2 stroke thread (with occasional F1 relevance!)

[manolis]

Hello MudFlap.


In the last posts there are three plots for the inertia and combustion torques; these plots were made based on the energy transfer between the cooperating parts of the engine.

The plots taken with the DOS balance program (older posts) were made based on the inertia forces, on the leaning of the connecting rods, on the resulting thrust forces onto the cylinder liners etc, etc.

Two substantially different approaches / solutions for the same problem.



Here is another plot for the Honda Civic engine revving at 9,000 rpm:



The yellow curve is the Position of a piston vs the crank angle..

The thin curves are the Cylinder Pressure inside the cylinders of the engine.

The cyan curve is the total Combustion Torque from all the four cylinders.

The orange horizontal line is the Mean Torque on the crankshaft of the engine (it is the mean value of the cyan curve). It is 170mN (the overall friction loss is taken as 20%) .

The white curve is the total Inertia Torque.

Adding the Combustion Torque (cyan curve) and the Inertia Torque (white curve) we take the Combined Torque on the crankshaft (green thick curve).

This is for the case the throttle is completely opened.

In case of closed throttle valve, the Combined Torque is the Inertia Torque (I.e. the white curve).

In simple words, when at 9,000rpm you release the gas pedal, the transmission line has to “deal” with a substantially larger torque: 3,200mN peak-to-peak versus 2,700mN peak-to-peak.



Hello J.A.W.

I.e. when the driver of a racing motorcycle powered by a 4-cyl flat-crank engine closes the throttle to decelerate before a curve, the rear tire “tries” to get what is “going on” (zero useful power, yet increased idling / useless power).

With the same engine modified to “crossplane” (like the Yamaha R1) the throttle opening is directly proportional to the torque being applied (provided) to the transmission and to the rear tire.

Thanks
Manolis Pattakos

[63l8qrrfy6]

Hi Manolis,

See below a similar concept where the kinematics can be adjusted on the fly to reduce any order. Probably not much use in flatplane l4 where the 2nd order appears to dominate in most practical situations but more appropriate for engine configurations where the dominating excitation changes between inertia torques and gas pressure torques with engine speed.

https://www.google.com/url?sa=t&source= ... 27jRhp3VoZ

[manolis]

Hello MudFlap

You write:
"See below a similar concept where the kinematics can be adjusted on the fly to reduce any order. Probably not much use in flatplane l4 where the 2nd order appears to dominate in most practical situations but more appropriate for engine configurations where the dominating excitation changes between inertia torques and gas pressure torques with engine speed.
https://www.google.com/url?sa=t&source= ... 27jRhp3VoZ"


From the above link:








The green and the while curves of the plot of my last post (i.e. the Total Torque provided by the engine of the Honda Civic at 9,000rpm at full load and at no load) can be interpreted as follows:

• Case A (green curve, full open throttle):
  the car is “pushed” strongly forwards, then it is pushed backwards but less strongly, then it is pushed strongly forwards again, and so on.

• Case B (white curve, closed throttle):
  the car is “pushed” strongly (as strongly as in the case A) forwards, then it is pushed strongly backwards, then it is pushed strongly forwards again, and so on.

That is, with closed throttle the forward pushes are as strong as with wide open throttle, while the backward pushes are stronger than with wide open throttle . . .

Thanks
Manolis Pattakos

[Tommy Cookers]

well thanks to Manolis for all these plots for us to get excited over ...

Hello MudFlap.
In the last posts there are three plots for the inertia and combustion torques; these plots were made based on the energy transfer between the cooperating parts of the engine.
The plots taken with the DOS balance program (older posts) were made based on the inertia forces, on the leaning of the connecting rods, on the resulting thrust forces onto the cylinder liners etc, etc.
Two substantially different approaches / solutions for the same problem.

Hello J.A.W.
I.e. when the driver of a racing motorcycle powered by a 4-cyl flat-crank engine closes the throttle to decelerate before a curve, the rear tire “tries” to get what is “going on” (zero useful power, yet increased idling / useless power).
With the same engine modified to “crossplane” (like the Yamaha R1) the throttle opening is directly proportional to the torque being applied (provided) to the transmission and to the rear tire.

