Exhaust manifold flowtest results

[Directconnection]

I have a couple pages of the data written down from pages of flowtests, but I will give a rundown on just the flow gains/losses in cfm with the manifold bolted to the head, and at .500" lift. Also, I bolted the manifold down to exhaust runner #2 on the bone stock g-head and simply used the same head runner for all manifold runner tests (and blocking off the other runners of course) this way there was no variance of different head ports being used (and less math, too). Used the same exhaust head port for all tests. I also had a 3" spacer between the headstand and head for exhaust manifold clearance.

Stock G-head *alone* at .500" lift: 127.7 cfm

Stock exhaust manifold:
#1) 1.4 #2) -17.2 #3) 7.4 #4) 6.9

Vendor A ported stocker:
#1) 1.7 #2) -17.7 #3) 6.9 #4) 6.5

Moparzrule ported stocker:
#1) 5.1 #2) -15.0 #3) 9.8 #4) 20.5

Vendor C ported stocker:
#1) -.3 #2) -16.7 #3) 9.3 #4) 4.3

Vendor A log header:
#1) 3.1 #2) -5.8 #3) 1.3 #4) 11.1

Vendor B log header:
#1) 3.9 #2) -3.5 #3)no test #4) 10.9

Vendor C semi equal length header:
#1) -1.3 #2) -2.8 #3) -3.7 #4) 3.2

Vendor B *LOG HEADER* PORTED by Steve Mercier:
#1) 11.1 #2) 10.5 #3) 17.2 #4) 12.7



I also performed another flowtest on these manifolds on a SF-300 at 10",28" and 45" but never got around to crunching the #s and also had a few test situations with 45" creating so much heat that the tape on the other ports would vibrate off on some and I tried my best to keep an eye on. Doesn't appear that at 45" that they flow alot more air as I expected, but it does appear that they do flow more, especially the larger volume ports.

A 1 cfm gain over what the head flows is a pretty big gain to be asking of, so take these small flow increments in terms of more like a stock head that flows 140cfm, and a big valve head that flows 200+

[BadAssPerformance]

Vendor C semi equal length header:

Nice study and information Steve!

semi?

[Directconnection]

Nice study and information Steve!

semi?

Hehe... I say semi because I never measured to confirm it, but it does look close enough

[nomadman2001]

I'm confused. The numbers are all over the place , can you explain them a little for me?

[Directconnection]

I'm confused. The numbers are all over the place , can you explain them a little for me?

The #s aren't all over the place, but maybe it's a bit confusing for some as they are tight knit.

I 1st flowed a stock g-head exhaust port which happened to be cylinder #3 and was 127.7 cfm. I now had my baseline. Then, I put on the exhaust/intake gasket, put heavy tape on runners 2,3 and 4 on the exhaust manifold of choice for the test and bolted the open #1 runner to the head. Now, for the stocker, as you can see, flow picked up to the tune of 129.1 which was a *GAIN* of 1.4 cfm. Then, peeled off the tape on runner #2 and put new tape on #1 and tested #2 which lost 17.2.


See the pattern?

[Frank]

Steve,
I dont think he is confused by your numbers and your method, but rather the reason why they do that.


Jim,
It has to do with header construction. Our flanges are placed such that the turbo actually fits. When you have to meet those demands, your runners are not equal and can very flow quite a bit.

[moparzrule]

I take credit for the Vendor B ported stocker. Since steve has given me the results a little bit ago before this post I've calmed down the #4 runner porting to make it flow a little more even to the rest of the ports. I also spend a little more time in runner 2 to make it flow better as it is hands down the worst port in our stock manifolds. I've learned a lot from this test, and I want to thank steve very much for helping me improve my manifolds even more than what they are.

[Directconnection]

I will change vendor B to your moniker.

Also, I edited my ported example as it was Vendor B, but Vendor B ported LOG header, not a ported moparzrule's manifold by me .

[SwiftTech]

did you test a tbi header?

[Ondonti]

did you test a tbi header?

You cant really use the flow numbers on individual runner manifolds to compare how well they will perform against a log.

Higher flow numbers on this test dont = more hp but you can notice that among the log style, some of them really suck with cylinder to cylinder flow so some people are trying to balance that out.

