Shifting gears with broken clutch

[raymondu999]

Hey all. This is not a fanboy thread of Freddielonso. It's about him (practically) running a whole race distance with no clutch. I don't get how he could have downshifted for corners and upshifted on the straights without a clutch. Wouldn't that have at least fried his gearbox? I can't imagine putting down 780bhp or so of power into a gearbox, switching gears, while not having a clutch the whole time...

[marcush.]

believe it or not ,you don´t need a clutch for shifting up or down.
If you ever had a motorcycle you would be amazed just how easy it is actually to shift up-no problem at all and also down ...a bit more finesse is needed..
If it is feasible with todays cars ,this is more a matter of hows the backup mode if the clutch is not able to disengage..as this is the one possibility of clutch failure that still keeps the possibility of carrying on...it has to be a means of lost clutch actuatiuon..the plates ,pressure plates etc still have to remain in perfect condition so the word running without the clutch is wrong altogether .He lost clutch actuation ..as you do not use the clutch manually anyways ,apart from
getting away from the start and from the pits....the main point is what does the car do when you do not have a clutch actuation? Upshift? Normal ,I´d say.
Downshift ..it is still blipping the throttle for synchronising the gears.. so it is
maybe harsh in action ..and it will be very hard on the engine and gearbox but if the electronics allow for changing gears without clutch actuation anyone can drive..

what about the antistall? ..could you use it to get away from the pits ?

[Mr Alcatraz]

There was a rare Ferrari radio transmission that went something to the affect of:

Engineer: I know you don't like driving like this~/

I believe he was referring to the fact that Fred actually had to hit the accelerator to downshift into the corners in order to find the synchros' which level the rev's and vitually takes away the advantage of engine braking. Without doing that the gearbox wouldn't last 3 laps It's the reason he had such a hard time passing through the field.
without that problem he would have easily passed Button, (instead of having to run wide at every corner) and been able to bring the beast home for the points. The blame lays firmly on the team as they are responsible for getting a car out there that the pilot has the opportunity to trash himself. On the other hand, how do we know that Fred didn't thrash it taking off on the reconnaissance lap. Either way, Ferrari failed and lost an engine, and a gearbox to boot.

[autogyro]

Downshifts are controlled electronicaly so the driver cannot damage them.
Clutch actuation faults would mainly give problems on stopping or launch.
It depends on how much Ferrari depend on electronic clutch control during downshifts as to whether this effected the downshifts.
If not much, it was probably ECU. How much of that is Ron Dennis I do not know.

[Edis]

There was a rare Ferrari radio transmission that went something to the affect of:

Engineer: I know you don't like driving like this~/

I believe he was referring to the fact that Fred actually had to hit the accelerator to downshift into the corners in order to find the synchros' which level the rev's and vitually takes away the advantage of engine braking. Without doing that the gearbox wouldn't last 3 laps It's the reason he had such a hard time passing through the field.
without that problem he would have easily passed Button, (instead of having to run wide at every corner) and been able to bring the beast home for the points. The blame lays firmly on the team as they are responsible for getting a car out there that the pilot has the opportunity to trash himself. On the other hand, how do we know that Fred didn't thrash it taking off on the reconnaissance lap. Either way, Ferrari failed and lost an engine, and a gearbox to boot.

The clutch is only used to start and stop the car, and a F1 car doesn't have synchros like a manual gearbox.

[autogyro]

It seems there could be a number of ECU problems this season.
The unit is supplied by McLaren to FIA specs.
The McLaren also had phantom neutral problems when coasting.
I wonder if any of the X Trac based problems have been because of similar causes?

[Edis]

It seems there could be a number of ECU problems this season.
The unit is supplied by McLaren to FIA specs.
The McLaren also had phantom neutral problems when coasting.
I wonder if any of the X Trac based problems have been because of similar causes?

Usually this kind of problem isn't caused by the ECU but by sensors or the hydraulic system which operate the clutch, gearbox and other systems.

The ECU supplied by McLaren is quite 'conservative'.

[marcush.]

It seems there could be a number of ECU problems this season.
The unit is supplied by McLaren to FIA specs.
The McLaren also had phantom neutral problems when coasting.
I wonder if any of the X Trac based problems have been because of similar causes?

Usually this kind of problem isn't caused by the ECU but by sensors or the hydraulic system which operate the clutch, gearbox and other systems.

