Content uploaded by Niloufar Emami
Author content
All content in this area was uploaded by Niloufar Emami on Feb 18, 2022
Content may be subject to copyright.
The Design and Fabrication of Variable Façade
Panel Systems
Tanner Theisen 1, Niloufar Emami 2
1 Louisiana State University, United States of America
ttheis1@lsu.edu
2 University of Illinois at Urbana-Champaign, United States of America
nemami@illinois.edu
Abstract. This study explores how alternative molding methods can be utilized to
construct a variety of prefabricated volumetric concrete panels for façades from a single
digitally fabricated mold. First, precedents were studied and panel variability was
classified into ornamental, geometric, or assembly categories. Then, a molding method
was proposed that improves upon traditional processes. Traditionally, geometrically
varied façade panels are realized through creating a single mold for every variant, which
is inefficient and wasteful. The proposed method allows for reusability of a singular mold
which can fabricate variated panels through utilizing interchangeable mold inserts. This
proposal was tested on a small scale through creating rapid iterations of molds
fabricated with Stereolithography (SLA) printing. Emerging big area additive
manufacturing (BAAM) technology allows for the proposed methods to be utilized at the
industrial scale where they can reduce the cost, labor, and time of fabricating varied
concrete panels while also creating complex geometries.
Keywords: Variability, Precast Panels, 3D Printing Mold, Custom Repetitive
Manufacturing, Facades.
1 Introduction
1.1 Limitations of Precast and the Case for Additive Manufacturing
Concrete precast building systems have become widely popular since the early
20th century (Sutherland, 2001). Adopting concrete prefabrication methods as
an alternative to in situ casting has many benefits, one being increased
productivity in construction (Nahmens et al. 2011). This increase in productivity
typically only largely applies to the fabrication of standardized objects through
restricting the variety of possible panel designs in order to simplify production.
In addition, precasting is labor intensive. Oftentimes, formwork elements
1257
SIGraDi 2021 | Designing Possibilities | Ubiquitous Conference
exclusive to the desired product are assembled on stationary tables by hand
(Pan et al. 2019). This requires the utilization of skilled laborers and limits the
ease at which complex geometries can be formed. From a different perspective,
the use of standardized façade panels by designers simplifies production, but
can lead to monotony and regularity amongst façades (Feng Li et al, 2020).
Given those restrictions, if a designer requires panel variants on their façade,
precasters would have to fabricate a different mold for every panel. Having to
fabricate a large number of molds is both time consuming and costly. For
instance, according to Love et al, the cost of fabricating a traditional wood mold
can be around $3,000. Typically, that mold can only produce twenty panels
before it becomes unusable, which means the cost per pour is $150 (Love et
al, 2019). In a highly varied façade a mold may be used as few as once,
meaning that the cost per pour becomes the full price of constructing the mold.
Despite the increase in geometric complexity of façade panels as a result of
computational design tools, precast concrete manufacturers at large do not
currently have a strategy to fabricate complex and varied panels without
fabricating a large amount of formwork. Therefore, there is a need to
conceptualize alternative fabrication methods that allow for variable production
of architectural panels without compromising fabrication efficiency. A possible
solution to the inefficiencies of precasting can be found in digital fabrication
(DF).
The exploration and application of DF for casting concrete has become a
popular field of research and is widely referred to as Digital Fabrication with
Concrete (DFC) (Asprone et al, 2018). Studies have been done on the
applicability of a wide variety of methods to either fabricate the formworks of
concrete through subtractive or additive processes, or to directly 3D print
concrete. These include but are not limited to robotic hot wire cutting of
polystyrene formwork (Martins et al, 2019), PLA fused deposition modeling
(FDM) for creating one-off formwork (Jipa et al, 2019), Smart Dynamic Casting
(SDC) of concrete (Lloret-Fritschi et al, 2017), and 3D concrete printing (3DCP)
(Asprone et al, 2018). This study focuses on the production of formworks for
precast concrete panels.
