Jump to content


Photo

Seamless Gearboxes


  • Please log in to reply
631 replies to this topic

#51 Engineguy

Engineguy
  • Member

  • 989 posts
  • Joined: February 01

Posted 22 February 2012 - 13:24

I was under the impression the zeroshift involved a bisaed, one-way or sprag arrangemnt that allowed the upshifted ratio to overrun the previous ratio. Maybe I'm thinking of some other invention.

Edited by Engineguy, 22 February 2012 - 13:24.


Advertisement

#52 cheapracer

cheapracer
  • Member

  • 10,388 posts
  • Joined: May 07

Posted 22 February 2012 - 14:47

I was under the impression the zeroshift involved a bisaed, one-way or sprag arrangemnt that allowed the upshifted ratio to overrun the previous ratio. Maybe I'm thinking of some other invention.



...about a billion automatic transmissions!

#53 24gerrard

24gerrard
  • Member

  • 2,008 posts
  • Joined: March 11

Posted 22 February 2012 - 16:54

...about a billion automatic transmissions!


Indeed Cheapy and of course ideal for your idea of a free overrun LSD.
However I am sorry to say that such a sprag (one way clutch) is of no use in a layshaft/laygear gearbox.

#54 Engineguy

Engineguy
  • Member

  • 989 posts
  • Joined: February 01

Posted 22 February 2012 - 18:54

...about a billion automatic transmissions!


... and the mighty Lenco unit... powershifts 3000 hp without touching the clutch pedal, without coming off full throttle.

Lenco 4-speed manual planetary transmission
Note: Watch the tach needle during the run!

Edited by Engineguy, 22 February 2012 - 18:59.


#55 rachael

rachael
  • Member

  • 110 posts
  • Joined: October 10

Posted 22 February 2012 - 20:26

The previous gear is immediately overrun/unloaded. This makes no sense.


It makes perfect sense - Engineguy got it spot on!

I posted this description of how F1 seamless shift is achieved in one of two ways a while back;

In the twin barrel system one barrel controls the odd numbered shift forks - 1 3 5 7 and the other the even 2 4 6 forks. To upshift from say 3rd to 4th the odd barrel rotates first to bring the 3rd gear fork out of engagement - the dog ring is held in engagement by the torque being applied through the gear so the fork becomes preloaded. The even barrel rotates to move the 4th gear fork and dog-ring into engagement and 4th gear starts to transmit torque - because it is a higher gear this unloads the torque from the lower gear and the preloaded fork flicks the dog ring out of engagement with third gear before the gear dogs hit the over-run side of the dog ring.

In the ratchet system each of the mainshaft gears has an assembly that allows the gears to spin freely in over-run on the shaft. For the same 3rd to 4th shift the 3rd gear is driving the mainshaft, the barrel rotates to move the 4th gear fork and engage the dog ring with 4th gear. 4th gear starts to transmit torque and the 3rd gear is allowed to over-run the shaft by the ratchet system. The barrel continues to rotate to disengage the 3rd gear fork and dog ring completing the shift. The ratchet is locked for downshifts so that the gear can drive the mainshaft in over-run.

Ratchet is another name for sprag btw.

#56 mariner

mariner
  • Member

  • 1,381 posts
  • Joined: January 07

Posted 22 February 2012 - 20:30

Not seamless and 42 yers old but still fun.

the owner bough it as 21 yr old in 1969 ... Oh happy muscle car days!

http://www.youtube.c...feature=related

#57 24gerrard

24gerrard
  • Member

  • 2,008 posts
  • Joined: March 11

Posted 22 February 2012 - 21:36

Posted Image

Yes I used to drive one and a dragster.
Lenco was OK bit crude, my own clutchflite epicyclic units were better and faster for the smaller (up to 700bhp) applications we developed.

I see Rachael is back still unable to accept the impossibility of transfering high torque levels through a layshaft shift overlap.
The description given again for zeroshift and other methods of gear engagement baulking are fine Rachael.
The shifts are fast and torque is transfered up to and down from the actual shift point.
Ratcheting out and sprag run over can also be included with bullets or pawls.
Wont remove the gap in the middle though no matter what you do.
It is a stepped gearbox with a layshaft and that is that.

Oh and ratchet is not another word for sprag anymore that zeroshift or seamless are acceptable terms for any layshaft gearbox.

Edit: I do not wish to decry the abilities and obvious technical expertise of any of the posters on this thread.
I hope that the posts are of use to readers and might go some of the way in helping others to understand the usualy forgotten geartrain.

Edited by 24gerrard, 22 February 2012 - 21:49.


