Garrett TiAL housings? If they're still using M6 turbine housing retention hardware, you need to run Inconel bolts, EDM drill the heads, and run Inconel safety wire, and you'll still want to go through it every 6 months and tighten everything up. BTDT.

 

Next time try new hardware and some form of positive locking (wire, for example) and torque the hardware to only say 10 to 12 ft*lbs. This will not stretch the metal as much, but when everything heats up/expands, it will tighten up from thermal expansion.

The goal being to tighten the bolts enough that it's not loose or leaking when cold, but not too tight so that you don't stretch the hardware on the track.

 

It was suggested to drill into the manifold/studs and secure them with set screws. Any thoughts about this option?

 

Assuming you are using studs and nuts, you don't need to secure the studs. Just drill a hole through the nuts and stud where it intersects, wire them together. Then they can't move.

Wire is how critical fasteners are locked in industry. (brake rotors, air plane parts, things like that).

 

Tell us more about drilling small holes through Inconel with a hand drill, please

 

I would not recommend trying to drill inconel. Use steel studs, those can be drilled with a hand drill.

 

That wont work. **** son, you're on page sixty god damned 6 of the epic stud loosening thread recommending steel studs.

 

I ran grade 5 3/8" steel bolts, wired them together, torqued to 10 ft*lbs, and they worked from 2008 till I took that turbo off in 2013, running 17 PSI up to 28 PSI on a GT3271

So I understand that nothing I posted will work, but in my experience, it worked for me. OP said he can't buy what he wants and was looking for an alternative. I posted one.

 

I think the phrase I'm looking for is "textbook engineering myopia"

 

Explain. You like throwing insults, go ahead.

 

I posted this video six years ago in this thread.

 

Pat, you are wrong in this case because you dont understand the reason these things fail. Read the thread.

 

What did you torque the nuts to before you wired them together? It's very important to *lightly* torque them before wiring them together.


I do understand why they fail, the fasteners are being stretched. I have one of those paper slips that says something about engineer I think.

Hence I said to lightly torque the studs/nuts before wiring them together. This will only stretch them a small amount when cold, and give them the best chance to not permanently deform when under high temps/loads.

I did not say just install steel studs and torque the **** out of them and there you go, obviously that does NOT work or you guys would not have a 6 yr old thread on this.

Again, I drove a boosted miata for years as mentioned doing what I wrote with success, infact I used bolts instead of studs. So while you all know more about this than I do, I have a solution that worked for me, for years, with zero failure so I posted what I did. Feel free to now tell me it won't work.

 

Unfortunately your solution didn't work for anyone else, or this wouldn't be the giant thread that it is throughout all these years. In fact, I tried that exact approach on the second set I installed, and it didn't work.

 

What are you lap times at a local road coarse? What are the spec miata times? Are you at least as fast or faster? If not then you were not driving your car hard enough to do the damage that we are talking about. Quarter mile times mean nothing.

 

What did you torque the studs to? Were they 8mm?

I just realized a lot of people run 8mm. I ran 3/8, which is basically 10mm equivalent. So it may very well not work on smaller hardware. 8 is so tiny that I'll say now, what I posted would not work for 8mm hardware. It would have to be bigger.

Just did some quick math, and 10 ft*lbs on a 3/8 grade 5 bolt is about 29% preload vs tensile strength. Typical is 75%.

 

Then my times mean nothing.

Deerhunter wanted an alternative. I posted what worked for me, and even explained why. I did it back in 2007 or 2008, and back then everyone told me it would fail too, but it never did.

Most people tightened the crap out of their bolts or studs because "they kept loosening" which is 100% backwards/wrong. You want them not torqued too much so they can stretch a lot under high load without permanent deformation. And use a form of positive stop like wire so they can not loosen.

 

My experience is like Pat's. I was able to street drive for several thousand miles on mild steel and safety wire. And I would consider auto-X or drag racing similar to street driving vis-a-vis sustained heat loading of the manifold/turbo connection.

Then I caught the track bug. I ordered TSE studs not long after that.

 

Same, I tried to beat the sh*t out of the car on the street and it held up fine. First track day I lost two studs in the second session.

 

I thought I'd update everyone on what's happened since my last post. Andrew was kind enough to let me know that he'd sourced a few more sets of his TSE 10 mm inconel studs and he rushed them over to me. Not having the time to futz around with the install, I dropped my car off at my local independent mechanic. With some trepidation (they know what a PITA it is to get the turbo off on my car), they dove right in.





All the important bits removed





Turbo is fine, with no significant play in the shaft and only some carbon tracking where the outlet neck is leaking





The FM 8 mm inconel studs were fine, although the EM threads were either stripped or well on their way





A close-up of the stripped threads. Since we were stepping up the stud size, repair was a relative cinch, only requiring the services of the machine shop across the street.





We discovered that the outlet neck was not only warped but also cracked all the way through. You can also see where it was warped (carbon tracking). Rather than repair it, this part was replaced.





All back together again