Fastest gearchanges were pre-1990's

[hardingfv32]

So then this is really all about the acceleration rate of the shifting collar or sleeve, if we are talking about a cog system. The more force applied the fast the movement.

1) At a minimum, any pneumatic or hydraulic shifting system can be made to apply as much force as a human. I would venture to claim even more force. So the current systems should be at least as fast as a human if not faster.

2) Strength and weight has to be built in to the shifting mechanism to coupe with all this force required to win this competition between man and machine. It will be my claim that the machine shift system can be weaker and lighter than, thus capable of fast acceleration of shifting components. A manual system must cope with tremendous deceleration forces as the human over-strokes the shift mechanism into its motion stops.

I claim the machine shift can be faster.

Brian

[flynfrog]

so a driver blipping the throttle doesn't count to the total shift time but it does when computer does it? Even on a bike with sequential shiftier with a ignition cut box is pretty quick much quicker than I can lift the throttle and blip the gear in. Sure I could just jam the lever but its not going to benefit anything except the factory worker that builds dog rings.

Anything you can program a driver to do a computer could be programmed to do it faster.

I will repeat the call where is your data? Others have presented data to the contrary of what you state. You claim their data is no good then bring none of your own.

[autogyro]

Again, it is not a question of how fast the shift 'mechanism' is operated.
It is the speed of the engageing/disengageing components which are limited by the rotational moment of inertia present in the rotating powertrain components as a whole.

The difference with a manual shift, is that the skill of the driver modulates the shift to suit circumstances. (A lost driver skill).
There is no need to 'blip' the throttle during upshift racing changes.
A good driver will match the minimum shift speed without excess modulation when the fastest shift is needed.
He will also slow the shift when there is no need for speed, perhaps mid corner.
Knowing where and when to choose was part of a race drivers skills in the quest for faster lap times.

The speed and power of automatic shift operating systems is irelevent, it is how well they are progammed to achieve a range of shift requirements that count.
No increase in shift power or speed will exceed the limitations in time of any particular powertrain design and its speed capability at shift overlap.
Shift modulation solely to 'smooth' a gear shift will be less efficient and usualy slower than a skilled manual shift. Road car systems are still very crude compared to a skilled driver with a manual shift. Jolts and engine rpm flares are still common.
Modern racing shift systems are I agree, very close to the ultimate shift capability designed to achieve maximum efficiency and vehicle control. The actual shift speed at overlap however will be no faster potentialy than for a manualy shifted version. The main limitation is the layshaft geartrain concept itself.
In F1 downforce masks most of the handling upset caused by violent and crude gear shifting.

[autogyro]

Anything you can program a driver to do a computer could be programmed to do it faster.

I will repeat the call where is your data? Others have presented data to the contrary of what you state. You claim their data is no good then bring none of your own.

I am well aware of the capabilities of computer systems used in F1 cars.
We were talking about a car system without a driver way back in the early 90's.
Many of the FIA regulations since have been drafted to hold off this possibility.

I have not claimed anyones data is no good. As far as I can remember all the data posted on the subject so far has been fair and accurate and in no way contrary to what I state.
I am simply saying it only gives part of the picture.

[machin]

Slightly edited re-post to make my explanations clearer, original post deleted:

I can only see four gears in this shift cascade.
The 1st to 2nd upshift looks identical for both gearboxes.
The manual shifted box needs lower third and fourth gear ratios.
The rev drop is to much at these shifts.

Auto are you even looking at the same data?! I'll explain it for you:

Both gearboxes are using clutch slip for the first 3 seconds (road speed increases but revs stay fairly constant).

The H-box shifts twice during the acceleration; at 5 and 7 seconds. They are clearly noticeable from the flat points in the acceleration curve (acceleration stops for about 0.25seconds) or from the rev data... the driver changes up when the revs hit about 9000rpm. revs drop by about 2000rpm on each shift.

