Fastest gearchanges were pre-1990's

[autogyro]

Reading Autogyros post earlier, perhaps I misunderstood, but...

Actually, the largest contribution to torque fluctuation is due to the change in engine speed during the shift.
Torque disturbance due to inertia torque of the engine can be over 1000Nm at the layshaft - these engines are "decelerating" (or accelerating for the purists) at over 150,000rpm/s.

Where you are not traction limited however, there is no need for torque cut or clutch modulation. XTrac \ Machin are 100% correct, the compliance of the entire drivetrain and tyre is sufficient to dampen the torque spikes, preventing damage to the gearbox and reducing risk of tyre slip.

The need for engine cuts or clutch modulation depends on many other factors in the over all car package.
It is possible to allow all the torque disturbance to be absorbed by the output of the geartrain, tyres etc in some circumstances.
However, the shift speed at overlap is still limited by having to convert the rotating inertia to torque (spike).
This means a manual shift using a modern (double barrel)shift mechanism and the same weight of geartrain could have the same speed as a seamless shift.
In fact it could even be a seamless manual shift.(xtrac have one I believe)
However a dog shift system would give better manual control over the shift and could also be as fast at its maximum. The driver could far better control the speed of the over all shift to meet track circumstances using a dog system.

The ideal would be a double barrel odd even fork control with dog engagement capable of both manual and semi automatic shifting.
One dog engageing at 'almost' the same time as the other is disengageing in seamless use.
For precise manual use the 'gap' at shift overlap would need to be under the drivers control.
Maybe a sequential mechanism with a delay available from how long the paddles or buttons were held during triggering. This would delay the movement of one barrel relative to the other. One odd one even of course, not two on the same slider hub.

[Tommy Cookers]

Thank you for your efforts to inform us, now I think we have a consensus, something we can all believe in more or less.

All this talk of dogs reminds me of seeing (on some British motorcycle gearboxes) modifications to the dogs (eg removing some, undercutting the rest) to bring them up to standard of the best makes. This free-and-easy dog design allowed near-instantaneous clutchless upchanges, by using the magneto cutout button at full throttle.
This behaviour was in part attributed to the lower gearbox inertia (the gearboxes and clutches were running at less than half crankshaft speed). Those few machines with the clutch and gearbox input shaft (or just the clutch) running at crankshaft speed all had slow changes.
When 'F1 designers'(Cosworth, Hesketh and Weslake) designed power units for motorcycles the resulting gearboxes all had poor changes, attributed to oversizing.

I think the dog design is the key to fast gearchanges, different dogs for different gearbox layouts.
The last F1 race in 1965 (1500cc) was won by the Honda, whose transverse V12 had a central power takeoff, allowing a slow-turning gearbox, motorcycle style.

[MrEngineer]

Hmmm I'm not sure if I'm misunderstanding you Autogyro.

Do you work with current F1 control systems?

[autogyro]

Reading Autogyros post earlier, perhaps I misunderstood, but...

What do you misundertand?

Actually, the largest contribution to torque fluctuation is due to the change in engine speed during the shift.
Torque disturbance due to inertia torque of the engine can be over 1000Nm at the layshaft - these engines are "decelerating" (or accelerating for the purists) at over 150,000rpm/s.

I agree and it is the change in rpm between each ratio that sets the minimum shift speed at shift/overlap.

Where you are not traction limited however, there is no need for torque cut or clutch modulation. XTrac \ Machin are 100% correct, the compliance of the entire drivetrain and tyre is sufficient to dampen the torque spikes, preventing damage to the gearbox and reducing risk of tyre slip.

Gear shift circumstances where there is no need for torque cuts or clutch modulation in F1 is because of the nature of the car.

It is not just that it may in some conditions on track be 'not traction limited'.
The cars are very light and this light weight gives the potential to use acceleration to dampen the shifts.

The rotating components in the engine, the clutch and the input part of the geartrain is much lighter than in other vehicle design, so the forced change in rpm of these components at the gearshift over/lap can be achieved faster and with less torque disturbance. This is exceeded in racing motorcycle gearbox shift systems, because they tend to being much lighter and as rightly mentioned in a previous post, often have a speeded up gear shift because the gearbox input is geared down. This however results in the need to design in a higher torque capability in the geartrain as a whole.

