2 stroke thread (with occasional F1 relevance!)

[J.A.W.]

This is a Kawasaki 750/3 H2R roadrace engine in a US dirtbike (flat tracker) chassis built by Erv Kanemoto.
The idea was to make 100+hp available to defeat the Harley-Davidson XR 750 twins of the day ( ~1974).

The available tyres & smooth 120` firing impulses/even-flow inertia torque made it difficult to gain traction.

One idea tried was to re-configure the Kawasaki's built-up rolling element bearing crank to fire as a single,
( "tringle", all 3 cylinders together) - it was not a success, both vibrating excessively & damaging the transmission
components designed to have the power impulses divided evenly, not all in one hit.

Kel Carruthers also built a TZ 750 equivalent for Kenny Roberts to help defend his AMA No 1 plate/National Championship, when his 4T vertical twin Yamaha was being out-paced/unreliable at hard-tune level.

On the TZ, they tried a roadrace 'wet' ( tread-patterned) rear tyre to help with traction, which was an improvement,
but also went to a kill-switch to cut one cylinder - when pressed, to regulate excess power.

These machines were making progress, so the AMA promptly banned engines of more than two cylinders.

[gruntguru]

A small comment on the term "inertia torque". Its a bit of a misnomer because the transfer of reciprocating Kinetic Energy to rotational KE and back, as the piston stops and starts, actually manifests itself as a crankshaft speed variation. This of course results in a cyclic torque variation as the drivetrain winds up and unwinds.

[manolis]

Hello Tommy Cookers.

You write:
“this is nonsense - just Yamaha advertising hype”

Their M1 (crossplane crankshaft Four-In-Line) dominates for several years in the moto-GP.

From Wikipedia:
7 World Championships won:
Valentino Rossi in 2004, 2005, 2008 and 2009
Jorge Lorenzo in 2010, 2012 and 2015



You also write:
“at very high rpm the reciprocational inertias are very large
they give a substantial positive-rotational, then negative-rotational, impulse twice per rev to the crankshaft
these impulses are almost completely opposing, so the net effect over the whole rev is only small, and actually a power loss from friction
all reciprocation does this whether the pistons reach tdc/bdc together as conventional or at intervals as do the Yamaha's”

No.

In the crossplane crankshaft In-Line-Four the energy reciprocating between the crankshaft (flywheel) and the set of the four pistons is practically zero. This is so because the increase of the kinetic energy of the two pistons equals to the decrease of the kinetic energy of the other two pistons.
The crankshaft does nothing else than interconnecting the four pistons.
The crankshaft needs not to accelerate / decelerate in order to provide / absorb kinetic energy from the set of the four pistons.

In the plane cranshaft In-Line-Four things are completely different. The set of the four pistons has a total kinetic energy varying from zero (at TDC and BDC) to extreme (when the piston pass from their middle stroke). The crankshaft / flywheel has a difficult work to do: it has to accelerate as the pistons approach TDC / BDC and then the crankshaft has to decelerate as the piston approach their middle strokes. I.e. a lot of energy reciprocates between the crankshaft and the set of the four pistons.

What does it mean in practice:

If you remove the cylinder heads and drive the crankshafts to rotate near, say, 10,000rpm, and measure the angular velocity of the crankshaft during a complete crankshaft rotation, in the crossplane crankshaft I-4 you will measure a practically constant angular velocity, while in the plane crankshaft I-4 you will measure a variation of the angular velocity.

The previous is equivalent to:
The free inertia torque in the crossplane I-4 is about zero, while the plane I-4 suffers from an extreme free inertia torque.



You also write:
“such impulses happening 400 times per second cannot propagate beyond the transmission shock absorber (usually it's in the clutch)
this works by attenuating and killing any torque variation faster than equivalent to above about 2 or 3 times per second
that's what shock absorbtion is”

For an engine running in a wide rev range, a shock-absorber has a difficult, if not impossible, duty:
Every time torque is “killed” (either it is inertia torque or it is combustion torque), energy (and so power) is lost.

The fact that the free inertia torque in the plane crankshaft I-4 is of second order and the total combustion torque has a big coefficient of 2nd order (one combustion per 180 degrees) makes things even more difficult for the transmission shock absorbers: if they are set to kill the inertia torque, they cannot help killing a good part of the combustion torque, too.

In comparison, in the crossplane I-4 the transmission transfers only combustion torque. Either at low revs, or at high revs.


