OOops! Scarey complication in crankshaft keyway repair.

[wrnchbndr]

I got a new crankshaft sprocket, key, timing belt guide washer, and bolt. I cleaned up the crankshaft keyway with a dremel followed by a thorough cleaning using naptha. It seemed that things were going to be fine because with the new sproket and key, the remaining original shape of the keyway (about 50%) was enough to positively align the sprocket with minimal rotational play (less than what seemed to be one degree) and a reasonable fit.

I went the JB weld route and did a number of test fits to make sure I could get it all together after I applied the epoxy. I even did a practice run using the new bolt and torqueing it down all of the way to 94 ft/lbs. The bolt easily threaded all of the way in with no restriction.

I installed a new timing belt guide washer properly and then the new key and backfilled the damaged side of the keyway with JB weld. I did not put what I would call an excess of epoxy on the nose of the crankshaft. It took a little fiddle with the key itself to keep it from sliding out of place backwards and binding the installation of the sprocket but I did finally get the sprocket to slide on with the key settled in place after a few trys -- I didn't need to wack it on -- it slid into place and the rotational play felt right when it did.

The crankshaft bolt screwed in without a problem. I had fabricated a tool to hold the crankshaft via the pully bolt holes. I tightened the bolt to 94 ft/lbs and let the JB weld cure for 24 hours.

Now the crank is excessively tight when I try to rotate it. It took 100 ft/lbs to get it to rotate initially. This is with the timing belt removed.

Yea, the car is out of gear. I didn't do any work to the crankshaft -- the oil pan was not removed.

The rotation, although tight is smooth with no binding or grinding feeling or noise. I think it might still be too tight for the starter to cope with.

I think that the damage to the nose of the crankshaft might have been such that the torqued sprocket is now in contact with the engine block. I hope that JB weld did not go under the timing belt guide and bond to the oil seal. Getting this sprocket off again is going to be a little bit of a problem and this was a passing thought when I decided to use JB weld -- what if I need to remove the sprocket again? I may be in trouble.

Thoughts???

This my plan. I'm going to try and start the engine up and see what happens. If its a small area on the front of the aluminum block, it will wear away rather quickly. I'll do short runs of thirty seconds or so and see if it loosens up so I don't create too much heat from friction and burn up the oil seal. If I've bonded to the oil seal, I know it soon enough.

Any more thoughts???

[Scudo]

Are the plugs out when you try to rotate the crank?

Edit: Sorry if I missed that in the post somewhere.

[Bex]

This might sound silly - fortunately I have never had to do a keyway fix - but did you not let the epoxy cure with the keyway in place, and file whatever excess off, before you put the crank pulley back on?

[aquanaut20]

My reading, the JB was wet when assembled, complete with cog and crank key...

I suspect some material was forced out the back of the crank/cog... Philip

[Bex]

That was kind of what I was thinking.....

[ranger88den]

X 3, rubbing on crank seal and washer.

[wrnchbndr]

Yep, the plugs are out. I got the idea of putting the cog on wet from watching a youtube video on Miata repair which was illustrating the same process -- not an excuse -- its kind of a dumb thing to do but if the repair didn't work I was going to replace the motor with a 16V I have on hand. The wet JB weld on the nose of the crank is to address the internal wear of the surface of the crank and there was some. Using JB weld goes against the grain and my dead ancestors are responsible for the complication. The two minutes it would have taken to smear a layer of grease over the seal and the surounding area would have prevented the problem. Maybe its a karma thing for not filling up the gas tank on my friends log splitter before I brought it back to him.

I have the motor all assembled now with oil and a new filter. I need to charge up my battery. If the battery won't turn the motor (it might now), I'll pull start it. It'll be fine unless the JB weld spins the oil seal.

If the oil seal gets toasted, my guess is that it'll take an oxy-acetaline torch to heat up the cog enough to break the JB weld. I can make a puller to pull the cog off.

[xxx]

why use JB weld, JB is NOT structural epoxy i even have the Loctite engineers
post on my page, why not read that. (he owns a 91 MX5 , and fixed his own car)
why not post a photo of your crank snout. not all bad ones are fixable. (damage comes in all different degree's seen most)
you lost before you started.

Repairing damaged crank and replacing seal. - Page 1 of 1


there it is, using the 660 real stuff. use cheap dollar store epoxy and get failed.
its not magic or bs.
the 660 does work.

Shear strength of #660, is 2000 and up to 4000 PSI. (tuff stuff) (but must be clean ,rough surfaced)

and only works, with no AIR.
unless you buy the r , fast cure kit from Loctite.

Shear strength of #660, is 2000 and up to 4000 PSI. (tuff stuff) (but must be clean ,rough surfaced)

[wrnchbndr]

Don't actually need any strength since the keyway is not really intended to bear stress. Its a poor design. Can't really say that there is a failure yet. We'll know when I get a battery in the car and/or after a few thousand miles if my dead ancestors and karma are cooperative. I have an honest 94+ ft/lbs of torque on a bolt that threads into the crank properly unlike the previous bolt which ate up at least 15ft/lbs threading it into the crank.

[xxx]

that is correct, the key has no stress (my page states that)if the TSB is followed. for sure.
i just want the key not to move for 4 more belt changes. and then the dead ancestors get it.

its not poor after 1996, and the TSB is retroactive.
so there would be no failures after 1996, if mech. can read. or soddy books can read before they print. yes, haynes still gets it wrong. 14 years straight running.

many cars have just a key and work great, the mx5 got a longer snout
the suz has a longer. snout. so that is not wrong.
many mechs refuse to use a torque wrench .
some don't want to lock the fly wheel with the A/T, so guess.
id bet those are all the reasons, not the design.
i bet that if you start with fresh parts, cog/bolt and 94lbs. id bet it never fails.
in fact never heard of one failing set right.
some bolts have been yielded by using a nut behind the impact gun. too.

imo

unlike the previous bolt which ate up at least 15ft/lbs threading it into the crank.

a victim that was, from being loose from last mech.