Summer running - 2000 Ducati ST2

[OddDuck]

Sorting out an issue with the Oxford Hot Grips - the left keeps coming loose. It was too loose a fit on the bars and despite using plenty of superglue it still breaks free, the latest time was just after coming back from the SI trip. This is the third time I've glued it on, hopefully it's the last...

I've increased bar diameter by using a rolled 0.10 mm shim. The shim material was bought through Blackwoods Protector, but it's widely available as kits at most engineering supplies shops - you get an envelope of common thicknesses. Marking and cutting was as simple as ruler, Sharpie and tin snips, rolling was done with a screwdriver clamped in the bench vise and bare hands.

I didn't manage to photo the gluing process due to having no time before the superglue would start setting, but here's the rough sequence:

Get shim fitted about 1/4 way into grip
Put glue on outside surface of shim, compress shim, push shim into grip
Dribble glue down split line of shim inside grip, run a ring of glue around inner perimeter of shim, then rapidly push shim onto bar
Wipe any excess off.

I wore safety glasses and gloves for all this of course (way too easy to glue fingers together or eyes shut etc) and had cheapo bath towels down over the bike's bodywork.

This done I then left it well alone while the glue set. It's very tempting to see it it's gone solid immediately after placing it by twisting the thing in order the test the strength of the bond, but if it breaks free, it'll never set properly. I'll be able to test it out next time I ride.

[OddDuck]

Fitting Bark Busters for winter.

It's going to be (ahem) interesting to see what comments I get, while out riding with these. I got caught in a southerly front plus rain during the SI tour and it would have been very nice to have had weather protection for fingers.

The guards themselves are the flexible textile versions of the usual ABS and aluminium Bark Busters. They're intended to flex when contacting fairings, so should go onto a range of sport / road touring bikes. They're obviously practical. Not sure about the look though...

They come with a fitting kit. I wanted to keep the bar end mirrors and this meant either stacking anodised or painted components and hoping everything held together, or making custom fittings. I've heard too many stories about bar ends falling off during rides so it was off to the workshop for me.

Bare aluminium on bare aluminium has an interesting property: it is a terrible bearing pair. The metal will instantly surface weld if there's any relative motion and then the surfaces tear each other up. The counterpoint to this is that it's a great clamping pair. The metal components will grip each other, even with relatively low clamping forces. Of course I'll have to see how I go with actual ride use and weathering.

[OddDuck]

Back into the clutch, as well. It's started juddering again, and there are notchy gear changes, so clearly the work done earlier hasn't fixed everything.

I stripped the cover, pressure plate and springs, then the plate stack, hub and basket. While doing this I noticed that the basket has a wear pattern on the tangs that I haven't seen before: it's sawtoothed in profile. What I've seen on Ducati baskets before has been semi-circled or dished in nice uniform curves matching the plate tangs.

The wear profile is also asymmetric, it's much more heavily worn in the drive direction than in on the engine braking side of the tangs. These other sides of the basket tangs are worn in the usual dish shape, indicating random clatter while idling in neutral, and this wear is much lighter.

The plate stack, carefully filed earlier, is not only mushroomed again but also angled in the drive direction. No wonder the clutch isn't releasing properly - the contact on the basket isn't square, it's angled, and will tend to hold the friction plates in instead of letting them slide outwards.

I'd noticed earlier that the basket seems to be loose on its bearings. After some thought, it looks like what's happening is that the basket is moving (ever so slightly) relative to the plate stack while riding with the clutch engaged, and this movement is causing this wear pattern on the tangs. It's happening under load, not while clattering in neutral. Or so I think. Certainly the basket shouldn't be this loose on its bearings. The friction plate tangs have worn very rapidly and gone from square filed to tapered, the damage I'm seeing has happened in just 2,000 km.

