Additive Manufacturing with Solid Wood: Continuous robotic laying of multiple wicker filaments through micro lamination
DC I/O 2020 proceeding by JULIAN OCHS, ZUARDIN AKBAR, PHILIPP EVERSMANN.
Contents
Abstract
Current commercial additive manufacturing (AM) methods with wood filaments are based on extrusion of thermoplastics that are mixed with ground timber (approx. 40% content). Therefore, these filaments do not have the inherent structural and mechanical properties of timber with its continuous fibers. The mechanical strength of the thermoplastics can also decrease with a proportion of ground wood (cf. WIMMER; STEYRER; WOESS; KODDENBERG; MUNDIGLER, 2015). Since the decomposition of thermoplastics in an uncontrolled environment can still take 100 to 1000 years (cf. ROYTE, 2006), renewable and more biodegradable alternatives are required. We have developed an AM method using a wooden filament made of a fast-growing local plant: wicker twigs. This technique enables direct and laminated wood filament extrusion without the use of thermoplastics. The material used in this new method maintains its natural tensile strength, which, due to the unique material, its processing and AM technique is higher than that of beech wood. The wicker filament is made by joining homogenized wicker strips. In a continuous process, the resulting endless filament is then coated with a contact adhesive, dried, and fed into a specially designed extruder that combines up to five upright wicker filaments to one single laminated square profile. The adhesive on each separate string is activated both through contact pressure and with the help of a Proportional–Integral–Derivative (PID) controlled heat block. Due to this technique, a fast layer-based printing method was achieved. The ability of wicker filament to bend in the direction of the layers enables the creation of curved surfaces with a minimum radius of 20mm. A new AM technique similar to additive fiber composite technique has been developed, which enables the fabrication of freeform components, that are otherwise known from FDM-based printing technology (fused deposition modeling), and are suitable for large-scale applications such as furniture or architectural components due to the strength of its continuous natural fibers.
Keywords
Design Automation, Recommendation, Space Planning, Architecture, Residential Design, Machine Learning, Random Forest Regression.
Keyphrases
apartment design (140), candidate space (110), normalised score (80), similarity score (80), architectural design (80), apartment layout (70), space planning (70), feature weighting (70), machine learning technique (63), residential design (60), spatial layout (60), regression technique (50), external wall (50), popular vector (50), residential layout (50), architecture firm (50), space syntax (50), spatial analysis algorithm (47), window wall length (47), spatial analysis (45), first principle (40), machine graphical communication system (40), design space (40), user interface (40), recommendation algorithm (40), space type (40), end user (40), desk layout (40), feature weight (40)
Topics
AI-driven development, Architecture, Artificial Intelligence in Design, Assisted Design Decision Making, Construction, Design Knowledge Capture, Design Support Systems, Future of work/unemployment.