The high levels of downforce ensure that (subject to how well set up the car is) the cars are using the tyre traction to the maximum available. This prevents any potential traction break away caused by a 'fierce' and non torque controlled shift.
Many of the losses of car control on worn tyres (resulting in going off track or worse) are however, as a direct result of these fierce unmodulated gear shifts taking effect on handling when there is no longer huge tyre grip (because of worn tyres) to mask the results.

A manually shifted system with better modulation would allow the driver more control in such circumstances.

[machin]

The high levels of downforce ensure that (subject to how well set up the car is) the cars are using the tyre traction to the maximum available. This prevents any potential traction break away caused by a 'fierce' and non torque controlled shift.

If the car is already using 100% of the grip than any extra force disturbance will make the tyre slide. The "high level of downforce" is irrelevent (the tyre doesn't know if its aerodynamic downforce, or simply weight of the car above it).

Many of the losses of car control on worn tyres (resulting in going off track or worse) are however, as a direct result of these fierce unmodulated gear shifts taking effect on handling when there is no longer huge tyre grip (because of worn tyres) to mask the results.

Evidance of this (part in bold)? Xtrac advise that the seamless shift mechanism is actually smoother than traditional gear change systems ;

The drivers say it is very smooth and actually unsettles the car less in corners

and they're doing faster lap times(i.e. average vehicle acceleration MUST be higher with seamless shift to achieve faster lap times).

[seamless shift] is typically worth 3 tenths a lap

[Tommy Cookers]

'3 tenths a lap'

What is that as a proportion of each gearchange time ?

[machin]

Maybe I'm guessing your thoughts wrong but remember it's not quite as simple as saying "time saved per lap/number of gear changes equals difference in gear change time per shift"... even when there's no torque being transmitted the car is still covering ground and therefore still getting closer to the end of the lap...

[machin]

I've done a quick formula 1 simulation of Silverstone (using my own software):

if I set the 1st car so that the gear change time is 0 seconds I.e. no torque break during the upshifts and then increase the torque break until the lap time around Silverstone (old layout) is 0.3 seconds slower that torque gap needs to be 0.05 seconds per upshift... interestingly this confirms Patrick Head's comments about non-seamless gearbox shift times;

“There are several advantages to the semi-automatic gearbox. First... blah blah....Second, you can change gear in 30-50 milliseconds.

The other way of looking at it is that this simulation validates the Xtrac claim that the gearshift IS seamless and there is no torque modulation.... assuming you believe my simulation ...which to be fair, if I were you guys I'd be skeptical too... so that just leaves what Xtrac and Patrick Head are telling us...

[autogyro]

I've done a quick formula 1 simulation of Silverstone (using my own software):

if I set the 1st car so that the gear change time is 0 seconds I.e. no torque break during the upshifts and then increase the torque break until the lap time around Silverstone (old layout) is 0.3 seconds slower that torque gap needs to be 0.05 seconds per upshift... interestingly this confirms Patrick Head's comments about non-seamless gearbox shift times;

“There are several advantages to the semi-automatic gearbox. First... blah blah....Second, you can change gear in 30-50 milliseconds.

The other way of looking at it is that this simulation validates the Xtrac claim that the gearshift IS seamless and there is no torque modulation.... assuming you believe my simulation

Nice work machin, a good simulation but proves very little.
Convenient results though, almost as though all the figures quoted by others have been worked out in a similar way and not taken from direct measurement of the shift time at shift/overlap and the torque 'gap' created.

We still have no direct comparison with the current twin barrel shift system, which is the actual difference, not the 'seamless mechanisms' and the previous semi auto or earlier manual shifts.

Without supporting evidence, it is a bit much to accept a 0.3 sec a lap improvement just from the use of a new shift mechanism in a layshaft gearbox, where the ultimate shift speed is limited by the rotational inertia of the engine and input components in the gear train.

To prove this you need a direct car comparison with and without the mechanism and all else equal. Not a 0.35 sec manual shift speed as shown in your own graphs either machin but something realistic to compare.
I know a manual shift of 0.08 sec was possible manualy in 1991, a manual shift today should equal the current semi auto of 0.03 to 0.05 sec. There should be no difference using similar shift mechanism or odd even twin barrel dog systems manualy triggered.

Meanwhile you can ask Button if his neck has healed from the huge jerk his gearbox gave him at the first upshift off the grid at the Australian GP. That was NOT smooth or without upset to torque transfer, even if X-track say otherwise.

