static pitot tube.

Changing the ratio between pipe diameter and orifice diameter will upset things quite a bit, and so will moving the measurement points along the pipe. It certainly can be made to work, but there is no theoretical way you can work out the expected results beforehand.

All your orifice sizes will be strange and have no fixed relationship to each other or anything else.

On the other hand placing a thin orifice plate between two infinitely large plenum volumes will give accurate, repeatable, and (acceptably) predictable results. A hole twice the diameter will flow exactly four times as much, and so on.

I wish you luck with the orifice in the pipe approach.

.

Posted: **Mon Oct 18, 2004 8:57 pm**

Crazyman,

I have a lot of respect for anything that Tony posts here, as he's very sharp on this stuff. There are many approaches that can be used, so it's natural for different people to favor different approaches. And Tony has a lot of experience with what works and what doesn't. I'm not there on the experience level yet, but I do have some information that may be helpful to you.

The location of the pressure taps on each side of the orifice play a major role in the determination of flow from the orifice plates. Reason is that there are different sets of calculations used in determining the flow, based on the tap location.

The "flange taps" are the most common, and as you were instructed, are supposed to be as close as possible to the orifice plate. There are also taps that can be placed at D and D/2 distances from the orifice plate. You also have to be sure that your plate is installed in the proper direction. The bevel needs to be downstream for a unidirectional orifice, while a bidirectional orifice will not have any bevel to it. The bevels are machine into the orifice to reduce the amount of non-recoverable pressure losses from the orifice (one of the orifice's disadvantages is high pressure drop).

If the calibration data for your orifice plate/s were based on flange taps, then locating the taps in a different location will indeed give you different readings.

There are specific relationships between the diameter of the orifice and the inside diameter of the pipe that the orifice is placed inside of. This is called the Beta ratio which is merely pipe ID/orifice ID. For Beta ratios below 0.6, the uncertainty (related to accuracy at 95% confidence level) of the orifice is supposed to be +/- 0.6%, while Beta ratios above 0.6 have an uncertainty of the Beta ratio. For example, a 0.75 Beta ratio will have an uncertainty of +/- 0.75%.

One of the concerns with orifice plates is that the error will increase if you are trying to measure flow below their turn-down ratio, where the flow stops being turbulent. Good orifice design for flow in pipes is based on the flow entering the orifice being fully turbulent (Reynolds number > 10,000), so that there is an adequate vena-contracta produced downstream of the orifice.

Applicable codes for further reading for measuring flow with orifices within pipes are ASME MFC-3M, and ISO/DFIS 5167-2.

It's important to note that to properly measure flow within a pipe with an orifice, the flow has to be uniform, with no spiral or tumbling motion to the flow, and the flow must have the proper velocity profile. What this means is that if you make a sharp turn (or two is worse) in your piping leading up to or down stream from your pipe orifice, that you will be introducing error to your measurements.

This is where flow conditioners and flow straighteners come in. A flow conditioner works to restore the velocity profile, while a flow straightener (properly designed) will stop rotational flow within the pipe. Flow coming out of an elbow will have a distorted velocity profile. Flow coming out of two "out of plane" elbows will compound that velocity profile and create swirl within the pipe flow. A properly designed flow straightener will run for a length of at least two pipe ID's, and will start at least one pile ID downstream of the second elbow.

While this code stuff is great to keep you out of trouble, there's a certain brilliance in the approach that Tony has taken with the measurement of flow with the orifice in the plenums. Having large plenums will allow the flow to stabilize and become uniform before entering or exiting the orifice plate. In addition, having large plenums in relationship to the orifice plate diameter, results in a very low Beta ratio, which will improve the readings.

There are a lot of subtleties in measuring flow properly with orifice plates, that most often gets overlooked. The beauty of the Engineering codes for measurement of flow within pipes is that if properly constructed you can predict flow very well without the assistance of calibration. The math outlined in the codes is used to calculate the flow from the orifice flow in pipes, while the calibration is typically used only to verify the proper manufacture of the orifice meter. (The orifice meter is the pipe, orifice and associated straighteners as a system).

Keep in mind that when flow labs use orifice meters for benchmarking flow, they typically go to undisputed extremes of running with 80-100 diameters upstream of the orifices. You will read in many manuals and books that all you need is 10 diameters upstream and 5 diameters downstream, but that's not the entire story. The required number of diameters changes with the Beta ratio that you are using. But if you are using approximately 0.6 Beta ratio, then you can get by with 10 upstream and 5 downstream diameters, if and only if you utilize the proper flow straighteners or conditioners as required, depending on the piping of your project.

By the way, it's my personal opinion, that if the Superflow 600 bench, and possibly others, were properly designed, there flow numbers would not change when the test head is rotated on the top of the bench. I've seen posts listed here that indicates this happens.

All of the above technicalities can be eliminated if you use a Laminar Flow Element, which is not sensitive to upstream and downstream flow disturbances. But they are very, very expensive and have high pressure drop associated.

The application that Tony is using does require a calibration to get a baseline in flow. I believe his design would be more in line with what the vacuum motor manufacturers use to test their motors in a plenum. But their application only has one plenum. For more information on this code, research ASTM F 2105 "Standard Test Method for Measuring Air Performance Characteristics of Vacuum Cleaner Motor/Fan Systems".

I have a lot of Performance Trends products including their Port Flow Analyzer Pro, software, black box, pitot tubes, etc. that I plan to use in the near future. What manual of theirs are you referencing? Did you purchase their quick flow set up?

