Mobius, just as well you're not flying aeroplanes
my 2 cents worth...looking at the clues, there was a lot of hammering on those bolt threads before total matal failure, and that was caused by a loose sprocket...not the 70 or 80 hp from the low performance motor, but I would also question the bolts themselves...to me the bolts are defenitely of inferiour quality or too brittle, you have 2 new spare bolts, give them a torque test and you'll find that they will snap way before other same bolts...as I said thats just my opinion

Thats what I'm thinking G, Also,not that it is a big difference the recess would lessen the contact area of the sprocket to the carrier. All this along with the additional rocking moment of moving the line of force away from the carrier,old rubber could add up to just enough to shear the bolts. May require a step up in bolt diameter or class or both. Just a thought

Dear Mobius

Too me, the failure of your bolts look like dynamic fatigue failures. I have been working as a test engineer working with fatigue testing of screw joints and screws in tensile and bending stresses for years (albeit screws made of titanium, but the surfaces look scaringly alike to yours!)

May I dare ask the following?

1. Is the "special anti-unlock" washers original or your own idea?
2. Most importantly: Did you use the correct torque for the screws (as specified by the Ducati workshop handbook or alike or for this screw dimension)?

What is extremely important for screw joints is the "preload" or "static tension" of the screw. A too low preload will lead to a high ratio between the unloaded situation and the loaded situation, and this is very aggressive for the fatigue life of a screw system sustained to dynamic tensile stresses. You would like to have a as small ratio as possible.

Thus it could be the case that you simply have used too low torque when mounting the screws (?)


Further information on the subject:
Applying grease when not supposed too, or not using grease when supposed to might danger a construction. Normally I whould say that a little grease is good on the threads - or used under the head, but this will effect (lower the friction) the amount of torque that is transfered to tensile stress of the screw. So one should use with caution. Especially for systems with tensile dynamic loading.


Normally I would say that the screws in this example should NOT be exposed to high dynamic fatigue. But using an altered driveline or similar... maybe...? I do not have your situation fully clear in my head...

What however appears to me is that this failure is most likely NOT shear related, and I bet my cents on dynamic fatigue.

/P

I agree that this looks to be a dynamic failure.Maybe the grease is the culprit maybe not
To me I think it is moving the chain/sprocket in relation to the carrier that is the prime suspect,all the rest contributing to the failure
The solution to me would be to weld the old recess' up machine new ones on the now outboard side of the sprocket and larger diameter bolts. Then either safety wire or loctite the new bolts All this also assumes your front and rear sprockets are running true in relation to each other.

It`s definetely a combination of things. Bigger bolts won`t be possible unless you completely modify the carrier to fit larger nuts. I don´t think that is necessary.
I still don`t like the fact that the holes in the sprocket are so much bigger than the bolts. If that is normal the only thing that could minimize slip would be that (as mentioned) the recesses for the sprocket bolts are a perfect match in diameter to the bolt heads meaning they have minimal tolerance. If then flipping the sprocket you`d lose that and the won tolerance would start "working" on bolts, sprocket and carrier.

G.

PS: excellent input /P

Mobius,

Cool, I'll take a look at the fracture face photos closely tomorrow, but I already think I saw some tell take beach marks that seem to point to fatigue along with the final overload area. I only glanced so should not be so conclusive already.

Anyway, cheers,

Mick

You are not alone . . . thats why we have this support group

Everyones having a go IMHO Top pic shows my rear sprocket assy and the bolt. I think you should be using a bolt with a bit of shank on it rather than threaded full lenght . I think the last thread can weaken the bolt just where yours have broken . The marks on your sprocket would suggest to me that the bolts have been loose ,those are wear marks , unless you did them up thight enough for the sprocket material to be deformed by the serrated washer ( I think not ) .I think that those wedge type washers are as likely to prevent you from properly tightening those bolts as they are to stop them loosening.
Bottom pic shows ( hopefully ) the hardened steel washers I used to pack out my rear sprocket , over 10,000k's no problems .Reversing your sprocket would leave the recessess intended for the bolt head against the very surface your trying to attach the sprocket to , at its most critical point ... under the bolt head . As you can see my washers are under the bolt head , and are the only point of contact with the carrier , but the sprocket is ( time proven )sufficently well mounted.. I moved the sprocket just to improve alligment & it is still " spiggoted " onto the carrier. washers are 1.5 mm thick

Wow people - I feel guilty with all the posts - and time taken. Thanks so much. I will try to address all questions asked.

Well, I got the wheel off (Remarkably easily in the end) and I am ecstatic to report that the drive rubber is in perfect condition. Phew. No grease was on the rubber nor in the enclosure for it.

The sprocket carrier is in decent shape too...

However, I am an idiot for forgetting I instructed the Ducati service guy to flip the sprocket to gain the extra 2mm clearance for the D250. Because... this morning I went out to see my extremely experienced Motorcycle engineer Max. It seems I made several basic mistakes in this small modification.

1) I was not aware of the countersunk recess for the sprocket bolts.
2) I was not aware of the type of bolt which should be used there.
3) I did not use Locktite.
4) I did not measure the torque I used to tighten the sprocket bolts

Here's what Max had to say, not necessarily in order.

1) Why are you using a cap bolt with no shank?

Bolts are prone to failure at the point where the thread begins, especially when torque is applied directly to that area. (And the thickness of my washers meant the thread started at exactly the outboard surface of the sprocket. This style of bolt was what Duane's Bikes installed when I asked him to flip the sprocket, so when I replaced them I thought I was doing the right thing by getting identical items. Ugh.)

