Car exhaust bracing - keeps falling apart!

[dean0]

Hi there,

Not sure if this is the correct section (or forum) for such a question but I'm after the advice of someone who knows their stuff about car exhaust systems. I'll get some pictures up over the weekend too.

I've got a Mazda Mx5 fitted with a (badly designed) aftermarket turbo kit. The turbine outlet has a cast iron elbow which leads into a downpipe which goes down into the transmission tunnel and then curves to run parallel with the driveshaft. On the horizontal section there's a flexi joint.

I've been having lots of trouble with the nuts which hold the cast iron elbow onto the compressor outlet. They seem to work loose after hard use (trackdays). I think it might be because the downpipe length before the flexi is quite long and therefore acts as a lever on the turbo. The nuts holding the turbo to the manifold and elbow to downpipe are fine, but are in a different plane.

I've thought of bolting the downpipe to the engine block to stop this levering action. My question is; is it better to clamp the exhaust between the turbo and flexi or just after the flexi? I think before flexi but I've seen both done. I know the flexi should be on the vertical section but I don't want to get into major exhaust fabrication just yet.

Also, because of clearance issues with the elbow the studs are so short that I can't use the nuts with locking "dimples" at the top as the stud thread only just reaches the top of the nut. I've found squeezing the nut slightly in a vice helps but was wondering if there is a better solution out there anyone is aware of?

Thanks,

Dean

 

[the snooper]

i would say that you should bolt the down-pipe to the engine before the flexi otherwise it would be working against the flexi re the nut issue have you tried the various types of locking washers available or maybe brass nuts

 

[rikrobson]

Are your nuts the correct thread for the studs? If you are squeezing them and making them slightly oval then they will grip in the narrow section. The heat from the turbo exhaust will then soften them and they'll become loose and the flex back negating the squeeze from the vice. It could be that they are slightly the wrong size and as they are only on by their own thickness you may not notice. other than that are they getting the correct torque when tightening? I would usually expect the full thread to go through the nut as you'll not get the full benefit of the thread as the to of the stud will taper slightly. Maybe worth getting a set of brass nuts. you may be able to get nuts with a closer tolerance for the stud's thread also.

The nuts shouldn't loosen due to leverege. It the down pipe is rigid to the turbo you'd be better making a bracket to brace to the turbo mounting.

 

[BillJ]

Had the same problem with an aftermarket downpipe on a Saab 9000. The vibration was rotating the turbo/downpipe joint and loosening the nuts. I braced the downpipe to the engine before the flexi and had no more trouble.

You want the engine/turbo/downpipe all to move as one piece, not relative to each other, so keep the flexi out of the equation. The original downpipe on my car was braced to the engine before the flexi joint

 

[penfold]

What you want is stiff nuts, other online auction houses are available...

 

[BillJ]

Despite the claim not to be affected by heat, I've found that they are. At least in the sort of heat you get in turbo applications.

 

[MrFluffy]

what about a old fix for the problem and just using two nuts per stud? when you tighten the second nut up it stretches the thread between it and the first nut and forms a heat resistant locking mechanism.

 

[tigler]

Normal practice is to brace the down pipe to the block. I would also suspect that the fasteners causing the problem are just normal high tensile which is not correct for the application. They should be a high temperature stainless steel , I can't remember the spec off the top. Should you not be able to source the correct type you could use these , http://www.nord-lock.com/

 

[Flathead_Dave]

Can't you lockwire them?

 

[hotrodder]

The way i read the OP is that the studs for the downpipe flange are too short and replacing the studs with longer ones will make assembly impossible so the common all metal locking nuts are out as are double nuts. I'm wondering if there's sufficient stud length for k-nuts to work as they're shorter than the usual cheaper self locking nuts

 

[Rick]

Lookwire is the only thing that'll keep them on perminantly IMO.

The housings warm up quicker than the studs, so they work loose K-Nuts are good though.

 

[tigler]

In my experience lock wire is not effective on this sort of application. Lockwire is more a safety method than an actual locking system , for stuff like drain plugs etc.
I have found that on turbo installs the fastener must be of the correct material. Jap cars use a special grade of stainless for bolts used on turbos.

 

[jimi_boom]

Hi Dean

If it was my set up i would simply add in a sepcial typ eof washer that is normally used for just such an application.

If you google Nord-lock they have a very effective type of washer that grips anything and can be exsposed to high vibration without it coming loose, i have used them before and they are great.

Have a look and see what you think

Cheers

 

[tigler]

I already mentioned those dude !!

 

[TheJoker]

Nord lock washers are amazing.

 

[jimi_boom]

Sorry Tigler i was skimming the thread and totaly missed the link at the bottom of your post.

But yes joker they are amazing. i met the rep and he had a few samples in his bag of tricks that you simply couldnt un do by hand, good stuff

 

[tigler]

They are one of my favourite things

 

[dean0]

Thanks for the replies everyone, there's plenty of suggestions to keep me going for a while and most of them sound pretty cheap.

As promised here's a picture of the offending item:

The join which keeps on coming apart is the one just to the right of the wastegate. I'm not sure it's a heat issue as I'd have thought that would affect the joint between manifold and turbo worst.

I'm going with BillJ's suggestion first as that seems to be a solution used by a lot of the guys building high powered volvos over on the turbo bricks forum. A lot of people on the miata turbo forum have tried lockwire and found the studs work loose instead or if lockwired in pairs they still work loose enough to blow.

The studs and nuts I'm using are m8 and were bought as matching set when I had to have the last studs drilled out of the turbo (they sheared). So I think they should fit properly. As you can see the nuts are tight against the cast elbow so I can only get an open ended spanner on it. I grease the threads on assembly to ensure they are torqued up enough.

If the brace doesn't work I'll try the K nuts, but I think stud length might still be an issue. If they don't work I'll think I'll have to get longer studs and grind some clearance into the cast elbow.

If that fails I think I'll see if I can get some oem studs/nuts for a turbo charged car.

 

[tigler]

If you are using standard high tensile carbon steel studs and nuts that is your problem. Standard high tensile steel studs will lose their strength at the temperature that the exhaust outlet reaches and then just stretch and loosen.
Get some OEM originals which will be off the correct material. Don't use normal grease when assembling , use never sieze or copper grease.

 

[dean0]

I made my bracing bracket and so far I've managed one autosolo without the studs working loose. That's fairly good considering the turbo was hanging off half way through the last time I ran.

Here's the results of my work, excuse the terrible welding, I've covered my shame with lots of black paint. I made it out of 3mm steel using a hacksaw, file, pillar drill and vice.

This is part of the unpainted bracket hanging on the downpipe during a test fitting

The two parts together, left part bolts to the bellhousing, right part ties the left part to the downpipe.

In place on the car. The two pieces mean there is some adjustment available for fine lining up of the components.