The perfect, unorthodox design

What if you could create the perfect part unlike anything that exists today? New CAD tools combined with 3D printing can make that happen, makers say.

Jean Thilmany, Senior Editor

No matter what you call the design method, having the computer generate the designs from am engineer’s directions may just be the future. Unorthodox, 3D printed shapes can be found in aerospace now. Soon, such designs will be all around you.

3D printing methods are enabling the development of shapes unproducible by other manufacturing methods. Now, CAD developers are including design tools that take full advantage of the capabilities of 3D printing. These tools are often labeled generative design or topology optimization. They enable engineers to use design software in a new way to best fit design needs.

In April, Autodesk released generative design to subscribers of its Fusion 360 Ultimate product development software. The design concept allows engineers to define design parameters such as material, size, weight, strength, manufacturing methods, and cost constraints–before they begin to design. Then, using artificial-intelligence-based algorithms, the software presents an array of design options that meet the predetermined criteria, says Ravi Akella, director of product management at Autodesk.

“Our effort now is in helping people define the problem they’re trying to solve,” Akella says. “That’s a shift in focus in this industry and makes people have to change the way they have to work.

“The software asks the user preliminary questions. ‘What sorts of materials would you consider for your design? Where does it connect with other things as part of an assembly? What are the loads? What are the pieces of geometry?’” Akella says.

This Elbo chair was designed using Project Dreamcatcher, which was the name Autodesk used for its generative design tool before officially unveiling it this spring.

The software then presents designers and engineers with an array of design options that best meet their requirements. Designers choose the best option. Or, if none of the options meet their needs, they can begin the generative process again, this time offering slightly different inputs.

The computer-generated (“generative”) designs might be unorthodox, new, and unexpected, with geometries that wouldn’t naturally occur to the designer. Yet, no matter how different, if the design is shown to work, it can be created through additive manufacturing,” Akella says.

The method adds value to the present way designers use CAD software, he adds.

“None of these generative questions are asking ‘What is your solution and please start documenting it,” he says. “Without generative design, it’s like engineers were using a piece of paper to explain the problem to themselves. Our job is to get all of that into software.”

He compares generative design with the job of the wine merchant.

By using Autodesk’s generative design and additive manufacturing technologies, engineers at Stanley Black & Decker shaved more than three pounds off this crimping tool attachment, reducing the weight by more than 60%.

“Someone walks into a wine store and wants a Cabernet Sauvignon,” he says. “To get the best version you go in and say ‘It’s summer and this is what’s on my dinner menu’ and you’re trusting the sommelier to present you with a variety or vintage you’ve never heard of.

“Generative expands your solution options, which sometimes aren’t intuitive,” Akella adds. “Users look at their results and think ‘I never would have thought of it. I’m not sure it’s the right answer but I’m going to check it out further.”

By any other name?

Akella takes issue with what he calls “technologies that masquerade as generative design,” which, he says, include topology optimization, lattice optimization, or parametrics.

“Topology optimization assumes you have a solution you’ve thought of and are making a better version of that solution,” he says. “But generative design expects the user to define the problem they’re trying to solve. Then we use cloud computing and other technologies to present them with a set of solutions that solve their problem in a practical, manufacturable way.”

Generative design produces many valid designs instead of an optimized version of an already-modeled part.

“Optimization usually involves removing excess material without any notion of how something is made or used,” he says.

Generative design also takes manufacturability into account, which reduces an engineer’s need to redesign products after manufacturing weighs in, Akella says.

But developers and executives at other makers of CAD technology may take issue with that depiction of their topology optimization features, which can radically change designs and reduce weight and slash costs, they say.

SolidWorks introduced topology optimization capabilities into its recent release of SolidWorks 2018 Simulation Professional and Simulation Premium.

“We expect the computing platform to anticipate your design goals,” said Gian Paolo Bassi, chief executive officer at Dassault Systèmes SolidWorks, when he spoke at SolidWorks World 2018 in January.

“The era of design and validate is about to end. We are entering the era of optimize and manufacture,” Bassi said.

That means designers specify the aspects of the part they absolutely need, including loads, constraints, boundary conditions, and manufacturing methods. The CAD tool then supplies many versions of a near-optimized part, Bassi says.

Topology optimization can be an additive or subtractive algorithm, meaning it can create parts based on user inputs like loads and boundaries or it can subtract from an existing design by essentially chiseling away at the part, says Robbie Hoyler, a SolidWorks elite application engineer for TPM, an engineering services and design provider in Greenville, S.C.

SolidWorks uses the subtractive method. It creates a meshed part based user-defined loads, constraints and boundary conditions. The software cuts out elements that offer few structural or manufacturing benefits. This process is then repeated until the part meets all constraint requirements, Hoyler says.

The optimized CAD design shows engineers the areas of the part that need to stay and the areas where material can be removed, Hoyler says. He cited an example in which SolidWorks topology optimization reduced the weight of an existing part by 50% without removing areas designers had flagged as necessary.

The part can then be saved as a mesh body in the stereolithography (SL) format for 3D printing or can be retraced as a new SolidWorks part.

Another software package, Inspire, also features generative design and topology optimization tools. It allows users to save the enhanced design as a CAD model (skipping the retracing step). The software is from Altair channel partner solidThinking.

