Acknowledgements: Thank you to Elliot Olney and Christopher Tallman for volunteering their time to help complete this book. Much gratitude goes to Renee Cheng for the opportunity to teach this studio and for promoting a culture where limits can be pushed. Special thanks to Sharon Roe for her pedagogical insight and her relentless directorship of the University of Minnesotaâ€™s BDA program, which has become such a fertile environment for experimentation and innovation. Thanks to our dedicated core of graduate Digital Assistants (Will Adams, Phil Bussey, Sam Daley, John Greene, Dan Raznick) for their technical assistance, and to our peerless workshop staff (Kevin Groenke, Justin Kindelspire, Keith Tucker, Patrick McKennan) for their expertise and patience with a messy class. Thanks also to our roster of guest critics (Dan Clark, John Comazzi, Lisa Hsieh, Nat Madson, Molly Reichert, Marc Swackhamer) for their critical insight and feedback throughout the semester. Finally, thank you to these sixteen remarkable students, who rose to every challenge I could possibly throw their way. â€” Adam Marcus, June 2013
MODULARVARIATIONS ARCH 3250 DESIGN WORKSHOP SPRING 2013 UNIVERSITY OF MINNESOTA COLLEGE OF DESIGN SCHOOL OF ARCHITECTURE BACHELOR OF DESIGN IN ARCHITECTURE PROGRAM
Instructor: Adam Marcus Students: Elizabeth Adler Sam Anderson Holly Hodkiewicz Emmett Houlihan Erik Jackson Thomas Kuhl Adam Lucking Jonathan Meyer Nickolas Mosser Elliot Olney Jorie Schmidt Alicia Smith Christine Stoffel Christopher Tallman Rythm Unnown Sharanda Whittaker Book Layout: Adam Marcus Elliot Olney Christopher Tallman
CONTENTS Introduction..........................................................................................................................06 Precedents......................................................................................................................... 08 Project One: Variable Tessellations.................................................................................... 12 Project Two: Controlled Variation........................................................................................ 18 Project Three: Instrumental Variation................................................................................24 Image Credits.......................................................................................................................64
Let us understand at once, that change or variety is as much a necessity to the human heart and brain in buildings as in books; that there is no merit, though there is some occasional use, in monotony; and that we must no more expect to derive either pleasure or profit from an architecture whose ornaments are of one pattern, and whose pillars are of one proportion, than we should out of a universe in which the clouds were all of one shape, and the trees all of one size. — John Ruskin, “The Nature of Gothic”, from The Stones of Venice, Vol. II The work presented in this book is the product of a studio led by Adam Marcus in the spring of 2013 in the Bachelor of Design in Architecture program at the University of Minnesota School of Architecture. The general premise of the studio—that buildings are made of parts, and that it is the architect’s task to select, design, and organize these parts into a coherent whole—was investigated through a series of full scale design-fabricate-build projects that explored ways for variation to be employed (or not) in the design process. These experiments were framed within the broader context of architecture’s still somewhat nascent embrace of computation and digital technologies, which easily enable mass customization within both design and fabrication processes, and which have, in recent years, contributed to a staggering ubiquity of formal differentiation within architectural production. This studio sought to foreground the issue of variation not as a given byproduct of the technologies we use, but as a designed, intentional, and instrumental strategy to advance specific architectural goals. The tension between standardization and variation—what parts are the same, what parts are different—was investigated specifically through material practices of molding, casting, and tiling. With the help of parametric design and digital
fabrication tools, the studio explored opportunities for introducing variation and unpredictability within processes of forming and casting systems that typically rely upon standardized repetition, while at the same time developing rigorous logics for how this variation can be deployed. The studio was structured into three separate projects, each of which addressed the question of variation through various material and computational strategies and allowed students to gradually build a fluency with ways of calibrating and controlling variation. The first project introduced techniques of parametric design and digital fabrication by asking students to design and build systems of laser-cut and folded paper modules that could tile repetitively while also accommodating variable behavior. The second project extended this logic more into a volumetric realm, with students designing variable, flexible, and/or reconfigurable molds capable of producing cast plaster modules that could stack yet also exhibit a range of formal properties. The final project, a team effort among all sixteen students, channeled the casting research into the design, fabrication, and construction of a full-scale wall prototype consisting of structurally repetitive yet individually unique concrete modules. Students were challenged to design variation in an instrumental and thoughtful manner, considering the system’s ability to address specific criteria related to site, program, or performance. The structure of this book mirrors that of the studio. The first section, a brief summary of historical and contemporary precedents researched by students throughout the semester, helps lay the groundwork for the first and second projects, which are included in condensed form. These two projects established design methodologies and material strategies that directly informed the final project, which is documented in its entirety. MV | 7
