[s-cars] sheared crankshaft-gear key

Fri Jan 23 08:02:59 PST 2009

 Dave:
My point here is to develop a procedure for *everyone* to follow.? Me, I could care less if it's happened to anyone else.? I know that I'm overly anal about timing belt procedure.? And I know many aren't.? This isn't a 'risk', this is a historical reference of failure (debating the cause seems tangential).?? I turned white as a ghost the first time I encountered shear on removal of that component.? Thinking of the potential carnage of that failure = catastrophic.? From then on I over-scrutinized that key on both sides by removing the gear from the balancer.? I found many keys with cracks in them, fully capable of functionally fit for use, but potentially catastrophically-not.

I'm not looking to debate your experience, nor I doubt you, mine.? I know this problem exists in the design, the tool, and the procedure, and has for many years.? My opinion differs with yours only that you opinion it's procedural, and I opinion it's a combined: hardware, tool and sealant problem.? Looking at this procedure vs the forces on the hardware (see page 23 of the article below - torque is the 332lb/ft), key-shear would never be an issue if Audi locked the pulley and the crank.? Then there would be no shear forces on the key involved in this procedure and failure would be attributed elsewhere.? The reality is that the crank is free to turn during removal and install, the only thing preventing shear is a cheap pot-steel key.? That to me is the flaw in the procedure (compounded by the use of sealant that turns to kryptonite with heat cycles).

That bolt is the center of attention.? I don't see the reuse as good practice given the potential forces it can exert on the key.?? Interesting to note too, that the v6 locks the crank (Tool 4433), not the pulley = same problem.? An I5 with both tools applied?

To avoid the dead horse syndrome, here's where we can agree?

Remove and install bolt using proper Audi Tools
Inspect crank gear for shear-force cracks on key (ala page 23 above) = replace if key missing or cracked
Clean land area of balancer
Clean bolt threads and reapply AMV-type sealant to threads and bolt land area
Torque Crank Bolt to Spec

That leaves only:? "Replace Crank Bolt" as the only 'insurance' issue up for debate.? I say it's part and partial to every Timing Belt change to properly attain the factory torque spec of that fastener.

(I'd encourage anyone to do? the touchy-feely experiment on that crank bolt.? Put the old one in and torque to spec, then pull it out, put a new one in and torque to spec.? I claim you can feel the difference between the two easily.)

Cheers

Scott J

-----Original Message-----
From: djdawson2 at aol.com
To: qshipq at aol.com; forgied at shaw.ca; s-car-list at audifans.com
Sent: Fri, 23 Jan 2009 3:10 am
Subject: Re: sheared crankshaft-gear key

 Scott,

Thanks for the info.? Again, the keyway article does not address clamping force.? Blau's "service bulletin" says it all in the first sentence... "inadvertent damage during removal."

You don't have to make a case with me, nor do you need to be scared.? I replace parts that do not pass a close visual inspection... and I replace fasteners that indicate stretch, indicated by thread resistance.? I don't EVER replace everything for the hell of it.

Period... done... nothing more to discuss, really.? I am not lying to you about the fact that I have not had one fail.? I am further not lying to you that I have seen a crank key cracked after I removed it (once, a VW).? I have never seen a failed cam gear, except those others have shown me... typically with a humble admission about forgetting to torque it down after assembly.

It is really irrelevant to me to hear all of the stories about mystery shops doing who-knows-what to remove a crank gear.? To me, that is not valid data, as it introduces far too many variables and we can establish no meaningful relationship between the method used and the resulting failure.? Ask any mechanical engineer willing to speak up on this list.

It's simple math/stats to me (and in industry)... my personal cam/crank key MTBF is infinite, with respect to a catastrophic failure.? It simply has not happened to me, and this is fact.? That alone invalidates any arguement to replace all parts, every time.

A common error in reliability engineering is to overlook the root cause.? I believe the root cause of the failure to be improper removal and/or assembly procedures, combined with inadequate inspection... and I think you're making the same mistake.

My own AAN puts out plenty of power.? It's rev limiter is set at 8500rpm, and it has been there many times at the strip and on the dyno (and at least 6 bounces when driven by B Rogers).? I've removed/installed the same crank gear at least 4 times myself... the cam gear many more than 4 to test exhaust cams.? These were the parts that came on the engine when I got it.? I'm rebuilding the engine now, and unless these pieces fail to pass inspection, they will go back on... and I have every statistical reason to believe they will not fail.

