I have always been fascinated with historic architecture and its building techniques. Mainly because most of the times the shapes of architecture were a direct consequence of the building techniques used in the project and their technical limitations (e.g. the rural houses in the city center of Rome never exceeded 6 m in width, as the maximum length of a timber beam in that area was 6 m). While living in Rome I had the chance to study and perceive many of the monuments of the Roman Empire and it was really interesting to see how it is possible to build so many different shapes by using the repetition of a single unit and cement: the simple roman brick.
I have to be honest: I have never visited the inside of the colosseum. Nevertheless I have walked in front of it many times and I’ve always taken a glance at the vaults which sustain the upper orders. These are elliptical continuous barrel vaults which by nature are a problematic structure as a barrel vault is the repetition of the same elements in a linear direction. I’ve always wondered: how did the Romans manage the bricks’ position in a structure with this geometry?
Shape. This kind of vault can be described mathematically by the upper portion of an elliptical barrel but can also be described geometrically by the extrusion of an arch through an elliptical base. The problem is that the arch is made of bricks and if the inside side of the vault is shorter than the outside one, being them concentrical. The management of the bricks needs to be accurately solved in order to have a statically functional structure. This fact brings me to the second observation on this structure.
Displacement. As the minimal unit of a vault is the brick the design of an asymmetrical vault like this needs necessarily the study of the displacement of these units. In fact while geometrically the elliptical structure is built with a simple extrusion of an arch, technologically it is not possible to repeat the brick structure of an arch through an ellipse as the bricks would overlapse on one side of the vault and wouldn’t even be enough on the other. The solution must be in the displacement of the bricks in order to have a vault with normal statical behaviour.