#58 24gerrard

24gerrard
  • Member

  • 2,008 posts
  • Joined: March 11

Posted 22 February 2012 - 22:03

Posted Image

This is my patented ESERU. Electric Shift Energy Recovery Unit.
It is a seven speed (F1) epicyclic transmission, that is also a seven segment electric motor/generator.
It is a stepped shifting gearbox with seven distinct ratios.
However the shifts use electrical induction from storage batteries and it is possible to transfer more torque during the complete shift between gears than when the unit is engaged and driven only by the prime mover.
The unit is also a KERS unit that only requires an electrical storage device or battery pack and an electrical control circuit to operate.
In direct top gear the only torque loss is from one support bearing, the geartrain rotates as one unit.
In all the lower ratios, only the planetary set engaged has gears moving relative to each other, all other gear sets are locked.
I need a few million, have you got it?

Oh yes and the size you are looking at is near the actual size for a complete geartrain.

Edit:-
Notice I do not claim this seven speed stepped gearbox is seamless or zeroshift, even though it is capable of transfering MORE torque through the complete shift than is available from the prime mover ic engine and it does so far smoother than any layshaft gearbox.

Edited by 24gerrard, 23 February 2012 - 09:57.


#59 rachael

rachael
  • Member

  • 110 posts
  • Joined: October 10

Posted 22 February 2012 - 22:05

I see Rachael is back still unable to accept the impossibility of transfering high torque levels through a layshaft shift overlap.


I'm not back, I've been here awhile trying to resist posting in a thread full of nonsense like 'If full torque is being transfered seamlessly as you keep trying to say is happening during gearshifts, then why in hells teeth is there a reduction in engine noise at the point of the gearshift.'

It's totally possible and occurs routinely - if the parts are capable of transmitting the peak torque pre and post shift why would they not be capable of transferring this torque during the shift? The torque during the shift is largely immaterial - it's the timing and control of particularly the twin barrel system that is critical to avoiding damage.



Advertisement

#60 cheapracer

cheapracer
  • Member

  • 10,388 posts
  • Joined: May 07

Posted 23 February 2012 - 06:20

... and the mighty Lenco unit... powershifts 3000 hp without touching the clutch pedal, without coming off full throttle.

Lenco 4-speed manual planetary transmission
Note: Watch the tach needle during the run!


Great to watch specialists equipment in it's environment.

Notice he gets it all out the way by half distance then uses torque to accelerate him through the tough half.

Wait, what .... :lol:


#61 24gerrard

24gerrard
  • Member

  • 2,008 posts
  • Joined: March 11

Posted 23 February 2012 - 09:40

I'm not back, I've been here awhile trying to resist posting in a thread full of nonsense like 'If full torque is being transfered seamlessly as you keep trying to say is happening during gearshifts, then why in hells teeth is there a reduction in engine noise at the point of the gearshift.'

It's totally possible and occurs routinely - if the parts are capable of transmitting the peak torque pre and post shift why would they not be capable of transferring this torque during the shift? The torque during the shift is largely immaterial - it's the timing and control of particularly the twin barrel system that is critical to avoiding damage.


I agree with your explanation as to the operation of the systems, I always have.
So explain the spike.
How can the torque during the shift be immaterial?
Timing and control is indeed critical.
So explain the spike. Why is it there and is it a torque spike?
If what you say (as to there being a constant transfer of torque) were true, why would there be a spike (either torque or power)?
On an unloaded test bed the zeroshift and seamless parts will operate fine and the timing and control will be enough.
Under load on the road at the point of shift overlap (whether ratcheted, sprag overrun or simply preloaded dogs) the components will not be capable of handling maximum torque. Control over the source of torque must be added.
Therefore non of these systems can be described as seamless or 'zeroshift'.
Unless it is accepted in marketing terms by those with vested interests that is.

Resisting 'nonsense' and then taking a hollier than thou stance could be seen as a sign of arrogance and an inability to debate, I am sure that this is not the case with your obvious engineering abilities.

EDIT: A change in engine note is a result in a change in engine RPM.
This fact on its own will be accompanied by a 'change' in torque output from the engine.
Therefore the shift is NOT seamless.

Edited by 24gerrard, 23 February 2012 - 09:50.


#62 24gerrard

24gerrard
  • Member

  • 2,008 posts
  • Joined: March 11

Posted 23 February 2012 - 09:44

:lol: You go girl! :up:


Nothing to add then? :well:

#63 Wuzak

Wuzak
  • Member

  • 3,534 posts
  • Joined: September 00

Posted 23 February 2012 - 12:04

EDIT: A change in engine note is a result in a change in engine RPM.
This fact on its own will be accompanied by a 'change' in torque output from the engine.
Therefore the shift is NOT seamless.


