Building Freeform 2014

With two openings supported by corners plates on both sides as two fins above the ground, the pavillion was named Manta Ray. The overall shape has a central symmetry with a semi-circular arch in the mid-plane, two opposite laced edges on the ground and two corner plates fixed to composite posts. The form-finding was done with Ferrari-sketch and the cutting pattern with a custom triangulation program on Grasshopper. From a technical point of view, the softness of the wooden supports on the ground did allow for a good pre-tensioning of the fabric, so that wrinkles are quite visible in some areas. The stiffening effect of double curvature was however clear and the project was a good initiation to tensile structure design. Beside the enclosed space centering the facility window was really nice.

The rigid gridshell was shaped as the head of an elephant and its geometry is made of the concatenation of three moulding surfaces with a common generator and three plane rails with a C1 continuity. This insures that all nodes are torsion free. Therefore a prefabrication of the different quad modules independently is possible and allows for connections along the edges and not at nodes. The modules have planar faces and their edges have constant height which allows easy construction without recourse to advanced numerical fabrication. The structural behaviour of the grid was studied with Karamba and some additional reinforced tape was added in the tensed areas. Unfortunatly, a default in the control process of the members geometry was detected late and the pavilion could not be finished on time. It is still a good lesson about free-form architecture.

2014, the elastic gridshell pavilion benefited from the numerical developments done around the Créteil cathedral and for the first time in the workshop it was properly analysed: the relaxation of the grid, the triangulation, the reinforcement of the members on edges, the reaction on supports were all studied. The shape is that of an undulating tunnel, with two free edges and two line supports on the ground. As usual, the assembling of the grid and the supports was the time consuming step of the construction process: the forming in itself and the connection to the supports took only one hour to the group. One more hour being required for the bracing. The control of the form was satisfying, but not perfect as the plumbing connections used had tendency to rotate (which among other aspects reduced the shear stiffness of the arches on the free edges…).