Bah ; charge pipe O rings blowing out

[spuggy]

Like title says; opened the engine bay and found the top charge pipe O ring seal sitting on the ledge behind the mounting console. This is not the first time...


Anyone have any idea why this happens? My best guess is that something is misaligned, the rubber isn't supported properly, ends up with the full charge behind it and just blows out.


Oddly, it seemed to be sealing fine on higher boost a few days ago - and still sealing on 1 bar more recently - but then I heard the metallic whistle under boost, and the BOV wasn't making as much chuff-chuff noise on shifts...

Is there any kind of tape that would conform to the shape of the charge pipe, compress down to close up the gap and stay in place? Hoping this might provide a better seal/avoid the O ring taking the entire load.

[flightlead404]

Too much boost. Dial it down. You probably only really need 2-4 psi!

lol

[RarlyL8]

I don't know what your setup is but when using the OEM up-pipe with the dual O-rings it is important that the pipe be tightly secured at the fan so it cannot move. Also important that the intercooler cannot move. The few times I had a problem that was it.

[spuggy]

Originally Posted by RarlyL8

I don't know what your setup is

Huh, yeh, I guess I should have said something about that...

K27 (w/ adapter in outlet), up-pipe (looks a little modified), 3.2 manifold w/ rotated throttle body, modified Kokeln; end tank outlet to suit TB and a tab on the back edge to accept a pair of factory rubber washers & bolt down to the manifold.

Front edge of the Kokeln uses the factory rubber-mounted L bracket bolted up to the coil bracket on the fan housing.

Originally Posted by RarlyL8

but when using the OEM up-pipe with the dual O-rings it is important that the pipe be tightly secured at the fan so it cannot move. Also important that the intercooler cannot move. The few times I had a problem that was it.

Thanks Brian. Yeh, it *seems* tight enough when bolted down all the way - after all, there's no metal-to-metal contact anywhere - and I've got the rubber doughnut on the intercooler as well as the 999.140.469.40 O ring (which PET says to use 3 of - WTF do the other 2 go?).

I tried using the orange silicon doughnut, but it interfered with the charge pipe and prevented it orienting correctly on the intercooler. As it seems to only be to seal the tin anyway, been leaving it out.

I suspect maybe sometimes everything isn't quite lined up perfectly, resulting in one of the joints being cocked over a little & perhaps leaving too much of the o-ring unsupported. Had the intercooler on and off a bunch recently, and entirely possible I got careless a time or two.

My only other concern is that the charge pipe looks to have been modified somewhat (weld bead around the bottom flange fitting). Pretty sure that's the one that came with the Kokelyn/K27 when I bought them way back when.

Maybe the current setup was intended to use/would do better with a bone-stock charge pipe. Maybe I should get get one and compare...

As a band-aid, I was toying with the idea of using silicon, hard pipe sections & t-band clamps - but this really seems like a PITA.

Maybe I should just try fabbing something up that seals better...

[RarlyL8]

For custom applications such as yours I prefer to use silicone tubing and clamps.
The O-rings really just work for the OEM setup as everything is designed for it. If you have proper alignment and the intercooler cannot pop up on boost the O-rings work pretty well. The additional rubber seals really don't seal anything, they are more of a cushion for the assembly so you don't have metal to metal vibration.

[Alan L]

A pic is worth a thousand words.
Alan

[philip j]

happened to me also...

[RarlyL8]

Often times a blowout occurs due to the o-ring being pinched/sliced during installation. You can't see it until you take the assembly apart. The picture above looks like such a situation.

[Alan L]

^^
Agree. Hard to see it slicing its way out. But being pinched and sliced at assembly is more common.
I always put some CRC on them at assembly.
Alan

[Tippy]

Yeah, I doubt you could blow out an O-ring with just boost. An 'unsupported' O-ring can hold a ton of PSI.

I suspect slicing/tearing during installation, too.

[philip j]

well, I had 2 o rings blow out while on the dyno last session... both times, I found the missing piece on the engine tins... I know its hard to believe...

[spuggy]

Mine always have what looks like a single cut, at an angle. But the end isn't clean like a razor, more like it was pulled apart.

At the risk of exciting Rawknee, there's always plenty of petroleum gel involved in assembly, and I always try to get both pipes started square at exactly the same time (yeh, right, that's easy to do - not).

This last time, I popped the lid for something else and a split o ring was sitting on the engine tin. I thought maybe it was an old one I'd dropped previously and not noticed, and the fan had blown it around.

But I pulled the I/C and the charge intake was entirely bare - it'd spat the entire top o ring out...

I'm leaning towards picking up a cheap bead roller and experimenting with silicon hose and hard pipe to get a better fit/seal. I can hear a boost leak and don't pick up anything over WG spring right now...

[dos531]

I was so happy to get rid of those damn o-rings. Silicone couplers or v-bands are the way to go.

[Alan L]

Spuggy - can you post a pic of where the problem is and maybe some measurements - ID of the IC neck and OD of the pipe?
Something not right here.
Regards
Alan

[spuggy]

Originally Posted by Alan L

Spuggy - can you post a pic of where the problem is and maybe some measurements - ID of the IC neck and OD of the pipe?
Something not right here.
Regards
Alan

Kokeln is pretty much "standard aftermarket long-neck intercooler" utilizing as many factory fittings as possible, and the turbo intake side on mine is unmodified.

So the I/C intake neck is the 3"/~76mm fitting accepting the factory o-ring, and at least the top of the charge pipe that mates to it is stock (I think the bottom of the charge pipe is modified, but I also think that's to mate to a K27 better - that end doesn't leak).