Surely the term "seamless" refers to there being no interruption to torque being delivered through the gearbox to the wheels, and not to change in the torque being transmitted?

#64 Kelpiecross

Kelpiecross
  • Member

  • 903 posts
  • Joined: October 10

Posted 23 February 2012 - 12:37

Surely the term "seamless" refers to there being no interruption to torque being delivered through the gearbox to the wheels, and not to change in the torque being transmitted?


I have to admit that I am not sure what this argument is about.

Is a seamless gearbox not strictly speaking literally seamless but because the change is so quick it really is virtually (or exactly) equivalent to being seamless or uninterrupted?

#65 24gerrard

24gerrard
  • Member

  • 2,008 posts
  • Joined: March 11

Posted 23 February 2012 - 13:31

No. I am aware I am not qualified enough to reply on the merit.

But she knows what she is talking about.


I recognise this but it does not instantly make her correct.
I have no problem with the basic operation of F1 stepped layshaft gearboxes.
It is the definition of seamless and zeroshift I find almost offensive.

#66 24gerrard

24gerrard
  • Member

  • 2,008 posts
  • Joined: March 11

Posted 23 February 2012 - 13:33

I have to admit that I am not sure what this argument is about.

Is a seamless gearbox not strictly speaking literally seamless but because the change is so quick it really is virtually (or exactly) equivalent to being seamless or uninterrupted?


Here it is again the 'virtual' world.
When you crash into a wall are you virtualy dead?

#67 24gerrard

24gerrard
  • Member

  • 2,008 posts
  • Joined: March 11

Posted 23 February 2012 - 13:42

Surely the term "seamless" refers to there being no interruption to torque being delivered through the gearbox to the wheels, and not to change in the torque being transmitted?


If the torque has to be reduced at the point of shift overlap, then the amount of torque at that point is dictated by the weakness of the components slipping, ratcheting or overrunning the gear dogs.
This can only be described as 'partial' unloaded engagement and forms a gap or seam if you must use this seamstress' clothing term.
I was the first to suggest the control over gearshifts in layshaft boxes by cutting or reducing the torque from the engine during shift overlaps.
The result of this consultation was the semi/fully automatic shift system in Mansels Ferrari.

I accept that since then a great deal of development has gone on to improve and speed up the gearshifts.
Non of this has changed the basic concept or limitations of all layshaft geartrains however.

#68 StressedDave

StressedDave
  • New Member

  • 8 posts
  • Joined: October 09

Posted 23 February 2012 - 15:05

It is the definition of seamless and zeroshift I find almost offensive.


Almost? To me this smacks more like an argument between engineers and scientists - while there is undoubtedly some form of step change in the torque delivery as a result of the shift which has to be reacted somewhere - XTrac seem to suggest it's via windup in the driveshafts etc. - there isn't the reduction to zero that there is in the standard dog ring transmission., hence to the driver it appears 'seamless'.

Or to use the vernacular it's seamless enough for all practical purposes (engineering) rather than being seamless to suit an arbitrarily anal definition (scientist).

IIRC, the research done by the various dual clutch transmission manufacturers shows that if shift time is less than 0.2s then the 'average' driver can't detect any interruption in the drive to the wheels.


#69 MatsNorway

MatsNorway
  • Member

  • 2,069 posts
  • Joined: December 09

Posted 23 February 2012 - 15:19

I have heard about rallycross cars that have utilised the flex in the driveshafts to get better launches of the line. But that has nothing to do with gearshifts..

Edited by MatsNorway, 23 February 2012 - 15:21.


#70 cheapracer

cheapracer
  • Member

  • 10,388 posts
  • Joined: May 07

Posted 23 February 2012 - 16:02

I have to admit that I am not sure what this argument is about.




#71 24gerrard

24gerrard
  • Member

  • 2,008 posts
  • Joined: March 11

Posted 23 February 2012 - 17:57


:up: Saves having to move to China :lol:

#72 carlt

carlt
  • Member

  • 1,064 posts
  • Joined: June 09

Posted 23 February 2012 - 20:08

I have to admit that I am not sure what this argument is about.



Is the argument not just about semantics ?

I am sure I don't understand the finer details of the transmissions being discussed here
but from what little I grasp it would appear that the next gear is engaged and an 'increase' in torque/power disengages the previous gear

therefore the change is not strictly 'seamless' in that there is a change in engine revs

but if that change in engine revs is due to an 'increase' in torque/power to the driven wheels , then the rate of acceleration is 'seamless'

?