You cab think of it differently:
The plane crankshaft I-4 has a crankshaft rotating at variable speed (during a crankshaft rotation, I mean), while the transmission and the rear tire try to rotate at constant angular speed.
Is there a way to much those two different angular speeds without killing a part of the energy?

Here is how:



The primary transmission gearwheels are properly changed . modified to allow the different angular velocities without overstressing and without killing energy / power.
The idea is described in a patent filed half a century, or so, ago.

If it is not clear how it works, let me know to further explain.

With such a primary transmission, the plane crankshaft I-4 could play and win the crossplane I-4.

It is a pity no engine maker yet “discovered” this simple idea.



(Hello J.A.W.
I wrote “Later they discovered the inertia torque” in my last reply to you.
The solution was there for several years (I know it from a Search Report).
A patent for a I-4 crossplane was filed by Kawasaki a few decades ago.
Only recently Yamaha “discovered” the advantages of the crossplane crankshaft I-4).



So, the I-4 with crossplane crankshaft is substantially different than the I-4 with plane crankshaft as regards the loading of the transmission, the loss of power in the transmission, the feeling of the rider, the side loading capacity of the contact between the rear tire and the road etc.

Thanks
Manolis Pattakos

[J.A.W.]

& of course Manolis, 2T 120` triples have always been 'crossplane', with even firing/inertia torque = to 4T inline 6 cyl.

Here is another look at the 180` 'twin-fire' Yamaha 2T inline 4 TZ engine..

http://www.classicyams.com/production-r ... tails.html

[manolis]

Hello Tommy Cookers.

You write:
“ok this might allow the necessary crankshaft torsional etc frequencies to be met with quite a slender crankshaft - if the load had relatively little inertia
but the load has great inertia (particularly so relative to the slender crankshaft)
and is worse than purely inertial because the (propellor) blades will oscillate in use at a natural frequency
and there will be cyclic bending loads of aerodynamic origin with eg 2 and 4 ? bladed props

aero engines generally have increased crankshaft and bearing dimensions because of these factors
and typically have tuned dampers eg 41/2th and 6th order
yours might suffer more than most in this regard
and reduction drive or only direct drive capability is likely to be a big factor for consideration in this regard”




Look at the diameters of the basic journals and of the crankpins of the flat-4 single piece crankshaft; look also at the degree the basic crankshaft journals and the crankpins overlap with each other.

Now look in the following drawing of the Bristol Hercules Sleeve Valve two-row Radial engine and compare:



There is no “overlap” between the basic journals and the crankpins of the crankshaft (all of small diameter).

See at the way the crankshaft is assembled.

Each crankpin serves 7 cylinders (which means 7 combustions per crankpin per two crank rotations).

The crankshaft drives a big propeller. The engine was / is one of the most reliable airplane engines ever.

Now look at the Cross-Radial PatAT (it has 8 combustions per two crankshaft rotations):



and let me know what makes it, according your opinion, less reliable than the Hercules.

Thanks
Manolis Pattakos

[manolis]

Hello J.A.W.

You write:
“& of course Manolis, 2T 120` triples have always been 'crossplane', with even firing/inertia torque = to 4T inline 6 cyl.


Not exactly.

The free inertia torque of a six in line is not so small.
It is about half than the free inertia torque of a I-4 having same pistons, same connecting rods and same stroke.
And it is of third order.

In a triple the inertia torque is half than in a six even firing 4-stroke.
The problem in the triple (either 2-stroke or 4-stroke) is the unbalanced inertia moment, which comprises a big 1st order constituent and a smaller, but still significant, 2nd order constituent.
Even if a first order balance shaft is used, the engine still has an unbalanced inertia moment to rock it.


In comparison, a crossplane I-4 two-stroke (with a first order counter-rotating balance shaft), can be considered perfect: even firing and fully balanced: no free inertia force, no free inertia torque, no free inertia moment.
As regards its overall vibration-free-quality, it is not worse than the best V-8 4-stroke engines.


The same is true for the Cross-Radial PatAT and PatATeco, but in this case no balance webs others than those two of the crankshaft are required (i.e. there is no need for an external balance shaft).

Thanks
Manolis Pattakos

[manolis]

Hello Gruntguru.

In the program balance.exe at http://www.pattakon.com/pattakonEduc.htm “Total Torque on Block” is the pair of inertia forces (rocking couple) applied on the casing in case the crankshaft is forced to rotate at constant angular velocity (say, by securing on it a huge flywheel).



