Educating architecture students on precast: Employing additive manufacturing to recreate historic precast facades

Introduction

Kiel Moe, in his article “Building Agnotology,” describes agnotology as a form of knowledge production that directs attention to what remains unknown and why. Agnotology considers how ignorance develops, gains momentum, and in the case of pedagogy, identifies topics left unlearned. Moe further argues that although a minority of faculty may acknowledge ‘building’ as a legitimate category of knowledge production, it often revolves around composition and assembly of architectural objects. This narrow pedagogical approach limits our understanding of building as a dynamic process. ¹ In fact, the networks of people, ideas, and materials directed through the process of building are far more complex, and potentially impactful, than the design of buildings. ² While current architectural education emphasizes high-performance and sustainable building enclosure design or complex assembly of building elements, precast building enclosures raise unfamiliarity regarding the interplay between module design, mold design, and labor. This paper argues that these aspects remain inadequately explored, if considered at all, in architectural education. There are multiple steps for casting a malleable material: creating a positive reference part, conceiving the formwork as the negative of the desired part, then pouring the liquid concrete, and finally, demolding the hardened part. The pedagogy presented here brings to the forefront the overlooked aspects of designing and fabricating precast façade elements, including the intricate design of modules, mold conception, and the implicit labor underlying the entire process.

This article begins by examining the past, present, and future prospects of architectural precast in facades. It then delves into the pedagogical approach implemented in a studio setting offered at the Illinois School of Architecture. Afterwards, the paper exemplifies three historic precast façade precedents by explaining their design and fabrication techniques. These selected precedents feature sculptural, complex, and custom-designed modules. Each precedent undergoes a historical review, followed by an analysis of the students’ approach to its recreation and fabrication. The discussion section discusses the pedagogy, incorporating students’ feedback on the project alongside the instructor’s observations on additional precedents. The results section examines the results, establishing connections between students’ initial skill levels, and the successful delivery of their projects. Finally, the Conclusion provides a summary of the outcomes and reflections on the pedagogical approach.

Past, present, and prospects of precast

During the 1950s, the United States witnessed the development of curtain wall systems featuring sleek glass facades. Architects, engineers, manufacturers, and developers were driven by the goal of creating an economically efficient system that projected a progressive image. ³ The growing prevalence of curtain wall facades raised concerns among critics who expressed unease about the potential alienation caused by the repetitive panes of these structures. In 1954, Saul Steinberg, a renowned artist known for his New Yorker cartoons, humorously transformed an actual piece of graph paper into a towering curtain wall structure, mocking the curtain wall’s impact on the traditional urban fabric. This artistic irony led to the term “graph paper architecture,” coined to describe these buildings, suggesting that anyone capable of creating a grid could design such structures. ³ Consequently, concrete, a material characterized by its plasticity that lends itself to expressiveness, was considered as a viable alternative to the glossy, reflective surfaces of the curtain wall. ³

The use of precast modules in building envelopes persists in today’s architecture, employing concrete or alternative materials. One notable distinction between the contemporary cases and the concrete modules of the 1960s era lies in the design variation among façade modules. When examining precast envelope systems constructed in the 1950s, variability in the design of envelope modules was limited. This limitation stemmed from the associated cost and labor of molds, necessitating their repeated use. In contrast, today, the use of computational and parametric design tools has allowed a multitude of designers to create non-identical elements for facades. ⁴ However, variable façade systems remain predominantly present in a select set of projects designed by star architects who typically undertake projects with higher budgets. From a different perspective, many of the precast elements of the 1960s era feature complex and sophisticated forms designed by sculptors and renowned designers, aiming to manifest concrete’s potential. Currently, the creation of complex precast forms is not widespread due to the industry’s emphasize on employing more streamlined and cost-effective production methods, coupled with budget constraints in most projects. Only a small percentage of the projects have the financial capacity to realize geometrically complex precast façade systems. Examples include the Broad Museum by Diller Scofidio + Renfro Studio, ⁵ Prosolve 370e by elegant embellishments Ltd, ⁶ and Morphosis Team’s design for Kolon Industries Incorporation in Seoul. ⁷

