M-DCT Semantics and Subtleties

[CanAutM3]

Originally Posted by mkoesel

Aha, nice work. So indeed it does behave the same way whether the car is moving or not - it just happens to not be the way it appeared.

I guess technically we were both wrong, but I'll take the blame on this one. My apologies. Thanks for tracking down the answer.

It's funny because after I wrote the part about how switching to neutral would technically only need to change the display to a "N", I started thinking to myself - yeah, actually, why doesn't it just work like that? It makes sense from the standpoint that, like I said, at zero pressure the clutches are open so when the system is off, there is nothing trying to force the car into gear (unlike a 6MT where the clutch plate is forced closed by springs).

So, it would seem that if they didn't want to give the user the ability to disconnect the engine from the wheels while in motion, in fact, the UI would not even require an "N" position at all. However, I think it is good to have it for safety since otherwise bumping the gas pedal at rest unintentionally would set the car in motion. Would be very bad in an auto wash, for example.



One note on this - and I'm going to qualify this one with "as I understand it" so I don't have to wipe egg off of my face again - to my knowledge there is only one clutch plate, so technically if one clutch is closed the other one would by necessity already be open. I believe the clutch pack design is such that this single clutch plate has a friction surface on each side and it simply moves from one side to the pack to the other. Think of it like if you took the upper bun of a hamburger off and held it an inch above the beef patty - so that represents the lower bun being engaged. Then, if you switch your grip to lift both the upper bun and the beef patty up, now the upper is engaged. The patty can only engage one or the other. Hey and maybe the mayo can be the DCT fluid? No bacon though because that would really screw up your transmission.

Apologies for making everyone hungry with my example, by the way.

IMO, we have to be careful how we interpret the schematics and accompanying caption provided by Carry.

The caption only says that when in neutral, both clutches are open. It does not say than neutral is obtained by opening both clutches.

Further, looking closely at the schematics for Neutral, all dogs (blue bars) are shown in their middle position, meaning the transmission is really in Neutral (no gears engaged).

In contrast, in the 1st gear pictorial, we can see the 1st/7th gear dog moved to engage 1st gear.

As for your second section, it is contrary to my understanding of the DCT clutches. My understanding is that the two clutches are complete units allowing them to "slip against each other and hence the seamless shift.

If it is as you say, there would be a slight pause in the transmission of power.

[CaryTheLabelGuy]

Originally Posted by CanAutM3

If none of the dogs are engaged, how do you call that ?

I don't know if the input shaft(s) were designed to be driven by the engine with no dogs engaged. Seems to me the clutches need to be open in "neutral" as the design doesn't seem to allow the DCT to free-spin with no dogs engaged. Also, if the design did allow for the input shaft(s) to free-spin, which clutch would be the clutch to close? Both? How would that work?

I'm obviously not entirely sure about any of this. But, now we do know both clutches are in the "open" position in neutral, which was the point I was trying to make, regardless of what is actually going on inside the DCT in regards to the gears.

[CaryTheLabelGuy]

Originally Posted by CanAutM3

IMO, we have to be careful how we interpret the schematics and accompanying caption provided by Carry.

The caption only says that when in neutral, both clutches are open. It does not say than neutral is obtained by opening both clutches.

Further, looking closely at the schematics for Neutral, all dogs (blue bars) are shown in their middle position, meaning the transmission is really in Neutral (no gears engaged).

In contrast, in the 1st gear pictorial, we can see the 1st/7th gear dog moved to engage 1st gear.

I saw that and understand that, but my point was that the clutches are "open", at least in this configuration of "neutral". Beyond that, I've no idea. Lol.

I was never a transmission guy. My understanding of DCT's is very limited.

[mkoesel]

Originally Posted by CanAutM3

IMO, we have to be careful how we interpret the schematics and accompanying caption provided by Carry.

The caption only says that when in neutral, both clutches are open. It does not say than neutral is obtained by opening both clutches.

Further, looking closely at the schematics for Neutral, all dogs (blue bars) are shown in their middle position, meaning the transmission is really in Neutral (no gears engaged).

In contrast, in the 1st gear pictorial, we can see the 1st/7th gear dog moved to engage 1st gear.