The sequential shifts 3 times during the acceleration: at 4.3, 5.7 and 7.3 seconds. There is almost no perceptable stop in acceleration, but the rev data and gear indicator show where the gear changes happen. Again the driver upshifts at about 9000rpm each time. At the first change revs drop by about 2500rpm and on the other two by about 1700rpm.

it is actually the sequential which I would suggest would benefit from a different ratio, not the H-box: these cars with 300bhp are traction limited even on slicks up to 50mph anyway so the short 1st doesn't give you any advantage (we've discussed this on another thread)... lengthening 1st would remove the 2500rpm drop from 1st to 2nd, giving something like the 2000rpm drop at each change like the H-box has.

You can see from the acceleration curves that the H-box gear ratios are actually slightly better than the Sequential ( notice the gradient on the speed:time curve is actually very slightly higher on the H-box than on the sequential except during the gear changes, and this backs up my suggestion about changing the gear ratios slightly on the sequential to keep the revs up).

Despite the extra gear change and slightly poorer in-gear acceleration by 7.5 seconds the sequential flat shift equipped car is travelling approx 7 or 8mph faster. if the in-gear acceleration is worse then the additional 7 or 8mph must be due to the quicker gear shifts (I.e. there is less time when the car isn't accelerating).

[Richard]

Anything you can program a driver to do a computer could be programmed to do it faster.

The point is that the shifting is limited by the drive train inertia and a racing shift doesn't require any throttle blip. So the human crashing the gear lever can engage cogs as quickly as a pneumatic gear box.

The only difference between the two lies in the human skill to time the change to match the road speed and engine revs.

Others (machin and the post I linked) have shown evidence to agree with this.

I think autogyro's argument is also dependent on the current F1 rule set. The size of the clutch pack and gears are limited to what is a sub-optimal arrangement for the drive train. That arrangement is tolerable due to other factors in the car design. However, if the rules were changed to make a smooth drive train more critical to the car performance (for example less downforce and manual gear shifts) then natural evolution would lead to a different drive train that was smoother and quicker to shift.

Less downforce means more wheelspin with clumsy gear changes, and more spins with poor gear selection. Something like we had 3 decades ago. Probably when Mansell was driving a Ferrari. Probably when a poster on this forum had indepth knowledge of the drive train performance and knows that the current configuration is suboptimal for the drive train, but optimal for the overall car-driver package.

We may not like the rhetorical style, but with he's probably got a point that could be valid.

[machin]

In F1 downforce masks most of the handling upset caused by violent and crude gear shifting.

The example sequential flat shift vs manual H-pattern box I have presented above is fitted to a Westfield (Lotus 7 style kitcar, - see my avatar). It has no downforce , furthermore it was developed by what is basically an amateur outfit, and yet the results are reportedly excellent; perfect changes every time... No its not a road car so I'm sure its not silky smooth, but its clearly not too harsh for this car in a race situation despite the lack of downforce.

[machin]

So the human crashing the gear lever can engage cogs as quickly as a pneumatic gear box.

The thing is Auto originally stated that the manual change from the 1980's was QUICKER than the current pneumatically operated gear change.... and yet can present no evidance of this.

[Richard]

as soon as you add engine or clutch modulation (seamless) the effective shift speed increases because of the inefficiency introduced to the torque flow.

If a seamless shift system fitted to this gearbox matched the 80 ms it would still be in effect much slower in transfering torque from input to output because of the pre and post overlap modulation.

This bit intrigues me. So you are saying that torque is sacrificed to ensure a seamless shift, but that would not occur with a manual box? That sacrificial torque results in a slower overall shift??

[autogyro]

So the human crashing the gear lever can engage cogs as quickly as a pneumatic gear box.

The thing is Auto originally stated that the manual change from the 1980's was QUICKER than the current pneumatically operated gear change.... and yet can present no evidance of this.