For posterity, what I eventually figured out (pretty much entirely by feel and experimentation, as eyeballing it doesn't help much) is that it all lines up best with a single bolt in the factory L bracket 930.110.223.02 (that lets the I/C bolt up to the coil bracket)..

This guy:



I can get the bolt started in the extreme left of the slotted hole easily. Then just gradually tighten down, maintaining alignment.

If I try to get a bolt started in the right-most slot (the open-ended one) in that bracket, moving the I/C over enough (literally, just a few mm) to allow the bolt threads to engage seems result in enough misalignment to cause the issue I saw.

It "looks" good afterward. But the charge pipe just doesn't seem to seal correctly with both bolts in. It clearly leaks boost at lower pressures (boost doesn't build as fast), and as you use more boost there's an audible leak. Eventually, this will result in it spitting out the o-ring. Or even the i/c outlet popping off the 3.2 throttle body.

On the other hand, with everything sealed tight and bolted down with only that one bolt at the front and a 0.8 bar spring, it's a different animal; boost builds much faster - enough for the BOV to vent on every shift over 3K RPM, even if you didn't think you were using enough welly to go positive in the manifold..

[Alan L]

i'm thinking you have answered your own Q. When you rock a circle on an angle you create an oval. I think you have a misalignment which is possibly even squeezing the o ring out of place. Then you build pressure behind it and it wants to push out via the gap.
I think you need to go back to the beginning and see if there is a better alignment. Some bits you can't change - eg the TB throat. So everything has to follow back from there. Remember you can alter the turbo end also.
Good luck.
Alan

[spuggy]

Originally Posted by Alan L

i'm thinking you have answered your own Q. When you rock a circle on an angle you create an oval. I think you have a misalignment which is possibly even squeezing the o ring out of place. Then you build pressure behind it and it wants to push out via the gap.

Yup, agreed... If the O-ring were supported evenly all around, it would take ridiculous amounts of pressure to blow it out. So it must be that the gap is uneven - and then pressure is concentrated on a section of o-ring that's unsupported.

It seems to hold together (and seal) pretty well the way its aligned now, though. Only took taking it off/putting it back on/running it a dozen or more times...

And now I have to start over with the boost control parameter tuning. Sigh.

Originally Posted by Alan L

I think you need to go back to the beginning and see if there is a better alignment. Some bits you can't change - eg the TB throat. So everything has to follow back from there. Remember you can alter the turbo end also.
Good luck.
Alan

Yeh, there's only a very limited range of adjustment at the turbo outlet - it still has to poke through the hole in the engine tin I'd say my outlet is as close to vertical as dammit.

As for the charge pipe, the 3" flange on the I/C inlet is something like 15, 20 degrees off horizontal - so the angle on the top of the charge pipe needs to match that. Which pretty much means that there's only one "correct" way to fit that (with the flange sloping down towards the rear cross-member).

Does anyone have a good measurement of the outside of a K27 outlet adapter? Best guess I can muster is 55mm, just because i can't get calipers down in there to measure exactly. But 2 1/4" would be 57mm, which seems more plausible.

I've been browsing the offerings from HPS with increasing determination; they have a range of flexible couplings/reducers that seems like that might be able to be cut down and do the job in the space available...

[Alan L]

One reason I asked for a pic is to see the whole assembly. I have silicon tubing between my turbo and up-pipe and up-pipe and IC. No O rings - that allows for a degree of misalignment. Only O ring is on the TB. And if I am off line on that it tries to squeeze the O ring out.
Alan

[spuggy]

Found this looking for something else, realized I never followed up when I found the issue.

Originally Posted by RarlyL8

when using the OEM up-pipe with the dual O-rings it is important that the pipe be tightly secured at the fan so it cannot move. Also important that the intercooler cannot move. The few times I had a problem that was it.

Ding-ding-ding. Both Alan and Brian were right. Winners.

So, yeh, the clues were all there in front of me the whole time. Factory up-pipe was/is tight on the fan side (pretty much only fits one way round - and then it's tight/right), only the throttle body o-ring was getting spat out.

There was this manky-looking piece of vintage-looking bracket used as a spacer on the intercooler mounting tab on the back (throttle body) side.

I had put this thing back multiple times, and eventually discarded it after checking that everything lined up fine (or better) without it.

What I hadn't realized was that, without that spacer, the bolt was actually 5mm or so too long for the hole in the 3.2 manifold it screwed into. So what felt like the intercooler bolting down hard on the manifold was actually the bolt bottoming out in the thread instead.

Leaving a 5mm gap under the intercooler mounting tab. D'oh!

A casual shake of everything didn't reveal movement of any concern. But given enough boost, the intercooler mouth was moving off the throttle body, tipping it enough that the gap around the throttle body was uneven - and it would spit out the o-ring.

Quick trip to a hardware store for a shorter bolt later, the intercooler was really bolted down hard to the manifold as intended - and hasn't spat out a seal since, even with accidental forays into much more boost than intended (you really need a dyno to tweak/dial in boost control PWM tables).

Goodbye persistent boost leak I mostly didn't even realize I had, you will not be missed...

[Bucketlist]

Originally Posted by Alan L

One reason I asked for a pic is to see the whole assembly. I have silicon tubing between my turbo and up-pipe and up-pipe and IC. No O rings - that allows for a degree of misalignment. Only O ring is on the TB. And if I am off line on that it tries to squeeze the O ring out.
Alan

Do you have a pic of your setup? I need to take mine apart this winter and may be ready to modify the up pipe while i'm in there. I've had problems cutting o-rings in the past.