This is common to the extent it has been seen many times before. It could be
a result of an oversized O ring or a join in the O ring which took up all
the available space - there needs to be expansion space.
It could also be metal fatigue of an old but refurbished housing. It could
also be a casing flaw in a brand new housing .
George (down under)

The piece failed, and thousands of others just like it failed because the Japanese cannot be told they have screwed up because they would loose face before their ancesters. So nobody says anything at all. 

 

Was the engineer who did the casting design told there was going to be an "O" ring groove so close to the edge, within a MM of the radius? Maybe not. It looks to me like there was never any intent to put the groove in the iron. Was there a real good reason? You would think that the tooling cost alone would rule that out. Would you rather machine aluminum or cast iron? The groove in the rotor housings worked perfectly.

Why change perfect, and easy to do?

 

So if we were doing that piece would it ever fail. No. Your iron man would go to the sand shop with new drawings, adding just a bit more iron behind and along the outside of the ring grooves, You could have the patterns changed and back in service yet today. and that would be it. Distribute the new dash number drawings to inspection, remove the previous dash number and go on with life with no failure reports from the field.  Pull all the irons still in the building and have warehousing return all of the old P/N

to go into the next days melt.

 

Ever wonder why there are the remains of the peripheral exhaust ports in the Renesis (Side Port) rotor housings?  

As soon as that guy dies or retires, those will go away. 

 

Lynn E. Hanover