Preliminary results

This website has described two different schemes for covering large venues with a retractable roof.

It has been found possible to cover a planar bar structure with rigid elements and it has also been shown that it is possible to design structures made from rigid plate elements alone. These structures do not need to be circular is shape thus allows for a large array of applications of which retractable roofs are only one.

However, for very large spans the inherent problem of low stiffness in flat structures has led to the development of several alternatives which are based on doublely curved shell elements cut from a sphere. These structures promises to overcome the problem with lack of stiffness as they both in their open and closed states have an increased stiffness due to shell action.

Through recent research we have developed a method for generating these doubly curved elements such that there is no overlap of these elements at any point during their motion as well as forming a gap free, continuous surface in the ekstreme positions. The later is what gives the possibility of using shell action to carry the forces in the roof.

Our current research is focused on developing mechanisms, which in an efficient maner allows the motion of the shell elements between the extreme positions without contact between the various elements in the mechanism.

We have build several physical models of our schemes and they have thus been verified both nummerically and physically.

Future work

In the near future we have several very interesting subjects to investigate.

To improve the useability of the planar solution we will be writting a piece of software that will allow quick and easy generation of the geometry for this type of structure.

We will finish our work on several larger physical models of both planer and spherical structures. These will demonstrate the schemes while also providing valuable information about various problems that will be encountered when building at larger scales than small concept models.

Our work on novel types of mechanisms will continue as these show great promise and might prove to be expandable from purely spherical structures to more free forms. This will allow these schemes to align themselves with the architectural trend of freeform blob structures.