Following the physical test of our new formwork and casting technique, we created a grasshopper simulation of the cast form which will be used to test the proposed tiles before they are made. Both of these tests will inform our decision about the range of the values we will use as inputs to the linear actuators which will control the degree of difference within each individual tile and between each tile, and influence how the entire series is perceived.
Design of the final formwork uses 3 different plates, two holes to attach the linear actuator too and the third plate to create space for the fabric. Our first pour using the formwork did not work so well. The fabric was too porous and most of the quickcrete seeped through. Also, the fabric does did not have enough elasticity to be pulled probably. Nonetheless, the idea of pulling 4 points using strings works with the formwork. Next step would be to try a different fabric and also use different material.
This week we attempted to make a test casting of a nested condition using spandex and a hollow condition.
For the nested version, we poured Rockite into a piece of spandex stretched over a box and created the bowl in the center using a bottle covered with spandex. We didn’t attempt to twist or move the bottom of the spandex, so the test only considered the stretch of the fabric and the displacement of concrete.
For the hollow attempt, we wanted to create a hollow center in the Rockite but we didn’t but in the square in the center before we poured so the material flowed into the bottom and closed up the opening that we were aiming for.
Next time we will consider making more defined edge seams that can show the rotation of the hollow center by sewing wire into the spandex.
Team: Tina and Jenny
In the final version of our formwork, cement will be cast over a series of four strings, following the same methodology as our first cast. The tension of each string will be controlled by its own linear actuator. The cast object will be created by the response of the concrete to each of these strings, creating five “bumps.” The shape of each separate part of the cast object (each bump) will be determined by the tension of string on either side of it. Strings with more tension will create a curve with a smaller radius, and the result will be a shallower three dimensional shape in the cast.
The dimensions of our form, and resulting tile will be 6″ by 3″. This shape is informed by our proposed site. Instead of a 3 by 3 grid, we would like our tiles to be in a single line and installed along a corner. We haven’t determined a final site, but this corner could be along the top of a wall below the celling, between two stairs on a set of steps, in a vertical corner between two walls, along the floor, or any combination of these spaces. One physical site we have in mind is the steps below Low Library on campus.
The proposed site also influences the top of our formwork. In order for it to be installed in a corner, the top should form a right angle instead of a flat plane. To achieve this, the top panels will be hinged allowing them to close once the cast material has been poured.
We started off by creating a 6″x6″ tile in grasshopper. Using grasshopper and Kangaroo we extracted the four corner points to manipulate them to see how a piece of fabric would react. We also used Weaverbird plug in for some commands. The Z-axis for each point can be controlled and the stiffness can be adjusted to better simulate real life.
Next we took the same process and produced 9 different tiles. Although we have no fully grasped the control of all those points when multiplied to that scale. So we used a random number generator or a series modifier. The problem with both these method is that they don’t create interesting form. The random number generator only generates one type of form and the series modifier creates form in a very linear fashion.
We still decided to take the idea forward and create a system where the linear actuators would be able to be used to create such a form.
But instantly after creating the system we realized it would not work the same way as our initial intent. By manipulating the exterior piece with the fabric draped in the middle the effect would be not the same. Also, with this method we would be manipulating two points at the same time and not the desired 4. At the end we realized that the method of manipulating 4 points in the specific way we wanted would not be possible using this type of system.
Our latest irritation is a system that will work. It is based on the same principle of manipulating 4 points, but rather than controlling the whole form we will be controlling the fabric from inside the form.