EPIC nuts/studs loosening thread (reposting stupid stuff without reading = warning)
#462
Elite Member
iTrader: (15)
Join Date: Dec 2007
Location: San Antonio, Texas
Posts: 4,847
Total Cats: 27
Pictures. No resbond yet, just a 'dry' fit
Picasa Web Albums - John - through bolt ...
Damn it, I hit the wrong key and wiped out my commentary of the pictures in the above link. I hate that. I could retype it all but I think I will have another beer instead.
Picasa Web Albums - John - through bolt ...
Damn it, I hit the wrong key and wiped out my commentary of the pictures in the above link. I hate that. I could retype it all but I think I will have another beer instead.
#463
Moderator
iTrader: (12)
Join Date: Nov 2008
Location: Tampa, Florida
Posts: 20,675
Total Cats: 3,017
Tex,
That's beautiful and looks like it took a lot of work. So I am reluctant to say I think I see a possible problem. I think by going with through-bolts instead of studs you might have inadvertently increased the area of expansion that will be trying to stretch your bolts.
In the picture the distance in red would be the thickness that would be stretching your studs if they were threaded into the manifold. The distance in green is the amount of material that will expand and work on a through-bolted nut and bolt arrangement. I hope it doesn't matter much, but I'm concerned that it might.
That's beautiful and looks like it took a lot of work. So I am reluctant to say I think I see a possible problem. I think by going with through-bolts instead of studs you might have inadvertently increased the area of expansion that will be trying to stretch your bolts.
In the picture the distance in red would be the thickness that would be stretching your studs if they were threaded into the manifold. The distance in green is the amount of material that will expand and work on a through-bolted nut and bolt arrangement. I hope it doesn't matter much, but I'm concerned that it might.
#466
Moderator
iTrader: (12)
Join Date: Nov 2008
Location: Tampa, Florida
Posts: 20,675
Total Cats: 3,017
True, true. I remember but... I dunno. Maybe you're right. I hope you are. It would be nice to have a solution. I would have to ditch my cast manifold and build a fabricated one, but it would be worth it if it worked. Good luck.
#467
Elite Member
iTrader: (15)
Join Date: Dec 2007
Location: San Antonio, Texas
Posts: 4,847
Total Cats: 27
There is another aspect of this too. The material added to the joint (the manifold flange) is the same material as the bolt. Assuming BEGI uses a 300 series stainless flange, the CTEs are probably the same (the bolts are 300 series SS, 316 IIRC). Therefore, if they heat the same amount, then they elongate the same amount. Following from this, since the manifold flange makes up a large percentage of the thickness of the joint (more than 50%), the overall differential thermal expansion is decreased, meaning less bolt stress if the turbine housing material does indeed have a higher CTE than the stainless bolt and flange.
#468
Roger Roger,
There is something intersting in what was just mentioned. Cast iron of the turbine housing will be between 5-7 microinch/F and the stainless steel will be 9-11.5 microinch/F
So according to this in a steady state test the studs will be relieved of preload which is clearly not causing failure. Therefore you have to conclude that the permanent deformation occurs in a non-stead state warm-up/cool down cycle.
Causes?:
The temperature gradient from the head. The exhaust manifold right at the head is probably no more then 300-400 degrees as heat is being transferred from the manifold and dumped into the coolant system.
Turbine housing is a massive thermal heat sink. So it will stay hotter much longer then the side closest to the flange. This just because of the bulk of the turbine housing. Some of these disproportionate cooling rates could be causing the issue.
Different thermal properties from stainless to cast iron in the way of specific heat capacity and thermal transfer.
Comments?
There is something intersting in what was just mentioned. Cast iron of the turbine housing will be between 5-7 microinch/F and the stainless steel will be 9-11.5 microinch/F
So according to this in a steady state test the studs will be relieved of preload which is clearly not causing failure. Therefore you have to conclude that the permanent deformation occurs in a non-stead state warm-up/cool down cycle.
Causes?:
The temperature gradient from the head. The exhaust manifold right at the head is probably no more then 300-400 degrees as heat is being transferred from the manifold and dumped into the coolant system.
Turbine housing is a massive thermal heat sink. So it will stay hotter much longer then the side closest to the flange. This just because of the bulk of the turbine housing. Some of these disproportionate cooling rates could be causing the issue.
Different thermal properties from stainless to cast iron in the way of specific heat capacity and thermal transfer.
Comments?