Another distinction between past and contemporary precast facades lies in the fabrication methods of concrete elements. Advanced digital fabrication methods have emerged, often registering marks on the concrete elements. For instance, a smooth curved surface in digitally fabricated concrete may indicate the use of molding method, while observing deposited parallel and continuous layers of concrete suggests using the liquid deposition modeling (LDM) method. Emerging digital fabrication methods have the potential to be widely adopted by industry, expanding the realization of geometrically complex forms while facilitating efficient production of these geometries.

Contemplating the prospects of precast façade elements, there is awareness regarding the negative environmental impact of concrete. ⁸ Therefore, material optimized concrete design and efficient construction methods are major research focus in the field. The key fabrication technique for transforming concrete construction is additive manufacturing (AM). Additive manufacturing technology can revolutionize two aspects of traditional precast construction: Either the formwork can be 3D printed, referred to as 3D printed formworks (3DPF), or the reference part used in mold making can be 3D printed.

Reviewing the literature on 3DPF, it has mostly been utilized to showcase the potential of bespoke fabrication. Examples include melting 3D printed PolyLactic Acid (PLA) molds with a heat gun, ⁹ 3D printing sandstone molds that remained in place, ¹⁰ or dissolving PolyVinyl Alcohol (PVA) formwork. More recently, with a shift towards exploring new design approaches and materials for repeatable use 3DPF, a range of new research projects has emerged.^11–14 Although 3DPF is a leading method for creating complex geometries and controlling the surface quality of precast concrete, approaches involving sacrificed mold parts are incompatible with economies of scale, while the materiality and durability of reusable 3DPF molds require further investigations.

The second disruption by AM in precast involves 3D printing the initial positive part — also called the reference part — to be used in mold making. This approach is currently underrepresented in literature. This disruption is significant from three points of view:

- First, as AM technologies continue to advance and become more prevalent, this method can promote the design of complex and optimized precast geometries due to the inherent efficiency of creating the reference piece.

Pedagogy

The pedagogical approach entails an examination of historical precast facades, followed by the recreation of these precedents using today’s advanced technologies, namely computational aided design (CAD), and AM. This pedagogy serves not only to bridge the past and the future, but, more significantly, to establish a direct engagement with the fabrication reality surrounding the precedents in both eras. The utilization of historic precedents in education implies a focus on technical and tangible design thinking. Additionally, the pedagogy draws attention to the overlooked aspects of education in precast building enclosures, particularly regarding the design and fabrication of façade’s modules.

Focusing on design, upon evaluating historical precast facades, it is recognized that sculptors often collaborated with architects in designing the façade modules. Today, in the context of REVIT product libraries, designing complex geometries for facades poses a challenging task for architecture students. Conversely, if students are willing to tackle the challenge of designing sophisticated geometry, their initial attempt might very likely involve a Grasshoppered Voronoi pattern. One of the objectives of analyzing historical precast elements is to enable students to comprehend how having control over geometry at a micro level, coupled with a deep understanding of the geometric constructs such as lines, curves, and cross sections, can lead to the creation of complex yet sophisticated elements.

Focusing on fabrication and construction courses currently offered at the architectural schools, two main approaches are commonly found: abstract construction courses that require students to create detailed drawings of building assemblies, and Design-Build studios where students participate in physically assembling building components. While both approaches in solo or in some level of combination hold value, courses that emphasize creating wall section detail drawings or a participating in Design-Build studios often tend to be disconnected from design in its artistic form, where aesthetics and proportions of the geometry play a crucial role. This disconnection arises because creating construction drawings or engaging in the construction process consumes a significant amount of time for students, demanding the majority of course time to be dedicated to these tasks.