I've left the friction plates as is for now, but have filed the basket tangs straight and square. This is a bodge fix at best unfortunately, the friction plate tang to basket clearance is now a mile wide and everything's going to rattle, causing further rapid mushrooming and wear. While doing this I noticed what look like flap disk marks on the basket, it seems that someone has been in doing this before. Certainly the 8 x M8 bolts holding the basket to its carrier haven't been properly sealed. On assembly these are supposed to be gasketed with Loctite 510, so that engine oil doesn't enter the clutch compartment.

[OddDuck]

Changing out the primary drive output wheel's bearings.

Clutch completely stripped (hub and basket both off), it's the engine cover next. Before doing that I pulled the spacer from the input shaft spline.

This wee beast has an O-ring on an internal groove. It grips the shaft tightly and there's no positive feature on the thing to pull or lever at. I'd ending up taking a Dremel to the one on the 900SS to make slots and then using a bearing puller, this time around I made myself a handle via two hose clips. The 1/4" spanner and screwdriver bit were needed for tightening these up, there isn't access for normal tooling unless slender wall sockets are available.

Once the pump cover is off, the output wheel slides out. I used my slide hammer set to yank the outboard bearing and seal in one go. This showed quite a bit of black gunk on the narrow outer bearing - it looks like this area of the engine doesn't see much in the way of oil circulation. Particulates build up on the seal side of the bearing. It's easy to see why the basket might have got a bit wobbly.

As to why this might have happened, given that it's the same design as the 900SS and that was fine until engine rebuild at 46,000 miles, my best take is that it's a combination of my clutchless changes during the East Cape tour, the 10,000 km oil change intervals, and sitting around. A lot of black gunk seems to be in the oil that I drained (I'm at 35,000 km now, i.e. 5,000 since the last oil change). If the bike's been sitting for long intervals, and I know that the previous owner wasn't putting high mileages onto it per year, it's also believable that the bearing wasn't being wetted enough to be lubricated properly during engine startups. It's then placed under immediate load as soon as first gear is selected.

I drove the replacement bearings home with a socket and hammer. Simple enough but there were two problems, both of them my fault:

1) the bearings in this gear wheel sit face to face via a matched circlip and central spacer, these move just enough that they aren't a solid base to drive against. Every time I drove one bearing home I'd knock the other out. I finally realised that I couldn't drive against the wheel with the bearing itself sitting unsupported except for a soft rag, I had to put something underneath both wheel and bearing.

2) any dirt left in the socket, if it's been sitting upright on the bench or similar, will get knocked out and go straight through the nice new bearing on the first hammer blow. Gotta clean that socket properly first... I ended up spraying CRC through the bearings and then re-oiling prior to assembly.

[OddDuck]

Changing the large diameter seal in the pump cover. This is early (in terms of km's) but since I've got the spare available and the cover off, it makes sense to. The original seal does show signs of wear.

As far as I'm aware neither Ducati nor the aftermarket make any kind of a tool for this job. I'm not even sure what the official factory procedure would be... I ended up using a lathe and a drill press to make my own dedicated tool for this job.

The tool is in three main parts:

A collar with a throat wide enough for the seal to fit into at one side and flanged down at the other
A press plate, wider diameter than the seal
A driving plate, very slightly smaller diameter than the seal and machined precisely to fit the metal body of the seal at the rear.

There's a bit of juggling of plates, seal and collar to make the thing work but the basic idea is to drive the old seal out with the driving plate and then drive the new seal in with the press plate. The press plate being larger diameter than the seal means that it'll drive to the correct depth relative to the pump cover.

An improvement to the tool would be a very large diameter threaded fastener fixed in the center (60mm or larger) instead of using 8 independent screws, but this works, albeit with a lot of gradual tightening and some tilting of the seal relative to the cover.

[OddDuck]

Since I've bought the bike I've been thinking it's got a slightly twisted front end. A minor whoopsie at some point, slight wrench of the triple trees, perfectly rideable as is but annoying all the same. Today I finally got onto straightening it out and had an unexpected result.

