Tractorsport Flowbench Forum Archive

Posted: **Sun Oct 15, 2006 6:09 pm**

Hi my first post on here, after months of research I decided to build my own orifice flow bench using the FP1 digital manometers for ease. I feel I know understand the principle behind the flowbench quite clearly. There is one thing I can't find an answer to though is there a preferred chamber volume for the primary and secondary chambers either side of the metering orifice? Are they best left the same volume as each other? Or can I assume providing the motors can pull the vacuum, bigger is better in order to provide a less turbulant airflow through the orifice? I'm hoping to start next weekend so any light than can be shed on the chamber volumes would be great.
Thankyou JP

Posted: **Sun Oct 15, 2006 6:53 pm**

Good question to ask.

I know of no specific ratio or size that is claimed to provide optimal performance. I can say from experience that if you build the cabinet per the MSD bench design, the chamber volumes will work BUT you would be well advised to install a baffeling system above the orifice plate.

If you're wide open to cabinet design concepts, you may want to consider locating the orifice away from the opening in the top of the bench [opening on one side - orifice on the other] to allow the air to calm down before it enters the orifice.

What you're after is pressure differentials that are not directly influenced by turbulence.

By using the FP1 setup, you will find that when it comes time to calibrate the bench, it's a straight forward operation with Cd value changes for the orifice under test.

On the exhaust side of the orifice [the center board where the orifice rests] make the opening large enough so that it has the least amount of influence on the air entering the orifice on exhaust tests.

If it were just slightly larger than the orifice, it could definitely affect the Cd of the orifice.

Some people make that center chamber with an additonal baffle plate below the orifice.

Posted: **Sun Oct 15, 2006 7:21 pm**

Thank you Larry for that answer, that confirmed a couple of ideas I had floating around my head.
I was going to incorporate the valve system of the msd bench to regulate inlet and exhaust flow as I felt this was a relativelly simple method to incorporate into a design. Main difference being that I'll be making a perfect cuboid as a box with the valve controls sitting on the top surface of the bench. I figured the simpler the lines the less turbulance. Next I was going to offset the top aperture and the orifice along with a baffle. I definatelly like the idea of placing a baffle in the middle chamber too like you just mentioned, I think I will try and incorporate this into the design as this will probably aid flow past the regulating valves with minimal turbulance.
The centre board, housing the orifice, is going to have a machined 1/8" metal pate inserted into it with a recessed lip of 1/16" with a groove and o-ring seal, again all the orifice plates would have the same 1/16" recess so the two should mate very cleanly.
I'm Uk based so most of the heads I'm doing are 4 valve and 5 valve per cylinder configurations with flow figures upto 275-300cfm, however most are around the 150-200cfm mark. So motor wise I've opted for 4 x 1200w wet and dry motors with the option of upgrading to 6 or 8 if the need ever arises. Likewise the inside of the box will be varnished to make it 100% water proof to give me the long term option of wet flowing aswell.
I would scan in a picture to show my ideas unfortunatelly my scanner is currently living in a removals box as I'm moving in the next 5 days to new house and new workshop premises hence building the flowbench as I finally have the space.
One more question though if I'm introducing the baffles in both upper and middle chambers would I place the tubes for measuring pressure differential inbetween the baffle and the orifice plate or on the opposing side of the baffle? Or again will this not make a difference?
Cheers JP

Posted: **Sun Oct 15, 2006 10:38 pm**

Quote:

"What you're after is pressure differentials that are not directly influenced by turbulence."

Gentlemen, I have had many conversations with Harold Betties (from Super Flow) about this attempt to remove Turbulence. His comments were that you will ALWAYS have turbulence on the upper chamber and they even go out of their way with the baffle to have more turbulence. You should probably have a baffle on both sides of the orifice plate if you are doing multi-directional testing.

The volume in the chambers is to slow down the velocity of the air and not necessarily eliminate the turbulence of the air. The benches I have worked with had anywhere from a 55 gallon drum as a settling chamber to a 8 cubic foot chamber (2x2x2) on each side.

Hope this helps.

Tom V.

Posted: **Mon Oct 16, 2006 4:53 am**

OK so if I understand correctly from the above, as the volume of the chamber increases, velocity of air will slow down. Which in turn will give me a more stable reading as there will be less high speed fluctuations past the sensors? So I assume from this that the pressure sensor is then best placed inbetween the baffle and the orifice to provide the most stable figures. If that's so I think I'm ready to start building

I'll take some picies along the way and post them up on reader's benches.

Posted: **Mon Oct 16, 2006 6:43 am**

Posted: **Tue Oct 17, 2006 3:44 am**

[color=#000000]In general, the onset of turbulent flow occurs at Reynolds numbers between ~ 2.3

Posted: **Tue Oct 17, 2006 7:01 am**

[color=#000000][quote="Greenlight"]In general, the onset of turbulent flow occurs at Reynolds numbers between ~ 2.3

Posted: **Tue Oct 17, 2006 8:17 am**

^^^^^ lol glad to inspire new discussions (edit) with my first ever post. I have to say makes definatelly for interesting reading this as it has opened my eyes up to altering my original designs, hopefully for the better. Still haven't got an answer though on the best pick up location with regards to the pressure reading for the fp1?

I have got another question in the meanwhile with regards to the 28" of water. I'm under the understanding that this has been introduced as the industry standard as it is felt that it best simulates the conditions created by an engine under use in a real life situation. I assume rigorous testing has come into play somewhere along the line to achieve this constant. Well I was contemplating on my latest car project to try and establish this figure in the field.
The car I'm working on should be ready early next year and will have a fully functioning PC build into it to realtime monitor the aftermarket management. So I was wondering whether the FP1 would be acurate in such conditions and what would stop me from tapping a pressure sensor into the intake manifold as close to the head as possible. Hopefully giving me real life figure as to whether the engine is actually pulling 28" of water or whether it varies depending on the driving styles? Would this be a worthwhile experiment to conduct?

Posted: **Tue Oct 17, 2006 9:30 am**

Posted: **Tue Oct 17, 2006 9:58 am**

Sorry been in London too long, you end up with a very brash interpretation of the English language

Bruce differing oppinions are good like you say, as otherwise there would be no need to push the boundaries.

Posted: **Tue Oct 17, 2006 5:10 pm**

Posted: **Wed Oct 18, 2006 3:32 am**

Larry and/or Bruce:

I would like to post a graph of Reynolds number vs. Discharge Coefficient that I have to help spark this discussion, but I don't know how. Can I email it to one of you for you to post it here?

Posted: **Wed Oct 18, 2006 6:31 am**

I"m sending you a PM with my email address...check your private messages...

Posted: **Wed Oct 18, 2006 7:58 am**

Since going away from pitot metering for my bench I've been getting bogged down in Laminar Flow theory.

In the laminar range pressure drop is proportion to flow.
In the turbulent range pressure drop is the sqaure of flow.

Now I tripped over some data that has transitional flow, between laminar and turbulent, progressing from proportional to square law in a predictable way. I wondered after reading that, what regime applied to orifi flow benches.

It seems greenlight is alluding to them being in the transitional and turbulent range ?

Anyone done the calcs for sf110 and 600 size orifi and pressure drops to determine the reynolds numbers at say 0 / 50 / 100 % of the range ?

Also is seems to me that the definition of laminar / transient / turbulent in relation to reynolds number is at a molecule level.
Whereas general discussion about flow straighteners and turbulent flow is on a level many multiples larger. More like streams of air columns going across the metering instead of concentric with it.

Tractorsport Flowbench Forum Archive

primary and secondary chamber volumes