Exhaust pipe diameters

Something which interests me is the diference in the exhaust pipe diameters which are used on varous bikes. I have always has a preference for small diameter pipes. My triton had small diameter pipes which were originally supplied as part of the 1953 factory racing kit. At that time all Triumph road bikes had large diameter pipes. I notethat the Steve Many two into one pipes are large diameter whereas I use a smaller diameter. Also some Manxes have been used with smaller diameter pipes than others in the same period.
I suggest that exhaust pipes are more about standing sound waves and mass transfer of gases then about flow. I found this article interesting :

http://www.fonema.se/qpipe/qpipe.htm

acotrel said:

I suggest that exhaust pipes are more about standing sound waves and mass transfer of gases then about flow. I found this article interesting :

http://www.fonema.se/qpipe/qpipe.htm

Click to expand...

Agree!

If memory serves me correctly, a Honda model offered in the mid to late sixties. had a small diameter pipe inside a larger tube. Honda knew the smaller diameter offered better performance, but the American market looked askance on weeny pipes. The pipe in a pipe satisfied both performance and the market idiosyncrasies.

Slick

1 1/2 pipes (stock style) on my Commando produced less torque and less power compared to 1 3/8 pipes.

But then again, the Maney system produced the best results, by a long way.

Tis a black art!

acotrel said:

I suggest that exhaust pipes are more about standing sound waves etc

Click to expand...

You keep using this "standing waves" phrase.
Understandable, but for clarity we should point out that PULSED FLOW would be more appropriate.

For 3/4 of the time, there is next to no flow in the exhaust (or inlet system either for that matter),
and then from a standing start again all hell breaks loose, until the next idle time and then the next pulse.
(There may be a slight reversion wave travelling back again, but this will be at quite a low gas pressure)

The gas waves have to negotiate all the obstacles twists turns and obstructions as a new experience each time,
they don't learn from the previous pulse, or leave any trails or markers to remember for the next time.

And, at different rpms, the flow responses can be quite different.
Or engines would develop max torque at all rpms, not just in specific rpm range(s).

N'est-ce pas ?

Sure, go ahead open that can of worms! I'll start it off with what little I know. Just rambling a bit, so it's maybe not too organized.

The Atlas and first Commandos had larger diameter pipes, but that was soon changed to the 1 3/8" diameter used from then on.

I've seen aftermarket and custom Commando pipes in 1 3/8", 1 1/2", 1 5/8", and 1 3/4". They all work ok in the right application. In general, the larger diameter pipes need to be shorter than the smaller ones to work properly in the same application.

The factory race team did some experimentation in diameter and length, and had pipe and megaphone combinations for 750 and 850 engines. They used 1 3/8" and 1 5/8" pipes, and probably tried other sizes, but I don't have data on that. The 1 3/8" pipes gave good mid-range power, but not as good top end as the 1 5/8" pipes. They had pipes and megas for max power, but also some to meet race track sound requirements, that made a little less power. They did use 1 3/4" straight pipes on some of the dirt track racers in the US.

Axtell did a lot of dyno work for Ron Wood's dirt track Nortons, and found the best power with 1 5/8" pipes with carefully sized reverse cone megas. I think Ron may have also run 1 3/4" pipes with the short stroke 750s in the mile races, but I'm not sure about that. It's been a long time, and I've only got an old guys memory now.

For a street bike, it's really hard to find anything much better than the stock pipes with the good peashooters. Of course that's just my humble opinion. Feel free to disagree. But, like for race bikes, it depends on your application. Something like Dunstall's last system, designed by Dr. Blair, would give more mid-range performance, enough that you could tell the difference, but gave up some top end for it. Probably great for most street bikes, but maybe not for the hard core who like to run their street bikes up around 7,000 rpm a lot. The 1 5/8" pipes with Axtell megas that I used on my race bikes gave great race track performance, but the huge flat spot in mid-range would be really annoying on a street bike.

Ken

texasSlick said:

acotrel said:

I suggest that exhaust pipes are more about standing sound waves and mass transfer of gases then about flow. I found this article interesting :

http://www.fonema.se/qpipe/qpipe.htm

Click to expand...