The ECU supplied by McLaren is quite 'conservative'.

if you define a sensor failure caused malfunction not ECU related ,then yes.Of course this is highly depending on the strategy the ECU is adopting if the sensor is
gone...you could apply a modelled input ,wich would possibly save your race ,or shut down the whole unit and create a dnf just to mention two possible strategies.

If the hydraulic system is at fault..effectively not disengaging the clutch...this is surely not a prob when the car is in motion ...if the ecu is able to react to it.But maybe todays systems are in need of a small dab of clutch to smooth out the driveline shock with all those downsized parts..
But it was the engine that gave up in the end?

[autogyro]

Smoothing out driveline shock is an important point marcush.
As the gearshift speeds increase with development and I know of one such box at least, which can change gear so fast that one gear is disengaged and the next engaged in the time between one gear tooth being under load and it being off load.
As you point out, the components are getting lighter and more likely to be damaged by shock loading. I have used torsional damping within the drive train, much in the same way the radialy fitted coil springs are used in a conventional clutch plate. Unfortunately this is not possible within an F1 clutch pack turning up to 18000 rpm and there is a big downside in over all weight increase and rotational forces if fitted anywhere in the rest of the train.
The gearbox designers try to achieve a fast enough shift to keep it between the gear tooth pressure points, so that the ECU does not have to be used to much to reduce torque input from the engine during the shift (a performance loss). It is all to easy to end up with a pile of bits however. The most likely hydrolic (or pneumatic) fault is in the system that feeds the scrolling meachanism responsible for selecting and moving the selector forks. This could be the McLaren problem with false neutrals that only occurs when coasting off load. The pressure off load might be insufficient to move a component or there could be an air lock in the system that compresses out under racing use or perhaps just a sticking part. It is all equaly possible to be an ECU fault however.

[xxChrisxx]

if you define a sensor failure caused malfunction not ECU related ,then yes.Of course this is highly depending on the strategy the ECU is adopting if the sensor is
gone...you could apply a modelled input ,wich would possibly save your race ,or shut down the whole unit and create a dnf just to mention two possible strategies.

A 'limp home mode'? Do the ECU's not do this already?

It's very rare that an ECU acutally goes heywire, as they are just look up tables that don't change. It's the inputs or outputs that generally don't do as they are told. As you can't write to the ECU on the fly (edit: in a F1 sense), I really don't see how an ECU can go wrong short of it physically breaking.

Also could someone clear this up for me? I realise they all come from McLaren tech, so at least the functionality is controlled. The ECU's are mappable by the teams aren't they?

[marcush.]

if you define a sensor failure caused malfunction not ECU related ,then yes.Of course this is highly depending on the strategy the ECU is adopting if the sensor is
gone...you could apply a modelled input ,wich would possibly save your race ,or shut down the whole unit and create a dnf just to mention two possible strategies.

A 'limp home mode'? Do the ECU's not do this already?

It's very rare that an ECU acutally goes heywire, as they are just look up tables that don't change. It's the inputs or outputs that generally don't do as they are told. As you can't write to the ECU on the fly, I really don't see how an ECU can go wrong.

Also could someone clear this up for me? I realise they all come from McLaren tech, so at least the functionality is controlled. The ECU's are mappable by the teams aren't they?

F1 is not a road passenger car... so you will run the unit at temps and levels of vibration individual electronic components were never designed for ,wich is also true for most sensors btw.
But my point is ,if road car software shows malfunctions in the field after literally years of development and approval before ever offering parts to a customer I see absolutely NO chance that a ECU by Mclaren will be free of bugs.
If those bugs would result in malfunctions or even be noticed by the teams is questionable ,but sure there will be some.

[xxChrisxx]

F1 is not a road passenger car... so you will run the unit at temps and levels of vibration individual electronic components were never designed for ,wich is also true for most sensors btw.
But my point is ,if road car software shows malfunctions in the field after literally years of development and approval before ever offering parts to a customer I see absolutely NO chance that a ECU by Mclaren will be free of bugs.
If those bugs would result in malfunctions or even be noticed by the teams is questionable ,but sure there will be some.

ECU's are just maps with basic firmware.

This was why I asked if they could be mapped by individual teams (I assume they must be as they don't run identical engines/components). If some teams specify differnt inputs or outputs, it could be their map thats the problem and not the ECU.

If it's an inherent software/hardware/firmware problem, surely you'd be seeing more problems across all teams. (maybe not, I don't profess to to have a full and all encompassing knowledge of ECU's)

[Edis]

It seems there could be a number of ECU problems this season.
The unit is supplied by McLaren to FIA specs.
The McLaren also had phantom neutral problems when coasting.
I wonder if any of the X Trac based problems have been because of similar causes?

