Yesterday, I interviewed Salomé Galjaard, the leader of a Dutch team from the global engineering company Arup that has been researching methods of 3D printing structural steel. She and her team have made huge breakthroughs in this research and the process they have developed could soon become a staple of engineering. I have previously written about how 3D printing could make waves in the worlds of architecture and engineering, so as soon as I learned that methods were being developed to print stainless steel rather than the less architecturally versatile plastic (as used in conventional 3D printers), I contacted Arup to learn more. Ms. Galjaard was kind enough to answer my questions about the process and its development, uses, and future. Key parts of the interview are transcribed below.
First, some context: the process came into development to solve problems that arose in the design of lighting structures in The Hague’s Grote Marktstraat. These lighting systems were designed in the form of “tensegrity structures”- compressed beams and a tensile net in cables interacting to form a rigid frame. The irregular design meant that each of the 1600 “nodes” connecting beam to cables was different. 3D printing these nodes, Galjaard’s team theorized, would be more effective than conventionally manufacturing them. Plastic nodes would have been generally weak: brittle, less heat-resistant, and short-lived, so the team wanted to use steel, which is stronger and has a better aesthetic. As could be expected, steel printing has its challenges, but the Arup team has made huge breakthroughs.
One of the 3D printed nodes
Me: How long have you been working on this project?
Salomé Galjaard:It started off as a relatively small research project on the side, so it was sponsored by Arup but not part of my full time job. We started looking at this two and a half years ago. We put the first results online about a year ago and are about to put more results online. It’s picking up pace, so we’re spending more and more time on this… it’s getting bigger and bigger, and we’re seriously considering now to develop it into a business, something we can offer our clients…. What I’m also hoping for is that architects, for example, will hear about this technique and possibly be inspired by it, and maybe even make different designs that are especially based on additive manufacturing.
So you think this process will become very widespread once you refine it?
Yes, but I don’t think it will be used on every project… For example, if a project could really benefit from weight reduction, then this could be really beneficial, it could be beneficial if you take a building apart and build it somewhere else. But maybe more standard structures might not need this at all, and then it’s too costly to implement it.
When you were developing the printer, what was the biggest challenge you faced?
The biggest hurdle was actually our minds! We were so stuck in traditional thinking that it was and still is difficult to let go of those traditional approaches and solutions. We are so used to working with sheets and struts and cables and combining them in certain ways. Free form is not a part of this methodology.
At this point, she showed me the two nodes designed for the tensegrity structure. She showed me the first, pictures of which were released along with the first information about the research, and a new optimized version, pictures of which have not yet been released. Whereas in the first node cables connected to the joint with a fork-and-pin component, they connect directly onto the new node, lightening the structure by forty percent.
How are you designing the objects that you print? What is the design process like?
We just started off with pen and paper… Every time we thought we had a solution, someone said “Oh no! We can’t produce this” or there was too much traditional thinking. We used Grasshopper to check the boundary conditions (of each design)… and at some point we used another software for optimization… but if you give the computer too much freedom you’ll end up with something you can’t use. Unfortunately, we don’t have the software now that allows us to design an optimization that’s really fit for production with 3D printing. You can come up with a blob that works perfectly from a structural point of view, but that doesn’t mean it’s easy to produce (with a printer).
Previous statements by Arup have emphasized how the size of the objects that could be printed was limited- you won’t be seeing full-size steel buildings being printed anytime soon. I asked how detailed the printing could be and was shown a cylindrical test model embedded with openings perhaps a millimeter or two in diameter.
Would you say the main benefit of this process is to create fine detail that allows lightness, et cetera, rather than large freeform objects?
It depends- big is expensive, but we work on buildings that cost billions, so the cost of one element might not be that extreme. It just has to add value; I think printing for the sake of printing is not such a good idea… Obviously, size is a problem, but what is nice about these printers is, because you have so much design freedom, that if you want to make one really big object, you can cut it into sections and design the interfaces and connections so everything clicks together. There are limitations, but I think that right now, if you’re a smart designer, you can work around them.
How soon could you be commercially using the printers?
One limitation at the moment is the certification… It might be that the first couple of times we do this, we will have to test every single product that we put in to make sure it’s reliable.
In other words, if the tensegrity structure was to be built, all 1600 individual nodes would have to pass tests.
Are there any other materials you are planning on 3D printing?
We are not at the moment looking at different alloys (other than stainless steel) but I wouldn’t be surprised if we do in the future. We are looking at completely different materials like plastic or concrete. No research projects yet, but we have some good ideas and are involved with plastic 3D printing projects with external partners.
Outside of other materials, how are you thinking of expanding on this research?
I have a lot of different ideas, but even if we keep the project fairly small there is a lot to do. We would like to talk to software developers… to create better support tools. As I said, I have ideas about printing in concrete, to create building envelopes or shells directly. This would be based on additive manufacturing, but require much larger machines than we use with the steel.
There you have it: engineers are on the brink of having the ability to build freeform with steel in a way that would have seemed absolutely inconceivable a couple decades ago. It’s an example of how not only 3D printing but also computers in general have revolutionized architecture; it’s impossible to say what the next breakthrough will be.
A view of the detail of the first node
The alternative: a manually manufactured node
Credit for all images goes to Arup
Thanks again to Ms. Galjaard for the interview and permission to publish it