Great point. I didn't study the picture closely, and just looked at the clutches, not the dog gears.

But now that I think about it, I had even said in my earlier post that, the DCT could very well additionally keep the clutches open since it has no reason to close them anyway. So I really should have looked closer.

As for your second section, it is contrary to my understanding of the DCT clutches. My understanding is that the two clutches are complete units allowing them to "slip against each other and hence the seamless shift.

If it is as you say, there would be a slight pause in the transmission of power.

Hadn't thought about the power transfer, but that makes sense. However, even with two independent clutches, the power will drop briefly as they are allowed to slip. The thing about having just one plate is you can move it back and forth as fast possible with no worry about designing in slippage to ensure two gears are not engaged at once. And, if the distance the clutch must travel from open to closed is tiny (a couple mm, say), it would essentially be an instantaneous shift. Just thinking out loud.

[mkoesel]

Originally Posted by CaryTheLabelGuy

I'm obviously not entirely sure about any of this. But, now we do know both clutches are in the "open" position in neutral, which was the point I was trying to make, regardless of what is actually going on inside the DCT in regards to the gears.

Fair enough, but I never contradicted that point to begin with. From my first post (the one you "took issue" with):

Actually, if you select N then the transmission really is in neutral. This means no dog gears inside the transmission are engaged just as if you shift to neutral with an MT. At this point the DCT TCU could close one of the plates on the clutch pack, just as you can allow the clutch to close when in neutral with an MT. But it likely doesn't since it doesn't have a calf muscle it needs to rest, plus it takes less energy since the clutch pack goes open at zero pressure.

[Tomm3]

I think the DCT is in "real neutral" while put on N.
I use that feature every day as I have quite bumpy road (about 200meters) to reach my home. There is slight descend at one place for about 80m so I put it on N and drive about 5km/h braking a bit not to accelerate. It's easier and smoother than in D as the drive is disengaged and does not accelerate nor slow me down.
The point is - when changing from D to N or vice versa I can hear double click under the car. So I strongly belief that both halves of transmission move their dog gears to neutral or then back into gears (1 and 2 at the same time).
I also think that operation of clutch is completely silent as they are in the oil and movement of the plates is very small.
What you hear are the dog gears actuators which require quite a lot of power to move.

[KevinGS]

Originally Posted by aboulfad

Some great reads on the DCT:
M5 F10 car blog
Chapter 4, Powertrain!
M5 technical Info: http://f80.bimmerpost.com/forums/att...3&d=1400504642
M4 technical info: http://f80.bimmerpost.com/forums/att...5&d=1406475520

The M4 has a parking lock, but no description of its details!

So much good info here, Ab, many thanks!

I swear this forum makes all the difference in whether I "like" my M4, or LOVE my M4. The Mercedes forums are cool, but this Bimmerpost forum is priceless. You guys rock!!!! SOOOO much info, I love it.

[CanAutM3]

Originally Posted by mkoesel

Hadn't thought about the power transfer, but that makes sense. However, even with two independent clutches, the power will drop briefly as they are allowed to slip.

IMO, it is possible to have a seamless shift, without any interruption of power transmission, with two complete clutch units.

While accelerating in the lower gear, the clutch of the lower gear is closed while the next higher gear is engaged but with its clutch open. For the higher gear clutch that is open, one half is turning at the same speed as the closed clutch since it is driven by the engine while the other half of the open clutch is spinning slower because it is driven by the rear wheels through a higher gear. During the shift, the open clutch starts to close and at some point it is starting to transfer power through dynamic friction. At that point, the lower gear clutch starts to open and also starts to slip as the engine slows down while the other clutch picks-up more and more power.

This is what I mean by the two clutches "slipping against each other".

[Tomm3]

Anybody has any interesting materials or links about how actually our transmission works in details?
That is - how exactly the clutches are actuated? By oil but from where, which point etc.?
How the sleeves moving dog gears are operated? Hydraulically or maybe electrically?
The fact is - when there is no pressure - no clutch is engaged because springs try to keep them open. That is why there is a locking pin blocking the transmission and the car from moving. It's not possible to leave the car "in gear" like with manual transmission.

[aboulfad]

Originally Posted by KevinGS

So much good info here, Ab, many thanks!

