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

Hi guys, I'm italian, my name is Francesco and I'm 27 years old; I'm studying engineering mechanical, working in my spare time with 2t engine and I follow this thread with great interest.

I've asked to Frits if he could help me (us) to project completely an head for our engine, how squish velocity, shape head combustion chamber, RDC, distance between spark plug and piston, exhaust timing, octane number in our gasoline are related, how we could choose the better parameters for our better use and if there's some matematichal equation that we could calculate ourself.

Thanks a lot Frits, you're help it's always very important.

I had already started writing an answer to the private message Francesco sent me, but then I realised that there may be many more forum members with similar questions, and when each of them starts asking their questions via private message, answering all those questions will use up far more time then I have available. Therefore I asked Francesco to repeat his question here, in the forum, so that I can give an extensive answer and at the same time avoid getting the same question over and over again from other forum members.

@Cocco83: I see that you joined the forum on 14/12/2010, so you may have missed what I wrote three weeks earlier, on 24/11/2010, about squish, turbulence and flame speed. You can find it here:
[Vous devez être inscrit et connecté pour voir ce lien]

The 1% squish gap I mentioned, is a guide number; you will also need practical experience. When I researched how far (or better: how close) you can go, it turned out that the old Rotax-124 kart engine happily sustained 0,4 mm at a mean piston velocity of 26,1 m/s (14.500 rpm with stroke 54,0 mm) whereas a Kreidler piston already hit its head at 21,2 m/s (16.000 rpm with stroke 39,7 mm), if the squish gap went below 0,7 mm. The main reason for this difference was the type of crankshaft bearing in the Kreidler: very light-running, but not very stiff...

The higher the squish velocity, the better. Combustion should be as fast as possible (apart from detonation) and squish turbulence helps.
So are there no negative sides to a high squish velocity? In theory no; in practice the ignition may not be up to it. I will try to explain later; first I need something to eat...

1/ cocco83, please, to be fair: take your time, read the full topic and avoir MP to ask for favor as soon you arrive...



2/ Frits, realy, you are a very nice guy...

3/ 26,2 m/s ? Why everebody write that 25 m/s is a maximum?

Hi Marc, yes I've read all the topic some days ago, sorry it's not my intention harm this thread and harm Mr Frits :)

Thanks Frits, I've just read also that part of the topic (and it has been very illuminating) but my major question is.... If I have to start the draw of an head from zero, how can I make the better project quite immediately, how have I to decide between the lots of shape combustion chambers, how can I predict the squish velocity quite good without expensive instrumets?

They are lots of questions, sorry but I'm very curious.

But first... good dinner :)

Hi, I would like to ask Frits what method of measuring the squish you use as I have worked with the rotax 124
and not managed to acheive below 0.6 without signs of the piston touching the head. I use solder over the piston pin area of the crown.
Your posts are extremely informative.

pfpraider a écrit:

Hi, I would like to ask Frits what method of measuring the squish you use as I have worked with the rotax 124
and not managed to acheive below 0.6 without signs of the piston touching the head. I use solder over the piston pin area of the crown.
Your posts are extremely informative.

solder is effective. nevertheless please bear in mind that the thinner solder you use themore accurate the value will be. solder is a bit 'elastic' after deformation. this could result in faulty measurement, especially the version with the harsh kernel.
for example you measure 0.6mm using a 0,75 thick wire, whilst you determine 0.7 when using a 1.5mm thick wire.

whereas the real value can easily be 0.55.

Marc a écrit:

Frits, realy, you are a very nice guy...

Forumrule # 1: the Admin is allways right

Citation :

26,2 m/s ? Why everebody write that 25 m/s is a maximum?

I don't know. In the last year that engine rpm in Formule 1 was still free, Cosworth reached 21.000 rpm with a 39,7 mm stroke (they all have that same stroke now). That gives a mean piston velocity of 27,8 m/s. But that is not the limit (I do not think there is a fixed limit). A friend of mine (BVE) prepares speedway engines: 500 cc methanol-guzzlers with bore and stroke 86 x 86 mm. They regularly run 12.000 rpm but some riders (probably heavily sponsored) rush them to 13.000 rpm without anything breaking. That's 37,3 m/s......

pfpraider a écrit:

Hi, I would like to ask Frits what method of measuring the squish you use as I have worked with the rotax 124
and not managed to acheive below 0.6 without signs of the piston touching the head. I use solder over the piston pin area of the crown.
Your posts are extremely informative.

