Developing a mold-free approach for complex glulam production with the assist of computer vision technologies
Introduction
Wood has been regarded as one of the most promising building materials for the future due to its many excellent features such as sustainability, lightweight and high strength [1]. Recent years has witnessed the increasing use of wood in a variety of building types, especially large-span and multi-tall buildings. One of the most important wood products in construction industry is Glue-Laminated Timber, or glulam, which is made from layers of dimension lumbers bonded together with durable structural adhesives. With timber grain running parallel to the length direction, glulam is allowed to be produced as large scale components with almost no limit in length, depth and width. Glulam has been widely used in a variety of structural applications due to its high strength-to-weight ratio and large dimensions [2]. However, along with the expansion of the application scope, complexity in timber structures are gradually becoming more prominent, posing great challenges for production and processing of complex glulam with high accuracy and efficiency.
Complex glulam structures such as Centre-Pompidou Metz [3], French Expo-Pavilion Milan [4] and La Seine Musicale [5] could only be realized with the technical support of advancing consulting firms and contractors, not only in the design and optimization process, but also in the fabrication process of complex structures, components and connectors. In such buildings, single curved glulam components are normally achieved with the aid of heavy molds, while double curved ones are mostly milled with Computer-Aided Manufacturing (CAM) machines like wood machining centers. However, great cost of time, materials and money in both mold making and glulam milling process is hard to ignore, since double curved components need to be milled from linear or single curve components. At the same time, the damage of the wood grain during the milling process also severely reduces the load-bearing performance of the wood. The necessity has been widely recognized to explore new approaches for producing complex glulam [6].
The main challenge here is how to produce complex glulam directly so as to reduce the immense material waste and time consumption in mold making and subsequent subtraction process. A straightforward mode of glulam production could not only conducive to simplifying the production process, but also significantly improve the material efficiency in glulam construction. Response to this challenge is of particular importance for sustainable use of wood resources in such environmental challenges that construction industry is facing.
Rapidly developing digital technologies are constantly re-shaping the way that buildings are conceived and produced. In this research, a mold-free approach for complex glulam production is developed with the aid of computer vision technologies. In order to avoid the use of molds, a mechanical equipment is designed to spatially shape the double curved glulam, with computer vision system employed to inform the fabrication process with digital design directly.
This paper is structured as follow: The second section summarizes the current research status of complex glulam production technologies and digital fabrication related computer vision technologies; The third section describes the aim and methods of this research; The fourth section introduces the proposed mold-free production technology from three aspects: equipment, vision system and digital integrated system; The fifth section presents the production process and experiments conducted with different vision systems; The meaning and importance of this approach are analyzed, and future steps are proposed in conclusion.
Section snippets
Background
The curvature of Glulam components could be divided into four categories, say Straight, Single curved, Double curved with no torsion and Double curved with torsion, with increasing degree of complexity [7](Fig. 1). For the production of ordinary straight glulam components, a mature production process in factory has long been formed from strength grading, finger jointing, gluing, laminating to finishing [8]. There are already specialized equipment for each process, forming an efficient
Methods
The technical challenges faced by the current double curved glulam production technology are mainly two aspects: how to accurately locate the virtual design in physical space, how to spatially bend, twist, and finally stabilize the laminas in place during the fabrication process. In response to the current challenges, this research is dedicated to developing a mold-free approach for double curved glulam production, addressing existing issues from three aspects:
- •
a mechanical system is designed as
Mold-free glulam production equipment
The way to produce curved timber could be analogous to that of a draftsman to create a Spline curves with a wood strip. A series of nodes, which could prevent translational movements while allow rotation, act as boundary conditions to hold the strip, resulting in a natural curve with minimal strain energy [10]. In this sense, the logic of single curve glulam production could be understood as shaping the laminas by controlling the location and angle of each press point on a plane. To extend this
Experiments of curved glulam production
A series of double curved glulam production experiments were carried out to verify the feasibility of the system proposed in this study. A scaled prototype of the mold-free equipment was developed for the experiments. The external dimension of the equipment is 1320*740*1160 mm, maximum clamping size of fixture is 160*160 mm, and moving range of the two-axis system is 400*350 mm. As mentioned before, Kinect V2 and HoloLens, as a representative of depth cameras and AR respectively, were used as
Conclusion
This paper presents a mold free approach for curved glulam production with the aid of computer vision technologies. Compared with the current approaches that rely on molds and subtractive processing, the approach proposed in this paper is more straightforward, which provides a universal mold system that adapts to different geometries and tasks. The intuitive benefit is that the consumption of materials, time and labor in complex mold making and material subtractive fabrication process could be
Declaration of Competing Interest
None.
Acknowledgment
The work was supported by the National Key R&D Program of China (Grant No. 2018YFB1306903), and Science and Technology Commission of Shanghai Municipality (Grant No. 16dz2250500, Grant No.17dz1203405).
We'd like to thank Gang Liu from Shanghai Fab-Union Technology Co., Ltd. for his support on the platform development. We would also like to thank Chunpong So from Tongji University for his assistance on the experiments.
References (27)
- et al.
The wood from the trees: the use of timber in construction
Renew. Sust. Energ. Rev.
(2017) - et al.
A review of structural, thermo-physical, acoustical, and environmental properties of wooden materials for building applications
Build. Environ.
(2017) Glue-laminated timber (Glulam)
- et al.
Digital in situ fabrication - challenges and opportunities for robotic in situ fabrication in architecture, construction, and beyond
Cem. Concr. Res.
(2018) - et al.
An interactive augmented reality tool for constructing free-form modular surfaces
Autom. Constr.
(2018) Materialising complexity
Archit. Des.
(2010)- et al.
Timber Gridshells: Architecture, Structure and Craft
(2016) - et al.
From lamination to assembly: modelling the seine musicale
- et al.
New workflows for digital timber
Design Implications of Rigid Timber Gridshells: A Method for the Planning and Manufacture of Load-Carrying, Rigid, Freeform Structures of Timber
(2019)
Parametric and generative design techniques in mass-production environments as effective enablers of Industry 4.0 approaches in the building industry
Autom. Constr.
Cam Design and Manufacturing Handbook
100 Jahre hetzer-patent
Bautechnik
Cited by (2)
Human-cyber-physical system for post-digital design and construction of lightweight timber structures
2023, Automation in ConstructionBending Properties of Glulam Made by Domestic Fast-Growing Chinese Fir
2023, Jianzhu Cailiao Xuebao/Journal of Building Materials