I swear this forum makes all the difference in whether I "like" my M4, or LOVE my M4. The Mercedes forums are cool, but this Bimmerpost forum is priceless. You guys rock!!!! SOOOO much info, I love it.

Glad you enjoy them, but check the stickies here for the F8x and in there are bunch of technical info as well the S55 engine technical guide. Maybe now you can join the rest of the transmission engineers in their discussions about Neutral

Tomm3 , for a start read http://f80.bimmerpost.com/forums/sho...&postcount=409

Other technical maintenance information is found in Rheingold which I have saved some docs, check this thread.

[SD ///M4]

Originally Posted by jwzimm

Interesting. I may have misheard this then. Your description makes sense.

Does the DCT have a parking pawl?

I don't know if the DCT has a parking pawl as described in the Wikipedia description, but in the Technical Training F80/82 Complete Vehicle document being discussed and linked elsewhere in this thread, on page 38 the diagram shows a "Parking lock" which likely locks the output shaft, thus keeping the vehicle from moving when in Park. The exact mechanism isn't described in detail.

[mkoesel]

You are probably correct, CAM3.

I am likely recalling informational texts with diagrams similar to what CaryTheLabelGuy (first rate user name, BTW) had posted where, from a high level, conceptual point of view, the clutch pack could be thought of as only allowing one side or the other closed at a time. In reality, the actual mechanics are more complex, probably similar to what you describe. And it may very well be the case that, in reality, the hardware, including the hydraulics and related valving is actually capable of closing both clutches at once. That being said, I would suspect that the microcontroller logic is programmed in such a way that the commands issued by TCU are only able to perform operations that make sense. This would include not allowing both clutch plates to engage and remain that way. This would be similar to how a computer's hard drive firmware may not allow one to create a device driver that spins the thing into oblivion and destroys the platters or somesuch. This is just an educated SWAG on my part though.

Originally Posted by CanAutM3

IMO, it is possible to have a seamless shift, without any interruption of power transmission, with two complete clutch units.

While accelerating in the lower gear, the clutch of the lower gear is closed while the next higher gear is engaged but with its clutch open. For the higher gear clutch that is open, one half is turning at the same speed as the closed clutch since it is driven by the engine while the other half of the open clutch is spinning slower because it is driven by the rear wheels through a higher gear. During the shift, the open clutch starts to close and at some point it is starting to transfer power through dynamic friction. At that point, the lower gear clutch starts to open and also starts to slip as the engine slows down while the other clutch picks-up more and more power.

This is what I mean by the two clutches "slipping against each other".

[aboulfad]

here is more details on the parking lock of the DKG (albeit from M5?):
A locking magnet parking lock is connected to the twin-clutch gearbox control unit (DKG). The parking lock is engaged when the locking magnet is supplied with current.

When the ignition is off, the locking magnet is no longer supplied with current by the twin-clutch gearbox control unit. The retaining tab moved back. The release lever becomes free. The parking lock is engaged on the twin-clutch gearbox through spring force via the cable.

1. Holding magnet, parking lock
2. Release lever
3. Release lever
4. Retaining tab
5. two‐pin plug connection

The automatic parking lock is engaged automatically for the following conditions:

- After stopping the engine with ignition off, if Position R or D is engaged.
- For ignition off, if D is engaged.
- At a standstill with engaged transmission position R or D if the seat belt of the driver's seat is put off, the driver's door is opened and the brake is not applied.

[lemetier]

Originally Posted by SD ///M4

I don't know if the DCT has a parking pawl as described in the Wikipedia description, but in the Technical Training F80/82 Complete Vehicle document being discussed and linked elsewhere in this thread, on page 38 the diagram shows a "Parking lock" which likely locks the output shaft, thus keeping the vehicle from moving when in Park. The exact mechanism isn't described in detail.

DCT does have a pawl gear

[Tomm3]

Thanks for pictures.
So our DCT hase the same locking capabilities as regular automatic. Which is obvious as without oil pressure the clutches are open and car would roll even if left in gear.

Do you have pictures and data about clutch pack?

[lemetier]

Originally Posted by Tomm3

Thanks for pictures.
So our DCT hase the same locking capabilities as regular automatic. Which is obvious as without oil pressure the clutches are open and car would roll even if left in gear.