Thanks for the flowers, pfpraider.
I too use solder: 0,8 mm wire with hollow resin core because that is softer and does not put so much strain on the big-end bearing when it gets flattened. I use two pieces of soldering wire simultaneously, left and right above and parallel to the piston pin, so the piston won't tilt in any way. That gives you the static squish gap.
To find the dynamic gap you need some nerve: just put it together, run it and open it up. My rule of thumb is: 'If I cannot clearly see that the piston has kissed the head, everything is fine'.
I will say a bit more about this later; I am preparing some more text about squish.

Frits Overmars a écrit:

The higher the squish velocity, the better. Combustion should be as fast as possible (apart from detonation) and squish turbulence helps.
So are there no negative sides to a high squish velocity? In theory no; in practice the ignition may not be up to it. I will try to explain later.

So here goes.
You must be able to rely on the ignition, even when the electrode gap has gotten on the big side, when the mixture is so weak, rich or diluted that you can hardly ignite it, or when the compression end pressure is high because of good cylinder filling or a high compression ratio.
All parts of the ignition system are important: the spark plug of course, but also the plug lead (never use carbon core leads), the plug cap (never use resistor caps; if there has to be a resistor in the high tension line, it sould be inside the plug, not in the cap, because caps fail), the coil (with Jamathi, Bultaco and Garelli the fat, blue, oil-filled Bosch coils were used despite the fact that they were almost as heavy as the engine. Replacing them with slim Japanese coils cost several hp...)

Let's take a look at the spark plug. What exactly happens there when the coil sends its high tension voltage? The gas molecules between the electrodes are ionized by the high potential. Starting at one of the electrodes, the ions begin to form a conductive channel through the otherwise non-conductive gas and as soon as this channel reaches the other electrode, a connection is established and a current starts to flow: the spark.
The ionisation needs a certain voltage per mm electrode gap, dependent on temperature and pressure of the gas. The squish turbulence can blow the ions away before they have had a chance to form a channel from one electrode to the other. Then, when the channel is finally completed, it is no longer straight but arched, also longer. Because of the required voltage per mm you then need a higher ignition tension.
Some time ago Honda RS125s would not run with more than 0,4 mm electrode gap, whereas Aprilia used 0,7 mm gap. Then Honda offered a kit with a different (better!) ignition system...

Talking about better ignition systems: our fellow forum member Emotracing offers something nice:
an ignition extension. The normal ignition takes care of the ionisation channel; then as soon as the current starts to flow, the Emot-set sends a big extra current through this channel. It looks like this: [Vous devez être inscrit et connecté pour voir ce lien]

And if you really want to know all about sparks, look here:
[Vous devez être inscrit et connecté pour voir ce lien]

There is one other possible objection against effective squish. When a gas comes in contact with a solid surface, it tries to cling onto it: it forms a boundary layer. This layer of stationary gas is a fairly good heat insulator. The piston crown enjoys this insulation, as the boundary layer protects it from the combustion heat. In theory the squish flow can blow the boundary layer away, exposing the piston directly to the combustion.
But there is always a far greater danger: detonation. The shock waves in a detonating engine are a thousandfold more effective in shaking the boundary layer loose. So I prefer a good squish that speeds up combustion.

