It looks like you verified that every combo of slave cylinder and rod will work. You already had the clutch releasing most of the way. Those two things together mean you stop looking for the problem in the hydraulics. The trick is to find the last thing that is physically dragging. You said you ran the clutch assembly through its travel on the bench and it never released, which i assume means you checked it at various points through its travel to make sure it wasn't releasing and then re-engaging again at full travel. If it just never released, there is going to be a contact point on the flywheel/disc/PP surfaces that remains touching. If you never really forced it to drag there at more than hand speed or for more than a second there may not be obvious witness marks, which is why i suggested letting the engine make it slip for a while to make your clues more obvious.
Anyway, from the picture of your flywheel the only thing i would look at on that side is if there is a slight lip on the flywheel surface right at the outside diameter of the clutch disc and whether your clutch disc might be just large enough in OD to have its outer edge sitting on that raised lip.
Vigo (Forum Supporter) said:
You said you ran the clutch assembly through its travel on the bench and it never released, which i assume means you checked it at various points through its travel to make sure it wasn't releasing and then re-engaging again at full travel.
Yep, I checked that but my method of turning the disc isn't the best. To do it better, I would need to use a press to compress the PP and use the clutch centering tool from underneath to spin the disc. I think it's a good excuse to purchase a press, but all the places will only re-open on Monday.
I'll go back to the shop today. I think that with all those tests, the hydraulics are OK. I'll check if I can see some witness marks somewhere. With the force on the pressure place from the 6 bolts holding it, I'm sure there will be some sort of mark if there was contact.
Is there a spacer or thrust bearing or something like that missing?
Rocambolesque said:
I checked the clutch disc thickness, 9.5 mm. I don't know if that is correct:
I have never seen a clutch disk's internal spring have that much free height. The purpose of the spring action is to make engagement more progressive and less of a switch. Normally they have less than half of that much travel.
Replace the clutch disk, preferably from a different manufacturer, I bet you will be able to disengage.
Pete. (l33t FS) said:
Replace the clutch disk, preferably from a different manufacturer, I bet you will be able to disengage.
I think this is my next step. I went back to the shop today and re-did the pressure plate test. I found out that at full travel, the clutch disc can be moved, I just had to put more force on the pick. However, it is not "free", in other words there is no gap between the clutch and the flywheel and pressure plate.
I read a bit about that marcel spring between the 2 linings of the clutch disc. I didn't know what was the purpose of that spring. The good side of taking this apart many times and studying the part is that I now know more about how the clutch works!
Just thinking out loud here, but hopefully it gives some food for thought.
Basically, as soon as the diaphragm starts being depressed and flexes to the point that it's overcoming clamping pressure and moving the pressure plate away from the flywheel, you need minimal travel for the clutch to release. Say you're getting 14mm of travel at the ends of the diaphragm fingers/throw out bearing and your diaphragm has aprox a 3:1 ratio, that's still 4.5ish mm of actual pressure plate travel. Now, based on your current measurements, and knowing each friction surface is 3.5mm thick and your over all clutch disc is measuring at 9.5mm, that leaves 2.5mm. of that, the center plate of the clutch would account for say 1-2mm..... Even if it was only 1mm, that gives you 8mm of clutch disc thickness when it's clamped. That leaves 1.5mm of travel before you begin breaking contact between the pressure plate and the clutch disc.... Going back to the above figure of 4.5mm, that would be more than ample for your clutch to disengage wouldn't it... Somewhere in the vicinity of 3mm overall clearance if you moved the pressure plate full travel.
So that leave one of two potential causes to your problem... Your hydraulics aren't, for any number of reasons providing sufficient travel to depress the diaphragm far enough to achieve full disengagenent. Are both master and slave new? If not, you can get hardening of the cup seals, meaning it will work ok after bleeding, but by the next day, the seals have moved away from the cylinder walls and fluid will bypass, reducing your travel. Or, maybe you've ended up with a mismatched slave/master combo. An oversized master/undersized slave will give you excessive travel and reduced mechanical advantage, a undersized master/ oversized slave will reduce your travel but give you increased mechanical advantage.
Or, you have something physically binding... Think spigot not locating correctly (causing loading on the input shaft and transmitting some drive despite the clutch being disengaged), clutch disc fouling on flywheel bolts, overpressing the diaphragm causing it to reapply pressure to the clutch disk, that sort of thing.
Both cylinders were new when I did the swap. I renewed both again when I started doubting the hydraulics. Both from LUK, part numbers check out, etc...
Like I said, at full travel of the pressure plate, the clutch isn't released. Even with the longest rod in the slave cylinder and no shim between the slave and the transmission (OE setup has a 2 mm shim) trying to max out the pressure plate, the clutch doesn't fully release. It has almost no clamping pressure, but still to much to engage the gears with the engine running.
I ordered a new Sachs clutch kit. We'll see how thick the disc is on that one when it gets here on January 7th!
Problem solved, I think... I received the Sachs clutch kit and did the same test where I bolt the PP and disc to the flywheel and compress is with a bolt. About half way down the PP travel the disc is clearly free inside and can be moved without forcing. Now there is a gap between the disc and the flywheel/pressure plate. I just need to figure out why the old LUK kit wouldn't work as both are not quite identical but the key dimensions are the same. The pressure plate spring seems softer on the Sachs clutch compared to the LUK.
I'll install in the car and report back. But I think it'll work now. Thanks a lot to everyone who participated in this thread!
That is awesome news. I noticed that marcel spring as well but didn't feel confident in my suspicions. Can you visually compare the two discs before it goes in the car for good? (for good because its gonna work! 
)
I measured the new disc with a measuring tape yesterday, it was around 10 mm. Old one is 9.5.
I'm just wondering if its visually obvious that the marcel spring on the new disc has much less 'travel' than the old disc.
If that is indeed the issue, it's an issue I've never run into before.
Let's play "spot the differences"
Old LUK on the left, new Sachs on the right:
I even re-did some more tests today. I tried the Sachs plate with the LUK disc and it releases. The LUK plate with the sachs disc doesn't release. The culprit is obviously the LUK plate. But all of this is not to discredit LUK's products. I've had great success with them before, the cylinders I have work great and a good friend of mine recently installed one of their Repset clutches (just like the one I have) in his Jeep and it works good. Maybe I just bad luk (see what I did there...)
In reply to Rocambolesque :
Must have been made on a Friday afternoon.
Weird stuff happens. I've installed dozens of LuK clutches and never had an issue.
To be honest, I wouldn't have suspected the clutch disk either, if I was the one putting it together, because I've never seen that happen before.
Thanks for the detailed follow up pics. It's interesting how much less mass/material seems to be in the 'plate' of the of the new pressure plate. Looks like you might be losing a couple pounds of rotating mass, at least from the visuals.
