Professor Jerry Connor, from the Civil Engineering Department, Massachusetts Institute of Technology, gave a special seminar at Ashurst Lodge on the Collapse of the World Trade Centre.
Jerry started by explaining the unique design of the WTC with a nucleus of large columns containing the lifts. This was to achieve space without columns. The weight of the floor is considerable and the wind loading is strong in that location.
It was designed in the early 1960s and constructed in the late 1960s and early 1970s. The interesting point is that to carry the load, the designer used not only the central core, but an exterior wall consisting also of very closely spaced columns. It is basically a tube structure. This results in the floor loading being roughly 50-50 distributed in the inside core and the columns on the periphery. The design at that time was unique.
Each floor was composed of a series of trusses, designed in such a way that prefabrication on the ground was optimised. This was the weak point of the design from the point of view of its collapse. The floors were fabricated using a corrugated steel decking and pouring concrete over it. This was a very different and easy system for prefabrication.
The building was designed against aeroplane impact in fact, but could not stand up to the fire damage. Each of the steel columns in the periphery was prefabricated and three storeys high. This helped to optimise erection time and made the work to be done above ground level.
The floors were supported on the columns by L shaped supports. The joining of the trusses to the L shaped supports was the weakest part of the design. This connection would not have been acceptable in an earthquake-proof design (Stage 1).
Soon after construction, it was noted, however, that the building started to move perceptibly under high winds. This was solved by using viscous dampers and connecting the bottom members of the trusses to the columns but this was done only for damping and not to transfer forces.
The collapse was produced, not by the loss of strength of the columns after the crash, although the hole in the external surface was substantial, but by the shearing off of the L supports. Those L clips were shown to have sheared. (Stage 2) The floors consequently collapsed and the floors above impacted on the floor below. (Stage 3) The resulting 'domino' effect produced the collapse of all the floors below. This resulted in practically a free fall of all the floors. The impact continued to shear off the clip-like supports all the way down.
The above is the theory put forward by Jerry Connor, although other people have put forward different ideas. They have proposed that the core buckled and that this produced the weakening of the building due to a dynamic effect. Jerry simply believes that the gravity load was enough to cause the collapse. This weak connection would not have been allowed on the West Coast, where earthquake requirements apply. Even nowadays, such a connection would be made into a moment-resisting-joint on the East Coast. It was a nice design with a weak point. The lecture was well attended and resulted in a lively discussion.