Winter Layup - 1995 Ducati 900 Supersport

[OddDuck]

Worn guides can be internally knurled and then reamed back to size so you stay with the standard valve stem diameter. It COULD increase oil flow down the guide as there is now a spiral track (knurling is a bit like a screw thread but it doesn't cut metal out) but if the guide seals are good it should be right.

Been doing a bit of reading, not sure about this. The spiral groove could flow way too much oil if the guide seal's not very good, and I'm now certain that the all-rubber Ducati OEM seals are simply crap. The taper onto the valve stem is probably OK but the body of the seal relies on rubber tension over the valve guide boss. I think it flexes under valve motion, drawing oil up like a pump. I've had problems with oil-fouled spark plugs since I bought the bike and I want this issue sorted out.

[BMWST?]

this is bloody interesting.
I like the knowing of how things are put together,not really interesrted in doing the actual doing though!
The ash you talk about.Is it just the residue from very hot oil,or is it actually some residue from combustion or the "smoke" from hot oil inside the engine.Presumably the "smoke" is burnt in the engine via the crankcase ventilation systen which is plumbed into the airbox?

[OddDuck]

this is bloody interesting.
I like the knowing of how things are put together,not really interesrted in doing the actual doing though!
The ash you talk about.Is it just the residue from very hot oil,or is it actually some residue from combustion or the "smoke" from hot oil inside the engine.Presumably the "smoke" is burnt in the engine via the crankcase ventilation systen which is plumbed into the airbox?

Ash: combustion residue, mostly. Burnt and carbonised oil is a small contributor, hopefully, but most of it is fallout from combustion of petrol and air. I think it gets into the crankcase primarily as blow-by of hot gases making it past the piston rings. There's particulate drop-out, condensation etc.

Smoking oil and the main part of the blowby gases are vented to the airbox, for cleanup via combustion. I reckon the crankcase vent valve is one of those unsung genius bits of technology. So simple and it's a win for everyone. Cleans up smog, better economy, motor has a chance of being oiltight, and it helps power too.

[Voltaire]

I've had my last two BMW racebike engines guides fitted with K liners. They drill out the guide and fit a sleeve which then has a tool pulled thru it to force it onto the guide. Advantage is does not disturb the guide.

[OddDuck]

I've had my last two BMW racebike engines guides fitted with K liners. They drill out the guide and fit a sleeve which then has a tool pulled thru it to force it onto the guide. Advantage is does not disturb the guide.

Who did the work, and how much did it cost?

Just having a look at the K Line page now. I'd prefer to have a go at it myself but the pricing on the tool kits needed is a bit out of my reach - $566 US - and it's out of stock at the moment anyway.

[Voltaire]

A friend of mine who used to do engine work had a set.
He let me have a go and its pretty idiot proof.
Did a Google and this place near you came up/
http://www.trademe.co.nz/services/tr...-816372118.htm

Process is:
Drill out the guide with a hand held electric drill- the tool is self aligning.
fit the sleeve.
Hammer a broach tool thru to press it into the guide
ream



He did a 3 angle valve job too and bike puts out a healthy 84 BHP

[OddDuck]

Thanks. I've used Kerry Lindsay before for a recondition on a car cylinder head, he's good... will have to get the rocker arms and camshafts out of the heads before going to see him though. It's not as convenient as the heads in your photos, since they seem to cleanly split rocker mechanism and valves.

In the meantime I'm still toying with the idea of trying to do this myself - keen to learn and there's nothing to get focused like getting hands-on.

Progress: got the swingarm and front subframe back from the powdercoaters today. Haven't unwrapped yet.

Got further into the now-split engine and had the first look at bearings. Most seem OK in terms of play but the crankshaft mains are definitely shot - there's (by eye) about 0.5mm free play on the LH bearing if I try to push-pull the inner roller axially at any point on the circumference. So it's probably appropriate to change the lot.

The plug screw for the mesh screen turned out to be made from straight aluminium, screwed into aluminium. It's a stripped thread waiting to happen. Don't mess with this if you don't have to, I want to repaint the casings so I wanted it out... if you're unscrewing this and it gets stiff, screw it back in. Work it back and forth until it smooths up. Otherwise if the threads stick, that's it, it'll ruin either the screw or tear the thread in the casing itself up.

I found bits of the main gasket sitting in the sump, too. They were all ovalled or kidney-bean shaped - looks like they've been tumbling in the oil and getting rounded off. Quite a bit of this gasket sits in a way from the casing wall, it's not hard to see how it gets brittle and breaks off inside the engine.

