Clutch problems, caused by modifications or bad parts?

[VitesseEFI]

About 20 years ago I got fed up with whining, grinding, leakingTriumph gearboxes and managed to shoehorn a Toyota W58 from a Supra into my Vitesse.

I couldn’t get much sense out any of the companies offering conversions (whether in NZ, Oz or US), and anyway I’m cheap, so I made my own. Some detail on that here

https://triumphowners.com/to-car/vitesse-mkiii-efi/

(Scroll down to special interest projects).

One of the challenges was the clutch. What I ended up with was the original Triumph pressure plate squashing a friction plate listed for a 70s 1.6 Hiace. It was 200mm rather than 215mm and a bit thick, but it worked. Spigot bush was a modified oilite top-hat bush fitted to the flywheel (oe bush is in the crank but Toyota shaft too stumpy to reach). The release mech including bearing and slave cylinder were standard Toyota Supra.

After about 10k miles it was becoming obvious that the clutch had problems. Odd noises when taking up and lots of shunt when trickling along in traffic accompanied by clanks and squeaks from the bell housing. Managed about another 5k and pulled it apart. Basically, the sprung centre was disintegrating and the springs utterly crushed. I put this down to the combination of the springs always being too light for the 2L torque and it being a rather crappy QH aftermarket plate. Managed to find a better quality one and slapped it back together. All good for about 10k miles (some of them very hard miles) before the noises started again. Engine was done too so everything came out for a refresh. Once again the sprung centre was collapsing with very crushed springs.

This time I splashed out and bought a bespoke 215mm friction plate built for the application from Helix Motorsport. It was expensive - seemed well made. Also used a brand new cover plate, machined the flywheel to take the OE Supra ball-race spigot bearing and spent some time checking concentricity and made some minor changes to doweling.  I also did some detailed measuring and determined that a 5/8” M/C should give the correct throw. I had been using 3/4”!

Once back together it worked beautifully. Light action, biting point mid-stroke, plenty of bite. No more excessive shunt. Only concern, slight squawk from the bellhousing on initial pedal pressure.  Anyhow, got some miles on the new much fiercer engine, mapped it. Went for a hoon around Europe. All good.

Except….. after about 5k, the biting point was near the floor. Then on the floor…. Then under the floor. Could find nothing wrong externally, so I cheated and fitted a 0.7” M/C. Problem solved. Slightly heavier pedal but otherwise good. Apart from that squawk.

Fast forward a few years and maybe 8k miles and preparing for another European mission, I have my wife drive the car as practice and she struggles because the biting point is once again on the floor. I cheat (again!) a stick a 3/4” M/C. Problem solved. It’s a bit heavy now though.

So, in the last 4 to 5 years and 12k miles the biting point has continued to go down. Or rather, it’s been getting wider. That is, you have to crush the carpet to get it to clear, but it doesn’t fully bite up until right near the top, and how much bite comes in at what point in between seems variable.

Yesterday I decided it was time to stopping ignoring it (apart from moaning about it) and investigate.

Wrestling it out is a bit of a mission. Some of the bellhousing bolts are very inaccessible, it’s probably my fault (consequences of various mods) but the Audi design guys would probably give me 8/10. Two marks deducted for not using Torx/triple hex heads.

This bit looks normal-ish. Bit grubby and the plate is very sticky on it ….. and there’s loads of backlash!

Spline in my expensive Helix plate is completely borked.

This is what it should look like (this is a 200mm Exedy plate)

So, that’s part of the problem. The plate hanging on the splines would explain the variability and possibly some of the long travel.

Also….

This looks distinctly unpromising. Especially since only about 28k miles in total. That’ll be the squawk then….

Friction surfaces all look fine. Plate is 1/3 worn at most. It’s never slipped.

So what’s going on? 
- Misalignment is a possibility though I’m struggling to see clear signs.

- It is possible (probable) that the release bearing profile is wrong. It’s the Toyota one for a 250mm clutch being used with a 215mm cover. The Triumph release bearing doesn’t fit the carrier (2mm smaller ID, very thin wall on the carrier)

- It’s possible that the cover assembly is poor quality. Can’t remember where I got it.Its marked QY13039. More QH crap?

-That spline though. What has done that? Never seen anything like it. I’ve still got the last plate I used and the spline on that is perfect. Expensive and crap?

-Will do some more measuring today to check the release mech geometry but I don’t think it’s the problem.  
 

I have new Sachs cover intended for a Saab that usually works very well, but obviously the friction plate and release bearing need more thought..  

Thanks for reading. Ideas welcome!

 

[Doug Hill]

Misalignment will wear splines like that. As will lack of appropriate lube.

Check crank thrust too. Then measure the clutch stack height and try find some proper specs for the parts you're using 

If the stack height is wrong the diaphragm may be touching the disk (almost looks like it has kissed the springs) and this will then eat the clutch and thrust bearings in the engine 

[Mof]

Have you got freeboard at the clutch master pushrod/ booster pushrod? If the pushrod is slightly pushing on the master  won't let new fluid into the system so when the clutch wears the pedal will go down. 

