Seamless Shift

[Cogs]

You are over complicating your explanation.

I disagree.

Yes the clutch can be allowed to 'slip' during the shift to allow the engine to increase in rpm to match the new gear engaged to output.

We were discussing the seamlessness of positive torque upshifts, actually the engine speed needs to decrease to match not increase.

A similar effect is achieved by either disengaing the clutch during the shift altogether or by just backing the ignition to reduce torque output from the engine. The engine can also be throttled down or even turned off completely, resulting in little or no torque transfer.

Agree, but this is not in my explanation and perhaps is the cause for confusion.

I cannot see a dynamic torque spike during a gear shift.

The spike is caused by the synchronisation of the inertia of the upstream components when reflected to the point of impact.

The 'gap' is a time period where there is little or no torque to be measured at the output shaft.

This 'gap' does not exist. There is always measurable torque at the output shaft the only change is from the output torque available from the current gear and the output torque available from the next. During the transition between these two values torque is being delivered without interruption. Please explain why you believe this output torque has to be zero or negligible. If we can agree on this then I believe the seamlessness will be resolved.

[autogyro]

In upshifts, yes the engine rpm has to decrease. Which was my 'bad' sorry, however the engine is also forced to change rpm upward on downshifts as well and although this is not under power the shifts can still unbalance the car if not carefuly controlled.

The spike you are measuring is at the engine output, gearbox input train.
It is a result of the engine being forced to decrease in rpm as the shift components or main clutch finish slipping and is when they 'fully engage'. It can result in a very short duration wheelspin as the engine becomes fully engaged in the higher gear. It depends on the method used and may not be present at all if complete disengagement during the shift and good timing and system control is used.

If you measure the torque at the output shaft during the shift however, there will always be a 'gap' where there is either no torque or very low torque.
There is low torque if the clutch or engagement components in the design are 'slipping' during the shift. There is 'no' torque if the engine is either completely disengaged during the shift or the fuel and ignition is backed off to control the shift.

Controlling the transition between the output low/zero torque during shifting and the sudden spike on full engagement 'you mention' is essential to ensure smooth traction.
IMO the torque spike does not have to or should not exist with modern systems, certainly not in so called seamless dual clutch geartrains.

[flynfrog]

It is not a rant and proves that certain people do not know the capabilities in how to change gear with a manual layshaft box.
The time it takes for a driver using lever actuated shift, to move a dog ring selector from one gear to another can be just as fast as any modern trick method.
The difference is the skill and experience involved. Now sadly lacking in most performance drivers.

Um no wrong again. A driver has a limited amount of speed he can move is hand and a limited amount of force he can apply. A poorly controlled air shifter with ignition cut is many times faster than a driver doing the same thing.

[autogyro]

Wrong. The speed of shift is limited by the operation of the mechanical engagement parts beetween the slider assemblies on the shafts and the gears.
It does not matter what is used to actuate the mechanism, human hand, hydraulic, pneumatic or elecro magnetic, the speed of shift given by the actual internal gear engagement mechanism will be the same.
When to change gear is (unless fully automatic) decided in advance in the drivers brain and the time taken to move the lever far enough to start the 'set' actuation mechanics time, can be pre determined in that brain.
There is no difference in the pre determined shift speed.

As a matter of interest, trying to trick the mechanical system with higher scrole ramp angles and faster actuation systems was the cause of at least one DNF this year for teams using a certain type of gearbox. I tried to tell them but they did not listen. If you exceed the actual gear engagement speed by using an external shift mechanism that is to fast you will get system jam ups and stripped gears.

[Cogs]

First of all I'm really enjoying this thread and will continue to post until the whole issue regarding seamlessness is resolved.

Regarding the latest tangent, again I totally agree with flynfrog.

Fact.....no human can compete with modern actuation, engine management and clutch control systems, especially considering that the controller can determine the optimum shift point for performance.

The common theme here though is a failure to accept that modern ECU and TCU technology when combined with innovative mechanical solutions such as Zero Shift provide alternatives!

I have attempted to propose a feasible explanation for a seamless shift but keep getting contradictory responses that continually reference "old" conventional methods of shift management and mechanical geometry.

Take a step back and accept that engine speed can be synchronised across a slipping clutch whilst supplying constant torque through engine management and clutch modulation. Then realise that with "unconventional" dog ring geometry torque reduction is not required prior to engagement. Then we'll continue the whole point of this thread.

The 'gap' is a time period where there is little or no torque to be measured at the output shaft. This 'gap' is always present in a stepped gearbox and is the reason why the description 'seamless' is untrue and simply a marketing term.

Then hopefully agree that this statement is completey untrue.

[Cogs]

Wrong. The speed of shift is limited by the operation of the mechanical engagement parts beetween the slider assemblies on the shafts and the gears.

Wrong. Yet again this refers to conventional dog box architecture. Realise that the next gear being engaged may not require the dog ring that is currently being driven. Provide either a control system that guarentees extraction before failure or as the case may seem the Zero Shift components and the rest becomes irrelevant.

They probably did not listen because you do not understand their system.

[autogyro]

There realy are NO modern mechanisms for gear shifting in a layshaft gearbox,only modern methods of operating the basic engage/disengage mechanism, which are in effect dog clutches.

Now lets look at it a different way and ask a question.