Analyzing historical precast facades for their design and fabrication followed by fabricating the elements using contemporary means, inherently limits the scope and the scale of the project. This limitation encourages students to focus on both design and fabrication. This middle-ground educational approach allows ‘design’ to maintain a balanced role alongside hands-on activity, rather than being diminished or eliminated altogether. Today, in the context of rapid prototyping where state-of-the art 3D printers are readily available to students, another objective of the studio was to demonstrate the potential of AM beyond model making. The pedagogy highlights the disruptive impact that these technologies can have on the building industry. This pedagogical method could serve as a precursor for introducing other innovative methods for creating precast elements, such as 3DPF.

A six-credit design studio, paired with a three-credit hour seminar, was offered at the Illinois School of Architecture of the University of Illinois at Urbana-Champaign, with access to Digital Technologies in Architecture (DiTA) Research Lab's ¹ resources. The syllabus for both the studio and the seminar was developed in tandem. The studio incorporated design and fabrication activities, while the seminar aimed to enhance students’ understanding of the field’s state-of-the-art knowledge through in-class discussions on selected readings and guest lectures from academia, industry, and practice. Additionally, the seminar provided computational design lessons to develop students’ computational design skills.

The design studio comprised two projects. The first project, lasting 4 weeks, allowed students to delve into historical precast facades, focusing on studying their design and fabrication technologies. The second project, spanning 10 weeks, tasked students with exploring the application of AM in the design and fabrication of self-standing precast elements. This paper focuses on reviewing the first project.

Project one started with the presentation of a series of precast facades and self-standing sunscreens featuring sculptural, complex, and custom-designed modules to the students. They were asked with selecting a precedent, conducting research on it, and reproducing the forms in a computational design environment by interpreting the images and found drawings. Additionally, the students were introduced to AM and educated about the file-to-factory workflow for transforming computational forms into 3D printed objects. They were also trained on operating 3D printers. Once the students had fabricated the module’s geometry, they were introduced to rubber mold-making techniques, specifically OOMOO-25 silicon rubber compound by Smooth-On, referred to as rubber mold in the text. Students were guided to create a one or two-part rubber mold for casting multiple instances of the precast modules before assembling them into a facade or a self-standing screen. It is noteworthy that Hydraulic Expansion Cements (HEC), particularly Rockite, were used for casting instead of traditional concrete made up of coarse aggregate. Rockite only requires mixing with water and sets rapidly, allowing demolding within an hour of casting.

From another perspective, various precedents had different module sizes, ranging from modules that covered a floor-to-floor height, to modules similar in size to a brick. To ensure a consistent workload for students in terms of the time needed to 3D print the positive part and the amount of material needed for creating the rubber mold and casting, all students were instructed to 3D print their case study modules to fit into a 5” × 5” × 5” (12.75 × 12.75 × 12.75 cm) cube. They were directed to ensure that the longest dimension of the cube could fit within the aforementioned dimensions. This approach facilitated the creation of cohesion among the level of detail visible in the modules while requiring roughly the same amount of material for the students’ projects.

Upon completion of the project, the outcome was evaluated based on four criteria:

- Analysis: Students conducted an analysis the geometry of the precast façade elements, assessing their repeatability and the reusability of the mold by evaluating the number of designed module types covering the façade and the number of required mold types.

Three historic facades are selected to be exemplified in the following sections. These three case studies serve as significant examples of precast building envelopes. The first precedent, the American Cement Building, formerly the industry’s headquarters in Los Angeles, USA, was designed to showcase the strengths of precast. The second precedent, the IBM building in Honolulu, is another instance of a corporate building integrating precast elements into its façade, highlighting its distinctive form. The third precedent, the Welbeck Car Park in London, UK, represents car park as a civic movement where precast modules were utilized to cover the building. Each historic precedent is subsequently followed by reviewing the students’ approach to its analysis, modeling, and fabrication.

