There is another piece of information that can be added to help make the true root-cause determination more likely, and that is the analysis of the oil/grease and debris found in the failed bearing.  While an outer race can crack due to significant weakening after a substantial amount of material has been removed, that is not the case here.  It is more likely that a significant piece of the broken cage got trapped between the roller and outer race and caused deformation to induce the crack.  The looseness due to the loss of engagement of the taper sleeve is much more significant than looseness created by heating.  It is most likely that the lock ring was not properly engaged or it fatigued from multiple uses.  (If previously someone removed the lock ring and re-bent the tangs to lock the nut, they would fatigue).  This might correspond to the retention of one of the tangs in the slot of the lock nut.  In the image, there is a very long extension line for the grease supply, while it appears that the bearing is fairly accessible.  When large bore grease extension lines are used, the grease can take a very long time to get to the bearing, and in a higher temperature or vibration environment, can separate in the lines and be dried out before reaching the bearing.  I would want to disconnect the extension tube from the bearing, and pump some more grease into the line and capture the grease exiting the extension tube near the bearing using ASTM D7718 as guidance.  The condition of that grease can be compared to the new grease using ASTM D7918 Die Extrusion method.  Only about 1 gram of grease is required.  I would also inquire about the procedures used to install the bearing.  Is there direction not to re-use the lock rings?  Is it clear that all of the tangs need to be bent into position on the lock nut?  Is the method to achieve this spelled out in the procedure?  If this was installed correctly, it is unlikely that any other failure mechanism would have resulted in the loss of the lock ring and loosening of the lock nut, without much more severe damage to the bearing.  The crack on the raceway could also be due to unusual forces imparted on the bearing by the looseness of the assembly, but I would also carefully check the fit in the housing against the outer race dimensions.  If the outer race was properly seated and fit into the housing, it would take very significant forces to create the crack.  If the outer race fit was originally poor, that could allow the unusual loading that occurs when the taper sleeve loosens up to create the conditions leading to the crack, in addition to the damaged race bits being run over by the rollers.

That's a really good point rich, to be honest I hadn't even thought about it, I should have really collected some of the old grease for that purpose. I'll remember that for next time. Yes so the grease line has been installed due to the access restrictions, it looks like it's achievable to just grease normally but it's actually about 7 feet off the platform, making it hard to reach. The grease line was checked after to make sure there wasn't any blockages. But I understand you're point , if there was a more direct way then we would be doing it to ensure the bearing was being lubricated.

The main thing is that during the installation of the adaptor sleeve no feeler gauges were used, this could have resulted in the bearing being overly tight , reduced the gap between the rollers and the race. 

Is there direction not to re-use the lock rings?  No there is not, and that's what I believe is a cause for concern, the Engineers are fire fighting breakdowns daily and I think that maintenance procedures are being ignored.

Is it clear that all of the tangs need to be bent into position on the lock nut? Again this is not being done, only one tang is nipped down. 

  Is the method to achieve this spelled out in the procedure?  From what I've seen and experienced it's not something that being implemented, that's what I will have to get them to start doing.

You're end point is very interesting and I will have a think about that because it makes alot of sense.

Cheers 

Jim P

Good catch with the feeler gage issue.  The bearing being improperly set up with insufficient clearance creates the hoop stresses needed to crack the outer race in the way that you showed here.

Rich Wurzbach

MRG Labs