Here is the comparison between this inertia torque in the typical I-4 with the plane crankshaft and in the same I-4 with a crossplane crankshaft:



The small one (less than 5% of the other) is of fourth order, the other is mainly of second order.

One can imagine it as a wave in the sea.
If the “crossplane crankshaft” wave is 1m high, the “plane crankshaft wave” is more than 20m high (we talk for two four cylinder engines having same pistons, same connecting rods, same stroke and running at same rpm).


If the crankshaft/flywheel/load is allowed to accelerate – decelerate, then the above inertia torque gives an idea of the resulting “crankshaft speed variation”.

Thanks
Manolis Pattakos

[J.A.W.]

Hi Manolis,
Yes, for sure the 120` inline triples have a 'rocking couple' characteristic, but firing-wise @ 120`also,
- the 2T matches the 4T inline six for power strokes.

Here is K. Cameron's article on the effect of the introduction of Yamaha's 'crossplane' crankshaft in Superbike racing : http://www.cycleworld.com/2010/06/01/th ... m-of-power

[manolis]

Hello J.A.W.

Quote from you interesting link at http://www.cycleworld.com/2010/06/01/th ... m-of-power :

“Yamaha engineer Masao Furusawa says the 90-degree crank is useful at full lean in turns, when the bike has little extra grip and the rider begins to first feed power. With a 180-degree crank, all pistons stop at either top or bottom dead center every 180 degrees of rotation. If the engine is turning 10,000 rpm at the bottom of its powerband, piston peak velocity is about 65 feet per second, so the engine’s approximately 3 pounds of pistons are trading some serious energy back and forth with the crank 333 times per second. The crank responds with an equally rapid rpm flutter, which is transmitted to the rear tire. The energy in each of these pulses is equal to the weight of a service automatic pistol falling seven stories. The rider is trying to feed power from full lean, but this crank-speed flutter is messing with traction in a rapid series of snatchy yanks, making the back tire “feel squirrelly,” so he waits. Riders on other makes aren’t waiting—they are gone.

Now, look at what happens in a 90-degree V-Twin. When one piston is at maximum velocity, the other piston is stopped, and 180 degrees later, that situation reverses. These pistons are trading energy, not with the crankshaft but with each other. As a result, the crank rotates more smoothly, and the rear tire is undisturbed by piston stop-and-start. The same is true of any Vee engine—V-Four, V-Five, V-Six. It is also true of a 120-degree Triple.

The main attraction of a 90-degree V-engine is its perfect primary balance. But that large Vee angle requires the crankshaft to be far back in the chassis, making it hard to get adequate weight on the front wheel. For this reason, many makers have chosen 60 degrees (previous Aprilia V-Twins, for example) or the KTM’s 75-degree Vee. These do require one or more balance shafts, but the overall compromise is attractive, and opening the Vee a bit from 60 allows more room for optimal intake setups. Aprilia’s V-Four has a 65-degree Vee.

But by rearranging an inline-Four’s crankpins at 90 degrees instead of 180, the same benefits result: When piston #1 is stopped at top or bottom center, piston #2 is near maximum velocity, and the same goes for #3 and #4. Suddenly, the engine is smooth. Our peerless star rider turns the throttle and the squirrelly feeling is gone. And so is he—down the next straight.”

End of Quote


Nice explanation of the differences between the I-4 with plane and cross-plane crankshaft.



Quote from the same link:
“Ducati has always given its Twins maximal valve area by going to the practical limit in big bore and short stroke. The four-cylinder makers have been less aggressive in this. Also, as bore is made bigger and stroke shorter, combustion tends to slow down because a tight chamber acts as a damper on flame speed. But, historically, Ducati’s engineers have been more clever at finding ways around this than have the four-cylinder brigade. To a great extent, development trumps design.”


Here we go again.

The 4-stroke PatRoVa rotary valve can further extend the “practical limit in big bore and short stroke”.

From the current 1.91:1 of the Ducati Panigale 1299, the bore to stroke ratio with the PatRoVa on the cylinder head can go substantially higher; say, at 2.5:1 or even more:



And this can change the rules of the game.

