Sacred Screw Reset

[GonzoB]

This has been very interesting. Congrats to Roy!

Just throwing a small spanner in the works, how do we know that the 500mm WC is providing a valid comparison to air going through the TB?

The fluid mechanical approach to this (I think) would be to try and get similar Reynolds numbers for the fluid dynamics. If the RN's are the same, then the flow patterns are likely to be similar. No guarantees, but likely.

The RN form applicable here is probably RN = M/(D*U), where M is the mass flow rate, D is the hydraulic diameter and U is the dynamic viscosity. D is the same (because we're using the same TB), so that cancels out. This means that we need to have the mass flow rate ratio about the same as the viscosity ratio.

Water has a dynamic viscosity about 50 times that of air, so we need to get a water mass flow rate about 50 times that of the air mass flow rate. Roy's water mass flow rate is about 30g/sec. To be comparable therefore we'd need an air flow rate of about 30/50 = 0.6g/sec. Is that in the ballpark of what a MG engine would be doing at idle? I guess you could work it out from the AF ratio if you knew the fuel flow rate...

If we could find that out, then Roy could adjust the water column to get the correct mass flow. Then we'd be a lot more confident of the method.

Gonzo

[Huzo]

Well..
If Roy has determined how much water passes through a known good TB in a certain time, then he can adjust the screw in a rooted one to achieve the same dump rate..
It therefore matters not at all what the Reynolds numbers, viscosity, s.g., values are.
The fact is that he is performing the test on the molested TB under identical conditions as the control one, farting around with equations in Physics is doing nothing more than giving a quantitative account of the results.
One would think that once adjusted, Roy could run a side by side dump test of the two bodies and they would flow the same volume/time.
A slam dunk...!
If I had the balls to get off my arse and do the same, I would have used metho or petrol because it passes through a small aperture with less impedence, but he has done the work not me.
Once calibrated and adjusted however, I’d think the final test could be run side by side with petrol for the final fine adjustments, but Roy is the MAN, and I have a lot of confidence in his work...

[tris]

+1 Huzo

This is seriously good work IMO

If we could get sufficient data points (30+) and dump the data into a control chart we could nail this issue

The best I'd found up to this point was this on GuzziTech https://www.guzzitech.com/forums/threads/stepper-motor-on-off-kit.3025/page-4#post-28926

Bravo Again 

[GonzoB]

Well..

If I had the balls to get off my arse and do the same....

OK Huzo, I take the challenge. By the way, I'm not knocking Roy's work - I think it's great. I just wondered how representative it was for air flow. Roy himself said that he chose the size of equipment because it was convenient to get a repeatable result. That in itself says that he's probably in the right ball-park, as the guys at the factory are going to be looking for something that is repeatable when they are working with air.

Anyway, here goes...

Fuel use at idle is something like 0.09 USgall per hour per litre engine size, which is 0.095 ml per sec
Fuel mass flow rate is therefore 0.071 g/sec  (fuel is about 0.755 kg/litre)
The air mass flow rate is therefore 0.93 g/sec (with a 13:1 air fuel ratio)

So, to get the same Reynolds number you'd have to have a water flow rate of 0.93g/sec *48 = 44 ml/sec  (48 is the ratio of the dynamic viscosities of water and air)

At 30ml/sec Roy is pretty close! To get 44ml/sec, which is ~50% increase in flow, so you'd need about 1m water column.

Note: The "fuel use at idle" is probably not very accurate for our size engines, but I doubt that it would be a factor of 2 out.

Conclusion:  I think Roy's model is pretty good.

Gonzo

[Kiwi_Roy]

I really must thank John aka Moto for taking my raw data and making sense of it , it almost looks like I knew what I was doing.
He also suggested a better way of measuring the flow rate using a graduated cylinder, I would certainly do that if I need to repeat the measurements.
The point of this exercise was to prove that it's possible to re-set the sacred screw without having to use an expensive flow bench. The test rig I made was put together using a few simple plumbing fixtures to take advantage of known properties
It should be possible for me or someone else to set another throttle body off a Griso 1100 to match the untampered one I have, all it needs is a simple rig to duplicate the conditions I subjected mine to, basically a 500 mm column of water from the butterfly shaft to an overflow point and a method of collecting and measuring the water that passes through the gap around the butterfly (the Air bleed must be closed to do this)
John suggested a better way to measure the flow, the detailed way he described this is what made me think he was a lab technician, Im sorry about that. We know the flow rate should be for this throttle body so it's a matter of making successive runs and adjusting the sacred screw until the fill time matches. Of course if we were able to check other good throttle bodies it would increase our confidence.

