ESE's works engine tuner

[Flettner]

Manolis, hi, good to see you are here and interested in the sleeve design. I had kind of given up since the Ryger thing came out but because the Ryger design information is is still far off being released ( "shortly" , six months ago) I'm back on the sleeve case, as it were.
The RR sleeve twostroke was tried with rings in the head as you have shown but it suffered thermal issues so they went back to Harry Riccardo's original (tested) design with the open ended sleeve. This sleeve is not a tight fit, in fact Harry said looser the better so long as it was not too loose to start (not enough compression). When things heat up the sleeve expands untill it starts transfering heat to the cylinder wall at which point it contracts untill it finds a happy heat transfer clearance, apparently.

Uniflow

[wobbly]

Glen, I think my point was more based around the fact that it revs out so well in 6th to near on 12,000, and thus you should be
doing this thru every gearchange as this will be alot faster , as you will drop back closer to peak torque - no matter what the final drive gearing is.

The Ignitech has no internal voltage "limiter ' except a failsafe detection circuit that shut it off at 18V +.
Thus as the input voltage rises so does the ionising voltage at the plug, and the "power " stored in the caps.
We spend ( or should ) big money on good plugs that make it much easyer for the CDI to fire a strong spark across rare earth electrodes, and using
a charging system ( be it RS250 or CR125 basically the same ) will hold the input at the much higher 14.2V, and then a small battery is fine.
The only way to increase the current draw ( and thus power in the gap ) is to use a very low primary resistance coil - such as RGV250/Aprilia RS250, or one with higher inductance.

[manolis]

Hello J.A.W.

You write:
“It was Bristol who really put sleeves on the map... Bristol ran one of their 6.5in sleeves up to 8,000rev/min
in a single-cylinder test engine without anything going awry,”

I suppose the Bristol single-cylinder was a 4-stroke.

In a 2-stroke sleeve-valve single-cylinder things are tougher because the inertia vibrations increase a lot.

The sleeve valve has a smaller, than the piston, stroke (say 40%) but it cannot help being heavy.
The sleeve valve reciprocates, more or less, in phase with the piston (in the animation, and in the “exe” program, the auxiliary crankpins are advances by only 10 degrees relative to the main crankpin; the resulting asymmetry between the transfer and exhaust is more than necessary).

Provided the balance webs of the crankshaft are optimized in each case,
when a conventional 2-stroke single-cylinder is modified to “sleeve-valve”, the vibrations will increase a lot.
It would be like replacing the piston of the original conventional by a two times heavier one.

A 2-stoke sleeve-valve needs a substantially longer stroke than a conventional 2-stroke (uniflow scavenging and short stroke do not fit), making things even tougher.

Thanks
Manolis Pattakos

[marsheng]

After a bit of help from Hasaberg I found a YZ250 rod was only 2 mm shorter than the Macio. The big end width is 20 on the YZ and 17 on the Maico. Is bigger better ? The YZ is supposed to be 48 HP.

I received my New ProX rod but there is quite a big difference in the wall thickness between the 2.



Is this an issue? I would have thought from a dynamic point of view the lighter the bigend the less the load on the pins however it makes it not quite as strong. However, the main strength will be required in compression which is the same for both rods.

I have put down the bigend failure to the oil slots not being radiused and hence they picked up on the rollers.

Lastly, what piston/crankshaft balance ratio is suggested and what to you normally weigh to make the bob weight?

Cheers Wallace.

[manolis]

Hello Uniflow.

You write:
“Manolis, hi, good to see you are here”

I expected more objections here (in this forum).


You write:
“This sleeve is not a tight fit, in fact Harry said looser the better so long as it was not too loose to start (not enough compression). When things heat up the sleeve expands untill it starts transfering heat to the cylinder wall at which point it contracts untill it finds a happy heat transfer clearance, apparently.”

The thin sleeve valve (2mm?), especially at its open top end, cannot help being flexible. If you put a high pressure inside it, it expands and its diameter increases.

Suppose you operate the engine at light load for several minutes.
The sleeve valve expands (due to the leakage of hot gas around its top end) until to touch gently the cylinder and to seal properly the combustion chamber.
The pressure during combustion is, say, 20 bar.

