[GP125] All that you wanted to know on Aprilia RSA 125, and more, by Mr Jan Thiel and Mr Frits Overmars (PART 5)

Yes, very nice pictures !

About flattening the pipe, on the 1984-1986 Yamaha 500cc MX works bikes (Carlqvist's ones) they had flattened the pipe so it could also play the role of a kind of air scoop (it was still aircooled while racing against the famous liquid-cooled Honda 500 RCM). Look at this picture.
[Vous devez être inscrit et connecté pour voir cette image]
Do you think it was detrimental to performance ?

WOW. Some very very nice Cad development software being used there. 
The mind boggles at the thought of moving the inner radius wall towards the outer walls radius whilst maintaining the volumetric area of the round pipe. 
You have me thinking so much more now Frits, I would need some decent cad software to quickly obtain  volumetric calculations, I don’t fancy the thought of hand drawings & the long drawn out process of hand drawn development of a pipe & subsequent pocket calculator area calculations

You need not worry about volumetric calculations RAW. As long as you keep all cross flow areas equal to those of a round pipe, the volumes won't deviate.

Round is not what we want for a bend, and the rectangular cross flow area I drew on the previous page would have an undesirable ratio of cross flow area to wall surface, losing much exhaust gas energy to the outside world, so we may decide on an ellips.
You may remember that the cross flow area of a circle is Pi/4* diameter² and for an ellips it's Pi/4 * height*width, so you just need to make sure that height*width of the ellips is always equal to diameter² of the equivalent round pipe.

Hello all, so I find myself here wich makes me very glad, apparantly it's 8 a.m. I built my first pipe some time ago now (Gilera 80 cc moped), and contrary to Frits' it neither fell off or impeeded power, the latter was completely due to luck. Well I have a lot of questions but also a few answers after partly unwrapping this Christmas gift.

Welcome Andreas. You must be the fastest reader I ever met, starting from the link I gave you yesterday afternoon,
to this page in one night!

Haha, not at all that fast, as much as rushing into. I'll be picking up where Jan Thiel holds a grudge against uncertain ex-duct floor optimum. but I'm confident there will be more late evenings.
Much obliged.

Yes, that's more than 200 pages, very impressive !

Cheers!

[Vous devez être inscrit et connecté pour voir cette image]" />[url=https://servimg.com/view/20010268/1][Vous devez être inscrit et connecté pour voir cette image][/url
Would it help scavenging to increase the c- port angle with retained flow area? I said I had a lot of questions, they have mostly been answered here, I particulary found the 8.72 value helpful. How did the experiment with the dual- steep to shallow cone baffle work out? Thank you again Frits.

no, you would lose much needed blowdown by opening the transfer way to early, resulting in exhaust gases still under considerable preasure entering the crankcase and weakening your exhaust pulse in the pipe.

Only the cylinder outlet would open early, the opening to the crank case would open simultaniously with the other transferports. I.e. the window in the piston will determin the timing.

oh yes, didn't thought of that. still : the flow from piston through port would only be very shortly turbulence free at BDC.
I think you will gain time-area at the C-port itself, but would flow less through the whole C duct.

The window would of course be as wide and probably as low as possible, but yes I also think it would flow less.

Hi Frits or suited to answer.
Having a greatly oversquare engine (66x50)and giving it the same time x area as an square one, would the formulas given above be enough or do I need to compensate even more for the 'flattened' space? I saw your calculations for the Maico, but wasn't sure if I could draw the answer from that.

andreas länström a écrit:

Having a greatly oversquare engine (66x50)and giving it the same time x area as an square one, would the formulas given above be enough or do I need to compensate even more for the 'flattened' space?

In an oversquare engine the ratio of port areas to cylinder volume cannot be as good as in an optimized square engine.
You can compensate for this by giving the ports more open time. Not by enlarging their timing, which could have undesirable consequences, but by lowering the rpm. Then you may achieve the same torque, but because of the lower revs you will not achieve the same power.

We need to look at our definitions here. Time.area is the sum product of port open time x port open area, and the time depends on engine rpm. In other words: time.area is undefined until the rengine is running.
Instead of time.area we should use angle.area: the sum product of port open angle x port open area.
Angle.area is a true constant value, defined by crankshaft stroke, con rod length and cylinder layout.

Now the answer to your original question: you need to compensate for the 'flattened' space by adjusting the axial scavenging angles. Here are some guidelines.
[Vous devez être inscrit et connecté pour voir cette image]

Thank you Frits. I perfectly follow what you said. My concern was that a less 'spherical' chamber would take longer to emty or fill, but either it doesn't or the effects are negligable!
Cheers Andreas.[Vous devez être inscrit et connecté pour voir cette image]
Low rpm indeed.

