WojtekSz
Thanks for your YouTube post, it was enlightening. I would encourage everyone to look at the Part 2 of this series:
YouTube - How to Build A Jumbo Jet Engine - 2/4 (Full Version)
About half way through, the process used to make the turbine blades is shown in good detail. Both the HPT & IPT blades are made this way (single crystal) with the IPT blade being solid, non-air-cooled. Also, note the machining section. After I looked at this:
35Year Pilot
I take back the idea the blades were pushed out of the dovetail slots.
lomapaseo is correct, the disc stretched, releasing most of the blade dovetails from the corresponding disc slots. Also, the blades are shrouded at the tip and are interlocked by means of the z-form shrouds. This means that all the blades probably fractured near the platform upon contact with the LPT nozzle inner band ring. The machining tolerances of the dovetail forms are so tight, they couldn't be pushed out of the disc slots, unless the IPT rotor blades all came out uniformly in mass because of the shroud interlocks, an unlikely scenario.
DERG
I am not a RR guy so the Trent is a slightly different animal from what I am use to seeing as far as the details go. But, basic design principles apply regardless of manufacturer, RR, P&W, GE or EA.
bearfoil & WojtekSz
I am going to give you my thoughts on the Trent 800 vs Trent 900 and IMO, what I might think of the Trent 900 design in a new post.