network partners:

Digital Crafting 3 Workshop Report Now Online

Read the Workshop Report and Interview with Workshop leader Asbjørn Sondergaard.

Digital Crafting 3 “CNC and Concrete: How to Mould”

The third Digital Crafting workshop “CNC and Concrete: How to Mould” was held at the Danish Institute for Technology in Copenhagen August 18-19 2010.  Asbjørn Søndergaard from Aarhus School of Architecture led the two-day workshop which allowed participants to work in small groups to create digitally fabricated foam formwork for a concrete prototype.  Discussions and lectures during the workshop related to the question of how designers can use morphogenetic tools to allow for structural optimization and how designers can design material logic and performance into digital models.

During the workshop, the first day was spent largely working in groups in the classroom, creating the digital files for the forms and the fabrication data.  At the end of the day this digital information was sent to the robot, and overnight the formwork was milled.

The design of the three prototype panels was carried out in groups. Each group designed one panel and then the panels came together to make a larger design, inspired by the game “exquisite corpse”.  The overall design was that the three panels then could lean against each other and the load paths were designed for this configuration.

The pre-optimized, overall geometry of these was given in Rhino at the start of the workshop.  These rectangular forms were taken into the topology optimization software where the load conditions were specified.  The software gave results for the optimized forms, and these forms were then brought back into Rhino.  This form was adjusted for manufacturing requirements, minimum thickness of concrete etc, using T-Splines, a plug in for Rhino.  These adjustments allowed the further optimization using the criteria of manufacturing constraints. T-Splines was used to create the final forms and those were then exported to the CNC milling software where the tool paths for the robot were created.  In groups, participants investigated different kinds of tool paths, which leave the mark of the robot´s tool in the form.   The CNC software then exports the tool path to the robot and the robot cuts the form.

The next day, when the moulds were cut. We developed three different strategies for milling the elements. Differentiating the milling time we created three levels of detail from the very fine to the very rough. This differentiation was a practical decision as well as a probing one. As the cutting time in the short workshop was limited we varied the cutting time from 8 hours to 4 and finally 2. The different levels of detail gave very different aesthetic expressions.

After milling the reinforcing was set into the moulds, release agent added and the concrete poured and leveled.  In the morning of the Seminar day, the concrete forms were carefully broken out of the milled moulds and cleaned off.

Report by Terri Peters

Performance, Scale and the Beauty of Optimization – Interview with Ole Sigmund

Ole Sigmund is a Professor at the Department of Mechanical Engineering, Section for Solid Mechanics, Technical University of Denmark.   He researches the design of extreme materials, smart materials, compliant mechanisms, MicroElectroMechanical Systems, crashworthiness, fluid systems and wave-propagation problems in acoustics, elasticity, nano-optics, meta-materials and antennas.

His lecture was about the applications of topology optimization, and about how materials can be designed to be better performing, solve complex design problems and be beautiful.   Following the lecture the group discussed optimization in relation to material and material design.  The group discussed some ideas about gradient rather than absolute material performance, and materials with potential for form change.

–Could you elaborate a bit on the design potentials of designing for optimization and the challenge of scale (for example nano-scale to building scale in material design) as it applies to your research?

OS: Of course there is a scaling issue since currently our nano-material based structures are very small. However, with improved manufacturing methods this will come soon. For example, one of our current research projects is concerned with surface structuring of plastic parts or bottles with the goal of saving costly painting. We do it using mass-manufacturable, nano-imprint technology we can stamp nano-structures into plastic surfaces, in turn changing their colors or making the hydrophobic.

–This relates to the toy example you described in your lecture – the Harry Potter figurine by Lego?  You mentioned that the most expensive part of the process is not the creation of the ten parts which are assembled to create this figure but actually the way that the glasses and face need to be painted on the figure´s head, separate from the injection moulding process? In your lecture you proposed a new way of creating this, using actual material deformation rather than pigment.

OS: Yes we are using the ideas from butterflies – where if you take a microscope and look at the surface they have nano-structured surfaces that actually create the colours that we see.  The butterflies actually designed their surfaces in order to absorb light at different frequencies.   In this project our job is to make the nano-structuing of the Lego surface so that we don’t have to use paint it just becomes an absorbing surface that makes it look like we painted it.  With regard to scale, of course from smaller plastic parts to building parts there is still quite a step. However, I am convinced that this gap will be bridged within the coming decade.

