Actual gains from exhaust modifications

Backpressure is a stupid word.

When people say that cars "need backpressure" what they should be saying is that a properly tuned exhaust will provide tiny amounts of restriction at the peak in favor of significant increases in torque everywhere else. Its all a trade-off. You can't have a properly tuned scavenge event at 3000 rpms and not have modest restriction at 7000. Conversely, you can reduce the restriction at 7000 with a monster exhaust, but you give up most of your scavenging everywhere else.

A race-only car like a drag car spends all of its time at peak RPM. In that case, bigger is better if it makes an additional 2 hp. On a street car, I would gladly give up 2hp to gain 15 tq average in the rest of the RPM band.

Supercharged S197. Before has longtube headers, aftermarket performance cats, OEM 2.5" exhaust, performance axle-back mufflers (loud!) and a custom roadcourse tune for 93 octane. After has a new 3" custom exhaust and two pair of mufflers (quiet!) and a tune by the same guys that did the first one with the same parameters.

Here are dyno pulls from our stock motored 2011 Mustang GT. During it's competition career we upgraded the exhaust, headers, cold air intake, and then another cold air intake. During this time the engine remained stock from the throttle body to the exhaust port of the head.

Here's the numbers from the stock car when we first picked it up. Same longblock, IM and throttle body. Same dyno and same operator.

And here is the car retuned with a CAI only.

Add headers, no cats, custom exhaust and a custom tuned length CAI, still tuned by the same guy with the same parameters (93 octane, roadcourse use, wide safety margin).

curtis73 wrote: Backpressure is a stupid word. When people say that cars "need backpressure" what they should be saying is that a properly tuned exhaust will provide tiny amounts of restriction at the peak in favor of significant increases in torque everywhere else. Its all a trade-off. You can't have a properly tuned scavenge event at 3000 rpms and not have modest restriction at 7000. Conversely, you can reduce the restriction at 7000 with a monster exhaust, but you give up most of your scavenging everywhere else. A race-only car like a drag car spends all of its time at peak RPM. In that case, bigger is better if it makes an additional 2 hp. On a street car, I would gladly give up 2hp to gain 15 tq average in the rest of the RPM band.

Basically, once you get a foot or two past the collector, you're good. Make it as open as you can.

Here is some completely unscientific evidence. On my old commute there was a long gradual incline on the interstate. In my 90 jeep with a 258 I6 with stock exhaust about halfway up the hill I would have to downshift to fourth to maintain speed. After deleting the cat and replacing the muffler with a cheapo high flow oval muffler I was able to accelerate up the grade in fifth.

I saw a 30HP gain at the wheels in a TR8 with a Rover 4.6 by installing a good X pipe, and a vacuum pump on the engine. Don't know which mod accounted for what, but 297 to 327HP is substantial. FWIW another buddy has a TR8 with a 4.9 stroker and much more work done to it including programable FI. No vacuum pump, no X pipe, same 327 peak HP number.

I've been wondering about this lately. My TR6 project needs a completely new exhaust system. I'm planning on running a 6-3-1 tubular header with a 2.5" output flange. I could adapt it to the stock single 1 7/8" or double 1 3/4" systems, but want to try running 2.5" custom pipe all the way to a single muffler. I'm running a stock cam (FSP rules), so no crazy cam overlap or particularly high lift is involved. My concern is a loss of low/mid range torque. I've heard many claim that the stock single pipe system actually improves peak HP, but that may be with an otherwise 100% stock setup.

In an GRM article about a year ago there was mention of a company that would give you a "recipe" for headers for your application. IE: You fill out a form with the specifics of displacement,number of cylinders/arrangement, operating RPM, goal (low end torque, max hp gain, etc) and they send you back the specifics of how big to make what how long (at least that's how I understood it) for more money they'd make the headers for you.

Has anyone done anything like this? Is there some authoritative reading material on the subject so someone can make a go themselves? Does a program exist that's "plug and play"?

Hungary Bill wrote: In an GRM article about a year ago there was mention of a company that would give you a "recipe" for headers for your application. IE: You fill out a form with the specifics of displacement,number of cylinders/arrangement, operating RPM, goal (low end torque, max hp gain, etc) and they send you back the specifics of how big to make what how long (at least that's how I understood it) for more money they'd make the headers for you. Has anyone done anything like this? Is there some authoritative reading material on the subject so someone can make a go themselves? Does a program exist that's "plug and play"?

WHAT HE SAID! We would all benefit from that kind of knowledge, if it is correct.

Years ago, when I was messing with the RX-7, I recall being very disappointed with the gains from pulling the entire stock-ish exhaust and replacing it with a 3" downpipe, 3" exhaust, and a 3" Borla XR-1 race muffler. It wasn't until I pulled the stock airbox (duh!) and MAF sensor, opened up the turbo inlet, and installed the Haltech that I really woke up the beast.

So yeah, address the choke point. It might not be the exhaust.

I think the gains really depend on what you are fixing. I bought an 82 Z28 new, and several years later I built a decent 350 for it, but couldn't afford to upgrade the exhaust at that time. Cast iron log manifolds and a single 2 1/4 inch pipe made for a disappointing ride. Edelbrock headers and a 3" exhaust picked up about six tenths on the dragstrip.