John further commented "If you have a concise argument about the properties of fluids that make a substitution of water for air appropriate here, it might help other readers. But I think it is already evident that what you are doing makes sense. Other readers are confusing the issue of approximating the dynamics of air with water, when what you are really doing is measuring the position of the butterfly valve accurately enough to set it at the factory value"
This is exactly what I'm doing, I know the properties of air and water are totally different i'm not trying to calibrate the throttle body all I'm trying to do is reproduce the setting in a way that anybody could do it, you don't need a science degree just reasonable care.

Further to the graph that John provided, the first third was done with a dirty body, its very important to give it a good scrub before you start,  I was thinking about just erasing this data but Imm glad I didn't John was able to use it.
Note how the second third is very stable.
https://ibb.co/wgs13Ws
The last third really shows how sensitive the Sacred screw is, this is good in a way it means its very easy to adjust this is what an owner would have to do with his tampered bodies set it in the rig and adjust it for 30 ml/second.

If you look at points 4,5,6 & 7 on the graph effectively what I am doing is substituting a different throttle body and dialing it in to match the Left hand one, the Bell crank screw serves the same purpose for the Right Hand body that the Sacred Screw serves for the Left Hand body. I never tried to get it exact, just made several step changes to show how sensitive the screw is to tweaking, If you ignore the fact that its the same set dealing in a different set is that easy.

I only have one set to play with, if someone in North America would send me a messed with set I will endeavour to get them back in line.
It would also be nice to test some other models like the Norge and the Griso 1200 if someone could make their throttle bodies available for a short time I will happily put it through the wringer.
I mentioned before that the 1100 Norge set may be similar to the 1100 Griso, does someone have any air flow numbers from the factory?

When I get some time I will sit down and do a drawing of my rig, its really very simple and there are only a couple of important dimensions.

[moto]

Thanks, Roy!

All,

I would like to re-emphasize the importance of the sizes of the step-like jumps in the graph of Roy's results I posted earlier. They are much bigger than his errors in repeated measurements during the 18 different individual trials, as a quick look back at the graph on the first page of this thread will confirm. Because his trials varied the throttle body settings by a single flat (1/6 of a turn) or a single thread (1 pitch) at a time, we can see that he can distinguish departures from the correct setting (whatever it is) to more accuracy than a single flat or a single thread. Based on my experience in balancing throttles, I expect that ham-fisted mechanics would make changes greater than a single flat or thread, and therefore that Roy's device, as used by him, is accurate enough to detect and correct them.

The key here is that Roy has a way of detecting throttle setting departures that is very precise. It does not matter, I believe, whether the water fluid dynamics mimic those of air, but only that his method is sufficiently precise.

Moto

EDIT: Oops. I see Roy has now modified his post to include this point. But I'll leave this post here in case it is still helpful, and since it emphasizes taking another look at the step sizes in the graph (and comparing them to the random variation around the lines within each trial), which I found most revealing.

M.

[GonzoB]

 It does not matter, I believe, whether the water fluid dynamics mimic those of air, but only that his method is sufficiently precise...

Hi Moto.

Yes, I believe you are correct. As it turns out I think Roy's model closely mimics the air flow. If, however he was trying to pass, say, 10 times the equivalent air flow, or 1/10th of it, then other effects may have come into play that are not obvious (eg: turbulent/laminar flow effects), and the changes to the adjustments may not have been as relevant. The calcs show that he is in the right zone, and I think it is a great result.

Nice analysis of his figures, BTW.

Gonzo

[Huzo]

Ok, moving right along..
Pizza Guzzi has given me a set of throttle bodies in nice condition, but molested. There’s a bit on at present, but a set to experiment with is a massive bonus. I still need access to a good set which I don’t have, but Pete Roper can loan me a set I think.
Anyway, here’s the thing...
I’ve long wondered how one can load a supply tube with an outlet and purge the whole system ,then start the dump and have a crisp cut off point where flow stops, there will be a period where dribbling of the test fluid occurs at the end of the run and as such, the results will be meaningless.
But salvation is at hand...!
Here is a schematic of a simple rig to run a series of test dumps for a control measurement, however there is a difference in what I have suggested in the past.



