factors affecting exhaust pulse duration

[jaguar57]

hi guys

for the past 3 years I've been developing my own spreadsheet for use in analyzing and designing expansion chambers. I used to be under the impression that most exhaust pulse durations (the time from the beginning to the end of the pulse) were around .8 milliseconds. But recently someone with a Rotax 493cc engine (two 247cc cylinders) sent me a pressure trace showing the exhaust pulse lasts 1.6ms at 8600 RPM. (12:1 compression, 191 exhaust port duration). Blair published a great paper about analyzing a karting engine. It was a single cylinder 100cc, reed valve intake, 10:1 compression, 175 exhaust port duration. Its pulse durations were .83ms at 8000 and 10,000 RPM, and .52ms at 16,000 RPM. 

 

Since the equipment to make your own pressure traces costs over $6000 I may have to wait another 3 years for more data. But I thought I'd take a wild shot here and put this out. Going only on those two examples I am prone to assume these things cause a longer exhaust pulse: 1) larger cylinder size, 2) higher engine compression, 3) longer exhaust port duration.

 

The larger cylinder size makes sense to me since an analysis of exhaust port area compared to cylinder size for a single port exhaust with 171 degrees duration showed a 100cc port had 8.8% of above-port combustion chamber area, whereas the same with a 200cc had 6.9%. Adding auxiliary exhaust ports to the 200cc brought it from 6.9% to 9.7% since the auxiliaries increased port area 40%. So comparing apples to apples, the larger the cylinder the slower the exhaust gases will come out due to a smaller ratio of exhaust port area to combustion area. (6.9% is 22% less than 8.8%)

Unfortunately I don't have specs on the exhaust ports for either of these two engines.

 

At first I thought a higher compression would give a quicker pulse but now I can see that in a different light. Higher compression causes a quicker burn which means the chamber pressure at time of port opening would be less, therefore taking longer for the gases to exit.

 

The third assumption is whats gets me. Why would a higher exhaust port contribute to a longer exhaust pulse? A higher port opens when the pressure is higher and so it should exit faster. Maybe I should toss that assumption.

 

Since the exhaust pulse duration affects the lengths of the return waves in the pipe this is a point of concern for someone who wants to properly design his own pipe. I am all ears for smart comments from 2 stroke experts.

 

Three other factors to consider: ignition timing, fuel used (since high octane causes a slower burn), and pipe back pressure.

 

ps- yes I know there are other pressure traces available on the internet but most all of them were generated by software so you aren't seeing the real thing, just someones idea of what it should be.

[grayracer513]

Without sorting through the whole issue, I see at least one flawed assumption that you haven't considered which is that octane affects the rate of burn; it does not.  In fact, octane number itself does not directly affect any attribute of the fuel's behavior beyond the resistance the fuel has to being ignited by sources other than flame or spark (heat, pressure, shock, etc.), and thus its resistance to detonation under those conditions.  The specific means by which octane increases are chemically achieved can sometimes have "side effects" that do change other fuel attributes such as burn rate, vaporization rate, burn temperatures, and all of that, but not necessarily, and the changes can go either way depending on the specific fuel chemistry.

 

In most two-strokes, combustion pressures peak well ahead of port opening to the extent that raising the port would likely not start off the exhaust cycle at a significantly higher pressure, so that one is questionable as well.

[JoeRC51]

And the point of this thread is what ? What are you looking to do ?

[turbo dan]

I expect that exhaust port and power valve design has a significant effect.  A wide, bridged exhaust port would presumably produce a strong, shorter duration pulse.  An oval shape port without a bridge would open more gradually and produce a longer pulse with less peak intensity.  A cylinder with exhaust sub ports might produce an even longer duration pulse since the sub ports require a longer flow path into the expansion chamber than the path traveled by the gases flowing through the main exhaust port.  I have seen sub ports timed to open slightly before the main port as well, which would presumably produce a double pulse kind of effect.