'Two substantially different approaches/solutions for the same problem' - ???
No, not the same problem, because ....

'the inertia forces ... the leaning of the rods .... resulting thrust forces onto the cylinder liners etc ...'
aren't relieved by eg the PatVRA or similar - (or any engine layout eg whether crossplane or V or whatever)
ie there's friction from the inertia forces that's only relieved by reducing rpm


'zero useful power, yet increased idling/useless power' - ???
No, the 'useless' power doesn't exist .... because ....

there's essentially no work done to the load ......
because there's purely elastic compliance of the transmission (an accidental but inevitable property)
well 99.9% pure and so 0.1% lossy
the 'useless' power is analogous to the electrically 'useless' reactive power known as 'Wattless power'
(the inertial force is analogous to reactance voltage - immaterial unless eg it increases cost or bulk of parts)

this elastic angular compliance could (mistakenly) be designed-out by using much heavier/bulkier transmission designs
yes minimal angular compliance may be very important in driving a gun turret


btw there was a 2-element flywheel in the 650cc Suzuki 'Tempter' motorcycle - a development of the 180 deg 450cc twin
afaik it used the whole flywheel around town but for quick gearshifts at high rpm only used partial flywheel
a twin that behaved like a four

[manolis]

Hello Tommy Cookers

I wrote:

• 'Two substantially different approaches/solutions for the same problem' - ???

You write:

• “No, not the same problem, because ....
  'the inertia forces ... the leaning of the rods .... resulting thrust forces onto the cylinder liners etc ...'
  aren't relieved by eg the PatVRA or similar - (or any engine layout eg whether crossplane or V or whatever)
  ie there's friction from the inertia forces that's only relieved by reducing rpm”

The “common” problem is the calculation / analysis of the inertia torque of the plane-crank 4-cylinder engine.

With the DOS-balance program the “solution” of the problem ”what is the unbalanced inertia torque”, is:



and the way to make this plot is to calculate the accelerations of the various parts, then to calculate the required forces for these accelerations, then to calculate the leaning of the connecting rods, then to calculate the eccentricity of the crankshaft center from the line connecting the big-end and small-end centers of the connecting rod, and so on.


With the other method (the energy method), the plot of the unbalanced inertia torque is made based on the energy transfer between the cooperating parts of the engine:



In the second method all we need to know is the displacement of the pistons relative to the crank angle.


Worth to note: the one plot is the Inertia Torque on the engine casing, the other plot is the Inertia Torque on the crankshaft (equal and opposite, "action re-action").




You also write:

• No, the 'useless' power doesn't exist .... because ....”

The strong idling/useless power maximizes (and minimizes) two times per crankshaft rotation; depending on the revs, it can be several times stronger than the combustion (i.e. the useful) power; the (over)loading of the relative parts, the vibrations, the noise, the worse feeling of the drive-tire hook and the consumption of mechanical energy (into friction loss) confirm its existence.

The elastic connection between the flywheel and the transmission line is a “solution” of the problem, but not a panacea.


Quote from the Rostock University presentation (previous post in this page):

• A non-uniformly transmitting mechanism is proposed that couples a flywheel kinematically to the crankshaft of a combustion engine in order to compensate fluctuating engine torques.
  . . .
  An advantage is that the oscillations are compensated directly at the crankshaft where they arise.
  …
  Thus, the principle of the mechanism works well, although the mechanical design has to be improved with respect to fatigue strength”

The typical plane-crank 4-cylinder engine (i.e. the most common engine in cars and big motorcycles) has the worst “torque fluctuations” among all engines: because all the four pistons stop simultaneously and, 90 crankshaft degrees later, all the four pistons move at high speed.

The PatVRA mechanism can compensate the 2nd order torque fluctuations of the 4-cylinder even-firing, being simple, cheap, lightweight and strong (i.e. reliable).

