My role as a research assistant was to produce all wooden pieces required and create a safe environment for the processing of wood with a saw blade tool. This was achieved through the rigorous programing and building of the tool together with students from the Material & Detail course.

We designed and structured the production of the wood pieces to optimize for fabrication time. We held an average of four minutes per piece, this also included the time it takes to replace each piece for the next one. As the final pavilion was created from three identical legs, we focused on processing each individual module piece in a series of three. This simplified the overall management required to keep track of what piece and which NC code we needed to run.

Typical CNC milling is never as efficient and productive as a saw blade when processing wood. The amount of sawdust is minimal as each cut was only 3,5mm thick, the speed at which we could move the robot arm was well below its limits. Grasshopper was used in addition to KUKA PRC to generate each NC file. The modular code allowed us to alter the code quickly for eventual problems with various robot motions.

The wooden pavilion was developed and built in collaboration between Chalmers University of Technology and Augsburg University of Applied Sciences. The project relied on the parallel development of the Chalmers ACE RobotLab where I was responsible for the development of hardware, software, and production of the project. This was only possible to our work developing the design, detailing, and material preparation as this allowed us to fix the final bugs in our production methodology and work holding setup.

The design focused on producing a grid shell that needed to span 6 meters yet be quick to produce and modular. They are detailed so each piece is cut on four sides with precise angles cut with a single slit that allowed for an additional piece of wood be used as a guide during assembly. Because we cut each of the four sides of the piece, the modules are able to be build flat. This greatly increased the overall accuracy and ease of assembly. The amount of material we used was also relatively little compared to overall strength and span of the structure. Six 2500x1200x21 sheets of plywood and 120m of 45x145 standard C3 building timber was used. Each piece was nested and prepared before the robot production.



The Reciprocal Shell pavilion is made up of 28 pentagons, hexagons, and heptagons in three identical collections creating each leg. These modules are designed to be put together flat and later assembled through bolting together with adjacent pieces.

The modules are comprised of two parts, one 21mm plywood frame and a series of 40x140mm reciprocal interior pieces. Each piece becomes angled so even though it can lay flat, a curvature can be created through the assembly of several pieces. Each piece had additional slots cut into them that allowed for an additional wooden spline. This allowed for extreme precision when assembling each individual frame as well as increased the strength of the joint. This detail made the assembly process possible, and it shows in the extreme tight tolerances of the final structure.

Through trial and error in the initial setup of the fabrication setup, we held a tolerance of ±0.5mm through the entire fabrication process.


 

<sup>Researchers: Amin Adelzadeh, Dr. Hamed Karimian, Karl Åhlund, Jonas Lundberg, Prof. Dr. Christopher Robeller (Chalmers Jubilee visiting Professor 2022)

Students: Agathe Ducos, Adrien von der Weid, Alexandra von Bartschikowski, Anne Wicklein, Clément Braekevelt, Edith Tamm, Edona, Ege Can Yazici, Erika Perleroth, Felicia Raunås, Felix Bossenmaier, Herman Ehrnberg, Jael Kahlenberg Caso, Jakub Maliborski, Katharina Harre, Lina Svantesson, Linn Appelgren, Lluc Chia Colomer, Lorenzo Mulatero, Lukasz Gnatek, Madalin Manu, Maija Virkki, Mauritz Renz, Natasha Zwarts, Nina Tran, Nora Reis, Parjaree Manuch, Rick Persson, Rikard Murgård, Sakurako Matsuoka, Samira Sarreshtedari, Sara, Murphy Colome, Selina Ahrens, Selina Christ, Simon Ciompi, Simon Wikström, Sofia Deixelberger, Theresa Haase, William Havner, Ziming Wang</sup>