[GP125] All that you wanted to know on Aprilia RSA 125, and more, by Mr Jan Thiel and Mr Frits Overmars (PART 4)

GtG001 a écrit:

Thanks Jan for your reply,
The loss of max revs might indicate that it pulled more charge out of Transfer A by being longer and cooled the pipe too much?
Maybe, the longer wall reduced turbulence and thereby increased the suction of the exhaust wave?

Regards
Allan.

It is a long time ago I did these tests! 1996
Until a certain number of revs the engine run very well, with good power.
And then it completely stopped firing at all, quite suddenly.
I do'nt remember if the exhaust temperature changed, but I think it did, with the engine not firing anymore
it would be logic!
The important thing would have been to see if the temperature was lowering before the engine stopped.
I really do'nt remember this.
The more I shortened the divider wall the better it revved.
At the time the auxiliary ports were not so big as they were to become later.
And I had to work in a hurry as I had to produce a lot of cylinders.
Another thing I remember was that on the flowbench I saw more flow when I flowed the main exhaust and the
auxiliary ducts separately.
But when I flowed them all together the flow was less!
So I think the flow of de various ducts disturbed each other.
And that the total exhaust flow became insufficient at a certain No.of revs.

So I had to accept the short divider walls as they were the quickest solution...
No time to really investigate everything!

Thanks Jan for the detailed account.
Another puzzle yet to be solved
Maybe, someone with CFD expertise will solve this for us

i worked on a few honda bridged exhaust style cylinders but i never thought to shorten the divider. i wish i would of atleast tried it once. i wonder if it would have even less dimensional stability from heat and push significantly into the bore causing a piston seize ?

Jan Thiel a écrit:

Yes, that is indeed the case!
There is a very big twist in the transfer tunnel.
The exhaust side of the A-transfer was directed as much away from the exhaust as possible!

Can you tell us where it met the cylinder centerline?

Haufen a écrit:

Jan Thiel a écrit:

Yes, that is indeed the case!
There is a very big twist in the transfer tunnel.
The exhaust side of the A-transfer was directed as much away from the exhaust as possible!

Can you tell us where it met the cylinder centerline?

I do'nt know, I never looked at that!
I never cared much about such things...

Hi Frits (and everyone else).
I have a question about your FOS scavenging concept.
(For everyone that had not saved the pictures to their HDD, I felt free to upload them again - so they do not have to search  

"old" scavenging modell:
[Vous devez être inscrit et connecté pour voir cette image]

new scavenging modell:
[Vous devez être inscrit et connecté pour voir cette image])

I wonder about what you said some days ago:

Frits Overmars a écrit:

[...]
By the way, instead of looking at the scavenging angles in my Pisa-story, look at the FOS scavenging concept I posted not too long ago. Those angles are more up-to-date.

I am just curious and ask myself: What would be the difference of the powercurves of two engines that where complete identical except for having the old/new scavenging modell.
Or the question more simplificated: What made you change the "old" scavenging modell to the new one?

Greatings and thanks a lot!
Tim

Tim, the first picture, the one that you call the 'old scavenging model', was not a scavenging concept but a representation of an existing 6,5 cc model engine. When I posted it, I explained the reasoning behind its geometry, especially regarding the axial scavenging angles, and why these had to be changed afterwards. Below you'll find the current scavenging concept.
[Vous devez être inscrit et connecté pour voir cette image]

Many Many thanks for all this information. If I understood well these are the scavenging angles of a 6,5cc nitro methane engine. Do they differ a lot with a 125cc petrol race engine?
Frits where you able to calculate the entrance speed thru the scavenging ports combined with the opening angles?

Maurice Specken a écrit:

Many Many thanks for all this information. If I understood well these are the scavenging angles of a 6,5cc nitro methane engine. Do they differ a lot with a 125cc petrol race engine?
Frits where you able to calculate the entrance speed thru the scavenging ports combined with the opening angles?

Hello Maurice, welcome to Pit-Lane.
The first picture posted by Tim Ey on top of this page was from a 6,5 cc engine. But that was not a nitromethane engine; it runs on methanol mixed with 20 (!) % oil, as prescribed by the F3D technical regulations.
The scavenging concept I posted in my above answer is supposed to be universal. Only the photo on the right still shows the cylinder of the 6,5 cc engine, just because it represents a good comparison with the concept angles' drawing.

The flow velocity through the transfer ports not only depends on the position of the crankshaft, but also on the rpm (and obviously also on the throttle position). So I cannot quote a simple number; I would have to present a three-dimensional graph. Calculating it involves some serious gas dynamics and my engine simulation program does not generate this graph as an output item; I would have to reconstruct the program in order to do that.

Now how about a little introduction of yourself?

Thanks Frits,
I didn't expect it was an easy question and answer, I was just wondering if somebody did take the challange already. Intoducing myself wil be à long story, if You look at the website of my son.
[Vous devez être inscrit et connecté pour voir ce lien] you will find were my interest is coming from.
My whole life I'm playing with two stroke engines, from model helicopter motors to racers but never on competition level like last years. Also scouting a lot on the Dutch websites like raceheldenvanweleer and visiting old racer meetings frequently (last one Capella a/d ijssel). Getting more humble by reading all this information but having a lot of joy when questions are clarified. I have my own test bench now which gives me a lot of fun ( even more than the races). Now if I could find some more time without having to much trouble with the wive, it would be the ultimate hobby. Just like you I like to solve and understand challenges with mathematical tools to find solutions. This is why I asked You if You have tried to calculate flow velocity at the scavenging ports.
Many thanks to everybody on this forum

Frits Overmars a écrit:

Below you'll find the current scavenging concept.
[Vous devez être inscrit et connecté pour voir cette image]

Thanks for sharing, Frits. I assume this is valid for a square engine layout? Also, would you say opening the A ports last, so that they actually flow first, because they have the longest path to travel would be beneficial in your concept? And last, are you not afraid of trapping a residual gas pocket by aiming the leading directional angle of the A port that far back?