You’ll see that this setup allows the rig to be filled (after closing the bleeds) and the air can be purged easily.
Now we have the entire system full with no air.
Opening the dump will result in a slow draining of the top tank, but the rate of drop will increase as the meniscus enters the thin transparent tube.
The stopwatch can start as the fluid passes the top mark and stop as it passes the bottom mark. Several runs can be made and the median can be found. I would hope that the dump rate would be the same over 10 runs, if so, then the efficacy of the process could be considered reliable.
If the drop time over the metre was say, 50 seconds, then the buggered one could be attached and manipulation of the Sacred Screw can be employed to achieve the same drop time.
Clearly if the shagged one dumps the same volume in 46 seconds for example, the screw can be adjusted to achieve the same dump time over several tests.
Then I think one would have to admit that the recalibrated one has the same flow rate as the good one...
Isn’t that what we all want....?
All that’s required after that is to attach the other one and wind the connecting rod to achieve the same rate...(I think).

[Kiwi_Roy]

I think that should work Huzo but here's the thing, if someone else wanted to reproduce your results they would have to duplicate your rig exactly, I think mine could be reproduced much easier, the biggest problem I had was getting a seal from the column to the throttle body, it didn't matter how much leaked out there because I only measured what passed through past the butterfly.
My rig doesn't have dimensions that are important other than the head (column height over the butterfly), I could make the column of 1/2" pipe or 4" pipe, the flow over time would be pretty much identical.
I used a height of 500 mm, I think 1000 mm would be too unwieldy and besides the flow rate is not significantly higher, I think the rule to double the flow calls for squaring the head, I'm sure we have some pump guys on here. So saying the difference in flow between my 500mm and your 1,000mm can be calculated
Your rig will slow down as the level drops and you cannot afford to have any leakage.
You are basically proposing to measure the flow into the throttle body, I measured the flow out, should be the same eh!
I used water because it's so much easier to measure than air which is compressible and effected by temperature.
Gasoline would be fun, I got soaking wet using water lol
I don't know why you would chose gasoline, its normally air that flows through the butterfly.
BTW
With my instrument designer hat on, control valves are usually specified by their Cv rating, the ability to flow water at 1 psi difference between inlet and outlet.

[Huzo]

There is no point that you have made that is not valid Roy.
However I would like to add some angles. The longer the transparent tube is, the longer will be the dump time over the test and so consequently the tolerance in the stopwatch accuracy will become less significant.
A half a second each end of 1 minute is not as significant as a half second in 30 seconds.
Also, the flow rate will alter during the dump run because the mass of fluid above the butterfly is reducing. However this will not matter because you have tested the dump time over 10 or so runs on the good TB.
We are only attempting to replicate the dump time on the maladjusted one.
Anyone who replicated my rig will get a different reading for dump time, but it will be THAT figure they are attempting to replicate, not mine.
For what it’s worth, what volume of water did you dump per minute ?
Also I’ll reiterate that one can put a 50 thou’ or so packer under the stop screw on the control TB and put that same packer under the one for testing. This will allow for a higher flow rate and I have an inkling that this will yield more accurate results.
But I cannot prove it..
Yet...

[Kiwi_Roy]

There is no point that you have made that is not valid Roy.
However I would like to add some angles. The longer the transparent tube is, the longer will be the dump time over the test and so consequently the tolerance in the stopwatch accuracy will become less significant.
A half a second each end of 1 minute is not as significant as a half second in 30 seconds.
Also, the flow rate will alter during the dump run because the mass of fluid above the butterfly is reducing. However this will not matter because you have tested the dump time over 10 or so runs on the good TB.
We are only attempting to replicate the dump time on the maladjusted one.
Anyone who replicated my rig will get a different reading for dump time, but it will be THAT figure they are attempting to replicate, not mine.
For what it’s worth, what volume of water did you dump per minute ?
Also I’ll reiterate that one can put a 50 thou’ or so packer under the stop screw on the control TB and put that same packer under the one for testing. This will allow for a higher flow rate and I have an inkling that this will yield more accurate results.
But I cannot prove it..
Yet...

Your rig may get more accurate results, the more volume you pass the less error you will get, I chose to use many tests of low volume rather than one of high volume. Ideally I would have used something like a 1 Litre graduated cylinder, I didn't have one handy so I used a 500 mL bottle and a stopwatch. The other nice thing about water is 1 L weighs 1 Kg so you could use a decent weigh scale instead.
You don't actually have to measure your Volume, you just measure time you don't even have to know the volume, that's one advantage.
The flow will slow down as the head drops not the mass, its all about the pressure drop, that's why i say a 1/2" column will give the same flow rate as a 4" column at any given head (500mm in my case).
But how would you convey that information to someone on the other side of the earth, they would have to duplicate your dimensions.
I did try simulating a screwed with sacred screw using feeler gauges between the screw and the stop.
Note: The throttle balance can also be checked, its just a matter of replicating the flow on the other body by adjusting the bell crank.
I look forward to an on-going discussion.