Thanks
Manolis Pattakos

[Tommy Cookers]

Hello Tommy Cookers.You write:“this is nonsense - just Yamaha advertising hype”

In the crossplane crankshaft In-Line-Four the energy reciprocating between the crankshaft (flywheel) and the set of the four pistons is practically zero. This is so because the increase of the kinetic energy of the two pistons equals to the decrease of the kinetic energy of the other two pistons.
The crankshaft does nothing else than interconnecting the four pistons.
The crankshaft needs not to accelerate / decelerate in order to provide / absorb kinetic energy from the set of the four pistons.

If you remove the cylinder heads and drive the crankshafts to rotate near, say, 10,000rpm, and measure the angular velocity of the crankshaft during a complete crankshaft rotation, in the crossplane crankshaft I-4 you will measure a practically constant angular velocity, while in the plane crankshaft I-4 you will measure a variation of the angular velocity.
The previous is equivalent to:
The free inertia torque in the crossplane I-4 is about zero, while the plane I-4 suffers from an extreme free inertia torque.

So, the I-4 with crossplane crankshaft is substantially different than the I-4 with plane crankshaft as regards the loading of the transmission, the loss of power in the transmission, the feeling of the rider, the side loading capacity of the contact between the rear tire and the road etc.

@ Manolis
you seem to write as if the Yamaha is not propelled by a series of explosions (it is not propelled by an external electric motor as you show)
these explosive energy pulses are in most engines most of the time greater than the inertial pulses (ok not at very high rpm - this I've said for years)
so the Yamaha is not totally different to other machines
their crankshafts all via their inertias absorb this energy and make it available at an almost steady rate
almost means they all have some rev-cyclic variations in rpm (ok the Yamaha at high rpm less than others - I have never said otherwise)
maybe 99.8% of the energy is in the steady component of rpm and maybe 0.2% is in the variation component
you write as if the road Yamaha doesn't have a transmission shock absorber
I imagine it does, and there's a large enough engine inertia and so a small enough variation component for the absorber to survive the warranty period
the crossplane crank's power impulses may in this regard be worse than the conventional crank's
at lower rpm the Yamaha may have a greater variation component and potential transmission loss than would a conventional even-firing machine

even if the race machine's variation component is eg the 333 Hz that you quote from a yamaha source in your recent post
and this 333 hz component travels through the chain and sprockets and reaches the wheel .....
surely only a small part of this very small component of the total energy will travel through the rear tyre to the road
the reason we use preumatic tyres is because they do not transmit energy variations at these high frequencies

so the engaging suggested explanation of the race traction benefits - is there any paper substantiating these effects ?
as I said, the explanation seems to discredit the supposed traction benefits on dirt surfaces of the single-cylinder engine

[manolis]

Hello Tommy Cookers.

Suppose:
50mm stroke
80mm bore (i.e. per cylinder capacity: 250cc)
100mm center to center con-rod length
500gr mass of the assembly of piston, wrist pin and top 1/3 of the connecting rod.
peak torque: 110mN at 12,000rpm
peak power: 200bhp at 14,000rpm.


At 10,000rpm a conventional four in-line (plane crankshaft) suffers from a free inertia torque of 700mN of 2nd order (it also suffers from a free inertia force of 14,000N (1.4ton), of 2nd order too, but this is another story).

At 10,000rpm, the cross-plane version of the same engine is actually rid of free inertia force and torque.


In a turn the motorcycle / rider is leaning as much as the turn and the speed allow.

Suppose that the rider needs, during the turn or near the end of the turn, only a part of the torque of his engine, say only 40mN.

He puts third in the gearbox, he engages the clutch and he opens slightly the throttle.

The engine is revving at 10,000rpm and is sending 40mN of “useful” torque (from the 110mN peak) to the transmission.

In the cross-plane version of the engine, the rear tire receives only useful torque (the 40mN provided by the engine multiplied by the total transmission ratio).

In the conventional (plane crankshaft) I-4, the rear tire receives the useful torque (the 40mN provided by the engine multiplied by the total transmission ratio), it also receives the useless inertia torque (the 700mN free inertia torque generated by the engine multiplied by the total transmission ratio).


With the typical shape / form of the instant combustion torque provided by the engine, in order to get an average of 40mN torque you need an instant torque varying from, say, -40mN to +100mN, i.e. a total variation of 140mN, while the free inertia torque of the conventional I-4 varies from –700mN to +700mN, giving a total variation of 1400mN (ten times higher).

Think for a second about it: the useful torque (i.e. the combustion torque) passes like noise inside a huge inertia torque that does nothing useful, only friction and stressing of the parts.