[Tommy Cookers]

Thanks for that

Many circuits were designed around cars of about 1000 bhp. With the current cars (lots of downforce but 730 bhp? ) a lot of the lap time is bound up with power limited corners, more than ever before.
Some think the rev limiting has made the engines more 'top-endy', or otherwise demanding more frequent gearchanges.

Assuming that any gearchange system is a tradeoff between speed and smoothness, surely the current situation (as above) encourages speed ?

Are drivers changing gear more (eg in corners?) now than in the days of 3000cc cars ?

[machin]

Are drivers changing gear more (eg in corners?) now than in the days of 3000cc cars ?

I doubt it'll be that different... the gearing is still required to cover the same road speed range (0-200mph), and they are still limited to 7 gears... so on average each 30mph speed gain will still require 1 upshift.

There will no doubt be some differences though... with less power it means that the length of 1st gear could be shorter (basically the low speed grip:power ratio has gone up: this means a shorter first gear can be used without fear of breaking traction). the downside of this is that it means the gears at the top would be more spaced out...

The changes to the aero rules in 2009 (less downforce), but the change to slick tyres (more grip) would also have a knock on effect to optimising gear spacing... but they're still limited by the required road speed range and the number of gears....

[machin]

We still have no direct comparison with the current twin barrel shift system, which is the actual difference, not the 'seamless mechanisms'

The twin-barrel system is a fundamental part of seamless shift; essentially they are one and the same thing, so talk of dual barrel gearbox is essentially talk of a "seamless" gearbox. The "Direct comparison" that you talk of will never happen... but what we do know is that the current F1 cars don't use pre-1990's manual gearchange systems... and that has nothing to do with the rules, and everything to do with lap times.

Without supporting evidence, it is a bit much to accept a 0.3 sec a lap improvement just from the use of a new shift mechanism in a layshaft gearbox

Without supporting evidance (but lots of anecdotal comments to the contrary from people who are currently in the F1 transmission industry) it is a bit much to accept that a manual pre-1990's gearchange was faster than today's seamless shifts, is it not?

There should be no difference using similar shift mechanism or odd even twin barrel dog systems manualy triggered.

As with many of these threads Auto you appear to be changing your argument now that you have realised the error of your ways.... You seem to be shying away from your original statement that manual gearchanges using pre-1990's gearboxes were faster than the present semi-auto seamless shift gearboxes despite that statement being in direct contradiction to your own definition of the limiting factor to gearchange times; rotational inertia of the parts involved. In addition you seem to now concede that the seamless system (with its dual selector barrel layout) is faster than the pre 1990's single selector ("non-seamless") arrangement.

If you do now agree that you have changed your opinion on this point, and this point alone, then maybe you could ask the Mods to change the topic title?

[autogyro]

The twin-barrel system is a fundamental part of seamless shift; essentially they are one and the same thing, so talk of dual barrel gearbox is essentially talk of a "seamless" gearbox. The "Direct comparison" that you talk of will never happen... but what we do know is that the current F1 cars don't use pre-1990's manual gearchange systems... and that has nothing to do with the rules, and everything to do with lap times.

Strange then that 'zeroshift' state their system using a shift mechanism on one hub is 'seamless'. They must be wrong using your definition machin.
Seamless is purely a marketing name not a technical description.
All layshaft stepped ratio gear shifts have a 'seam' at shift overlap.
There is no way to eliminate it.
To achieve a reading of constant torque at the output shaft you either have to use a 'sledgehammer' shift or engine and or clutch modulation.
The mechanisms either soften the 'bang'and or help the non manual modulation.
Neither is ideal for ultimate efficiency in torque transfer.
There is also absolutely no reason to prevent any type of shift mechanism in a layshaft gearbox from being manualy 'triggered' using a gear lever.
The problem with most current twin barrel systems is that they only allow for external modulation of the shift and do not allow for modulation at the shift overlap components as in a manualy shifted dog systems.
You either suffer a fierce engagement or have to reduce the torque going through the shift system.

I stand by my statement that it is possible for a pre 90s manual gearshift to be shifted faster than the current 'seamless' gear shifts.

[bigpat]

Having actually raced a proper RACING H pattern gearbox, having worked with sequential gearbox single seaters, and knowing a little of F1 technology, I can tell you a 80's or 90's box is NOT quicker than today.