Preced Erwin Hauer, Chicago Hall, Vassar College (1956)
Ron Resch, Tessellation study (c. 1970s)
Matthew Shlian, Extraction Series: Extruded: White (2012)
William Morris, Design for a printed textile: ‘Wey’ (1882-3)
A number of precedents—formal, structural, material—informed the work of the studio throughout the semester. Inspiration was drawn from both historical and contemporary case studies, from William Morris’s repetitive yet seamless textile patterns to artist Daniel Widrig’s CNC-routed plaster tiles. The studio’s design research was also influenced by architecture’s long tradition of modular masonry construc-
Daniel Widrig, C. Tiles (2009)
tion, from standard CMU block typologies to more contemporary experiments with flexible and reconfigurable formwork. Of particular interest to the studio were examples of modular assemblies that introduced unpredictability or variability into processes of mass production, maintaining the efficiencies of repetition while allowing for a wider range of variation in form or performance.
MV | 9
Erwin Hauer, Jerusalem Tower (1969)
Typical concrete masonry screen units
10 | Precedents
Mid-20th century concrete block screen wall (c. 1950s)
Morphosis, Perot Museum of Nature & Science (2012)
Miguel Fisac, Centro Cultural, Castilblanco de los Arroyos, Sevilla (2000)
Andrew Kudless / Matsys, P_Wall (2009)
Adam Marcus / Variable Projects, Modular Variations Prototype I (2013)
MV | 11
Variable Tessellations This project explored principles of rigorous, rule-based variation through processes of tessellation and tiled patterning. The objectives of the project were to (1) develop systematic ways of relating 2D and 3D geometries, (2) understand how subtle and incremental variations at one scale can produce larger change on a different scale, and (3) introduce methods of using parametric design software to iterate and test different ways to deploy these variations. Students began in an entirely analog mode, using only flat Bristol paper and a blade to develop a system of geometric operations (limited to cutting, scoring, and folding) that yield significant formal transformations of the material in three dimensions. The intent was to develop a rule-based logic for effecting and controlling these transformations across a number of iterations. Students then transitioned
to the computer and re-modeled their system using the Grasshopper parametric modeling software package. The use of Grasshopper was limited and strategically focused; students reconstructed the 2D system of cut and score lines in a relational model that enabled the testing and refinement of the rules and parameters that govern the system. The project was highly self-referential in the sense that there was no outside information or performance criteria by which the iterations were evaluated; aside from negotiating the material parameters and limits of the paper itself, the students remained purely in the formal realm, and the variation remained merely for variationâ€™s sake. But the simple overlay of analog and digital computation foregrounded the false dichotomy between the two, and the project established a framework for thinking about variable geometry in a controlled manner.
MV | 13
14 | Project ProjectOne: #1 Variable Tessellations
MV | 15
16 | Project ProjectOne: #1 Variable Tessellations
Christopher Tallman Modular Variations Project One
MV | 17
Controlled Variation This project continued to explore the relationship between standardization and variation through a rigorous process of designing and fabricating mold systems that produce variable cast components.
able with casting methodologies, (2) test how parametric software can be used to design and fabricate molds / mold components, and (3) develop compelling models for the mass production of variable components.
In teams, students designed dynamic and/or reconfigurable molds that produced cast modules capable of both individual variation and seamless tiling. The objectives of the project were to (1) become familiar and comfort-
The project introduced intensive and iterative processes of plaster casting into the studio workflow. Each team constructed a small full-scale wall prototype that demonstrated the systemâ€™s capacity to vary.
MV | 19
Group 1: Reconfigurable + Flexible Piston Mold
Elizabeth Adler, Emmett Houlihan, Nickolas Mosser, Rythm Unnown
Mold design: Three pairs of adjustable pistons produce a variety of extruded modules that can stack together interchangeably. The pistons serve as an armature for a thin plastic formwork which conforms to global shape parameters yet allows for some measure of uniqueness in the surface geometry and texture of each individual module.
