[GP125] All that you wanted to know on Aprilia RSA 125, and more, by Mr Jan Thiel and Mr Frits Overmars (PART 1) (Locked)

Frits Overmars a écrit:

One cannot quote a cone angle for this phenomenon, Allan. The stronger the initial pulse and the greater the distance from the reflector back to the cylinder, the greater the risk of the return pulse developing into a shock wave before it arrives at the cylinder, and the shallower the reflector needs to be.

Thanks for your reply Frits,
This is the first I have not heard about this Phenomenon and would like to read more about it. Do you know of any books, software or sites that explain more about advance pipe design for 2 strokes?

I have played on the Mota program for many years and I have built many pipes for myself in this time with good results but it does not teach the reasons behind the solutions. I have appreciated using this program even if it not perfect as it has allowed me to learn a lot about exhaust pipes by trial and error. I also have Blair's book but I find it too simple in the pipe design side for serious consideration.

GtG001 a écrit:

This is the first I have not heard about this Phenomenon and would like to read more about it. Do you know of any books, software or sites that explain more about advance pipe design for 2 strokes?

You would need a book on gas dynamics; forums or sites that go deeper into this subject do not come to mind.

Citation :

I have played on the Mota program for many years and I have built many pipes for myself in this time with good results but it does not teach the reasons behind the solutions. I have appreciated using this program even if it not perfect as it has allowed me to learn a lot about exhaust pipes by trial and error. I also have Blair's book but I find it too simple in the pipe design side for serious consideration.

There are several Blair-books. Blair developed a very elegant method of dealing with particle behaviour in engines and from memory (I'm abroad; can't look it up for you) I would say this is dealt with at length in his latest book on two-strokes.
This being said, it is true that neither of his books is much help at practical pipe development. But they may help you gain the understanding you need to move forward under your own steam.

If you know Frits, you don't need complicated formulas and calculations.
With this Aprilia pipe drawings the best basics are given.
You can scale the dimensions and it will works!
"We" (some friends and other 2stroke-tuners i know) develop with the Frits-pipe. It is the best i know and it will work to 90% perfect. The remaining 10% come with testing on the dyno and the track, therefore you need some experience and your own motor facts, so no book or software or person can tell you how to calculate the 100% pipe.


Edit: I started a summarization of (in my eyes) important facts, mostly given by Frits Overmars. Generally designed in answer-question-style.
I stopped at page 41 here but will go on one day.
And please forgive me, but I ignored all french postings because of can't reading them

[Vous devez être inscrit et connecté pour voir ce lien]

Frits Overmars a écrit:

One cannot quote a cone angle for this phenomenon, Allan. The stronger the initial pulse and the greater the distance from the reflector back to the cylinder, the greater the risk of the return pulse developing into a shock wave before it arrives at the cylinder, and the shallower the reflector needs to be.

Thank you again for your reply.
Maybe, I can work backwards in this regard. What characteristics do you experience if the reflected wave has turn into a shock wave? Is it a lack of power and a requirement for an unusual ignition map or excessive fuel curve to gain some resemblance of a power curve?

Frits Overmars a écrit:

You would need a book on gas dynamics.

GtG001 a écrit:

Maybe, I can work backwards in this regard. What characteristics do you experience if the reflected wave has turn into a shock wave? Is it a lack of power and a requirement for an unusual ignition map or excessive fuel curve to gain some resemblance of a power curve?

Wouldn't it be better to work forward? I can hardly believe you already finished that book.
And before you ask, this forum is not the place for a comprehensive lecture on gas dynamics.

Frits Overmars a écrit:

The three-part reflector is a relic of evolution, like an appendix. I think a single cone can be just as effective.

Okay thank you! But the conical middle section isn't a relic of evolution I guess!? This is the thing I find most interesting with this exhaust pipe.

Regards

Yes, that is a relic too, Daniel. Sometimes a modification gives a positive effect which in turn may lead to a false conclusion.
For example, if you enlarge the diameter on one side of the cylindrical belly of a pipe, the engine may run better. Then you may think it's because the belly has changed from cylindrical to conical, while in reality the engine just runs better because the pipe volume also changed.

