ESE's works engine tuner

[breezy]

Wobbly, would you think that if transfer ports could be controlled at the cylinder/face/ wall( reed type valve) that having the transfer ports high in the cylinder would enable the filling of the cylinder to be more efficient, using the the opening of the exhaust port to both expell combusted fuel and pull a fresh charge in behind it, with the transfers being open to atmospheric pressure while open? could this lessen the amount of fresh charge being pulled straight out the exhaust port as in the way transfers are positioned in the cylinders we have now? with this in mind if the piston was used only to compress the fresh charge and was not using the return to bcd to influence the goings on below wouldnt this enable the piston to move more freely?

[TZ350]

Page 1230 ......

ESE's number 1 works rider has scored herself a ride at Suzuka on board this Beast, moonlighting must be ok

Ken Oconnor racing has a good video on rebuilding tin can cranks,he cuts the rod and pin in half for easy removal..

https://www.youtube.com/watch?v=hs5Au5rrlUw

Compression Ratio

TeeZee I dont understand why you would spec a combustion chamber with a huge squish - 0.8mm when 0.65 is fine.
And no squish width, that gives no squish velocity at all at 18.7M/Sec - this is just throwing away alot of power everywhere in the useable band.???.

As the Ex height has a big influence upon the trapped dynamic com, I use the trapped ratio. For your air cooled 125 on AvGas a ratio of 7.4 will work reliably no matter where the port is.
Here is the output with the Ex at 82*, double the squish velocity, and near on the same pumping pressure prediction.

Wobbly, can I assume the trapped ratio you refer to is the CR calculated from exhaust closing rather than the simple geometric CR based solely on swept volume?
If so, do you have a guideline for a water cooled 100 on 98 pump gas?

The trapped ratio takes account of radical differences in Ex port height, but when dealing with your normal race engines the Ex port is usually in a fairly small window of timing.

Thus your 100 on Avgas could quite happily go around 16.5 full stroke CR (geometric CR based solely on swept volume), 98 pump would be pretty much maxed out at 15.5.

Exhaust port Duct length and Shape.

The optimum transition length. There will be a direct relationship to how long the well cooled duct wall is, to the length of the A/F slug sitting outside the port waiting to be shoved in by the returning wave front.

All I can say is that in any normally dimensioned cylinder we usually end up with a duct and spigot around 2X bore length, and that a slip joint
spigot is usually around 1/2 bore - and that works.

Technically correct as Frits has pointed out, would be to derive a relationship between duct exit area and the blowdown needed to develop the target bmep. But in any scenario keeping the duct volume down by lifting the floor and filling in the bottom corners, then restricting the exit and connecting that to a transition WITH NO STEPS will make better power than any other solution that has actually been tested to date.

This has been playing on my mind I couldn't reconcile this with what I had read about not removing the step in the exhaust duct. Why would the A kit not have the step if it's supposed to reduce power. Clearly I don't properly understand the function of the step.



So I went back and tried to connect the dots from the bits I remembered from the ESE thread: so no steps at the flange is better, so why does the NX4 RS125 lose power when the steps are cut out? Is this a function of the duct entry to exit ratio changing?

Step or no Step in the exhaust duct.

For anyone still wondering about steps ( or not ) in the Ex or if a race plug is worth buying then this dyno sheet should give you the idea.



Green is a straight round spigot with steps, brown is a CNC oval/round transition,red is the same transition but replacing the R7376-10
plug with what is commonly used alot in USA, a BR9EIX.

The transition needed 2 jets leaner to hit the same 650*C egt in the tests. It might cost 4X the price but the race plug is the cheapest 1Hp gain in a box stock engine you can buy.

Fancy Plugs.

The plug result just confirms that having a rare earth fine wire center AND a fine wire platinum ground electrode makes the ignition power required to initiate ionization under heavy compressive loads is way less with the race plug , when compared to the simple iridium plugs cast iron plank hanging off the side of the body. The dyno doesn't lie - check the overlay at the bottom of the curves, near as damn it identical in the range where the plugs efficiency makes no difference.

This 125 engine only just cracks 40Hp so isnt really stressing the conventional ignition that much, but as soon as things get tough the good plugs configuration makes a difference.

A low bmep engine with rezo spark would as you say make bugger all difference, as it simply doesnt need the extra grunt. When you really stress the hell out of the system, as in the Aprilia where simply changing to a bigger coil made more power, or when running big com on rich methanol engines the power difference can be really worthwhile. Ages ago my best TM125MX kart engine went up 2Hp to around 50 when changing from a BR10EGV.

So here it is as simple as can it be made - NO STEP IS BEST - but this does NOT mean grind the duct to make it round OR bigger in any way.

If its a 3 port or a T port, then the duct exit should be around 75% of the total effective Ex area, then the spigot should transition to 100% round diameter.