If detonation occurs, it happens in the end phase of combustion, when the pressure in the combustion chamber is high and the temperature of the unburnt mixture in the edges of the chamber is raised by the heat radiation of the burning mixture.
Effective squish flings burning parts of the mixture to the edges of the combustion chamber, igniting all mixture before detonation factors there can reach a dangerous level.
Finally a simple advise for junior tuners: make the squish gap so tight that there is no more room for mixture; if it is not there, it cannot detonate 8)

yeah, so if I've correctly understood I measure the squish distance in coold engine in order to have approximately the 1% of stroke with the right tin thick, mount all, used it and remeasure the squish a hot engine? And verify if there's or not squish distance between head and piston right?
So although squish distance is very low we do not be afraid of detonation or preignition along the squish crown.... or better.... If we take an 1% of stroke as squish distance and during the functioning the engine will stretch some parts, this distance will trend to approximately zero and this is fine if we haven't contact between the two parts.... So if I make correctly this and I see that around the squish crown I find detonation I've to find cases elsewhere right?

sorry, I've posted after your explanations.... now I will read your response dear Frits.

Ok, fantastic, I've read all.... perfect explanation, and great two final lines.... no space for gas, no danger for detonation.... simple and efficacious. The problem is to make this space thinner and thinner.... piston has it's camber, and it's difficult to have the same profile along the squish band

perfectly the same profile might be useful but maybe not needed, when you go really squish tight the high surface/areavolume ratio should keep end gases from detonation.
at a given pressure and temperature, detonation occours after a certain delay so if you don't squish enough you have to burn them before they become critical.

Frits
Do you think that the torriodal shape has less cyclical variation(less variation between a 'good' and 'bad' combustion cycle)? Allowing more consistent jetting and ignition timing
I have seen a few CFD simulations, that show a great deal of turbulence (small fast spinning eddies) coming off a small radius at the inner edge of the squish band.

Frits pensez-vous que le torriodal forme a moins variation cyclique(moins de variations entre un «bon» et «mauvais» cycle de combustion)? J'ai vu quelques CFD les simulations, qui montrent que beaucoup de la turbulence (petit rapide filature tourbillons) venir off Fr

Institute of TwoStrokes a écrit:

Frits, Do you think that the torriodal shape has less cyclical variation(less variation between a 'good' and 'bad' combustion cycle)? Allowing more consistent jetting and ignition timing.

The Toroidal shape has proven to give better power; that might come from fewer bad combustion cycles. It can be measured, but I have no relevant data available.

Citation :

I have seen a few CFD simulations, that show a great deal of turbulence (small fast spinning eddies) coming off a small radius at the inner edge of the squish band.

CFD programs have improved tremendously over the years, but the saying still applies:
garbage in = garbage out; it all depends on what you feed these programs. I do believe in these
micro-eddies at the inner edge of the squish band though; it is mr. Coanda making himself noticed.

At Aprilia Jan Thiel performed a series of tests with different radii at this edge. The outcome was crystal-clear: R=0 was best and the bigger the radius, the worse the result. It is all the more surprising that Honda has always stuck to a radius (sometimes as large as 5 mm).

Thanks a lot Frits (and nasone too), this is an important answer to start making something good :)

Hi guys,
until now I have not had time to continue at the 24/7 valve project.
But I am shure that I can surprise you with something beginning the new year.
So I want to wish you all a merry christmas and a happy new year

Hi Michael, I hope you are right: surprise us!
For a systematical research it may be best to start with a smaller carburettor (maybe 28 mm instead of 35 mm) and a long intake tract between carburettor and valve. Then you can search the minimum rpm where the engine will run with the valve open. From there you can then shorten the intake tract to synchronize intake and exhaust.
EDIT:
With the valve open the crankcase is filled in 360° instead of <300°, so the mean flow velocity through the carburettor will be lower. Therefore you may need much bigger main and needle jets.
Or maybe you can play with the choke
Merry Christmas and a happy New Year to all
[Vous devez être inscrit et connecté pour voir ce lien]

Citation :

Merry Christmas and a happy New Year to all

Same for you and your family !!!
and all of pit-lane members

merry christmas to all!!! and thanks everybody for the interest in this passion

cocco83 a écrit:

merry christmas to all!!! and thanks everybody for the interest in this passion

this forum is like 'Hotel California". you can check in or out any time you like but you can never leave.
all the best for 2011 everyone - thanks for sharing so much knowledge and ideas.

cocco83 a écrit:

merry christmas to all!!! and thanks everybody for the interest in this passion

thank you to all for the research, i think we can develop faster than the factories who are stopped developpements [Vous devez être inscrit et connecté pour voir cette image]
SO i wish you all the best for the next Year [Vous devez être inscrit et connecté pour voir cette image]

Frits what's your opinion on high inertia crankshafts(Aprilia has lots of tungsten inserts)? I know they even out the crank speed variation, effectively increasing port durations, but I have recently found they have the inertia to over come the high compression 'brickwall' that seems to limit rpm?

