Everyone knows each generation of students supercedes the last with their technical knowledge. James Roberts, recently graduated from Liverpool University and now a much appreciated colleague of mine, sheds light on his experience of using Generative Components in both academia and working practice.
During my final year of studying on the BA course at the University of Liverpool School of Architecture, I was introduced to the dark world of parametrics and Generative Components by Prof. Andre Brown, Mike Knight and Martin Winchester. Knowing nothing about it, I started from scratch, re-writing the GC tutorials for a piece of coursework, with the aim of making them more accessible to complete beginners, and the following paragraphs are taken from that.
“Parametric design revolves around a basic knowledge of geometry and
relationships between entities. These properties can then be altered, giving different outcomes of form. For example a line has two parameters – its length and its direction, altering one of these factors gives you a different form. A polyline has the previous two factors plus the positioning of its vertices and if any of these are altered a different form is given, and so on.
Expanding on what was previously mentioned; parametric design is the use of geometry and parameters which define form via a series of relationships. These parameters can be defined from a vast list of possibilities; they could be taken from data on wind speed or rainfall for example. The parameters basically affect something somewhere along the line. Within Generative Components, such relationships can be explored without the need to re-draw, which provides freedom and flexibility within the design process. It also allows some startling geometric results, as well as the ability to connect the model to an Excel spreadsheet. This sounds banal but the program developers give the example where someone involved with the financial side of the project can see how the cost of the project relates to the form. For example a tower that has a vertical rotation of floor plates can be seen in terms of cost and how a very twisted form costs more than a not so twisted form – then the client can weigh up cost verses form which eventually gives the company an image."
Parametric design misunderstood
Nine months in practice and eighteen months since I was first drawn in by the graphics of Tom Wiscombe, Hernan Diaz Alonso and Achim Menges to name a few – I still make no claim of being anything other than a novice. In fact, in my nine months of practice I have used GC once on a project based task, which is half the problem. GC and parametrics is just not ‘understood’ or accessed by the masses yet; many people see what it can do and the glossy images of Zaha show that, but for most practices, someone in the office can do what is needed in another program without the cost or time for training, albeit less efficiently. I use it whilst I eat my breakfast at my desk and that’s about it. But that’s not to say it’s a lost cause – it is far from that. The future is almost undoubtedly going to head that way, university’s are getting big on it and eventually it will become a standard skill for a graduate, as using a CAD drawing package is now.
It may seem like I am saying the next generation of architect’s will ensure our streets are filled with blobby buildings, but fear not, parametrics have been used in other industries for years, for example in the naval industry, yet boat hull’s still look the same to me as they always have. Yes, mind boggling geometry can be achieved (I shall come onto this in a second), but a big advantage of parametrics is it can be incredibly efficient on time and work load. If you alter one element and everything will alter to fit the change, and when I say everything, that includes the drawings. The key to using something like GC is understanding what you want to achieve and not expecting immediate results. You may want a cube, but it may take a certain amount of instructions to get it.
Inherent complexities
The problem, or rather, difficulty, with parametrics in architecture and buildable buildings is the whole model. Buildings have floors which have voids which house stair cases which have handrails and they also have walls which have openings for windows. This level of factors makes the modeling very complex.
Going back to the geometry statement, the design in terms of the form will come down to what route you take. I recently went to a talk at the RCA with Patrik Schumacher (Zaha’s number two) and Lars Spuybroek (NOX) where Schumacher explained his view on the matter which I translated as..
Architecture vs Engineering
Style vs Technique
Form generating vs Form solving
Zaha’s Funicular in Innsbruck vs Foster’s British Museum in London
No repetition vs repetition
I’m not saying any of these ‘dogmas’ are correct or incorrect, or one way is better than the other, but I think it illustrates the vast possibility, approach etc that is possible through parametrics. It can even be adopted by the (rightly) eco-friendly designers; the AA has some great work coming out of Achim Menges and Michael Hensel’s unit based on morpho-ecology; which is a whole conversation on it’s own and admittedly nothing has been built based on that theory yet.
Theory
And theory is something parametrics is very rich in. The AD has had various issues written on it and is certainly a good place to start. And if you do so it should be read with the fact that although in its ‘early days’, parametric design has been in used by architects in projects, some of which very successfully. Some of these projects would include Grimshaw’s Waterloo International Rail Terminal, Foster’s Swiss Re, London City Hall, The Sage and Beijing Airport projects, Greg Lynn’s New York Presbyterian Church, KPF’s proposed Bishopsgate Tower and endless Zaha Hadid projects. Coincidentally these are all big firms. This is not to say small firms are not using it, in my mind, some the more interesting work is coming out of the small, young offices which are well versed in parametrics and scripting; but at the moment, parametrics generally means big budgets which means big clients who want big names.
Getting back to GC, the best place to start is the tutorials that existed under the help tab in the versions up to v8i. These gave a pretty good understanding of how things worked. If you have v8i or have completed the GC tutorials and want a further task, then please download the tutorial that will be attached to the next blog for a bridge similar to the one designer by Arup and Hodder Associates in Manchester, based on a paraboloid. (see http://www.arup.com/facadeengineering/project.cfm?pageid=1810). This tutorial is the main one I wrote for my coursework whilst at the University of Liverpool Architecture School and has been updated and amended for the readers of this blog . This task will cover most things the previous tutorials covered, but in a more ‘realistic’ set up. The .gct file can also be downloaded, and by turning on the transactions systematically, the build up can be followed. Watch this space for the follow up blog.
Should you enjoy this blog please feel free to download my tutorial in the next post. I will be more than happy to take questions on this or the GC tasks and do my best to answer them.
