Complex Geometry in Architecture

Rephrasing a project title as a strategic decision

Having started my PhD only a few months back, it feels as if I am really at the beginning of my research. This is probably a common feeling as well as a correct statement and it led to the question whether I made any research relevant decision I could write about so far or if I was just collecting information and trying to understand the process and the project my work will be embedded in. The answer is that I permanently made decisions influencing my “research path” by choosing which articles to read, which people to talk to (or not) and by redefining the main objective of my project.

The most interesting, influencing and challenging decision though, was to adjust the original title of my PhD from “Komplex Geometry – B-processor” into “Complex geometry in architecture based on building information modeling”. And it is almost amusing how simply rephrasing the title, freed my mind and turned a – without doubt challenging – task into a research project. It might also serve as an example of how I try to make judgment on questions and take decisions for my PhD.

B-processor is – simply put – CAD software. It is also a research project of the Arhus school of architecture in cooperation with Alexandra Institute, investigating new ways of integrating building information modeling (BIM) in an architectural design workflow. B-processor was developed from scratch as all existing packages seemed to be too restrictive. This led of course to a rather slow development process not only because any object or tool had to be programmed but also designed, discussed and reviewed. Today B-processor works as a very simple and intuitive tool but it is missing elementary functionality to be tested in real life. This is the lack of any curved element: from arcs to freeform surfaces.

Before being able to work on any complex geometry a lot of ground work will need to be done by implementing an extended object structure in B-processor, allowing curved surfaces of any kind. This will affect existing tools and objects because all of them need to interact (to be trimmed, intersected, copied, etc.) The risk is to get lost in a task that cannot be defined as research, is time consuming and could better be done by a real programmer than a researching architect with programming skills. At the same time, it offers the opportunity to develop B-processor in a way and direction that allows complex geometry to be created in a new way. Looking for these new workflows and questioning how curved surfaces are created in software so far can therefore become research. This thin line between a programmer’s homework and a research project will constantly need to be checked.

The decision I made was to widen the field on the one hand side and narrowing it down on the other. Instead of researching complex geometry in B-processor I first decided to research complex geometry in building information modeling in general. This will allow me to look at B-processor as one specific BIM tool with certain advantages and disadvantages. It might fail to use B-processor in the intended way and I could still research using other software.  It also means that the basic work in B-processor can be done in parallel to a general research that is not necessarily depended on the outcome of the task of implementing curves and still creates valid research outcome.
There are still many reasons for not abandoning B-processor completely from my research project. I see it as a prepared ground that is ready for almost any seed to be planted. It is a relatively unknown project that I could form in the way I can make the most use of it.

Secondly I narrowed down complex geometry in general to “complex geometry in architecture”. It might seem like a minor change but it helps me to exclude all the freeform modeling fields that rather deal with visualization and animation and is important because so far, architects tend to use tools that originally were created for other purpose (movie, car and airplane industries). I however would like to focus on architecture-specific tools and workflows.

In combination, the two phrases also exclude the formal questions why complex geometry is used in architecture at all. My research starts at the point when a decision for double curvature is made and concentrates on ideal workflows from design to production instead of questioning the reasons. I don’t think that looking at the tectonic derivation of a specific architectonic shape isn’t worth a research project it is rather a project on its own. Again a decision was made towards a more technical approach, excluding (or better minimizing) the formal background. The hypothesis that an intuitive and optimized workflow to create this kind of geometry is independent from the formal reasons might though need to be looked at.

Last but not least I replaced the simple hyphen in the title with “based on”. Even though I am not yet fully satisfied with this, it is a step in the right direction as it starts defining how the two elements, “complex geometry” and “B-processor” or “building information modeling” should interact. The question I would like to research is how the use of building information modeling changes and influences the workflow of creating complex geometry.

The essence of the above process is that I made quite a strategic judgment. The decision opens up a wide range of possibilities and new research directions while it narrows it down on the other side to limit options and not getting lost in a jungle of too many potential directions. I guess it will often need this kind of decision to balance the specific and the general. And it will be exciting to see how quickly a chosen path can be evaluated.
It seems to be a process of trying to steer the right way while reaching a dead end is not a failure of research rather than an exclusion of one possible solution. It is also a strategy of breaking the problem into solvable parts, a method of temporarily narrowing the range of directions and to focus on a more specific problem before looking at the big picture again.
I also decided to work in parallel on different questions as it can be quite frustrating, being stuck with a certain problem. A solution often can’t be forced. Switching track and working on something different is a trick to re-motivate myself and can potentially untie a “mental knot”.

I envisage to theoretically investigate on complex geometry and building information modeling and to sketch potential ways of how the two can be joined to create new workflows.
The results will be evaluated and if found to be promising translated into case studies in either B-processor or another software. This will allow testing as a form of evaluation. The results of the case studies will hopefully drive the next level of theoretical decisions.