Agree!

If memory serves me correctly, a Honda model offered in the mid to late sixties. had a small diameter pipe inside a larger tube. Honda knew the smaller diameter offered better performance, but the American market looked askance on weeny pipes. The pipe in a pipe satisfied both performance and the market idiosyncrasies.

Slick

Click to expand...

Double pipes were as much about avoiding the blues as they were about size. Appearance in any case. Nowadays, with bodywork coverage, they use exhaust-grade stainless or titanium for the high-grade pipes.

At each junction (change of diameter) of an exhaust system, a negative pressure wave moves back towards the exhaust valve seat at the speed of sound, and the proper timing of these negative pressure waves help extract exhaust pulses and suck intake charge in. It's a sort of reverse-supercharging. The same effect is used in intake tract length design, where a positive pressure wave moves toward the intake valve seat, again at the speed of sound, and helps ram more charge into the combustion chamber. So, where the pipe joins the muffler, where the expanding muffler volume hits the tail cone and where the exhaust pulse exits, each creates a negative pressure wave. Variable-length intake setups move the carbs or throttle bodies farther away from and/or closer to the intake seats to create power at different rpm, where a static system creates best power over only a short range. Internal exhaust system valves do the same thing on that end. Not really black art, just numbers.

Pipe diameters are determined by the compromise between restriction (small) and weakening of the reverse waves (large). If there was a way to make them blow up like balloons and shrink back down, depending upon rpm range, that would be another tool to create more power over a wider rpm range.

So, as a general rule, smaller-diameter, longer overall length systems make better low-end power and shorter, larger-diameter ones make better top end power

What about the double diameter exhaust pipes that are 1 5/8 at the port then 1 3/8 at the muffer :?:

Anyone thought of reading the relevent section in the book Performance Tuning in Theory and Practice (Four strokes) by A. Graham Bell ? ISBN 0 85429 275 6.
I had my 2 into 1 exaust system calculations done using the book checked by a rather well known tuning Company and their computor gave identical results........

J. M. Leadbeater said:

Anyone thought of reading the relevent section in the book Performance Tuning in Theory and Practice (Four strokes) by A. Graham Bell ? ISBN 0 85429 275 6.
I had my 2 into 1 exaust system calculations done using the book checked by a rather well known tuning Company and their computor gave identical results........

Click to expand...

Any pics? :idea:

I always thought the Commando pipes looked small. Especially when seen next to my P11 which from the factory had a decent size.

My 1 1/2" pipes on my 750 are 2.5" longer than stock. My low end response is remarkable and has pulled nicely past 7000rpm.
The sweet spot seems to be 4500rpm at 80 MPH.
Sometimes I question if a 1 3/8" would be better, but I have tried for so long to achieve these results that, although I cannot believe I am saying this, I am totally satisfied and don't want to dick with it anymore.

JS1, Beehive springs, 40mm pumper, tuned (advertised as) 1 1/2" exhaust, shaved RH1.

J. M. Leadbeater said:

Anyone thought of reading the relevent section in the book Performance Tuning in Theory and Practice (Four strokes) by A. Graham Bell ? ISBN 0 85429 275 6.
I had my 2 into 1 exaust system calculations done using the book checked by a rather well known tuning Company and their computor gave identical results........

Click to expand...

I am an aerodynamicist, and I would not attempt tuning an exhaust system without study of reference material such as this. Anyone who tries to do so without such study will be "shooting in the dark". At the very least, do not waste your time and money without having some benchmarks or guidelines developed by someone else thru trial and error, and/or in conjunction with theory.

I fully agree that exhaust tuning is usually done to increase the mid-range torque, and this boost is paid for by a loss at the top end (First Law of Thermodynamics ... you can't get something for nothing!).

Slick

First law of Norton tuning, Luck is HUGE!

Bernhard said:

What about the double diameter exhaust pipes that are 1 5/8 at the port then 1 3/8 at the muffer :?:

Click to expand...

Probably just another attempt at having your cake and eating it too. Flattened-section "Cobra head" pipes were the rage for a short time. Didn't last.