#73 24gerrard

24gerrard
  • Member

  • 2,008 posts
  • Joined: March 11

Posted 23 February 2012 - 20:27

Those who can use a dog ringed layshaft gearbox and have the skill to achieve 'racing' gearchanges will confirm that upshifts can be achieved with the engine at full throttle and no lift through an 'almost' instant shift.
Bang and its done.
Doing so puts high load and wear on the dog rings and most who use the technique only use it in the most competitive of circumstances.
This manual 'racing' change can match the modern so called seamless and zeroshift systems for speed.
There is definitely a gap or 'seam' in the shift.

#74 carlt

carlt
  • Member

  • 1,064 posts
  • Joined: June 09

Posted 23 February 2012 - 21:28

Those who can use a dog ringed layshaft gearbox and have the skill to achieve 'racing' gearchanges will confirm that upshifts can be achieved with the engine at full throttle and no lift through an 'almost' instant shift.
Bang and its done.
Doing so puts high load and wear on the dog rings and most who use the technique only use it in the most competitive of circumstances.
This manual 'racing' change can match the modern so called seamless and zeroshift systems for speed.
There is definitely a gap or 'seam' in the shift.



Thats a totally different argument

#75 rachael

rachael
  • Member

  • 110 posts
  • Joined: October 10

Posted 23 February 2012 - 22:18

Those who can use a dog ringed layshaft gearbox and have the skill to achieve 'racing' gearchanges will confirm that upshifts can be achieved with the engine at full throttle and no lift through an 'almost' instant shift.
Bang and its done.
Doing so puts high load and wear on the dog rings and most who use the technique only use it in the most competitive of circumstances.
This manual 'racing' change can match the modern so called seamless and zeroshift systems for speed.
There is definitely a gap or 'seam' in the shift.


This is all true for a standard dog ring manually shifted box but does not apply to F1 boxes because two gears are not engaged at the same time. Dog damage occurs in these boxes because no matter how skilled the driver he cannot do a perfect shift every time - occasionally there will be a dog on dog engagement or partial engagement and this is when the damage occurs. F1 boxes use clever electronics to learn the position of the dog windows and so time the shift to perfection so minimising the dog wear or damage.


#76 24gerrard

24gerrard
  • Member

  • 2,008 posts
  • Joined: March 11

Posted 23 February 2012 - 23:04

This is all true for a standard dog ring manually shifted box but does not apply to F1 boxes because two gears are not engaged at the same time. Dog damage occurs in these boxes because no matter how skilled the driver he cannot do a perfect shift every time - occasionally there will be a dog on dog engagement or partial engagement and this is when the damage occurs. F1 boxes use clever electronics to learn the position of the dog windows and so time the shift to perfection so minimising the dog wear or damage.


It was simply a comparison Rachael.
I know that the timing can be so fine in a modern F1 system, that a shift can be undertaken between one pair of gear teeth coming off load and the next pair coming under load. That is a very very small gap.
However, the dog out to dog in time during a manual racing change using a dog ring manual box can be just as fast.
Why then do they not describe a manual dog ring gearbox as a 'seamless' or 'zeroshift' gearbox?

Your understanding of engagement puzzles me.
You state that partial engagement can occur in a dog ring manual box and it can often damage the components. I fully agree.
Surely a so called 'seamless' box is therefore also 'partialy' engaged in all conditions other than when the gear set selected is 'fully' engaged.
How then can you state 'both gears are engaged at the same time'?
In this condition you know the gearbox would grenade.

Surely it is more precise to describe the system as having two gears partialy engaged through the gearshift, so long as there is a wind up mechanism to account for the differences in shaft rpm.
This mechanism will only operate properly if the torque being tranfered through it is carefuly controlled in balance with the level of engagement throughout the shift.
The result is a smoother shift than the original torque off pulsed shift with a distinct air gap, we originaly played with.
It would be interesting to see if the later systems are actualy more efficient at transfering torque, or for that matter faster than a manual dog ring box.

Edited by 24gerrard, 23 February 2012 - 23:24.


#77 gruntguru

gruntguru
  • Member

  • 5,372 posts
  • Joined: January 09

Posted 24 February 2012 - 03:19

the next gearset actually has to begin turning the output shaft faster than the previous gear before the previous gear can dis-engage

And just how does it do this with the output shaft connected to the rear wheels which control the rpm the output shaft turns at??????????????????

Didn't read my post properly huh? Just how it does "this" is initially by:
1. "winding up" the drivetrain, tailshaft, axles, tyres, engine mounts, suspension bushings etc etc. Then by-
2. tyres breaking traction and/or
3. Vehicle speed increasing

Edited by gruntguru, 24 February 2012 - 04:07.