I have used repaired and developed improved performance gearbox systems for over 40 years. Re read the post heading, I made no direct comparison to F1 technology.
The jury is out on when the fastest F1 shifts were in use. It is not the point.

Those boxes contained BIG gears, with a lot of rotational inertia, compared to the tiny gears of today. Also, flat changing is quick, but the dog rings still have to slide the 10-200 mm across to engage. With human effort, and play in linkages etc, 0.2 sec for a completed shift. A modern 'non-seamless' box can do it 0.05 sec. No arguement.

There were plenty of geartrains in use with similar gear weight to todays F1 gear sets. Your shift timing is 'wrong' I have timed manual gearshifts in dog ringed boxes at 0.08 ms. A grandmother in a shopping car could meet your figures of 0.2 sec.

Also, with the H pattern boxes, you always lose time going diagonally across the gate.

True but it is a loss of 0,02 ms at the worst for a skilled driver.
In anycase manual gearshift mechanisms are not limited to H pattern.

With a conventional box, you always lose drive through the shift, even flat changing. I change clutchless in a racecar, and only lift of say 15-20% throttle on an up change ( flat shifting is reckless, especially if you want reliability), but, there is a reduction of torque to the back wheels, compared to full throttle. A seamless shift sees no perceptible loss in torque measured at the back wheels.

There is NO potential difference in the mechanical shifting mechanism chosen at the hub for either manual or so called seamless shifting, it is the method of triggering the shift that varies and the design of the gear lever geometry.
In fact the 'seamless hub mechanisms are more complex and work in more than one direction and are potentially slower than a strait dog hub.

The seamless shift systems use engine cuts and clutch disengagement to 'smooth' the shift actuation. Where this modulation occurs other than at shift overlap it reduces the efficient transfer of torque through the system and increase the time taken to transfer torque. The shift speed is not decreased by modulation because the limit to shift speed is the rotational inertia of the input components, not the shifting mechanism.

Sequential and seamless shifts were NOT created to eliminate driver input of skill, as I believe gerachanging speed/skill is always over-rated by people( most of whom have never driven a race car). No, these systems were introduced (as with all things)because they are quicker in laptime, full stop.

Sequential shifting was a sensible improvement to manual gearshifting. I advised Garry Anderson on this very point for the first Jordans, where they used a 7 speed manual sequential geartrain. Seamless is simply a method to soften and control gearshifts sufficiently to allow a fast gearshift using automatic systems. In F1 cars are held to the ground with huge downforce and therefore do not suffer the unbalance the use of a close to 'sledgehammer' shift would otherwise apply to the car. Without huge downforce these gearboxes would make the cars undrivable.

It is a fact that an F1 seamless shift box reduced laptimes by up to 0.3 sec per lap over a conventional box. That is an average of 15-18 secs over a race, which is massive.

You have no data to confirm this and it is simply untrue. The shift overlap and engagement spike may be measures at between 0.05 and 0.08 ms but this does not take into account any powertrain modulation that effects the TOTAL time taken for the change between fixed ratios.

Personally I think today's drivers have it tougher. No they don't shift gears, ( but the cars are quicker which means more physical loading on the body)but they must have presence of mind to use all the buttons and knobs on the wheels, and activate all the systems on a modern GP car, at full speed, listening to someone in your ear. Also, tactics are much better understood today, with everything calculated down to the second, where they expect the edriver to punch out times like a machine.

Back in the day, it was more man vs man, and generally laptimes varied a little more. The sport wasn't as precise as it is now...

1.I understand you have measured manual dog changes at 0.08 /sec, but I can't see that being consistently done over the course of an entire race. Could you?

2. I seriously doubt whether the best transmission companies in the world would overlook a quicker gearchange system ( conventional gearbox) than what they currently use.

3. It IS true seamless shift is worth 0.3 sec. Toyota F1 team publically confirmed this during 2009. When they supplied their powertrain to Spyker (Force India), they too saw a laptime gain.....

Maybe we're all wrong, and we'll be back to H pattern boxes in the future.......

[machin]

and another one:

The advantage a seamless-shift gearbox confers is reckoned to be
in the order of less than half a second per lap - 0.3sec or
thereabouts.

full article: (LINK FIXED)

http://formula-one.speedtv.com/article/ ... gearboxes/