The Parametric Wood seminar is an advanced digital design research seminar focused on contemporary timber fabrication and long-span structures enabled through parametric design in Grasshopper. The seminar will be broken down into research teams each with a specific focus: timber gridshells, timber lamellas, panelized construction, variable plate connectors for straight timber segments, and continuously curved rainscreen cladding. Through a basic working knowledge of these research areas through a precedent survey, research teams will exploit the possibilities of these material typologies through parametric design, testing out through scaled physical prototypes and full-scale assemblies. By precisely taking control of variations in material bending and/or versioning of joints enabled through parametric design, the focus of this design research seminar is to exploit these timber typologies through physical prototyping.
As a design research seminar, this course has very specific goals and expectations. Note that “search” is the core aspect of re-search, in which the expectation of re-search is to search again and again. This not only an iterative or cyclic process but one that is funnel shaped – narrowing from more general to being very precise. This search within research should be considered as an integral aspect of design, however research also requires a rigor of investigation, an accountability, and a verifiable demonstration that is often not articulated in design (cf Glanville 1999). In this seminar, rigor of investigation develops through a very focused research project, which I have outlined for you to focus the design research in this seminar. Accountability is developed through your clear articulation of your process in both graphic and written form. In this seminar each research team will continually document their efforts in an on-line lab journal through this blog format. As a lab journal, each team member will make at least one post a week documenting how they have pushed their research further in that week. Failure is an integral part of research, and documenting what doesn’t work or the gaps in your own knowledge can help others in their research efforts. Developing skill and precision with tools and procedures (particularly with new tools) is a form of tool-driven research that is part of the accountability process as well (cf Gallison 1997). Therefore, your lab journal will take accountability of this research process through the weekly commitment to further develop the research, documenting failures as well as successes, as well as documenting your developing precision with tools and materials. Keep in mind we are not operating in a cloistered environment, but will rely on the internet for precedents, user-groups and forums which should be integrated (cited through links!) into your on-line lab journal. Finally, verifiable demonstration is the core product of this design research, which is not at all part of the design process as you have been familiar. By verifiable demonstration others can pick-up the research, and come to a similar understanding through testing-it-out. Parametric design offers a unique context for verifiable demonstration as others can employ your parametric definitions, while at the same time, this requires a rigor of documentation within your Grasshopper definitions to clarify the procedural steps within the definition. In scientific language, to be verifiable your definition cannot be a “black-box” hiding its internal workings to only reveal inputs and outputs. This is essential to emphasize, because of course parametric design as an algorithm is a black-box. The rigor of documenting the course of your logic within the definition then enables inspection for others, making its internal logic transparent to others, and therefore verifiable.
To summarize, as a research seminar you need to maintain a rigor of investigation which develops through a clear research proposal (in this case largely given), you need to develop an accountability of your research through clear documentation not after the fact, but in process (through our on-line lab journal), and finally, the verifiable demonstration should contextualize research as a form of knowledge for others to pick-up and develop from. As design research, it is important to emphasize that this research is ultimately useful for the designer in the design process and that we are designers engaged in research. Research should not seem foreign to you, and at the same time should shift your thinking in what is produced in this seminar. Ultimately, this design research seminar is about the production of new knowledge in connection to timber fabrication, long-span structures and envelopes enabled through parametric design that can be demonstrated in artifacts and shared through parametric definitions. The products you are familiar with, sketches and models, are integral and essential vehicles in this re-search (and by all means the beauty of these artifacts will demonstrate your rigor of investigation) but keep in mind that the ultimate produce of this research is new knowledge for others to understand, pick-up, and employ. How you are able to communicate this knowledge to others is core to design research.
Evaluation / Deliverables
I envision this course as a think-tank of mutually engaged and dedicated people who have chosen this seminar because they are interested in it and self-motivated enough to pursue a design research agenda. As a core component of this course is research outside of typical class time, the evaluation of your work will be based on what you are able to document as outlined above in design research. Therefore, there are three principle views into your design research plus a final exhibit which evaluation will be based on:
- 1. the on-line lab journal (blog),
- 2. process models, drawings, and diagrams, and
- 3. clearly articulated Grasshopper definitions.
- Finally, we will have a concluding exhibit which will include both final physical artifacts with well documented graphic boards.
As largely self-guided research, I do not want to require a split percentage of each for some ideal grade, rather, everyone is required to work in all of these areas, but as team research projects, I understand certain people will be driven by different aspects of these deliverables. As a team research project, all of these deliverables will be met. As an individual in a team, make sure that your work is clearly documented not simply for evaluation, but that you are accountable through documentation within this design research.
Digital Design Process
As an upper division elective in advanced digital techniques, students must have a working knowledge of Rhino software and experience with Grasshopper is highly desirable, thought not required. Basics of Grasshopper and parametric approaches to design will be introduced, while each research group will advance their development through numerous on-line resources. This is neither a Rhino nor a Grasshopper course, and in fact my hope is that you become the experts at Grasshopper. Nonetheless, the course is enabled by these tools to exploit the opportunities within timber fabrication. I have a basic working knowledge of Grasshopper, but am also well aware of the time-suck that Grasshopper can entail. My role is not to troubleshoot your definitions, but to continue to guide your research process. Grasshopper is not magic, but will require you to invest a fair amount of time on the forums (see blog side bar) understanding what others have done, trying their definitions etc. This is part of the research process and be sure to document this in your lab journal. The key to success here is to break up the bigger problem into manageable chunks with two clear rules of thumb: start small and just get it working. In other words, don’t expect some master-definition that will do everything, but break up the problem, tackling it bit by bit, getting each bit working as you proceed to connect to other bits. I have outline the research projects below with this in mind.