#78 gruntguru

gruntguru
  • Member

  • 5,372 posts
  • Joined: January 09

Posted 24 February 2012 - 03:32

You CANNOT do it using small mechanical components between the shafts and gears especialy when these parts are at their least engaged and least torque capable geometric condition, one set on one gear and one set on another. The available torque has to be reduced to allow the change in ratio to be achieved and a new safe path for torque transfer established.

The Zero shift has a negatively bevelled dog which cannot transmit any torque until it is fully engaged and fully "torque capable". The available torque only needs to be reduced to the extent required to avoid a "spike" using any of a number of suitable methods as with DSG type transmissions. Torque to the rear wheels can be maintained without any gap or dip.

#79 gruntguru

gruntguru
  • Member

  • 5,372 posts
  • Joined: January 09

Posted 24 February 2012 - 03:46

Is a seamless gearbox not strictly speaking literally seamless but because the change is so quick it really is virtually (or exactly) equivalent to being seamless or uninterrupted?

I gues the true meaning of the word "seamless" when applied to gearboxes could be the property of the person who first married the term but I think it is farly clear that it should mean a stepped shift in ratio where the step is not actually felt because torque to the drive wheels is maintained during the shift, at or above the levels available before or after the shift. Both DSG and zeroshift are capable of precisely that. Incidentally a conventional dog box is not capable of that - there will always be a gap however tiny.

Edited by gruntguru, 24 February 2012 - 03:48.


Advertisement

#80 gruntguru

gruntguru
  • Member

  • 5,372 posts
  • Joined: January 09

Posted 24 February 2012 - 03:51

If the torque has to be reduced at the point of shift overlap, then the amount of torque at that point is dictated by the weakness of the components slipping, ratcheting or overrunning the gear dogs.
This can only be described as 'partial' unloaded engagement and forms a gap or seam if you must use this seamstress' clothing term.

The input torque may need to be reduced to avoid an inertia "spike" but the relevant "torque" is output shaft torque which does not have to reduce during the shift for DSG and zeroshift boxes.

#81 24gerrard

24gerrard
  • Member

  • 2,008 posts
  • Joined: March 11

Posted 24 February 2012 - 09:10

The Zero shift has a negatively bevelled dog which cannot transmit any torque until it is fully engaged and fully "torque capable". The available torque only needs to be reduced to the extent required to avoid a "spike" using any of a number of suitable methods as with DSG type transmissions. Torque to the rear wheels can be maintained without any gap or dip.


You are correct, 'The Zero shift has a negatively bevelled dog which cannot transmit any torque until it is fully engaged and fully "torque capable".'

So why the contradiction? Torque to the rear wheels can be maintained without any gap or dip.

#82 24gerrard

24gerrard
  • Member

  • 2,008 posts
  • Joined: March 11

Posted 24 February 2012 - 09:12

I gues the true meaning of the word "seamless" when applied to gearboxes could be the property of the person who first married the term but I think it is farly clear that it should mean a stepped shift in ratio where the step is not actually felt because torque to the drive wheels is maintained during the shift, at or above the levels available before or after the shift. Both DSG and zeroshift are capable of precisely that. Incidentally a conventional dog box is not capable of that - there will always be a gap however tiny.


Incorrect, if a dog ring layshaft box changes gear fast enough, the 'illusion' is the same.

#83 24gerrard

24gerrard
  • Member

  • 2,008 posts
  • Joined: March 11

Posted 24 February 2012 - 09:32

The input torque may need to be reduced to avoid an inertia "spike" but the relevant "torque" is output shaft torque which does not have to reduce during the shift for DSG and zeroshift boxes.


If you reduce the torque through the shift components, how can the output shaft still transfer torque to the rear wheels?
It will continue to rotate at road speed rpm.
The output shaft does assist in 'dragging' the engine rpm down to match the next higher gear during upshifts.
It does this through the shift mechanism during the shift and it is a ' potential reversal' of the torque path.
It only becomes a reversal of the torque path if it is allowed to exceed the torque from the prime mover ic engine.
The situation then becomes overrun engine braking.

Remove the engine and you have potential for 'full' energy recovery from vehicle inertia.
Unfortunately at present F1 bolts a motor/generator on the crankshaft and subtracts the parasitic losses from the ic engine from the recovered energy.
This makes balancing engine braking, KER and wheel braking a difficult and pointless exercise.

#84 24gerrard

24gerrard
  • Member

  • 2,008 posts
  • Joined: March 11

Posted 24 February 2012 - 09:46

Didn't read my post properly huh? Just how it does "this" is initially by:
1. "winding up" the drivetrain, tailshaft, axles, tyres, engine mounts, suspension bushings etc etc. Then by-
2. tyres breaking traction and/or
3. Vehicle speed increasing


Many clutch plates in ordinary powertrains have radial shock absorbing springs.
I have designed such radial springs into powertrain designs either in clutch plates or as seperate units for high performance applications.
The idea can be used to allow wind up in shift systems and even act as a form of traction limiting.