If its a single port, the duct should be around 90% exit area, with no step and the header the same diameter if its round. If you have a really dumb 3 port or T port with a round exit, THEN CHANGE IT.


Re the angle area of the RSA. The answer is all to do with finely balancing all the variables to optimize the end result. Do you REALLY think that if going to 204* was actually better then why the hell wouldnt Jan do exactly that. We have had the same question on here about the pipe dimensions,why not simply make it shorter and fatter, well if that was better do you really think that it would not have been done.

I have been running thru pipe designs for a 250 KTM for a road racing kart, and discovered in the sim that going over 132mm belly diameter makes
NO MORE power.

The cylinder pressure at exhaust opening can be as high as 7 bar,

Maybe Jan was under the thumb somewhat with the pipes, but still they were up to Tubo 120 or something so plenty of experimental ideas were a failure.

But something he did say was that they didnt pay enough attention to the changes in fueling when trying out the new pipe ideas. In that many pipes with changes to dimensions made little difference, but may have been better/worse if the carb tuning was strictly optimized to suit.

I have taken this onboard and find now that I need to change jets for nearly every single small change in any part of the engine, just to get back
to my baseline peak power egt on the dyno.

Re the KTM road racing pipe. My best pipe so far has a diffuser end ( belly front ) of 135 diameter and a rear cone diameter of 132.
Going up to 138 with either steeper diffuser angles, or a shorter or longer rear cone to suit, was always within no more than 1.5Hp everywhere across the powerband. Sometimes a bit more down low and less on top or visa versa.

The 135/132 has the best average everywhere, and as these things have a wide spaced MX gear ratio split with something like a 1500 rpm drop when running to 10800 and selecting 5th, the front side power slope is just as important as the overev.

As I have stated before I believe that we have finite energy available from the advancing wave front, and in this case the best result I could get from the engine involved a slightly steeper diffuser, but as soon as I shortened/lengthened ( reangled ) the rear cone I lost power - thus I ended up with a slightly tapered mid section just as the Aprilia had.

Hello guys Frits has already explained about port directions and axial angles.
Aprilia A-ports had about 28* axial angle and B ports had 7*.What about the opposite?

I was watching old Yamaha cylinders (B ports directed upwards and A ports are perpendicular) I was wondering what effect will these axial angles have with A ports having more duration (no exhaust power valve). Cheers!

All latest generations of engines with normal stagger ( A port first ) have the same axial relationships as the Aprilia.

Flat A ports are long dead.

Welding the roof corner is needed in the Yamaha B port to flatten it out as it goes into water, the A port simply needs epoxy
to fill in the top radius and increase the upward tilt.

The C port is always in the 50 to 60* range, depending upon the end use, the bore/stroke, and the port width in relation to the B width.

Crank pins and Mallory metal plugs.

The alloy plugs cant be weld retained ( but several pressed in dimples on the circumference helps ) so i press them in with around 0.03 to 0.05mm interference. The Mallory inserts can be retained by simply melting the crank material into the plug in a couple of opposite arcs on each side, so the press fit can be reduced to 0.01 - 0.02 . You have to be careful with this as I have had a crank web split out from the Mallory hole, to the wheel circumference ,as the press fit was too tight due to the plug being slightly oversize.

Welding of the center axle on twins is often needed - " advised " when there are two press fits on top of each other as we see in rebuildable RD/RZ/Banshee cranks. The inner press fit simply expands outward into the hole for the big end pin. This prevents the two inner wheels from turning in relation to each other and wreaking the phase, but usually this only occurs with a seize on one side.

Cranks like RGV and Aprilia etc have integral pins on the inner wheels so should not need welding of the axle if the press is sufficient.
If the big end press fits are not correct then welding is sometimes the only way to keep them true - but is really butchery of the first order.

Interesting thing to do! I once fitted a tungsten insert that started moving in the crank, by deforming the edge of it with a center punch.

Sorry for my curiousity but I really do not like to weld on cranks: Has anyone a hint how high I must go with the press fit without using any special tricks?

For some Numbers: I just got my calculater working:
The Force of my tungsten insert will be 16,2KN @ 15000RPM (WOW!)
Incredibly high - perhaps I calculated it wrong...
Radius of the Center: 32mm
Diameter of the Tungsten: 22mm

Space to the edge of the crank: 3mm (is this enough?)
Space to the neighbouring Tungsten: 2,5mm (same question here)

Cheers!Tim

Having any sort of working press fit in a flywheel that has ANY hole only 3mm from the circumference is simply asking for failure to occur. Simple rule of thumb for this is that TWICE that is needed.

All that occurs with insufficient material is that if the core of the wheel is still under say RC40 hardness the hole will elongate outward to the wheels circumference and the press fit goes away - if the core is any harder the material will split outward from the hole.