The ideas I state are not mine, nor are they based on speculation, Scott... they're based on principles used by the largest fleets in the world.? Those businesses live and die by the reliability of their equipment.? I've worked with 3 of the top transit authorities in the US, the largest 4 trucking companies in North America, as well as the largest airline in the world.? Millions of dollars are spent doing reliability analysis and engineering, and the solution is rarely to replace everything in sight with new components.? If that is what you elect to do... it doesn't bother me at all.? It's just not what I will do.

This has been beaten to death a bunch of times on this list... and this is my last thump on the carcass.

Dave

-----Original Message-----

From: qshipq at aol.com

To: djdawson2 at aol.com; forgied at shaw.ca; s-car-list at audifans.com

Sent: Thu, 22 Jan 2009 5:19 pm

Subject: Re: sheared crankshaft-gear key

 Dave:

I sure wish this phenomenon was rare, it's just not.? For understanding on how Audi designed that key to fail, take a gander at pages 23-27 of this article:

http://www.mfge.atilim.edu.tr/Courses/mfge306/Material/Course%20files/MFGE306_Lecture%205-Locking%20Devices.pdf

 I claim that key was never designed to take the reverse shear load in removal of the sealant coated (audi glued) crank bolt.? The key is designed to fail, or it would be a long square key to spread that type of load and shear stress.? If you do some of those calcs vs 332lb/ft of bolt torque (times a multiplier on removal with hardened sealant), we should expect to see more failures of that key not less!? My theory is that key failure happens more often on removal than on install.? Never seen one?? I can't be that unlucky, but I also remove that gear and inspect it under a mag-glass, because a hairline fracture is all that's needed.? Again, I claim Audi was aware of the issue, and made the TB R&R on the v6 to be without removing the crank bolt or gear to reduce this chance of key failure.? I also claim it's not an 'old' concept unique to the I5 either, this later Audi crank gear key is wider than the I5:

http://www.blauparts.com/audi_technical_tips/audi_timing_belt/audi_crankshaft_pulley.shtml

300 belts without 1?? Can't fathom that frankly, maybe it's a salty-chicago thing, but I really doubt it.? I witnessed my first key failure well over 10 years ago, and have seen many since, and just won't take the risk.? The threads and contact surface of the crank bolt should be coated with Audi sealant AMV 188 001 02 (watch yer wallet!), which over time turns to a really hard cement (part of the key-shear on removal problem IMO - see page 23 above).? Torque of the crank bolt on the 20vt is 350NM (with 2079) or 450NM (without 2079).

If you are doing all this without replacing the bolt and sealing (by definition: 'lube' on install) it, the torque spec you use is no longer significant, nor does it correlate to how Audi put these parts together at the factory.? That bolt has an elastic phase to it, and on reuse a more dangerous plastic phase to it.? 

Minimal Bottom Line:? Buy the new bolt, pull that gear off the pulley and inspect it carefully.? Sealant on the crank bolt threads and bolt head face, and torque to spec.? 

You guys scare me sometimes.

SJ

-----Original Message-----

From: djdawson2 at aol.com

To: qshipq at aol.com; forgied at shaw.ca; s-car-list at audifans.com

Sent: Thu, 22 Jan 2009 2:00 pm

Subject: Re: sheared crankshaft-gear key

 Agreed... if there is a bolt that reluctantly threads in, it has yielded, effectively changing the pitch. That bolt should be replaced... crank, cam, or anywhere else a bolt is used.

It should also be noted that more responses to this post are coming to me off-list than on-list... due to "sheer" (pun intended) desire to avoid entering the debate. Most recently, one shop, that I am vaguely familiar with, touched base claiming in excess of 300 timing belt procedures without a single keyway failure... and elaborating that they haven't even seen one. To me, this is just another set of data points to consider.

Again, for me it's a numbers game... and the numbers don't justify modification to the method/materials. It makes no difference to me if everyone on this list makes it part of their standard practice, it just isn't part of mine. I've shared my opinion, you've shared yours... and anyone reading this stuff can try and make an informed decision.

