Much has been said since my first post regards the building. And since then I've seen pics of the machine's removal which widen my understanding of the building. "robdean", I am not an engineer, just a builder, but I have done quite a bit of self-study on calculated strengths of building materials and in failure analysis when they let go. I have much experience in construction and repairing buildings, and I have personally seen many damaged structures but none quite like this. I know what math to do and how to do it regarding strength and failure. It's complex, tedious, and coincides with my experience so I'd rather not go there myself since the investigators are going to do it anyway. All I post here will be based on my experience alone which I can prove through valid calculations if necessary. This is going to be a long post, my apologies for that but it's necessary. If you can't be troubled to read it all, please read the italicized parts.
The pics I saw showed the newest roof used lightweight bar joist construction. The strength of this kind of roof is well known to me, and it alone would not have held the static weight of this machine. "Technet101" notes this. I'll get back to this part later, it's important.
The pics showed another older roof structure underneath. "G-CPTN" noted this, and that makes it plausible that there would have been a delay between impact and the machine dropping to the floor. We know this happened. And with this being a wooden structure, my earlier post concerning staged slow failure seems to have been correct after all even though there was much more structure here than I anticipated. What would happen with and excessive load on this roof is the beams would deflect, the wood fibers stretch until they lost molecular adhesion with surrounding fibers, then they would slip apart. This would occur from the bottom of the beams upward, the upper sections initially being under compression then going to tensile extension as drop induced stretching. Everyone has seen this in the frayed ends of broken wood. Under slowly applied weight this would occur slowly, but under impact would occur almost instantaneously. More on this later too.
The floor beams under this was quite a surprise to me. They would have been the strongest framing members involved and they are what did the most work here. It is very likely that there was direct contact between this and the roof members above, causing this part of the structure to act and fail in unison with the old roof above. This structure together could possibly held this machines weight- I feel that it's likely. Yet the machine still broke through here and that seems to indicate some substantial downward velocity at impact with this part of the structure, causing partial failure of the wood, thus weakening it allowing further failure to occur. Said more simply, the wood cracked but did not initially break, and this explains the delay before the fall to the floor. This slow rate of failure is what saved the most lives below. (It's also the basis of why wooden wing spars work so well in light FW aircraft, within their ultimate limits they will 'take more punishment' without distress than metal does before ultimate failure occurs). Back to the end of my last paragraph, from this failure rate occurring as slowly as it did, one can reasonably ascertain the force needed to cause this failure, and one can reasonably calculate the rate at which this force was applied.
"Chronus" has posted some figures regarding the force of impact - I'll take them as being accurate. In this we can see that if the entire structure could have held the static weight of the machine as I believe it would have, then there had to be more energy involved- the added force from mass in motion. Now we get to the tricky part- the bar joists I first mentioned. In the pics, I saw where these were perpendicular to the machine and also appeared to be severed. On impact they would have twisted under load, sagged downward as strength was lost with their structure deforming, then snapped through necking when the ultimate tensile strength of the components they are made of was exceeded. This would have slowed the downward momentum, acting as slings under the machine, until they parted. And with this support being applied only at intervals and not across the whole area under the machine, it is probable that one of them was under the tailboom at it's point of separation, causing it to part from the cabin section at that point. With the downward momentum slowed as this all happened, the bar joists would have abated some of the force which was thereafter applied to the structure underneath, allowing it to hold where it might have instantly failed otherwise. Had these joists been parallel to the machine, it would have slipped right through immediately. The roof members under the bar joists were oriented similarly, also reducing the chance of this 'slip-through' scenario.
Knowing the strength of wooden floor and roof joists which I saw some of in the pics, knowing how the bar joists affected the impact, and knowing the approximate weight of the machine, it seems clear to me that there was a free-fall drop from considerable height involved here, perhaps not from the 300ft "Chronus" calculated but still quite a lot regardless. It would probably have held intact even with a free-fall of say 10-20 feet. This seems to agree with there being no significant MR blade damage from rotation and witnesses to the event saying there was no rotation. This would all imply that nothing was done to stop the fall, which would certainly have happened had there been enough time and any ability left to do that.
The practice of doing a roof-over is purely economic, but in this case it also gave a cushioning effect and prevented a larger catastrophe. I'll never call them a cheap way out again. This was critical in saving lives underneath, but it also likely prevented the crew from leaving the cabin were they otherwise able to do that. Much had been said of the designed resistance of g-force on impact of these machines, but we need to understand that this is calculated based on a flat, relatively unyielding surface like the ground and this was an entirely different scenario where the forces were applied to the machine locally and not generally. With these impact forces being concentrated, unless a seat was directly over a joist the designed impact resistance could not have functioned as well as it was designed to, and there being more space than joist here, this arrangement would be very unlikely. No fault to the safety design engineers for you can't foresee every possibility and this was a very unlikely one.
Sorry to have bored you with all this, but it does explain and seemingly clear up several points of contention being discussed here, which I hope will bring about a better understanding of this tragedy. In my analysis of the bar joists and flat roof construction in general, I hope the pilots here now understand their danger to you should you ever be so unlucky as to need somewhere to set down right away. As buildings go, the Clutha was a great anomaly, not at all usual by any stretch of the imagination, and this along with the orientation of the machine on impact certainly saved many lives inside, even as it may have also doomed the crew. Having nothing more I can add I'll drop back to read-only mode in this thread now. Thanks for allowing me in your presence.