Small Exhaust Leak @Spring Joint w/New Gasket?!

[Anon.II]

Recently, my driving became prominent to anyone within earshot

when parts of the original joint gasket fell out. 1991 Legacy L wgn, auto.

(Thanks to a prior poster for the image of what/where:

http://img.photobucket.com/albums/v684/Setright/cat2.jpg )

 

I got new parts from Advance Auto -- bolts, nuts, springs, and gasket.

With a friend's help, we cut off the old bolts (Dremel) and installed the

new gasket (which is conical, looking like someone soaked a brillo pad

in zinc or ...); but it leaked. The bolts can be tightened only so far

--the threaded end just clears the nut, and the thicker shaft snugs

up to the plate. (I should remark: the old bolts had the same size

EXCEPT for having longer threaded ends -- which would make starting

the tightening easier (we used tools to bring the pipes close enough)

--; the thick part abutting the plate to the broad flat part compressing

the spring is same length.) (I neglected to compare springs; the old

ones I see are now as long as the thick/unthreaded part of the bolt.)

 

Well, we added first one, then a 2nd washer (thin: 2mm@?). But

at the end of the day (and it was that), there remained a small,

felt w/one finger, leak. Quite quiet and all, but he doubts that the

emissions tester will think so, or the testor will detect something.

 

---------

Are these gaskets such that with some use --vibration, continued

spring compression, heat-- , the rear/cat-side pipe will press ever

more into the gasket and finally, completely seal?

 

My sense is that there is a side --and it must be of the gasket,

for it shifted between our efforts (1 washer, than 2nd)-- from

which the leak comes; i.e., one could mark the gasket at the

point, and then *aim* it by rotating the gasket (when loose).

So, some imperfect match of roundness of pipe & gasket

concave/convex surfaces.

 

I suppose we could try adding yet a greater washer thickness,

so that upon tightening, the springs would be even more compressed.

But pulling in two broad surfaces that imperfectly match seems to

be something that could resist considerable force, at least initially.

 

--Anon.II

[porcupine73]

The surfaces must be clean of excess rust particles for the gasket to seat properly, especially the mating surface on the mid pipe. Here's my pics from when I had this issue '00obw

 

can't have an uneven surface here or it will leak

 

[avk]

Edit: what he said.

[davebugs]

My solution is simple.

 

The local exhaust bending place fixes these for 40 bucks. Almost what I have in parts especially when I need to buy those goofy clamps to mimic a flange.

 

They cup off both pipes and weld a new piece in - maybe 6 inches. I do this after all work has been done on the car. Lets face it these joints are wimpy. Even if you fix it now within a few years the exact same problem is likely.

 

On the rare occasions that I do just put a new donut in there (I'm working on 95-99's) Walker does make a slightly larger donut that I find works better than the one listed that is often too tight (although I always start with the listed donut to try). The generic hardware kits with springs are useless. If you'd like some factory springs and can wait a few days and pay for shipping plus a couple of bucks I can send you some (again off 95-99's). PM me.

 

Dave

[EVOthis]

My experience with these is this...as mentioned make sure both mating surfaces are clean and rust free...install gasket..and over time they usually work themselves into place..hence the use of the springs/bolts....Ive never really had any problems with them leaking much...

[johnc]

Eeerily familiar, AnonII! top of my list is to add 3 more washers to the existing shim stack from January and tighten the bolts some more, because I just noticed a trace of soot on the exposed part of the subaru donut. I looked as there was a slight increase in noise level but I couldn't feel any gas leak.

 

I had used an Autozone flange repair so the subaru bolts weren't the right length at all.

 

John

[Anon.II]

Porcupine73, good photos! How is that photo'd condition, IYHO?

--i.e., one can see rusty imperfections, but how much sanding/etc.

did you do to that, or is that the workable state (really rough & loose

stuff gone) ?!

(I confess to having not given much thought to this, but recall that

I don't think my pipes were much worse really.)

 

ALSO, again, it seems as though my leak has a direction, which I

think is related to the **gasket** --rust, OTOH, is fixed in place on

the pipes. So, that suggests some imperfect roundness, IMHO.

(i.e., again, the gasket I think rotated some between initial install

and re-installs w/added washers; the direction (& volume) of leak

changed.)

 

Davebugs, thanks; I'll keep your kind offer in mind. I do have time

(emissions test due August). So, you find the non-Sub. parts to have

inferior springs, insufficient tension? --not redressable by washers?

(Though, as I noted, given the short threaded part of the replacement

bolts --JUST enough to go through length of nut, really zilch beyond(!!)--,

increasing the compression will entail some added effort/technique to

bring pipes close enough to get a fair start threading the nut (after

which there is only so far to tighten_.)

 

I'm also remarking at the different appearances of the gaskets -- what

PP73 shows being quite *meshy*, and the Advance being more *solid*,

covered mesh.

 

BTW, my buddy suggested trying one thought I had about inverting

the gasket, as that would put its thicker side towards the Cat-pipe's

concave compressing edge; but re-thinking that I figured it would

also quickly compress & CRUMPLE the gasket into pieces!

(and that WWII bomber airplane sound would come again)

 

I'll go check the Advance springs to get some idea of their coil

vis-a-vis the Sub ones (which had ends overlapping about 1/4"

and 5 full turns/coils between).

Will have to give a little chance to seeing the gasket get compressed

into a complete seal. --my preferred repair method, wishful thinking.

 

Thanks,

--Anon.II

[porcupine73]

Not sure, I just wire brushed the surfaces and wiped them off before putting on the new donut. I don't remember if those pics were before or after the wirebrushing. I also gave it all a good coat of antisieze. It didn't have any leaks after.

 

[ron917]

My son's '99 Outback Wagon (formerly mine) has normal bolts with washers (at the head to capture the spring) instead of those shoulder bolts. That allows the springs to be snugged up real tight without disassembling to add washers.

 

I bought the car used, so I assume it did not come from the factory that way.

[OB99W]

[...]Here's my pics from when I had this issue '00obw

Looks like you had a camshaft to seal and a starter that needed contacts as well.

[avk]

Yeah, same contacts as I changed in the minivan some weeks ago. They should have stayed with Mitsubishi starters.

[porcupine73]

hehe yah when I was taking pics of the exhaust parts the o-rings and starter contacts were making a big fuss about being left out so I let them be in the pic too.

[OB99W]

hehe yah when I was taking pics of the exhaust parts the o-rings and starter contacts were making a big fuss about being left out so I let them be in the pic too.

 

I only mentioned it because I thought someone might think there was a trick to sealing a leaky exhaust connection using those parts.

[avk]

Well, the bolt holes in the contacts are 8 mm, so if you need extra washers, in a pinch...

[Anon.II]

Ha! I recognize that disintegrated heat-shield corner!!

 

Nice, er, *belt* you put on the waist. I just marveled that mine was holding

up sans the two bolts (both corners) at that end, and will have to look into

doing some rigging of some sort I guess.

 

[porcupine73]

hehe the belt was to keep those heat shields from rattling. I must have at least a half dozen belts on the exhaust parts to stop all that ratling. I like my cats to be well dressed.