Blow through intake manifold flow testing

[Scotthatch]

Charlie, I tried the same thing as you several years ago on 3 or 4 intakes and made notes on each test. Then I turned them around with the air being pulled in as it normally would be. Every test in the blow through method gave higher numbers. I think this is due to the way the air acts at the runner entrance areas, but I am not sure.
A simple way to block the ports is to use a straight 1x4" wooden plate with a piece of high density foam and 2 quick clamps and make it long enough to cover any head you may test
I have made an adapter for the bench but I don't have the computer skills to send pictures, sorry.

Good post


Do not use duck tape ... air flows through it and will throw your number off ...

If you are doing a quick test and don't have a plate to block the other runners use high adhesion clear packing tape or the aluminum tape used for hvac duct work....

[turbo2256b]

Charlie, I tried the same thing as you several years ago on 3 or 4 intakes and made notes on each test. Then I turned them around with the air being pulled in as it normally would be. Every test in the blow through method gave higher numbers. I think this is due to the way the air acts at the runner entrance areas, but I am not sure.
A simple way to block the ports is to use a straight 1x4" wooden plate with a piece of high density foam and 2 quick clamps and make it long enough to cover any head you may test
I have made an adapter for the bench but I don't have the computer skills to send pictures, sorry.

This is the same thing we found out higher numbers blowing through than sucking. its because of a pressure differential one pushing one sucking.
also found best if drawing through a cylinder head. the left side of the intake needs to be ported a bit different than the right due to push rod location. Most intakes reduce the quite a bit once bolted to the head. Getting within 20 cfm less than the head flows is about as good as it gets installing an intake and flow testing. Except for a sheet metal intake which could match flow and well done could even increase air flow through the head.
Example heads on my car flow 235 CFM at .500 lift. Intake installed that i spent 2 weeks bolted to the heads flow testing and porting flows 215 CFM on each port it is mated to in the engine assembly.

[gmrocket]

How about tightly wadding shop rags into seven of the ports from the plenum end? The worst that could happen would be turning the manifold into a tee shirt cannon...

I used some plate aluminum over the tape and clamped them to the manifold. Looked ugly but worked well. I would not want the wadded rags to change the flow in the plenum. Thanks, Charlie

Your fooling yourself if you think blowing through only one manifold port, with the others blocked off, gives you an accurate indication of how it flows..

Why not port the entry to that runner so it flows even more? Of course not, because it would make the entrance of the port beside it worse. That proves it's a bogus test

[groberts101]

This is the same thing we found out higher numbers blowing through than sucking. its because of a pressure differential one pushing one sucking.
also found best if drawing through a cylinder head. the left side of the intake needs to be ported a bit different than the right due to push rod location. Most intakes reduce the quite a bit once bolted to the head. Getting within 20 cfm less than the head flows is about as good as it gets installing an intake and flow testing. Except for a sheet metal intake which could match flow and well done could even increase air flow through the head.
Example heads on my car flow 235 CFM at .500 lift. Intake installed that i spent 2 weeks bolted to the heads flow testing and porting flows 215 CFM on each port it is mated to in the engine assembly.

Your fooling yourself if you think blowing through only one manifold port, with the others blocked off, gives you an accurate indication of how it flows..

Why not port the entry to that runner so it flows even more? Of course not, because it would make the entrance of the port beside it worse. That proves it's a bogus test

Is it possible due to the different pressure differentials seen on either side of the intake manifold during flow testing, that the manifold may actually achieve higher flow potential once pressurized at atmospheric pressures?

IOW's.. if said cylinder head flows xxx amount of air on the bench and achieves anywhere near its theoretical limit for final power production.. but the manifolds flow numbers choke that number down below what is theoretically possible.. wouldn't that be proof that the flowbench number achieved through the blowthrough testing maybe has more relevance to actual power production?

[turbo2256b]

if one used calcs to figure HP based on cylinder head flow vs flow sucking with the intake mounted, the intake mounted to the head flow will be more accurate.

Blow through would flow figures would be way off

[Scotthatch]

How about tightly wadding shop rags into seven of the ports from the plenum end? The worst that could happen would be turning the manifold into a tee shirt cannon...

I used some plate aluminum over the tape and clamped them to the manifold. Looked ugly but worked well. I would not want the wadded rags to change the flow in the plenum. Thanks, Charlie

Your fooling yourself if you think blowing through only one manifold port, with the others blocked off, gives you an accurate indication of how it flows..

Why not port the entry to that runner so it flows even more? Of course not, because it would make the entrance of the port beside it worse. That proves it's a bogus test

So we need to flow the whole intake at once and just divide by the numbers of runners ? That would take a big bench ....

Mostly I started doing this to see how far off it was ....

If we take this thinking another way does it mean we should never test on the bench without a intake attached because it's a bad number ?

[groberts101]

if one used calcs to figure HP based on cylinder head flow vs flow sucking with the intake mounted, the intake mounted to the head flow will be more accurate.

Blow through would flow figures would be way off

Sure, I get that and assume why most everyone does it that way when balancing cylinder to cylinder power and high resolution results are paramount. But what is considered "way off"?