American Cement Building’s screen in Los Angeles

Some sculptors, such as Malcolm Leland, collaborated with architects to design modular volumetric elements for building facades. The American Cement Building in Los Angeles was designed by Daniel, Mann, Johnson, and Mendenhall (DMJM) in collaboration with Malcolm Leland (1961). The executives of the American Cement Corporation aimed to construct a new headquarters that would promote reinforced concrete as the building material of the future in dramatic fashion. The Corporate Office Buildings were converted into loft buildings in 2011. ¹⁵ Daniel, Mann, Johnson, and Mendenhall designed the 13-story building as a two-unit structure: a 4-story base for parking, ground floor shops, lobby, executive offices, and a 9-story tower ¹⁶ . Kiewit Construction won the construction contract through a design-build procedure deemed novel at the time. ¹⁶ “Some of the structural elements were poured in place using wooden forms; others were poured in place using plastic-faced wooden forms to produce a textured effect. Some of the individual members (X-members, garage grill elements, floor beams) were cast away from the site, hauled in, and hoisted into place”. ¹⁷ The precast concrete X-members are both dramatic and functional. The 11-foot (∼3.35 m) reinforced members take lateral forces on the north and south sides of the building. Therefore, no interior columns are above the fourth floor. The 450 X-members weigh 4500 pounds (∼2041 kg) each. ¹⁸ The corners of the elements allow them to hook onto the edge of the upper and lower slabs. The pre-installed rebars in the precast member appear to be in place to create a monolithic union through on-site concrete pouring once modules are placed.

The other module designed for the building was the sculptural precast screen enclosing the garage that provides ventilation, protection, and a rich surface texture that is harmonic with the X-member wall. ¹⁷ The 250 sculptured concrete garage grille members weigh 3000 pounds (1360 kg) each. ¹⁸ “The parking annex façade is non-load bearing” ⁶ (p. 55).

The student who reproduced the X-members used computational modeling (Figure 1-left) and 3D printing to create the module, with its geometric proportions being very close to that of the precedent. According to the fabrication report submitted by the student, their main challenge was designing, fabricating, and casting a mold. The student first made a two-part mold for casting the pieces. But since the rubber mold material and the activator were not thoroughly mixed, the rubber mold never cured. In addition, the student dismissed applying the mold release agent to the mold parts during production; thus, the two parts of the mold became homogeneous and never came apart. In the second attempt, the student designed a one-part mold (an open mold) and observed how eliminating the process of producing a one-part mold increases the production speed. After successfully making the mold (Figure 1-center), the student’s initial attempts to cast it all failed because of the module’s narrow legs, prone to breakage. In this stage, the student paid closer attention to various cement-to-water ratios. The 2:1 ratio created brittle elements, whereas the 4:1 ratio created successful intact cast pieces without using any reinforcement. In this phase, the student also gained a qualitative understanding of concrete mix design, where a higher cement portion in a mix makes the cast piece more brittle. In contrast, a higher water portion can slow down the solidification process. Finally, the student adhered the cast pieces together to create the standing façade (Figure 1-right).OPEN IN VIEWER

Upon completion of the project, the instructor realized that learning about precast joinery is crucial in this project. Although the geometric proportions of the module and assembly were a close replica of the reference case study, the method of construction and module attachment to the slabs was not reflected in the final assembled model. Even though the student modeled, and 3D printed the notches behind the X-member’s legs for connecting the module to the floor slabs, they filled it out with clay upon casting them to increase the strength of the cast piece. Therefore, they used glue to assemble the modules into a facade instead of creating an assembly that closely represents the construction method. Changing the modules’ scale in the model and increasing the duration of this project will allow students to dive deeper into the connection details and construction techniques of the precedents.

