Stag,
I would actually expect case stretch, datum-to-boltface, in high-pressure cartridges like the 350/5.56: firing pin strike drives the case inward to max seating, pressure glues the forward/thinner sidewalls, and the case head stretches back to boltface.
That the stretch and/or case movement back to the bolt could then produce "statistically-feasible" uncovering of a weak enough portion of the case web to allow gas breakthrough into the action... is troubling to say the least.
FWIW: I ran QL for the 180 Speer at 42ksi for both Lil'Gun (incipient failure/"belt" produced) and Norma-200 (
totally normal). Lil'Gun's rise-time is peaks earlier relative to bullet movement, and remaining pressure as the bullet moves past the port is lower -- yet still that powder produced the anomaly.
Truth in lending: The Lil'Gun anomaly occurred using the LEE 358-200
cast at 42ksi indicated -- and measured velocity was waaaay outta whack* to where I couldn't even jigger Lil'Gun's burn parameters to match it at all -- but the incipient failure mode was the same as yours.
... bothersome that it could be at luck-of-the-draw bolt/chamber/case design (stacking) tolerance levels..... Yeah, cast seals better/a bit higher pressure resultant, but not totally-off-the-charts higher* as indicated by the LABRADAR on two successive shots -- only one case of which "belted" on me.
* I
still can't explain why I (and you) saw those velocity/pressure spikes