The modules are stackable in two different ways. Oriented consistently, they can stack like bricks in a regularly spaced pattern. They can also be stacked in a more ad hoc fashion, akin to dry-stacked stoned walls, which achieve structural stability through density and maximizing the surface area of module-to-module contact.
20 | Project Two: Controlled Variation
Group 2: Crochet Fabric Formwork & Vacuum-Formed Molds Holly Hodkiewicz, Adam Lucking, Jonathan Meyer, Alicia Smith
Fabrication workflow: The process began with casting plaster within flexible, custom-crocheted formwork. As reuse of the crochet formwork proved to be difficult, the students used vacuum-formed molds of the original casts to mass-produce the intricately patterned modules.
The transition to vacuum forming preserved the rich patterns produced by the crochet fabric, but also allowed for much greater ease of production. Variation of the modules is achieved by using different permutations of the plastic molds for both the front and back of each cast.
MV | 21
Group 3: Canvas Fabric Formwork and Reconfigurable Mold Sam Anderson, Thomas Kuhl, Elliot Olney, Christine Stoffel
The mold consists of interchangeable, stacked planar elements that serve as an armature for a taut fabric form. Reconfiguring the armature pieces yields a range of cast modules that produce a variety of apertures when stacked. The performance of the canvas fabric during the casting process adds another layer of texture and pattern to the modules.
22 | Project Two: Controlled Variation
Group 4: Rotationally Reconfigurable Formwork
Erik Jackson, Jorie Schmidt, Christopher Tallman, Sharanda Whittaker
This system utilizes a rotational mold that yields a set of variable modules that stack in a radial manner. The modules join together using a small, standardized connector piece. The rotational logic produces variety across the wall assembly, and the systemâ€™s two-sided nature provides enough structural redundancy such that certain modules appear to spin or float freely in space.
MV | 23
Instrumental Variation The final project built upon many of the techniques and strategies developed in Project #2. As a group, the studio designed, fabricated, and constructed a site-specific, self-supporting wall prototype composed of variable, modular masonry units. The process was similar to that of Project #2 in its reliance on iterative material testing, yet it addressed several new significant constraints. First, the studio transitioned from plaster to concrete, which introduced a number of new material concerns and parameters into the fabrication process. Second, the studio’s approach to designing variation—which in the first two projects was simply for variation’s sake—was now driven by two more quantifiable forces: site and performance. Students were challenged to (1) identify a site for intervention in or around Ralph Rapson Hall (the home of UMN School of Architecture) and (2) identify specific performance criteria (light, visibility, etc.) which the intervention could address. The
studio selected an exterior covered terrace adjacent to the building as the site for the prototype. The high visibility of this location from both inside the building and the surrounding campus provided ample opportunity for testing variable transparencies of the cast modules. At the outset of the project, four teams engaged in a one-week charrette, structured in the format of a design competition, to generate design ideas and fabrication strategies. The studio then collectively determined a single direction and self-organized into four teams—Prototyping, Computation, Representation, and Documentation—for the remainder of the project. This collaborative model was fundamental to the completion of the project; the division of labor facilitated a much higher level of resolution than otherwise would have been possible with individual projects in such a short timeframe.
MV | 25
Charrette Scheme 1: Variable Interstitial Space
A dynamic mold produces stackable, brick-like units that vary in thickness to modulate the interstitial space between units, allowing for variable conditions of transparency across the wall.
26 | Project Three: Instrumental Variation
Charrette Scheme 2: Hexagonal Module with Variable Void
A standardized, hexagonal module is punctuated by variable apertures that are produced by a dynamic bladder that spans the inside of a rigid mold. Different materials used for the bladder have dramatically different effects on the form of the resulting void.
MV | 27
Charrette Scheme 3: Reversible Hexagonal Shells
A reconfigurable mold produces thin, cast shells that have different sized apertures on either end. The variation is deployed such that the shells stack vertically. The scheme also includes partial extensions of the shells, to provide shading and collect water.
28 | Project Three: Instrumental Variation
Charrette Scheme 4: Stackable Vaulted Modules
Stackable vault-like modules with variable interior geometries are produced using either a flexble fabric form or a series of interchangeable inserts.
