Don Terrill’s Speed-Talk

Yes, another exhaust back pressure question On another site a guy is testing Harley Davidson exhausts on a dyno. The engine is moderately tuned, not a race engine. Open pipes or minimal baffling produces a sag in the torque curve at lower RPM's. The guy doing the testing says there's not enough exhaust restriction....He adds more restrictive baffles and the torque dip disappears with a slight drop on top end..... I call it exhaust reversion, something my highly tuned 650 race Triumph has.......Perhaps Harleys have unique tuning problem...but... there are better ways to eliminate power curve dips than exhaust restriction?

about the only time a more restrictive exhaust equates to more torque is when the exhaust events are too early for a more 'flow-capable' exhaust system (header pipes and muffler as a system). the exhaust events may be too long or ECL too early for engine's intake parameters (flow and valve events), therefore the motor is down on cylinder pressure and torque and lacks low/mid range grunt.

the opposite argument could be made for a engine that has shorter and later exhaust events. An exhaust set-up with increased cross-section and flow capability can increase low/mid range end torque.

So it's a situation of mismatch parts...

To be honest, most people are disappointed when they spend their money to lose power or place it an rpm band they didn't want. But if you think about it. Its good to make a change for the worse because you can be assured that you were closer to optimum before, for the as-built combination. So, you atleast know the way you went was the wrong way.

Truckedup wrote:So it's a situation of mismatch parts...

It seems to me that a Harley bike would be pretty tough to get optimum on the exhaust because you only have two cylinders and the pipe would tend to act more like zoomies do on a V8. With one pipe per cylinder you have fewer pulses to work with and the pipe tends to become more in tune and more out of tune at various places across the rpm band of the engine. I'd say that what's happening with the restriction is that it's either changing the tuning of the pipe by making it behave less tuned. It's not as bad when its out of tune and its not as good when it is IN tune either. What does a Harley run like with a 2 into 1 exhaust setup it seems like that would work better than two straight pipes.

Single or twin cylinder engines are very sensitive to exaust pipe lenght and generally are very engine specific.

Commonly a open pipe situation lose torque due the shorter lenght compared to a full exaust, as the exaust system see the muffler as a longer pipe, torque may improve due the additional pipe lenght , not the muffler restriction itself.

Any details on the carb, and whether the "jetting" was changed when the baffling was changed ? A/F ratios might be revealing.

It is criminal to choose low end torque over the lives that would be saved if only the baffles were left on the shelf.

http://i.ebayimg.com/00/s/NTE5WDU4Mg==/$(KGrHqVHJEIFBnjJpLt7BQl8DtrGJ!~~60_35.JPG

i once did a test on the dyno where i put a small orifice in the tailpipe to see the effect of restriction. i put it as far back as possible so as minimize interference with wave tuning. the results showed that the restriction hurt everywhere from 2000-6500rpm albeit at low rpm it was a only 2% increasing to 7% at peak. was a dual 2.25" exhaust and orifice was 2 x 1.0". probably should have tested even smaller orifice to see a bigger effect.

i also disconnected the exhaust so no cats, resonator and muffler in the system and topend went up by the same amount as what the restrictor lost but now some bottom end torque was lost. the exhaust length was altered so that is where the loss probably comes from rather than lack of back pressure as the other test showed that more back pressure did not help.

had i removed the cats, resonator and mufller with same pipelengths best of both worlds probably, except for noise....

Newer HD's will respond a little differently, But...

My 80 inch Shovel I switched from standard drag pipes to 40 inch pipes and picked up a munch of bottom end torque, but lost too much top. So I cut the pipes down to 36 inches, regained top end at a slight loss of bottom.

Drilled and installed 1/4 bolt across the pipes 1 inch in the from the end. All Good Now.

CharlieB53 wrote:Newer HD's will respond a little differently, But...

My 80 inch Shovel I switched from standard drag pipes to 40 inch pipes and picked up a munch of bottom end torque, but lost too much top. So I cut the pipes down to 36 inches, regained top end at a slight loss of bottom.

Drilled and installed 1/4 bolt across the pipes 1 inch in the from the end. All Good Now.

When I'm racing at the ECTA track in Ohio I see many highly tuned Harleys with the bolt near the pipe opening. There's also no consensus on exhaust design, some have two into one ,some have two individual pipes. None are long pipes. Some two cylinder Brit bike guys have a small "lump" in the head pipe near the head port. This allows the gases to pass out but gives a restriction to exhaust pulses trying to return. This can help with reversion in some cases. Us Brit Bike guys call "reversion" the sag in low speed power from an over rich carburetion caused by exhaust scavenging pulses messing with carb mixture. I don't know if the problem on Harleys is lean or rich? But on the one mile standing start competition low speed response is not very important.

I always thought it was common knowledge that street engines and stock engines like a little back pressure for low end torque
I had a 454 bbc 9.25:1 compression O49 ovals stock, comp 280h cam 230-230 .520 lift 110lsa 106ilc
Edelbrock C400 dual plane intake, Holley 750 and 1.875" primary headers and dual 3" exhaust
I swapped to 1.75" primary headers and I was very surprised at how much torque I gained, i picked up over 2 tenths in the 1/4 mile from 12.65 to 12.43
Im sure this engine was not optimized and probably had too much exhaust to begin with but I was 15yo and i didnt know what i do now thats for sure

The 454 example with oval port heads and edelbrock C400 intake, with the 280/280,230/230,110 camshaft is going to be needing an exhaust flow capability in the 175 (28") cfm region. This is well within the reach of a 1 3/4 primary to a 3" exhaust system. A 1 7/8 primary to a 3" will typically have around 190 cfm, with the 3" pipes being somwhat of a restriction for that primary size. So yes, the math is showing that the motor was definitely a little too exhaust capable before the header swap.

Had the camshaft had a somewhat delayed exhaust events (narrowing the LSA to a 108), the header swap may have results in less performance.

Drag pipes are not street pipes.
Straight single pipes with cutoffs at a certain length are very tightly tuned for performing at certain rpms. If you tune them for high rpm hp, they often have a loss in the lower or midrange rpms, depending on how the out of phase wave plays into the timing.

The "bolt in the tailpipe" or other similar mods are wave-breakers, and this helps break up the wave so that it doesn't do so much harm in the lower rpms, but it also doesn't do as much good up top.
The anti-reversion cones and dams at the head outlet interfere with the returning wave at the head, as a similar thing in a different spot.

These are all "band aids" for bad exhaust tuning.
A properly made exhaust system for the application would not need any of that stuff.
Restrictive exhausts, or "back pressure" are not helpful. There is a diameter and length that will be appropriate for any application, and it is up to the engine builder to get it right.

another example of this is honda cx 500 bikes.The stock setup has the 2 individual pipes dumping into an expansion chamber (fooling the pipe into thinking it dumping into open air) .Then the mufflers come off of this expansion box, when the box is removed for individual pipes power usually goes way down. Functions like the 0 loss exhaust system David Vizard uses.

The one thing you have to watch with looking at other racer's stuff is there is really only a very small handful actually doing development. Most everyone else is just copying what they say without understanding the why.

I was always under the impression that adding 'backpressure' to the system to make it work better was a band-aid for the header length and/or diameter being wrong and/or cam timing events being wrong for the application. I don't see how backpressure could be a good thing with all other things being equal.