Jump to content

# factors affecting exhaust pulse duration

## Recommended Posts

However, Blair's numbers as reported do disagree with your best guess analysis:

Quote

Its pulse durations were .83ms at 8000 and 10,000 RPM, and .52ms at 16,000 RPM.

Truthfully I do not know precisely, but the above is time duration and not duration of crankshaft degrees. I was perhaps not clear, but I meant to suggest the relative degrees of duration would extend as the RPM increased do to the shorten time window allowed for blowdown during the exhaust event. The equation to me suggest that pulse time will diminish with RPM assuming a fixed blowdown window in duration is present, but maybe not by much if the blowdown duration extends also as RPM rises.

Paul

#### Share this post

##### Share on other sites

At 8K RPM, the crank would rotate 39.8 degrees in .83ms.  At 16K, it would rotate 49.9 degrees in .52ms, so that part holds up.

#### Share this post

##### Share on other sites
Grayracer, why did you say “Blair's numbers as reported do disagree with your best guess analysis” after I said the 8,000 and 16,000 RPM exhaust duration times which were .83ms and .52ms were derived from Blairs own exhaust pressure traces? (which is real life results, not calculations)

My thoughts on why the shorter time of exhaust happens at double the RPM is that at double the RPM the port is obviously more quickly opened to relieve the pressure quicker. (and the port being roundish and not square [which is impossible for ring life] accentuates this)

I made a graph of port open area vs piston descent and it was almost a straight line which I thought was a cute mathematical quirk. I theoretically then squared the top of the port as much as I’d dare and that didn’t really help the graph very much. The stock port is fairly roundish to maintain a broad powerband. (The cylinder needs auxiliary ports to make much of a difference)

Of course for the graph to be more true to life the horizontal scale should be in equal segments of degrees of crank rotation. (The first mm of piston descent was 2.3 degrees, the 7th is 2.65, and the 13th is 3.4, which shows the non-linearity of it.)

I also did an analysis of resultant exhaust duration times if that depended on time-area (see below) and there was little difference* (5.1%) in duration between 8,000 and 16,000 RPM. So either that theory is unusable or I was far off in my educated correction of Blairs pressure traces. Suggestions for a different theory are welcome.

* .697ms at 16,000 RPM and .735ms at 8,000 RPM

(blowdown has little to do with it since when the transfers open some of the pressure theoretically is lost by its pressurizing the crankcase but I doubt it lowers the pressure much in real life because crankcase space is so limited)

——————————

Analysis at 8,000 RPM

each row is the result of 1mm piston descent more

1             2         3         4

degrees  time    port     time x

open       open   width   area

2.3          .048    12mm  .576

2.35        .049    19        .93

2.4          .05      25      1.25

2.45        .051    29      1.48

notes:

1. degrees of exhaust port opening for each row (each 1mm piston descent)

2. time in milliseconds related to degrees at 8,000 RPM

3. port width in millimeters along the top of the piston edge

4. number 2 times number 3

Edited by jaguar57

#### Share this post

##### Share on other sites

Read my post; I was responding to pcnsd, not you, and that was because he was talking about the pulse being longer in terms of degrees of crank rotation, not time.

Follow

I take issue with this, however:

(blowdown has little to do with it since when the transfers open some of the pressure theoretically is lost by its pressurizing the crankcase but I doubt it lowers the pressure much in real life because crankcase space is so limited)

Without knowing an exact number, I'm thinking the transfers open a good 40-50 degrees after the exhaust, by which time the blow down phase will have advanced to the point that crankcase pressure will be higher the cylinder pressure.  If it isn't, if the cylinder pressure is higher, then the transfer charge will become stalled.  This should only happen at speeds well below the effective operating range.

Also, there is the fact that with a properly designed exhaust, the negative pressure wave created by the divergent cone should have arrived at the port as the blow down ends to reduce residual pressure even farther than it would be on its own, so basically the scenario of residual combustion gas pressurizing the crankcase should never happen except at impractically low RPM.

#### Share this post

##### Share on other sites

regardless if it stalls or anything else I still don't believe the blowdown has any relevance on the exhaust pulse time. It has relevance on engine power.

I think on most motocross bikes at high RPM the transfers open when the cylinder pressure is still too high which delays the transfer of intake charge. As rpm increases this becomes worse and worse which is why the power drops off and you are forced to upshift. Even if the pipes divergent cone starts early to start the suction wave at the same time as the transfers opening, it still isn't much because the cone angle is very shallow which doesn't produce much of a suction. It helps a little but not much.

Edited by jaguar57

#### Share this post

##### Share on other sites

I agree with your first sentence, disagree with the last three.

