nTopology, based in New York, NY, USA, makes software that aims to improve advanced manufacturing outcomes with its 3D modelling design technology.
The software is based on a proprietary algorithm that uses equations to represent a 3D solid by calculating the distance to its surface from any point in space. To render, the model is raytraced so that no matter how complex it is, it can still be visualized as it is edited in real time, the company says. The software automates tasks that take hours, days or weeks with traditional design tools, due to its ability to build reusable workflows.
nTopology, established in 2015, recently formed a number of partnerships with AM companies, including EOS and Renishaw, to improve AM techniques as well as direct transfer of more accurate design data to 3D printers.
According to the company, their nTop Platform software can help position, orient and prepare parts for AM from a set of common build platforms, add lattice support structures more easily and quickly, and slice part data—avoiding stereolithography (STL) files, which nTopology says can be error prone. At the end of the design optimization process, the software can export the sliced data directly to the machines in the format that drives the laser during printing. (Figure 1.)
Performance criteria
nTopology recently updated its nTop Platform 2.0 computational-modelling software into automated toolkits that package and streamline important design capabilities for AM and a variety of other advanced manufacturing technologies. Toolkits include lightweighting (via part consolidation, shelling, conformal ribbing), architected materials (engineered for performance at any scale), topology optimization (optimizing for a variety of performance criteria— including stress, displacement and stiffness—under multiple loading conditions), AM (including a large library of lattice structures) and design analysis (integrated with simulation tools such as Ansys, Abaqus and Nastran). (Figure 2)
This latest development was funded with US$20 million raised by the company in July 2019 to expand its design capabilities and increase its customer base, bringing nTopology's total funding to US$31 million, at a valuation of some US$90 million. In September the U.S. company opened its first European office in Regensburg, Germany. ‘Our expansion into Germany is the logical next step since that is the base for much of European manufacturing,’ said Bradley Rothenberg, founder and CEO of nTopology.
I spoke to Blake Courter, CTO at nTopology, about the company’s software and what it could mean for metal 3D printing.
What are the challenges around computational modelling for 3D printed parts?
In traditional design, the design engineer needs to find out only two things: the shape of the part and the material it is made of. AM offers a third choice of fine, ‘mesoscale’ structures that can make powerful contributions to overall part performance, acting as architected materials. At the same time, AM brings new challenges in quality control and repeatability. The new capabilities open up an enormous design space that can be intimidating to navigate, and each new process brings new control challenges. Engineers need a modelling and design system that can be used to prove out new concepts, standardize them, and roll them out throughout an organization.
How does 3D printing integrate with computer aided design (CAD)?
The main CAD systems on the market today were designed for drawing shapes, and they include specialized tools for specific processes like sheet metal and injection molding. The process of designing for advanced AM applications can use those tools, but typically users want to provide different functional treatment, such as strong versus lightweight, in different regions of the design. Although the CAD tools can be adapted to maintain these models, no mainstream CAD tools on the market today contain a comprehensive set of tools to imbue parts with the specific, new superpowers enabled by 3D printing.
Our computational modelling software features CAD geometry combined with process knowledge and simulation to synthesize functional, high-performance geometry with a documented, repeatable and scalable process.
Tell me about 'architected materials' – how important is this aspect of the process?
Metal 3D printing could be divided into two kinds of applications: parts that could be manufactured traditionally, but for which, for logistical reasons, AM is preferred; and parts that use architected materials to achieve performance breakthroughs or other superpowers. (Figures 3 and 4.)
It can be difficult to keep up with the ever-increasing technologies and applications enabled by architected materials, but some recurring themes with our customers include:
- Creating parts that are lightweight, stiff, and interchangeable with existing parts
- Heat sinks and heat exchangers, which can experience massive efficiency increases
- Bio-scaffolding and the creation of biocompatible surfaces
- Applications in antennae, wave guides, and radiofrequency modulation
- Improving the 3D printing process itself through more functional support fixtures.
In the last case, even when printing parts that themselves do not include a mesoscale structure –such as supports – you can still create more accurate and cost-effective results.
Is metal powder particularly challenging as a material to be assessed at the design stage?
Engineers’ trust in materials evolves over time, as data comes in. Powder metal manufacturing brings new challenges in repeatability and reliability, so engineers will likely find applications as the risk/reward profile evolves.
We have partners and customers who produce AM hardware, produce raw powered metals, and operate service bureaus catering to specific verticals. These vendors realize the overwhelming complexity of the space can be a barrier to growth, and they are taking it upon themselves to research and develop application-specific workflows and then distribute them to end users. Such workflows not only make it easier for them to adopt their technology, they can add enough value to be sold as products. Our software is both a development and distribution platform for such intellectual property (IP) products. (Figure 5.)
Where do you see metal 3D printing in ten years’ time?
I think the best way to predict the future is to make a list of problems that need to be solved (and maybe start working on them). In terms of what has to improve, users should be able to achieve reliable, deformation-free, metal additive parts. I’m not sure that there is a single solution to this challenge, but we should be able to produce shapes the customers want, reliably. Moreover, simulation and real-world data should guide the creation of high-performance geometry. At the moment, end users have to make significant investments in research to achieve cutting-edge results, but this can be commodified by technologies such as our nTop Platform.
Hardware will consume more intelligent data and use more closed-loop processes to produce repeatability in otherwise impossible situations. Embedded systems should be able to self-correct for deformations and other errors, as well as provide assessment of build quality. Slice data will be accompanied by more metadata, such as intended phase and thermal gradients, so the hardware can make better decisions.
What still needs to be done in order to scale up the 3D printing process, and have it accepted more universally?
We need to work together as an industry to make metal AM as easy as plastics prototyping became in the last decade. By creating application-specific workflows and tools, more and more end users will be able to benefit from the technology without having to make large research investments up front.
How important are the learning and adapting aspects of computational design?
It’s reasonable to say that the complexity of the space can be overwhelming. Design engineers aren’t used to having as much freedom as is now available. The solution is to narrow down the problem to a certain domain, and then start, slowly, to explore the space. One needs a system that can capture knowledge from the earliest phase of getting the process working, then add data and engineering learning to refine the process. Also, it’s important that the acquired knowledge be easily reused as well and understood and adapted as needs evolve.
This article was originally published in June 2020.