Thanks
Manolis Pattakos

[gruntguru]

I'm not sure if the following point has been made. The PatVRA assembly is not heavily loaded. It does not "see" or absorb the "inertia torque". It simply allows the crankshaft to move with the irregular angular velocity required to maintain constant kinetic energy of all the moving parts in the engine.

(This is why I dislike the "inertia torque" concept. With a kinematic solution like the Pat VRA, there is no torque fluctuation anywhere in the entire system so the "inertia torque" is not a "real" torque.)

This characteristic means the PatVRA mechanism is not heavily loaded - seeing only the combustion torque - and does not need to be as robust as one might think.

[manolis]

Hello Gruntguru.

You write:

• “(This is why I dislike the "inertia torque" concept. With a kinematic solution like the Pat VRA, there is no torque fluctuation anywhere in the entire system so the "inertia torque" is not a "real" torque.)”

This “not real torque” is the simplest way to describe the problem and give an idea of its severity.
It is the unbalanced Inertia Torque in the case of constant angular velocity of the crankshaft.
One could translate it as the Torque that can be avoided in case a proper kinematic mechanism (say, like the PatVRA) interconnects the crankshaft with the flywheel.

Instead of the above unbalanced Inertia Torque, one could use the fluctuation of the crankshaft angular velocity during a cycle; but how many people can get what this means, and how severe the problem is?

• Using the “not real torque” things become a lot easier to get and compare:
  
  With the Honda engine revving at 9,000rpm, the Inertia Torque is six times larger than the 2nd order component of the (full open throttle) combustion torque, and this 2nd order component of the combustion torque is the – by far – dominant component of the combustion torque (as shows the plot at the bottom of the previous page).
  
  
  
  The above plot shows the combustion torque harmonics of a four-cylinder Diesel engine (compare their relative sizes with those of the Honda Civic approach), and is taken from https://www.researchgate.net/publicatio ... r_velocity

Regarding the compensation of some harmonic components of the combustion torque (Rostock University, kinematic mechanism, link in previous post), the kinematic mechanism must be widely adjustable because:
the inertia (compensating) torque increases with revs squared,
and because:
the combustion torque varies widely with the throttle valve position.




The PatVRA eliminates, at all revs and loads, the Inertia Torque passing to the transission line of the four-cylinder even-firing four-stroke, making it a better engine than the four-cylinder uneven-firing crossplane (Yamaha-R1 like).

Thanks
Manolis Pattakos

[Tommy Cookers]

these inertia torques ........
(that could significantly pass to the load only if the transmission had hypothetical high frequency response)

would they pass to the load (hypothetical transmission as above) if said transmission was of ....
the 1924 Mercedes (4 cylinder Targa Florio) functionality ?
ie gearbox driving a differential having 2 concentric output shafts each driving its own crownwheel and pinion ....
one pinion driving on its left its crown wheel - the other pinion driving on its right the other crown wheel ....
as in Tipo B & Bimotore (2 engines/opposite rotation) Alfa Romeos(ok 8 cyls so no hypothetical inertia torque issues)


inertia-related engine frictional moments on engine 'casing' ........
yes seem to be eliminated by 2 opposite-rotation crankshafts (ie coupled by direct gearing) .....
eg Ariel square 4s and c.1980 Kawasaki 250/350 WC winners and road replicas (also the Valveless 2 stroke car ?)
(thus according to MZ the piston side load helps exhaust port action)
though yes frictional moments aren't eliminated in typical GP V4s eg RG500s ie same-rotation crankshafts


btw/NOTE TO SELF .....
not only did 1950s BRM 2.5 litre 4 cyl GP car have 4 main bearings without problems - though 5 were used later
Vanwall did the same (mbs posns 1 and 2, big counterweight posn 3, mbs posns 4 and 5)
until conventional 5 mbs introduced in 1958
Stewart Tresilian being 'Mr 4 mains' from the 1930s
was the higher rpm used on the transition to 'pump fuel' (=Avgas) in 1958 the reason ?
amusingly Vanwall won using 108/135 Avgas in 1958 (though historians all seem to think 100/130 Avgas was the ruling)


and ....
'Mr Jetpack' Bentley was on TV advocating jetpack racing - over lakes (UK-style 1 metre deep 'gravel pits' presumably)

[J.A.W.]

these inertia torques ........
(that could significantly pass to the load only if the transmission had hypothetical high frequency response)...

