Intake manifold flow - Correct way to test?

[2seater]

Not quite sure how to pose this question, but here goes. I am looking for a good way to test air flow through a fwd V6 intake manifold. This not connected to a head, simply open ports. I have tried measuring the flow through each port, one at a time, with a handheld pitot tube, adding the results together and compared that to the measured total flow. For reference, all ports are open and flowing at the same time. Results are unsatisfactory, the total is close but always less than the measured incoming flow, plus the flow is highly unbalanced. My assumption is the process of measuring causes a restriction that redirects the flow somewhat. Six identical pitot tubes connected to all ports would probably be more correct, but that is a daunting task. My second idea and the source of this question, would simply reading the pressure available at all ports be a valid test? Relatively simple to construct a six port manometer, kind of like a water version of carb stix, and plug them into the port for the injector. I have given this a rough try, and it does give an indication, which looks like a graphic equalizer in operation. The second part of the question is, should the ports be restricted? The rough test was with plates over the ports with an area of 25% of the total port area (.70" diameter opening). Best guess to simulate a 180 degree cam @ .050" lift. The other thought would be to allow each port 1/6 of the area of the area of the valve body, which would be a little larger than the first test, about .80" diameter. Surprisingly, the rough test with restricted ports showed fairly even air flow, at several different incoming flow rates. The total estimated air flow will be 360-400 cfm at 7 psi of boost in finished form. Any thoughts greatly appreciated.

[Tony]

I doubt if anything really meaningful can be assumed by just testing a bare intake manifold runner in isolation. The reasons being that you cannot test it under any sort of realistic operating condition that corresponds to a real running engine.

I read somewhere that each intake runner when fitted to a cylinder head should not restrict the maximum full lift flow more than ten percent. Someting much less than ten percent would be better. This is compared to just the bare cylinder head when fitted with a properly radiused entry.

Steady state flow is not really a good indication for a multi cylinder manifold. The pulsing in the runners and plenum can cause cylinder robbing and very uneven air (mixture) distribution, especially in high boost forced induction engines. Many production engines have strange shapes, baffles, steps, humps and other features in the plenum or the runner entries of particular cylinders. This is done to even out the air distribution under actual pulsing and high velocity operating conditions.

By removing these features to gain additional flow, and maybe doubling or tripling the airflow velocity into the plenum (with a turbo), you can get yourself into some real trouble. Exhaust gas temperatures on the dyno are what you should really be looking at to get the cylinder to cylinder airflows even.

A production manifold designed for 200Hp, will quite likely work perfectly well at 250Hp, but at 700Hp, the cylinder to cylinder air distribution may be truly horrible and need to be corrected.

At one stage I had seriously thought of hooking up an intake manifold and cylinder head to my flowbench, and driving the overhead inlet camshaft with an electric motor to various Rpm. Airflow to each cylinder could be measured with multiple pitot tubes and multiple manometers. The idea was to try to simulate real engine running conditions as closely as possible.

After discussing this idea with someone that has actually tried it, I was told not to bother. Apparantly piston motion, fluctuating temperatures, and exhaust reversion in a real engine running under full load will significantly change the operating conditions. Just opening and closing an intake valve to a fixed depression creates nothing like the violence and fury needed to come anywhere near close to real running conditions.

It seems that doing it for real on a dyno is the only reliable way, no short cuts unfortunately.

[larrycavan]

Tony raised some good points on your problem. I wouldn't disagree with his point of view. Still, all things considered, I think the flowbench can be used to obtain useful information in what you're after.

You certainlly can't duplicate every effect the intake manifold would have on a live engine but you can at least get some local velocity measurements in the runners [at least where it's reachable] and get a picture of how the total flow is influenced with the manifold attached to the head.

Just as you cannot encompass all information into single readings, you should never dismiss single readings as being non-effective on the total operation.

What's good to do is to establish testing paramaters and to take extensive notes on the results. What today may seem insignificant may in time prove to be useful. Trends show up over time and trends without supporting data are difficult to understand.

Perhaps you'll discover that a certain range of FPS at a given test pressure proves to be beneficial on any particular head. It's information you won't have unless you obtain it and document it.

It may seem fair to say that most speed secrets / discoveries for any given engine have been pretty well covered and exhausted by the "experts". Maybe YES.....Maybe NO.

You may discover that certain runners in a given manifold are faster or slower than others in any or all sections of the runner. That's useful information.

A wise man uses every tool he has available to obtain information he can apply to solving a problem. The information may seem insignificant to others but to the man that gathers it and is persistant in his efforts, it can prove to be otherwise.

JMO

Larry C

[2seater]

Thanks fellas. I appreciate the input. I agree it will almost impossible for the home builder to simulate an actual running engine, and that would be the ultimate goal. It definitely is a dynamic situation, but at high rpm, the events occur so often it approaches steady state more closely. Still can't simulate the wave actions and ram effects from valves moving and runner reversion. Larger plenums are generally used on the really high horsepower setups, my logic says that is to slow the high velocity air flow from the throttle body to allow it to actually turn into the closer runners. I do have a reasonable idea of what the total air flow will be at the target boost/hp range although I cannot reach the pressures the manifold will be seeing at those flow rates. I do not know if what I am doing will yield real data that can be used for meaningful changes. Gathering data can do no harm, and may show some sort of pattern. I just don't know until I try. The one thing I do know, at least from the turbo Buick clubs is that the end cylinders, furthest from the inlet, do tend to run lean, pretty much in line with what I have been finding. Real world with individual EGT probes for the exhaust will tell the tale as you have suggested. Thanks again.

[larrycavan]

[2seater]

Thanks for the suggestion. It does make sense to couple the manifold to a head to get the real information on total flow actually available, not just a guess. I am flowing the manifold in the correct direction, blow mode through the stock MAF and throttle body. Fixturing it will be interesting, but should make for a good winter project.

[Hopper]

[Hopper]

I forgot to hit track topic.

[2seater]

I will try the individual port flow as suggested, that is relatively straightforward. I do have a couple of questions. Do you measure the actual pressure inside the manifold plenum to compare to the depression of the bench? I'm not sure if it is important, or maybe critical, and I imagine it would only indicate the pressure loss through the MAF/throttle body combination? Mine is a turbocharged application to an N/A engine. Is your plenum larger, perhaps tapered? Very curious as to your dyno results and I do appreciate the input.

[Hopper]

I just tested with the bench depression.
The new intake has a smaller plenum, with larger runners.

[2seater]

[larrycavan]

There's a good thread over on speedtalk that you guys might be interested in

[2seater]

[DaveMcLain]

Here's a picture of a fixture that I built for intake testing on my flowbench, nothing fancy. I just use it with the bench set on exhaust with the carb body of choice mounted to the bench using the fixture. Then I block all but the runner I'm testing with duct tape. Depending on the quality of the tape you will probably have to limit your testing to about 8 to 10 inches of water.

I think it works good and it makes it easy to find the good and bad runners. I don't have a lot of experience with all of this but to me figuring out how the intake interacts with the cylinder head is one of the toughest porting problems. Maybe this will be of help to you.

[larrycavan]

There ya go.... a setup like Dave's will help a bunch. I got some spring loades squeeze claps from a hardware store that might allow for clamping a piece of plate over the unused ports and allow you to test at higher pressures.