#470
Elite Member
iTrader: (15)
Join Date: Dec 2007
Location: San Antonio, Texas
Posts: 4,847
Total Cats: 27
Travis here is what I get from Matweb.com
For 316SS
CTE 9.72 µin/in-°F (@Temperature 32.0 - 1000 °F)
For 'ductile iron'
5.89 - 10.4 µin/in-°F (this is for the range of ductile irons)
For 'cast iron'
4.31 - 10.7 µin/in-°F (this is for the range of cast irons)
So, to your point, I think I agree, with the caveat that it depends on which alloy the turbine housing is made out of. Unless the alloy is at the very upper end of the range, it probably has a CTE that is the same or less than the 300 series stainless. Therefore, the differential thermal expansion may not be the culprit here after all. In fact, it is at least possible that Garrett picked an alloy that had a favorable CTE match with typical turbo fastening hardware.
I do not know why missed this before so thanks for pointing it out.
On another note... I was adding to my latest point above when my internet connection went down and I lost my addendum. So I'll get to it now. There is also the transient thermal condition to consider. In fact, the WOT on/WOT off nature of the track would tend to increase the frequency of the transients, if not also the magnitude.
Allow me to express this next point simply for the sake of a brief illustration. When going into WOT, the heat transfer path is: exhaust gases -> manifold and turbine flange -> bolts and nuts -> ambient air. So, the manifold and turbine flanges would tend to heat up before the bolts do. Even if there is a CTE match, if the flanges are hotter, they will expand more than the bolts will, and thus have the same effect as a differential CTE.
Now the quantitative analysis of this is pretty complex, so I cannot say offhand how significant a contribution this could be. One, we are talking about a transient condition, two, there are a lot of variables at play, and three, the heat transfer mechanisms are complex.
For 316SS
CTE 9.72 µin/in-°F (@Temperature 32.0 - 1000 °F)
For 'ductile iron'
5.89 - 10.4 µin/in-°F (this is for the range of ductile irons)
For 'cast iron'
4.31 - 10.7 µin/in-°F (this is for the range of cast irons)
So, to your point, I think I agree, with the caveat that it depends on which alloy the turbine housing is made out of. Unless the alloy is at the very upper end of the range, it probably has a CTE that is the same or less than the 300 series stainless. Therefore, the differential thermal expansion may not be the culprit here after all. In fact, it is at least possible that Garrett picked an alloy that had a favorable CTE match with typical turbo fastening hardware.
I do not know why missed this before so thanks for pointing it out.
On another note... I was adding to my latest point above when my internet connection went down and I lost my addendum. So I'll get to it now. There is also the transient thermal condition to consider. In fact, the WOT on/WOT off nature of the track would tend to increase the frequency of the transients, if not also the magnitude.
Allow me to express this next point simply for the sake of a brief illustration. When going into WOT, the heat transfer path is: exhaust gases -> manifold and turbine flange -> bolts and nuts -> ambient air. So, the manifold and turbine flanges would tend to heat up before the bolts do. Even if there is a CTE match, if the flanges are hotter, they will expand more than the bolts will, and thus have the same effect as a differential CTE.
Now the quantitative analysis of this is pretty complex, so I cannot say offhand how significant a contribution this could be. One, we are talking about a transient condition, two, there are a lot of variables at play, and three, the heat transfer mechanisms are complex.
#471
I have the software to do it, but I would need cad models from popular manifold manufacturers, and a 3-d model from garrett.
I think barring a 2 week research project by maticulously recreating these manifolds and turbine housings, and then running a thermal analysis with convective heat transfer outer, finding whatever convective constant the exhaust gas has, adding in the conductive effect of the exhaust system, head heat dump, getting before and after turbine EGT's to understand how the turbine extracts energy adiabatically from the gas, and then melting my processors trying to crunch that... The next best thing is just to try the solutions to the problems, and see how we come out.
I think barring a 2 week research project by maticulously recreating these manifolds and turbine housings, and then running a thermal analysis with convective heat transfer outer, finding whatever convective constant the exhaust gas has, adding in the conductive effect of the exhaust system, head heat dump, getting before and after turbine EGT's to understand how the turbine extracts energy adiabatically from the gas, and then melting my processors trying to crunch that... The next best thing is just to try the solutions to the problems, and see how we come out.
#473
Boost Pope
iTrader: (8)
Join Date: Sep 2005
Location: Chicago. (The less-murder part.)
