Sealing ball flange

[AA-ron]

I've been struggling to make a good seal on my ball flange. I know this topic has come up before, but I thought I'd post my questions.
Because of the proximity of the O2 sensor, the ball flange needs to make a near perfect seal. I've been struggling with engine performance that is traced to leaks in the exhaust. I've pressurized the system via the shop-vac method and it's clearly the ball flange that is leaking, which in turn is throwing off my O2 reading. I've repositioned the flange clamp and no difference. I put a thin layer of high temp RTV and that helped for a while, but the problem returned. The mating surfaces are pristine as the car only has 3000 miles on it. Both mating surfaces of the ball and socket are smooth as well.
Soooo, I'm at a loss. This must be a fairly common due to the sensitivity to leaks with systems like the Holley Sniper, which I'm using. You can see the leak show up in my data log, so I know it's the culprit.
Any thoughts out there on a solution?


Thanks

[JohnK]

One possible problem is the ball flange shifting over time as it is carrying the full weight of the exhaust. I didn't have a leak at my ball flanges but I was struggling to keep my side pipes from sagging so I added supports for the ball flanges. Details in my build thread here and here. I have to believe that if they're being forced to carry the full weight of the side pipes, they will shift and eventually leak.

[BEAR-AvHistory]

+1 on the additional exhaust support. Used turnbuckles headers to frame.

[steno]

When I get my headers to fit, I'm going to tack weld the joint, take off the clamps and tig weld the whole joint. It'll fit perfectly, not leak and look better overall.

[cc2Arider]

I have ambitions like steno, but haven't actually TIGed SS before. It's been many years since I TIGed anything...Methinks I'll need lots of practice

Craig C

[Jeff Kleiner]

Craig,
I know a guy…



Jeff

[cc2Arider]

NICE!

When I took my TIG class at Ivy Tech, the Instructor told us about that challenge (if you can TIG aluminum cans together, you're doing alright). I attempted with my meager skills and blew thru the metal quickly. Later, I discovered that my Tungsten was too big anyway...but it was a fun demonstration of the skill involved in "doing it well"

Craig C

[steno]

I'm lucky that I know a guy too!!

[AA-ron]

exhaust picture.jpgside shot 2.jpg
I'm not running standard side pipes, this is the exhaust system I built with the muffler in the nose. So I don't have any weight on my ball joints. I'm just looking for any type of sealing advice on how to keep the balls from leaking. Even the small leak I have is messing with my O2 sensor.

[Sdonnel]

I had two tack welds put on the bottom side of the ball sockets. I have tried to move them with no visible change in position.

[JohnK]

If you're reasonably sure it's not movement of the ball flange that's causing the leakage then there must be some sort of irregularity/deformity of the sealing surface. See if you can identify where exactly the leak(s) are occurring, and look for high or low spots or out-of-roundness in the ball and flange. IMO, adding sealant is just a band-aid. They should be sealing fine without it.

[Jeff Kleiner]

How about wrapping the ball with a couple layers of thin malleable copper or aluminum then install the bell and pull it tight. Kind of the same principle as the laminated exhaust gaskets like Remflex, etc. in that when the joint is tightened the layers will “extrude” to conform to irregularities and seal. It might work???

Jeff

[Dave M]

I seem to remember this system or one just like it from awhile back. If the system doesn’t have hangers that will flex easily with engine movement I suspect your ball joints are being stressed enough to break seal.
The nature of the design looks like it would be very rigid.

[Railroad]

I do not think the ball flanges are tight enough. If he used an rtv sealer and it started leaking again, it was not tight enough, IMO.

[AA-ron]

Well the problem is solved, I guess. Only miles will tell for sure.
First, I may be a bit anal when it comes to checking for leaks and expecting a bit too much out of a ball joint. My data logs show an exhaust leak according to Holley, so I set out to make sure there are NONE. The leak I was seeing was a very very small soapy bubble at the joint, but according to Holley, "any leak throws off O2 readings".
I removed the exhaust to take a long hard look at the situation and the ball and cup looked ok, kind of, but I can see where there is some deviation in the compression marks left on the cup, and I suspect that one of the two surfaces is slightly out of round.
The fix:
I made a male/female mold out of stranded fiberglass to match the ball and socket and used them to press form a piece 24 gauge copper sheet into a gasket that exactly matches the profile. I put it in place and the leak seems to be gone. Again, I may be a bit too OCD here, but it seems to have worked.
Also, regarding the comment on the exhaust system being rigid thus putting stress on that joint is a great point. I went to great lengths to double check that it moves with the engine easily. There are rubber exhaust hangers at every connection point and it does move freely. I appreciate that note, because it wasn't something that I had considered as a root cause. However, everything is still nice and flexible, so I should be in the clear there.
I'll let you all know how my copper gasket holds up. It better, because any other solution involves a ridiculous amount of work