However under full torque transfer, (maximum acceleration), these radial shock absorbers placed in the torque path will already be fully compressed.
Any other parts of the powertrain capable of wind up will also be fully wound up. (bit like me I suppose :lol: )
Therefore non of the items you list can play a part in assisting the shift mechanism when it is asked to transfer torque during your suggested maximum torque upshift.

#85 Kelpiecross

Kelpiecross
  • Member

  • 903 posts
  • Joined: October 10

Posted 24 February 2012 - 10:41

I gues the true meaning of the word "seamless" when applied to gearboxes could be the property of the person who first married the term but I think it is farly clear that it should mean a stepped shift in ratio where the step is not actually felt because torque to the drive wheels is maintained during the shift, at or above the levels available before or after the shift. Both DSG and zeroshift are capable of precisely that. Incidentally a conventional dog box is not capable of that - there will always be a gap however tiny.


I still can't quite comprehend this "seamless" shift busimess. DSG I think I understand - here it is two friction clutches momentarily sharing the transmitted torque - essentially the same situation as a traditional automatic gearbox changing ratio. But "Zeroshift" - the information on the internet says two gears are engaged at once so there is no interruption in torque flow. But traditionally two gears engaged at once means that the system locks up or goes "crunch" . How can two gears be engaged at the one time? Is the change so fast (8 milliseconds is quoted as being the shift time) that the gears don't have time to rotate enough to begin binding?
Presumably the shift is carried out electrically/electronically somehow - what if the electrics suddenly go dead (Mr. Lucas - Prince of Darkness or a similar problem) in the middle of a shift with the car travelling at high speed - is there any failsafe system etc?

#86 24gerrard

24gerrard
  • Member

  • 2,008 posts
  • Joined: March 11

Posted 24 February 2012 - 11:30

I still can't quite comprehend this "seamless" shift busimess. DSG I think I understand - here it is two friction clutches momentarily sharing the transmitted torque - essentially the same situation as a traditional automatic gearbox changing ratio. But "Zeroshift" - the information on the internet says two gears are engaged at once so there is no interruption in torque flow. But traditionally two gears engaged at once means that the system locks up or goes "crunch" . How can two gears be engaged at the one time? Is the change so fast (8 milliseconds is quoted as being the shift time) that the gears don't have time to rotate enough to begin binding?
Presumably the shift is carried out electrically/electronically somehow - what if the electrics suddenly go dead (Mr. Lucas - Prince of Darkness or a similar problem) in the middle of a shift with the car travelling at high speed - is there any failsafe system etc?


So called 'seamless' systems use a slipping or ratcheting mechanism operated through the shift 'overlap. in the same way you understand the overlap during a shift in a DSG gearbox.
The difference is the slip/ratcheting/overrun clutch parts are contained in or on each main gear, in a gearbox with a single laygear, rather than in two clutch packs on two shafts as used in a DSG gearbox. The slip being in the two clutch packs one on each shaft.
The downside of the DSG is the much larger torque loss and oil windage with the extra shaft and gears and the larger size and weight.
The DSG is also far more robust and is for road vehicles. The systems used in F1 only have to achieve race distance/allocated number of races.

The internet description is not good enough. Two gears are not 'FULLY' engaged at the same time, they are partialy engaged and there is a reduction in torque 'flow' (whatever that is) if not a complete gap in 'flow'. At the center point of the shift when both gears have the shift mechanism at its least capable operating geometry there is very little chance of noticable torque 'flow' in this condition.
Looks like the internet is now comparing the concept to water dynamics instead of the equaly incorrect tailoring analogy.

The speed of shift does indeed help the operation by limiting how far the gears rotate during the shifts. It is clever engineering, no doubt of that.

Edited by 24gerrard, 24 February 2012 - 11:33.


#87 jrobson

jrobson
  • Member

  • 62 posts
  • Joined: March 09

Posted 24 February 2012 - 11:35

Perhaps the marketing department is referring to the acceleration of the vehicle rather than the torque applied to the drive train as being seamless, because the end user is more interested in the former than the latter. It does not matter if there is a torque spike or dip for that matter, if the shift is quick enough acceleration would continue, the only true seamless system, by your definition is already invented and is installed on a lot of road cars, but is banned in F1.

#88 24gerrard

24gerrard
  • Member

  • 2,008 posts
  • Joined: March 11

Posted 24 February 2012 - 11:55

Perhaps the marketing department is referring to the acceleration of the vehicle rather than the torque applied to the drive train as being seamless, because the end user is more interested in the former than the latter. It does not matter if there is a torque spike or dip for that matter, if the shift is quick enough acceleration would continue, the only true seamless system, by your definition is already invented and is installed on a lot of road cars, but is banned in F1.