Existing DFC techniques that are concerned with the production of concrete
formworks are successful in a number of ways. They succeed in casting
geometries far more complex than those produced through traditional methods,
and the labor required to assemble the form is reduced through automation of
its fabrication. (Jipa et al, 2017) However, the methods are mainly used for
bespoke fabrication and thus lack reusability since the forms can typically only
cast one specific geometry. Many of the current bespoke fabrication methods
eliminates the possibility of a variable product from a single mold print. Further,
the vast majority of formworks created through these methods require time-
intensive production. These shortcomings will prevent the current DFC state of
the art from completely replacing the current precasting methods. However, the
strengths of additive manufacturing (AM) cannot be ignored. AM can be applied
to aspects of the currently existing precasting methods to improve the possible
geometric complexity of cast objects. In fact, studies have been done that show
when utilizing mold inserts fabricated through big area additive manufacturing
1258
SIGraDi 2021 | Designing Possibilities | Ubiquitous Conference
(BAAM) the cost per pour to produce a panel can be reduced by over a third
(Love et al, 2019). While AM methods can be feasibly utilized for architectural
scale production, additional design exploration is required to implement
variability to the fabrication process.
1.2 Objectives
Given that neither the withstanding precast methods nor the emerging DFC
techniques are able to cast a large number of varied products without
fabricating a large number of molds, a new approach is needed. The question
is how can panel variability be designed considering that digital fabrication will
be employed for realizing the design?
This study aims to explore how a digitally fabricated mold can produce
varied façade panels. The objective of this study is to formalize the way that
variability in façade panels can be created considering digital fabrication
techniques for realizing the panels. More specifically, our objective is to design
and fabricate a variety of geometrically complex panels that are realized
through interchangeable mold inserts. This new approach to precasting will
allow for the rapid casting of varied products, which is a necessary step in
making varied façade panels more accessible to designers. Computational
design platforms namely Rhinoceros NURBS modeling platform are used for
design, digital fabrication techniques, namely 3D printing, are used for
prototyping. The methods are described in detail in the following section.
2 Methodology
The study began by classifying different types of variability using observations
made on built facades. When succinct definitions of different types of variability
were developed, the design and fabrication of façade panels and molds began.
Two facades were fabricated through an innovative approach for creating
variability, and were investigated further. The Le Vérone Building and the Perot
Museum of Nature and Science were two case studies that we focused on since
they both utilized mold inserts for creating variability through a single mold.
These two case studies created the base for exploring the design and
fabrication of variability. Inspired by Le Vérone Building’s panels, Panel A was
designed. It had a prescribed geometry which we introduced variability to
through a new molding method. Inspired by the Perot Museum of Nature and
Science, Panel B was designed. It had an opposite approach, in that the design
was based solely around the new molding method. Rhino 3D was used for
computational design, and custom repetitive manufacturing (CRM) with AM was
used in fabricating the molds, which were then used to cast varied panels. For
the casting we utilized Rockite expansion cement due to its fast set time, and
1259
SIGraDi 2021 | Designing Possibilities | Ubiquitous Conference
because its mechanical properties are similar to that of ultra-high-performance
concrete (Rockite Cement, 2017)(Portland Cement Association, n.d.).
2.1. Observing and Categorizing Variability
It is important to critically assess, rationalize, and provide definitions of
variability before attempting to design and fabricate varied panels. To do so,
general observations on built façade panel variability were made. While the
variations between panels were unique from project to project, it was observed
that there are similar overall strategies to create that variance. To start, every
project’s panel variants did not dramatically differ from other panels on their
respective facades, there was always a holistic aesthetic. Further, variations
can oftentimes be observed as a change in the size and/or distribution of the
additions to and/or subtractions from a singular overall panel geometry. Panels
may also vary through an elimination from or addition to the overall geometry
of the panel. These observations are not, by any means, an all-encompassing
list of the possibilities of creating variations between facade panels. However,
it seems that there are two different broad classifications of variability amongst
built facades: ornamental variability and geometric variability. While thinking on
the possibilities of variability beyond modification of the panel’s geometry or
finish, a third classification was conceptualized: assembly variability.
- Ornamental variability occurs when panels have the same overall
geometry but a range of different ornamentation on one or both of their
surface(s).
- Geometric variability occurs when panels have a range of different
measurements for their final cast panels. With this classification, the
measurements of the outer boundaries of the panel changes between
panel variants.
- Assembly variability refers to the panel’s qualities upon on-site
assembly. Similar to a brick that is agnostic to orientation in certain axes,
a panel with assembly variability has a design that allows for it to be
placed freely on a facade to create a number of different overall facade
arrangements. The massing and surface quality of the panels would be
the same, but the design would be ambiguous enough to place the
panels anywhere on the façade without compromising the overall
design.