I still believe that these failures are due to the use of impact tools.? I've seen a couple of keys damaged that I replaced, but I was not the last person that serviced the car.? I have NEVER pulled one apart that I had assembled and found a failure.

Peace.

Now, let's talk politics...

Dave

-----Original Message-----

From: qshipq at aol.com

To: djdawson2 at aol.com; forgied at shaw.ca; s-car-list at audifans.com

Sent: Thu, 22 Jan 2009 11:07 am

Subject: Re: sheared crankshaft-gear key

 The only place I would take exception to your presentation below, is with not routinely replacing the crank-bolt Dave.? It is a torque to yield.? Think about what the crank "locking" tool is locking.? It's locking the *crank pulley*? - not the crank, in a given position....? Now, you take a 'already' yielded bolt with a 1mm pitch and torque the crank pulley to the crank with 400lb/ft+ of torque, the stress of that locking tool and the shear forces of tightening is exactly and fully on the key.? You have the face of the bolt on the pulley, the threads of the bolt in the crank, and the only thing preventing shear between the crank and the pulley, is a crappy pot steel key.? The more heat cycled and prior-yielded the bolt, the more shear on the key.? The omission of the 100USD insurance I have a tough enough time grasping.? Impossible to think of a stretched+heat-cycled hardened bolt locking a crappy pot steel key in an attempt to prevent some serious shear forces as that bolt is torqued to spec.

The very physics of this assembly process dictates failure of that key.? Or (IMO) just darn good luck in your case.? I don't look at the procedure as dictating the potential failure.? I see the hardware and reuse of that same hardware as the potential failure of the procedure.? If that bolt is torque to yield, what is the torque spec you advocate when reusing a heat cycled/torque yielded bolt vs new?? Next time you do the procedure, try putting in the old bolt, take it back out and put in a new one.? You can feel the difference = yield.? BTDT.

I'm not an engineer looking at the practice of failure, I'm just well seasoned at reverse engineering failure to practices.? I don't even keep the broken crank keys anymore, but have over a dozen in 15 years.? Forgie has a picture up of what I see often when it's blatently obvious, and I have had that failed bolt pic on Audiworld for over 8 years.? I kept that bolt to remind myself what would have happened had that bolt sheared fully with the leftovers in the crank....

OR, Force Majeur sucks?

Scott J

-----Original Message-----

From: djdawson2 at aol.com

To: qshipq at aol.com; forgied at shaw.ca; s-car-list at audifans.com

Sent: Thu, 22 Jan 2009 11:36 am

Subject: Re: sheared crankshaft-gear key

 Again Scott, I think you know I respect your opinion. For those thinking this is an argument, I beg to differ... this is good natured debate.

I suppose if I were operating a shop and repairing other people's cars all the time, I would do the same thing. After all, it is cheap insurance that your customer is paying for.

Perhaps I will change my mind if I have one fail. However, as stated before, I've got 26 years of experience with the exact same setups, and haven't had a failure yet.

I guess I'll choose to think of it like this:

Insurance... the concept of paying a small amount of money on regular intervals to eventually cover for the big event *if* it ever happens. You have selected to have yourself or your customer pay $100 at each event to prevent the "large" failure.

I have chosen to not pay on my policy, and deal with the big failure "out of pocket" *if* it ever happens. If I were to count the number of times I've either done a timing belt or removed the gear to swap a cam etc... I would conservatively estimate it at 100 occurrences during 26 years. 100 x $100 = $10,000.? Even *if* this failure ever bites me, I'm still a solid $8000 ahead of the game.

Component failure is all about statistics, that's what I do for a living. I don't disagree that the failure *can* happen, I disagree with the rationale of replacing those parts at each maintenance interval, because the frequency of failure doesn't justify it... if the job is done properly, and the parts are carefully inspected prior to re-installation. I *have* replaced each of those gears on some occasion... but they looked like Dave F's picture, and the disaster never happened because the clamping force kept the gear stationary.

If I had a preventive maintenance engineering client with a fleet of Audis (that'll never happen!) I would not advise them to include replacement at each interval as a part of their PM program.? The cost is higher than the benefit.

Again, this is something that each person has to evaluate for themselves.? My engine is currently torn down... I'll be happy to tell you when I have my first failure.

Dave