A bad thing for the plane crankshaft is that the basic frequency of the combustion torque is the same with the frequency of the free inertia torque. A high frequency torque absorber in the transmission cannot “kill” selectively one of them (even if there were such a shock absorber, by killing the free inertia torque you actually kill a part of the power; think of it).


So, while in the one case (cross-plane) the engine feeds the transmission exclusively with combustion torque (i.e. useful torque), in the other case the engine feeds the transmission system with the same useful (combustion) torque plus a ten times higher and useless inertia torque.

Even if the full throttle is open at 10,000rpm, the useful torque is still four times smaller than the inertia torque.

If you increase the revs to 14,000 of the peak power, things get two times worse because the inertia torque doubles ( (14,000rpm/10,000rpm)^2=2) while the useful combustion torque falls slightly.


So, there is a significant difference between the two arrangements.
The drawbacks of the cross-plane I-4 is the uneven firing and the need for an external counter-rotating balance shaft.


The Cross-Radial PatAT / PatATeco is rid of the drawbacks of the cross-plane I-4 because it does not need an external balance shaft and because it is even firing.
As for its crankshaft, it has only one crankpin and only two basic journals which, by the way, are rid of inertia loads.
The crankshaft is both: lighter and stronger.


Thanks
Manolis Pattakos

[J.A.W.]

Of attached significance, is the fact that of the current Moto GP machines ( limited to 4 cyl 4Ts of ~80mm bore)
is that nearly all are of V-configuration, or effectively mimic it, by dint of 90` 'crossplane' cranks, for inline 4s.

The weighty engine inertia characteristics of these litre capacity 4Ts has had to be tamed by intrusive ECU,
& by such devices as 'slipper' clutches - to prevent excessive back torque effects ( that GP 2Ts didn't suffer)..

[Brian Coat]

PATRoVA:

Has the thermo-structural FE analysis been done which confirms that the structure and its sealing will work under the required operating conditions?

[manolis]

Hello Brian Coat

You write:
“Has the thermo-structural FE analysis been done which confirms that the structure and its sealing will work under the required operating conditions?”


No.

Only conventional analysis has been done and it shows that the sealing will work under the required operating conditions.


Suppose that the rotary valve and the cavity on the cylinder head are both made of steel or of spheroidal graphite iron (modulus of elasticity: 200*10^9 N/m^2, linear temperature expansion coefficient: 12*10^-6 m/(mK) ).

Take 1ton (10,000N, 2,200lb) force acting on each oppositely arranged front.
The two fronts are secured to each other through the big diameter (say of 10cm2 section area, which means 35mm diameter) and short (say 25.4mm / 1’’) hub / shaft.
For simplicity forget the bending loads.


The tension load causes only a tiny extension of the hub (i.e. an increase of the distance between the two oppositely acting fronts): less than 0.01mm.


Take 100 degrees Celsius (180 F) maximum temperature difference between the PatRoVa rotary valve and the part of the cylinder head between the two oppositely arranged fronts (i.e. the cavity).

Such a temperature difference causes a 0.03mm change in the distance between the two oppositely arranged fronts (i.e. 0.015mm per side).


The above figures show that without sealing means other than the small clearance between the cooperating surfaces, the PatRoVa can keep the leakage inside acceptable limits.
No sealing means simply means: elimination of the friction in the cylinder head and no rev limit as regards the cylinder head.


It is not required, however there is the option of using INVAR, or other low coefficient of thermal expansion alloy, for the rotary valve and the cavity.



In practice there are some engine arrangements proving the previous “theory.

For instance the following OS Wankel model engine (no side seals):



Read at

http://www.f1technical.net/forum/viewto ... &start=795

and

http://www.f1technical.net/forum/viewto ... start=1020

the comparison of the leakage in the PatRoVa and in the OS Wankel model engine.

After reading the part with the calculated leakage from a “modified” to PatRoVa C50 Honda engine (the leakage, as a percentage of the entered mixture, is ten times less than in the Wankel model engine), think what this means in the case of a normal size (say 500cc per cylinder capacity) PatRoVa engine.

Thanks
Manolis Pattakos

[Brian Coat]

"Only conventional analysis has been done".

In 2016 thermo-structural FEA is not unconventional - it is taught to most / all undergraduates and open source code is readily available.

The trend in engineering is to more CAE prove-out with less/no prototypes and one of the big drivers of this trend is cost.

Would using open source CAE tools (FEA, CFD ...) provide a low cost way to further progress some of your your feasibility studies?