Wanna Talk About Port Shapes?

[jar944]

Is this from that half blocked (non english language) link ?
If so, Any chance of seeing the pics?

Pretty Please, for this barely monolingual, untraveled Yank?

ill see if i can get them to display, but in teh mean time... here is some more nuggets of info from the other half of the dominating Overmars/Thiel team (BTW I didnt put this summary together)


A few words of wisdom from Mr Thiel


"We did not flow the principal and secondary transfers separately.
In fact flowing the transfers makes little sense!
The important thing is more their direction and how
they influence upon each other."


"No pressure transducers were ever used in our engine development.
And time/area was never calculated.
The port timings remained practically the same during 15 years!
What we did was trying different angles and radiuses, mainly on
the transfer ducts. I think we tried 40 different types of transfer ducts
that did not chanche the time/aerea. It was all about in which direction
the charge entered the cilinder and how the tranfer streams influenced upon
each other! Also about 200 different exhaust pipes were tried. After 2004
nothing much was changed but we improved with different power jet and
ignition mapping. It seemed nearly impossible to improve the transfer ducts
any more. The exhaust ducts were CNC machined, using different programs,
mainly to reduce exhaust duct volume. Also about 100 head designs were tried.
This work was done over a period of 12 years.
In september 2004 the design of the RSA started, at Derbi, and the first dyno
test was in october 2005. In the meantime Aprilia had been bought by Piaggio
which also owned Derbi, so we could use Aprilia cilinders.
The engine went very well almost at once!"

" Very little computer simulation was used, as far as I remember only for the cilinder and head cooling.
We once had an exhaust design software on loan from a very well known company, the results were useless! It was mostly cut and try!
Cilinder development was done without any computing, just logical thinking.
The exhaust ducts were CNC machined.
All improvements were fully understood."

"We never had pistons seize during our steady state tests.
Working on the dyno continuously 5 days a week!
I am 100% convinced our engine could have run for 6 hours at max power without seizing.
The problems arise when you close the throttle, or run part throttle!
The piston is mainly cooled by the transfer flow.
And at part throttle there is less transfer flow, causing detonation (auto ignition)
The entering fresh charge is ignited by the remaining, hot, burned gases!
You can see very severe damage to the piston after maybe 10 seconds at 20% throttle.
This still is an unresolved problem! I was thinking about a way to reduce engine power without closing the throttle. But how can you do this? I did not find a solution before I retired.
And nobody else was really interested.
At 100% throttle the engine was undestructible!
By making the transfer ports as wide as possible we had very good piston cooling"

"The part-throttle detonation was caused by the scavenging pressure dropping soo low that burnt gases entered the transfer ducts, notwithstanding a blowdown time.area that was sufficient for full throttle at high revs. It was not a situation that could have been cured by altering the ignition timing"

"we tested various oils and fuels, there was not much difference.
I think they were all 'copied' from ELF.
1:20 was used, we tried more and less oil: no big difference!
Long ago, at Bultaco, I tested Bell Ray oil 1:50.
The result was 2 seized big-ends in 1 day!
Later we used MOTUL with very good results"


"The most time was spent narrowing the bridges in the exhaust duct.
These could not be CNC machined:the inside was unreachable.
Preparing for plating took about half a day.
And preparing for dyno testing another half day.
A cylinder 'untouched' after plating would give 1,0 to 1,5 Hp less.
All cylinders destined for 'works' riders were dyno tested.
And reworked in case of missing HP.
We could dyno at maximum 3-4 cylinders a day.
Difference between 'best' and 'worst' cylinders was about 0,4 HP.
There were 2 people grinding cylinders daily.
I only did necessary corrections personally.
There were 2 dyno's working every day.
About 300 cylinders were made every year.
And apart of this we also constantly tried to improve HP.
Mostly batches of 30 cylinders were cast, 25 'as before' and 5 with small changes.
Between casting and dynoing took about 3 month's time.
We also dynoed replated cylinders.
They were almost never as good as new one's!"
rsss396
08-04-2013, 11:36 AM
Flow testing thetransfers proved a bit useless.
So we made many different types and tested them.


Why do you want tomake such a short stroke engine?
Square bore and stroke dimensions would be FAR better!
It gives you more port surface, which is power determinating.
And with a short stroke the piston will become very hot.
Personally I would make a slightly long stroke engine.
I cannot see 1 single advantage for a short stroke!