Idea to aid scavenging in the 360 exhaust engine.[Vous devez être inscrit et connecté pour voir cette image] And oversquare for the sake of scavenging efficiency.
And also a tuned length variator.[Vous devez être inscrit et connecté pour voir cette image]

Frits Overmars a écrit:

We need to look at our definitions here. Time.area is the sum product of port open time x port open area, and the time depends on engine rpm. In other words: time.area is undefined until the rengine is running.
Instead of time.area we should use angle.area: the sum product of port open angle x port open area.
Angle.area is a true constant value, defined by crankshaft stroke, con rod length and cylinder layout.

Is there a way to determine an A.A number to shoot for ?

my simple brain suggests :

A.A = STA(rsa)*RPM*(Swept Vol + Chamber Vol)/1000 where STA is from the RSA and the rest from our own engine.

but then a slightly more inteligent part of my brain suggests that would be dumb as one might as well use the STA number instead of the A.A number  
or should we multiply by RPM of RSA ?

Frits Overmars a écrit:

Now the answer to your original question: you need to compensate for the 'flattened' space by adjusting the axial scavenging angles. Here are some guidelines.
[Vous devez être inscrit et connecté pour voir cette image]

are these guidelines universal, by which I mean independant of the type of stagger used ?
and I presume they are only for high BMEP engines ? I mean : for my lowerish BMEP'ed Mx 50cc moped running only about 8500rpm max, I presume it is better to have lower exit angles to give the gasses a longer path to travel ?

JanBros a écrit:

Frits Overmars a écrit:

We need to look at our definitions here. Time.area is the sum product of port open time x port open area, and the time depends on engine rpm. In other words: time.area is undefined until the rengine is running.
Instead of time.area we should use angle.area: the sum product of port open angle x port open area.
Angle.area is a true constant value, defined by crankshaft stroke, con rod length and cylinder layout.

Is there a way to determine an A.A number to shoot for ?

Yes there is. You'll find it somewhere in this thread; it has been quite some time ago. But you'll have to search for it yourself; I'm way behind with my work because of the holidays.

JanBros a écrit:

Frits Overmars a écrit:

Now the answer to your original question: you need to compensate for the 'flattened' space by adjusting the axial scavenging angles. Here are some guidelines.

are these guidelines universal, by which I mean independant of the type of stagger used ?
and I presume they are only for high BMEP engines ? I mean : for my lowerish BMEP'ed Mx 50cc moped running only about 8500rpm max, I presume it is better to have lower exit angles to give the gasses a longer path to travel ?

I try to avoid stagger at all. And you are more or less right about low BMEP and longer flow paths. But low-BMEP-engines usually have lower pressure differentials, hence lower flow velocities, which more or less compensates for the lower piston speeds that these low-BMEP-engines also usually have.

Jan, there was an error in the picture showing the axial scavenging angles. I substituted the picture in my post with the correct one. Could you please delete the erroneous picture in your above post? Thanks in advance.

tnx, and edited the pic ;-)

JanBros a écrit:

tnx, and edited the pic ;-)

Jan Thiel a écrit:

The pipe as fitted to the bike usually gave slightly more power than the straight pipes used for testing.
Bends in pipes must always be tested very carefully, as they can destroy power quite a lot..
Of course RSW pipes were tested on the RSA, and vice versa.....
The RSA pipe proved to be a bit worse, and caused some more detonation.
I think due to the necessary longer 26mm tailpipe...
Without this tailpipe there was no detonation at all!
The pipe under the engine heated the crankcase.
So a carbon shield was made to prevent this.
A rear exhaust would gave been better of course.
But that was quite impossible as the carburetor was already there....

Did anyone tried to fit the slightly curved pipe on different axial angles? Like turning the pipe 45degrees on the collet and map the curve, repeat +45deg...etc till reached at least 180deg then see all the graphs together?

Andreas......thought you would find this interesting regarding your "tuned length variator"......https://patents.google.com/patent/US3703937A

This was a good solution at the time, but for sure can be done better today.

Look forward to seeing your progress.

Yes. I see It's already been thought of. How very beautifully displayed! Perhaps it's a good idea to have the valve arched in order to not alter the cross-section area too much. Originaly I thought the valve should close at a 90 degree angle, but I understand this will create a too strong/inefficient return pulse.

Regards Andreas

Hello, I am trying to understand more in regards to 3d mapping
For example ignition maps for each individual gear including both power valve maps & pulsed power jet maps.
Would anyone have some real race information showing a complete set of maps for a given circuit & a set of maps for a different circuit to enable a comparison to better understand how gear ratio / circuit loading differences require a completely different setup of pararamaters for the ignition in each gear, power valve opening & fuel delivery via the pulsed power jet
Kind Regards Rick