[Huzo]

You don't actually have to measure your Volume, you just measure time you don't even have to know the volume, that's one advantage.
The flow will slow down as the head drops not the mass, its all about the pressure drop, that's why i say a 1/2" column will give the same flow rate as a 4" column at any given head (500mm in my case).
But how would you convey that information to someone on the other side of the earth, they would have to duplicate your dimensions.
I did try simulating a screwed with sacred screw using feeler gauges between the screw and the stop.
Note: The throttle balance can also be checked, its just a matter of replicating the flow on the other body by adjusting the bell crank.
I look forward to an on-going discussion.

Nah mate..
The flow WILL slow down imperceptibly because there is less water above the butterfly so therefore less mass. Imagine the head was 10 metres high, you will get a greater flow than I metre, but that is immaterial because you only need a control value from YOUR good TB, then replicate THAT figure on YOUR maladjusted TB.
Pressure is mass/area.
Also.
You do not need to convey your dump rate to anyone else, because they can do multiple test runs on THEIR good TB.
The discussion regarding column diameter is interesting but again does not apply to my model, because the variables are the same on the calibration run, as the control run.
In addition.
My model only needs a good seal between the input column and the TB and I can witness and measure the descending meniscus in the tube. It will result in a time of say 60 seconds and repeat runs will reveal a real time indication of the predictability of the test.

[Kiwi_Roy]

I agree if you are working with one good body and one screwed body your rig will be fine.

Look back at Moto's post ~22, he took my crude data and massaged it to show a pretty constant 30 mL per minute for the readings between run no 30 and 60
Runs 0-30 were done with a dirty throttle body showing how important it is to scrub the throttle body and 60-90 were done with feeler gauges to simulate a fiddled with sacred screw.
If you had a 2 Valve 1100 it should flow 30mL per minute, at least my Griso still has the paint seal intact on the screw.
I'm sure you could extrapolate my results to match yours but the math required is beyond my 4 function calculator brain.

Pressure is mass/area, I don't follow that, pressure is head, nothing to do with Volume or weight (mass), I'm probably confused.
What I mean is the pressure at the bottom of my 500mm column is exactly half the pressure at the bottom of your 1,000mm one whatever the profile.
We are probably saying the same thing.

Whatever method you use i'm sure it will have good results, you will find it easy to dial in the screw.

[reidy]

I don't have a bike with a sacred screw, but still found this a good read. If I am interpreting the graph , the point I take out of this is how much a dirty throttle body will effect flow rates at idle or small throttle plate movements.

It would effect all bikes with throttle bodies, and if one was to become more dirty than the other due to crankcase venting or something similar it will upset the balance by quite a bit.

Steve

[Huzo]

I agree if you are working with one good body and one screwed body your rig will be fine.

Look back at Moto's post ~22, he took my crude data and massaged it to show a pretty constant 30 mL per minute for the readings between run no 30 and 60
Runs 0-30 were done with a dirty throttle body showing how important it is to scrub the throttle body and 60-90 were done with feeler gauges to simulate a fiddled with sacred screw.
If you had a 2 Valve 1100 it should flow 30mL per minute, at least my Griso still has the paint seal intact on the screw.
I'm sure you could extrapolate my results to match yours but the math required is beyond my 4 function calculator brain.

Pressure is mass/area, I don't follow that, pressure is head, nothing to do with Volume or weight (mass), I'm probably confused.
What I mean is the pressure at the bottom of my 500mm column is exactly half the pressure at the bottom of your 1,000mm one whatever the profile.
We are probably saying the same thing.

Whatever method you use i'm sure it will have good results, you will find it easy to dial in the screw.

Yeah we probably are saying the same thing, but I see no significant errors in your thought process.
The main thrust of both approaches is to replicate the performance af a newly adjusted TB with a known good one.
What would be a great test would be to get two known good TB’s and see if they both flow the same rates.

[Motormike]

I've always wondered about this "sacred screw" stuff.  You'll have to enlighten me...what are the consequencses of someone tampering with said screw?  The bike won't run? Won't idle" Runs poorly? All the above. The reason I ask is most V11's I've looked at show signs of the "Italian locktite" (yellow paint) on said screw having been broken.  When you consider the multiple owners these bikes go thru, the odds of someone not fu*king with it seem low indeed.  When Moto Guzzi was making V11's and other models with this throttle body, I can't imagine there was someone in the corner of the factory pulling each one out of a box and putting it individually on a flow bench.  Kind of slows down the assembly line just a bit, don't you think?  More likely they were set at the supplier before shipping.   