Nickolas Mosser, Christine Stoffel, Emmitt Houlihan, Christoher Tallman
Infinite fabric form.
Nickolas Mosser, Christine Stoffel, Emmitt Houlihan, Christoher Tallman
MV | 29
Mold Development At the conclusion of the charrette exercise, the studio agreed to proceed with a hexagonal module that incorporated a variable void, not unlike a camera lens. Consensus was to build upon Scheme 2, which the group found compelling for both its structural resilience and its aesthetic potential. Specific aspects of the other schemes would also be brought forward into the next stage of prototyping. The process of refining a viable mold focused on developing simple methods to control the size of the aperture. The students determined that the dimension of the final modules would be twelve inches (see mold photos above), but the initial explorations during the mold design were produced at 1/2 and 3/4 scale (prototypes at left). 30 | Project Three: Instrumental Variation
The studio explored several methods of modulating the aperture: fixed collars, latex bladder displacement with filler material such as foam nuggets and glass marbles, and incremental rotation of the form faces in order to produce a twist in the latex bladder. The mold design team also tested the addition of a third variable through lateral manipulation of the aperture in elevation using a choking wire around the stem of the latex hour glass, yet this ultimately proved too difficult to control. Ultimately the simple twist of hexagonal faces was chosen as the primary means of controlling the variation of the interior void. With each successive iteration, this method maintained controlled consistency of the cast product, and maximized the legibility of the variable apertures. MV | 31
Mold Side (wood) Mold Side (Wood)
Sleeve Cap Sleeve Cap(MDF (MDFor or acrylic) Acrylic)
Mold Walls Mold Walls(wood) (Wood)
Mold WallLiner Liner(duct (Duct tape) Tape) Mold Wall
Variable Void Variable VoidSleeve Sleeve(Latex) (latex) Geometry of void twists and aperture size changes based on rotation of mold sides.
32 | Project Three: Instrumental Variation
MV | 33
Computational Development Computational design, via Grasshopper for Rhino 5, was leveraged in this project at several stages within the design process. First, the parametric functionality of Grasshopper was employed to assist in form finding and idea generation. Once the general focus of the project was established (controlling a dynamic void within a fixed geometry), students used the software to test possible variations to influence the directions of exploration. The computerâ€™s capacity to quickly produce numerous iterations expedited the design process and allowed the studio to quickly focus the material prototyping based on design directions explored in digital space. Later in the process, Grasshopper became the primary tool for generating construction documents, diagram content, and visual representations (see diagrams at right). The algorithm used to generate digital forms earlier in the process was developed further into a kind of proto- building information model (BIM) that incorporated the refinements and improvements made to the mold design during the prototyping stage. During the fabrication process, the Grasshopper design model produced instructions for void sizes, which was communicated to the casting team in terms of how much to twist the bladder in each mold. During the assembly process, the digital model generated assembly drawings to direct the stacking of the physical blocks. A vital and easily overlooked component of the process was the integral give-and-take that existed between the digital work of the Computation team and the physical work of the Prototyping team. The efforts of the two teams, though simultaneous, drove and informed the work of each other, contributing to a highly productive feedback between digital and physical.
34 | Project Three: Instrumental Variation
MV | 35
36 | Project Three: Instrumental Variation
MV | 37
38 | Project Three: Instrumental Variation
This diagram explains the logic of the variable formwork: as the rotation angle of the mold sides increases, the twist of the internal latex bladder increases, and the resulting aperture in the cast module becomes smaller.
MV | 39
Material Refinement The students in the class had very little prior experience working with concrete and none with casting concrete masonry units. The extensive experience with plaster of Paris during Project #2 had developed a basic competence with casting and formwork design, but this did not fully prepare the studio for the use of concrete as a material or the scope of the final project casting: sixty-six 12â€? concrete tiles for a wall, in only four days. Issues arose from the addition of sand and gravel aggregate, as well as the volume of casting (eight tiles every twelve hours). The studioâ€™s approach required a high degree of discipline vis-Ă -vis the processing of the mold from assembly through casting, un-molding, cleaning, and reassembly. At the final scale of twelve inches, the latex bladder was materially at its limit. Even the merest Portland cement residue could cause the failure of the latex upon reassembly. This fragility led to explorations of alternative aperture materials, including spandex, woven plastic sheeting, and landscape mesh fabric. Though the studio ultimately returned to the latex due to the superior finish it imparted upon the concrete, these tests allowed for a rigorous investigation of the properties of concrete (see images at right). The final concrete mix used a 5000 psi readymixed concrete with the addition of one liter of Portland Type II cement per 80 lbs. bag. Molds were treated with canola oil and hand filled. In the end, all tiles were within an eighth inch tolerance and required very little post-cast cleaning.