Power is mainlydetermined by blowdown, which is better with a triple port.
A bridged exhaust is wider at the bottom.
Which means the transfers have to be smaller.
This gives less power


When you havesufficient blowdown there is no exhaust flow anymore when the transfers open.
I suppose this happens at max. torque
After max power the blowdown becomes insufficient andexhaust gases penetrate the transfers.
You can see this when you take off the cylinder: thetranser ducts become black.
Some of the fresh charge goes into the exaust, this helpspiston cooling.
But when it becomes too much exhaust temperature becomeslower and you lose power and revs.
Until at a certain no. of revs the engine completelystops.

 

[jar944]

continued


Normally highcrankcase cv. likes short timings and low crankcase cv. likes longer timings.
I think there is also a limit of how low you can go, butI do not know where this limit is!
The flow from the inlet into the transfer ducts is veryimportant but little known I think.
It is impossible to simulate on a flow bench!
And difficult to understand, with the moving conrod,piston and crankshaft.
One thing I am sure about is: always give priority toflow over crankcase volume!


We sometimesmeasured under-spark plug temperatures, they proved too high so we modified thehead insert,
bringing the water nearer to the plug. This was very succesful, afterwards wenever bothered with the sparkplug temperature anymore!

The best way towork on plated cylinders is by using dental diamond burs.
There were never any problems with the piston rings withunchamfered transfer ports.

Chamfers at thetransfers gave less power.
A radius at the exhaust was better.
But this also depends on your exhaust pipe.
Many years ago we used unchamfered exhaust ports on the50 and 125/2 engines.
And kept the exhaust port as low as possible.
I discovered this did not work on a 125 Aprilia cylinder.
Maybe because the exhaust pipe was already made for ahigh exhaust port!
A radiused exhaust port flows better, but I think a sharpedge may give a stronger pressure wave.
You only arrive at the best compromise by testing.

You should always prepare your cylinders yourself beforeplating.
NEVER permit the plating shop to do this!
Because they have NO idea about the importance ofchamfers etc.
Once Gilardoni destroyed my best 50cc cylinder this way!
30 years ago, I still get angry when I think about it!



First the bottom ofthe C-port was lower than the pistonring at BDC.
Then the C-port was made narrower, to be able to make theB-ports wider.
The idea was that the near horizontal B-ports would flowmore than the steeply upward angled C-port.
This gave a good improvement, and was done in 1997.
Later the bottom of the C-port was raised until nearlyBDC
We made 4 cylinders, just to try.
They gave better power and caused no problems at all.
This was done in 1999, halfway through the season.
After that all cylinders were made like this.
The bottom of the C-port is not quite as important as thebottom of the A and B-ports.
Because of its upward direction it does not cool thepiston much!
A still narrower C-port was tried, (12mm) but this lostsome power.
Also a cylinder without C-port, and still wider, bridged,B-ports was tried, with negative result.
rsss396
08-04-2013, 11:36 AM
a thermostat was never used.
On the dyno the lower the temperature, the more HP!
I do not know the lower limit, if something 'strange' happens if you go too low.
But at a certain piont you might get fuel evaporation problems.
At Aprilia HP still improved going from 45° to 40°.
And at Bultaco I used tap water for cooling, about 20°, with good results and without any problem.
Lots of water moving rapidly was clearly the best.
Some people believe cooling the exhaust duct 'too much' would cost power.
This proved to be nonsense.
The more we cooled the more HP!
Reducing wall thickness also proved to improve HP.
The limit is mechanical reliability.

I just meant to say that reducing wall thickness improved power.
And especially in the hottest places of course.
Like the top of the cylinder, and around the exhaust duct.
The limit, of course, is mechanical strength.
And, in your case, maybe cooling capacity!

You want the aux ports as big as possible to have more blowdown.
You want the most blowdown with a limited exhaust timing in order not to lose power at low revs.
But there are limits.
You have to keep away from the transfer ports to prevent fresh charge going into the aux. ports.
That is why their bottom is inclined.
The same happens if you make the aux. ports too wide.
You loose fresh charge.
This also cools the exhaust gases, so you loose revs.
Symmetry is also important.
We tried a piston with a closed pistonpin hole on one side.
With this piston power was very bad!
Worse than with a 'normal' piston!
Closed piston pins did not improve power at all revs.
Only at some points in the power curve.
Mostly at high revs.
At some points in the power curve you loose something, but not much.