Damn. I had hoped to have found some new power sources

But I don't really understand, why does the latest exhaust pipe still contain so many "relics"? Was there no time to test a 0° mid section and a 1-stage reflector?

Regards

Daniel A. a écrit:

why does the latest exhaust pipe still contain so many "relics"? Was there no time to test a 0° mid section and a 1-stage reflector?

Simple answer: no. For example, if you substitute a three-stage reflector with a one-stage reflector, you always change more than one factor, like cone angle, cone length and pipe volume. You may have to test 10 one-stage reflectors before you find one that is better than the three-stage reflector.
But there are always other experiments waiting as well, with cylinders, piston shapes, combustion chambers, inlet disks, ignition curves, carburettor settings, con rod lenghts, crankcase volumes and so on. Racing = lack of time.

Dear Frits, if you're not exhausted by all the questions you unleashed... did you ever try a multi-stage reflector with the other ordering, ie steeper part first?

Back when I was young and dumb(er) I built some 250MX pipes where the 2nd stage of the reflector had a quite gentle angle. So it was a bit like a diffusor mounted backwards, based on the intuitive principle that dArea/Area should be closer to constant.

The result didn't produce lots of power but it did seem to provide a nice smooth powerband, so far as my buttocks could judge.

Pipe design is very interesting to me. It would seem that if Jan tested over 200 pipe designs then the latest design would be an evolution of the best results. With 200 designs there would be variations that fell flat and some that were superior. The superior designs would be analyzed and the best pipes zeroed in on and the designs studied to see what works and what doesn't. When I test pipes I always make a full size template of each pipe and overlay them. I also print the dyno sheets and overlay them. This way I can see how each difference in the pipe design changes the power curve, the air flow and the fuel flow. If I had the luxury of testing full time on someone elses money then surely the effect of a tapered center would be understood and if it were not advantageous then would be substituted for a better design. Obviously if this is the latest design, the record of success on the track speaks for its self. I have never tested a tapered center but when I think about it it allows for a longer baffle cone of the same angle with the same tuned length which would give a slightly earlier and longer return pulse. Because I hydroform my pipes it will be easy for me to make a new template from an existing design and blow a pipe with a tapered center as the only change. Yes it will increase the volume but it should not change the tuned length. Stay tuned as I am going to the shop today to fabricate a set and compare them to the straight center set.

Of course if you are hydroforming, you have the option of a continually varying taper in the diffuser. Theoretically that must be superior to a collection of different cones... although the benefit might be microscopic compared to the fabrication issues...

GrahamB a écrit:

... did you ever try a multi-stage reflector with the other ordering, ie steeper part first?
Back when I was young and dumb(er) I built some 250MX pipes where the 2nd stage of the reflector had a quite gentle angle. So it was a bit like a diffusor mounted backwards, based on the intuitive principle that dArea/Area should be closer to constant. The result didn't produce lots of power but it did seem to provide a nice smooth powerband, so far as my buttocks could judge.

Yes, I did. When developing the 6,5 cc MB40 model aero engine, we encountered the problem that we had to fit very light propellers in order to pass through the torque dip just before the onset of the power band.
Once the engine ran in the power band, these propellers were much too light, so the engine ran right past the power peak onto an rpm where most of the power was not present any more.
If you want to go fast, you have to keep the engine at max. power rpm, so you need a heavier prop. But with a heavier prop you will never get past that torque dip....

