Capacitor Container Failure

I found an interesting puddle of liquid under my Tesla Coil this
evening.  I really hoped that it was water, but I wasn't that lucky.
Turned out to be good ole Shell Diala-AX.  I removed my rolled LDPE
caps, started spreading the kitty litter, and took the caps to the work
bench.  My capacitor containers are the typical 6" SDR sections with a
regular flat end cap on the bottom and an end cap modified with a
window, terminals, and vac attach fittings on top.

The failure turned out to be a hairline crack in the side of the lower
end cap.  The crack extends 1.5" vertically from the base.  It doesn't
appear to go along the bottom of the endcap, but the paint may be
obscuring it.  I am puzzled by this failure.  This doesn't seem to be a
highly stressed area.  Being on the side, it is twice and thick as
anywhere else and definitely didn't see the stresses during vacuum
degassing that the bottom did.  My only guess is that the PVC cement
softened it and somehow weakened it.  The coil (6" sec, 12KV, 120ma, 50"
arcs)  has been run  5 times and the caps are only a couple of months

Has anyone ever seen a rolled LDPE cap with a PVC container fail in this
manner before?  Is painting PVC considered a bad practice?  Are there
some crazy PVC glue dynamics that I need to be aware of?

Ross Overstreet
Huntington Beach, CA
ICQ# 20762411