After understanding and categorizing variability in the selected cases,
fabrication methods for accommodating variability at the assembly level were
studied.
2.2. Existing Fabrication Methods of Variable Facades
It has been observed by the authors that the number of panel variants on any
façade with cast panels is typically low in order to reduce the need to fabricate
multiple molds. However, some precasters have developed strategies to reduce
the required number of molds. Among the studied built cases, there were two
projects where variability was created by utilizing mold inserts. These projects
1260
SIGraDi 2021 | Designing Possibilities | Ubiquitous Conference
include Le Vérone Building and the Perot Museum of Nature and Science
fabricated by Techni Moulage and Gate Precast respectively.
Techni-Moulage is a small company based in France with a singular facility
that produces mostly architectural precast objects. In this facility they fabricated
the panels for the renovation of Le Vérone (Fig. 1), a building in Saint-Denis
designed by Wilmotte & Associés Architectes for the online retailer Vente-
Privee. The façade is composed of a fishnet pattern designed by Pucci de
Rossi, an Italian artist and friend of the Vente-Privee CEO (Wilmotte &
Associés, n.d.). In fabricating the 12 panel variants for the curved façade,
Techni-Moulage used only a flat and a curved mold. To create the variations
required, certain channels in the mold would be blocked off with reusable mold
inserts, which minimized the need for more molds (Techni Moulage, n.d.).
Figure 1. Le Vérone Building Facade. Source: Photo by authors
The second project was the Perot Museum of Nature and Science. Gate
Precast worked closely with Morphosis in conceptualizing a fabrication strategy
to construct the façade for the Museum (Fig. 2). The intent of the façade design
was to evoke geological striations, and included 656 panels with unique designs
(Stephens, 2013). To fabricate those unique panels within budget, the team
composed of both Morphosis and Gate Precast designers developed a mold
system that utilizes interchangeable parts. These parts could be inserted into 4
differently shaped formworks to create 39 different geometries in the panel,
either recessed or protrusions (American Institute of Architects, 2014) There
were 12 different patterns, sometimes continuing from panel to panel, but when
one views the façade the patterns seem completely random. Their advanced
approach was different from traditional formwork construction, which allowed
them to reduce the number of required molds during the fabrication process.
Overall, the fabrication strategy of the design team was highly successful in
creating variability for this project.
1261
SIGraDi 2021 | Designing Possibilities | Ubiquitous Conference
Figure 2. Perot Museum of Nature and Science Façade + Molds. Source: From
“Perot Museum Exterior Architecture” [Photograph], by Jonathan Cutrer, 2017, Flickr
(https://www.flickr.com/photos/joncutrer/37585013761). CC BY-SA 2.0;
The next sections describe in detail how Panels A and B were designed with
inspiration from Le Vérone Building and Perot Museum of Nature and Science
respectively. The former panel (panel A) prioritizes Design by conceptualizing
a fabrication method that can accommodate a specific geometry. The latter
panel (panel B) prioritizes Fabrication by exploiting design possibilities that the
conceptualized fabrication method offers.
2.3. Panel A: From Design to Fabrication
Variability in facades can be designed with both a prescribed geometry, or
with a prescribed fabrication technique. For panel A, the former approach was
taken and the fishnet geometry of Le Vérone building was mimicked. Because
the design aesthetics were prescribed, the major hurdle when designing the
panel became organizing the geometry so it allows for the mold design to
accommodate the digital fabrication of variability. Because of this, one must
consider how to fabricate the panels at every step of design past
conceptualization, and thus the new molding method was born very early into
design exploration. To increase the variability of the panel in a controlled
manner, it was found that the mold could be split into even segments that are
both interchangeable and rotatable along the XY plane. This concept was
applied to organize the geometry of Panel A by setting up the design with a 2x2
square grid. On all four sides of each individual square cells, 2 touchpoints were
placed equidistant from the center of the square (Fig. 3). The geometry of the
panels was developed around these guides, and with this controlled
segmentation of the design the mold could be composed of 4 pieces that are
all interchangeable and rotatable (Fig. 3).
1262
SIGraDi 2021 | Designing Possibilities | Ubiquitous Conference
Figure 3. Initial organization of panel A (left) and two variants (middle and right).