Posts: 33,072
Total Cats: 6,625
Forgive me if it's been mentioned (I haven't read all four hundred and seventy-something posts in this thread) but I happened across this just recently:
Castelated Nuts / Saftey Wire Nuts BEGi
Related thought: If the studs themselves are turning in the manifold, why not cross-drill through the manifold flange into the stud and drive in a pin to lock them in place?
Castelated Nuts / Saftey Wire Nuts BEGi
Related thought: If the studs themselves are turning in the manifold, why not cross-drill through the manifold flange into the stud and drive in a pin to lock them in place?
#475
Elite Member
iTrader: (15)
Join Date: Dec 2007
Location: San Antonio, Texas
Posts: 4,847
Total Cats: 27
Yeah Joe IIRC there was some discussion on that idea in here somewhere. I mentioned I think some permanent assembly aircraft studs I have used before that are staked in place with special tooling. They are a permanent assembly at that point, and are used for example for aluminum housings where tacking is not an option. And, you absolutelymotherfreakingpositively do not want the stud to come out again, ever, like on the critical aircraft actuator I was working on... I think the hope is that the resbond threadlocker will take care of at least the stud retention issue.
Travis I agree with you on the complexity of the model. I have done some thermal modeling, but not one like this. Sounds like you have.
I also think sometimes it is better, even easier, to just try out the actual idea than to model/analyze it. I think this is one of those cases, especially when there are so many here that are trying out or going to try out so many different approaches. Ain't nothing like the real thing.
And, I know all the non-engineering types reading through this must be thinking, what in the hell is such a big deal about a bolted joint?
Travis I agree with you on the complexity of the model. I have done some thermal modeling, but not one like this. Sounds like you have.
I also think sometimes it is better, even easier, to just try out the actual idea than to model/analyze it. I think this is one of those cases, especially when there are so many here that are trying out or going to try out so many different approaches. Ain't nothing like the real thing.
And, I know all the non-engineering types reading through this must be thinking, what in the hell is such a big deal about a bolted joint?
#476
Forgive me if it's been mentioned (I haven't read all four hundred and seventy-something posts in this thread) but I happened across this just recently:
Castelated Nuts / Saftey Wire Nuts BEGi
Related thought: If the studs themselves are turning in the manifold, why not cross-drill through the manifold flange into the stud and drive in a pin to lock them in place?
Castelated Nuts / Saftey Wire Nuts BEGi
Related thought: If the studs themselves are turning in the manifold, why not cross-drill through the manifold flange into the stud and drive in a pin to lock them in place?
Sav, do you have before and after pics of your safety wired hardware? What size/spec safety wire did you use?
Spooky or ZX, do you know a source for metric pre-safety wire drilled bolts or nuts? My only civilian source is Spruce and they're Std. Same with my military hookup. Though, I can get some big *** titanium bolts for a rotor head.
Chris
#477
Former Vendor
iTrader: (31)
Join Date: Nov 2006
Location: Sunnyvale, CA
Posts: 15,442
Total Cats: 2,100
Best I can do is the video I posted earlier, which shows all the attempts I made before pulling the rip cord. The safety wire is from Coleman, leftover from my karting days. Came in a blue plastic cylindrical container, like 500ft worth or something.
Last edited by Savington; 07-26-2009 at 02:56 AM.
#478
Tour de Franzia
iTrader: (6)
Join Date: Jun 2006
Location: Republic of Dallas
Posts: 29,085
Total Cats: 375
I have the software to do it, but I would need cad models from popular manifold manufacturers, and a 3-d model from garrett.
I think barring a 2 week research project by maticulously recreating these manifolds and turbine housings, and then running a thermal analysis with convective heat transfer outer, finding whatever convective constant the exhaust gas has, adding in the conductive effect of the exhaust system, head heat dump, getting before and after turbine EGT's to understand how the turbine extracts energy adiabatically from the gas, and then melting my processors trying to crunch that... The next best thing is just to try the solutions to the problems, and see how we come out.
I think barring a 2 week research project by maticulously recreating these manifolds and turbine housings, and then running a thermal analysis with convective heat transfer outer, finding whatever convective constant the exhaust gas has, adding in the conductive effect of the exhaust system, head heat dump, getting before and after turbine EGT's to understand how the turbine extracts energy adiabatically from the gas, and then melting my processors trying to crunch that... The next best thing is just to try the solutions to the problems, and see how we come out.