The short answer is that the triple tree / steering stem / forks etc are all fine. I've got a bent RH clip-on. It's folded enough that the outboard end is about 10mm off where it should be. No wonder the bike feels a bit weird sometimes, or I've felt like I'm riding twisted up.

This is something quite new to me. I've certainly had bikes drop and rotate their clip-ons around the fork leg, or similar... actually bending one of these would take some force. They aren't tubular, these clip-ons (if that's the right name?) are solid aluminium castings. Given that they bolt fairly rigidly onto the top triple clamp plate though and there's no way to slip, it's believable that a relatively minor spill could do this.

The giveaway was eyeballing the line of the ruler against the line of the grip. I did this on both sides and the LH gives a parallel line between ruler and grip, with the RH as pictured. Took quite a while to get there though... I spent ages mucking around with lining the wheel up and trying to twist the bars. The giveaway was was that the bar bases measured the same on both sides (with wheels in line via string and measurements taken via tape measure) but the bar ends didn't.

As to the wheel alignment method... it's a bit ghetto but it works, albeit with a serious drawback. The string is placed in line with the rear tyre's edges, then the offset from the front tyres edges is measured via a ruler and the front wheel turned in-line with that. The bar ends positions can then be checked against a datum point on the frame (these datums have to be symmetrical about the centerline of course). It greatly helps to be able to sight down the line of each string from vertically overhead, which I couldn't quite do here. The contact against the rear tyre forward edge has to be bang on, just touching and no more. Previously I've just squinted and sight-lined against tyre edges, which is fuzzy at best. The string is an improvement over this.

The drawback is that there's no clamping force on the front wheel whatever. If the front end is off the ground, as it has to be with the fork clamp bolts released, any attempt at an adjustment means having to get down to the ground and realign everything before measuring whether or not the alignment worked. There's a lot of up-and-down. It's OK for one bike but if I had to do this several times I'd probably look into making a couple of full length beams with offsets (for the front tyre) and through bolts for clamping.

Anyway... a very long day to find out that one component is bent and needs replacing. I'm not keen to just bend it back. I have no idea how much of this kind of abuse the cast alloy bar can take and I'd really rather not have the thing come off in my hand after hitting a decent bump in the road.

While the front end was up, I took the time to check front and rear wheel bearings, head bearings, and swingarm bearings. Everything came in fine. That's just push-pull while sitting on the garage floor of course.

Some handling issues may be down to wear on the tyres of course - I replaced both front and rear just after purchasing the bike, meaning that the Bridgestone T30 Evo's have now done approx. 9,000 km.

[AllanB]

The throttle side may be slightly longer than the clutch due to the extra hardware for the throttle where there is only a one switch block on the left. Which would give a different measurement.

[OddDuck]

The throttle side may be slightly longer than the clutch due to the extra hardware for the throttle where there is only a one switch block on the left. Which would give a different measurement.

Thanks Allan but I'm not measuring length here, I'm looking at whether things are in line and straight or not. Hopefully the photo shows that the grip isn't in line with the ruler's edges. The left hand grip was straight and true by the same test, in comparison.

[OddDuck]

The gearchange is still a bit notchy so I've been doing some investigation.

There's a very good writeup of dry clutch behaviour here:

http://www.desmotimes.com/clutch.pdf

Snyder only lightly touches on clutch throw, taking the view that if there's enough to separate the stack then it's OK. I think it's a bit more important than that, if you don't have enough throw then you will have some problems:

1) gearchanges are going to go through with a bang, thus shock loads on dogs, gearbox bearings, gears themselves, the crankshaft etc
2) finding neutral can be difficult or even impossible
3) the clutch itself will get thrashed in relatively short order.

If the clutch isn't releasing properly, there's still torque being coupled across it when a different gear ratio (up or down) is selected. That means a very abrupt speed change, thus a rotational shock load across basket, plates and hub. Basically the tangs will get hammered forward or backward on every gear change, and so will the splines of the hub. Mushroomed tangs, dished baskets, and notched hubs follow, leading to problems with the plates not sliding axially any more.