Source: Photo by authors
For digital fabrication of the mold, SLA desktop 3D printers were used. It
was decided to utilize an elastic resin because the thin geometry of the panels
likely required a flexible mold. The first iteration of the mold was unsuccessful
in casting trials because the design was essentially just the thin geometry of the
panels subtracted from a box. The large mass of the mold pieces prevented it
from being very flexible, and thus prevented the cast object from being removed
from the mold without a significant overpour. The design was revisited after
being informed by a few other studies into thin-walled formwork (Naboni et al,
2018), and a thin-walled mold iteration was printed that was able to produce the
delicate panels without breaking them during the demolding process (Fig. 4).
However, when arranging the pieces in a tray the contact points between the
thin-walled geometry were tricky to perfectly align and proved to be leaky. To
prevent that, a number of clips were printed that could be placed at the contact
points of the pieces to both align them and prevent leakage.
Figure 4. Picture of four 3D printed mold pieces placed in a tray. Source: Photo by
authors
The experiments of designing and fabricating panel A with a prescribed
geometry proved to be successful in that it limited the number of unique mold
pieces to four, while the number of unique panel variants surpassed that.
Although we did not experiment with creating panel variants using the same
mold pieces inside a tray, this could also be done to further increase the
possible number of unique panels.
1263
SIGraDi 2021 | Designing Possibilities | Ubiquitous Conference
Figure 5. Pictures of panel A variants. Source: Photo by authors
2.4. Panel B: From Fabrication to Design
For panel B the idea of interchangeable and rotatable mold pieces was
expanded upon. The conception of the design started around the idea of using
a square grid larger than the previously used 2x2 grid to organize the geometry.
This time, however, the design of a particular connection point was completely
avoided when arranging the 4x4 grid (Fig. 6). The panel’s design developed to
become ordered around 6 different inserts that vary in both the height and
complexity of their top surfaces, but have the same length and width (Fig. 6).
All but one insert have a doubly curved surface, and the heights along the upper
edges of the inserts correspond with the height of one or two other inserts. The
correspondence in height presents the opportunity to cast a panel with a
complex and smooth surface. However, the correspondence in length and width
present the inverse opportunity, the inserts may be moved interchangeably or
rotated in the mold to create jagged edges. The opportunity to construct a
façade that has patterns continuous from panel to panel becomes present with
this design, similar to that of the Perot Museum of Nature and Science.
Figure 6. Elevation of panel B inserts and two variants with 0.5mm topological
overlays. Source: Photo by authors
Because of the nature of mold B to act more as a form liner than a formwork
it was decided to fabricate it out of clear resin, as less flexibility was required.
The lack of flexibility of the mold pieces did, however, cause problems. We
found that the thickness of the pieces needed to be a certain depth to be
1264
SIGraDi 2021 | Designing Possibilities | Ubiquitous Conference
successfully removed from their printed supports. Otherwise, the brittle nature
of the clear resin would cause the pieces to break from the force of the supports
being removed. In the process of removing the supports, multiple of the thinnest
pieces were shattered. Further, because of the difficulties removing the
supports the pieces did not have even bottom surfaces. When casting, we
elected to use a 3x3 grid of mold pieces instead of 4x4 as a result of the broken
pieces. To allow for the top surfaces of the pieces to align when casting we
elected to place them in a box of sand, thus eliminating the need for a finished
bottom surface. An alternative approach could be to finish or reprint the bottom
surfaces so that the pieces are level, which is likely a better option for quicker
mold assembly (Fig. 7).
Figure 7. Picture of mold B. Source: Photo by authors
Figure 8. Pictures of panel B variants. Source: Photo by authors
3 Discussion
All design and fabrication was done in house with the use of Rhinocerous 3D
and additive manufacturing respectively. Specifically, Rhino 6 was used to
design the panels and Formlabs Form 2 SLA printers were used in the
fabrication of the mold pieces. The mold pieces are made of multiple different
resins that contain different mechanical properties. While the fabrication of
reusable molds with additive manufacturing is a much less labor-intensive
practice than the creation of traditional formworks, there were still hurdles that
1265
SIGraDi 2021 | Designing Possibilities | Ubiquitous Conference
had to be cleared to successfully fabricate the molds. The problems
encountered are unique to additive manufacturing, including the maneuvering
of tolerances between pieces that must fit together and working with the
material constraints of our chosen resins.
