Its a classic fatigue crack, you can tell by the “beach” ripples on the crack surface before the last part failed in tension. The crack has been present and growing slowly for some time hence the rust.
The bottom of my rear strut looked the same as yours and I just brazed a new end plate on the cap as it had worn a hole when my hub shaft broke on the track at about 120kmh
cheers
Rohan
I agree with Rohan. It is a fatigue failure that seems to emanate from the thread root and then has progressed across approx 80% of the area and finally failed in overload. The thread should be in tension such that any bending stress is lower than the tensile stress, ie it won’t fail in bending. The failure indicates the tensile stress was lower than the bending stress, ie the nut was not tight enough, as you have stated.
If you count the beach marks it will tell you the number of cycles to failure, but probable getting a bit too techy here!
As the nut is also under tightened on the other side it is highly likely that the thread is also cracked. If not it will be fatigue damaged and have a reduced life, so simply tightening to the correct torque is not the answer. It should be replaced.
If the taper is not accurately seated then it can later ‘bed in’ reducing the torque loading and thread tension. Best to retorque the nut after 100 miles or so.
Thanks to all for the feed back and guidance.
I am ordering TTR billet shafts, extra strength hubs, bottom wishbone, bearings and will either repair or replace the upright.
Will make sure the hubs are well lapped to the shafts.
Will make sure hub nuts are torqued correctly and then check frequently during the first few runs out.
Thanks
Martin
I’m learning things here that I hope I never need to know. How would it be possible to spot the crack before failure?
I think you’re unlikely to get cracks if the nuts are torqued up correctly, and wheels coming off is very rare. However, if you really want the check you need to get a None Destructive crack detection Test (NDT).
This is standard industry practice but you can probably buy the chemicals to do yourself. The process is to clean and dry the area, then apply a dye penetrant, then apply a developer. The penetrant enters the crack and then shows the crack when the developer is applied. No indication means no crack. I’m sure someone could do this for you.
This is what I use on any components under load and liable to crack. You will find it or similar products at most industrial supply stores
rocol.com/products/flawfinder-kit.
The biggest challenge with the rear hub shafts is you need to have the hubs apart to use it. It is also challenging to get a true indication of a crack at the root of threads but possible.
cheers
Rohan
Dye penetrant as above is a great diy tool if you follow the instructions carefully. It’s also easy to cause misinterpretation of the results if you don’t follow the directions exactly.
Magnaflux (trade name) is a dye penetrant process that magnetizes the part and then shows possible indications under a UV light. Much smaller defects (indications) can be detected. Most good automotive machine shops can do this for you for a modest cost.
The important part of the threaded portion of the axle is the radiused transition from the threads to the face of the shaft at the taper. The radius helps prevent the formation of stress raisers between the threaded section and the tapered section of the shaft.
Imho you need a friend in the Aero Space Industry in the QC department who can test and interpret correctly.
It’s not a diy job
Alan
I’m sure there are people out there with the skills to weld repair that but agree that as replacements are available that’s the way I’d go. I actually expected it to look worse than it is given the speed you were doing, but I guess that’s one of the benefits of a light car.
Definately a good trophy to hang on the garage wall! 
Too bad lessor priced units aren’t being made. Thou maybe junk if they were
I think i would get it welded but i’m lucky to have a friend who is a welding specialist.
Alan
