dual-clutch transmissions 3

[pontiacjack]

I've only recently been reading about DCTs, and wonder about a couple of things.

1) In a high-performance car (say one that uses the Getrag that will live with more than 500 ft.lb. of engine torque), what is a typical limit value of clutch engagement speed from vehicle standstill? Does sudden full-throttle (guess I should really say full-pedal-depression) give approximately the same result as "dumping the clutch" of a manual trans car?

2) If the driver makes a downshift when the controller is anticipating an upshift, how slowly will the shift occur? And- following a downshift, does the controller typically anticipate another downshift?

 

[BrianPetersen]

My dad has one, although it's the VW/BorgWarner unit and it isn't in the torque range you mention, but there's no reason why it would be different.

1. If you mat the accelerator from a standstill, it lets the engine rev up to the bottom of the torque-peak range and feathers the clutch out, holding RPM at that level until the clutch is fully engaged. It's very much like a smooth takeoff by a good driver with a manual gearbox. It will not let the engine go to very high revs and abruptly "dump" the clutch. It does a smooth (but fast) takeoff. But, obviously the details of this would have to do with how each individual application is programmed. You COULD program it to let the revs go way up and then abruptly let the clutch out, but as an OEM, why would you want to? More delay before starting (waiting for the revs to get to a higher level), more shock load on everything (more warranty claims), more tire wear, etc.

2. Too quickly to be noticeable. Even if the mechanism has to change cogs inside the gearbox, the time for completing the entire process is still in milliseconds - just more milliseconds than if it just has to swap clutches. It's not noticeable during anything resembling normal driving. I have no idea whether it anticipates further downshifts and it doesn't really matter.

Delays programmed into the controller intentionally to smooth out minor abrupt actions by the driver are much more noticeable than the actual time it takes to change gears. Most people wouldn't want an instant downshift/upshift if they accidentally have a little twitch with their foot, or if their foot moves a fraction because the car went over a bump. The consequence of that is that if you are driving along in "drive" at a constant speed, and mash the pedal to the floor, it still takes a moment to decide to do the downshift (just like every other traditional auto-tragic does), but once it does, it's very quick.

These are my observations from driving my dad's car occasionally; I still prefer to row my own gears and operate my own clutch at my own discretion.

 

[tbuelna]

pontiacjack,

One of the most difficult things for a transmission controller to do is modulate a friction clutch, and the most difficult clutch modulation is the one at a standing start. Wet clutches are easier to modulate than dry clutches, since they tend to have more consistent (albeit lower) friction characteristics. I believe the Getrag unit you mention has dry clutches. The attraction of the dry clutch automated DCT is that it is less expensive and more efficient than a conventional AT with a torque converter.

The modern driving public generally demands a very smooth launch/shift event from an automatic transmission. Until the advent of modern, digital electronic engine and transmission controls/software, getting a commercially acceptable shift quality from an automotive DCT or AMT was not practical. But with integrated engine and trans controls, the manipulation of the gear shift, engine speed and load, and clutch engagement can all be coordinated at a very high frequency. Since the speed and torque of the engine can be closely controlled by the ECU, a full throttle launch would not likely overstress the modern DCT, like "dumping the clutch" would with a conventional manual trans.

As BrianPetersen noted, the gear change in a DCT is very rapid, since the gear change (and torque interruption) does not occur until the next gear has been pre-selected.

Hope that answers your question.
Terry

 

[dgallup]

Most of the high horsepower DCT cars these days have launch control as well as all the other nanny electronics (traction control, stability control, ABS, don't even get me started on lane departure, backup, etc.). So you just sit there with one foot on the brake and one on the throttle, let off the brake and accelerate at the limit of traction. Really takes all the fun out of it.

 

[pontiacjack]

Thanks for the responses.

 

[FeX32]

Check out VW launch control DGS tranny. It certainly it not feathering so to speak. Comes standard on GTI.


Fe

 

[Warpspeed]

If you look at the ratios of these DCT transmissions, you will notice that first gear is always a very low ratio.
That considerably eases the thermal stress and wear on the clutch when starting off.
First gear is really just to get the vehicle moving.
Once on the move, the shift to successively taller gears becomes far less of a problem.

 

[evelrod]

Interesting concept. I'm a bit skeptical about the 'abuse' problem in the Ford setup. By not having any temp sensors in the clutch, they have instead programmed the ECU to detect excessive slippage (that could produce excessive heat?---no sensor?) to warn the driver of a potential problem.

As explained to me, if the ECU detects the trans being used to "hold a hill", the clutch will release briefly (to allow the car to briefly roll?) to alert the driver (to use the brake?)!

Hmmmm. Further study is in the offing, I suspect. I have exactly zero faith in any power transmission method that is not completely 'bullet proof'.

Rod---feeling a bit old fashioned and out of date!

 

[theoldwizard1]

Believe it or not, I actually worked on the first prototype of Ford's DCT back in the mid 1980s ! Obviously it took awhile to get all of the bugs out and make it cost effective !!

A lot of the questions you are asking have a lot to do with the "application". Are we talking a production vehicle where smoothness of launching and shifting as well as durability are critical or are we talking true race car (F1) ?

My responses assume we are talking production.