The evidence is in the physics.
The speed of gear shift cannot exceed the speed limitations placed upon the geartrain by the rotating components and their ability to change engine rpm and match the new torque path when fully engaged.
In this case around 0.08 ms.
If to achieve this speed of 'shift overlap', or to 'smooth' the shift, the engine is cut or any other modulation of torque flow is undertaken, if this modulation occurs prior or post shift overlap, then it will reduce the efficiency of torque transfer from engine to drive wheels. The result will be less acceleration than without the modulation.(the gear shift will be slower over all)

A sequential operating system makes changing gear easier but not actualy faster, in most cases the actual components in the shift mechanism at the dog slider will be the same as in the manual system anyway.(potentialy the same speed of operation)
Complex 'seamless' geometry designed into multi path syncro hubs will further slow things. (They do allow easier torque modulation however).
In the graph shown, the sequential gearbox is useing an extra gear in the acceleration curve shown. This on its own will level out the road speed curve and achieve a higher road speed no matter what the shift mechanism.
There is no control in this case on the gear changing skill of the driver operating the H pattern shift system.

[Billzilla]

So the human crashing the gear lever can engage cogs as quickly as a pneumatic gear box.

The thing is Auto originally stated that the manual change from the 1980's was QUICKER than the current pneumatically operated gear change.... and yet can present no evidance of this.

Because it is quite simply incorrect.
The current F1 gearboxes shift gears in single-digit thousandths of a second (maybe faster) and that is tens or hundreds of thousandths of a second faster than a human can do it.
End of story.

[hardingfv32]

Again, it is not a question of how fast the shift 'mechanism' is operated.
It is the speed of the engageing/disengageing components which are limited by the rotational moment of inertia present in the rotating powertrain components as a whole.

Please expand BUT remember most of us are not transmission engineers.

1) What components and at what time during the shift does 'rotational moment of inertia' come into play? Are we concerned with rapidly changing the speed of some of the engagement components? Can these components be strengthened to absorb more force?

2) The limitations of the 'rotational moment of inertia' would seem to be the same for human and machine shifting. Is this correct? If not, why?

Brian

[flynfrog]

he also forgets that today's boxes are much lighter than pre 90s boxes since they don't have a human crashing the gears

[autogyro]

So the human crashing the gear lever can engage cogs as quickly as a pneumatic gear box.

The thing is Auto originally stated that the manual change from the 1980's was QUICKER than the current pneumatically operated gear change.... and yet can present no evidance of this.

Because it is quite simply incorrect.
The current F1 gearboxes shift gears in single-digit thousandths of a second (maybe faster) and that is tens or hundreds of thousandths of a second faster than a human can do it.
End of story.

Think again.
The time the 'actual' change takes is dictated by the syncro hub mechanism and the rotational inertia of the rotating components in the gearbox.
What is used to trigger a shift is irelevent.

[autogyro]

Again, it is not a question of how fast the shift 'mechanism' is operated.
It is the speed of the engageing/disengageing components which are limited by the rotational moment of inertia present in the rotating powertrain components as a whole.

Please expand BUT remember most of us are not transmission engineers.

1) What components and at what time during the shift does 'rotational moment of inertia' come into play? Are we concerned with rapidly changing the speed of some of the engagement components? Can these components be strengthened to absorb more force?

2) The limitations of the 'rotational moment of inertia' would seem to be the same for human and machine shifting. Is this correct? If not, why?

Brian

1) The main rotating component that has to rapidly change its rpm is the crankshaft of the engine. It has to do this to match the next gear ratio selected. The output part of the geartrain stays at road speed. The crankshaft and all connected components that are rotating with it have to reduce their rpm, usualy by at least 1000rpm in a very short time period. In an F1 powertrain, the crankshaft and these other 'input' components are much lighter than in say a road car. This means the minimum time needed to slow these components in an F1 car is less, so the 'potential' shift speed at the 'shift overlap' will be faster.

2) This absolute limit to shift speed caused by the rotational moment of inertia is indeed the same no matter what method of shift mechanism and triggering is chosen.

The problem comes when we try to compare a manualy triggered shift which is solely reliant on driver skill, with a 'modulated' semi or fully automaticaly triggered shift, where the modulation is applied before, during and after the actual shift 'overlap'.
In the later case, the modulation that occurs 'not' at overlap must reduce the torque being transfered from the engine to the road wheels and has to be added to the effective 'over all' shift time.
Of course few drivers today have the skills to effect a shift speed as fast as the modern systems but it was and is possible.