I was just more curious if the actual manifold flow measured on a bench with the head attached.. exactly correlated to specific output on the running engine. It just seems, at least on the surface.. if by measuring them as currently being done is choking flow to the cylinder heads(and pushing that number further downwards.. and then when using those actual flow numbers generated with it attached to the heads.. is ultimately generating an excessively high VE factor for that particular final dry flow number.

[turbo2256b]

You want to block off the runners not being flow tested. It acts like there is an infinite plenum. A running engine is stealing flow from other runners. closing up the runners not being flowed gives a closer idea as to what actually happens in the engine not perfect but a lot closer.

[BILL-C]

For getting "quickie" flow numbers for an intake manifold, i prefer to use clay radius, or better yet,same plastic flow guide used for head testing, and suck through intake. Your curves are so gradual in the middle of the runners that the flow direction doesn't matter much there. The tape stays on much better also. Also helps to speed up testing so you can spend more time on more important measuring, like cross sectional areas. Basically, the bigger you make the intake, the better it flows on the bench. This IS NOT what makes all cars fast. I believe "line of sight" and getting the cross sectional areas the right size for the intended application are the 2 most important things to spend your valuable time on when porting intakes.

Bill, this reads like you install the radius on the intake and then flow air from the runner into the plenum. Is that what you do? Thanks, Charlie

Yes

[Carnut1]

So flowing from the runner to the plenum shows you what? Is it usable info? Please explain. Thanks, Charlie

[BILL-C]

I think that sucking through the intake give a little better info on what's going on with the short sides of curves and roof of runners. That line of sight deal that i mentioned.

[gmrocket]

I used some plate aluminum over the tape and clamped them to the manifold. Looked ugly but worked well. I would not want the wadded rags to change the flow in the plenum. Thanks, Charlie

Your fooling yourself if you think blowing through only one manifold port, with the others blocked off, gives you an accurate indication of how it flows..

Why not port the entry to that runner so it flows even more? Of course not, because it would make the entrance of the port beside it worse. That proves it's a bogus test

So we need to flow the whole intake at once and just divide by the numbers of runners ? That would take a big bench ....

Mostly I started doing this to see how far off it was ....

If we take this thinking another way does it mean we should never test on the bench without a intake attached because it's a bad number ?

flow it with the max your bench can muster with all the ports open,,and convert to 28" if you want. then block off one at a time and note the difference. do that for all 8 and see where your at for there being an equal loss of flow through each runner. i think this test is more about equalizing flow when its a blow thru.

[user-30257]

Or you could save yourself a bunch of time, make the ports as straight as possible. Try not having huge differences in CSA through the port. Chose a length that suits rpm. And let it eat. You can tune the carb if need be or nitrous/ fuel jets..etc heck even individual cylinder timing.

If you are worried about a casting not being right.

You started with the wrong part. And manipulating it any more than you have to, will only show you a fabricated manifold would have been cheaper, than the labor on and off the bench..

Not to mention why not flow through the cylinder with the heads and intake pistons and valves at the correct phase? Everything bolted together. Wouldn't that tell you more?

[GARY C]

Or you could save yourself a bunch of time, make the ports as straight as possible. Try not having huge differences in CSA through the port. Chose a length that suits rpm. And let it eat. You can tune the carb if need be or nitrous/ fuel jets..etc heck even individual cylinder timing.

If you are worried about a casting not being right.

You started with the wrong part. And manipulating it any more than you have to, will only show you a fabricated manifold would have been cheaper, than the labor on and off the bench..

Not to mention why not flow through the cylinder with the heads and intake pistons and valves at the correct phase? Everything bolted together. Wouldn't that tell you more?

That is my thought, considering how much difference there is in an intake alone vs a head alone vs the 2 combined, what # would you use intake vs head for a combined combo?
I could see trying to equalize a duel plane by itself but without testing what is needed then how do you know whats needed?
But even flowing the package together you still need actual engine testing to confirm.

[gmrocket]

The current most common flow bench test pressure is 28" H2O/2.05" Hg./1.00 psi. Standard sea level air pressure is 408" H2O/29.92" Hg./14.7 psi. Air flows as a result of pressure differential, so testing at sea level, when sucking through a manifold runner the plenum entry pressure is 408" H2O/29.92" Hg. and the runner end pressure is 380" H2O/27.87" Hg., a delta of 28" H2O/2.05" Hg.

Now let's move the bench to a higher altitude, one at which standard pressure is 2.05" Hg. or 28" H2O lower, per this chart: https://www.avs.org/AVS/files/c7/c7edae ... 54f87b9e.p , 27.82" Hg. @ ~ 2000 ft. This time though, we will blow through the plenum, so the inlet pressure is now 27.82 + 2.05 or 28.87" Hg./408" H2O and the runner end pressure is 28.87" Hg./380" H2O, a delta of 28" H2O.

Hmmm... The inlet pressures are the same for both tests and the outlet pressures are the same, so the delta is the same. I wonder if perhaps the flow will also be the same?

Wouldn't the flow bias on a blow thru test be more along the floor and long wall of the port? Compared to the short side walls and roof of a suck thru set up?

so wouldn't the flow be different,, more or less?