...inertia-related engine frictional moments on engine 'casing' ........
yes seem to be eliminated by 2 opposite-rotation crankshafts (ie coupled by direct gearing) .....
eg Ariel square 4s and c.1980 Kawasaki 250/350 WC winners and road replicas (also the Valveless 2 stroke car ?)
(thus according to MZ the piston side load helps exhaust port action)
though yes frictional moments aren't eliminated in typical GP V4s eg RG500s ie same-rotation crankshafts


btw/NOTE TO SELF .....

and ....
'Mr Jetpack' Bentley was on TV advocating jetpack racing - over lakes (UK-style 1 metre deep 'gravel pits' presumably)

Kawasaki GP 250/350 'tandem twin' 2T engine crankshafts were initially phased for 180 degree firing,
but were revised to 360/0 degree (simultaneous firing*) due to the the former arrangement causing
intrusive/destructive vibration frequencies, which were harmful/fatiguing for both bike & rider.

(AFAIR, W.O. Bentley advocated a minimal practicable number of main bearings in his designs,
which duly exposed their limitations in AM-Lagonda race-usage at extended rpm/comp-ratio.)

Some time back I suggested that Manolis might contact RedBull racing to assist him with a view
to including his 'personal flyer' as addition to the existing RB-sponsored FAI air-racing series...


*An ultimate big bang 'twingle' set-up, & one which Kawasaki tuner Erv Kanemoto had previously
tried in 'tringle' form, by re-phasing his 750 2T triple from 120 to 360/0 degrees in an attempt to
gain traction on dirt, for US Grand National flat-track racing. However in this instance, vibration
was found to be excessive, plus transmission damage/failures resulting from the torque pulse, were
'deal-breakers' (plus the US AMA banned GN racing bikes with more than 2 cylinders, shortly after).

[manolis]

Hello Tommy Cookers

You write:

• “these inertia torques . . . would they pass to the load (hypothetical transmission as above) if said transmission was of ....
  the 1924 Mercedes (4 cylinder Targa Florio) functionality ?
  
  ie gearbox driving a differential having 2 concentric output shafts each driving its own crownwheel and pinion ....
  one pinion driving on its left its crown wheel - the other pinion driving on its right the other crown wheel ....”

The inertia torques pass to the wheels no matter what transmission is utilized.

If you look it from the energy view point, it is quite simple.

The crankshaft (together with the flywheel) accelerates and decelerates to absorb and return kinetic energy from/to the set off the “four pistons – connecting rods”.

No matter what the architecture of the transmission is, the drive wheels / tires (which turn at the same direction) follow these accelerations / decelerations trying to accelerate / decelerate the vehicle.

I.e. excluding the pistons / connecting rods, the rest vehicle is used as a kinetic-energy accumulator (or tank):
when the pistons / connecting rods decelerate, the vehicle accelerates to absorb the kinetic energy lost by the pistons / connecting rods,
and 90 crankshaft degrees later, the vehicle decelerates to provide kinetic energy to the accelerating pistons / connecting rods.

The Mercedes 1924 (Targa Florio) had a 2 lt in-line four-cylinder engine providing 68bhp at 4,500rpm without the compressor (which means 110mN mean (i.e. combustion) torque) and 120bhp at 4,500rpm with the compressor on (which means 190mN mean (i.e. combustion) torque).

The engine was operated, from time to time, at 6,000rpm.

With 129mm stroke and 70mm bore, at 6,000 rpm the mean piston speed is 25,8m/sec! (a century ago)

Supposing a piston-assembly mass of 1Kg (250mm is a reasonable "center-to-center" connecting rod length for 129mm stroke ), the following plot “talks” for an inertia torque of 3,300mN:



Even with the compressor on, this inertia torque is more than 15 times larger than the mean combustion torque.