Another student worked on modeling and fabricating the precast module used in the garage grillage of the American Cement Building. The geometry of this module was complex and challenging for them to model computationally. According to the fabrication report submitted by the student, one of the lessons that they learned was how to analyze complex geometries and break them down into simple shapes or curves before modeling them. The student was successful in making a two-part mold following all the instructions. However, they experienced failure during the casting phase. They, too, oscillated between making a thick mix, not allowing concrete to move in the mold, and creating a runny mix resulting in fragile cast pieces. It was on the third attempt that they got the mix right. The student also observed the importance of applying pressure to mold pieces, so the mold does not leak. They also realized the importance of vibrating the molds for outing the air bubbles by shaking the mold after casting. Finally, the student used cement to connect the modules to one another. They had correctly understood that the modules are non-load bearing and are stacked before being bound between the lower and upper slabs of the building (Figure 2).OPEN IN VIEWER

An important observation the student made was relating the three-step mold-making process to real-world precast concrete production, except that more advanced tools and reinforcement materials are used in the industry. This observation is crucial and demonstrates that making the small-scale mockup has engaged the student in fabricating a component and constructing an assembly which is beyond making an architectural model. In fact, the student understood the act of building.

IBM building in Honolulu

The IBM building, located in Honolulu, Hawaii, was designed by Vladimir Ossipoff and Associates Inc. and was built in 1962. The 6-story office building is framed with prestressed concrete joists and composite poured-in-place concrete slabs and beams. A honeycomb concrete brise soleil inspired by Polynesian culture intending to resemble the punch cards used in the computer industry wraps around the building. The decorative concrete grill comprises 1360 precast pieces surrounding the upper floors, providing an effective sunshade ¹⁹ (Figure 3-left). By inspecting the construction images, each vertical row of the concrete grill consists of smaller precast sections before being attached to the slabs (Figure 3-right). For module-to-module connections, the top faces of the modules have insertion holes for incorporating rods that vertically connect the grill modules to each other. Figure 4-left shows the modules laid down on the construction site where the insertion holes are visible on the top surface of the module. Figure 4-right shows the construction detail drawing. According to construction documents, the insertion holes are 12’ 2” (∼35.56 cm) deep, and it is instructed to fill the ¼” (6.35 mm) holes in the lower part of the precast module with non-shrink grout just before positioning the upper module. For module-to-slab connections, there is an insert for passing stainless steel rods onto the side of the precast modules (Figure 5-left). Figure 5-right shows a construction drawing that instructs to fix the joint by inserting the steel rods into the concrete slabs.OPEN IN VIEWER

OPEN IN VIEWER

Figure 4.

Module-to-module connection: the precast module laid flat on site, showing two insertion holes for inserting rods (left | image courtesy of Dr. Don Hibbard’s donation to the Haigo & Irene Shen Architecture Gallery, University of Hawaiʻi at Mānoa); IBM Building, drawing of module-to module connections in elevation as seen in the façade (right | Image courtesy of the Ossipoff & Snyder Architects Collection, University of Hawaiʻi at Mānoa).

OPEN IN VIEWER

Figure 5.

Module-to-slab connections by inserting stainless steel rods into floor slabs (left | image courtesy of Dr. Don Hibbard’s donation to the Haigo & Irene Shen Architecture Gallery, University of Hawaiʻi at Mānoa); IBM Building, drawing (right | Image courtesy of the Ossipoff & Snyder Architects Collection, University of Hawaiʻi at Mānoa).

The student approached working on the project by modeling the smallest geometric module that was making up the honeycomb facade (Figure 6-left). They first modeled this smallest module in proportions larger than the actual module size and created a two-part mold. By casting the two-part module, they realized that using one half of the mold and using it as an open mold would not only create a module with the correct proportions but would make the casting and demolding process straightforward. The 3D printed piece used as the positive part for making the rubber mold and the open mold are shown in Figure 6-center. Afterward, the student used the open mold to cast the smallest module of the façade multiple times. They then glued the smaller pieces together using hot glue to make the grillage assembly before connecting them to the lower and upper floor slabs (Figure 6-right).OPEN IN VIEWER