[pcnsd]

Speaking strictly to the exhaust pulse length. I think you will find RPM is an important factor since it is units divided by time and variable. My guess without actually spending effort to build sheets, is the input factors are RPM and port window duration in degrees of rotation. JMO

[jaguar57]

yeah I would of assumed that too but Blairs pipe had a duration of .83ms at 9,000 RPM* and at 16,000 it had a duration of .52ms which is more than half of .83 which it would have if your theory was correct. RPM is having an effect but I need to reanalyze this in light of the increasing port area as the piston descends. Since the port is rounded this causes a non-linear effect.

 

* the exhaust pulse and diffuser wave were "blending" at 8,000 and 10,000 so I was doing educated guesswork about their theoretical duration if the pulse was not mixing with the diffuser wave (which shortens both of them). Now I'm sure the non-mixed exhaust pulse at 8,000 would be a bit longer than it is at 10,000. Just now realizing this means I have to modulate the exhaust pulse duration according to RPM in my spreadsheet.

Edited January 16, 2016 by jaguar57

[jaguar57]

OK I traced the exhaust port of my Suzuki 100 and it has an oval single exhaust port at 65% of the bore for its width. Its duration is 171 degrees. Spending hours to do all the meticulous computations for the time-area each 1mm down the port, and assuming an exhaust pulse duration of .7ms at 10,000 RPM I came up with how the duration changed with RPM. It wasn't near as drastic as I thought it would be. Here they are:

RPM  exhaust duration

16K   .695ms

10K   .700ms

8K     .72ms

6K     .76ms

4K     .81ms

Maybe Blairs notable decrease in exhaust duration had something to do with minimal delivery ratio at such a ridiculously high RPM of 16,000.

Edited January 17, 2016 by jaguar57

[JoeRC51]

Can I ask the point of the thread or is it just an exercise ?

[clappedoutkx]

It certainly is interesting.

[JoeRC51]

It certainly is interesting.

 

Well sure it is , to what end ?

Edited January 17, 2016 by JoeRC51

[grayracer513]

Can I ask the point of the thread or is it just an exercise ?

 

 

... for use in analyzing and designing expansion chambers.  [...]

[SDet]

Could the efficiency of the intake have an effect?

As the rpms go up, you're filling less of the cylinder, so smaller explosion, and less exhaust gas, so a shorter pulse. Unless that's going in the wrong direction.

[JoeRC51]

There are programs to determine all of the intricacy's of expansion chamber design dimensions based mostly on RPM range , desired power curve and port timing. So far you're just peeing into the wind here.

[grayracer513]

Plugging numbers into software and watching it spit up results will only teach someone so much.  He's trying to examine and understand the effects of various elements of the equation, and IMO, any furthering of one's education or understanding is an admirable pursuit much different than upwind urination.  Even if he's on the wrong track, it's beneficial to learn why. 

[JoeRC51]

He's no pioneer , It's ALL been done before.

[jaguar57]

agreed, it's all been done before, but not by me. My doing it causes a more perfect understanding of all that's involved, by me of course.

Like just this morning I realized that it's not "pumping losses" that cause hi compression engines to not do so well at high RPM. It is because the higher compression causes a longer exhaust pulse duration which limits "transfer" of intake charge up thru the transfers because the transfer has to wait till the exhaust pulse is around 5psi before it can enter the cylinder. Yes pumping losses exist and are important but they don't increase their percentage of loss at higher RPM. The percentage stays the same.

 

My purpose was stated in my first sentence "I've been developing my own spreadsheet for use in analyzing and designing expansion chambers". How anyone could ignore that and go on to ask "why" is beyond me. 

 

SDet, the "efficiency of the intake" is basically what I was referring to with the term "delivery ratio". The less efficient the engine is at filling the cylinder with intake charge, the less combustion pressure there will be and the shorter the exhaust pulse will be.

 

Joe, yes there are other programs but the only good ones cost $400 which the basic tinkerer/DIY person will not spend. I like to fight for the underdog so part of my motivation is to make something good that is available to everyone free of charge.

Edited January 17, 2016 by jaguar57

[SDet]

agreed, it's all been done before, but not by me. My doing it causes a more perfect understanding of all that's involved, by me of course.

Like just this morning I realized that it's not "pumping losses" that cause hi compression engines to not do so well at high RPM. It is because the higher compression causes a longer exhaust pulse duration which limits "transfer" of intake charge up thru the transfers because the transfer has to wait till the exhaust pulse is around 5psi before it can enter the cylinder. Yes pumping losses exist and are important but they don't increase their percentage of loss at higher RPM. The percentage stays the same.