Several USA tuning firms have done stroker cranks for Jetskis with only 4mm and they explode continually, even with only intermittent use as in Freestyling. They botched the problem by welding them initially, but later they got a bit cleverer and made eccentric pins,and welded those - still a dumb solution.

It is FAR better to have a light fit in your Mallory and tig melt a couple of small arcs of weld for retaining it, than having to use any excessive press that automatically creates a stress concentration.

[chrisc]

This has been playing on my mind from a comment that you made to me via pm Wobbly:

Also the [Honda RS125] A kit Ex manifold was the same as the later 250 with an oval to round transition in the flange, no step at the cylinder face.

I couldn't reconcile this with what I had read about not removing the step in the exhaust duct. Why would the A kit not have the step if it's supposed to reduce power. Clearly I don't properly understand the function of the step.


So I went back and tried to connect the dots from the bits I remembered from the ESE thread:

Peewee, you are of course going to be tig welding the inside of the steel spigot to grind and match it to the duct exit shape,
by making an oval to round transition - yes.

is the roof and floor of the manifold supposed to flat like the cylinder exit ? i was just going by how the rsa looked.

Yes, you need to shape the manifold floor/roof as per the exit, and this shape as well as the side port ducts need to taper
down smoothly all the way to nothing at the face where the header starts. I have tested this numerous times and a proper oval to round transition always makes more power than having the steps.

ok so the manifold floor and roof should match the exh exit, no steps. then transition to round at the exit of the manifold. got it . i hate to even ask but why was the rsa125 not that same way ?

The RSA indeed had steps at top and bottom of the cylinder/manifold junction; something to do with evolution I'd say, like why men have nipples. But I am with Wobbly on this one. Moreover, the RSA's exhaust port shape as it has become known through the numerous pictures, is not the latest version; the final version had the port floor raised even higher.

So no steps at the flange is better, so why does the NX4 RS125 lose power when the steps are cut out? Is this a function of the duct entry to exit ratio changing?

[Frits Overmars]

I mentioned that rod length because you gave up angle area. If a 2 stroke is angle area limited, why give any of it up? around 10% for rings, around 15% for crank bearings. Will you gain 50% more power with ultimate ports?

"Why give any of it up?" is the mentality that wins races, so I had to look into your point of view, Jonny.
As I wrote earlier, shortening a 120 mm con rod to 100 mm gives 2% more blowdown angle.area and 2,9% more transfer angle.area.
The shorter rod gives about 26% more friction. That seems like a lot until you realize that at 90° before & after TDC a 54,6 mm rod would stand almost perpendicular to the bore in an engine with a 54,5 mm stroke, and give almost infinitely more friction.
Ring friction and bearing friction won't be heavily influenced by the con rod angle, and I can go along with a piston skirt friction in the area of 20%.
If this 20% is augmented by 26%, the piston skirt friction will become 25,2%, so the total friction loss in the short-rod engine will be 5,2% more.
That engine can rev 2% higher because of the increased angle.areas, which is not enough to compensate for the increased loss. Dammit Jonny, you are right.
Moreover, the short rod decreases the crankcase volume (OK, that can be compensated somehow) and the increased piston friction will mean an increased engine temperature, which will cause a secondary power loss. How much? Hard to say .

[Frits Overmars]

I couldn't reconcile this with what I had read about not removing the step in the exhaust duct. Why would the A kit not have the step if it's supposed to reduce power. Clearly I don't properly understand the function of the step.

That A-kit paper is from 1992. You can tell by the picture and the drawing that show the coolant hose nipple still attached to the cylinder.

So no steps at the flange is better, so why does the NX4 RS125 lose power when the steps are cut out? Is this a function of the duct entry to exit ratio changing?

Cutting the steps out increases the duct volume, which is bad. You may remember my bicycle-pump metaphor.
Filling in the steps until there is a smooth transition to the header is good.

[Frits Overmars]

husaberg, have you come across any 2 stroke engine designs that in corporate reed controlled transfer ports, with the transfer ports high in the cylinder above the exhaust port/s. transfer ports connected directly to the inlet and not through a crankcase?

The cylinder pressure at exhaust opening can be as high as 7 bar, so if you put your transfer reeds even higher up, they must be able to cope with more than 7 bar.
This means strong, heavy reeds that will be very reluctant to open at the very limited scavenging pressure differential that the exhaust suction can generate.
And how would you start the beast? No running = no pipe suction = no starting...

Wobbly, would you think that if transfer ports could be controlled at the cylinder/face/ wall( reed type valve) that having the transfer ports high in the cylinder would enable the filling of the cylinder to be more efficient, using the the opening of the exhaust port to both expell combusted fuel and pull a fresh charge in behind it, with the transfers being open to atmospheric pressure while open? could this lessen the amount of fresh charge being pulled straight out the exhaust port as in the way transfers are positioned in the cylinders we have now? with this in mind if the piston was used only to compress the fresh charge and was not using the return to bcd to influence the goings on below wouldnt this enable the piston to move more freely?