Haufen a écrit:

I assume this is valid for a square engine layout?

Yes, but the differences with other bore/stroke-ratios are small and they concern only the axial angles.
Here is their complete story:

A-port: axial scavenging angle = ATN(0,46 x stroke / bore)
B-port: axial scavenging angle = ATN(0,20 x stroke / bore)
C-port: axial scavenging angle = ATN(1,40 x stroke / bore)

Citation :

would you say opening the A ports last, so that they actually flow first, because they have the longest path to travel would be beneficial in your concept?

I've seen this remark before and I can't understand why anybody would say that. The port that opens first, flows first.
Maybe not in the desired direction, when a transfer port opens before the blowdown phase has completed, but that is a question of mismatched time.areas.
And why would the A-ports have the longest way to travel? Their flow should meet the flow from the other transfers ports in the cylinder center, and the path differences are negligible.

Citation :

are you not afraid of trapping a residual gas pocket by aiming the leading directional angle of the A port that far back?

The A-ports are aimed that far back to prevent fresh charge from making a U-turn and taking the short route to the exhaust. Some of the charge will try to make that U-turn anyway and expel those end gases.

one question from my side regarding the areas of a port.

every port has a certain area, is that area kept trough the whole port from the bore side till the bottom of the cylinder? or does it increase/decrease?

koenich a écrit:

one question from my side regarding the areas of a port. every port has a certain area, is that area kept trough the whole port from the bore side till the bottom of the cylinder? or does it increase/decrease?

Usually the port window in the cylinder is the smallest cross area of any duct.

ok, so the area decreases from beginning of the duct (bottom of the cylinder) till the end, means also the hydraulic diameter decreases? do you have the areas for the rsa at hand as an example?

appreciate :)

koenich a écrit:

ok, so the area decreases from beginning of the duct (bottom of the cylinder) till the end, means also the hydraulic diameter decreases? do you have the areas for the rsa at hand as an example?

Not right now.

[quote="Frits Overmars"]

Haufen a écrit:

I assume this is valid for a square engine layout?

Yes, but the differences with other bore/stroke-ratios are small and they concern only the axial angles.
Here is their complete story:

A-port: axial scavenging angle = ATN(0,46 x stroke / bore)
B-port: axial scavenging angle = ATN(0,20 x stroke / bore)
C-port: axial scavenging angle = ATN(1,40 x stroke / bore)

Mr Overmars , in your formula calculation, i dont understand "ATN" , .......so what is this value?? 25° for b-port and 10° for a-port......

pagi a écrit:

Mr Overmars , in your formula calculation, i dont understand "ATN"

ATN(n) = ArcTangent; that is the angle for which the tangent is n. Example: the tangent of 30°=0,577, so ATN(0,577)= 30°.
Hopefully this answer will be satisfactory because this forum is not the place to go deeper into the basics of goniometry.

OK THANK YOU

Hello Jan and Frits,

In full design of an admission by rotary disc for engine of solex of competition, I like to know how to define the good timing opening and closing of the valve?
What did you note like profit and loss compared to a standard valve admission?

There is it a large difference of adjustment of the carburettor and it's diameter?

Thank you for all what you do for us, thin preparer.

granjoie a écrit:

In full design of an admission by rotary disc for engine of solex of competition, I like to know how to define the good timing opening and closing of the valve?  What did you note like profit and loss compared to a standard valve admission? There is  it a large difference of adjustment of the carburettor and it's diameter?

A rotary disc engine can produce about 10% more maximum power than a reed valve engine. The carburettor diameters can be roughly the same.
Your rotary disc should begin to open at 40° after Bottom Dead Center and should be fully closed again at 75° after Top Dead Center. Afterward you can experiment with earlier opening and later closing timings.
The above timings are milder than in the Aprilia RSA, but for Solex-racing you will need a much, much broader power band.
A trombone exhaust pipe would be ideal for a Solex. And a good clutch setup will be more important than horsepower.

...

hi frits i have a question. with inline twin cylinders spaced close together it puts the transfer tunells in a poor shape. mostly straight up and sharp turn toward the cylinder, hardly any radius on the inner tunell wall either. in a situation like this i was thinking it might be better to have the A and B ports aimed slightly higher than normal, mostly to decrease the sharp turn into the cylinder. im thinking it should also prevent the mixture from slowing down as much

if the piston top was 10* the B ports would be set between 15*-20* upward angle. the A ports would be set between 30*-35*

does all this sound logical enough to give it a try or is my thinking wrong ?

With such straight-up transfer ducts the flow will simply refuse to make that sharp turn and will tend to enter the cylinder at an axial angle of about 45° anyway. That is too steep already, so it's no good aiming the ports higher. That would only reduce their cross flow area.
Because the steep axial flow from such A-ports cannot sufficiently balance the flow from the B-ports, the B-ports must be made smaller and/or be aimed higher up to prevent them from pushing the A-flow right into the exhaust. But then the central flow column will have too much axial velocity and insufficient density.
Cramped transfer ducts are a lose-lose situation.