[Huzo]

The sacred screw is used to adjust the butterfly at the factory, to achieve a nominal flow rate.
The idle is adjusted by the stepper motor automatically in daily use by the ECU after it crunches the incoming data.
Buggerising around with the sacred screw to set the idle speed on the throttle stop as you would with carbies, will introduce anomalies into the system, most of which are beyond my ability to explain or describe.

[lucky phil]

I've always wondered about this "sacred screw" stuff.  You'll have to enlighten me...what are the consequencses of someone tampering with said screw?  The bike won't run? Won't idle" Runs poorly? All the above. The reason I ask is most V11's I've looked at show signs of the "Italian locktite" (yellow paint) on said screw having been broken.  When you consider the multiple owners these bikes go thru, the odds of someone not fu*king with it seem low indeed.  When Moto Guzzi was making V11's and other models with this throttle body, I can't imagine there was someone in the corner of the factory pulling each one out of a box and putting it individually on a flow bench.  Kind of slows down the assembly line just a bit, don't you think?  More likely they were set at the supplier before shipping.

The V11's dont have the "sacred screw"

Ciao

[Kiwi_Roy]

I think it's just the CARC bikes that have the sacred screw, Pete Roper has made a point of telling us not to touch it.
Some owners or ham fisted mechanics think it's the idle speed adjuster and move it to find the bike runs like a bag of aholes.
As Huzo says the idle speed is set by a stepper motor and its also thrown out by misaligning the Bell Crank screw that sets the RH body opening.
The CARC bikes are completely different to setup compared to the earlier 2000 era bikes.

[Kiwi_Roy]

Huzo,
        I think I may hav misunderstood your sketch, I thought you were timing the funnel shaped section but on further reading I see you are just timing the straight section below it.
 



To that I would say if your straight section was reasonably large you don't need the enlarged section above, the flow I measured is only about 30 mL/second so you just need a few inches to give you time to get the stopwatch ready. My rough calculation says the level would drop about 1 inch per second for 1-1/4" pipe, I probably have that wrong.
Do you have some clear pipe in mind, it would be good to use something commonly available so that others could reproduce it or perhaps a glass window in a common plumbing fitting like an 1-1/4 tee.
Do you really need a valve below the throttle body, perhaps a cork or just bung your hand over it while filling.
One problem I had was adapting the column to the throttle body to get a watertight connection, it wasn't so important for mine since I was measuring the flow past the butterfly. Since you have a lathe you could make something to go inside or outside the body sealed by an "O" ring, you can't afford any leakage there.
I had to use the rubber adapter off my bike, not the best solution as you don't want to take a bike out of service.
I'm sure someone here is clever enough to compare my results to yours.
I look forward to see what you come up with.

[Huzo]

Yeah mate..
The reservoir above the clear section, is just so there is adequate time to prepare as the dump run section is timed.
Also I like the idea of a valve up the blurter, because that allows the system to be filled and properly purged before the run.
As for replicating my rig, again this is not necessary because any rig that you build can be calibrated by doing a control run on a known good TB.
You then have the figure that YOU need...
BTW..
30 ml/sec is a lot more than I imagined the nominal flow rate to be.

[Kiwi_Roy]

Yes, if you look back on page 1 at the graph Moto did that's where its sitting, I had to enlarge the spout from 1/2" tube to reduce the back pressure
Remember this is the normal air flow passage for ~1100 RPM
I think its important that others can replicate your rig, they may not have the luxury of a good TB to measure, but if they have your good body data they don't need it they can just bung on the abused one and dial it in to match your good TB data. 
And don't forget the rig can be used to set the right hand TB balance as well using the Bell Crank screw.

[Huzo]

Yes, if you look back on page 1 at the graph Moto did that's where its sitting, I had to enlarge the spout from 1/2" tube to reduce the back pressure
Remember this is the normal air flow passage for ~1100 RPM
I think its important that others can replicate your rig, they may not have the luxury of a good TB to measure, but if they have your good body data they don't need it they can just bung on the abused one and dial it in to match your good TB data. 
And don't forget the rig can be used to set the right hand TB balance as well using the Bell Crank screw.

