The story of a hybrid structure

by Mladen Babic and Paul Carneau

As a material for construction, stone has been used for centuries by all know human civilization. From house, to bridges, aqueduct, pyramids, historic examples are visible almost everywhere in the world which show the structural efficiency, sustainability and esthetic aspect of its use. Stereotomy or the art of cutting stone has evolved over time to allow the construction of more complex, more slender structures. From the stacking of cubic stones (cf. Egyptian pyramids) to the erection of long span vaults, it was the number one material for construction until industrial revolution brought to us more flexible and adaptive materials such as steel and concrete. From that moment, the number of project made out of stones has decreased continuously to almost nothing. The main drawbacks of the stone being the lack of industrialization process (no controlled flexibility), the difficulty to procure the material and the complexity of assembly of the discrete elements (temporary formwork, heavy lifting).

Nowadays, stone structures are making a slow come back thanks to new computational power and digital tools. Interesting design is emerging by enhancing the craft art of stereotomy with computer made algorithm. Very slender vault like the armadillo vault [1] or complex structure like the “stonematters” pavilion [2] are exciting examples of what can be achieved using “old” technique and material with new computational methods.

Figure 1: The Armadillo Vault by BRG
Figure 2: Stone Matters by AAU Anastas

The second half of the XXth century has also seen the birth of a complete different type of structures characterized by its light weight, slenderness and ease of installation: the gridshells. Again, thanks to development of computational tools, and apparition of new composite materials, it got easier to model, parametrize and calculate this kind of complex geometry. At the beginning, this was only feasible using real model at a small scale to try to understand behaviors.

Figure 3: Elastic Gridshell in composite material

Therefore, we have in our hands two very distinct types of structures, old and new, heavy and light, stone and composite material. But they both have been used to create double-curved geometries, or vault. Our idea then, is to combine them into one hybrid ultra-efficient structure. Using the advantages of both to compensate the problems that each faces on its own.

For stone vault to be build, it always requires a complex and expensive formwork, which is also temporary. In our case, we would build the gridshell first, and then use it as support for the stones, changing something temporary into permanent. The gridshell will also enhance the structural behavior of the stone vault. Stone constructions have been oversized in the past. That way it was correct to consider only the dead load in the calculation. A stone vault designed to work only in compression, with a minimized thickness would behave badly under dynamic or asymmetrical load. However, the presence of the elastic gridshell, when oriented correctly will help for those specific load cases.

Figure 4: Symmetrical tessellated vault

The gridshell would serve as formwork for the installation of the stones, so no need for an expensive temporary support. It should also enhance the behavior of the vault under asymmetric and dynamic load thanks to its resistance to bending moment, the stone vault working mostly in compression. The gridshell reinforcement of the stones should also provide a plastic rupture behavior in case of collapse instead of the characteristic fragile behavior (i. e. all the vault collapse if you remove only one piece). In addition, the stones bring a remarkable structural esthetic, and solve also the issue of the cladding of the gridshell which is usually as complex to generate as the gridshell itself.

Figure 5: Gridshell generation on the vault surface

Form-finding tools (RhinoVault) will help us to first generate the global shape of the vault with funicular loading. Then the structural analysis of the vault will guide the processes of tessellation and voussoir design, as well as the choice of the generatrix for the elastic gridshell generation.

Proofs of concept will be done thanks to small scale models of the vault using 3D printing of the voussoirs. The gridshell tubes will be modelled by the filament (1.75mm diameter). Two different models will be build: a first model as close as possible to reality, combining the gridshell and the stone voussoirs. And a second vault with only the stone part. The idea being to show the behavior differences between the two and proving that the gridshell highly increase the vault structural efficiency.

Next step would be building a full scale structure. We believe that this hybrid structure will bring some new thinking in modern stone architecture and inspire further merging of traditional knowledge, vernacular architecture and technologies.