Levitas – The PLAN Journal, 2021

TECHNICAL INTRODUCTION

The research underlying the design and structure of Levitas followed simultaneous, parallel paths: a classic modelling/analytical approach as well as a physical/trial and error approach using mock-ups and real sections in wood. This “phygital” approach combined computer-based analytics/calculation and virtual geometry with real-world experimentation, and was particularly important in view of the geometrical and material constraints imposed by the artwork’s unique geometry. The result is a tectonic artifact which emerges from the ground “like a new mountain”, as the Art Director of Arte Sella, Emanuele Montibeller, loves to say.

During the research process it became clear that the lines of the pure parabolic surface in the computer models, where virtual strings (axis) defined a certain geometry, would be subject to transformation when the virtual models were applied to material reality. The woven structure of the flat wooden strips and the “small geometry” of the lattice elements’ rectangular sections affected the “big geometry” of the built sculpture, and created a surprising change in the whole geometry of the artifact. The sculpture acquired a sense of living interconnectivity from the unexpected shape created by the interaction of the wood lattice and the gravity loads distributed within it and supported by the tall arches on three hinges and pin-point ground supports: two lateral points on each side and one “whale tail.”

The magic of Levitas also derives from the embodied process of its creation, during which a kind of “neural connection” grew among all the partners involved, tuned to Ian Ritchie’s wavelength, achieving a unique structure-cum-work of art within a beautiful, natural sylvan landscape.

Material Considerations

The sculpture is constructed of 9 m [29.5 ft.] long wood strips of small section (60 mm x 10 mm – 2.4 in. x 0.4 in.), jointed with wooden bands and M5 metal bolts. It was particularly important to consider the durability and mechanical characteristics of various wood species before deciding which to use, and the choice between two locally sourced woods: black locust (pseudoacacia) or red oak (Quercus Rubra), was considered during the design phase.
Black locust is not considered a commercially important wood species and it has some potential as a more sustainable alternative to tropical hardwoods. It is a durable, attractive, and naturally rot resistant wood, superior to pressure-treated lumber while non-toxic to the environment. This makes it a locally grown alternative to chemically treated lumber, endangered tropical woods, and decay prone woods.
Limited information is available regarding the mechanical properties of black locust and the species is unaddressed by design codes and grading agencies. Load testing has been performed only on red oak (the wood finally chosen) using full-scale members and small holes for the connections in order to substantiate average values reported in the literature and to validate model FEM analysis. Local availability of both woods is good, as are lattice elements of 6 m length and over, although transport becomes more complicated when lengths exceed 6 m.

After the realization of the first mock-up, black locust was chosen as the preferred option due to its resistance to weathering and durability when compared with other local woods. In the end, however, red oak – which has similar characteristics of resistance and durability – was the final choice. This was because the quality of the available wood was better, and a rapidly available supply of sufficient quantity was needed – which then had to be subjected to the necessary cutting/drying procedures – in order to have all the material ready for erection in time.

Joints

The original design of the lattice is a 9.6 m x 9.6 m [31.5 ft.] “membrane,” so the red oak lattice strips had to be jointed. It was possible to realize longitudinal joints with complete resistance restoration by coupling with wood-continuity plates. Initially, a section of 60mmx10mm wood elements had been assumed.
The metal elements, laser cut, had been used in the prototype in order to guarantee a consistent curvature because they have the same stiffness as the wooden elements. Finally, wood plates were placed in order to disguise the joint. This joint is easy to manufacture and, as required by Arte Sella’s guidelines, does not require the use of glues. Their positions in the membrane had been well studied in order to harmonize with the overall design. Longitudinal joints were calculated according to Eurocode 5, and the number of bolts was calculated in accordance with the stresses to which the elements would be subjected to.

Boundary Joints

At the edges of the membrane, laser cut and perforated metal plates that can be coupled with wooden elements have been used. In this way, joints with excellent resistance and precision were created. Tests were done of the loads on connection strips, in order to optimize dimensions and quantity.