Hydraulic clutch gets heavier as engine warms

[Wing Commander]

I have recently replaced the entire hydraulic clutch system on my Saab 9000. This included the master cylinder, the slave, the clutch including thrust bearing and also updated the hose to a braided variety.

Clutch and brakes share a reservoir.

I have bled the front two brakes but the back two bleed nipples were stuck so will do that another day. Have also bled the clutch through.

So why does my clutch start off being all silky silky smooth and then get progressively heavier as the car warms up. It's never mega hard, but it's harder. Also, the clutch pedal takes a fraction longer to return too.

No slipping occurring, and no leaks anywhere.

Any thoughts??

[Mroad]

From experience, albeit not with a Saab, I would say it points to a fluid moisture issue or trapped air. Try rebleeding everything and, if not already, make sure you use fresh fluid (fresh from the shop).

[Wing Commander]

The 9000 clutch is particularly difficult to bleed, requiring a pressure bleed kit. I have been using one but maybe I still have a bit of air in there. I will bump up the feed pressure and also get my trusty assistant to pump the clutch a couple of times.

Will keep you posted.

[Wing Commander]

OK, update.

This problem has now started to manifest itself in a slightly more serious manner.

A few days ago, I bled the clutch system again and things improved quite a bit. But the pedal still got a bit more stiff after a longer run.

Today, driving through town, and I couldn't engage any gears - the reason being the clutch pedal would not go down far enough. A few depressions later and it still wasn't playing ball. I turned the engine off, engaged first, and all was working again. But within a few seconds, after another gear change, the clutch was still depressed, even when the pedal was released. Engine revving but no drive.

The whole clutch system was replaced recently by myself, but I am thinking that the clutch slave cylinder may be sticking in some way. This would explain not being able to fully depress the clutch (slave not extending) and then no drive (slave not fully retracting).

What are people's thoughts? Could a slave really be defective from new??

[Wing Commander]

After doing some google investigating, it turns out that a few batches of slave cylinders had a crap type of grease applied to them. It works fine when the engine is cold but when heat soak from a motorway run etc warms it up, it's far less effective. Results in a stiff clutch. Mine must have just been sticking quite a bit.

Same happened today. Lovely when cold, got worse as it got hotter.

My question is this. Surely a hydraulic slave cylinder is lubricated by the hydraulic fluid? Every mention on the forums say the problem disappears a few weeks after fitment. Why would grease be on the slave if the fluid is used as lubricant???

[Wing Commander]

Update time.

I have been driving the car on a complete mixture of roads and have clocked up at least 750 miles since the clutch change, if not more.

The clutch is now fine to use, in terms of reliability. But it is still slightly heavier and slower to return after an A road drive at 50-60 or motorway at 70-80.

Bare with me on this. Seeing as I haven't managed to get the two rear brakes bled yet (bleed nipples are stuck in), and the brake and clutch share the same cylinder, could the following be feasible?

When starting from cold, the clutch works fine. The reservoir level is ok, and everything is good in the world. As the car warms up and gets driving down A roads/motorways, the brake/clutch fluid warms up. The old fluid that is in the rear brake lines is old and manky meaning it has moisture/air in it which now expand. This forces up the reservoir level and acts upon the clutch in some way. For example, if the reservoir is now under pressure, it would take longer for the excess fluid in the clutch line to bleed back up to the reservoir, hence the slow return on the clutch.

Have I made any fundamentally unsound errors in my thinking??

[Adam205]

The fluid reservoir can't be under pressure as it should be vented.

All these complicated explanations are all well and good, but its 100 times more likely that it's 1) a dead slave cylinder, 2) a dead master cylinder or 3) lots of air in the system.

[Wing Commander]

Adam205 said:

The fluid reservoir can't be under pressure as it should be vented.

All these complicated explanations are all well and good, but its 100 times more likely that it's 1) a dead slave cylinder, 2) a dead master cylinder or 3) lots of air in the system.