Both panels involved open molding techniques and CRM, but the form of
Mold A acts more as a self-supporting formwork while Mold B acts as more of
a traditional form liner. There was minimal labor involved in the fabrication of
the molds through AM, which allowed for multiple mold pieces to be printed with
minimal fabrication turnaround time. This meant that our mold design process
turned into a cyclical and iterative process. There were failures in the first
iterations of the molds, but we were able to quickly find them through casting
trials and then go back and make changes to the design. This aspect of
immediate feedback was extremely beneficial in developing the molding system
for each project, and is only possible through AM and other forms of digital
fabrication.
When printing a piece, 3D printers are usually unable to print the prescribed
geometry of a CAD model to the exact dimensions present in the model. This
is due to the inherent limitations of the additive printing process. For example,
if a printer can only produce layers that are 1mm in depth, then the nominal
dimensions of the printed object will be +/- 1mm from the CAD model. The
ability of a printer to produce very thin layers and thus a more geometrically
accurate print is called its resolution. Form 2 printers do maintain an extremely
high layer resolution, with the lowest layer size being 50 microns and the max
being 100 microns. The extremely high resolution of the printers did not
eliminate the need to work around the tolerances of the prints. In particular, the
fitting of clips for mold A proved to be a process that required multiple iterations
before finding a tight fit.
The material constraints of the resins we used were discovered largely
through print failures. Both the Elastic 50A Resin and Clear Resin developed
by Formlabs were used. The mechanical differences between the two are
primarily in their elasticity as the elastic resin has a much higher flexibility than
the clear resin, which is solid and brittle (Formlabs, n.d.). However, the
mechanical properties of the elastic resin made it much harder to print with.
Oftentimes when using the elastic resin the 3D printer fails to cure a certain
layer or section of resin enough, which disallows the next layer from properly
forming and eventually causes the entire print to fail.
Although there were some challenges during the fabrication process, both
mold designs are able to produce a number of panel variants from a singular
mold. To perfect the casting process, however, more mold iterations are likely
needed.
4. Conclusion
The lack of widespread variability in facades today can be partially attributed to
the fact that cheaply and efficiently fabricating a large number of façade panel
1266
SIGraDi 2021 | Designing Possibilities | Ubiquitous Conference
variants is difficult. After exploring the limitations of both modern precasting
methods and those of digitally fabricated concrete, neither alone can meet the
levels of varied output desired by contemporary designers. However, when
paired together they can meet design and fabrication demands of both
architects and precasters. The use of mold inserts inspired by observed
traditional processes was paired with digital design and fabrication methods to
create volumetric precast panels based on built contemporary designs. These
newly developed variable mold inserts produced multiple panel variants from a
singular mold. Further, BAAM practices for mold inserts in modern precasting
methods reduce the cost per pour of producing architectural scale panels. If the
proposed methods are paired with BAAM, the ability to economically produce a
wide variety of volumetric façade panels becomes possible. However, the limits
of this method are not yet fully understood. In the future, this study will further
consider the principals of designing panels that offer variable assembly
potentials. Other fields of study such as mathematics, specifically aperiodic
tiling patterns, will be sources of knowledge for expanding the work.
Acknowledgements. Thank you to both LSU Discover and the National
Science Foundation’s program, SMART Polymer Research Experience for
Undergraduates (CHE-2051050) for providing financial support to this study.
Further, thank you to the LSU School of Architecture for providing the facilities
and resources to fabricate the molds and panels. Finally, thank you to the
faculty and student body of the LSU School of Architecture for providing
encouragement, support, and guidance throughout the entire study.
References
American Institute of Architects. (2014). AIA Tap Awards 2014 [PowerPoint slides].
https://network.aia.org/HigherLogic/System/DownloadDocumentFile.ashx?Do
cumentFileKey=3b78302a-cf30-4e4f-bd2c-f9eddd49d902&forceDialog=0
Asprone, D., Menna, C., Bos, F. P., Salet, T. A. M., Mata-Falcón, J., & Kaufmann, W.