40 | Project Three: Instrumental Variation
42 | Project Three: Instrumental Variation
MV | 43
Elizabeth Adler Sam Anderson Holly Hodkiewicz Emmett Houlihan PROTOTYPING
Erik Jackson Thomas Kuhl
Adam Lucking Jonathan Meyer Nickolas Mosser Elliot Olney Jorie Schmidt Alicia Smith
Christine Stoffel REPRESENTATION
Christopher Tallman Rythm Unnown Sharanda Whittaker SCHEME 4
Collaborative Workflow Throughout the course of the five-week project, the students self-organized into a number of task-based teams in order to most effectively iterate and produce. 48 | Project Three: Instrumental Variation
FRIDAY 5/3, 9:00AM
FRIDAY 5/3, 9:00PM
PROJECT #3 — CHARRETTE THURSDAY 5/9 PROJECT #3 — PROTOTYPING SATURDAY 5/4, 9:00AM
SATURDAY 5/4, 9:00PM PROJECT #3 — PRODUCTION
PROJECT #3 — COMPUTATION SUNDAY 5/5, 9:00AM
SUNDAY 5/5, 9:00PM PROJECT #3 — DOCUMENTATION
MONDAY 5/6, 9:00AM
MONDAY 5/6, 9:00PM
TUESDAY 5/7, 9:00AM
TUESDAY 5/7, 9:00PM
MV | 49
Final Review 50 | Project Three: Instrumental Variation
MV | 51
52 | Project Three: Instrumental Variation
MV | 53
54 | Project Three: Instrumental Variation
MV | 55
56 | Project Three: Instrumental Variation
MV | 57
58 | Project Three: Instrumental Variation
MV | 59
60 | Project Three: Instrumental Variation
MV | 61
IMAGE SOURCES Unless otherwise noted below, all images in this book have been produced by the spring 2013 Modular Variations studio at the University of Minnesota School of Architecture and are licensed under a Creative Commons Attribution-NonCommercial-NoDerivs 3.0 Unported License: http://creativecommons.org/licenses/by-nc-nd/3.0/deed.en_US Page 8: Erwin Hauer, Chicago Hall, Vassar College (1956) http://erwinhauer.com/eh/installations/vassar-college-chicago-hall-poughkeepsie-ny Page 9: William Morris, Design for a printed textile: â€˜Weyâ€™ (1882-3) http://www.telegraph.co.uk/culture/culturepicturegalleries/8296065/Pre-Raphaelites-The-Poetry-of-Drawing.html?image=10 Ron Resch, Tessellation study (c. 1970s) http://jane-kate.blogspot.com/2011/01/j6-origami-tessellation.html Matthew Shlian, Extraction Series: Extruded: White (2012) http://www.mattshlian.com/ Daniel Widrig, C. Tiles (2009) http://www.danielwidrig.com/ Page 10: Erwin Haeur, Jerusalem Tower (1969) http://www.francisfrost.com/hauerjerusalem.html Typical concrete masonry screen units http://www.flickr.com/photos/buxwal/205513449/ Mid-20th century concrete block screen wall, c. 1950s http://thuydaoarc.blogspot.com.au/2011/06/mid-century-decorative-concrete-screen.html Page 11: Andrew Kudless / Matsys, P_Wall (2009) http://matsysdesign.com/2009/06/25/sfmoma-update/ Adam Marcus / Variable Projects, Modular Variations Prototype I (2013) http://www.variableprojects.com Miguel Fisac, Centro Cultural, Castilblanco de los Arroyos, Sevilla (2000) http://matsysdesign.com/studios/compositebodies/2010/03/miguel-fisac-facade-in-sevilla/ Morphosis, Perot Museum of Nature & Science (2012) http://www.flickr.com/photos/redblank/6808926793/
64 | Image Credits
MV | 65