–There was some discussion about manufacturing constraints and how this must be built into topology optimization design, how do you design with these limitations?

OS: Different manufacturing methods have different limitations. For example in concrete casting it is not directly possible to introduce internal voids. In many cases void regions inside the structure would be structurally good and hence the results of the optimization would contain holes. Therefore we introduce constraint in the optimization that hinder the creation of internal holes (at the cost of worse performance/weight) of the structure.

–This led to a discussion about the aesthetics of optimization – how can and should something “look” optimized or perhaps look “not” optimized?  You said “Whenever I see a structure with circular holes I know it has not been optimized”. What do you think about the relationship between aesthetics and optimization?

OS: I certainly think that an optimized structure is beautiful. However, due to my training I see many flaws in “optimized structures” that ordinary people would not see. Hence, a structure with many circular holes may look light and efficient for many people, however, in my eyes I see stress concentrations and waste of material. Also if I see a curved bar that is supposed to support longitudinal forces I know that the structure is not optimal. Unfortunately one of the workshop structures has such features –this I partly attribute to bad post-processing steps in the used software.   A good example of this faulty optimization is the CCTV tower in China. The outer structure is claimed to distribute the forces in an optimal way, however, to me it is clear that it is by no means optimal and that a much better (and possibly even better looking) could have been obtained using topology optimization. Unfortunately I never found time to test it but I will try to find some students who can perform the optimization study.

–What did you think were the most interesting aspects from this Digital Crafting Seminar?

OS: It was interesting to see the broad range of speaker topics –from my very basic engineering structures that fulfill well-defined optimization goal to the very artificial and complex structures produced by various digital processes. For me, structural beauty is a natural bi-product of the structural optimization process. I hope that this message will be remembered by the participants. The artificial digital processes are also very interesting but I think they should be hooked up with some measures of efficiency to become well accepted in a world that becomes increasingly aware of limited natural resources.

Report and Interview by Terri Peters

Interview with Asbjørn Søndergaard, Workshop Leader “How to Mould”

During the workshop, participants visited the topology optimized concrete structure produced during the Unikabeton research project by researchers Per Dombernowsky and Asbjørn Søndergaard. Fabricated using large-scale industrial CNC-milling facility at Danish Institute of Technology, the structure represents the first realized topology optimized concrete structure. A full-scale version of the optimization experiments undertaken at the Digital Crafting Workshop 3, the prototype reflects the morphogenetic principles of design and conceptualization facilitated by the method of topology optimization.  For more details of the research project see

–As workshop leader, what were your intentions with “How to Mould”?

AS: The idea was to offer a platform for a 1:1 experience with the morphogenesis of topology optimization in relation to robotic fabrication, in order to facilitate a discussion of the implications of the field in regard to related theoretical discourses and technological aspects of production.  In the workshop, we explored the cycle of optimization, remodeling, full-scale milling and casting all within 3 days.  I think the most successful part was delivered by the participants in their dedication to the workshop content, and the discussion that arose from it.

–In your two introductory lectures during the workshop, and in the Seminar on the final day, there were discussions about topology optimization and the aesthetics of optimization.  What parts of this discussion do you think are the most relevant for designers?

AS: It is often the case that a type of formal language emerges as the result of intensive work of experimentation and reflection – and then this language is adopted by others that take interest in the appearance, but not in the process behind it. The language then becomes a self-referring, self-explanatory image of the original thought, but without its coherence.

This sometimes happens with so called “optimized” structures. Actually, the first large-scale topology optimized structure to be realized – the Qatar convention hall by Mutsuro Sasaki – is a good example of this. The structure was originally conceived by a process of optimization, but then simplified into an internal steel rod skeleton clad with non-load-bearing steel plate that imitates the original optimization output.  This means that it lost most of it initial structural logic in the process of realization, although it still formally appears “optimized”.

–How can we interpret the optimized results? Can something be really “optimized”?

AS: We discussed two opposite positions on this: the first position is to say the most interesting results arises directly from computation, without the designer interfering or polluting with his formal preferences. The other position is that the most interesting results arise as a work of the interpretation of the designer, and that the computational process should only secondarily contribute to the appearance of the design.  In my opinion, the first position does not take into account that the premise for any computational process is manmade, and so subject to inter-subjective conventions. No matter how strictly mathematically the process may be, there will always be a modeling setup preceding it, in which several design threads can be pursued, tried and discussed. Also, the computational results need subsequent interpretation in preparing the shape for production and manufacturing – and this is also an area of aesthetic evaluation.