I also remembered I had put a version of a 24/7 inlet in a few old piston port engines(open 150deg ATDC) an just put one into a racing chainsaw engine.I'll post some pics soon

Frits quel est votre avis sur la haute inertie vilebrequins? Je sais qu'ils même hors la manivelle variation de vitesse, accroître effectivement port durées, mais j'ai récemment découvert qu'ils ont l'inertie à plus venir la compression élevée «brickwall» qui semble limiter

J'ai également rappeler j'avais posé une version d'un 24/7 inlet dans quelques vieux piston port moteurs(ouvert 150deg ATDC) un vient de poser un dans une course tronçonneuse moteur.Je vais poster quelques photos bientôt

Institute of TwoStrokes a écrit:

Frits what's your opinion on high inertia crankshafts(Aprilia has lots of tungsten inserts)? I know they even out the crank speed variation, effectively increasing port durations, but I have recently found they have the inertia to over come the high compression 'brickwall' that seems to limit rpm?

I also remembered I had put a version of a 24/7 inlet in a few old piston port engines(open 150deg ATDC) an just put one into a racing chainsaw engine.I'll post some pics soon

Frits quel est votre avis sur la haute inertie vilebrequins? Je sais qu'ils même hors la manivelle variation de vitesse, accroître effectivement port durées, mais j'ai récemment découvert qu'ils ont l'inertie à plus venir la compression élevée «brickwall» qui semble limiter

J'ai également rappeler j'avais posé une version d'un 24/7 inlet dans quelques vieux piston port moteurs(ouvert 150deg ATDC) un vient de poser un dans une course tronçonneuse moteur.Je vais poster quelques photos bientôt.

The Aprilia 250 twin needed high-inertia cranks because of its 0°-firing order (see [Vous devez être inscrit et connecté pour voir ce lien] ), not to increase port durations.
With a nominal engine rpm of 13000, the rpm fluctuation within one crankshaft revolution lies between 12970 and 13030 rpm. That is +/- 0,2 %. With heavier cranks maybe you could gain another 0,1 % of port duration; hardly worth risking your gearbox for (the gearbox gets the most punishment from downchanges, and heavy cranks aggravate that).

The 'high compression brickwall' is no longer as hard as it used to be. With leaded fuel the compression ratio was 19,5; nowadays it is about 15. The 'brickwall' arose from the high expansion ratio (which is identical to the compression ratio). The higher this ratio, the more the burnt gases in the cylinder cool down before they enter the exhaust, thus lowering the resonance frequency of the exhaust system.
Riders complained that the engine would not rev, especially not in the lower gears, where the revs rose so quickly that the rising wall temperature of the exhaust pipes could not keep up with them. So you had cold exhaust gas and cold pipes.
The problem was to a large extent solved when the solenoid-controlled on/off-power jets were replaced with stepper motor-controlled jets that could continuously adjust mixture strenght, allowing the engine to rev more freely.

Regarding the Aprilia RSW/RSA125 single, Jan Thiel told me that in seven years of testing the racing department had not been able to establish which was best: low or high inertia. I would choose low...

"stepper motor controlled jets" is that the DAS system?

This is it. The stepper motor itself originates from a Fiat Uno where it regulates the idling rpm

[Vous devez être inscrit et connecté pour voir cette image]

We always learn with you speak Frits, silly me I thought that was just a plain solenoid

Nous apprenons toujours avec vous parler Frits, idiot moi, je pensais que c'était juste un solénoïde plaine

You are not that silly, Lawrence. Not too long ago is was a plain solenoid, like this:
[Vous devez être inscrit et connecté pour voir cette image]