But there are two more-serious problems here.

One is the fact that the connections between structural elements tend to be stronger than the elements themselves, yet Sunder says they failed before the elements did.

However, a more fundamental problem is one of apparent scientific fraud relating to the collapse of Column 79, the column NIST claims initiated the failure -- the linchpin of NIST's "thermal expansion" argument.

NIST claims that heat caused the floor beams supporting the 13th floor near Column 79 to expand sufficiently to push the girder on which the beams were resting off its connection to Column 79, so that the floor fell, pancaking the floors below and robbing the column of support until it collapsed and fell, causing adjacent columns to fall.

Unfortunately for this conclusion, WTC 7's blueprints indicate that those beams were locked into the concrete-and-steel floor decking by the use of many shear studs. This floor decking wouldn't have expanded to the extent and at the speed NIST claims the steel beams did, and thus would have prevented the beams from doing so. So how could NIST claim the beams expanded? They simply omitted the shear studs from their model. This can hardly be considered accidental, in a model that allegedly modeled every element and connection in the building.

But it gets worse. The heat diagrams generated by NIST's model of the fires in WTC 7 do not match the photographs it provides. For example, this picture of the north side of the building was taken at 4:38, about 45 minutes before collapse:

Fire can be seen raging on this north side of the 11th floor, presumably heating the 12th floor slab. But take a look at the diagrams for this 12th floor slab at about the same time and later -- 4:30 and 5:00 (just before the collapse at 5:25):

This slab should be affected by fires raging at that time in the above photo. (Note: these diagrams are labeled 5:30 and 6:30, both of which were after the building collapsed; all of NIST's diagram times appear inexplicably to be labeled an hour late.) The north (upper) side shows no indication of having been heated by the blazing 11th floor fire in the photo. As to the inferno modeled on the east and south (right and bottom) sides, keep in mind that there is no photographic evidence for what would have been happening on these sides. (Note: the key Column 79 is the green post just up and to the left of the rectangular gray area.)

More importantly, we should consider the crucial 13th floor's floor slab, whose beams supposedly started the whole thing. It should be affected by any fire on the 12th floor. Here we have a photo of the north face of the building a little before 4PM:

Note that floors 12 and 13 appear to have been on fire earlier, but have burned out. Here are the heat diagrams for a short time later -- 4:30 and 5:00:

The earlier diagrams show fire spreading out from the lower right along the east (right) and south (bottom) edges, but never reaching the north side. Yet the photo shows that fires had actually been there, and then gone out. They burned out long before the building collapsed, and could not have contributed to any "thermal expansion" in the area. As to that inferno heat shown on the east side, remember, there's no photographic evidence for it. It appears to have been inserted into the model to make a case for thermal expansion that otherwise couldn't have happened.

In addition, people familiar with the floor layouts (rooms, etc.) in WTC7 have pointed out that NIST's description of the floor layouts in many cases bears little resemblance to the actual floors, and their estimates of combustible loading (the amount of burnables) are likewise frequently inconsistent with the actual reality.

Finally, NIST appears to have ignored the common parameter that office fires tend to burn out in any area in about 20 minutes, moving on to adjacent areas where possible, leaving little time for major heating of beams or other elements that are part of a huge network of heat-dispersing steel. All of this further calls into question NIST's description of the fires.

PM apparently didn't look at NIST's report sufficiently to see these crushed-beer-can pictures. Anyone who watches the videos (taken from the north and northwest) of the collapsing building's exterior sees what looks like a perfectly rectangular box being sucked into the ground. The discrepancy between that and NIST's crushed beer can is so great that one is tempted to think NIST hoped no one would actually look at their report.

But maybe this is just a glitch in the model or illustration? It doesn't matter. High-school physics teacher David Chandler used free high-school physics software to calculate the acceleration of WTC 7's roofline as the building collapsed. For a period of a little over 2 seconds during the start of the collapse -- 10 floors worth -- it fell at free-fall acceleration. In the preliminary report (which, in fairness to PM, is what PM would have seen if it had looked at it), NIST used inappropriate beginning and ending points for their calculation and claimed the collapse was much slower than free-fall. Faced with the obvious proof of free-fall, NIST eventually back-pedaled and admitted to free-fall in the final report. But free-fall means that nothing (no intact structure) was present beneath the upper part of the building to slow it down. Ironically, Sunder had made precisely that point in his initial comments, saying that that is why free-fall couldn't have happened -- i.e., it was inconsistent with their model.

The point here is that NIST could not have missed the nanothermite chips or the microspheres if they had looked for them, as is legally required in the standards set forth for fire investigations where there is significant structural dismemberment -- standards which explicitly mention thermite. But NIST didn't look for them, because they claimed that getting such materials into the buildings was "unlikely." Likely or not, they were there. As an engineer once pointed out, "We know that there is methane on Jupiter. We don't know how it got there, but that doesn't change the fact that it is there."

But more interesting is the fact that the fire department had pulled all its people out of the building and given up fighting the fires hours before the building collapsed. This information was certainly available on the day, contrary to Sunder's claim. Larry Silverstein (WTC7 owner and WTC lease holder) claimed that he and the fire chief decided to "pull" the building to avoid further loss of life. "Pull" means to run a controlled demolition, in spite of Silverstein's later claims that he meant "pull out the firemen," who had, in fact, been out for hours. Avoiding loss of life by demolishing a building implies knowledge that it has been evacuated.

As to the debris being analytically worthless because it had not been labeled (why not?), a demolition would have left massive amounts of evidence in the debris, whether that debris was labeled or not. But because Sunder wasn't about to consider demolition, he feels he can dismiss the destruction of "unlabeled" evidence as irrelevant, even though it is an unprecedentedly gross violation of fire-investigation rules.

Given that Sunder was just quoted as having said that NIST knew from the beginning that the study would be difficult, it's hard to understand why (a) he then "underestimated" the effort, and (b) NIST declined to add investigators to an understaffed operation.

But let's not fault him for that kind of incompetence. What he does need to be faulted for is absurdly saying that his conclusions -- see the crumpled building above, so utterly at odds with the videos of the collapse -- "go along with what was observed." And he should be held responsible for the afore-mentioned fraud committed by the omission of the shear studs from the model.

It is certainly understandable, given all this, that NIST refuses to permit the scientists involved to discuss the report in public with its critics, though that refusal is highly unscientific. But it is unacceptable, bordering on criminal, that they refuse to release their model's inputs for the usual scientific examination, on the grounds of "public safety." They are leaving construction engineers without the information they need to decide if their long-held building standards are in need of radical revision. So much for public safety.