What's new
CK5
Register an account today to become a member! Once signed in, you'll be able to participate on this site by adding your own topics and posts, as well as connect with other members.

Problem with Banks turbo crossover flange

BadDog

BadDog

SOL
Staff member
Super Moderator
Author
Joined
Feb 24, 2001
Posts
8,003
Reaction score
232
Location
Phoenix, AZ
In my copious spare time (heh) I'm working on a new desert truck (Sad Suburban build). This one's an '85 Suburban outfitted for long range camping trips and totally focused on expedition style trips without the gratuitous rock crawling and such. As such, I'm building a Turbo Diesel to put in it for reliability and fuel efficiency for resulting longer range. The exhaust manifold is a Banks Sidewinder, and right now it's the final turbo fitment that's causing issues.

This Banks turbo exhaust manifold flange appears to use a proprietary 3 bolt trigon flange. The flange is very close to a Walker 31305 gasket (link to specs), but the hole is at least roughly 1/4" too large (pipe is 2"), and the bolt holes are about 1/16 too far out. I've got that gasket and modified the bolt holes to fit, but that leaves very little (less than 1/8") between the bolt hole and the gasket hole.

The OEM Banks crossover flange was removed from the old tube on the lathe and the old weld beard turned down to more or less what it looked like when first used. It was warped pretty good, so I stabilized/averaged the flange face with shims to surface grind the back side flat, then flipped it using the now flat back side as a reference to grind the face with minimal removal. It took about 0.008 to clean the seating area, leaving the ears just a bit low (which I think is good). But then I had to weld it onto the new custom crossover pipe taking pains to minimize warp, which certainly warped it again to some degree, though I doubt near as badly.

FYI: The Banks gasket I got with the manifold and crossover was clearly leaking pretty badly between the flange ears on one side. Hard to say if warped by being installed by gorillas with impacts, or welding, possibly both.

So I have 2 question:

1) I know lots of you guys work in different industries. And while I can't find an automotive number that matches the Banks flange, I'm betting that Banks didn't actually have short runs made JUST to lock customer into their "only way to get it is a kit that will cost over $55 shipped". That kit only contains 2 non-common/cheap gaskets, and this is one of them. So if any of you have suggestions for other sources where I can look I would much appreciate it.

2) Given that the flange has certainly warped to some degree, and it's utterly impractical to resurface after welding, it really needs a good crush gasket. The banks unit appears to have a crush ring in the ID of the main gasket, but the Walker gasket is just a steel/fiber flat gasket with no ring. I don't have great hopes for it holding, and it's not an easy gasket to replace. I've got ultra copper high temp RTV, but given that Turbo EGT's routinely hit 1000+ *F, I don't have much hope of that holding up. What would you guys suggest to help that gasket hold?
 
Hopefully @AgDieseler knows something and may lead you on the path to enlightenment.
I know when we had this issue on the Sidewinder for the 7.3L units, we went back to Banks
 
I think I'll gnaw off my leg before going back to Banks. From my perspective, this was an intentional lock-in move on their part. Like inkjet printers and gold plated ink refills, if the printers were also gold plated and overpriced.

When I called their tech support hoping to either get the one gasket for a decent price, or maybe a part number, they guy went into "they make me say this" parrot mode. He just kept repeating, "I can't give you that". I got the distinct impression that they have orders to respond with that exact phrase because it might as well have been a recording. Exactly the same phase to every question regarding any option other than ordering the $55+ golden gaskets set. All the gaskets except 2 (possibly just the 1 flange gasket, not sure about that 6 bolt turbo exhaust flange) is readily available and cheap from any local source.
 
I’ve only ever used the Banks gasket, and have not had a problem that would drive me elsewhere. Candidly, a fiber-only exhaust style gasket should work just fine up the max EGT of 1100. Coat it permatex copper if you’re really concerned.

No gasket will compensate for warp, so clamp the wee out of it and control the movement as best you can. You can always experiment, and go back to Banks if you have to - both for the gasket and that half of the crossover.

David
 
Can you manually block sand the flange?
At least one crazy guy resurfaced a warped head using a block sander.
Granted, it was aluminum.
But I have to think; the flatter you can make it, the better.
 
When I made the new cross over I extended the pipe up past the flange to act as a bit of shield, so resurfacing isn't practical for that reason. It's also at a compound angle on the end of a twisty turny cross over, so no way to mount/grip for mill or surface grinder. I suppose I could go caveman and make a sort of surfacing ring with ID to match the 2" tube and flat surface. Then blue the surface to see how bad warped. If more than a little, glue emery cloth to surface and drive with a hand drill motor. But without a locator, that would be a dancing wobbly PIA to run.
 
BadDog said:
When I made the new cross over I extended the pipe up past the flange to act as a bit of shield, so resurfacing isn't practical for that reason. It's also at a compound angle on the end of a twisty turny cross over, so no way to mount/grip for mill or surface grinder. I suppose I could go caveman and make a sort of surfacing ring with ID to match the 2" tube and flat surface. Then blue the surface to see how bad warped. If more than a little, glue emery cloth to surface and drive with a hand drill motor. But without a locator, that would be a dancing wobbly PIA to run.
Click to expand...
It is an exhaust pipe.
You may want to consider that you are over-thinking it.
If you extended the the pipe to extend past the flange, you already have protected the gasket.
I'd run that bitch.
 