Suppose now the throttle turns suddenly wide open.
In the next cycle the combustion pressure increases to 50 bars.
The top end (more correctly the top external edge, which is hot and has the biggest diameter) of the sleeve valve abuts heavily on the cylinder (because it is flexible there and because it is loaded by the high pressure acting on a big inner surface of the sleeve valve).
The sealing is too good, but for several cycles the sleeve valve top end will be in contact with the cylinder (scuffing).


It is also the scavenging of the top cylinder. The short distance from the transfer ports to the top of your sleeve valve give the chance to a lot of mixture to escape to the exhaust (short circuit); the engine will operate with lots of residual gas (trapped in the top and around the core of the cylinder), even at full throttle.


It is also the vibrations / inertia loads issue (reply to J.A.W.) and the ability of the frame to hold the engine.



If all this trouble (new cylinder, sleeve valve, auxiliary connecting rods, new crankshaft etc) is in order to put exhaust ports at the top (or more correctly at the middle) of the cylinder of a 2-stroke, why not to use an exhaust PatRoVa rotary valve?



(more at http://www.pattakon.com/pattakonPatRoVa.htm )


Thanks
Manolis Pattakos

[J.A.W.]

Hello J.A.W.

You write:
“It was Bristol who really put sleeves on the map... Bristol ran one of their 6.5in sleeves up to 8,000rev/min
in a single-cylinder test engine without anything going awry,”

I suppose the Bristol single-cylinder was a 4-stroke.

In a 2-stroke sleeve-valve single-cylinder things are tougher because the inertia vibrations increase a lot.

The sleeve valve has a smaller, than the piston, stroke (say 40%) but it cannot help being heavy.
The sleeve valve reciprocates, more or less, in phase with the piston (in the animation, and in the “exe” program, the auxiliary crankpins are advances by only 10 degrees relative to the main crankpin; the resulting asymmetry between the transfer and exhaust is more than necessary).

Provided the balance webs of the crankshaft are optimized in each case,
when a conventional 2-stroke single-cylinder is modified to “sleeve-valve”, the vibrations will increase a lot.
It would be like replacing the piston of the original conventional by a two times heavier one.

A 2-stoke sleeve-valve needs a substantially longer stroke than a conventional 2-stroke (uniflow scavenging and short stroke do not fit), making things even tougher.

Thanks
Manolis Pattakos

Actually Manolis, most current high performance 2-strokes are close to 'square' bore & stroke-wise,
unless they are either overbored, or stroked - as increased capacity variants of existing engines..

The Bristol aero-engines were undersquare, like the R-R Crecy, but the Napier Sabre was oversquare,
& it was intended by design - to be capable of running as a 2-stroke..

I expect that current metallugical design techniques could produce a functional sleeve valve unit..
there must be a reason why the FIA banned them for F1..

[husaberg]

After a bit of help from Hasaberg I found a YZ250 rod was only 2 mm shorter than the Macio. The big end width is 20 on the YZ and 17 on the Maico. Is bigger better ? The YZ is supposed to be 48 HP.

I received my New ProX rod but there is quite a big difference in the wall thickness between the 2.



Is this an issue? I would have thought from a dynamic point of view the lighter the bigend the less the load on the pins however it makes it not quite as strong. However, the main strength will be required in compression which is the same for both rods.

I have put down the bigend failure to the oil slots not being radiused and hence they picked up on the rollers.

Lastly, what piston/crankshaft balance ratio is suggested and what to you normally weigh to make the bob weight?

Cheers Wallace.