Thanks
Manolis Pattakos

[manolis]

Hello J.A.W.

Here is a racing Rotax tadem:



Simultaneous combustion in the two cylinders.

In case of “even firing”, the unbalanced moment of the typical 2-stroke even firing twin (crankpins at 0 and 180 degrees) is cancelled out, in expense of a heavier (due to the necessarily longer offset of the cylinder axis) inertia torque.


With the two crankshafts phased (simultaneous combustions, as in the Rotax), the inertia torque on the engine casing is cancelled out ****, in expense of two times heavier combustion torque pulses; by the way: with the two crankshafts counter-rotating the first order inertia force is completely cancelled out without external balance shafts.

• **** The inertia torque on the crankshaft / transmission is not cancelled. It actually increases as the following plot shows:
  
  

You write:
“Kawasaki GP 250/350 'tandem twin' 2T engine crankshafts were initially phased for 180 degree firing,
but were revised to 360/0 degree (simultaneous firing*) due to the the former arrangement causing intrusive/destructive vibration frequencies, which were harmful/fatiguing for both bike & rider.”


Think of having two such, 350cc each, 2T engines directly secured on your body.


In the OPRE Tilting of the Portable Flyer:



the "basis" (i.e. the body of the pilot) is rid of all kinds of inertia vibrations.



And due to the kind of the load / transmission:



its "basis" is also rid of combustion vibrations of all kinds.

Thanks
Manolis Pattakos

[Tommy Cookers]

You write:

• “these inertia torques . . . would they pass to the load (hypothetical transmission as above) if said transmission was of ....
  the 1924 Mercedes (4 cylinder Targa Florio) functionality ?
  ie gearbox driving a differential having 2 concentric output shafts each driving its own crownwheel and pinion ....
  one pinion driving on its left its crown wheel - the other pinion driving on its right the other crown wheel ....”

what I had in mind was the reaction moments (at the gearbox mounts) due to the inertia torque moments in transmission
(ok I didn't make that clear)

for 2 reasons .....
1. claimed benefit of PatVRA as reduction in vibration from alternating torques around the eg engine/gearbox mounts
2. famous benefit of 'TF' Mercedes axle - no loss of traction due to 'torque effect' (before invention of limited slip diff)

so doesn't 'TF' cancelling at gearbox mounts transmission torque around gearbox shaft cancel torque vibration as in 1. ?

[manolis]

Hello Tommy Cookers.

You write:

• “what I had in mind was the reaction moments (at the gearbox mounts) due to the inertia torque moments in transmission
  (ok I didn't make that clear)
  
  for 2 reasons .....
  1. claimed benefit of PatVRA as reduction in vibration from alternating torques around the eg engine/gearbox mounts
  2. famous benefit of 'TF' Mercedes axle - no loss of traction due to 'torque effect' (before invention of limited slip diff)
  
  so doesn't 'TF' cancelling at gearbox mounts transmission torque around gearbox shaft cancel torque vibration as in 1. ?”

Quote from https://primotipo.com/2020/09/11/benz-tropfenwagen/





The engine is the six cylinder version, however the rest (gearbox / differential, TF axle) seem to be the same.

I can’t find photos / drawings of the “internals” of the differential / semi-shafts. If you have some link or photo, please post it.



Regarding the unbalanced inertia torque on the engine casing:

With the gearbox / differential directly secured to the engine casing, and with the semi-shafts (of the rear wheels) arranged perpendicular to the crankshaft rotation axis, the inertia torque about the gearbox shaft is cancelled out; however another inertia torque is created, and this torque tries to vibrate the engine / gearbox / differential assembly about the rear wheels axis two times per crankshaft rotation.

And it doesn’t matter how many pinions and ring gearwheels are used in the differential.

This is so because the two rear wheels (i.e. the load) turn at the same direction.