The wood typologies we will work with stem from the material itself: in segments (members) or in panels. These timber members or panels can be straight or curved (single curved or double-curved). Furthermore, we will describe curvature as continuous or segmented, and likewise those timber elements that make-up this curvature as continuous or segmented. However, no matter continuous or segmented, all timber construction is assembled from smaller components and therefore a fundamental aspect of timber is understanding the problem of the joint. Please do not underestimate the difficulty of the joint, and how the joint is parametrically developed.
The research projects outlined below develop through the research I have developed over the last 5-7 years, but before the accessibility of parametric design afforded through Grasshopper. The research I have developed has enabled me to understand and outline the work that needs to be done, but it does not make me an expert nor does it mean I have solved the problems I have outlined. On the contrary, these are not trivial problems and in the case of Gridshells and Lamellas, not fully realizable given the state of existing tools. I have outlined the research areas below, from most difficult to more basic, depending on your skills and interests.
Gridshells (continuous curvature) (advanced GH)
Gridshells rely on very specific constant curvature to create a shell structure that is comprised of thin elements. Gridshells are not diagrids, don’t get this confused. Gridshells, like diagrids, externalize their structure in the exterior envelope, but gridshells are shell structures, diagrids are not. Core to gridshells of all kinds is a balance between form-finding and the pattern of geodesic curves that sit on the surface (a geodesic curve is the shortest distance between two points on a surface). We will distinguish between gridshells in general and lattice shells as a particular form of gridshells. Lattice shells develop from continuous straight members and the arching action of these straight continuous members organized in a matt-like system raised or lowered into the shell shape (cf Cabrinha 2008). New approaches to gridshells closely connected to contemporary timber fabrication and digital design can be seen with Shigeru Ban’s Pombidou in Metz which maintains the continuous curvature of gridshells, but uses CNC milled heavy timbers (e.g. rigid) to form the structure instead of the traditional arching action of straight segments in lattice gridshells. In either approach, this research team will take on a three part strategy: surface relaxation or form-finding through physics plug-ins to Grasshopper (Kangaroo and/or Geometry Gym), understanding of geodesics and parametrically defining their lattice-like pattern, and finally parametrically defining the multi-layer lath offsets including their thickness, depth, and joints.
Lamellas (straight-segmented curvature) (advanced GH)
Lamellas appear similar to gridshells both for the grid-like appearance and their shell-like long-span structures. However, the fundamental difference from gridshells is that lamellas ingeniously are built-up from short straight segments in a centripetal pattern. Whereas the multi-layer laths of gridshells and constant curvature allow for a simple pin connection where these members overlap, the basic centripetal building block of the lamella develops around the joint from piece to piece. Based on their simplicity of construction and economy of straight segments, lamellas are evocative structures but complex to design, engineer, and even more so to develop parametrically. This team will need to first develop a basic understanding of lamellas through physical prototypes to understand the basic unit, and then move into parametric design. Different approaches and strategies will need to be developed as there is not as clear a path (to me) as there is for gridshells. After initial material and parametric investigations, this team will outline what needs to be done in general with lamellas, and what aspect(s) in particular they wish to develop (methods of fabrication, simple form with parametric control, or more challenging forms etc).
Variable Plate Connectors (straight segments, variable steel connectors) (intermediate GH)
Most timber construction develops through straight segments in which the form development is focused at the joint. Traditional timber to timber connections (mortise and tenon, dowels) are possible with CNC timber fabrication, but in larger structures wood to wood connections are not strong enough to take their associated increased loads. It is now commonplace to insert steel splice plates at timber joints or steel connectors at joints as can be seen in our Simpson Strong-Tie building. The intent for this research group is to understand variable joints / plate connectors as a jig from which the building’s form develops. Parametric design can enable the mass customization of these plate connectors developing 2d cut files direct from the 3d model. This group will develop large-scale laser cut joint studies and potentially full-scale assemblies.
Panelized (straight and curved)(basic GH)
Pre-fabricated panelized systems are existing technologies in the USA in SIPS panels, with far more sophisticated approaches developed in Europe through Cross Laminated Timber panels. These systems are often seen in prefabricated housing, multi-family housing in Europe, and recently through 7+ story residential towers. This group will begin with a thorough survey of existing products, technologies, and case studies using panelized systems, to then generate a parametric tool that can either optimize existing designs into panelized units, or conversely, develop a design tool that develops panelized systems from a simple line set-out drawing.
Rainscreen (continuous curved geodesics)(basic GH)
Rainscreens are pressure equalized open joint cladding systems. Wood is often used for rainscreens as the open joint system does not allow moisture to build-up and therefore decay the wood, while wood rainscreens certainly have a desired aesthetic effect. Although wood can curve fairly easily, these rainscreens are often only developed in straight segments. This research project focuses on the geodesic curve to develop rainscreens from straight segments that can be curved in place during installation. The benefit of geodesics in this application is that when these curved surfaces are unrolled, they will be unrolled into straight segments therefore giving a striking aesthetic facade with an economy of means. This team will develop a parametric definition for geodesic rainscreens, including their associated back-up support structure, and demonstrate these through large-scale models and possibly full-scale mock-ups.