Let's assume for round numbers that you have a total exhaust port opening of 180 degrees and transfer opening of 90, which would mean opening at 90º BBDC on the exhaust and 45º BBDC on the transfers.  At 8000 rpm, Blair's pressure pulse lasted .83ms, which accounts for 39.8º of crank rotation; the transfers are not yet open.  At 16000 RPM, Blair's pulse duration of .52 accounts for 49.9º crank rotation, so the transfers would have been open for 4.1º, which would make them roughly 1mm open, "just" at the point of being effectively open.  Then, you have to consider the pressure pulse as graphed on a curve over it's duration.  It will rise sharply, peak quickly, then fade, probably over the last 65% or better of its duration.  So I seriously doubt you're going to find it capable of raising crankcase pressure once the ports actually become effectively exposed.

In the expansion chamber, the negative wave begins to develop as soon as the positive wave from the port enters the divergent cone, and its specific purpose is to aid in residual gas evacuation.  That's why the header is so short.

Power fall off results from outrunning the reflected positive wave from the convergent cone, dumping fuel/air charge into the exhaust and closing the port before that wave has a chance to push it back into the cylinder.

#### Share this post

##### Share on other sites

for one, no engine ever has as short a transfer duration of 90 degrees. Please use a realistic example.

regardless, the transfer port location has no affect on the exhaust pulse length of time.

I've been studying pipe design profoundly for a few years and can say your belief is a bit off. If you make the header short then a very weak beginning of the diffuser wave will help a tad with residual gas evacuation. What really helps a lot with evacuation is when the header diameter is not too big and the outward blast of the exhaust thru the pipe causes a vacuum "pulse" that follows it which does the work of evacuation. It is about 2/3 the time of the exhaust pulse and 1/3 its strength.

Yes power falls off for more than one reason, one being the RPM is higher than the pipes powerband. But I was talking theoretically as if there was no expansion chamber, talking just about the engine dynamics.

anyway we are getting off-topic. I just wanted to chew over the main factors that influence exhaust pulse time duration, not all the intricacies of expansion chamber design. We can start another thread for that if you'd like.

#### Share this post

##### Share on other sites

for one, no engine ever has as short a transfer duration of 90 degrees. Please use a realistic example.

Provide one, then.

As far as headers go, all expansion chamber headers are short relative to 4T exhausts.

120° ?

#### Share this post

##### Share on other sites

115 to 130 is the normal range of transfer port duration.

Please don't compare a 4 stroke pipe to a 2 stroke pipe. It's like comparing apples and oranges. When I finish the new version of my expansion chamber design spreadsheet I will be able to provide the visual illustration of how little a short header helps.

#### Share this post

##### Share on other sites

They're more alike than you think, just designed to accomplish something different.

Thanks for the specs though.

#### Share this post

##### Share on other sites

Well it does look like RPM is a main factor in exhaust time. At 8000 RPM it is hard to tell what the original exhaust wave length is so I looked at just those from 16,000 and 10,000 from the pressure traces Blair put in his paper. It turns out the ratio of them (.54/.88ms) is almost exactly the same as the ratio of the cycle times* of the two RPM (3.75/6.0ms). So if I used that same method I can derive that the exhaust length at 8000 RPM is 1.1ms.

* here's the formula for finding the time of one cycle (one crank revolution): cycle time=1/(RPM/60)

#### Share this post

##### Share on other sites

data for the two samples of real life pressure traces give me engine compressions of 12:1 and 10:1

What does that normally equate to in psi of cranking pressure? (kicking it over. rings not worn down)

#### Share this post

##### Share on other sites

well I think the advantage of having a short header and long diffuser is more than I thought.

On a test analysis if the diffuser wave is made to begin at the same time as that of the transfers opening then it does shorten the exhaust pulse a tad which is beneficial. It also increases the strength of the exhausts trailing vacuum pulse which is advantageous as long as it doesn't suck out of the cylinder some of the intake charge along with the last of the exhaust gasses. (you can tell by looking at the tell-tale pattern on top of the piston.)

#### Share this post

##### Share on other sites

I apologize, when I got to the second page I skipped to the end. I think you may be chasing your tail or are confused on what you are trying to measure. I may be missing your point also.