A CVT transmission is the only type of gearbox that is truely seamless.
Unfortunately they need lots of energy to operate and only give a 1:1 to 4:1 ratio range on average.

The nearest gearbox in a grand prix car to be seamless, using todays incorrect term, was the Wilson pre-selector fitted to the earliest ERA's of the 1930's.
We have such a gearbox in a 1935 Armstrong Siddeley once owned by Harold MacMillan.
ERA won races with it until they ruined the powertrain with a conventional crash box. (Colloti was it? I forget, Tony?)
Alec Stokes prefered the Wilson although it was heavy. We discussed modern versions of this concept on many occasion.
Tony Rudd used to drive Remus one of these Wilson ERA cars up Goodwood hill most years.
I projected a bevel epicyclic geartrain to achieve a 25 speed gearbox based on it. That had power on upshifts.
That was well into the 'seamless' definition as wrongly applied today but I would still not describe it as such.

Edited by 24gerrard, 24 February 2012 - 11:58.


#89 Kelpiecross

Kelpiecross
  • Member

  • 903 posts
  • Joined: October 10

Posted 24 February 2012 - 13:31

[quote name='24gerrard' date='Feb 24 2012, 22:55' post='5548251']
A CVT transmission is the only type of gearbox that is truely seamless.
Unfortunately they need lots of energy to operate and only give a 1:1 to 4:1 ratio range on average.

Do you think an all-mechanical/all teeth and gears/chains etc. postive engagement/rigid connection between input and output/works both in input driving output and vice versa type of CVT possible? No friction belts etc. or electrical gizmos, or ratchets allowed either - purely mechanical only.
This is something I have taken a lot of interest in over the years - generally speaking you can't get an opinion on this subject from an engineer.

#90 24gerrard

24gerrard
  • Member

  • 2,008 posts
  • Joined: March 11

Posted 24 February 2012 - 14:35

Do you think an all-mechanical/all teeth and gears/chains etc. postive engagement/rigid connection between input and output/works both in input driving output and vice versa type of CVT possible? No friction belts etc. or electrical gizmos, or ratchets allowed either - purely mechanical only.
This is something I have taken a lot of interest in over the years - generally speaking you can't get an opinion on this subject from an engineer.


There are a number of variations on the CVT principle.
The Van Doorne system originaly called the variomatic used cones and belts.
This type was modified into the chain and cone type used by ford in a limited edition fiesta and a small volvo.
Most modern CVTs use this concept with additional gearing.

The Prius hybrid uses an epicyclic gear set with input from two electrical sources and an ic engine and balances the rpm between the inputs and outputs.
A useful idea for a hybrid or a hybrid with range extension.
It has a high torque loss and limited ratio range. Like the other types it uses a fairly large amount of energy to operate the ratio mechanism.

The toloroidal system uses tilting disks and relies on drive friction by using a special friction fluid.
This system was adapted by flybrid to drive an energy storage flywheel for KERS F1 and hybrid road car use.
The toloroidal is suitable for instrumentation and light use but not realy high torque applications.
It is better to use a sealed flywheel in a vacuum and to drive it and recover energy from it using electrical induction without breaking the seal.
Patrick Head swears by this idea but then he swore by the Williams CVT F1 car especialy when the FIA banned it.

There is a varied range of hydraulic CVT systems mostly used in industrial, plant agricultural and heavy mover vehicles.
Usual method is an engine driven oil pump driving an hydraulic motor, (they have been in use for decades,) although there are some linear and hydro/pneumatic types which show promise with reduced weight and size.
I built an hydraulic diff for F1 on these principles, which could have been developed into a dual path CVT powertrain. (banned by the regs as usual)

Another problem with cvt transmissions is the lack of feel they give the driver. Without the positive engagement you have with a stepped ratio gearbox the feeling is of constantly changing gear, a soggy indirect feeling.

Edited by 24gerrard, 25 February 2012 - 10:47.


#91 cheapracer

cheapracer
  • Member

  • 10,388 posts
  • Joined: May 07

Posted 24 February 2012 - 16:53

Do you think an all-mechanical/all teeth and gears/chains etc. postive engagement/rigid connection between input and output/works both in input driving output and vice versa type of CVT possible? No friction belts etc. or electrical gizmos, or ratchets allowed either - purely mechanical only.
This is something I have taken a lot of interest in over the years - generally speaking you can't get an opinion on this subject from an engineer.


There's about 10 different patented versions over the last 100 years that I've seen and a couple in development - "Investers Wanted" ....... one guy had contacted me a while ago and hasn't gotten back to me.

