Thursday (8 October) we splitted up the group between Delft and Amsterdam VUmc.
We had an important meeting with an expert on Computer Graphics and Visualization, Elmar Eisenmann (http://graphics.tudelft.nl/~eisemann/ ), at the TUDelft. He took a look at the STL model and concluded the model looked like a 2.5D structure, which will make it possible to take a 2D image (a slice of the STL-model) and multiply it to get a bigger pattern.
Specific algorithms can be used to generate new patterns without boundaries, because they take small parts of multiple pattern-images that fit together best and “glues” those together. To read more about this, here are some of the papers he referred us:
- Fast Texture Synthesis
- Texture Synthesis over Arbitrary Manifold Surfaces
- Solid Texture Synthesis from 2D Exemplars
The group in Amsterdam worked on dividing the mesh in 2D layers and export these as pictures. We also worked with the staff at the 3D Innovation Lab to make a solid mesh to be able to invert it or to find the original scan so we could make a new and hopefully better model. This turned out to be more difficult task than it first seemed and we couldn’t finish it in one day.
The next step is to convert the larger model into a 3D solid, by combining 3 patterns in 3 different planes. Click on the images below to visit the links. For that we need to experiment with different 3D-model programs.
Friday (9 October). Working with layers.
After the layers were processed in illustrator and photoshop we were able to create a new bigger random cartilage pattern. We did this by using software called “Pixplant” that uses the algorithms to create a new pattern from different patterns combined.
Four layers of cartilage combined into one new seamless pattern looks like this:
After some work and discussions, we tried to take and simplify the averaged layers with Photoshop and Illustrator, in order to make 3D models with Rhino. To make it possible for the cells to move into the structure from the sides, we figured that we can make one big solid out of multiple layers taken from different averages, and put them onto each other. The layer averages where taken every 5 slices of the original cartilage (1-5, 6-10, 11-15, 16-20, 21-25, 26-30), finally having 6 different slices one onto each other.
The challenge of this method will be how to print it, probably we need dissolvable/detachable support structure.