Slave and Master were both replaced, by me, about 6 weeks ago. The clutch is engaging and disengaging with no problems, but gets heavier when the system is heated.

I thought the reservoir might not be pressurised but thought it the most likely possibility is the vent hole might be blocked.

Having air in the system - would that make the clutch more spongy or more firm. I guess spongy when cold (air being compressible) but when heated up, the air could increase in pressure giving the feeling of a much firmer pedal. Would any trapped air ever get to this sort of pressure though?

Edited by Wing Commander on Wednesday 15th September 11:29

[Kitchski]

Clutch fade NFTW!Until you bleed the rears you won't know but being that they're closed circuit I think the chances of cross contamination are pretty slim. Easiest to put right though! Otherwise that box is coming off again!

[Wing Commander]

Kitchski said:

Clutch fade NFTW!

Until you bleed the rears you won't know but being that they're closed circuit I think the chances of cross contamination are pretty slim. Easiest to put right though! Otherwise that box is coming off again!

I have the replacement bleed nipples but just need to find the time to get the car jacked up, blow torch and other trickery out, and to sort them. I have to wait until I know that I do not need the car for a few days, just in case I shag something up.

Like you say, once I have done the rears, I can bleed it all through with some brand new fluid and see what changes, if any, it makes.

Really, really do not want to take the box off again. That was 3 days of hard graft when I don't have an engine lift!

[guru_1071]

did you replace the rubber hose between the master and slave?

they can perish internally and can start to act like one way valves (particually if people have nipped them with grips to stop fluid leaking when replacing slaves)

[Wing Commander]

guru_1071 said:

did you replace the rubber hose between the master and slave?

they can perish internally and can start to act like one way valves (particually if people have nipped them with grips to stop fluid leaking when replacing slaves)

Yep, replaced it and, because of this pipe's known problems, I replaced it with the Abbotts Racing braided hose version.

[Vidal Baboon]

If you've replaced all hydraulic lines & still experience clutch drag when warm, have you taken a look at the mechanical side of it?

It would be pretty daft to have a linked clutch & brake system wouldn't it. So air in the rear brakes would make no difference to the operation of the clutch IMHO.

Maybe you have bent clutch fingers or lever arm & as the friction plate heats up (expands) it is dragging on the flywheel.

Does the clutch move freely on the input shaft?

I remember doing a clutch on an old battered Cavalier as an apprentice & getting a bking for putting the plate in the wrong way- you couldn't select any gears with the clutch depressed afterwards.

Turned out I'd actually done the job right but the crank thrust bearings were so shagged & the whole crank was shifting as pressure was applied to the clutch

Edited by Vidal Baboon on Wednesday 15th September 12:50

[DrTre]

Vidal Baboon said:

It would be pretty daft to have a linked clutch & brake system wouldn't it. So air in the rear brakes would make no difference to the operation of the clutch IMHO.

That's the way it is. The reservoir has a baffle to prevent a leak in one system terminally affecting the other.

[Vidal Baboon]

DrTre said:

Vidal Baboon said:

It would be pretty daft to have a linked clutch & brake system wouldn't it. So air in the rear brakes would make no difference to the operation of the clutch IMHO.

That's the way it is. The reservoir has a baffle to prevent a leak in one system terminally affecting the other.

Ok, so the air that potentially is stuck in the rear lines is going to make it's way up to the master cylinder, go against gravity just before breaching the surface & enter the clutch line?

If you split a brake line are you telling me the clutch won't work either?

Or that if the clutch line breaks, all braking pressure will be lost through the split clutch pipe?

Come on, really? I thought manufacturers had designed systems out like that since the late 70s- like having a cross-over in the brake lines.

A baffle? Surely you mean they share a reservior but each operation is completely seperate & will be fed from said reservior.

Edited by Vidal Baboon on Wednesday 15th September 12:37

[Dr Doofenshmirtz]

The braking hydraulics will have absolutely NO, nothing, nadda effect on the clutch hydraulics - all they do is share the same header tank - nothing else.
Was everything OK before doing the clutch?
You should be able to pump fluid through the system using the old two person method...open bleed nipple, depress clutch, close bleed nipple, raise clutch, rinse and repeat ad nauseum.