We'll have to wait for Flettners sleeve valve uniflow engine to give us a partly answer. Flowing top-down or bottom-up won't make much difference in itself.
It will still breathe through the crankcase though. It would be easy enough for Flettner to open the transfers direct to the atmosphere, but then he would have to fit a fistful of carburettors (like I once did with a Rotax-124 cylinder) and make a provision for crankshaft and piston lubrication. And again: how would you start it?

[chrisc]

That A-kit paper is from 1992. You can tell by the drawing that shows the coolant hose nipple still attached to the cylinder.

Cutting the steps out increases the duct volume, which is bad. You may remember my bicycle-pump metaphor.
Filling in the steps until there is a smooth transition to the header is good.

I was wondering if someone would notice that it was a NF4 example, good eye Frits. It was the only one I had on hand but have read the same thing regarding the NX4 steps.
Great, so it is the change in the duct which is causing the loss in power, rather than the step itself. Noted.

[husaberg]

Frits are you able to access your records at the moment?
I have a few questions in regards to the Reed Valve Derbi that Jan tweaked.

[Frits Overmars]

Frits are you able to access your records at the moment? I have a few questions in regards to the Reed Valve Derbi that Jan tweaked.

I'm abroad but I carry some historical paperwork with me. What can I do you for?

[husaberg]

I'm abroad but I carry some historical paperwork with me. What can I do you for?

I was wondering what the specs were on the Derbi
What the crankshaft and the reed valve and tracks looked like. I assume the cylinder and pipe was the same or very similar to the original RSA prototype

[ken seeber]

Frits,
All this current angle area stuff got me thinking. A question on the exhaust vs transfer mixture in-cylinder interface. In the classic and generic current Schnurle scavenging layout, one could imagine that there is a vertical plane, in line with the crankshaft axis, in the centre of the cylinder. One on side of the plane is the downward exhaust flow, the other side seeing the upward scavenging mixture. Yep, I’ll concede this is pretty simplistic.
Disregarding the length of this interface, its width is the bore diameter. Now in a FOS, or other similar concepts, the rising gas flow might consist of a column that is, say, 50% of the bore diameter. This being the case, the perimeter length would equal pi*bore diameter/2, which is over 1.5 times the length of the Schnurle perimeter.
Do you think that this increased perimeter length might result in increased intermixing, hence more fresh mixture lost and more exhaust gas retention, neither good? Not a criticism or trying to give you the shits or anything though.

[jasonu]

I was wondering if someone would notice that it was a NF4 example, good eye Frits. It was the only one I had on hand but have read the same thing regarding the NX4 steps.
Great, so it is the change in the duct which is causing the loss in power, rather than the step itself. Noted.

My mate Owen Wilson ground out the step to match the exhaust spigot in his 1989 NX4 RS125 and made it go slower.

[peewee]

The short rod decreases the crankcase volume (OK, that can be compensated somehow).

then wouldnt you end up with something like the yamaha twins ? large voids in the crank cheeks. what good would it do to have your volume down there ?

[adegnes]

then wouldnt you end up with something like the yamaha twins ? large voids in the crank cheeks. what good would it do to have your volume down there ?

Or you could do like in the Peugeot speedfight 3 scooter - very wide and short transfer passages. It has a short rod, may be why they designed the cylinder and crankcase that way.

[Haufen]

Increased piston velocity would have been no issue at all in the RSA; the crankshaft was bulletproof.
There have been experiments with a smaller bore for the RSA, but only with 52,8 mm instead of 54 mm, because the special small-bore piston had to be made out of a standard forging. Making a special piston forging was considered too expensive (in a period where millions of euros were wasted on development of the three-cylinder foul-stroke MotoGP engine).
Changing something in a highly-developed engine will almost invariably lead to a power loss and it will take quite some development to establish whether the change was worthwhile, but Jan Thiel did not get the opportunity to carry out this small-bore development because it was aborted by the Great Leader when there was no immediate improvement.
I expect that at the very least the axial scavenging angles would have to be adapted to the altered bore/stroke-ratio.

So the scavenging angles were the same as with the square engine? And the crankcase volume, too?
I experienced the same when scaling engines. Once you've done it right, the theoretically expected results will show up. But it takes some time to find the correct parameters to modify, as not all of these are obvious.

Cutting the steps out increases the duct volume, which is bad. You may remember my bicycle-pump metaphor.
Filling in the steps until there is a smooth transition to the header is good.

If I remember correctly, the RSA was just like that, with a step in the header. Was the step better on that engine? What was the theory behind the reason for this?