[tsofian]

I just reread the section in the module. It makes even less sense. The towers are surrounded by a steel lattice, which fills in the space between the wider top and bottom of the structures. This takes the additional load. The walls are described as built to "keep the ground outside".

So we now have a steel lattice structure which appears to be within an area of backfill. We have that backfill in direct contact with the concrete walls of the towers. We have the towers rigidly fixed to the bottom of the shafts, but also in full contact with other very heavy elements.

So now instead of a freestanding structure in a vertical shaft (which I incorrectly always assumed due to the diagram on the floor plans, which now seems rather misleading) we have a structure that is rigidly fixed to one surface which will be moving and is surrounded by a vast load, which will also be moving, but in ways that are different from the surface below. This is possibly the worst of all possible world. Any shock transmitted into the mountain will cause the bedrock to move in one way and the backfill to move in another. The towers can't be on springs or roller to allow motion because they are surrounded by backfill. That backfill will also transmit any ground shock into the towers.

Even in a period of a few decades the towers will be presented with huge amounts of stress just from the shifting of the backfill and the movement of the rock layers within the mountain. An Earthquake will send extreme and unpredictable stresses through the structure. There is no way this would work as described.

For the tower floors to be attached to the shaft walls the same goes, unless each attachment point is designed to flex in all three dimensions and to allow for compression, rotation, and tension between the tower and the living rock of the walls.