TBI header might flow less then anything else tested but that doesnt mean its going to lose power. The method of testing cant be used to compare log vs individual runner manifolds IMO but it does show how the individual is pretty balanced and it took work to fix a log manifold to balance it out (so maybe the log without porting is not such a great idea etc) The stock log headers still do a much better job of helping cylinder #2 flow.

[SL#189]

So an optimal header/manifold would have equal flow through all runners independent of actual flow quantity (with the understanding that more flow is better)? Is it possible to do this with our stock manifolds?

[Directconnection]

I could make a stocker flow close to the log header, but it would require welding, and I wouldn't trust it after that.

Also.... the stock manifold is a log manifold and much like the log headers except for lesser volume.

Lastly, an equal length or individual runner manifold has it's advantages. But... before we should care so much about tuning and it's effect of the pulses drawing other cylinders via scavenging during overlap (which may not be as critical as NA since we have little overlap and huge backpressure...but that's another story on it's own...) ....we shouldn't overlook that the meat and potatoes of the actual job the manifold has: and that's to move exhaust gasses through the port, 1st and foremost. Efficient engineering via proper tuning is the icing on the cake... and proably pretty thick, as well.

.

[tvanlant]

Looks like the log headers flow much more evenly than any of the stockers, but still very nice numbers on Matt's.

Have the other vendors requested that this test be anonymous?

[Directconnection]

Have the other vendors requested that this test be anonymous?

No, it was my doing. Also, most of them were collected samples, not vendor donated. Hint about the log headers...both are not being made anymore. So, it is not a TU that I tested.

[johnl]

Seeing how much better #4 flows in this "static" measurement of flow by individual header port test, where the inertia and flow of cylinder #s 1,2 & 3 are not working against #4, makes me wonder whether #4 needs the extra flow to overcome the other 3. In a "dynamic" sense it is 3 against 1, pushing uphill.

It's another one of those - "if only we had back to back dyno test" - deals.

[BadAssPerformance]

BTW,

Vendor C semi equal length header:

is my prototype... "semi" cuz Steve did not measure them, but the runners are within .25" of each other...

In a static bench measurement it flows worse than a stocker, yet much more balanced. At 7000 rpm I think it works just a little better... dyno needed to prove this.

[cordes]

BTW, is my prototype... "semi" cuz Steve did not measure them, but the runners are within .25" of each other...

In a static bench measurement it flows worse than a stocker, yet much more balanced. At 7000 rpm I think it works just a little better... dyno needed to prove this.

Do you think that the added bends needed to make it equal length are what causes the reduction in flow?

[BadAssPerformance]

The pipes are just slightly larger cross sectional area than the ports, so adding length of a similar cross section (even without bends) adds restriction to flow. Velocity is maintained better tho...

[moparzrule]

Seeing how much better #4 flows in this "static" measurement of flow by individual header port test, where the inertia and flow of cylinder #s 1,2 & 3 are not working against #4, makes me wonder whether #4 needs the extra flow to overcome the other 3. In a "dynamic" sense it is 3 against 1, pushing uphill.

It's another one of those - "if only we had back to back dyno test" - deals.

That was excatly the same point I said to Steve when we were first discussion this testing VIA PM's and he said he'd be testing each runner individually. But we concluding that testing all 4 at once simply couldn't be done on the flowbench plus it would take 10 times as much work to do it with all the fabrication needed so I wasn't expecting steve to do it that way. He's done enough IMO, this took a lot of work and he did it all for free so you won't hear complaints from me!
I already knew my manifold flowed better than and FM unit because it was my own personal experience and I had a friend do one of the old leaf blower test on mine against FM against a stocker. But that was flowing all four at once and for that reason he had to go backwards through the manifold. The test was just for kicks, but mine kicked FM's azz all over the place. But in my real world testing, I had 200-300 RPM faster spoolup and a definate seat of the pants upper RPM gain. Before that I never knew the manifold made such a difference. It might have been amplified because I was running close 300 WHP, perhaps a closer to stock car wouldn't have felt as much of a difference. But no other changes were made besides the manifold swap. Unfortunitely I never got to re-dyno because my car got totaled about 2 weeks after I changed the manifold.

[MiniMopar]

I have one of those leaf blowers that can double as a giant leaf sucker. If you hooked that up from the turbo end and measured negative inches of water at the turbo outlet, you could do a quick comparison while flowing all four runners. Not a scientific test for sure, but could test overall flow differences.