I had the first shot at cleaning gasket surfaces up. I tried with a knife until I found a rolled edge, then it was block and 400-grit paper with CRC as a cutting fluid (very carefully).

The block was made earlier from a piece of scrap aluminium plate, levelled against increasing grades of wet'n'dry on a proper surface table. The block is nimble, easy to work with, but it really is for localised cleanup, not any serious attempt to plane a flat onto something this big.

[OddDuck]

Gears next. I had the inevitable loose shim turn up - no idea where it came from of course.

Ducati assemble the crankshaft with shims on each end against end float. They use a lot more on the shift drum and at least one of the gear shafts, too. Just to make life more fun, the shims are stock sizes like 1.0 mm, 0.6mm, etc. So if they want an intermediate size, they stack up several shims, in one place. Mislay even one and you have problems.

That said, it might be a blessing in disguise. One of the reasons that I got this far was that the bike's been shocking for refusing to shift properly into 5th (or was it 6th?) Anyway: never does it first time. I have to hammer the shift peg a couple of times, and even then, sometimes it'll pop out again. Did it coming out of a fast corner on the Rimutaka Hill once, did it a few times on my one day at Manfield as well. Not good.

So I'm wondering if someone hasn't mis-assembled it with a shim in the wrong damn place. Not sure yet how I'll check it, I'll have to understand how the gear system is assembled first.

[OddDuck]

Tonight: preparing crankcase halves for bearing removal. I started by pulling the locating dowels.

I did some reading and it turns out that there are at least three methods, all of which keep the outer surface in good nick.

1) use an E-Z Out
2) use a taper (first) tap
3) drill through and fit a suitable pin, turn and pull with that.

I gave the E-Z Out a go and it worked a charm. Sutton No. 5 fits nicely for OD 10 and 14 mm Ducati dowels. Getting the dowel off after was done with careful use of a pair of vice grips. Any ridges left in the dowel outer surface (the first one was a bit stiff to unscrew) sanded out.

This meant I could now block and paper the gasket surfaces that had dowels in the way yesterday. A closer look and I'm thinking the full surface table treatment might be justified. There's been some nasty dents and scratches over the years.

Taking the bearings out is done after heating the entire case half to 100 C, according to the Ducati factory manual. I'll have to heat in the oven upstairs, wrap and box the half to keep it warm, then carry it down to the garage for the bearing pull work. Old oil stinks, I eat food out of that oven, so I've had a go at washing the casings by using turps and a paintbrush. Hasn't been 100% successful with the baked-on golden stuff but it did get a lot of loose grit off.

The bench surface is vinyl flooring, I'll need a decent chunk of chipboard or something similar as a surface to rest the casing half on. The vinyl is normally fantastic to work off, it's tough, cheap, wipes down with meths, and if it's getting carved up then I just sand it down again with 180-grit and more meths to get a new surface. I don't think it'll handle a hot casing though.

The cleanup got dirt out of the way and showed damage to one of the swingarm bearings - not sure what's happened here, will have to strip it and see.

[OddDuck]

Checking the conrod main bearings - trying to get the end caps unscrewed. With the crankshaft assembly on the bench, I thought I'd be clever and avoid any strain on the bearings, or the crankshaft itself, by clamping the conrod shaft in the bench vise and undoing the cap bolts against that.

I then watched the conrod twist a whole lot easier than I expected, the end quite happily rotated by 2 degrees or more. I think I stayed in the elastic zone and the metal simply sprung back, I pray that the conrod isn't bent or twisted, but in the home garage there's no way to be sure apart from replacing it. DO NOT UNDO THE BOLTS THIS WAY.

Oops.

[OddDuck]

Doing it the usual way - use the crankshaft as the counter-torque element and just go for it. Rattlegun on the bench worked.

Something that I learned just a few days ago: impact wrenches really should be driving impact sockets. Using the bright chrome vanadium normal sockets is taking a risk, especially 3/8"s as I'm doing here. Normal sockets apparently can shatter. Impact sockets (usually black-finished) are thicker walled, softer, and tougher, in order to take shock loading. I've been getting away with it so far, but I'll have to go shopping, the sooner the better.

Split conrod bearings checked. It was pretty obvious that they were due for replacement but I thought I'd get the micrometer and transfer gauge onto them anyway. The caps were refitted and torqued to specification for the measurement. As suspected, they're outside specification. Clearance is supposed to be around 24 to 56 micron, these are running between 60 and 70. The photo of the conrod in the vise, with breaker bar, is after I'd done everything and decided to leave end caps loose for bearing shell replacement. For measurement and subsequent loosening the vise was closed across the conrod big end, not the web.

