Fabric Formwork

Hi Daniel,

I tried following the steps you mentioned above, also ran the simulation as per your instructions but am still not getting the desired result. Could you please go through the attached files and suggest a solution.

Also, I am working on my dissertation which deals with structural optimization techniques with fabric formwork, similar to what Mark West from University of Umanitoba (http://www.umanitoba.ca/cast_building/assets/downloads/PDFS/Fabric_...) has done. Is there any plugin related to grasshopper which can be helpful for such experiments.

Any suggestions from anyone will be of great help.

Thanks.

Attachments:

forblog.3dm, 57 KB
forblog.gh, 8 KB

Replies to This Discussion

Permalink Reply by Suraj Suthar on June 18, 2012 at 7:36pm

Hi Daniel,

I tried working with the script you made but i think it is not much of use in this case. We tried to do a physical experiment with basic cylinder shaped in different angles to understand the deflection when plaster is cast into them. I want to simulate the experiment digitally as well based on the observations. I have tried to form the basic relaxation definition, but its very difficult to reach the desired results.

Please find attached the images of the experiment and the rhino file as well the definition, any inputs from anyone will be appreciated.

Attachments:

for blog2.gh, 12 KB
for blog2.jpg, 332 KB
for blog2.3dm, 153 KB

Permalink Reply by Diederik Veenendaal on June 25, 2012 at 2:29am

Ok, I haven't really looked at your files in detail, but I could give possible issues with recreating your physical model digitally:

(1) The starting mesh for each column has to be the same shape and relative size of the cutting pattern or flattened layout of your fabrics prior to casting. Kangaroo uses elastic deformation of sorts (the spring stiffness in Kangaroo is equal to EA over Lo, where E is Young's modulus, A is sectional area and Lo is the initial length of the spring/bar prior to loading), which means that the initial shape of your mesh carries information through to the end result (i.e. different starting mesh, all other parameters being the same (density of the fluid, boundary conditions, etc.), produces a different result).

(2) The springs are difficult to relate to a surface membrane like a fabric. Ideally you would have springs in the mesh along the orthogonal weave directions of your fabric, and the diagonal springs of lower stiffness to model any shear stiffness the fabric might have (typically lower than the orthogonal stiffnesses, low for uncoated fabrics, higher for coated fabrics). So in other words, because Kangaroo models a surface as a network of springs (which uses the mesh topology as that network), a different mesh, all other parameters being the same, will produce a different result.