Previously I've checked for clutch disengagement by centerstanding the bike, and with the engine off, turning the rear wheel by hand with the bike in first gear. Pull clutch in, overcome initial stickyness, presto the rear wheel turns freely and there's some very quiet rattle from the clutch, indicating motion inside the housing.

If the clutch cover is off then it's also possible to watch the plate stack move. The result here was disappointing: the plate stack didn't follow the pressure plate. Most of the travel wasn't used, with a clear 1mm gap left between pressure plate and the topmost plate in the stack. The clutch still seemed to release OK though, with light torque between rear wheel and engine. Unfortunately this isn't a test that matches running conditions, since the basket isn't turning.

This time around I tested by doing something dangerous: I took the cover off, centerstanded the bike, put it into the highest gear, and ran the test with the engine running and the basket spinning under power. This allowed me to watch what the clutch did in its natural habitat, so to speak.

I was very careful to keep the rear brake on during this test, keeping the rear wheel stationary at all times. The clutch was never fully released, I'd let it bite lightly at the most to test the action, but generally kept the lever pulled all the way in. Previous experience of allowing the rear wheel to spin up had been that the chain started whipping and the bike started jumping on the stand.

Having the bike jump off the stand, while in gear, engine running, with a moving parts cover off... hmm. Not good.

I was also very careful about things like shoelaces (just waiting to get sucked into that clutch, if actuating the rear brake by foot), wristwatches, loose sleeves etc. Anyone with long hair: tie it back and wear a beanie or similar as well, it's really tempting to get in close for a better look and that spinning clutch is just waiting to grab and yank anything it can.

Like I said, this is dangerous. All it would take is one second's worth of moving before thinking.

Anyway, the result: the clutch behaviour is very different when spinning. The plate stack separates beautifully and follows the pressure plate. Torque transmission seems different to stationary as well, it's still transmitting torque even though the stack has clearly opened up. I couldn't test for this without risking spinning the back wheel up, though.

After shutdown (and pulling the ignition key) I ran a few tests with the clutch cover off. Throw was measured via the vernier caliper between the spring cup and the pressure plate, for the following situations:

Engine cold, full spring set: 1.62 mm

Engine cold, half spring set: 1.75 mm

Engine hot, full spring set: 1.66 mm (this last was done after a half hour ride, I didn't just warm up in the garage)

I was checking for squash on the hydraulic lines. Hot vs cold should be obvious enough - is the clutch fluid suffering local boiling - but the half spring set may want explaining.

Measuring at full spring, then half, allows extrapolation to zero return force - ie a measure of clutch travel if the hydraulic system was perfect. If I cleaned and bled the system perfectly, this travel would be what I could expect.

Full: 1.62 mm
Half: 1.75 mm, ie a difference of 0.13 mm
Zero: 1.75mm + 0.13mm = 1.88 mm, or thereabouts. A little bit of angular deviation on the vernier during measurement would knock the precision about so there's a fair bit of uncertainty on these numbers.

It's a lot easier to run this measurement and calculation than it is to spend a few hours stripping everything, methodically going through the clean up / bleed procedure. The reason I did this was because I suspected that there simply wouldn't be enough clutch throw anyway.

There are 7 friction plates interleaved and bookended with 9 steel plates, in the stack. That's 14 friction surfaces used to transmit torque. On a 1.62 mm throw, that's a separation between plates of just 0.116 mm.

That's not much of a tolerance band for plates which aren't perfectly flat. If they're even slightly dished, bowed, warped etc, then there'll be an area of contact somewhere. If there's any force on that area of contact, the clutch will transmit some torque. The only time clean disengagement will happen is when the plates have zero force connecting them together, and for that to happen, there has to be enough throw.