I have been accused of over thinking things in the past. But when I thought about it, I disagreed... :cool:

Analysis paralysis is often my problem. But in this case my motivation is in trying to make sure I don't have to go into that joint again in frame starting with a somewhat questionable modified 31305. And it clearly didn't hold up with the OEM Banks gasket that included that inner crush ring. Plus the only thing that comes close to dripping diesel oil in annoyance is listening to an exhaust leak.
 
I am similar.
That is also why my 6.2 diesel still starts, without block heat, at minus 37 f.
Agree on exhaust leak.
No way.

What about two fiber gaskets stacked?
 
I thought about that too. But the common problem with exhaust manifolds is that the gasket provides spacing, then blows, resulting in a leak. Common practice then becomes installing manifolds without gaskets because any small irregularity isn't going to provide enough to really make a leak (presumably due to fluid dynamic effects providing high resistance at that scale). So that train of thought nixed the stacked gasket idea, particularly with the small amount of ID width left after fitting to the manifold stud pattern.

Still hoping someone can provide a lead on a better gasket, perhaps figure out the original Banks source (just no way I see them actually having these produced, not worth it). That would go a long way to easing my concerns. But I'll give my current gasket a go before being extorted by Banks for their gasket that is still substantially too large on the ID.

This is what it looks like. The speckles are overspray from the VHT I sprayed the pipe with, it was surface ground before welding.

My current soft plan is to make the flat piece I mentioned (with 2"+ hole to fit over protrusion) to do a blue transfer and check for flatness, potentially quantifying with feeler gauges to determine if size/location will hinder sealing, and potentially provide a flat reference for flattening the flange if required (sandpaper etc).

And don't look too close at the welds. Two not quite the same shape of round butt welded 16 ga mandrel bent tubes don't make for pretty welds using MIG, at least not with my skill level. But I did leak test it under low pressure. Found and fixed one pin hole out of 5 butt welds, so I counted it a win...

IMG_20190218_104343.jpg
 
That DIY gasket option is something I've been pondering. And I found some 6"x6" 18 ga dead soft copper sheet that looks like a good option. It's also been suggested to use stainless sheet, but that seems a good option only between cast iron where you don't have that welded warp issue. And the final option in that train of thought is something like Mr. Gasket 5960 Ultra Seal Exhaust Gasket material available in a 10"x10" for about the same cost as the copper sheet. Ugh, decisions...

I wish it were cast iron to cast iron. Then I could just surface grind both surfaces and install it dry just like the manifold to head...

An interesting update is that I believe I've found the flange used by Banks. It appears to match AP Exhaust Flanges 9152S exactly. Unfortunately, calling their tech line indicates they don't have any matching gasket numbers. But if they are selling flanges (which seems to be the only 3 bolt flange for 2" pipe?) I'm surprised that there are no gaskets to be found. Then again, that may just be a flange for the flared end? The pictures seem to show a taper in the bore.
 
Looking back at 1st post, I'm not sure copper will hang at 1k F. I love it for all my header/exhust port sealing . That pipe really gets that hot?
 
From what I read, copper sheet should be good up to something like 1800*F. But my concern is what kind of chemical reactions are likely to take place with the exhaust gasses near 1k*F? I'm not sure.

As for temps, turbo diesels pretty routinely hit 1000 *F EGT when loaded. My Duramax is set with a programmer to defuel at (IIRC) 1300*F to prevent tubo damage. The little 6.2 won't be making near that power, or fueled that heavy, and has a fairly breathable exit exhaust path, so my working theory is around 1000*F at the top end. But I likely won't know an exact max number until I get it all set up and see the EGT running loaded up a 6* grade. If it goes too high, I'll be backing off on fueling.
 
Oh, and regarding copper, random searches and such has informed me that the "tuner" (aka ricer) community has a lot of problems with turbo exhaust inlet flange burning out. The (supposed) cool kids running insane boost in that world seem to have all settled on copper gaskets as THE fix. There is even a cottage industry grown up to support that market with premade copper gaskets (none to fit mine, I checked). And of course it seems that is probably cast iron to cast iron (?) so may not apply to a welded steel flange.

Then again, there is a huge market for gold plated jumper cables used for "high fidelity audio", and K&N filters, so the copper gasket thing may be equally just drinking the cool aid.:doah::whistle:
 
I think I know more of what's going on, and why there seems to be no gasket.

When I was looking at the flanges I found while trying to figure out what industry might provide a gasket, on one of the clearer images, it appeared to be beveled in the ID. That made me wonder if maybe it wasn't actually a gasket flange, but rather a flare end donut flange. While talking to a friend last night (Nick) about that evolving thought, it occurred to me that the donut flange on the other end is about the same size. Not sure why I didn't make that connection before...

So I looked at the other end and, sure enough, it's exactly the same flange as the stock flare and donut use! Not close, EXACTLY.

The bevel wasn't apparent on my "flange" because they just flipped it over and used the bevel as a weld groove. When I cut it off and turned/ground it flat, obviously the bevel no longer existed. Not having been produced originally for a flat gasket seal might explain why the surface wasn't even, and might explain why it seems those "not meant for that flange" gaskets don't hold up.

That also explains why I had such a horribly loose fit to the new pipes when I rewelded it onto my new cross over. It was a bit of a pain to weld, but just what you need for a slip fit donut flange. But in any case, the core-hole in the manifold casting I have does not lend itself to switching to a donut on that end. I really wish it did.

That makes me wonder if maybe they originally tried to put a donut at both ends, but found it unstable or something? Or if they were just punching up their margins by using a cheap donut flange that they are now buying in twice the bulk. My opinion of Banks just keeps dropping and dropping...
 
You must log in or register to reply here.

Latest Posts

Top Bottom