TS TIME BOMB

[MD]

If this problem hasn't visited your 8v twin spark engine yet, it's just question of time. It's killed two of my engines. The photos explain the problem and the solution. Take heed.

[Duk]

That's some bad design right there.
But I would stay away from welding. Chances are the sprocket is cast iron and you're more than likely going to warp/bend the shaft.
Having a circlip groove machined into the shaft and using a circlip to keep the key in place is 1 possible solution.
Maybe have the keyway deepened so a taller key is captive within the shaft.
Or a radiused bottom keyway could be cut and a woodruff key used.

[MD]

Hey Dukster. It's been an eternity.
Just to clarify, the TS version has no key-way whatsoever or other captive method of fixing the sprockets onto the shaft other than an interference fit. That is the entire problem.
Other than that, I am not aware that there is any sideways migration of the sprockets along their mutual "axle" and so retaining circlips are not necessary if that is what I am interpreting from your reply.
I believe the sprockets are steel. A couple of small welds would lock them together if you are racing farmer on the go but if you want a more elegant solution go with the Dukster 

[vin sharp]

Er, yes a poor design idea particularly when they previously keyed them on all models for 30 years. In theory the interference fit should do the job, but that assumes all manufacturing diameters & bores are correct fit & a large hole/small shaft combo don't come together...
I've not personally seen one come loose yet, but have pinned a couple between the pairs just in case.
The sprockets & shaft appear to be steel & I've used a 4mm dia hardened steel pin to key it in place. This should be sufficient to stop it potentially working loose on the shaft. In both cases I did a test press (in the removal direction) to check that the pair were at least secured well before even bothering to attempt pinning. DON'T block off the oil supply thru the centre of the shaft when pinning!
PS, T/Spark are closer on the chains centre-line spacing, so you can't just toss in in a Nord one unless you can press it apart & machine a face to narrow it.

[GTV-074]

So whats the best preventative measure to take, or do we just wait until it goes 'boom'?

I am guessing its an engine out job?

I just purchased a TS recently (had one years ago) but haven't got around to doing anything to it yet, so just asking...

Cheers,

Paul.

[MD]

Well the purists would not like it but here's what I would other than the obvious which is to pull the engine down and take the opportunity to fully refresh it to trade standards

Remove battery connection
1 Radiator out.
2 Alternator off
3 Front pulley off
4 Cam cover off
5 Disconnect all hoses to water pump
Undo the two side bolts under the engine head
Remove all peripheral nuts holding on the front cover.
(the fun part starts here)
Remove front timing chain cover. Sump and head gaskets will get damaged in the process around the front cover portions.Don't worry about it. Substitute with a gasket in a tube on reassembly.
Once the sprockets are exposed, clean the contact area between them with acetone (brake cleaner).
Make a small but solid weld between the sprockets at the 9 & 3 o'clock position. ( Be ready to get wiped off Vin's Christmas list for the weld job) 
A full weld is not required.

Refit in reverse order and connect the battery.
Personally, I would take the engine out and fix it sweet throughout. Depends how deep your pockets are.

Failing that, do nothing. Nothing may happen, then again do you want to wonder if it will and when?

[vin sharp]

I would be afraid welding in-situ might deposit a stray weld-spatter ball in the timing chain or elsewhere that eventually gets into & stuffs the
oil pump or worse...
Personally I'm not about to break open my T/S because of worry. I've never actually seen one do it (yet). Perhaps you're just extremely lucky to have had 2 of the rare ones MD! 

[MD]

Yes Vin, I must say darn unlucky. Having said that, my European contacts have experienced the same.

Good caution regarding the weld spatter however to get you home from Julia Creek in a dust storm ( ) a bunch of wet rags wrapped around the chains and blanking off the sump openings would get the job done. Any port in a storm...

Vin. Just doing a bit of lateral thinking, would a roll pin inserted in parallel with the shaft be a reasonable option as it has no chance of penetrating the oil gallery in the shaft?

[Colin Edwards]

Well not exactly a poor design if the repeatability of the respective bore and shaft dimensions can be guaranteed
If the machining acccuracy improved in the later production runs (very likely), doing away with keys and keyways makes sense. 
Lower manufacturing costs due to less machining stages would be developed by the manufacturer.
Not introducing keyways into shafts and bores will reduce the likelyhood of stress cracking.
IF the respective surfaces are consistent enough, very high elastic averaging will result and contribute to a very reliable joint.
No idea what interference Alfa or their parts supplier chose, however 0.1% < 0.15% would by likely.

Could it be these engines that have "failed" have been "tuned"?
Could it be increased loading on this joint due to camshaft design, valve mass and spring rate have increased valve train loads excessively?
Have these engines been over-revved?
Have these engines been over-heated?

I'm with Vin on this one.
Why pull down a perfectly good engine and interfere with this joint when the likelihood of this particularl failure is statistically very low?

[105gta]

Totally agree Colin, I suggested on the 'other forum' to this situation valve train loading but I don't think I was met with much reception.

[vin sharp]

I have heard of it before on a stock road car, (but never seen personally). 
In theory all should be good, but obviously the later tolerances ARE NOT GOOD ENOUGH & are probably the ones most in need of keyway/keys just in case. Even if over-revs, and or modifications like stiffer valve springs have been made, the things should never be close enough to design margin cause a problem. I have never ever heard of a keyed Nord one even loosening.
The ONLY Alfa cranks I have ever encountered with manufacturing faults ( multi-degrees of out of phase on big ends) is on T/Spark 75s.
The only Alfa conrods I have ever seen fail in the centre of the beam (without any bearing seizures) are T/Spark 75.
The only out of parallel (.006" +) liner-seat diameter to top deck I've ever seen is on T/Spark 75.....me thinks there was a period (FIAT takeover period?) where the quality control was not effective. Not the period to decide to leave a keyway out for the sake of cents...
However, 35 years later & most are probably still in service somewhere without too much problem. Just key the sprockets when you have one apart for peace of mind.

[GTV-074]

Great discussion Gents.....

Its interesting, the 75 TS seems to clock up some big mileages 2-300K with a seemingly low number of issues.

Whilst looking for mine over the last 12-18 months, the high K's on them seem testament to their robustness but
as Vin notes they do have the odd failure.

MD, any 'pattern' in the failures, like build year etc.?

Cheers.

[MD]

Well dismiss the loaded valve train straight up. Standard, valves and original springs.
High revs? No more and no less than a 116 or 105 engine would normally handle.
Ditto for both engines.
It's a bitter pill but the Bean Counters get the gurnsey.

As for the mileage these engines typically do if they are correctly maintained is usually at least twice the mileage of its carburetted cousins. I put that down to better cylinder lubrication due to fuel injection. Sheit, there better be some compensation ! 

No pattern that I am aware of.

[MD]

Here's why it failed.

[kaleuclint]

Thing is -- there are so many vehicles with 'that' issue.  The sixth gear problem in the 159's Getrag box.  A certain BMW motorcycle that needs a groove to be cut in the outlet shaft and a circlip fitted like they used to, or else bearing failure is waiting to happen.  All cheap enough to fix as far as parts go (like a couple of dollars); just they all require hours of labour and/or mucking around to sort out.  All potentially catastrophic if unresolved.