PHASE I
Print to Fabric
Fabric formwork has the ability to create unique geometric forms, but it is difficult to control to the degree of accuracy required for a building element. This phase of our prototyping focused on printing onto fabric as reinforcement.
Our setup utlizes a 3-d extruder attached to a 7 axis Kuka Robot. The initial test involved the printing of a rectangular lattice and then framing it in order to pour concrete. The fabric type, and extrusion layer count were changed to test overall elasticity, print strength and concrete finish.
PHASE II
Auxetics
In phase two the research turned towards 3-dimensionality. With these tests the focal point is on controlling the final form through the alteration of cell geometry. We had begun to research auxetic patterns because of their ability to form geometries contrary to the forces applied.
The prints were subjected to gravity by hanging them from wooden frames allowing for the relaxation under the weight of concrete. PETG, PolyCarbonate, and PLA were tested as the print material. On the whole the tests were to strong and did not exhibit the slumping hoped for although the patterns are a quite striking.
PHASE III
Smart Mesh
Given the difficulties present within the creation of the auxetics especially in regards to the transition into 3-dimensions and the unexpected rigidity of the created geometries, we decided that instead of creating a stiff object that attempts to exert too much control over the final form, we would use gravity and the weight of the concrete to our advantage. We were inspired by Frei Otto’s hanging models which he used to create complex 3-dimensional forms. And although these models were used for form finding and not created as part of the construction process we saw potential to use them directly in the fabrication process
For this phase of the prototyping process the print bed has been scaled up and the print material changed. The printed material is now a rubberlike TPU with an elasticity around A50. The first attempt focused on changing the cell size to create greater slump along the larger cells.
PHASE IV
Panelization
Following the initial tests based on changing the cell size, work began on another option which relies on a python code to smash a 3-dimensional object into a flat mesh. The code analyses the given surface along UV points. It then measures the distance between the points and arcs on a corresponding 2-dimensional grid until the lengths are equal. The arcs will then straighten under the weight of the concrete to produce the needed geometry.
To test this mesh creation method a simple compression only vault was created in RhinoVault. The final form successfully recreated the predetermined geometry. The largest problem we had was the inconsistent edge condition created along the unbounded edges. This is due to the lack of a printed exterior wall which had previously been part of our prints. The other thing to notice is the relative lack of pillowing within the mesh cells. This is because of the use of 100% polyester and is regulated by the chosen fabric.