Ok yes, fair point.
Given that the only quantitative testing I am taking is between the top and bottom marks on the transparent tube, that realistically speaking, is the only bit they’d need to replicate. The area of the inside diameter x distance between the top and bottom mark, describes the volume, that divided by the time is the rate of flow of the fluid.
A piece of glass tube with accurate I.D. Would be obtainable most anywhere and this is the “measurement section”, none of the rest of it is doing any measuring.
My phone is +61437070946 if you want to txt me yours, I’ll call you...

[Huzo]

Oh, just one more thing.
I think that methylated spirits is a better choice than water, given that metho will pass through a small aperture more readily than water, thereby more closely resembling air.
Probably not necessary, but a move in the right direction.

[Kiwi_Roy]

Oh, just one more thing.
I think that methylated spirits is a better choice than water, given that metho will pass through a small aperture more readily than water, thereby more closely resembling air.
Probably not necessary, but a move in the right direction.

An air molecule is larger than a water module by half again, Water 2.75 Angstroms, Air 3.7, you may get more volume of air but not more mass.
Water has the weirdest properties. The company I used to work for sold filter media that was so fine it would let water through but stop air in its tracks.
Meths, I think you must be a pyromaniac at heart lol

[Huzo]

An air molecule is larger than a water module by half again, Water 2.75 Angstroms, Air 3.7, you may get more volume of air but not more mass.
Water has the weirdest properties. The company I used to work for sold filter media that was so fine it would let water through but stop air in its tracks.
Meths, I think you must be a pyromaniac at heart lol

I have seen mesh filters that will let metho through, but the surface tension of the water stops it passing.

[GonzoB]

Against my better judgement, but to correct a couple of errors I'm wading in again....

My analysis earlier this year had some errors that the members here have kindly decided not to point out (to avoid embarrassing me). I didn't correct for the Griso cylinder volume, and I used dynamic rather than kinematic viscosity.

I have re-done the calcs, and have included ethanol as the working fluid. Essentially the calcs take the (web obtained) idling figure of 0.09 USGall of fuel per hour per litre engine size, and find the airflow volume, then adjust that for the change in kinematic viscosity for the various fluids, keeping the Reynolds number the same in order to maintain the same flow characteristics.



This shows that Roy was almost spot on in his selection of 30ml/sec (well done Roy!).

It also shows that if you wanted to use ethanol (meths is mostly ethanol) you'd have to increase the flow by 50%, which means the pressure (i.e the head) would have to go up by about 2.25 times. Makes for a big apparatus!

The surface tension is irrelevant, as we are talking about a flooded system. It's only the kinematic viscosity that's important.

Gonzo

[Huzo]

Ok Gonzo Thankyou.
Are you suggesting that for a given head, metho will not flow as freely through an aperture as water in the same apparatus ?
It would surprise me, but I have been surprised before.
Do you have a comment regarding my proposal ?

[GonzoB]

Ok Gonzo Thankyou.
Are you suggesting that for a given head, metho will not flow as freely through an aperture as water in the same apparatus ?
It would surprise me, but I have been surprised before.
Do you have a comment regarding my proposal ?

No, what I am saying is that to get the same flow characteristics (ie Reynolds number), you will have to pass more ethanol through the orifice than you do water. If you have to get 50% more through the hole it will likely need a significant increase in head.

I don't know if maintaining the Re number is important because I don't know the characteristics of the orifice. It may be that the test will work at much lower flows, but without better information, maintaining Re is a safer approach.

Gonzo

[Kiwi_Roy]

I'm having a hard job following the technical discussion, what does it matter what the Reynolds No, Viscosity or any of the other characteristics are we are not trying to compare one fluid to another, all we are trying to do is run water through one throttle body and set another throttle body so it flows the same amount, apples to apples.
I chose water because we have it in abundance, its cheap and not likely to burst in flames or create an environmental disaster.
I'm an Instrument Technician used to measuring stuff, air is really hard to measure because you can't see it and its greatly effected by the conditions, water is dead easy because you can catch it or time the drop in level as Huzo proposes. The actual flow rate is not important I could fill a coconut shell in X seconds then switch to the other body and adjust the screw to give me X second fill time and its done, I don't need to know how much the coconut shell holds and I don't need to know the properties of various fluids.
The only time the quantity is important is if someone else on the other side of the earth needs to tweak a throttle body, he could easily reproduce the column and do his own flow test I could of course send him my coconut shell and say fill it in X seconds.
Heres where Huzo's column has merit, if you duplicated it closely all you would need is a stopwatch (cell phone)