I had a loose design of one myself but found a patent of it already that was pretty much the same - I'm now suing him for 'physic intellectual theft'.


#92 Wolf

Wolf
  • Member

  • 7,881 posts
  • Joined: June 00

Posted 24 February 2012 - 17:24

I know how you feel Cheapy- happened to me once with my new and revolutionary brake concept of reducing number of brakes by no less than 25%... I was all set to set the automotive world ablaze, only to find out someone already stole my idea and put it on a F1 car decades before. I still maintain there was something fishy about it, and am not ruling out a scheme involving time travel... :p

#93 cheapracer

cheapracer
  • Member

  • 10,388 posts
  • Joined: May 07

Posted 24 February 2012 - 20:45

I know how you feel Cheapy- happened to me once with my new and revolutionary brake concept of reducing number of brakes by no less than 25%...


Not only did BRM beat you to it, the Speedway boys beat you by 100 years :lol:

BTW, did you know many Toyota light commercial trucks and Landcruisers since the 60's have their handbrake on the tailshaft?


#94 Wolf

Wolf
  • Member

  • 7,881 posts
  • Joined: June 00

Posted 24 February 2012 - 20:52

So you're saying it was a conspiracy?!? :evil: That explains it... :lol: In the meantime, I've heard of Toyota... (and speaking of Toyota, I might drop you a mail about something related)

#95 Catalina Park

Catalina Park
  • Member

  • 5,729 posts
  • Joined: July 01

Posted 24 February 2012 - 22:53

BTW, did you know many Toyota light commercial trucks and Landcruisers since the 60's have their handbrake on the tailshaft?

And they copied British trucks.

The Australian made International C1100 ute had 4 wheel disc brakes back in the late 60s. It was the first Australian made vehicle with 4 wheel discs (the Renault R8 was not made here, it was assembled here) It was easy to fit rear discs when the handbrake was on the back of the gearbox.

Posted Image

Edited by Catalina Park, 24 February 2012 - 22:57.


#96 Lee Nicolle

Lee Nicolle
  • Member

  • 6,028 posts
  • Joined: July 08

Posted 25 February 2012 - 00:19

Not only did BRM beat you to it, the Speedway boys beat you by 100 years :lol:

BTW, did you know many Toyota light commercial trucks and Landcruisers since the 60's have their handbrake on the tailshaft?

And Dodges and Inters and fords etc. The only thing is when you use it as a emergency brake the tailshaft falls on the road!! been there and done that with 22 tonnes of truck and load [it was a 60s truck with single circuit brakes] Luckily it was a light hill and I used the plowed road edge to stop.

#97 Lee Nicolle

Lee Nicolle
  • Member

  • 6,028 posts
  • Joined: July 08

Posted 25 February 2012 - 00:26

As for this thread with seamless changes. Whatever it will never happen. Auto, constant mesh whatever, You cange gear and the accelaration rate changes, however slightly.
My experience riding in a new Porker with flappy paddles is that it was slow and inneficient and a normal manual box was far better. Even watching in car with modern sports cars it really does not seem any faster.
Bring back clutches and proper gearboxes, for racing make a driver earn his keep and for street just to enjoy the experience, and the efficiency. If you dont wanna change by an auto!

#98 gruntguru

gruntguru
  • Member

  • 5,372 posts
  • Joined: January 09

Posted 25 February 2012 - 04:33

You are correct, 'The Zero shift has a negatively bevelled dog which cannot transmit any torque until it is fully engaged and fully "torque capable".'

So why the contradiction? Torque to the rear wheels can be maintained without any gap or dip.

No contradiction at all. The previous gear continues to drive until the new gear is fully engaged and driving "harder" causing the drive dog to disengage from the previous gear.

#99 gruntguru

gruntguru
  • Member

  • 5,372 posts
  • Joined: January 09

Posted 25 February 2012 - 04:35

Incorrect, if a dog ring layshaft box changes gear fast enough, the 'illusion' is the same.

Sure but it is only an illusion. DSG and Zeroshift have NO gap in torque transmitted.

Advertisement

#100 gruntguru

gruntguru
  • Member

  • 5,372 posts
  • Joined: January 09

Posted 25 February 2012 - 04:41

If you reduce the torque through the shift components, how can the output shaft still transfer torque to the rear wheels?

It is not essential that the torque be reduced. If the zeroshift box is upshifted without reducing engine torque the is simply a spike in output shaft torque prior to settling to the steady-state torque available in the new gear. There is NO GAP and no reduction in torque to the output shaft. Torque reduction is only required to smooth the spike for comfort, driveline durability, driver control etc.