The student’s approach to analysis and fabrication is both similar to and different from Ossipoff’s design for the IBM facade. In Ossipoff’s design, there exists a base precast module (each containing six filleted edges), and then three of the precast modules are assembled to create a vertical grillage covering floor-to-floor which is fixed onto the slabs. This vertical grillage assembly is then connected to the other vertical grillages. The student correctly interpreted this logic. However, they broke down the smallest unit even more, and modeled the base precast module with two filleted turns instead of six. Therefore, the student had to connect three of their pieces to get the base module in Osiipof’s design. This interpretation error resulted in having three times more connections in each vertical grillage compared to the precedent. The second divergence in student’s analysis compared to the precedent was related to the size of the modules. These two deviations in analysis affected the aesthetics of the student’s project. Finally, the student’s approach to the method of connecting modules together was different from the precedent. The student glued the pieces together instead of embedding holes for inserting any dowels representing connecting rods for module-to-module and module-to-floor connections. It should be noted that students did not have access to historical drawings, nor was it strictly required in the project description. However, the student found the construction images towards the end of the project’s duration. Since there needed to be more time to redo the mold fabrication and casting, they only addressed the analysis deficiencies in their final report. Communicating the failures and lessons learned has been an essential part of this project.

Welbeck Street Car Park, in London

The Welbeck Street Car Park (1968–70) behind the Oxford Street branch of Debenhams department store in London was a fascinating patterned grid. ²⁰ Unfortunately, this building was demolished despite many pleas to save it. ²¹ Built in 1970 by the little-known Michael Blampied and Partners, this was a car park as a civic movement. ²² It had eight full upper floors for parking, including the rooftop. The exterior walls above the ground were built using triangular precast concrete units, carrying floors made up of precast concrete beams and slabs with a topping of reinforced concrete poured in situ. ²³ The highly permeable elevations on three sides were required to achieve the necessary natural ventilation. ²⁴ The modular geometry of the wall units has been an obvious appeal of this parking. The primary module type that is used in the façade is V-form. In his book, The Architecture of Parking, Simon Henley explains that the initial scheme of having structural columns and non-structural cladding was rejected due to economic reasons. Instead, “a system of precast concrete, load bearing Y-columns were developed to support the building perimeter. The Y, which required additional restraint [at] mid-story at the apex, evolved into a V that would span and support successive floors”. ²⁴ (p. 125) The V-form is not consistently solid. Henley (2017) elaborates that “from the outside, each V-form is hollow, the cables behind it [are] making it possible to carve out the center. The diagonal array creates a diamond pattern, in which each diamond consists of a part-solid triangle below (the structural V) and an inverted empty triangular void above (the space between two structural Vs)”. ²⁴ (p. 129) The pattern conceals the conventional section to portray the structure as a single, wicker-like frame that measures 68.89 feet (21 m) high, 118.11 feet (36 m) wide, and 131.23 feet (40 m) long. The modular unit was reduced to a 6.5 feet (2-m) built unit to fit all three elevations. In the development phase, models were made at 1:12, 1:4, and a full-scale polystyrene mock-up. The latter model was left outside to test weathering and marking. ²⁴ The final units were made by Atlas Stone Co. Ltd, incorporating drainage grooves at the junctions of the facets. The designers addressed the safety aspects of the openings by adding stainless steel wires threaded vertically through the building to act as a barrier, which is unseen from the outside. ²³ Although the façade was covered with V-modules, there was a second type of precast module used at the corners of the façade. This particular precast component is used at two-story intervals and links the regular pattern of V-sections. “Each is again a V but realized in three dimensions”. ^{24 (p. 131)} The result is a sculptural building made of tessellating forms and faceted surfaces.