you are trying to understand how an "expansion chamber" or "tuned exhaust" works on a two stroke engine. the effectiveness of an expansion chamber on an engine is not affected by ports, fuel, timing, compression or any of the above. what I believe you are referring to as a "pressure pulse", is, in fact, a "sound wave". Very simply put, when the exhaust port, or valve, first opens, there is an escaping "bang" if you will. that "sound wave" travels out of the port and into the pipe. the first section of which is called the diffuser. the diffuser is tapered at a specific angle. the sound wave travels along this taper like a donut, rolling along the outside, struggling to escape. as it rolls along the diffuser the, donut becomes a little weaker and starts to "spread out" along the pipe. the diffuser empties into the center section, which is usually a straight piece of pipe. the sound wave donut rolls along the center section and doesnt change much in intensity or speed. The center section dumps into the Conicle baffle, which is shorter and angled steeper then the diffuser and is angled in the opposite direction. the sound wave enters the conicle baffle and because the angle is steeper and the length is shorter, begins to speed up and become a more compacted donut (like a stepped on fried cake). at a point very near the end of the baffle, is a spot in the diameter called the "point of mean reflection" where in, physics require the sound wave to collapse on itself, and reverse direction, where this whole process starts all over. The sound wave heads back toward the cylinder, dies down to a slow, fat donut, cruises through the center section, then begins to speed up and gain intensity again in the diffuser, sliding easily through the hot, thin exhaust gas that followed it out of the cylinder, BUT, crashing violently into the cold fresh intake mix that tried to follow the exhaust out of the cylinder, and ramming it into the cylinder, just before the piston closes it.

So what you are measuring, is the speed of sound, inside an environment of varying pressure and temperature, and being manipulated by a metal tube, of varying diameter and thickness, to its point of reflection and back again.

Factors that affect the "time" it takes for the wave to do this are

Overall or "tuned" length of the pipe.  "tuned" referring to tuning the sound.

length of the diffuser and the conical baffle. Angle has more affect on the difusing and intensifying of the wave then on speed.

length of the center section.

Temperature inside the pipe. hot =fast, cold=slow

Pressure inside the pipe (not airflow) low pressure=slow, high pressure =fast (sound travels faster in dense air. no sound in outer space with no pressure)

Thickness of material pipe is made of. thick =slow, thin = fast, too thin = broke pipe (sound waves tear it apart)

The tuned length is determined by the RPM that the engine needs to develop peak power at.

Angles determine the "width" of your "power band".

Tailpipe length and diameter determine "back pressure"

Sometimes the tailpipe can come out of the center section so that it does not interfere with the reflection point.

So if you can calculate the speed of sound, based on the available variables that affect it, you will find that they match your time tables.

The perfect invention would be a pipe that has a variable tuned length, like a sliding center section. instead, all the manufactures just decided to go with a variable height exhaust port. more on that in chapter two if necessary.

Scooter

• 1

#### Share this post

##### Share on other sites

P.S.

2t pipes and 4t pipes are designed to do the exact same thing in regard to managing a sound wave. the difference is that 2t pipes want that sound wave to come back, 4t pipes dont. usually refered to as anti-reversionary. very complex also. returning pulse not necessary because port timing is not directly fixed to crank timing. just change the cam.

#### Share this post

##### Share on other sites

I am assuming by pulse duration you mean the returning compression wave that rams intake charge back into the cylinder. If so the duration of this event is mainly controlled by the blow down duration (ex open to top of transfer open) - no one wants positive pressure when the transfers open so by the time the piston reaches the top of the transfer port nearly all of the pressure is gone. The exhaust gas temperature influences the time for the returning compression wave but this has little influence on the duration of the wave. The length of the convergent cone will strongly influence the length of the compression pulse and its magnitude. A short steep angle convergent cone creates a steep and short duration returning compression wave, a long and shallow cone creates a long pulse with low magnitude. There are also expansion and compression waves still bouncing around inside the pipe from previous cycles these "residual" waves also affect the duration and magnitude of the "ram charge" pulse. Ultimately, one has to simulate these waves in an unsteady gas dynamic program to even begin to track all the wave interactions.

• 1

#### Share this post

##### Share on other sites

My god people have you never seen an exhaust pressure trace to know what I am talking about?

for your continued education please look at one from Blairs paper. It is the 2nd graphic on this page of mine: http://www.dragonfly75.com/motorbike/validation.html

An exhaust pulse is not a "sound wave". It is the actual pressure of the combustion gasses exiting the cylinder and how long it lasts.

XR, you are singing to the choir. I know more about this subject than you obviously can imagine.

Also you talk as if a sound wave is not just another type of pressure wave. Look "sound wave" up in Wikipedia. Also the Mirriam Webster Dictionary says a sound wave is a "longitudinal pressure wave in any material medium"

Edited by jaguar57

#### Share this post

##### Share on other sites

Nope, never saw an exhaust pressure trace. Thanks for introducing me. I believe your initial question was, what factors influence this mystical pressure wave. Since you were so kind to point out that a pressure wave and a sound wave are the same, then I believe I gave you some qualified suggestions. A change in material the pipe is made of is going to impact the wave much more drastically then the fuel your burning, and temp inside is going to have a bigger effect then transfer size or cylinder pressure.

#### Share this post

##### Share on other sites

I know more about this subject than you obviously can imagine.

Can't top that.

• 1

## Create an account or sign in to comment

You need to be a member in order to leave a comment

## ThumperTalk

Register a new account Sign In Now

×