Venturi Flow Element?

[HDgyro]

Hi Guys,

I've seen pitot, MAF, floating pressure and other alternative bench types discussed, but very little discussion of venturis used as flow metering elements. It looks like tapered PVC reducer couplings, or even a piece of straight PVC pipe reshaped using heat and pressure to constrict CSA in the middle, might allow creation of a useful venturi metering element.

Assuming a flow path similar to the PTS bench was arranged, would not a venturi still be ratiometric? What other pitfalls await?

I wouldn't be considering this, but my application only requires about 100 CFM of capacity, and PVC construction looks appealing in something this small. Thanks!

[Tony]

How do you calibrate it ?

That is the problem.
Many possible ways to measure airflow, an automotive mass air sensor or even a small anemometer will give an output that varies with airflow, but its pretty useless in a practical flow bench unless its repeatable and can be calibrated.

[HDgyro]

...its pretty useless in a practical flow bench unless its repeatable and can be calibrated.

If Bernoulli's math is correct, I should be able to determine with decent precision the correct tap point(s) along the slope of the venturi's cone based on inside diameter to produce the drop my manometer wants to see at any expected flow rate. Afterward, a calibrated orifice plate on the bench's test port should allow touching things up if they're slightly off by adjusting the slope of the inclined manometer, right?

The only issue I would expect would be errors produced by irregular shapes inside the venturi, but the pipes are uniform extrusions, and molded PVC pipe reducer couplers I've seen have been very uniform and smooth. I'll concede that a homebrew "flattened pipe" venturi could be problematic in this area, but once I got the tap point right, I'd expect it to hold calibration.

For that matter, starting with two different pipe diameters joined by a coupler, with pressure taps in the straight sections of both pipes, should be expected to produce accurate measurements based on the ratio of their inside diameters. I might not need to tap the tapered section of a coupler at all if the two pipe diameters got me close to 100% on the manometer at the max flow I need.

I don't know whether tweaking the coefficient of discharge in a digital manometer would be a useful calibration technique, since a venturi has a CD very close to 1. But otherwise, how would calibration or repeatability of this style bench be different from an orifice bench?

[Tony]

Flow meters that use any of the the "fluid velocity in a pipe" measurement techniques usually requite an absolute minimum of ten to twenty pipe diameters of dead straight pipe up stream, and a lesser amount down stream to ensure something like an even non turbulent flow.
While that works reasonably well in industry where you have existing long straight piping runs to monitor, it becomes a bit impractical for a home flow bench through sheer physical size.
Changing flow ranges means a whole different complete piping setup.
You cannot just swap over a short measurement section the way you can swap to a different orifice plate.

The humble settling chamber and orifice plate works, and is compact and is repeatable.

[HDgyro]

Thanks, Tony!

I like to understand pros and cons as much as possible before deciding how to proceed on something like this, and I knew there had to be a reason why venturi metering is not more widely used in this application.

This is exactly the kind of experience-based info I need.

[mk e]

This is the way I built mine.

Many years ago before I could just google it and learn the problems I built one thinking I'd use math to cal it...I kind of got it working with a very fine screen cone right below the cylinder but that's long gone. But when it was time to have another I returned to that experience and the one a built a few years ago I just made a set of orifice plates I lay on top, plug the readings into excel and I THINK its calibrated....think.

Its just a 4" pipe to a a 2" pipe. I was using zero + 5 orifice sizes but convinced myself zero+3 was better because the r^2 value comes out 1 by definition the formula is exact and comparing I didn't see I was missing anything not using the extra points. Its time to dust it off again ...so if I've done something really bad maybe someone will set me stray

[HDgyro]

Thanks for sharing details of your bench. I've seen other examples of venturis made that way, with no taper between the sizes, some even in laboratories where specific projects required improvisation. It seems they can be accurate, but the sudden step can hurt flow and produce turbulence.

Even if I needed 10 - 20x diameter to straighten the flow, that seems do-able in my garage. I might have to ask the Mrs. to move the Prius each time I want to roll it out from its storage spot against the wall, but that's not the end of the world. She'll probably want to leave the house when I run it anyway.

[mk e]

The standard pipe connector has about a 30 degree taper in the reducer which is text book perfect angle for min for restriction. Then on the expanding side I skipped the long hard to make 6-8 deg taper since the distance is so short the loses couldn't be huge (although it does expand at the Tee) and I figured I really didn't much care about flow losses after the measurement. Anyway it seems to work and is certainly able to tell me it I made the port better or worse.

I just I think stuck a valve and smashed the piston and all 4 valves to bits, yuck!.....so I'll be rolling the flow bench out of the corner soon to get the repair matched back up to the

Edit: The turbulence thing....I have 4 pressure taps on the 4" and 2 on the 2"just in case, well really assuming the flow is not uniform....I probably should add a straightener or 2 but it seems to work fine.

[1960FL]

Anyway it seems to work and is certainly able to tell me it I made the port better or worse.

For what most of us are working on this where the thinking should be, to many looking for the bench to be within .00001% when they can't even hold their hand still while trying to port a head with an old drill and stone....

Great job on the bench Mk e how many steps (sizes) have you tried calibrating across in orifice dimensions/CFM?

Rick

[mk e]

Yeah I built it with a specific project in mind so the ranges are sized for that.

I made 8 orifice holes in 2 plates, but small ones (below the 67cfm@10") I really don't use as my resolution isn't great down low....my slant manometer is a 10x multiplier but that's not enough to really trust at low flow numbers where there is not much signal, I may try a digital.

The way I do it is I work on the development port until I think its as good as I can get it, then I match the others to it....litterally. I reflow port 1 anytime I'm ready to make real measurements on another port so I know the manometer readings match. Not ideal for a business in a hurry or trying to duplicate work that's out the door, but I only have 1 project to worry about, condensed build thread is here:
https://grassrootsmotorsports.com/forum ... 959/page1/

The port rework was pretty extreme but I think after about 100 hours on the flow bench all 12 cylinders are within 1% total error.