Intake manifold flow - Correct way to test?

[cspeier]

Hey Guys, haven't posted for awhile. I have a good friend who works for a major head porting company. He gave me a few pointers on a manifold I was doing. His exact words to me were if the manifold drops the port more than 5 cfm, it needs work. If the manifold makes the head go backwards, it needs work.

Each runner seperate, the others taped off. Manifold flowed thru the head. I worked on the radius and runner length so much, my manifold actually drops less than 3 cfm. I also had a few runners that flow more air. This is due to the air being straightened out thru a long runner.

Just my experience.

Chad

[Tony]

Still not convinced I am afraid......

We all know that a properly developed set of tuned exhaust headers with correct diameter and length, feeding into an optimally designed collector can pick up really big power on a normally aspirated race engine.

So if I bolt on any old pipe over my exhaust port, and as long as the flow numbers on my flow bench show all the exhaust runners are flowing well, I know the exhaust system is as good as it can possibly be ? I really doubt it.

Same with whole induction systems. Steady state flow at low air velocity on an airflow bench will give no real indication of air distribution or cylinder filling under dynamic conditions.

An airflow bench may easily "prove" that a pair of very large four barrel carburettors would be best on a small four cylinder engine. Because the flow numbers prove it.

A flow bench is certainly a very useful tool. But it is not the complete answer to every problem, it can very easily mislead the unwary or the inexperienced.

[cspeier]

Funny you should mention exhaust flow. I know for a fact some of the very best head porters in the country currently don't care what the exhaust flows. It's all exit velocity and exit size vs throat. I also know that some of the Pro Stock intake ports are currently being developed that way. Velocity only!!!

The smart guys have found that a flowbench cannot duplicate a live engine. However, from experience velocity can be the most useful and most accurate measuring tool.

IMO, the flow bench is becomming one thing only. It's the lifeblood for your pitot tube!!! VELOCITY!!!! It's all about velocity!

Why do you think it's critical that a head keeps flowing to 1.000 or beyond?

[larrycavan]

[2seater]

I wish I could find a carb style manifold to fit my engine, but it doesn't exist to the best of my knowledge. Mine is the typical fwd fuel injected manifold where the air enters from one end of the plenum. I do know, from seeing it demonstrated on a similar manifold design, and connected to both heads, that the air flow does want to travel in the straightest path to the far cylinders, there is simply more air flow out the far ports, than the close ones. This is not scientific proof of how the actual engine operates, but it does indicate a tendency for the air to pile up at the far end at high flow. Testing the manifold divorced from the heads may not be entirely accurate, but the head is a relatively controlled environment where it allows the air in from a single direction, and while I am not optimizing the head in this case, the flow capability should be relatively similar if the air supply path to each is equal. That is my goal at this point; to optimize the supply to the head. My setup looks very similar to the carb'd manifold in the photo, except it stands on end with the throttle body connected to the bench. I have not tried the single port testing yet, but I intend to. It will be intersting to see if the same pattern is repeated as with simultaneous flow. I thank you all for the input.

[DaveMcLain]

It still seems possible to me that you could devise some sort of fixture that would allow you to flow test the intake in the forward direction on the bench. How about a plate that allows you to hook up a very large vac. sweeper hose, like a dust colletor uses and then another adapter at the other end to fit the intake inlet.

[cspeier]

True Dave, although I don't think it matters. Start thinking in terms of the runner of the manifold and cylinder head as ONE. Basically the plenum to the seat. It makes a major difference in how the flow get's to the chamber depending on the shape and design. You could probably tune a manifold to flow the same on each runner, but when you put it on the head all the work is lost. It's best to look at it from the perspective as the manifold/head is one.

[larrycavan]

I agree with Chad on part of this issue, regarding thinking of the flow as an assembly. It is one and as such, it can have far different dynamics than when each component is treated individually.

The unfortunate thing with having an intake manifold with a common plenum is that it's dam near impossible to get a full velocity map of the entire intake tract assembled.

Things can change dramatically with the full intake track installed. Case in point: Port mapping certain motorcycle heads will give you specific local velocities that will mislead you. Once the throttle body is attached with the rubber intake boot, local velocities can change to the tune of 100FPS. I'm not talking down the centerline of the port. I'm talking about nearer the port walls.

If flowing the head by itself, one could be misled into thinking the port was too slow and might be tempted to start filling it in. That would be a mistake.

Sections that measure low when flowing just the head will spark to life when tested as an assembly.

I also agree with Dave and think that in order to investigate the dynamics fully for the assembly, it should be flowed in the operational direction of flow in order to give you a true picture of velocity changes and the rest of the dynamics of the assembly.

As the flow winds it's way though the entire intake tract assembly, it encounters bends and area transitions. Each of those will have an effect on fuel distribution and atomization. The flow pattern, the surface finish, obstructions such as valve guides, etc., will all influence fuel distribution as well as flow.

Any opportunity to improve fuel atomization should be taken advantage of. Before you can improve it, you have to discover what it's actually doing, where its going, what it's bouncing off....

There's not always time to discover every last hidden opportunity lurking inside every cylinderhead and port but with each one you at least attempt to learn from you gain knowledge and understanding that will stay with you. In turn, you will apply that knowledge [even if in bits and pieces] to other porting work.

JMO

Larry C

[2seater]

I agree these are valid opinions. I would think the opposite would be true for head work. It isn't completely valid without the intake tract leading up to it? I know from testing air boxes and things like it that there can be very large differences in the flow rate if just the parts or the whole assembly is tested. I am flowing in the correct direction, blowing through the throttle body, but in some ways, sucking through the head would in some ways be easier, as well as better. More to think about

[DaveMcLain]

I belive what you find with the intake attached is that instead of approaching the port straight on like the air does when flowing just the head is that with the intake attached the air tends to approach the head opening from an angle induced by the shape of the intake runners. I would say the best approach would be to do some testing each way and see if you can correct some of the influence the intake manifold has on the flow numbers. Tuning of the intake, dynamics is a whole other issue, but I do think that getting it to work with the head on the flow bench is important. It's certainly suprised me how bad some intake designs really are...

[2seater]

Yes, I noticed the flow was certainly different from various ports. Some biased to the roof, others on the floor or relatively even across the whole thing. This was using simultaneous flow from all ports, which no doubt will help bias the flow by itself. This was true no matter what the flow rate. The other thing I noticed was when the flow was restricted at all ports with an orifice with 25% of the area of the port, the flow became almost completely even with no other changes. This did reduce the flow, but other than that it illustrated clearly that what is attached to the outlet does make a large difference. It certainly isn't a simple situation.