The optimization result within architecture is something derived by both structural and aesthetical consideration.  To misunderstand this is to repeat the modernist attempt to avoid the difficult but necessary question of aesthetics by claiming a false objectivity to the process. I think the question is rather: how do we affect the process of optimization prior to its execution?  And how do we choose to interpret the optimization results formally? I believe many answers can be developed to these questions, varying on both cultural and technological conditions.  The optimization process may actually result in unexpected design discourses that could influence the spatial concept.

Report and Interview by Terri Peters

Workshop 3: Concrete and CNC

Workshop: How to mould

Autumn Semester 2010: August 18 – 19. 2010

The workshop investigates how digital fabrication can lead to new principles of construction for their realisation. The workshop investigates ways of thinking the relationship between the joint and the cast, the mould and the form, the mono-material and the composite.

The workshop guest is Asbjørn Sondergard.

Digital fabrication is changing the making of architecture. Through mass customisation the manufacture of bespoke elements is becoming economically viable allowing for the development of more complex building solutions. But digital fabrication also leads to new structural opportunities. As the direct interfacing between design and production facilitates a higher degree of detail by which the material is addressed this new technological platform is questioning our tectonic traditions and fundamentally reforming the material practices of architecture.

In working with the wet processes of casting and printing digital fabrication has provided new ways of thinking fabrication. Here, a single fabrication logic can produce vastly diverse geometries. On the one hand the interfacing of digital design with CNC milling has allowed a new generation of highly complex individualised moulds to be produced directly from the CAD design package. This facilitates the development of bespoke formwork at a highly reduced costs and material intensity. On the other hand the development of rapid-manufacturing technologies suggests a new practice where materials are directly specified and detailed as part of the design process. In these additive processes the need for formwork creating a radically new position for in which the architect becomes part of a material design practice.

In this workshop we will examine the guiding research questions for understanding a new generation of digitally designed materials. We will explore the making of bespoke formwork the casting of structural elements and question how the increased level of detail and complexity can lead to new structural forms. We will look at how a new generation of morphogenetic tools can allow for structural optimisation and discuss how this material logic can be encoded into digital models. Finally we will discuss and frame the understanding of how full scale printing technologies can become part of our building practice.

Workshop: the role of the mould

The workshop pursues two investigations into a novel understanding of the potential of robotics and concrete:

The structural surface: morphogenetic processes and topology optimisation
The workshop is structured around a set of hands-on experiments working with topology optimisation and the industrial milling of large scale foam based moulds

Topology optimisation is a digital design tools that allow for a material simulation by which the structural form of materials can be optimised for minimum weight and maximum stiffness. The design process uses generative processes by which the given design constraints of structural load, span and activation are calculated. Where topology optimisation is increasingly common in mechanical engineering and aeronautics it remains rare in architecture as the produced formwork is highly complex and often results in bio-morphic geometries.

In this workshop we will interface the structural logics of the topologically optimised with the fabrication of structural surface in concrete. Making use of the Technological Institutes industrial scale CNC robot we will investigate how to inform the making of complex formwork, how to explore their tectonic potentials and how to detail their joinery.

The workshop asks:
– How can a tectonic principal be encoded in a digital model?
– How can we understand the new restraints by which CNC milled formwork can be designed?
– What is the role of detail and surface in this new material logic?

The mark of the tool:
As a subtractive process, CNC milling carves away material to reveal the intended shapes. In this mode of production, the milling of shape is confronted with the logics of creating cutting tracks, the strategies for incremental removal of material. Though theoretically any shape can be milled, significant decrease in production time can be gained by choosing rough cutting tracks, leaving traces of production in the surface.

Schedule and detailed program: 03_WorkshopSeminar03_concrete_sendout.pdf

Rhino template files for the workshop:

Indesign template for participants portfolios:

Venue: The workshop takes place at the Danish Technological Institute – Gregersensvej indgang 3 – 2630 Taastrup –  Google maps link here

The way to get there by public transport is to take the train from the central train station IC 137 fra København H towards  Århus H and get of at Høje Taastrup st. From there you can take a bus 400S (toward Ballerup), 154E or 216. The cabs at the station are very scarce. You can also wolk – it is a bit more than a 10min walk. You can use : to find the right directions from your departure place.

When you are at the Danish Technological Institute you should find entrance 3 and from there room 35 (workshop days) or room 36 (seminar).