Post a pic of your crank webs.
if the crank webs are pork-chopped balanced and or shrouded or semi shrouded you can gain some better cooling for the big end with a narrower rod and more HP as there is less resistance to flow.
Conversely the crankshaft can be stronger with a narrower rod. as the big end can have a longer areas supported. It's a compromise. You could machine the crank to suit the piston and then use off the shelf conrods forever or mod the rod each time.
Personally I would plumb for the one where you have the better available Flat bearing cadge that is suited for higher revs.
I will add some pics latter.
This is the best HP set up. Courtesy of Robs Gp125 set up.
Robs is balanced with Aluminium plugs holes the Aprilia is balanced with tungsten.
I don't know for sure if a solid ended big end is better but I would suggest solid or plugged like a Four stroke Honda. judging by the Aprilia pics

Good point, Dave; in fact several good points.
Most of the volume is concentrated in the transfer ducts. Then there is the volume inside the piston of course, and the 1 mm shear-avoiding clearance at all surfaces of the crank.
But that is not nearly enough volume. If you take another look at the Aprilia crank below left, you will notice that the space between the crank webs is the same as the space needed for the big end bearing. In other words: the crank webs have flat insides, good for another 60 cc or so, if I remember correctly.
Additional benefits: the con rod has an easier time pushing the mixture aside as it moves between the webs, and the big end bearing gets a lot more cooling and lubrication because it is not shrouded in any way.
Because there are no overhung bobweights, the crank webs are stuffed with tungsten to get the balance factor right.
In the RSA125, the con rod was lengthened from the RSW's 115 mm to 120 mm to create even more crankcase volume.

The paddling is a mixed blessing; it creates aerodynamical drag but it also greatly improves the homogenity of the mixture.
Smooth, full-circle crank webs have the advantage that there is little mixture hiding in nooks and crannies. An example of it's importance: in a certain engine there were 20 mm spaces between the crankshaft bearings and the seals. these ill-accessible volumes acted as pneumatic dampers on the crankcase pressure fluctuation. Filling those volumes with plastic bushes gave a measurable improvement.

Summary: you need a large crankcase volume. Ideally all of this volume should be situated in the transfer ducts. In real life you will also need to lodge part of this volume between crankshaft and piston, i.e: use a long con rod. Avoid nooks and crannies. Crankshafts should be small and smooth. Big end bearings must never be shrouded by recesses in the crank webs or by stuffers.

The picture right below shows, wait for it, an RSA125-crank with stuffers...
After Jan Thiel went into retirement in 2008, some geniuses at the factory grabbed their chance to 'correct' the errors that Jan left behind, without even testing the result because 'everybody knows the smaller the crankcase volume the better'. But they never could understand why a 2011 RSA125 was slower than a 2007 model (just look at the 125 cc top speeds on any GP-track). O, the joy of working with Italians.....

They can't have been Aprilia racing cranks then, as these have no holes at all, just a lot of slugs. But some are tungsten, some are very light.
The color difference in the picture below shows the asymmetrical distribution of light and heavy slugs. This is a give-away that these are crankshafts from the 90°-twin engine; their balance mass center of gravity has a 45° offset whereas the crankshafts of the Aprilia singles have no balance mass offset.

You (not you, Wobbly) may also wonder why there are so many slugs; couldn't it have been done with a much smaller number? Yes, but the large number serves a second purpose, apart from balancing. Putting light slugs near the center and heavy slugs near the circumference of the crank webs increases the crankshaft inertia without increasing its weight.
Attachment 310031

Also if i can be so bold i was looking at a pic of a RSW or RSA crank the other day and it seemed to have what looked like Brass rings around the Crankpin pressed in the crankcheeks?
Is that what they are? and why are they there?

Attachment 315351

.Yes, that is what they are (well, not brass but bronze) and they are there to reduce axial friction between the con rod and the crank webs.

Amazing Frits how complex logic can get, and more amazing that in reality its all simple when you finally get to understand.
Of course Lambda only reads the O2, what a dumb shit.

Anyway more importantly I have now for the first time seen the inside face of an Aprilia crank.
Apart from the bronze ring ( does that replace a silver plated washer ? ) I now see a huge amount of inserts in the crank face.
These you dont see in the many pics that show a ton of what is obviously Mallory around the circumference increasing the rotational inertia.
I assume the inserts on the inside are filled with lighter material ? in the pic I cant tell if its alloy or even more Mallory heavy metal.

This was discussed on the Aprilia FB page Wob, not sure if it was Jan or Thijs that said it was tungsten with a magnesium cap. Thrusts are still there.