The student approached working on the project by modeling the modules seen on the flat sides of the parking façade (Figure 7-left). Next, they slightly manipulated the exterior face of the modules to capture the movement of the 3D printer and control the pattern that will be created onto the 3D printed reference piece, thus the cast pieces (Figure 7-center). They then designed an open mold for casting multiple module pieces before connecting them to a wooden base to make up the façade (Figure 7-right). This project demonstrated how students can recognize the effects of reference piece’s manufacturing method on the mold, and then on the cast pieces. One area of improvement for this project would have been investigating mold making approach for the corner module where the façade turns. Although the student correctly analyzed the number of module types, thus mold types required for casting them, the project duration did not allow time to investigate the second module in depth.OPEN IN VIEWER

Discussion

Several additional case studies beyond the three precedents examined above were assigned to students for analysis and recreation. These cases encompass examples of self-standing screens created by sculptors like Erwin Hauer, who crafted a series of modular structural sculptures featuring prominent interior voids bounded by continuous surfaces used as self-standing room dividers. ²⁵ Other case studies include more contemporary façade screens such as Prosolve370e, a decorative architectural module made of lightweight thermoformed plastic attached to a steel system in Mexico City. ⁶ The instructor’s observation is summarized under three categories discussed below.

Results

A total of 10 students were enrolled in the studio: six first-year graduate students and four second-year graduate students. At the beginning of the semester, students’ familiarity , comfort , and proficiency with computational design and digital fabrication tools was assessed. Familiarity was defined as having encountered the tool before by having opened it at least once. Comfortability was defined as being able to use the tool to some extent. Proficiency was defined as having the ability to effectively use the tool. Rhino NURBS modeling platform ²⁶ was suggested as the preferred CAD platform to be used by the students. Regarding CAM skills, AM was the key skill that was surveyed. The students’ skillset in CAD and CAM prior to the start of the studio is summarized in Table 1.

OPEN IN VIEWER

Table 1.

Students’ level of proficiency, comfort, or familiarity on a scale of 0 to 3 with CAD and CAM techniques.

	Student 01	Student 02	Student 03	Student 04	Student 05	Student 06	Student 07	Student 08	Student 09	Student 10
CAD skills survey- Rhino	3	2	1	1	2	1	3	2	3	2
CAM skills survey- AM	1	1	1	1	1	1	1	1	1	1

Synthesizing the information presented in Table 1, all students expressed having some level of familiarity with Rhino. Four students out of 10 stated that they are comfortable with it, while three students indicated that they are proficient in using it. Regarding digital fabrication techniques, all students expressed that this studio has been the first digital fabrication studio that they are taking. Students’ knowledge on using 3D printing was explicitly assessed. All students expressed familiarity with the concept of 3D printing. However, they had never used it in any of their school projects, and they expressed that they had never directly operated a 3D printer.

After project one was completed, the projects were assessed based on four criteria: analysis, modeling, AM, and mold making/casting. Table 2 summarizes the assessment results under three categories named ANALYSIS, MODELING, and FABRICATION. It should be noted that FABRICATION score encompasses both AM and mold making/casting. The criteria for evaluation are set as Successful, Borderline, or Unsuccessful.

OPEN IN VIEWER

Table 2.

Assessing students’ projects on Analysis, Modeling, and Fabrication.

	Student 01	Student 02	Student 03	Student 04	Student 05	Student 06	Student 07	Student 08	Student 09	Student 10
ANALYSIS	S	B	B	S	S	S	S	B	S	S
MODELING	S	B	U	S	S	B	S	B	S	S
FABRICATION	S	B	U	S	S	S	S	S	S	S

Looking at the ANALYSIS assessment in Table 2, all students received either a Successful or Borderline assessment. Regarding MODELING, only one student (student 3) received an Unsuccessful assessment. Upon revisiting the initial CAD skills of this student in Table 1, it is evident that they were only ‘familiar’ with the CAD tool. However, two other students with the same level of CAD familiarity at the beginning of the semester (students 4 and 6) achieved a Successful and Borderline MODELING assessment. Respectively, even within the small sample size, there is considerable variability in the correlation between students’ initial CAD skills and their success in delivering a CAD model. Therefore, further studies with larger sample sizes are necessary to explore this correlation.