Do you have pictures and data about clutch pack?

Correct as the hydraulics operate opposite of SMG.

This is the Training Manual from the E9x that is still notated as the current reference for further information.

http://www.m3post.com/forums/attachm...4&d=1317583307

Most of it has carried over unchanged but there are some differences including:

- Some early cars still have the interior emergency parking lock release however it requires complete removal of the center console to access. Later cars and M4 GTS do not. A separate release is located on the right side of the transmission housing and is even more difficult to access.

- Emergency Hydraulc Operation while running had been removed. Major faults or loss of power to the mechateonic control unit causes a gradual loss of pressure to zero and disengages the active transmission (the logic is Transmission A - Clutch1 / Transmission B - Clutch2). the car can still be driven to a close stopping point, but will no longer proceed beyond that point.

- A backup hard wire sensor to limit Vmax has been added and cannot be bypassed. This is above the Primary TCU Vmax limiter.

[Tomm3]

Thanks a lot!

[jwzimm]

And here is why the DCT (and other "complicated park engagement" transmissions) scare me: https://consumerist.com/2016/06/20/j...-actors-death/

Granted it is not the BMW DCT in this case but it is a good example of how an ambiguous or difficult process to engage Park can result in catastrophic consequences. Why a "P" position on the shifter cannot exist I will never understand!

[MFNATIK]

Originally Posted by jwzimm

And here is why the DCT (and other "complicated park engagement" transmissions) scare me: https://consumerist.com/2016/06/20/j...-actors-death/

Granted it is not the BMW DCT in this case but it is a good example of how an ambiguous or difficult process to engage Park can result in catastrophic consequences. Why a "P" position on the shifter cannot exist I will never understand!

Yeah, so sad. This is why I still treat my DCT as if I was driving a manual. I always pull the parking brake when I park. I'm not going to trust it and I also figured it would take some pressure off the parking pin at the same time.

[CaryTheLabelGuy]

Originally Posted by MFNATIK

Yeah, so sad. This is why I still treat my DCT as if I was driving a manual. I always pull the parking brake when I park. I'm not going to trust it and I also figured it would take some pressure off the parking pin at the same time.

^^^^ This.

Alway use the parking brake. Always.

[lemetier]

Originally Posted by MFNATIK

Yeah, so sad. This is why I still treat my DCT as if I was driving a manual. I always pull the parking brake when I park. I'm not going to trust it and I also figured it would take some pressure off the parking pin at the same time.

This isn't the same situation though. DCT (and SMG) are automated MANUAL transmissions. If you exit the car and don't apply the parking brake, it may roll and that's not a fault of the car. With the FCA vehicles, and what occurred in this case, when he opened the door, the car did not automatically engage park, and moved under power. It didn't just roll away. This is why BMW's (and Audi's that use the same gear selector as that Jeep) engage park automatically if a door is opened.

[jwzimm]

Originally Posted by miamiten

This isn't the same situation though. DCT (and SMG) are automated MANUAL transmissions. If you exit the car and don't apply the parking brake, it may roll and that's not a fault of the car. With the FCA vehicles, and what occurred in this case, when he opened the door, the car did not automatically engage park, and moved under power. It didn't just roll away. This is why BMW's (and Audi's that use the same gear selector as that Jeep) engage park automatically if a door is opened.

You just contradicted yourself here. You stated that:

If you exit the car and don't apply the parking brake, it may roll and that's not a fault of the car.

But then:

This is why BMW's (and Audi's that use the same gear selector as that Jeep) engage park automatically if a door is opened.

So unless you are thinking of "Dukes of Hazzard-ing" and climbing out the window, which is it? Does the DCT automatically engage the parking pawl when you exit or not?

My point in this is that, while we enthusiasts may understand that the DCT is really just a manual transmission that has a computer controlled clutch (or clutches), the average lay-person will not. They will see it as any other automatic transmission. That could make for some unfortunate results in valet parking lots or other similar situations such and lending your car to someone. All of this would be solved with a simple button with a "P" on it that would allow you to indicate to the computer that you wish for the transmission to be placed in Park and not move. Instead we are given a complex procedure to follow to engage park.