Robotic Knit Printing
My graduate student, Pınar Kahya, and I have begun experiments on material and robotic printing of mycelium-based composites. We are running tests to understand the knit-patterned alternating print, which is very common in robotic 3D printing, and to determine its parameters. We are trying to determine the optimal pattern by integrating material properties, geometry, and printing conditions. Therefore, we need a lot of trial and error. To this end, we created a Grasshopper code. Later on, we switched to a different print mode, but it was still worth testing this knit pattern anyway.



The extruder we use is a small yet capable device gifted to our lab by Doruk Yıldırım. Rather than material recipes, I am more interested in geometry and robot path design. In essence, this code wraps a sine graph onto any given curve. To achieve this, some parameters of the robot and the tool calibration must be completed. On the geometric side, after tangent planes are arrayed along the curve using Horizontal Frames, a sine function is executed on the +Y and -Y axes of each plane. Here, we used the shift and amplitude inputs from an example Pınar found online. Afterward, a series is created. The second sine curve in the -Y direction is reversed with the help of the Reverse List component. This makes it possible to continue seamlessly from the upper row in prints like walls. However, if a periodic path, such as a circle, is to be created, the Reverse List component needs to be bypassed. Ultimately, all points are generated using the given layer height and layer count. A representative view of the output is created with the Pipe component, and the corresponding robot code is generated and saved using Kuka PRC. To reiterate, the calibration of the tool must be done both on the robot and in Kuka PRC. As you can see, in small-scale print jobs like this, having the robot invert its wrist and extend its arm downwards in this manner brings many conveniences. I use this holding posture because the tool is a delicate device prone to tipping over. The axis values (Home position) I use here are A1=0, A2=-90, A3=140, A4=-180, A5=50, and A6=180.


The nice thing here was the patterns that emerged beyond the printing parameters. The initial print trials were not very successful because we had to experiment to simultaneously adjust the robot speed, compressor pressure, and extruder speed to achieve a consistent print. In this type of printing, rather than the material strength, we care about the wall thickness and the small cavities it creates. Our printing studies will continue to understand how living materials like mycelium will behave in these cavities. Below is the Grasshopper code we developed. Kuka-PRC is necessary to run it.