With aux. ports you can use a lower exhaust port, so that you have more power at low revs.


What really counts is blowdown.
But discharging of the burned gases may not be complete until BDC.
In this case you get 'short circuiting' of the fresh charge from the transfer ports into the exhaust.
By using auxiliar exhaust ports you can improve blowdown quite a lot.
So that you might be able to raise the exhaust's 'floor'.
This helps improve blowdown flow, as seen on a flowbench.
In 2007 we started to raise the underside of the exhaust port.
The first results were very promising!
But at the end of 2007 I retired, so I could not finish what I started.
The idea was to raise the exhaust underside as much as possible until power dropped.
And then, with a smaller exhaust port underside it might have been possible to widen the A-ports more
without losing the fresh charge into the exhaust.
The ideal situation would, of course, be that all the burned gases are discharged from the cylinder before
the transfer ports start opening.
So it is difficult to have too much blowdown.
But in trying to achieve enough blowdown you can arrive at a too high exhaust port.
Which first causes power loss at low revs, and if exaggerated still more also a loss of max. power.
Because of shortening the power stroke.


120mm is NOT particularly long, it is quite a normal length.
After much testing 50cc's arrived at a best length of 85mm
85:39,5=2,15
An MBA 125/2 had a 90mm conrod.
90:41=2,19
So a 125 conrod with 120:54,5=2,20 is about the same!
Even calculated in the '50cc way' it would be 54,5X2,15=117!
So the conclusion is that 125's always used way too short conrods.
Why?
Probably to reduce total engine dimensions.
At Aprilia we started with 110, and finally arrived at 120.
Each time the rod was lengthened the power improved.
Also if we kept crankcase volume the same!
Making the crankcase volume bigger increased power even more!
A lot of time and money could have been saved if they had asked, and listened to,
to someone with 50cc experience!
Instead rod lengths of 110, 113, 115, 118 and finally 120 were tried!


The squish band should follow the piston radius.
That was our conclusion after many tests!
rsss396
08-06-2013, 12:55 PM
Flow Testing of flushing channels. (transfer ducts)
I doubt very much to its usefulness.
We did it for each cylinder, mainly to see if there were strange anomalies.
They were never there.
But there were some changes eg cylinders flow more gifts.
But less power
It turned out in the end that it was best to test the changes. Simply power (dyno test)
Was there ever such Aprilia czech flow bench, Jante principle.
Also had little effect.
In all of these tests, flow does not take into account the differences in pressure during the rinse. (scavaging of the cylinder)
The best test for different coil channels was therefore the power test! (coil channels are transfer ducts)
The most useful had the flowen exhaust port and channel
When flowen of the outlet channel, the test was done with the piston in different positions.

1-2-3 mm open,Blowdown, Completely open.
We also had buses that you could test the auxiliary ports separately and also the middle gate.
We searched for the best possible flow with the smallest channel content.
The blowdown was a lot better with padded channel bottom.
A cylinder with center arrow had less flow than with auxiliary ports.
The dams between middle and auxiliary ports gave away less flow.
Make the exhaust port higher bottom gave slightly less total flow, but did not seem bad for power.
I was here not quite done with it when I retired
You could say that the outlet channel a flow bench is almost indispensable.
For the coil channels not. (transfers tunnels)
Bed and think much testing is therefore important.
And of course, the ability to be able to make. Many cylinders
I had that opportunity at Aprilia ample.
If you make little cylinders per year, it is almost impossible to try everything! Sufficient
What remains a question is how the influx from the crankcase goes into the coil channels.
It did not work for me as a test system to think.

Was important for you the total flow or the flow rate in terms flowen?

The total flow was actually the most important.
Searching for the 'narrowest point' in the engine.
These are the transfer ports, minimum flow and shortest timing!
So maximum transfer ports are very important.
But the timing is not so long, and you can not get too close to the exhaust port.
Because then bends to the flow, and the exhaust in!
So it is very important that most, preferably all, the cylinder gas burnt off before the flushing begins.
As far as the lower edge of the exhaust port than is currently interest for?
I so can not try! Until the end
The idea was then to increment to the assets less the bottom was.
And then if necessary
. to be able to make the coil. ports wider