Also, grease your shaft, stop the spline wearing and the disc hanging up on the input shaft. 

Check measurements. I was thinking the release bearing is bottomed out on the nose cone, and still just touching the pressure plate fingers. That would cause the wear on the fingers 

[mjrstar]

I have messed around with numerous clutch setups that have been piecemeal together from different cars.. they can be trickey to get juuust right..

 

It kind of sounds like the release bearing is backing off too far from pressure plate, or it's getting air in the system, or maybe a spongy clutch hose. Ie you don't get the travel at the release bearing that you apply at the pedal..

 

Finger damage probably occured when you fitted the bigger master which is over-throwing the release bearing, assuming the bearing profile and diameter are correct...

[VitesseEFI]

Thanks. Appreciate the input.

Some of these I can answer, some I’m going to have to check.

The spline was definitely lubed when assembled. Grease used was from a tube of special LUK spline grease that I’ve used on every clutch I’ve done in about the last 15 years. Seems to work elsewhere! There’s no trace of it now though.

Misalignment. Yeah. I’d like to think not but…. What I can say is that when last assembled the engine was out of the car and it went together really nicely. Clack! Straight up to the engine, no pulling it up with the bolts. With the ball-race spigot bearing and the gearbox bearings still being decently tight, I just can’t see that happening without alignment being decent.  I will revisit…

Crank thrust I will check. It is a known weakness on these engines - though this one has PB thrust washers so I’ll be (very) disappointed if it’s this.

Hydraulics I’m confident are good.  All hard-piped (with a coil for flex) and there’s travel to spare in both directions.

I’ve been measuring the release arm geometry and running it through a full stroke to characterise it. It’s running roughly in the middle of its range. Haven’t plotted the results yet but, quick and dirty, I measured the slave stroke before dismantling at 20mm and this seems to equate to about 9mm at the bearing. I measured a very similar clutch a couple of years back when designing the concentric setup on my GT6 and that needed 6mm for release,  so pretty sure I have more than I should need. So “over-throw” yes, almost certainly.

Various stacking heights still to check. Definitely no actual contact between the friction plate and the diaphragm, even with the over stroking.

Release bearing profile is definitely open to question. More work needed here.

[Muncie]

Are the thrust washers (in engine) shagged? Check endfloat of crank I battled clutch problems for ages because I didn't want pull a running engine out.

They can fall into sump if your real unlucky engine will run fine for a while but you will add a couple off mm to your clutch stroke.

See below not much too them

 

[yoeddynz]

11 hours ago, Mof said:

grease your shaft

[VitesseEFI]

Update on this.

Checked crank endfloat. 0.2mm. My build notes say it was 0.17mm when first assembled 28k miles ago.  So that’s good.

Did some testing using the mill

That’s the combo that’s just come out. The friction plate is only about 1/2 worn at 7.35mm.  It starts to release about  6.5mm but needs 8 to fully release as the spring has gone lopsided. That, with sticky splines will be the wide biting point I think. The same friction disc with my new Sachs cover starts to release at 3.5mm and is fully free at 4mm.

I gathered up the wreckage and took it to my localish clutch shop. He was very polite, but I think he was telling me his competitors friction plate spline was made of cheese. No doubt on his thoughts on the QH cover…. Bin food (that was before I fitted it!). They made me a replacement friction plate in an hour…. And it was half the price of the last one. Top service!

The new friction plate is 8mm thick and with the Sachs cover needs 5.5mm to start to release and is fully free by 6mm.

Meanwhile, I put my hybrid bell housing on the mill. Engine side face was a little wobbly so I faced it. Took less than 1mm to clean.

Fitted the backplate, flywheel and bell to my spare engine. Then posted the gear box in from behind. Spigot picks up in the pilot bearing first then the gearbox/bell dowels engage. Little bit of a wiggle and shove and on it pops. Alignment can’t be too bad.

But to check another way…. 

pop the backplate on the bell and it’s very tempting to think that 4” (3.997” in fact) hole might be intended to be concentric….. but is it?

Being both lazy and cheap I spent 10 minutes rummaging through my junk looking for a ready made tool….

Turns out I have this 3 jaw chuck which measures 3.993” and easy to nip onto the spigot.

Pops right on, no wiggles needed. I think she’s good.

Have also clocked the backplate on the mill. It’s a bit dished (0.5mm), but evenly so.  Within 0.2mm. Better than my 2 spares….

Made an alignment tool

Running out of excuses not to refit, except I tweaked my back a week ago and the refit is intensely physical…

 

[RUNAMUCK]

A local clutch guru once streesed to me that getting all the fulcrums/strokes/throws set up correctly is beyond critical when setting up hybrid/mish-mash installs...

[VitesseEFI]

He’s quite right obviously. I’m pretty confident that this aspect is good and the initial measurements I did tended to prove it. Will re-measure more accurately before refitting and may even plot a graph…..

Other thing is that I have had it working beautifully in the past …..  Even this particular collection of scrap worked very nicely when first installed.  I’m coming to the idea that this failure was mostly due to crap parts.