WIXLIB

Integrated Parametric Design and Production Planning:A Luminaire Design Case StudyPeter Ferschin1, Georg Suter212TU WienDigital Architecture GroupVienna, Austriapeter.ferschin@tuwien.ac.atTU WienDesign Computing GroupVienna, Austriageorg.suter@tuwien.ac.atABSTRACTThe paper describes how to extend the parametric designprocess by including feedback from the production processand iterating a design towards a realizable object,considering real world constraints such as production costsand manufacturing time. In this paper, we report on ourexperiences with integrated design and productionworkflows from teaching a Master level digital design andproduction class, in which we give students a task to designand produce a luminaire based on a given productioninfrastructure (3d printers, laser cutters, 3-axis millingmachines). Production processes are simulated and andprovide feedback to designers about the feasibility of designalternatives regarding material use, cost, machine time. Wedescribe the didactic and technical concepts and concludewith a discussion of open issues.Author KeywordsParametric design; digital production; 3d printing; lasercutting; milling.ACM Classification Keywords Applied computing Arts and humanities Architecture(buildings) Computer-aided design Computing methodologies Modeling andsimulation Simulation support systems Simulation tools1INTRODUCTIONAdvanced digital design and fabrication technologiesprovide significant opportunities for designers to realizeinnovative buildings and architectural products withimproved qualitative and quantitative properties regardingshape, material, cost, usability, comfort, or sustainability. Inthis paper, we investigate the challenge of integrating diversedesign and fabrication technologies in order to achieveworkflows that let designers rapidly and seamlessly explorethe feasibility of design alternatives with respect to givenfabrication constraints. Specifically, we explore the linkingof state-of-the-art visual scripting based parametric designtools and production planning simulation tools for milling,laser cutting and 3d printing. We describe the necessaryknowledge and skills that designers must acquire in order touse these tools productively. In order to mimic realisticSimAUD 2020 May 25-27, Online 2020 Society for Modeling & Simulation International (SCS)565conditions for architectural designers, we use the design of aluminaire as a case study.Our work contributes to the areas of parametric design anddigital fabrication by focusing on skills and knowledge, andby considering the integration of production planningsimulation for multiple production methods in the designexploration process. Previous related work investigatesdesign knowledge (Woodbury, 2010; Oxman, 2017) andpedagogical aspects of parametric design, and the role ofvisual programming and scripting languages (Aish & Hanna,2017; Celani & Vaz, 2012). Further related is work on digitalworkflows (Wortmann & Tunçer, 2017), design formanufacture and assembly (Austern, et al., 2018), and multicriteria design (Imbert, et al., 2013)2TEACHING DIGITAL DESIGN AND PRODUCTIONWe teach integrated parametric modeling and digitalproduction in a set of related, Master level courses (or‘module’) to develop and deepen knowledge of topics andskills for integrated parametric design and production.Knowledge of geometry optimization and discretizationconcepts and techniques is essential for shape designs thatare computationally efficient as well as material andproduction friendly. Such designs may be created withparametric modeling and programming functionality ingeometry modeling and CAAD software. Students learnbasic data processing concepts and develop programmingskills in computational geometry exercises. They learnconcepts of digital production methods, including additive,subtractive, and formative fabrication and are aware ofworkflows, material and production parameters. Acquiredknowledge and skills are applied in a small digital productionproject which we describe in detail in the following sections.3PREPARATORY EXCERCISESTo teach the necessary skills, required for an integrateddesign and production process, we defined various exercisesthat expose the students to important concepts like surfaceand solid modeling, visual and text-based scripting as well asdigital production methods. We emphasize the main idea ofparametric thinking, that allows the generation of numerousdesign variants, which could be evaluated either by humanexpertise or by computer simulations (Oxman, 2017).