Hungary Bill wrote: In an GRM article about a year ago there was mention of a company that would give you a "recipe" for headers for your application. IE: You fill out a form with the specifics of displacement,number of cylinders/arrangement, operating RPM, goal (low end torque, max hp gain, etc) and they send you back the specifics of how big to make what how long (at least that's how I understood it) for more money they'd make the headers for you. Has anyone done anything like this? Is there some authoritative reading material on the subject so someone can make a go themselves? Does a program exist that's "plug and play"?

Headers By Ed, I believe.
http://www.headersbyed.com/hc_buildbetter.htm

The calculations are pretty widely available in a number of pretty dry books, and of course there are online calculators of unknown provenance. It's not that mysterious, but you need to know quite a bit about the engine you're strapping it to.

When I did my high performance custom header, I started down that path. Then I realized that many of the small details were not making a big difference to the end result, and I was converging on a particular length. So I did an 80% job, cross-checked it against similar cars and built it.

Here's my header (blue) vs a Racing Beat unit (red). This is a straight A/B test, it was pretty early days in the engine's development. That dip in the lower rev range was characteristic of this individual engine, I never did completely figure it out.

http://targamiata.com/images_lrg/RB_vs_Keith_header.pdf

Oh, and here's what that header looked like. I had to package about 10' of tubing in a small space.

1988RedT2 wrote: Years ago, when I was messing with the RX-7, I recall being very disappointed with the gains from pulling the entire stock-ish exhaust and replacing it with a 3" downpipe, 3" exhaust, and a 3" Borla XR-1 race muffler. It wasn't until I pulled the stock airbox (duh!) and MAF sensor, opened up the turbo inlet, and installed the Haltech that I really woke up the beast. So yeah, address the choke point. It might not be the exhaust.

The Mazdaspeed Miata is like that. The stock intake prevents the turbo from making any more boost - to the point where the stock boost control just gives up at 5000 and lets the intake restriction limit things. I think Mazda did it to protect their warranty against owners with a boost controller made out of Home Depot parts.

Keith, looking at the dyno chart, would you say your header was "worse" than racing beats? Unless you got the colors backwards?

Burns is the best in the collector business and does a lot more work to get header design right for the application. Design is free with collector purchase.

We swapped a no name 2" collector to a burns and picked up torque everywhere and almost 10whp at redline. Thats JUST the collector with no other changes.

HiTempguy wrote: Keith, looking at the dyno chart, would you say your header was "worse" than racing beats? Unless you got the colors backwards?

You're right, I got the colors backwards. The chart is even clearly labeled. Mine is "better".

It really depends how bad the original pieces are too.

On a Friend's 65 dart, The PO had installed a really crappy dual exhaust setup and instead of notching the transmission cross member to run it right, they pinched a tube down to probably 1.5" to make the bend and not hit anything. So when we added the Doug's headers and a smooth flowing dual exhaust, it really woke that 318 he had in there up. It was really a runner when he added in the mild cam.

The PO's excuse for not doing it right? They didn't want to cut up the transmission mount.

In reply to bmw88rider:We've been talking about choke points. It sounds like you found one!

it doesn't hurt to improve the exhaust at the beginning.....

Also helps have BIG displacement, or a ton of boost. I have picked up 20-25hp with just bolt-on exhaust with my last viper. Now the stock unit is terrible and they did just about every trick to get it under the dB rating I am sure.

I'm going through the planning phase on the exhaust portion of a motor swapped car now. There is a lot of good information out there about the basics. But something else I'm finding is that it really is motor-specific, so I need to test my general conclusions against what other people experienced in actual implementation. Sometimes the CW doesn't apply, for unexpected reasons. This is one reason I like to use popular engines for swaps. Lots of people have already tried lots of things out for you.

Cars tend to run leaner when the exhaust is opened up. A lot of stock tunes run richer than ideal for max power. So, gains due to the exhaust might also be partly due to running leaner than you did with the original exhaust.

This makes it hard to determine what only the exhaust may be contributing, because very few people bother to lean out the stock system first for an accurate comparison.

GM' U Bend: http://www.zzperformance.com/blog/u-bend-replacement-test-by-jim-wierzbicki/

and a replacement with a high flow cat:

http://shop.zzperformance.com/store/p/249-ZZP-Downpipes.aspx

I've replaced exhaust probably on every car I have owned, at least with a better muffler but often the bigger gains are up front - depending on the vehicle - either the ypipe or front pipe area or the cat.

My wifes 2000 Sunfire at about 100K I replaced the cat with a Magnaflow unit (downpipe and cat together bolt on). The car picked up 2mpg in town and 5mpg on the highway. One could state the factory unit was clogged, but it was getting the same mpg as it did when we bought it with 8K.

Back when V8's had restrictive manifolds, ypipes, etc - when swapping to headers and the proper exhaust diameter -almost always required re tuning of the carb

i tend to do exhaust at the same time as an intake but when replacing the crap factory parts on the sundance the bigger down pipe and mid pipe made a noticeable difference in daily driving!

curtis73 wrote: Backpressure is a stupid word. When people say that cars "need backpressure" what they should be saying is that a properly tuned exhaust will provide tiny amounts of restriction at the peak in favor of significant increases in torque everywhere else.

No restriction is ever good.

People confuse "small diameter" with "restriction".

What is important is velocity. You can have high velocity pipes that have no backpressure. Conversely, if you go too big on pipes, velocity drops, so exhaust inertia drops, so the exhaust flow doesn't pull the exhaust out of the cylinders, so it takes more effort to get the exhaust out of the engine so backpressure goes UP with large exhausts until the engine is flowing enough air to get the velocity up.

Backpressure is never ever good. Ever.