The extreme variation of the kinetic energy of the set of the four pistons of the 4-cylinder engine (Mercedels Targa Florio, 1924) is translated into an extreme variation of the push of the drive wheels/tires to the road (and into an equally extreme variation of the push of the road to the vehicle).


*************************************************************************

EDIT / corection

The above images are not the Mercedes Targa Florio of 1924.

The following images from https://www.motorsportmagazine.com/arch ... o-mercedes show the Mercedes Targa Florio 1924 (engine at the front):







Regarding the inertia torque, things are worse becasue the differential is away from the engine, so it cannot be secured to the engine casing.

• As before, the extreme variation of the kinetic energy of the set of the four pistons of the 4-cylinder engine (Mercedels Targa Florio, 1924) is translated into an extreme variation of the push of the drive wheels/tires to the road (and into an equally extreme variation of the push of the road to the vehicle).

END OF EDIT

*************************************************************************


Compare it to the OPRE Tilting of the Portable Flyer wherein the two propellers counter-rotate cancelling out (on the basis of the structure, I.e. on the shoulders / torso of the pilot) every kind of inertia and combustion torque (the inertia forces are also perfectly cancelled out because of the symmetrical architecture of each engine).



Thanks
Manolis Pattakoe

[Tommy Cookers]

Hello Tommy Cookers.You write:

• what I had in mind was the reaction moments (at the gearbox mounts) due to the inertia torque moments in transmission
  for 2 reasons .....
  1. claimed benefit of PatVRA as reduction in vibration from alternating torques around the eg engine/gearbox mounts
  2. famous benefit of 'TF' Mercedes axle - no loss of traction due to 'torque effect' (before invention of limited slip diff)
  so doesn't 'TF' cancelling at gearbox mounts transmission torque around gearbox shaft cancel torque vibration as in 1

....Regarding the unbalanced inertia torque on the engine casing:
With the gearbox / differential directly secured to the engine casing, and with the semi-shafts (of the rear wheels) arranged perpendicular to the crankshaft rotation axis, the inertia torque about the gearbox shaft is cancelled out; however another inertia torque is created, and this torque tries to vibrate the engine / gearbox / differential assembly about the rear wheels axis two times per crankshaft rotation.
And it doesn’t matter how many pinions and ring gearwheels are used in the differential.
This is so because the two rear wheels (i.e. the load) turn at the same direction.
The extreme variation of the kinetic energy of the set of the four pistons of the 4-cylinder engine (Mercedels Targa Florio, 1924) is translated into an extreme variation of the push of the drive wheels/tires to the road (and into an equally extreme variation of the push of the road to the vehicle).
Regarding the inertia torque, things are worse becasue the differential is away from the engine, so it cannot be secured to the engine casing.

the Benz is totally different to the Mercedes (they were rival companies at that time)

the Mercedes gearbox isn't fixed to the engine
the Mercedes differential seems fixed to the gearbox not fixed to the axle - the Motorsport article P.83 saying ....
'at the rear of the gearbox is the universal joint and spherical bearing of the torque-tube type axle ....
typically Mercedes with a crown wheel integral with each halfshaft, the differential being between the 2 pinions'
and remember the chain drive cars c.1900-1908 had their differentials at the gearboxes
(afaik the 1924 Mercedes functionality was later emulated by Alfa Romeo in their Tipo B/'P3' and 'Bimotore' cars)
http://www.grandprixhistory.org/alfa-romeo-bimotore.htm

the point of the Mercedes axle is that each wheel/road contact load is unaffected by the engine power being used
its design eliminates the undesirable effects of other axles in this regard - hence my question
presumably those silly old chain-driven axles had the same characteristic
(chain-driven axles were generally replaced by 'Panhard' ie torque-tube axles - partly treating those undesirables)

[saviour stivala]

Alfa Romeo tipo A (Jano) 1931 was powered by 2 6C-1750 engines mounted side-by-side and driving the rear wheels through a ‘pair’ of transmissions, drive shafts and differentials into a single solid rear axle.