Important is the concept of identifying key design parametersas independent and dependent variables.3.1 Surface ModelingNon-uniform rational B-splines (NURBS) modeling is aquasi-industry standard geometric modeling method forcurves and surfaces which is widely used in the developmentof innovative designs, including industrial design, arts andcrafts, and architectural design. Its popularity is due to easeof use in rapidly creating and modifying complex free-formshapes based on intuitive parameters, such as control pointsand weights. Continuity (or smoothness) is a key qualitycriterion for free-form shapes. Analysis methods that areprovided by geometric design tools include curvature graphswhich let designers assess the smoothness of transitionsbetween curve or surface segments. We use the NURBSmodeling functionality in Rhino to study surface modeling.3.2 Visual ScriptingVisual scripting interfaces facilitate graphical rather thantext-based program definition. Thus they appear useful fordesigners who tend not to be expert programmers. Examplesof visual scripting environments for computer-aidedgeometric and architectural design include RhinoGrasshopper and Revit Dynamo. Visual scripting is usuallybased on the dataflow programming method. A programconsists of a directed, acyclic graph where the nodes processdata and edges represent data flow between nodes. Each nodeaccepts input data, processes the data, and sends output torelated nodes. The graph is automatically evaluated wheninput data are modified. This is beneficial for design taskswhere it is necessary to rapidly modify an evolving designby varying input parameters. We use the Rhino Grasshopperenvironment for visual scripting.3.3 Python ScriptingWe additionally expose students to text-based scripting, inparticular with Python scripting in Rhino as well as inGrasshopper. We choose Python as a language, as it is themost common scripting environment in 3D designapplications. We introduce a small subset of the languagePython, geometry creation in Rhino and how scripts can beintegrated into the Grasshopper environment to extend thepredefined functionalities. Each lecture is accompanied bysmall exercises to verify that the provided scripts of thelectures can be applied with small adaptations. The firstlecture is focused on parametric two-dimensional curves,simple 3d solids and extruded geometry from curves. Toteach first steps in programming, we implemented amodified version of a Turtlegraphics Python library forRhino. The exercise for the students consists of writing theirfirst name as a parametric function, additionally supportingdifferent graphical styles (Figure 1). The implementationshould define a Grasshopper python component, with a listof polylines as output parameter that can be manipulatedfurther (e.g. by extrusion).566Figure 1. Parametric first name – Lena Roth, 2019The second lecture focuses on constructing NURBS surfacesfrom NURBS curves and by defining parametricmathematical functions that specify design parameters andimplement surface meshes with adaptable resolutions. As anexercise, students have to define suitable mathematicalformulas and design parameters to create parametric disheswith design variations as mesh structures implemented asPython components in Grasshopper.

3.4 3D PrintingIn a separate lecture (“Digital Production”) we explain theprinciples and possibilities of 3d printing and whatrequirements for geometric models need to be considered(e.g. solid geometry, closed volumes, consistent normalvectors). As an exercise, students have to prepare a terrainmodel for printing. For the simulation of the printing processUltimaker Cura is utilized, parameters that influence thespeed and quality of the printing process are explained.Figure 2. Parametric dishes – Lena Roth, 2019The third lecture explains how parametrically definedmeshes can be converted into buildable structures bygenerating solid elements at vertices, edges or faces of amesh. The parametric dish functions of the previous exercisehave to be converted into buildable structures.Figure 4. Preparing a terrain for 3d printing – Lena Roth, 20193.5 CNC CuttingIn the next step, various CNC cutting strategies areintroduced (e.g. laser cutting, waterjet cutting). As anexercise, the same terrain needs to be prepared to begenerated as layers on a laser cutter. An additional exerciseintroduces the task of designing a parametric Kirigami (theart of paper cutting and folding) layout. To simulate thecutting process the tool VisiCut is used to check the layoutfor the laser cutter as well as to estimate the cutting time.Figure 3. Parametric dish constructions – Lena Roth, 2019567

to the stock material is visually inspected. Additionally, thetime required for milling is evaluated.Figure 7. Preparing a terrain for 3-axis milling – Lena Roth, 2019Figure 5. Preparing a terrain for laser cutting – Lena Roth, 2019Figure 6. Parametric Kirigami design – Mihael Barada, 20193.6 CNC MillingNext, the principles of CNC milling are introduced and aPython library for generating CNC code is provided to thestudents. The terrain has to be prepared for a 3-axis millingmachine. An additional exercise – to generate parametricmilling patterns – has to be implemented as a Pythoncomponent in Grasshopper. For the simulation of the 3-axismilling process the software Camotics is used. As the millingpaths are generated by the students with Python, it isimportant to check for possible errors that could damagetheir work piece or even the milling machine. This is doneby evaluating the milling paths in Camotics and verifyingthat no collisions with the milling machine will happen.Furthermore the plausibility of the milling paths in relation568Figure 8. Parametric milling patterns– Lena Roth, 20194PARAMETRICPRODUCTIONLUMINAIREDESIGNANDWe illustrate the parametric design and productionworkflows outlined above with an example project.4.1 Design BriefThe design task involves the design of a luminaire for aspecific activity, such as dining or reading, and for a specificroom, such as dining room or office. To ensure diversity inluminaire designs, these functional requirements were