Regarding FABRICATION, all students received either Successful or Borderline assessment except for one student. Relating the FABRICATION assessment to the initial skill level in 3D printing does not explain this correlation, as all other students were initially unfamiliar with AM. However, an Unsuccessful MODELING assessment correlated with Unsuccessful FABRICATION.

From a different perspective, although some students received Borderline assessment in CAD MODELING, they were able to compensate for that during FABRICATION phase. It is possible that through iterative hands-on activities, they manually improved upon their initial modeling approach without necessarily returning to the CAD environment for making corrections. For example, one student filled out notches of a 3D printed model with clay instead of revising the CAD model. Similarly, another student accidently discovered that one part of a two-part mold provided correct geometric proportions, without needing to revisit the CAD model and reprint it. This may explain why many students achieved Successful FABRICATION assessment, regardless of whether their MODELING scores were Successful or Borderline.

Conclusion

Almost 70 years after Saul Steinberg mocked curtain walls as ‘graph paper architecture’, architectural education mainly emphasizes educating the future generation of designers on creating high-performance curtain wall systems. This paper presents a pedagogy that foregrounds the overlooked aspects of design and fabrication of building enclosure, by focusing on studying precast façade modules. Today, in the era of computational design of non-identical elements and emergence of advanced fabrication techniques, the design of sophisticated and complex geometries for facades can extend beyond the current slim percentage found in the works of star architects. Considering the raised awareness on the negative environmental impact of concrete, future precast focuses on optimized design methods and efficient fabrication techniques. Additive manufacturing in particular, can transform precast by 3D printing the reference piece that is used in the mold making process.

This paper presents a pedagogy that bridges the past and the future by reinterpreting historical projects and recreating them using AM. In the context of REVIT product libraries or Grasshoppered Voronoi patterns, one of the objectives of analyzing historical precedents is to allow students to understand how having control over geometry at a micro level with deep understanding of the geometric constructs such as lines, curves, and cross sections, can lead to the creation of complex yet sophisticated elements. In the context of rapid prototyping where the state-of-the art 3D printers are readily available to students, another objective of the pedagogy is to demonstrate to the students the potential of AM beyond model making. The final objective of the instructor is to create a middle-ground course that goes beyond creating detail drawings of wall sections but does not become a design-build studio either. This course was envisioned to balance design’s role in its artistic sense with hands-on activities that demonstrate the labor and constraints of fabricating and constructing precast facades.

Focusing on students’ approach to CAD MODELING, and FABRICATION of precast facades presented in the precedents, the findings are summarized below:

• Even within the small sample size, there is variability among the correlation of students’ initial CAD skills and their success in delivering a CAD model. This correlation requires further studies with larger sample sizes.

The general observations on the design-fabrication iterative loops are summarized below:

• Teaching about the ignored aspects of building enclosure such as design of precast, design of mold, and fabrication allows students to develop an insight into interweaving architects’ role as designers and builders.

Modifications are planned for future project development.

• Learning about precast joinery is crucial, as many projects relied on perfect surfaces that could be created weak and delicate for connections. Therefore, the digital fabrication of components needs to be understood in parallel with connection details and joinery of precast elements. This includes understanding how modules are connected to the primary structure (slabs, frame) and each other.

By combining historic examples with contemporary technologies, pedagogy introduces students to the rich overlaps between the field’s subdisciplines, productively blurring boundaries between design, fabrication, assembly, and history. The resulting work inspires deep discussions about ornament, detail, form, and technology. But it is also impactful, showing students how digital technologies can be part of a wide palette of tools available to designers—and constructors. Digital fabrication methods result in revised understandings of materiality and structural textures expected by traditional craftsmanship or industrialized production. These understandings are valuable preambles for introducing new ways of fabricating precast elements towards making it more cost and labor efficient.