Thought I would chime in and clear some of the questions here. I apologize in advance to be so long winded !
First of my background is 15yrs as engineer specializing in welding metallurgy
Annealing - to say that you require annealing with a GTAW ( TIG ) or GMAW ( GMAW ) for this application ( carbon steel thin wall pipe ) is simply incorrect. We have clients welding 2" thick mine car wheels with the MIG process and are in service in as welded condition . These wheels look like ovals when they are running. No issues
Almost all carbon steel and stainless welding on cars / trucks is done with GMAW in as welded condition ( no annealing )
Heat Affected Zone Cracking - although at first glance at your pictures we may be able to conclude that welds are "heat affected zone cracking" which is delayed hydrogen cracking which causes brittle HAZ but this is not the case
Reason is GMAW / GTAW process is a low hydrogen welding process. Base material most likely does not have high Ceq ( like an alloy steel ). High quench rate is not possible when welding since material is very thin
Therefore the none of the 3 criteria required for true HAZ cracking due to brittle weld are being met
Oxy Acetylene Welding - there is a very common misconception that a oxyfuel weld is not as brittle as a proper TIG and therefore more forgiving. This is simply incorrect
Please follow the following links to the "certificate of conformance"...the legal document that welding filler material manufacturers require to show their products meet the requirement set forth by the American Welding Society
This is a typical oxy fuel deposit. Please note elongation
http://www.mylincolnelectric.com/LEE...mes=US9025.PDF
This is typical GTAW weld using the proper ER70S-2 filler material
http://www.mylincolnelectric.com/LEE...5_20090522.pdf
As you can see the GTAW filler results in a much higher elongation property. In general terms we can conclude that the GTAW deposit is far less "brittle"
Possible Problem ( highly unlikely however )
- often people who use the GTAW process use the incorrect R45 / R60 filler ( same filler used for oxy acetylene ). You cannot tell the difference looking at the filler. The filler is typically coined or flagged with its type but not always
- the use of oxy acetylene filler with TIG will result is a very poor , highly contaminated and brittle weld
- I see this on a regular basis out in the field. Having said this its highly unlikely Superformance would make this mistake
The most likely problem has already been suggested by others
Quite simply the tubes are experiencing far more stress than the wall thickness can stand.
Evidence that supports this - welds are
both in the weld centerline and adjacent to weld. This is typical evidence of a overly stressed weldment NOT OF A BAD WELD. The fact that some of the welds cracked in HAZ is simply coincidental
Why is the flange overly stressed -
As already suggested by others, the high temp creep strength of a carbon steel is relatively poor. If , as others have suggested you are running excessive EGT, the yield strength takes a nosedive
This is , why catalytic converters ( higher temp than the rest of the exhaust ) is typically a Type 409 stainless. Obviously the best would be a high nickel material such as Inconel 625 but that is very cost prohibitive. This is why you only find Inconel in high end race cars
Conclusion - most likely conclusion is a
combination of loss of strenth due to high EGT, combined with stress and cyclic loading
My SPF sidepipe hangers have both cracked. Although the design of the hanger should hold and withstand the vibration, they both cracked at a stress riser ( concave weld profile )
It does not take much of a stress riser to cause a failure in a weldment
We do lots of robotic welding of exhaust systems and all the in service failures are not due to bad welding practice but rather weldment is overstressed pass design specifications
Since I stare at welds daily, send them for chemical, mechanical testing etc, I am fairly qualified on discerning between a good and bad weld. I am pleasantly suprised on all the chassis welds of the SPF. They all look pretty darn good !
If anyone is concerned about possible start of cracking on your headers / exhaust the most cost effective is purchase a can of Magnaflux dye penetrant ( don't forget a can of the developer as well )
Hope this helps