I'd like to have a perfectly flat plate stack. They might have been that way going in. A few months of use, some stop and go traffic, keeping the bike in gear at the lights, and the occasional launch off the line at the green and I don't really think there's much chance that the plates will stay flat. Steel warps when it gets hot. I already know that there isn't enough throw, with this setup. Provisionally I'd guess that around 2.5 to 3.0 mm of throw would be needed to sort out the issue.

This all brought me to the next question: what's the relationship between clutch master cylinder and slave? Which component would I have to change, to get this amount of throw, if simply bleeding the fluid won't do it?

The whole system is effectively just a lever. Motion at the grip is converted into motion at the clutch, high motion and low force becomes low motion, high force. The relation between the two is governed by volume moved at the master being the same as volume moved at the slave, so:

Piston Area x Travel (master cylinder) = Piston Area x Travel (slave)

Ducati clutches are notorious for heavy actions, hence the thriving aftermarket in light-effort clutch slave cylinders. It's not just the lighter clutch lever effort. For whatever reason, Ducati OEM clutch slave cylinders have been of low quality for a surprisingly long time. They're notorious for blowing seals at relatively short mileages. The latest generation might (finally) be OK in terms of durability, but if an owner wanted reliability as well as reduced lever effort, the aftermarket was the way to go.

The ST2 has one of these, fitted by a previous owner. From my reading, the OEM piston is 28 mm, most aftermarket clutch slaves are 30 mm, trading reduced effort against slightly reduced travel. I've got an Oberon slave on the 900SS and have had no issues with disengagement on that bike (the Oberon's supposed to be 29mm btw).

The ST2 shouldn't have this issue, then, given that the 900SS is running basically the same setup and hasn't had any issues with clutch drag. Unless the master cylinder is different, of course...

Apparently Ducati addressed the high-effort clutch situation in the 2000 model year, by reducing the clutch master cylinder piston diameter from 13mm to 12mm. The ST2 is a 2000 year bike. It helped, but people would still fit aftermarket clutch slaves, as has been done on this bike. Right. I've taken a master cylinder apart once before (curiosity got the better of me) and it's not entirely trivial to do, or to put back together again. So: is it possible to confirm that it's a PS12 quickly and easily?

I used the verniers to estimate travel at the master cylinder, approximately 12mm, give or take. I guessed that the aftermarket slave is 30mm diameter, then ran numbers via the volume in equals volume out relation for perfect travel of 1.85mm, to get 11.77 mm diameter for the master cylinder piston. If working backwards from a PS13, the travel on a 30mm slave comes in at 2.25 mm, which is a lot more like what I want.

It sounds like a lot of fuss and bother but I really want my clutches to last as long as possible. It's not difficult to drop over a grand on these if everything has to be replaced in one go. It'd also be really nice to have a gearbox that changes with a snick instead of a clank.

A couple of other things found out during this, as covered in the photos...

It's possible, with a worn clutch, to have the friction plates start binding on the hub. Once they get loose enough in the basket, they'll start riding on the outside of the hub splines. The poor-man's fix is to take some 240 grit paper and bevel the inside diameter of the plates, taking the edges off the friction material (these were razor sharp 90 degree edges on mine) and the steel underneath. Smooth travel, without binding the plate stack up, follows.

Anodised hubs work brilliantly at first but can have issues if the anodising flakes off with wear or impact - this can catch the steels. This happened on one of my hubs and I had issues with slipping under power and / or clutch drag during changes. The field fix was to shut the engine down and work the clutch, while rocking the bike back and forth with the bike still in gear, to allow plates to unstick and settle back to where they needed to be. Properly chamfering the steel's splines sorted the issue.

Anyway, I've been through much fun and games to find an answer for this particular problem on the ST2. Shorty levers offer a cheap solution - shorter lever means more angular swing, which means more master cylinder piston movement - but somehow I think they'd just not quite work on the ST2, both in terms of style and function. At this point the best answer (definitely not the cheapest) is to replace both the clutch master and slave, a PS13 master and an Oberon 29mm slave, in order to get enough travel. Just the clutch master would possibly be enough. It'll be a while, I'm still clearing a credit card, and the current plate set and basket want replacement soon too.