(2018). Rethinking reinforcement for digital fabrication with concrete. Cement
and Concrete Research, 112, 111–121.
https://doi.org/https://doi.org/10.1016/j.cemconres.2018.05.020
Cutrer, J. (2017). Perot Museum Exterior Architecture [Photograph]. Flickr.
https://www.flickr.com/photos/joncutrer/37585013761
Formlabs. (n.d.). Compare Formlabs SLA 3D Printers Tech Specs. Retrieved July 7,
2021, from https://formlabs.com/3d-printers/form-3/tech-specs/#data-sheets
Jipa, A., Bernhard, M., Dillenburger, B., Ruffray, N., Wangler, T., & Flatt, R. J. (2017).
skelETHon Formwork: 3D Printed Plastic Formwork for Load-Bearing Concrete
Structures. Proceedings of the 21st Congreso Internacional de La Sociedad
Iberoamericana de Gráfica Digital, 3(12), 345–352.
https://doi.org/10.3929/ETHZ-B-000237357
Jipa, A., Giacomarra, F., Giesecke, R., Chousou, G., Pacher, M., Dillenburger, B.,
Lomaglio, M., & Leschok, M. (2019). 3D-Printed Formwork for Bespoke
1267
SIGraDi 2021 | Designing Possibilities | Ubiquitous Conference
Concrete Stairs: From Computational Design to Digital Fabrication.
Proceedings of the ACM Symposium on Computational Fabrication.
https://doi.org/10.1145/3328939.3329003
Lloret-Fritschi, E., Reiter, L., Wangler, T., Gramazio, F., Kohler, M., and Flatt, R.J. 2017.
“Smart Dynamic Casting. Slipforming with Flexible Formwork - Inline
Measurement and Control.” In HPC/CIC Tromsø 2017, Paper no. 27. Oslo:
Norwegian Concrete Association. https://doi.org/10.3929/ethz-b-000219663.
Li, F., She, L., & Wang, H. (2020). Standardization and Diversification of Precast
Concrete Facade Panel. IOP Conference Series: Earth and Environmental
Science, 455, 012108. https://doi.org/10.1088/1755-1315/455/1/012108
Love, L., Post, B., Roschli, A., Chesser, P., & Hun, D. E. (2019). Feasibility of Using
BAAM for Mold Inserts for the Precast Concrete Industry.
https://doi.org/10.2172/1606893
Martins, P., Fonseca de Campos, P., Nunes, S., & Sousa, J. P. (2019). The Tectonics
of Digitally Fabricated Concrete. A Case for Robotic Hot Wire Cutting. In RILEM
Bookseries (pp. 311–322). https://doi.org/10.1007/978-3-319-99519-9_29
Naboni, R., & Breseghello, L. (2018). Fused Deposition Modelling Formworks for
Complex Concrete Constructions. SIGraDi 2018, XXII Congress of the
Iberoamerican Society of Digital Graphics, 166–173.
https://doi.org/10.5151/sigradi2018-1648
Nahmens, A., & Mullens, M. (2011). Lean Homebuilding: Lessons Learned from a
Precast Concrete Panelizer. Journal of Architectural Engineering, 17(4), 155–
161. https://doi.org/10.1061/(ASCE)AE.1943-5568.0000037
Pan, M., & Pan, W. (2019). Determinants of Adoption of Robotics in Precast Concrete
Production for Buildings. Journal of Management in Engineering, 35(5),
5019007. https://doi.org/10.1061/(ASCE)ME.1943-5479.0000706
Portland Cement Association. (n.d.). Ultra-High Performance Concrete. Retrieved
August 25, 2020, from https://www.cement.org/learn/concrete-
technology/concrete-design-production/ultra-high-performance-concrete
Rockite Cement. (2017). Specifications: Rockite Cement. [Data set]. from
https://www.sculpt.com/2017/Tech/MsdsTechSheets/TechSheets/TECH_SHE
ET_rockite.pdf
Stephens, S. (2013). Sheared and shirred, surfaces and solids: Thom Mayne and his
firm, Morphosis, turn a corner for the Perot Museum of Nature and Science in
Dallas. Architectural Record, 201(1), 78–85.
Sutherland, J. (2001). Introduction. In J. Sutherland, D. Humm, & M. Chrimes (Eds.),
Historic Concrete: The Background to Appraisal (pp. 1–10). Thomas Telford.
Techni Moulage. (n.d.). Production: Verona Private Sale. Retrieved July 7, 2021, from
https://www.techni-moulage.com/realisation/10#realisations-5
Wilmotte & Associés. (n.d.). vente-privee - Tour Le VéroneProject . Retrieved July 7,
2021, from http://www.wilmotte.com/en/project/440/vente-privee-tour-le-
verone
1268
SIGraDi 2021 | Designing Possibilities | Ubiquitous Conference