2) Why are you using a high tensile bolt?

If it were my bike, I'd be inclined to use a non-high tensile bolt as by the evidence on the threads of the snapped items, there is significant work-hardening occurring at the point where the thread/wear starts, further enhancing the likelihood of a failure.

3) Why did you not take up any slack in the connection and then tighten them to the correct torque?

That can really affect the connection strength and endurance of it. (I'm an idiot who doesn't own a torque wrench, never torqued anything in my life, and just assumed that "as hard as I can do them up by hand" would be sufficient.)

4) The recess in the flipped sprocket concentrates the torque in half the space, higher up the bolt, with a gap between the point loading and the sprocket carrier.



(I had already thought of this, and had even thought to fill the recesses and redrill the holes...)

Max is smarter than me: So, we'll just drive 5 new holes for you...



So, the solution seems to be:



Max said: "it'll be all right now, I reckon."

What do you think?

An interesting question has been raised: was it my idea to use the never-undo washers? Err, yes. Yes it was. Because the original bolts came undone. I thought that made some kind of sense...

.. but maybe not!

I am now wondering if I should use any washers at all, given that I'll be using a lock nut + locktite.

Oh, finally, how many pounds of torque SHOULD I use on the sprocket bolts? I purchased a torque wrench today.

--------------------------

Claude, (My Grandfather's name!) thanks. I think you are correct due to the issues mentioned above. It would seem my attempt was pure fail.

I've only flown planes four times, but I have completed over 5,000 paragliding flights and once held a distance record here. No solid parts in those bad boys though, except carabiners and the like.

Here's what I would do,

1) Turn the sprocket back the right way and get a proper spacer made to go between the sprocket & carrier.
2) Check the alignment between front & back sprockets. Anything over 1mm needs attention.
2) Get the right size high tensile cap head bolts that have a shank(the original bolts were HT). If you like, copy Gerhard's idea and drill the heads and thread wire through them.
3) Loose the washers.
4) Put loctite on the threads to stop them from undoing.
5) Use your new torque wrench to correctly tighten the bolts.
Put the bike back together, go for a decent ride and enjoy.

How frustrating.. my thunder was stolen at the very end - I patiently and politely read through the whole post rather than jump to the end in case someone else had 'got it' although it was clear that there was a lot of 'barking up wrong trees' going on.. The two threads above by Mctool and Mobius are the most relevant by far* (IMVHO,WTFDIK?). The answer was crying out to be said and Angus was the first to draw attention to it :- Of course the sprocket bolts are in shear, Of course you should not be using 'screws' to attach the sprocket and of course the dimension of the hole in the sprocket is critical.. I haven't bothered to mention the other fundamentals of bolt quality and tightening torque. Aftermarket sprockets manufacturers are often 'sloppy' when it comes to fixing hole dimensions so that gets things off to a bad start. Next, the fixing 'bolts' should have a plain shank that extends as far as possible until thread is required. I (personally) would use a thick plain washer, not a hardened lock washer type, and I would (and have happily used) 12 grade cap-head bolts and Loctite. Torquing up to the recommended figure on a dry thread is adequate, the bolts are in shear and it's all down to how much clearance the bolt shank has as it passes through all the flanges present - i.e., there shouldn't be too much clearance - a slide fit would be perfect but it's all about mass production and holding a tolerance from the sprocket maker's point of view - any sprocket flange that isn't full width at the point where the fixing bolts go through should be avoided. You can see from the wear on the thread faces just below the shear line on your failed screws, that they have been walking around for some time - it's all about surface area here, precious little of it in this case and so the thread edges have been chewing away on an already 'over-clearanced' sprocket fixing hole and making things even worse every time you accelerate and lift-off/brake.
Glad you (finally) got to the bottom of it anyway.

*Haven't read any subsequent, and haven't time to edit accordingly.

I dont agree that a full circle sprocket is needed. I think that if said sprocket ( thats a bugger of a word to type.. letters all over the place ) is properly supported where it mounts it will be ok.
Just as an aside , someone made a comment re the quality of some after market sprockets, I was fitting a new chain and sprockets to a young lady's bike recently and I had to send her back to the bike shop to get a rear sprocket with less than the 1.5mm radial runout that the 1st new one had

I believe that G, Higgy and Giocast all have good inputs here (sorry for using wrong word Giocast: I meant bolt, but since I am not speaking english since birth, mayby I can be forgiven for writing screw? )

That the bolts are in shear tension is obvious. I was initially thinking of pure shear failure, like cutting paper with scissors, and the failure surfaces do not look like I would expect them to look when "cut". The failure surfaces rather look like dynamic fatigue, showing striations(?) and all.

If the load situation for the bolt is more tensile than shear is impossible for me to say without maybe a FEM calculation? But the flipped sprocket with the "inner hole" from the bolt head recess suggest that at least a part bending and tensile stress should be expected. But, OK, I can agree that shear loading is important, and perhaps the largest part. Still, the failure is not likely a "load to failure" static shear failure ("cut"). Even if one bolt fail after each other, you would not notice until the last one fail, right? (mind you: could be very very short time between the last failures)

OK, let's leave that booring theoretical discussion. Static vs. dynamic.


Regarding your question, Mobius, on the "never undo washer". I would not use them for this situation. Proper torquing and using the correct bolts, and too be sure, locktite, is what I recommend. But proper torque for the bolt is key!

Good luck

/P