I suggest that when building a commando. the first thing you need is an appropriate gearbox. Then you need to set a rev limit for your motor based on what you can afford in the way of engine modifications. If you are rev-limited to 7000 RPM the exhaust system serves less beneficial purpose if its length and diameter are optimised for best performance above 7000 RPM. Another factor is about the use of megaphones, if your motor comes on song at 4000 RPM a megaphone will intensify the transition from little power to lots of power, so the bike becomes more difficult to ride smoothly around tight circuits.
For Rowan about standing waves. When sound passes through an open ended pipe it reflect backwards from the open end and forms a standing wave. the wavelength is equal to the speed of sound divided by the frequency. An fixed length exhaust pipe really only operates at maximum efficiency at one rev level of the motor. Within the pipe there are high pressure areas and nodes, and as the rev levels change these move up and down the pipe so the pipe is effective over a range however rarely at optimum efficiency. Gas moves through the pipe by mass transfer in the same way that a tsunami crosses oceans. With a tsunami, a wave does not cross the ocean, it is mainly energy passing between pressure points across the nodes. The last pressure point is the wave that gets sucked off the beach and then blasted through the houses. You can call it pulsed output if you like, however I don't think of it in those terms.
With a two into one pipe, I suggest the tail pipe length should be optimised to vibrate at twice the frequency of one of the header pipes. Again there is probably only one rev level at which it will deliver maximum power. If the collector diameter causes back-pressure a two into one pipe is often less effective in delivering more torque, so diameters are important. I have also advanced my cam so more of the charge gets fired down the pipe.

acotrel said:

For Rowan about standing waves. When sound passes through an open ended pipe it reflect backwards from the open end and forms a standing wave. the wavelength is equal to the speed of sound divided by the frequency. An fixed length exhaust pipe really only operates at maximum efficiency at one rev level of the motor. Within the pipe there are high pressure areas and nodes, and as the rev levels change these move up and down the pipe so the pipe is effective over a range however rarely at optimum efficiency.

Click to expand...

To put acotrel's explanation in other words:
When the pipe or tube subjected to pulsating flow, hits the resonant frequency, it then is "tuned", and the mass transfer flow meets minimum resistance. At the resonant frequency, the pulsating waves appear to be stationery, or "standing". As Alan states, this occurs at a very specific frequency, and is optimal only over a limited range.

An analogy to standing waves in a string helps make this phenomena clear. Refer to an animation on string standing waves in the following link:

http://www.physicsclassroom.com/mmedia/waves/swf.cfm

Exhaust tuning is simply the art of twiddling the pipes length and diameter to alter the pipes resonant frequency so it (the resonant frequency) coincides with an optimum torque rpm range on the engine's performance curve.

Slick

acotrel said:

For Rowan about standing waves. When sound passes through an open ended pipe it reflect backwards from the open end and forms a standing wave.

Click to expand...

For Ellen, we are all familiar with this concept.

The problem in this comment is that exhaust gas flow in exhaust pipes is a PULSED FLOW.
For 3 quarters of the time, there is not much flowing in the pipes - so the 'standing wave' dissipates almost as soon as it forms - stand behind a Commando with peashooters sometime, you can feel the puffs of gas emerging.

And, its not sound waves that we are interested in to harness for better exhaust scavenging etc, its the pressure pulses of exhaust gas and the associated reversion waves, as Tex mentions above. There is a difference.
The sound is merely the musical accompaniment...

I do not pretend to have any facts or numbers to back these ideas and observations. I do believe and have heard air box resonance, along with exhaust pipe tuning, for the cam selected completes the compromised efficiency of engine breathing over operating RPM. Many of our Nortons (especially 2A's and 3's) are not running any air box at all, but are using the ever popular K&N pod type of filter for twin Amal carbs, myself included. Not scientific I know but on my MK3, replacing the original black caps with straight thru pea shooters resulted in a stumble at 4000 RPM or so. Gone with the K&N though.

And just a further observation. V12 Rollls-Royce Merlin aircraft engines of WW2 had virtually no exhaust system at all, just a flattened stack at each port, but they were tuned to run at a pretty constant RPM of around 2800.