Hope this helps,

Terry_Zakis

Posted: **Mon Oct 18, 2004 11:21 pm**

I'd like to clarify one thing in regards to the orifices I have supplied. They were machined based on a flow co-efficient of .62, tested on the top of my pitot style flowbench and cfm numbers from my bench compared to the calculated numbers.

My parts are not made to a level of calibration which has been discussed nor are they sold for that level or usage. They are however for the DIY'er who wants to build a simple repeatable flowbench in their own shop without breaking the bank. Please keep this in mind, we are all building flowbenches based on a level of accuracy veruses cost we are comfortable with.

Hope this clears up any confusion.

Posted: **Tue Oct 19, 2004 8:39 am**

Thx for the replies guys. Gives me something to chew on and try a few more things.
I was hoping to not have to make a plenum as that would mean a redesign.
I was hoping to be able to get by with a performance trends type of setup, but if needed I will add a plenum.

Bruce,
I don't think anyone here is expecting your calibrated orifices to be accurate and tracable to a known standard in a lab environment. I did not imply that they were exact in accuracy. They have however been tested by a repeatable bench and have measured quite closely to the theoretical flow. Thats much more accurate than I can claim at the moment on my bench. It is just what I need to start up my bench calibration and troubleshoot issues.
The reason I needed them is to see just how far I was off in my measurements. I know that 20+ cfm is definetly an issue. I have a set of heads here that are flowing 20+ cfm low also.
That might mean a few things:
-Orifice tap position is no correct
-As mentioned above, the flow coeffecient I'm using for the orifice plate may be incorrect for my setup
-flow is too turbulent.
-other...
- combintaion of all of the above.

I'll only be able to determine what the issue is if I do some more testing.

The performance trends manual I found was on the web by doing a quick google search.
I didn't buy a performance trends unit.
Here is the link if interested

P.S. Has anyone stumbled across plans for a homemade laminar flow element?
detailed diagrams of it construction?

Posted: **Thu Oct 21, 2004 8:50 am**

I did some more testing.
I was able to get some reasonably accurate(close to theoretical) numbers when I switched to a smaller orifice(about 2.5"). That may indicate the biggest one I have(about 3.xx") is getting to turbulent to measure properly.
I was about 7cfm high on high flows (around 210cfms) and about 3 cfms low on lower flows (around 100cfms).
I'm going by memory at the moment, I'll have to post the actual data later.

I noticed that the inclined manometer does move slighlty during the measurment. What do people usually see in terms of +/- cfm flucuations with orifice based flow benches?

Posted: **Sun Oct 24, 2004 5:18 pm**

Hi Again everyone...
This topic is quite similar to what we discussed a year or so ago, about the plenum volumes and where to put the orifice compared to inlet hole (head) and suctionchamber...
I did a lot of testing back then, and decided to tear off the top of my bench and build a new one... I am very glad I did today, since the bench is now very repeatable, and hopefully soon accurate, since a friend of mine ran a couple of orifices on Autoshop's (florida) bench...
The only thing I did, was to relocate every hole in the bench as far from eachother as possible...
From the beginning I couldn't get any vacuum in the upper chamber because the orifice was located right under the inlet hole.

hope I could be of some help...
best regards, Jocke

Posted: **Sun Oct 24, 2004 5:32 pm**

Fluctuations (around 1-2 Hz) is quite common on high flow in my bench, you can even hear the sound changing 'directions' in the port (different tune) due to turbulence, this only happens one some heads...

forgot to mention in the earlier post, that the 3 pickup tubes are connected as far as possible from any hole inside the bench...and I do NOT have the test pressure tube connected in the cyl. adapter. It's connected at the same point (with a T) as the inclined manometer is, in the upper chamber below cyl. adapter and above orificeplate. The other end of the inclined manometer is connected to the chamber below the orifice.

/Jocke

Posted: **Tue Oct 26, 2004 3:37 pm**

Jocke,
So what configuration did you end up with? I could't make out from your post what exactly you had before and after the changes were made.

Did you have a simple pipe system (no plenum) before and change to an open plenum before the orifice?
or did you have a pipe only system and keep it that way only to change the tap locations?

It seems you only have 2 taps right now. How far away from the orifice are you taps? and how far if the first (inlet tap) from your test piece?

Posted: **Thu Nov 04, 2004 4:21 pm**

Hi again!

Sorry for late reply...
I had to move the orifice hole, because it was located right under the inlet hole (build a new top) we simply couldn't get any vacuum in the top chamber, because air moved through the two holes kinda like it was in a pipe!

There are open plenums above and under orifice.

The taps are located as far from any hole as possible in the two chambers maybe 8-10" or so...

The inlet tap is about the same distance (8-10") from the inlet hole/ head adapter.

Posted: **Fri Nov 05, 2004 8:35 am**

[color=#000000]Hi folks just read through this thread and it seems to me that some questions here are similar to my newly acquired ones. Did some posts on the old board, but anyways, after a year i finally finished my bench, found out that i had to build three orifices, buut from the beginning i had on my setup(SF600 clone, the cylinderheadsteady on the right side and orifice in a big tube(~11"dia)(to avoid turbulences) on the far left top plenum. I also have slight fluctuations, and a way to low read out on quasiinclined manometer.(stands linear and straight to avoid percentagecalc.) So i guess, from the formulas, i take my 11"testpressure, take read out from orif.manometer, and feed into the 13.55 x squareroot of orif.manom x orif.diameter squared(And those numbers i would have to convert to 28"in order to get cfm-numbers?? Is that right??. Playing a lil with, pressuretablocations didn

Tractorsport Flowbench Forum Archive

Tap position in Orifice based flow bench?