IMHO, you will never see DCTs in high torque applications for several reasons.
[ul]
[li]You just can not beat that 1.5-2.0 times torque multiplication factor you get from a torque converter for launching a vehicle. Without that, you need more gear steps which adds weight and cost.[/li]
[li]Clutch durability is ALWAYS an issue. Customers expectation are more than 100K+ miles with no repairs.[/li]
[li]Launching at WOT doesn't really happen anymore, because most (all?) vehicle now use electronic throttle control. Sure engineers want to go fast, but the throttle and clutch will be modulated for smoothness.[/li]
[li]Anticipating incorrect upshift/downshift (also called "change of mind") is a big deal. Lots of modeling, testing and software algorithms. Short answer, if the transmission "pre-positions" for an upshift, but a downshift is required, it will take longer as the gear forks need to move the gears (which are now incorrect) before engaging the clutch.[/li]
[/ul]

Ford/Getrag actually have 2 very different transmissions. The one used in EU uses wet clutches (and can handle more torque) while the one used in the US (built in Mexico) uses dry clutches. The EU one uses hydraulics to "stir the gears" (heavier) and the US one is all electric motor actuation (light, but a bit slower).

 

[theoldwizard1]

...if the ECU detects the trans being used to "hold a hill", the clutch will release briefly (to allow the car to briefly roll?) to alert the driver (to use the brake?)!

Early prototypes could produce clouds of clutch smoke when doing a hold on a steep hill in about 30 seconds !

 

[dgallup]

This high enough torque for you?

"the wet multi-plate clutch DCT in the Bugatti Veyron is designed to cope with 1,250 N·m (922 ft·lbf)"

 

[theoldwizard1]

"the wet multi-plate clutch DCT in the Bugatti Veyron is designed to cope with 1,250 N·m (922 ft·lbf)"

When near infinite money is available, different solutions are available.

Ford, and the rest of the high volume manufacturers, simply could not build an affordable car with wet multi-plate clutches. I'll bet that vehicle is RWD, not FWD, like most of the "real world" DCTs. I'll also bet it as more than 6 speeds. With a RWD vehicle it is a lot easier to "package" a large transmission.

The point I was trying to make is, I don't think you will see DCT in any light duty/medium duty trucks anytime in the near future.

 

[dgallup]

It's AWD. Packaging on that vehicle is a nightmare.

I agree you won't see any DCT's on light duty trucks. What would be the point? DCT's only appeal to sports/sporty car drivers.

 

[theoldwizard1]

I agree you won't see any DCT's on light duty trucks. What would be the point? DCT's only appeal to sports/sporty car drivers.

Ford is using it in the 2011MY US Fiesta and the 2012MY US Focus strictly because of its superior fuel economy compared to a conventional automatic transmission.

Remember, "paddle shifter" are not required for this transmission, but sure would be fun !

 

[BrianPetersen]

"Dual-clutch transmission" is a separate thing from having a "manual shift mode".

You can have a DCT automatic without paddle shifters (VW TDI, Ford Fiesta) or with paddle shifters (VW GTI) and you can have a torque converter automatic without paddle shifters (examples too numerous to mention) or with a manual shift mode, whether by paddle shifters or otherwise (Mazda 3). It isn't a "sportiness" thing, it is a "fuel consumption" thing.

A DCT transmission is not appreciably different in size from a manual transmission with the same number of speeds.

For that matter, whether or not there is a torque converter doesn't really have to do with what type of gear-set is used. If you have a planetary gearset then there's no point using the dual-clutch concept ahead of it. But a planetary-gearset doesn't *have to* use a torque converter (Hyundai Sonata Hybrid puts the electric motor and a mechanical clutch in place of the torque converter) and a torque-converter transmission doesn't *have to* use planetary gears (every Honda automatic).

Improvements in regular auto-trannies have made it less likely that DCT transmissions will take over the world, though. Lock up the torque converter in a normal automatic as soon as the vehicle moves away from a stop, and do the gear-changes quickly but firmly, and design the innards of the thing properly to minimize friction, and you have the same FE as DCT and the same "road feel" as DCT but with the same pull-away from a standstill as a normal automatic. Or do what Hyundai did, and put the hybrid-system motor-generator and a regular clutch where the torque converter was.

 

[simple77]

Hi Guys,
I actually drive Dual Clutch VW R32 with a turbocharged VR6, 3.2L motor. Stock car is 250HP, I have it boosted 14PSI to 400hp. It been tested and proved to be good to 550hp, at which point OEM clutch will slip.

Gearbox can take a lot more TQ than simmular manual type simply because it shifts very smooth. It prohibits damaging actions that can be done by novice drivers that strip gears in manual versions. The gearshifts it self are seemless like in AUTO transmission, I loose no BOOST during upshift. Upshift at 7000RPM takes... hm.. very litle time. I think oficially it's 0.2sec and in tuned form 50% less.

You can start of the line very agresevly by using 'launch controll' - it is an programmed option of the "clutch dump" at 3500RPM. Turning ESP off, allows the motor to apply all power disregarding all four wheel spin

Stock dual clutch have lasted me 80,000Km in stock form and another 15,000km in boosted form. Still going strong.

Hardware wise this gearbox is major step forward, it can take major power upgrades. The key to this technology is the programming which is very tricky as already mentioned above in this thread.

Manufacturing parts in China

http://www.pa-international.com.au

 

[FeX32]

Wicked ride simple77 !!
I drive a mk6 GTI. Not DSG though, 6sp manual.




Fe