[Wing Commander]

Vidal Baboon said:

If you've replaced all hydraulic lines & still experience clutch drag when warm, have you taken a look at the mechanical side of it?

It would be pretty daft to have a linked clutch & brake system wouldn't it. So air in the rear brakes would make no difference to the operation of the clutch IMHO.

Maybe you have bent clutch fingers or lever arm & as the friction plate heats up (expands) it is dragging on the flywheel.

Does the clutch move freely on the input shaft?

I remember doing a clutch on an old battered Cavalier as an apprentice & getting a bking for putting the plate in the wrong way- you couldn't select any gears with the clutch depressed afterwards.

Turned out I'd actually done the job right but the crank thrust washers were shagged & the whole crank was shifting as pressure was applied to the clutch

Edited by Vidal Baboon on Wednesday 15th September 12:27

When you say clutch drag, what do you mean? The clutch operates just as it should. The only 2 things that are out of the ordinary is that it gets heavier to operate when the car has been on a run, and it sometimes takes a bit longer for the pedal to return to the top.
I'm happy that the clutch and brake systems are not interlinked critically, and the only thing they share is the reservoir. My concern was that the pressure release valve (for want of a better word) in the reservoir might have been blocked. Any air in the brake lines would heat up under use, pressurise and push fluid back to the reservoir. Since the air in the reservoir can’t vent, it essentially pressurises the whole system, which would slow any clutch operation somehow.
Like I say, that is a long shot and just me thinking out loud. I am most likely miles off, mechanically speaking.
The clutch moves fine on the input shaft, with the engine off or on. I do remember that not too long after I did the work, when I had the engine on and I was doing repeated depressions, it would sometimes get harder to press. I guess this points towards the clutch fingers.

DrTre said:

Vidal Baboon said:

It would be pretty daft to have a linked clutch & brake system wouldn't it. So air in the rear brakes would make no difference to the operation of the clutch IMHO.

That's the way it is. The reservoir has a baffle to prevent a leak in one system terminally affecting the other.

I don’t believe their is a baffle, per say. The reservoir is shared, and the clutch feed is taken from half way up the tank so that a clutch failure would not wipe out the feed to the brakes.

Vidal Baboon said:

DrTre said:

Vidal Baboon said:

It would be pretty daft to have a linked clutch & brake system wouldn't it. So air in the rear brakes would make no difference to the operation of the clutch IMHO.

That's the way it is. The reservoir has a baffle to prevent a leak in one system terminally affecting the other.

Ok, so the air that potentially is stuck in the rear lines is going to make it's way up to the master cylinder, go against gravity just before breaching the surface & enter the clutch line?

If you split a brake line are you telling me the clutch won't work either?

Or that if the clutch line breaks, all braking pressure will be lost through the split clutch pipe?

Come on, really? I thought manufacturers had designed systems out like that since the late 70s- like having a cross-over in the brake lines.

A baffle? Surely you mean they share a reservior but each operation is completely seperate & will be fed from said reservior.

Edited by Vidal Baboon on Wednesday 15th September 12:37

See above. I didn’t intend to imply the air would work its way down into the clutch system. More that the air in the reservoir might be pressurising and affecting the clutch operation. Most likely wildly out on this.

[Vidal Baboon]

Wing Commander said:

Vidal Baboon said:

Maybe you have bent clutch fingers or lever arm & as the friction plate heats up (expands) it is dragging on the flywheel.

Does the clutch move freely on the input shaft?

I remember doing a clutch on an old battered Cavalier as an apprentice & getting a bking for putting the plate in the wrong way- you couldn't select any gears with the clutch depressed afterwards.

