Generatively designed space exploration lander created with NASA’s Jet Propulsion Laboratory.
When bionic brake calipers and 3D printed high fashion have become your norm, it’s always good to go to a non-additive event every once in a while to get a true sense of where the technology fits into the wider manufacturing world. At Autodesk University (AU) London earlier this week, the technology was ever present throughout talks and exhibition space but in more surprising – and often subtle – ways.
As in previous years, generative design was a big focal point of the software leader’s annual London event. The “poster child for Autodesk Research” according to Erin Bradner, Director of Autodesk Robotics Lab, during a roundtable session where she spoke about agile robots being a key focus for her team which looks at technologies that are five to ten years out. Generative is a prime example of an Autodesk Research project that successfully went from idea to whitepaper to product and has been leveraged by companies like General Motors and Lightning Motorcycles.
Since generative’s introduction, the majority of case studies (and marketing materials) have fuelled the belief that these strange, organically formed parts could only be produced via additive manufacturing. While AM lends itself quite nicely to generative’s natural structures, we now know there are many ways to make a generative part.
A prime example of this is Claudius Peters, a manufacturer of industrial machinery for cement, steel, gypsum and aluminium industries. The company used generative design to optimise a piece of concrete cooling equipment in an effort to reduce weight and save costs. Working with Autodesk, the company ran simulations and initially 3D printed the new “alien part” – as Thomas Nagel, Chief Digital Officer at Claudius Peters called it – in metal but the process proved too expensive. So, they reverse engineered the additive part and scaled it back to a design that could still be produced through traditional methods. This resulted in a 25% reduction in material and improved the overall structural integrity of the cooler.
Design iterations for Claudius Peters' optimised concrete cooling equipment.
This example came during AU’s opening keynote where just moments before, the idea of metal 3D printing a functional space lander vehicle was referred to as still “science fiction” according to Autodesk’s VP of Cloud Platform Sam Ramji.
At this point, you might be thinking this was a bad day for additive. Actually, I’d argue, it was quite the opposite.
Taking centre stage in AU’s exhibition area was a space exploration lander created with NASA’s Jet Propulsion Laboratory. It’s a huge alien-looking structure designed to travel millions of miles and carry scientific tools into space - the more mass you save on the vehicle, the more instruments you can carry on board. The legs and internal structure were created using Fusion 360 and Netfabb and achieved a mass reduction of 35% compared to a conventional design. The lander’s legs were optimised to be milled while the larger chassis which houses the payload boxes was created using 3D printing, not for the final parts, but for huge moulds for casting. “Generative design is now practical,” Ramji concluded.
Get your FREE print subscription to TCT Magazine.
Exhibit at the UK's definitive and most influential 3D printing and additive manufacturing event, TCT 3Sixty.
In the same way 3D printed jigs and fixtures have become the unsung hero of major production facilities at the likes of Ford and Volkswagen, this is a prime example of how the technology is being used in creative ways throughout the production process to make those so-called “impossible parts” a manufacturable reality.
Like 3D printing, generative design comes with its own misconceptions. The idea that you pop in your goals and constraints, the software churns out a range of options and all the engineer has to do is pick which of the design results they like best certainly isn’t the case. In another example shown this week and what is believed to be the world’s first production chair created with artificial intelligence, French artist Philippe Starck worked with Autodesk and Italian furniture company Kartell to create a generatively designed chair.
Chair created using artificial intelligence with Philippe Starck.
The first set of results produced a design that met each of the structural performance and manufacturability constraints but was ultimately deemed too “ugly” by Starck. It really was, the chair looked like it might consume the person sitting in it. But that’s because algorithms are not furniture designers or artists. Through an iterative process, the team was able to go back and refine the design within the software which led to a much more simplistic and ergonomic chair that will now be produced via injection moulding and go on display this summer. The point is, while automation and artificial intelligence may be the supposed enemy of the design to manufacturing world, human interaction is crucial for tools like generative design to achieve their full potential.
During the event, Autodesk also announced a number of updates to its generative design workspace in Fusion 360 including new manufacturing constraints for CNC and a pricing feature powered by aPriori which will allow users to view manufacturing cost elements for each design result.
Speaking about the company's plans for the future of manufacturing, Greg Fallon, Vice president of Design and Manufacturing Strategy at Autodesk commented: "It's really about trying to give designers, engineers and manufacturers tools that help take out some of the burden in the process of evaluating multiple designs."
The convergence of these technologies is a positive thing. The fact that generative designs can now be fulfilled by more traditional manufacturing systems, many of which are already in machine shops all over the world, means companies can be more agile as the industry progresses. Likewise, the purposeful application of additive whether on a final part on in more behind-the-scenes steps represents a shift in mindset and brings home the point that none of these processes are out to replace one another - it's about finding the right fit for the use case.
Autodesk VP, Ian Mitchell opened this year's AU stating that the goal of the event was "to have attendees think differently about what’s possible in the design and make space." I'd say, mission accomplished.