To my mind, it's false economy to not do it. The bike will simply keep on eating plate sets if I don't.

This will result in a slightly heavier clutch pull, as well. There's just no getting around this:

The clutch has to transmit full engine torque, this means a certain amount of spring force;
To disengage the clutch, the spring force has to be overcome;
To separate the plates cleanly, there has to be enough travel;
Travel trades off directly with clutch pull force.

Springs / force / travel. They're all interlinked. Any change in one affects one or both of the others.

Anything more requires improving the basic clutch design, which when it comes right down to it, is limited by the human hand. Higher coefficient of friction materials would mean lighter springs, for example. A flatter plate set would help. There are already things like ramp clutches available (these clamp tighter under drive load, loosen up under back load, and have non-linear release force) but they're pricey. Maybe at some point we'll see power assist clutches, similar to vacuum brake boosters on cars. There are quickshift blippers which already avoid the clutch issue completely, of course.

[OddDuck]

Right, enough theory. Being curious I took a couple of wet winter evenings and took the clutch slave cylinder apart, on the idea that maybe it had sludge inside and cleaning this out would help.

First things first: whoops, I'd made an assumption and now it's time for some humble pie. It's definitely an OEM Ducati slave, 28mm bore, nicely made in the pattern of an aftermarket cylinder but nothing fancy. The total lack of a conspicuous brand name should have tipped me off. It's quite different to the previous Ducati slaves I've seen. It wasn't sludged up, although a few specks of black stuff came out with the fluid.

Instead, I think I found a design flaw. The piston seal has a 1 x 1 mm groove in it, the seal itself runs in a 0.25 mm clearance fore and aft, and the piston-bore clearance is around 0.10 mm.

There's pressure across this from operation, of course, but that's in line with the piston. There isn't pressure from top to bottom for any air caught in the seal groove or length clearance, apart from boyancy. Anything caught behind the seal would have to get out past the piston skirt, that'd mean the bubble would have to break apart into micro-bubbles around 0.05 mm or so, and fluid surface tension would fight that. If the fluid surface tension is stronger than the boyancy, the air will never escape.

I ran the numbers for lost volume due to compression of air in this space and got a match on the travel I've been getting versus the travel that should be happening. Right. Time to open up the rear of the seal so that any air would escape naturally.

The piston came straight out with a few lever squeezes, it wasn't hard to remove. The orange stuff inside the bore is presumably what happens with old brake fluid.

[OddDuck]

The slave had been crushed inward a bit where the chain had run on the outside - this was relatively straightforward to open up again via Dremel and then fine paper. There isn't really much I can do about the scoring on the cylinder bore or the piston skirt aside from smoothing the worst of it down, but this hasn't happened in an area which the seal normally sweeps.

The modification to the piston itself was done with a Dremel and small diameter cylindrical grinding wheel. All I wanted to do was to open up a path to or from the back face of the seal and the seal groove. I wanted the path to be a minimum of 1 mm wide, to allow bubbles free passage out. I was very careful to not cut into the front face or inner diameter of the piston seal's groove since any damage will allow a leakage path.

There are a few other ways that the same thing could be done - lathe the back diameter down, then taper the back face of the groove, or use a drill press to put a ring of access holes in from the back - this isn't the only way to solve this problem. I was careful to take sharp edges off after grinding, I don't want to cut into that seal on reassembly.

Reassemble, refill, bleed the system, test the travel and presto, limited success. Travel has increased from 1.6 ish mm out to just under 2.0 mm, similar to my calculations but not quite matching. Possibly I still have air trapped in the system. A test ride will show if this has worked in terms of getting the clutch to release properly, or if it's time to spring for a PS13 master after all.