I checked taper and ovality on the main journals as well. The limit for this is 10 micron, for either requirement. In practice that's measured by taking a number of diameters, at various positions front to back, and angles of rotation, making sure that they're all inside that 10 micron spread.

It was OK on one side, it wasn't on the other. That's the side with the oil galley plug which came loose, and also the side with the heavily worn main bearing. The spread was around 14 to 20 microns. I'll probably just run with it.

My workshop manual says that the two main journals also need to be checked for misalignment, by putting the crankshaft between centers and using a pair of dial gauges simultaneously. I can do this with borrowed gear at work (there's a lathe which can be set up with centers), but there's no chance here.

[OddDuck]

Checking the gearshift mechanism, starting with the drum.

The drum slots are supposed to be between 8.00 to 8.09 mm wide, with an 8.19 mm wear limit. Initially I tried with a vernier, but switched to transfer gauge and micrometer. Nothing came within normal service, most of it is only just above the wear limit - 8.165, 8.180, that kind of thing. It actually didn't look that bad so the numbers were a surprise.

The shift fork pins, despite visible wear, came in OK and well above their wear limits. I was careful to keep the micrometer inside the worn area. Clearances of shift fork pins in the drum slots (measured with a feeler gauge) were a little high but within tolerable limits.

There's a fairly generous end float allowed on the shift drum of 0.200 mm. If I read this right, I'll have to measure that inside the cases, with a new gasket fitted or allowing for the paper thickness of the old one.

There are some checks for gear fork width and slot width but I settled for feeler gauge tests. Everything came in fine. All three tested at 0.203 mm, with an allowable limit of 0.400 mm.

So the selection mechanism was probably not the issue. I had a play with the two shaft assemblies and found that all the roller bearings inside the gears seem to be loose. There's a lot of end play as well, haven't measured it yet, but there's way too much for it to be normal.

That got me looking. I noticed a massive amount of end play on the spacer shims at the end of a shaft assembly, then I looked closer. The end cap has been forced open, with a clearance of around 0.63 mm between end shims and the gear. If it was like this inside the engine cases, with allowable end float on these two shaft assemblies specified at 0.10 mm, this might explain the problem.

[OddDuck]

Carrying on with the gears, I spent a few hours working out how to take the chain sprocket shaft assembly apart and methodically measuring and documenting what went where.

The bearing cup mentioned previously turned out to be a very light interference fit, removable almost with bare hands. The bearing puller was total overkill, as I found when I could pull it by twirling the leadscrew between thumb and forefinger.

The snap rings proved tricky. They kept on slipping off the plier jaws. In the end I found that forcing them onto the snap ring pliers from the back of the ring worked. The rest of the assembly was simple enough to take to bits.

Bearing surfaces were checked visually and measured, with micrometer and transfer gauge. Everything came in fine.

I had it almost completely apart before it ocurred to me to check availability of shims, and guess what... it's still possible to get the obvious sizes, but the intermediates aren't made any more. So re-shimming the individual gears on the shaft probably isn't going to happen. There's a lot of free play, but little wear evident - the original planing marks are still visible on each shim face. Ducati's use of straight cut gearing means zero axial force anyway. Shifting aside, end shimming isn't critical in the way it would be with helically-cut gearing.

The bearing cup took very little force to push back on, so it looks like yesterday's theory about this causing massive shim errors for the rest of the assembly is wrong.

[OddDuck]

Next step: abandon hyper-detailed work and just reassemble the cases with the shift mechanism, then have a play.

The main case gasket came in with a thickness of 0.30 mm. Assembling without means zero or -0.10 mm play clearance for the shift drum and lightly clamped clutch and output gear shafts, but things still turned when I tried them.

So I had a play with the shifting mechanism while looking at the assembly through the horizontal cylinder opening. Everything seemed fine, but then I started getting the trouble shifting again - from 4th to 5th or 5th to 6th, either up or down. Nothing seemed wrong inside so I had a look at the shift claw and the shift drum's end pegs.

The shift claw is hitting the crankcase wall (polished metal visible in the last photo) and also failing to turn the shift drum the full 60 degrees needed to make a clean gear shift. It's nearly doing it, sometimes it does make it without problems, but most of the time it's getting nearly there and then stopping.

I pulled the claw mechanism and had a close look. The hook actually seems mostly OK. There's a planed surface, but the original guillotine marks are still visible. The elbow and shoulder bushings are very loose, and the stop peg is hitting part of the spring plate that it shouldn't be, losing a couple of degrees of motion either up or down.

[Ocean1]

I admire your attention to detail, I'd be in too much of a hurry to get the bloody thing back together again and miss things.