Turned out I'd actually done the job right but the crank thrust bearings were so shagged & the whole crank was shifting as pressure was applied to the clutch

Edited by Vidal Baboon on Wednesday 15th September 12:27

When you say clutch drag, what do you mean? The clutch operates just as it should. The only 2 things that are out of the ordinary is that it gets heavier to operate when the car has been on a run, and it sometimes takes a bit longer for the pedal to return to the top.
I'm happy that the clutch and brake systems are not interlinked critically, and the only thing they share is the reservoir. My concern was that the pressure release valve (for want of a better word) in the reservoir might have been blocked. Any air in the brake lines would heat up under use, pressurise and push fluid back to the reservoir. Since the air in the reservoir can’t vent, it essentially pressurises the whole system, which would slow any clutch operation somehow.

The cap should be vented- there should be a little hole in the top.
Clutch drag- the friction plate does not completely clear the flywheel.

You would immedeately feel air in the system- hot or cold. What happens when fluid get old is water enters the fluid through moisture in the air- as it's hydroscpoic.

Being a design engineer you will know you cannot compress a liquid, when the water heats up through friction of the brakes, that water boils & turns into a compressable gas.

Anyway, I don't want to teach you to suck eggs, maybe someone will read this & make use of it. If the pedal is not returning quickly, then check that the friction plate is moving freely & being pushed back onto the flywheel by the clutch fingers.

Get someone to press the clutch while you check it's returning properly. If it's taking it's time, as you say, it could be a blocked vent hole (possible, but probably unlikely) or the clutch is sticking.


I hate to ask you this, but are you sure you put the friction plate in the right way around?

[The Wookie]

Have to say it just sounds like a master or slave cylinder issue to me, with the fact that it's heat dependent swaying me further towards the slave.

Personally completely flush the system through to make sure it's not fluid related, and if that fails just bite the bullet and swap the slave.

[Wing Commander]

Vidal Baboon said:

The cap should be vented- there should be a little hole in the top.
Clutch drag- the friction plate does not completely clear the flywheel.

You would immedeately feel air in the system- hot or cold. What happens when fluid get old is water enters the fluid through moisture in the air- as it's hydroscpoic.

Being a design engineer you will know you cannot compress a liquid, when the water heats up through friction of the brakes, that water boils & turns into a compressable gas.

Anyway, I don't want to teach you to suck eggs, maybe someone will read this & make use of it. If the pedal is not returning quickly, then check that the friction plate is moving freely & being pushed back onto the flywheel by the clutch fingers.

Get someone to press the clutch while you check it's returning properly. If it's taking it's time, as you say, it could be a blocked vent hole (possible, but probably unlikely) or the clutch is sticking.


I hate to ask you this, but are you sure you put the friction plate in the right way around?

Can’t see any hole but assume it’s vented in some way. Will go and have a quick look in a minute. No clutch slip or drag is happening. The engagement/disengagement is all working fine, its just the transition that is slow when hot.

To be honest, I am just clutching (see what I did there ) at straws with the moisture in the brake lines. What you have said makes sense though – it would produce the opposite results to what I am getting ie hard when cold, getting spongey when hot.

Don’t worry – know eggs being sucked. I came on here for advice and that’s what I am getting!
Checking the clutch operation might be slightly difficult. It only really manifests after higher speeds. Will see if I can manipulate it into playing up at a time when I can then pull over and check it.

Pretty sure the friction plate went on the right way – this was my first clutch change, but I just replaced the bits the same way they came off. I guess there is a remote possibility it went on back to front. How would this effect the operation of the clutch?

The Wookie said:

Have to say it just sounds like a master or slave cylinder issue to me, with the fact that it's heat dependent swaying me further towards the slave.

Personally completely flush the system through to make sure it's not fluid related, and if that fails just bite the bullet and swap the slave.

I have flushed the system through quite a few times. No real change. The 9000 is quite difficult to bleed effectively though, apparently. I have been using a pressure bleeder (which is a must on this car).

The thing is, the slave was brand new. And its a box off, clutch out job. So really only want to drop 2 days or so and lots of grief on doing this if I really have exhausted all other options. Which I am starting to do...