Do we all agree that during an upshift, the engine has to reduce in rpm over a controlled time period? (dictated by the internal shaft/gear mechanism, not the control system)
If you have to answer yes to this question, then during this time period the engine is not transfering torque because it is reducing in rpm and is not under load accelerating the drive wheels.
If this reduction of engine rpm is not achieved by controlling fuel flow and or ignition and is achieved by forcing the engine to reduce in rpm using a slipping main clutch or slipping gear engagement components, then torque will be converted in part into heating up and wearing the slipping components.The remainder of the engine torque will be stored as inertia at a higher engine rpm than when fully engaged to the output shaft and on release this inertia will create the torque spike you mention which if uncontrolled could break drive wheel traction and unbalance the car.

[Cogs]

There realy are NO modern mechanisms for gear shifting in a layshaft gearbox,only modern methods of operating the basic engage/disengage mechanism, which are in effect dog clutches.

How can you possibly assume this without being in contact with every single transmission design house on earth, as well as knowing exactly the internal architecture of every transmission that claims to be able to perform a seamless shift?

Do we all agree that during an upshift, the engine has to reduce in rpm over a controlled time period?

Yes

If you have to answer yes to this question, then during this time period the engine is not transfering torque because it is reducing in rpm and is not under load accelerating the drive wheels.

Wrong!!!!!!!!!!!!!!!!!!!!!!!!!!!
The engagement has caused the input shaft to synchronise to the new gear speed, the remaining delta speed can be used without fuelling the engine to maintain input torque via clutch control. As the engine and input speed near synchronisation engine torque can be restored and clutch capacity can be increased back to driver demand. During this time the input shaft torque can be maintained.

The result..............A SEAMLESS SHIFT

[autogyro]

Wrong. The speed of shift is limited by the operation of the mechanical engagement parts beetween the slider assemblies on the shafts and the gears.

Wrong. Yet again this refers to conventional dog box architecture. Realise that the next gear being engaged may not require the dog ring that is currently being driven. Provide either a control system that guarentees extraction before failure or as the case may seem the Zero Shift components and the rest becomes irrelevant.

They probably did not listen because you do not understand their system.

I have read most of the data on the zero shift system and also designed similar systems.
Such systems can smooth the shift and increase shift speed (so long as the engine inertia is low enough).
There is no way to get over the time it takes for the engine to change rpm between gears however no matter how hard you market such mechanisms.

In anycase layshaft gearboxes are both ancient technology and far from the best method of transfering torque from input to output. Dual clutch or dual shaft gearboxes labeled 'seamless' are even less gear efficient as they have dual gearsets dragging in oil with mostly unused gears in each gear ratio applied.

Electric vehicles need far more efficient powertrains and the future will see totaly new gearbox technology with far higher efficiency than layshaft boxes seamless or otherwise.

[autogyro]

There realy are NO modern mechanisms for gear shifting in a layshaft gearbox,only modern methods of operating the basic engage/disengage mechanism, which are in effect dog clutches.

How can you possibly assume this without being in contact with every single transmission design house on earth, as well as knowing exactly the internal architecture of every transmission that claims to be able to perform a seamless shift?

Do we all agree that during an upshift, the engine has to reduce in rpm over a controlled time period?

Yes

If you have to answer yes to this question, then during this time period the engine is not transfering torque because it is reducing in rpm and is not under load accelerating the drive wheels.

Wrong!!!!!!!!!!!!!!!!!!!!!!!!!!!
The engagement has caused the input shaft to synchronise to the new gear speed, the remaining delta speed can be used without fuelling the engine to maintain input torque via clutch control. As the engine and input speed near synchronisation engine torque can be restored and clutch capacity can be increased back to driver demand. During this time the input shaft torque can be maintained.

The result..............A SEAMLESS SHIFT

Here we see you totaly missing the fundamental point.
The engagement has caused the input shaft to syncronise to the new gear speed.
A TIME DELAY that occurs as the bullets ride up the ramps.Exactly the same as dog clutches or syncro rings, just a tad faster. IT IS NOT IN ANYWAY SEAMLESS.

[autogyro]

There is only one possible 'seamless' gearbox.
That is one shaft joined to the crank at one end and the drive wheels at the other.
(I have one that has this at least in top direct gear).

There is no way that any stepped gearbox can be seamless and there is no need to look at every gearbox ever made to know this.

Seamless is nothing more than a marketing ploy.

[Cogs]

Here we see you totaly missing the fundamental point.

No sir, YOU have completely missed the fundamental point. The new gear has caused the input shaft to synchronise the previous gear has become instantaneously unloaded. There is no "time delay".

To even mention the function of a conventional synchroniser ring in this context is yet another example applying old principles to modern technology.

[xpensive]

The only true "seamless gearshift" there ever was in Formula one was Wiliams CVT, sadly banned by those in power.

http://www.forix.com/8w/altpower/williamscvt.jpg

[Cogs]

There is only one possible 'seamless' gearbox.
That is one shaft joined to the crank at one end and the drive wheels at the other.

If this is your definition of a gearbox then I really am on a hidding to nothing.

There is no way that any stepped gearbox can be seamless and there is no need to look at every gearbox ever made to know this.

Perhaps you just have to take a look and understand the ones you haven't seen yet.

[Cogs]

Before we continue and to get this thread back on topic, can you confirm that you believe the following STILL to be true?

The 'gap' is a time period where there is little or no torque to be measured at the output shaft. This 'gap' is always present in a stepped gearbox and is the reason why the description 'seamless' is untrue and simply a marketing term.

If so I will make the effort to get my point across [-o<