Its something that has been revolving in my mind as they are cracked... they seem to be actually pressed in the full distance of the pin rather than a thrust washer insert replacement? there def seems to be an full insert (steel looking inner one at that on the left side picture)

The bronze rings are only about 2 mm thick. Their advantage over the usual thrust washers is that they can get rid of friction heat easier and won't vibrate themselves to pieces.
The combination of Mallory slugs and hollow light-alloy caps serves to combine a low total mass of the cranks with a high inertia that made the 250 cc twins more rideable. The same setup was used for the 125 singles, but as Jan Thiel said: 'In seven years of experimenting we have not been able to establish what is best: high or low inertia'.
My approach: when in doubt, choose low; it will be a blessing for the transmission and the rear tire.

[Flettner]

Manolis, how does your rotary valve head seal?

[Frits Overmars]

Its a well proven fact that reving a race engine to just past it power peak, so that you drop back closest to peak torque thru every gear change is by far the fastest technique.

That is a good approach. However, theoretically the object is not to drop back closest to peak torque through every gear change, but to change gear at the rpm where the power before the gear change and the power after the gear change are equal. Or to express this graphically: where the current gear's rear wheel thrust curve intersects the next gear's thrust curve. Remember: power is not only torque x revs, it is also force x velocity.

The graph below shows the acceleration of an Aprilia RSA125 over a distance of 800 m from an initial speed of 50 kmh. The Y-axis scale is determined by the grip coefficient of the tire.
In first gear the clutch is slipped at 12517 rpm which is the rpm of maximum torque. By the way, slipping the clutch will be favourable each time the revs drop below max.torque-rpm after a gear change, provided the clutch can take the heat and the rider can spare the concentration. But with a racing gearbox this won't be necessary after first gear.

In both the graph and the numeric picture you may notice that in the higher gears the engine is revved a little bit beyond the intersection point, as a compensation for the loss of velocity due to the air resistance because of the power interruption during the gear change. You'll notice this loss between the final velocity in a gear and the initial velocity in the next gear.

[marsheng]

The thickness of the boss is 4 mm so taking off 1.5 mm on each side will be easy.

[manolis]

Hello Uniflow

You write:
“Manolis, how does your rotary valve head seal?”

No sealing means are used in the first PatRoVa prototype .

The clearance between the inner flat fronts of the rotary valve and its mate flat lips (on the external side of the combustion chamber) is quite small.

The material used in the prototype of the video is graphite iron (for both parts: the rotary valve and the cylinder head).
With DLC coating it would be better (long term reliability).

The “Sealing” paragraph at http://www.pattakon.com/pattakonPatRoVa.htm explains it better.

In a 2-stroke things seem even easier because the mixture has oil droplets (an oiled screw compressor is capable for way higher compression than a dry screw compressor).

Thanks
Manolis Pattakos

[ken seeber]

Well, things are happening here:
• cylinder is almost finished
• got some more profiling to do on the spacer plate, unfortunately vertical machining centre is tied up at present
• piston and bits are nearly ready
• Schenck eddy current dyno in place
• fuel flow meter, HC and NOx analysers coming next week

Just waiting on the final details from Harry.

Come on Harry, we’ve been waiting for ages Promise we won't tell anyone

[peewee]

In both the graph and the numeric picture you may notice that in the higher gears the engine is revved a little bit beyond the intersection point, as a compensation for the loss of velocity due to the air resistance because of the power interruption during the gear change.

frits i use this same technique in my pickup especially with a strong head wind and to make matters worse it has 24/7 fwd so the front axle is always turning, causing a big momentum lose each gear change

[peewee]

:

after seeing that pic it makes perfect sense why engmod keeps telling me every one of my cylinders has far to small of inlets. that inlet looks bigger than the bore

[husaberg]

after seeing that pic it makes perfect sense why engmod keeps telling me every one of my cylinders has far to small of inlets. that inlet looks bigger than the bore

What does it say about your carb size being far smaller the 70% of your bore.

The thickness of the boss is 4 mm so taking off 1.5 mm on each side will be easy.