The Inspire’s generative feature is easy to learn and is ideal for small and medium-size businesses with little or no simulation experience, says James Dagg, Altair’s chief technology officer for user experience.

Solid Edge, from Siemens PLM Software, of Plano, Texas, also includes a generative design feature that brings topology optimization to the Solid Edge 3D product development toolkit, according to the company. With the feature, designers define a specific material, design space, permissible loads and constraints and a target weight, and the software automatically computes the geometric solution.

The results can be immediately manufactured on 3D printers, or further recreated as a Solid Edge model for traditional manufacturing. Designers can run multiple weight targets, load cases and constraint scenarios simultaneously, according to Siemens PLM.

Refining the real world
Recently, engineers at automaker General Motors began putting Autodesk’s generative tool to the test to cut weight from GM vehicles. Lighter cars use less fuel, emitting less carbon.
Since 2016, the automaker has launched 14 new vehicle models with a total mass reduction of 350 pounds per vehicle, says Ken Kelzer, GM vice president of global vehicle components and subsystems. The 2019 Chevrolet Silverado, for example, reduced mass by up to 450 pounds as compared to earlier model years.

To further lighten the load, as it were, in May the automaker announced an alliance with Autodesk that will use additive manufacturing and Autodesk’s generative tool to develop future cars and trucks, Kelzer says. The pairing of additive and generative capabilities is a natural for the automaker, Kelzer adds.

GM has used additive technologies for more than 30 years to print 3D parts. The automaker has more than 50 rapid prototype machines that have produced more than 250,000 prototype parts over the last decade, Kelzer says.

And the generative capabilities now included in the Autodesk CAD systems put those printers to work in unique ways, he adds. “When we pair the design technology with manufacturing advances such as 3D printing, our approach to vehicle development is fundamentally different; to co-create with the computer in ways we simply couldn’t have imagined before, Kelzer says.

But the design of formerly unimaginable parts doesn’t mean the engineer lacks ingenuity. Rather, those reduced weight, and perhaps rather odd-looking shapes are the whole point of the generative process, which provides thousands of solutions to one engineering problem, Akella says.

He gives the example of a designer who wants to create a chair. Typically, the designer would start with some geographical representation of the chair humans have taken for granted for centuries; that is, four legs, a seat, and a back. But if the designer were to begin by specifying the amount of weight the chair must support, the materials it will be comprised of, and its cost, “the designer will get hundreds or even thousands of options he or she couldn’t have conceived of on their own,” Akella says.

The nature of the creation process also allows for a part with such complex geometries that it can replace multi-part assemblies. And they can be created with 3D printing, he adds.
The process is also being tested in other industries that design with CAD tools.

For instance, architects at Arup, the building and infrastructure design consultancy in the Netherlands, paired topology optimization and additive manufacturing to redesign a steel node for a unique, public lighting and artistic tensegrity structure.

Needle Tower, public art by American sculptor Kenneth Snelson demonstrates the concept of tensegrity. The piece is located outside of the Hirshhorn Museum and Sculpture Garden in Washington, D.C.

Buckminster Fuller coined the term tensegrity to refer to a structure that uses the principle of floating compression, with parts compressed inside a net of continuous tension with cables or tendons delineating the system. Think, of course, of his famous geodesic domes.
Arup designers created their trio of tensegrity structures for a shopping street, the Markstraat, in The Hague. Unveiled in 2013, the “urban chandeliers” integrate street lighting and add an artful element to the area.

Arup architects designed several variations of the node using conventional and optimization techniques. The third figure, on the right, is the final, lightest shape attained through topology optimization.

The urban chandeliers are beautiful. But they weren’t easy to create, says Salomé Galjaard, an Arup senior designer for the project.

Due to the irregular shape of the structures most of the 1,600 nodes that connected the cables to the struts, were different due to the more than one thousand variations in angle and position of the attached cables, Galjaard told attendees at the 2015 Future Visions symposium in Amsterdam.

The Arup architectural firm created this 3D-printed, optimized node for a study of nodes used in its urban chandeliers street-lighting project in The Hague.

“This ‘uniqueness’ inspired us to learn more about additive manufacturing,” she says.
Curious as to what optimization could have done for them on a project like the urban chandeliers, Arup designers conducted a study. Both topology optimization and additive manufacture have been little used in the architectural world, so this seemed like an excellent opportunity.

After performing topology optimization, using the Optistruct software from Altair, they found the node they’d modeled traditionally closely resembled the optimized node. And yet, that optimized design reduced the weight the node from 44 pounds to 11 pounds, a 75% drop, without compromising the functional and structural performance of the product, Galjaard says.

Still, the designers spent much more time working with the complex, optimization software than they normally would, and the process could be frustrating, she says.

“Our research illustrates that 3D printing can have a positive impact on the design and production process and the functional product,” she says. “The resulting costs of future construction products could be decreased significantly, whereas architectural freedom will be increased dramatically.”

Generative design and topology optimization can bring the same design freedom and cost reduction to engineering and other types of design of course. Imagine a complex, oddly shaped part that is printed and performs as an assembly. In other words, a part beyond your imagining. That’s the promise of generative design and topology optimization.

Altair
www.altair.com

Autodesk
www.autodesk.com

Dassault Systèmes
www.3ds.com

Siemens PLM Software
www.plm.automation.siemens.com