use the Fel-Pro permatorque gaskets and add 5-10 ft-lbs to the final torque value (though on a non-turbo it probably doesn't matter, they don't blow gaskets unless they're run out of coolant) - they don't require a retorque and are highly utilized by many of us. i wouldn't touch fel pro's for newer stuff, but this is one engine where they shine. i actually have a set i probably won't need any time soon. i'll ship them to you for a bit less than cost.

 

A backyard hack wrote the above. Consider this:

EXCESSIVE bolt loading will cause problems. Many backyard mecanics think that if 50 Ft. lbs of torque is good then 60 is better. In fact dynomometer testing has shown that less is best in most instances. Indeed extra horsepower maybe gotten by being "torque frugal". Why??? The more torque that is applied onto the block the more chance of distortion. This distortion is usually seen at the weakest places at the narrowes point of the bore and at the top of the cylinder. At the top of the cylinder where the compression pressures are always greatest any excess distortion will nullify any benefit of that extra clamping force. Blow-by of gasses will cause premature gasket burn through and less horsepower. "Less maybe best".

Doug

That's a good point about the block distortion - hadn't considered that myself.

 

The stock final torque value on the EA82's is 47 ft/lbs..... but the stock Subaru gaskets require a retorque where the bolts are loosened, oiled, and retightened after a run-in period. The Fel-Pro's supposedly don't require this extra step. As an added safegaurd against having to retoque the gaskets (because it's a huge pain in the rump roast), I typically do a final torque of 55 ft/lbs after the 47 just to be doubly sure. Emperical evidence sugests that this is not a detrimental practice at least in the case of the EA engines as both the 55 ft/lbs and the Fel-Pro gaskets have proven quite reliable in everyday testing. In any case 8 ft/lbs is not a lot and I did it as much because I was using a cheap $20 torque wrench and wanted to make sure I wasn't on the loose side due to any innaccuracies in the tool..... which was probably an unfounded supposistion as they are actually quite accurate but I didn't know that years ago when I was messing with EA stuff and doing lots of HG's on them. I rarely touch one anymore - mostly I work on EJ's and I use the stock HG torque from the FSM

 

GD

All this chatter and no one saw the cracked valve seat... gasket slap is a horrible idea on any turbo motor. Shame on all of you for suggesting it!