[AllanB]

The clutch slave appears to be a common bitch. Abused by chain crud they wear out.

The aftermarket has it covered and they tend to use larger pistons as well to reduce the clutch pull. Expensive little buggers though.

Reminds me must be time to pull the front sprocket cover and clean behind it.

[OddDuck]

Cash being a bit tight at the moment, I did some reading and thought I'd give something else a try: clean up the steels. Apparently they can get glazed and a bit sticky after high mileage.

The photos show the setup: I simply used methylated spirits, wet'n'dry paper (600 grit) and a reasonably flat metal plate. Most of the steels were the thick 2mm type and didn't need a dedicated hold-flat tool, working by hand was enough. It was certainly possible to feel the dark stain on the surface come off with the change to bright steel underneath.

BTW everything I read about other people trying this said: use a non-oily working fluid for the cleaning / scouring / sanding. Brake cleaner, meths, electrical contact cleaner etc etc were all fine... just as long as it didn't leave an oily residue or cause the steels to start rusting. The sanding can be done with a Scotchbrite pad, an orbital sander, wet'n'dry and so on, it really doesn't seem to matter as long as it doesn't carve the steel itself up too much. Caution was needed if trying this on the frictions though, quite a few solvents will start dissolving the pads.

The photo of me holding one side of the disc down, and the visible lift on the other side, is a basic test for flatness. This one isn't flat, it was dished slightly. I didn't have enough of a set of spares to replace it, so have put it back in for now. This sort of thing won't help separation in the disc stack - any discs with a ripple or dish in them will act like springs. As the stack expands when the clutch is pulled in, they'll expand outwards and maintain contact with the friction plates. A similar flatness test is possible with the frictions as well, provided that the tangs haven't mushroomed too far.

I've taken the bike for a couple of quick about town rides to try this out and see how well it's worked. The short answer is: it helps. It hasn't completely fixed the clutch disengagement issue, but it has helped.

As to why it helps... based on my reading, and what I've seen while working with the discs, it looks like the steels end up getting a very thin coating applied from the friction pad material. If the coating then sticks or bonds back onto the friction pad material, it follows that the clutch wouldn't disengage properly. It couldn't, not if the discs were gummed together.

Why would the coating on the steels get sticky? I think that somehow it got old. Oxygen? Maybe. Heat, and subsequent chemical changes? Possible. Best guess, and it is a guess, is that the coating became impregnated with oil. Well. Not sure.

About all I can say is that it looks like the steels benefit from being cleaned up every once in a while. Normally they'd get replaced with a clutch pack change, but having gone to the effort of properly chamfering the splines I'd quite like to keep them going as long as possible.

[pete376403]

something to be said for a wet (oiled) clutch. The oil will help to move the crud away from the places, prevent corrosion, make things a little easier where the steel places bear on the hub, keep the noise level down, add a tiny measure of shock damping on engagement. So what's so great about a dry clutch again? Racer rep? Anything else?

[OddDuck]

something to be said for a wet (oiled) clutch. The oil will help to move the crud away from the places, prevent corrosion, make things a little easier where the steel places bear on the hub, keep the noise level down, add a tiny measure of shock damping on engagement. So what's so great about a dry clutch again? Racer rep? Anything else?

Fair comment. Two things:

1) they're relatively easy to get into and service, and
2) you don't get friction material through the engine oil and thus the bearings and gears.

oh and 3) apparently playing hot potato with a wet clutch plate set trackside is not fun.

I'm going to thrown in my own as well..
4) it's possible to get full disengagement on a dry clutch in a way that isn't possible on a wet clutch, there's just too much fluid drag between the plates for any practical clutch rod throw. At least based on borrowing my mate's RF900R, which always goes into first with a hell of a clunk despite clearly having sufficient movement in the pressure plate.

But yes there certainly seems to be a lot of dry clutch fetish going on among Ducatisti, personally I think the whole open clutch cover, spend gazillions on clutch bling thing is a bit silly.