I reasoned a two-stage reflector with a shallow second stage might be the solution. But developing anything on these 6,5 cc engines is a nightmare because the glow plug ignition and the coarse mixture control make it impossible to get repeatable power readings, so I performed these experiments with a 125 cc engine.
The picture shows the 125 cc test pipe, a one-stage 6,5 cc pipe and a 6,5 cc cylinder.
[Vous devez être inscrit et connecté pour voir cette image]

quote]Wouldn't it be better to work forward? I can hardly believe you already finished that book.
And before you ask, this forum is not the place for a comprehensive lecture on gas dynamics.[/quote]

I am sorry if I have offended you Frits with my question that was not my intention – please forgive me.
I asked you this particular questions because I have experienced these characteristics when I was experimenting with dual stage convergent cones of large angles. While the power improved at the peak; the overrun dropped very quickly and required large amount of retarded ignition timing get it to over rev even a little. The EGT also changed a lot at this RPM by more than I would have expected and detonation was experienced. It seemed to me that something had caused the wave pattern in the pipe to become unstable, maybe a shock wave as you had mentioned.

Again, it was not my intention to offend you; I was just very interested in the subjects that are present in this forum.

Frits Overmars a écrit:

Exhaust 202°, A-transfers 130°, B- and C-transfers 132°.

Hi Jan,
In regards to the exhaust port timing of the RSA given generously by Frits; did the aux exhaust ports open at the same time as the main exhaust port? The main exhaust port has a lot of duration and it surprised me that you could get such good bottom-end power from that duration. Can you tell us the rpm range that the servo open the power valve over please?

Many thanks.

The aux. exhaust ports were 2mm lower than the central one.
The exhaust valve was fully open at 12.000 rpm.

Thanks Jan.
The "A" transfer port is lower in timing than Transfer "B and "C" which always seemed to make more sense to me when you take in account that the exhaust port is evacutating the cylinder. What was the main advantage of delaying opening of the "A" port; over rev?

Many thanks.

GtG001 a écrit:

I am sorry if I have offended you Frits with my question that was not my intention

No need to apologize, Allan; I did not feel offended at all. I just tried to make clear that one can only dive so deep into gas dynamics in a general formum. Well, judging from your reaction I may have made it a bit too clear .

Frits Overmars a écrit:

I may have made it a bit too clear .

Unfortunately, sensitive people are usually offended before the insensitive ones have even noticed a hint. I try to be sensitive on Mondays, Tuesdays and Fridays and insensitive the others...

The A-port was not lowered, but the B-port was raised 0,5mm.
This gave a bit more power at low revs.
Overrev remained the same.
During 10 years of development the port heights remained always the same.
But the widths changed a lot!

Jan what correction method did you use when measuring the horsepower?

Hi Frits ,
What criterior do you apply for proportioning the dimentions of a pipe diffuser , a single taper seems reasonably straight forward
but a multi stage with differing angles and lengths are a bit of a puzzle . Any guidance will be most gratefully recieved .

Regards , Melvyn

Frits Overmars a écrit:

...this forum is not the place for a comprehensive lecture on gas dynamics.

Frits Overmars a écrit:

....I just tried to make clear that one can only dive so deep into gas dynamics in a general formum.

melvyn trevor a écrit:

Hi Frits, What criterior do you apply for proportioning the dimentions of a pipe diffuser , a single taper seems reasonably straight forward but a multi stage with differing angles and lengths are a bit of a puzzle . Any guidance will be most gratefully recieved .

Somehow I get the impression you guys just don't read....

Formulating an adequate answer to a 'simple' question like yours would take days. Days that I haven't got .

Oh dear,
Apologies for my lack of erudition Frits, but i`m just a beginner with much more enthusiasm than knowledge !
However, i do try my best to understand these complex topics

Thankyou for the reply, regards Melvyn

Jan Thiel a écrit:

The A-port was not lowered, but the B-port was raised 0,5mm.
This gave a bit more power at low revs.
Overrev remained the same.
During 10 years of development the port heights remained always the same.
But the widths changed a lot!

Thanks Jan for the reply.
For me, to be able to ask questions on this forum helps me unravel a little more of the mystery about 2 strokes that did appeared veiled in some areas before.
There are not many people who have the knowledge you and Frits have, and even less that will share it. I am sorry if our endless questions are causing you and Frits to be uncomfortable.

We are all hungry for the truth and advanced knowledge that you have.
Is it possible that you could write an article each month on 2-stroke development and sell a subscription to those who are interested? Maybe Marc could host it?