Usually this kind of problem isn't caused by the ECU but by sensors or the hydraulic system which operate the clutch, gearbox and other systems.

The ECU supplied by McLaren is quite 'conservative'.

if you define a sensor failure caused malfunction not ECU related ,then yes.Of course this is highly depending on the strategy the ECU is adopting if the sensor is
gone...you could apply a modelled input ,wich would possibly save your race ,or shut down the whole unit and create a dnf just to mention two possible strategies.

If the hydraulic system is at fault..effectively not disengaging the clutch...this is surely not a prob when the car is in motion ...if the ecu is able to react to it.But maybe todays systems are in need of a small dab of clutch to smooth out the driveline shock with all those downsized parts..
But it was the engine that gave up in the end?

An ECU is an ECU, and a sensor is a sensor. As for sensor failures, they often can be handled without forcing the car to a stop. But that is of couse dependant on what kind of back up there is for the sensor, and how important the sensor in question is.

If you have trouble with the hydraulic system it often means you have more issues than just the clutch, if it's not the servovalve that controls the clutch that is the problem. This since the hydraulic system also powers the throttle, gearchange and servosteering.

Vibration levels and temperatures are probably lower in F1 than for production cars. F1 cars tend to have the ECU mounted in the chassi, for instance in a sidepod. That way they have direct access to cooling air. Production units are usually mounted directly on the engine, uncooled. But production cars use 'hybrid chips' developed for this purpose.

F1 is not a road passenger car... so you will run the unit at temps and levels of vibration individual electronic components were never designed for ,wich is also true for most sensors btw.
But my point is ,if road car software shows malfunctions in the field after literally years of development and approval before ever offering parts to a customer I see absolutely NO chance that a ECU by Mclaren will be free of bugs.
If those bugs would result in malfunctions or even be noticed by the teams is questionable ,but sure there will be some.

ECU's are just maps with basic firmware.

This was why I asked if they could be mapped by individual teams (I assume they must be as they don't run identical engines/components). If some teams specify differnt inputs or outputs, it could be their map thats the problem and not the ECU.

If it's an inherent software/hardware/firmware problem, surely you'd be seeing more problems across all teams. (maybe not, I don't profess to to have a full and all encompassing knowledge of ECU's)

Application code is autogenerated using Simulink which should reduce the risks for bugs. But I assume that the teams are not allowed to create their own application code, only adjust the values in the maps. Otherwise it should be faily simple to create your own traction control for instance.

[madtown77]

It's very rare that an ECU acutally goes heywire, as they are just look up tables that don't change. It's the inputs or outputs that generally don't do as they are told. As you can't write to the ECU on the fly (edit: in a F1 sense), I really don't see how an ECU can go wrong short of it physically breaking.
...

You'd be surprised, I've seen our team blow an onboard spark driver with no real explanation and no fault was given by the ECU. Stuff can and does go wrong. Sometimes it is controls, but likely the teams can only calibrate the ECU and make 'user interface' changes for their specific electrical components on the steering wheel.

The circuits on the ECU can break. I would say that the ECU in an F1 car is probably very carefully isolated from such vibrations and shocks and that Mclaren have probably designed it to take a lot of abuse

A small short in a sensor wire (or any wire for that matter) can cause the ECU to trip into fault modes (low voltage is a B*#%$!). This is especially true when you are trying to save weight by reducing redundancies (i.e. no backup wiring, which your road car usually does on critical components)

[riff_raff]

madtown77,

I engineered with a race team about 20 years ago, when ECU's and sensors were a lot less sophisticated and reliable than they are now. To be honest, I don't recall any DNF's that were caused by the ECU. The primary cause of DNF's were usually due to improper maintenance work at the shop or at the track by the mechanics. It was usually things like improperly installed fasteners, wiring or fluid fittings coming loose during a race. The mechanics were always in a hurry assembling things, and so these type of problems were common.

I would imagine that with the amount of resources an F1 team has, and considering the amount of money at stake, a modern F1 ECU and it's software code, wiring, and sensors have at least one level of redundancy, if not more. F1 probably uses the same control system architectures as modern aircraft, which have at least three levels of fault tolerance on all critical systems. It's not difficult to do and doesn't add much weight.

It reminds me of the old rule for drivers commenting on DNF's. Never blame the problem on a part of the car paid for by a sponsor!

Regards,
riff_raff