My purpose was stated in my first sentence "I've been developing my own spreadsheet for use in analyzing and designing expansion chambers". How anyone could ignore that and go on to ask "why" is beyond me.

SDet, the "efficiency of the intake" is basically what I was referring to with the term "delivery ratio". The less efficient the engine is at filling the cylinder with intake charge, the less combustion pressure there will be and the shorter the exhaust pulse will be.

Joe, yes there are other programs but the only good ones cost $400 which the basic tinkerer/DIY person will not spend. I like to fight for the underdog so part of my motivation is to make something good that is available to everyone free of charge.

I still question the assumption that pumping losses are constant. Higher rpm, higher piston velocity. Higher piston velocity, higher gas velocity. As the velocity increases, the drag effects will not increase linearly, so you end up with lower intake pressures, and thus a less full cylinder.

I feel the entire engine has to be modeled as a system, from air cleaner to the end of the exhaust. Then you can start to simplify what does and does not have an effect.

[SellYour4Stroke]

Have you tried getting a hold of a manufacturing engineer?

MMI is tight with the manufacturers and may be able to give you a few names and numbers to work with. Just brainstorming. Sounds like you have graduated from thumpertalk forums and are moving into the manufacturing engineering world.

Acoustics, harmonics, waveforms...heavy science. Another idea might be a professor that may teach some acoustics engineering. Either way, they may be able to point you to another person that may be deeper in the industry.

[pcnsd]

Have you tried getting a hold of a manufacturing engineer?

MMI is tight with the manufacturers and may be able to give you a few names and numbers to work with. Just brainstorming. Sounds like you have graduated from thumpertalk forums and are moving into the manufacturing engineering world.

Acoustics, harmonics, waveforms...heavy science. Another idea might be a professor that may teach some acoustics engineering. Either way, they may be able to point you to another person that may be deeper in the industry.

 

ManuEng's are involved in the design and refinement of production systems and processes but are not for the most part not involved in the design of products. Depending on company culture, consulting with the Design authorities takes place during pre-product implementation to insure the product does not carry difficult to manufacture features or unnecessary assembly steps. (Design for manufacturing and assembly).To learn more about Manufacturing Engineering check out: www.sme.org.

 

For transportation systems design check out: www.sae.org

 

That said, here is what I came up with.

 

Time in Seconds=(RPM/60)/(DUR/360)

 

Where:

RPM is the revolution count per minute of interest.

DUR is the duration in crankshaft degrees that the exhaust blowdown event occurs over.

 

The blowdown event that creates the pulse is only a portion of the total exhaust time window. My best guess without using a pressure log to record and review is this event gains slightly in total duration as RPM increases and will also vary based on the stroke/rod length ratio.

[grayracer513]

The blowdown event that creates the pulse is only a portion of the total exhaust time window. My best guess without using a pressure log to record and review is this event gains slightly in total duration as RPM increases and will also vary based on the stroke/rod length ratio.

 

The first sentence is obviously correct if you take the quoted numbers from Blair's work as correct.  For example, 175 degrees of exposure at 8000 RPM translates to roughly 3.6 ms, and the pulse lasts only .83 ms as reported.  Even the time from opening to BDC is much longer than the pulse time at 1.8ms. 

 

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

 

 

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

 

Nevertheless, it's interesting to consider what exactly constitutes the pressure "pulse", as it would seem to me that any condition that causes gasses to move out through the port is an indication of positive pressure in the cylinder relative to the exhaust.  It seems logical that there is a burst of pressure into the exhaust as the port opens, and that there is a finite life to that burst, I just wonder what the pressure curve looks like as it trails the peak point, especially since there is a outflow into the exhaust that includes some of the incoming F/A charge.  

 

While I understand how the rod/stroke ratio affects port open time, I don't see it affecting the burst pulse more than a very small amount, as the pulse is primarily a function of captive pressure remaining at the time of the port opening, and that would not change more than a few single digit degrees in the case of most real world adjustments to rod length.