I added in some other stuff to my reply above Wallace. Note the inserts and clean inner wheels. Order a billet of EN30 or 4330 and slap it in the CNC.
You know you want to.
I can't remember what Neil Uses?

Been back to ESE head quarters this evening to balance my crankshaft. TeeZee set me up with some scales, some welding wire and a pile of washers. First up the rod, piston, pin and small end were weighed giving us a total of, I think 232gms, TZ works on the 50% theory so the wire and washers were weighed accordingly and hung off the end of the rod, this was then placed on the very technical looking balancing device where it was found to be PBC, the fly wheel was drilled in four places to get the correct balance.
Before all this we put it in the lathe and using a dial gauge we checked that it was running true, not too bad just a light tap with the truing device and we were happy.
Thanks to Tee Zee and Culley for their time.

Let me start by complimenting you on a very clear description plus illustration of the balancing act. But you should not have drilled those holes where you did. As a rule of thumb, a good press fit requires that a big end pin hole should be surrounded by at least half its diameter in material. So for a 20 mm pin there should be at least 10 mm of crank web material everywhere around the hole. As you can see, the large original balancing holes already intrude into this zone, and with the small holes you added there will be little stiffness left.
Removing material in the blue circles would have been a safer approach, although it might not haved raised the balance factor sufficiently.
The best way would be to remove material from the inside faces of the crank webs, around the big end recesses. That would have two additional benefits: it would improve lubrication and cooling of the big end bearing and it would enlarge the crankcase volume.
Now that I am grumbling anyway: big end pins should be massive. The large void in the pictured pin is not exactly promoting a good press fit.

My pleasure Martin. Take a look at the above picture. Jot down the scales' reading, remove the piston from the con rod, let the con rod rest on the scales again and add a wire hook and a collection of nuts or washers to the scales until the reading is what you want it to be (50% of the original weight in the above example).
Attach the wire hook with the nuts to the con rod's small end and let the crankshaft rest on two beams, like in the picture below. First make sure that the beams are exactly horizontal by putting a piece of tubing (or a bottle) on them and adjusting their heights until the tube does not want to roll any more.
Attachment 309191

Now if we put the crankshaft on the beams, it will probably roll until the bob-weights are either at the highest or at the lowest point, in which case we will have to add or otherwise remove material from the bob-weights.
Let's assume that the bob-weights move to the lowest point. Then we add extra weight (some more nuts or washers) to the wire hook until the crankshaft will stay in any position on the beams without a tendency to turn. We have now created the torque that is necessary to keep the crankshaft in equilibrium.

Example: We have an engine with a stroke of 54,5 mm. And we had to attach 80 grams of extra weight to the hook in order to reach equilibrium.
This weight acts via the con rod on the crank pin radius (half the engine stroke), so it generates a torque of 80 * (54,5 / 2) = 80 * 27,25 = 2180 gram*mm.

The bob-weights had a tendency to move to the lowest point, indicating that they are too heavy, so we have to drill holes in them.
The total mass of these holes, multiplied by the distance from the holes' centers to the crankshaft center line, also generates a torque.
Let's assume that we can drill a hole in each bob-weight sidewall, opposite the crank pin, with its center 35 mm away from the crankshaft center line.
In order to generate the same 2180 g.mm torque, the total mass from these holes must be 2180 / 35 = 62,28 gram.

One hole in each bob-weight, that's two holes in total. One hole should then equal 62,28 / 2 = 31,14 gram.
The specific mass of steel is 0,00785 gram / mm³, so the volume of each hole must be 31,14 / 0,00785 = 3967 mm³.
Let's assume that each bob-weight has a thickness of 20 mm. That is how deep we can drill (through and through).
Then the cross-section area of each hole must be 3967 / 20 = 198,35 mm².
Its diameter must then be SQRT (198,35 / (pi/4) ) = SQRT ( 198,35 / 0,7854 ) = SQRT (252,55) = 15,9 mm.

Obviously the further away from the crankshaft center line you can drill these holes, the smaller they need to be.

Piece of cake, isn't it?

Now how about that teaser? Any takers?

Very nicely illustrated. I would have used those pics if I'd had them .