For polymer and composite additive manufacturing (AM), users have had two primary options – powder-bed systems that use heat to melt powdered material into shape, or filament-based systems that melt the edge of a thin polymer wire and stack the melted layers on top of each other.
With metal AM, systems such as selective laser melting (SLM) laser cusing, and direct metal laser sintering (DMLS) dominate because of a lack of usable filaments. Stoughton, Wisconsin-based The Virtual Foundry (TVF) is changing that, offering metal filaments that can be used on the same equipment designed for polymer AM.
TVF has been working with a range of manufacturers, testing its trademarked printing filament – Filamet – to gage suitability, and expand AM applications while lowering costs. Company executives recently spoke to GIE Media editors about the state of AM and how they are disrupting the market.
Direct metal printing
Unlike traditional metal AM technologies, such as selective laser sintering (SLS) or powder bed fusion (PBF), TVF’s direct-metal printing (DMP) technology does not generate a finished part out of the printer. Filamet – produced in widths of 1.75mm or 2.85mm – is a mixture of plastic pellets (the binder) with a metal powder. It’s similar to a green gear or other powdered-metal (PM) part that must undergo furnace sintering to harden.
As the printed parts heat up, and before sintering fuses the particles together into a solid whole, the binder material evaporates. Because Filamet is encased in the binder, it doesn’t require respirators, solvents, specialty chemicals, or special handling equipment, just heat, making it safer than existing laser-based metal 3D printing solutions.
Technology inventor and company founder Brad Woods says the two-step process of printing then heat-treating hasn’t been as much as a deterrent as he feared when he developed the technology. At the time, he was concerned that industrial users would stick to costly systems that could produce finished parts.
“I expected people who had access to this technology were satisfied, but they weren’t,” Woods says. He adds that aerospace defense giant Lockheed Martin was the first to test the system, “and they were very interested in this even though they had access to all SLS machines, as much as they wanted.”
Even with the furnace step, filament-based printing tends to be faster and requires less specialized training, making AM technologies more accessible to manufacturers who don’t have large numbers of specialists.
Also, because Filamet works in small desktop machines up to the massive filament-based systems that can produce entire car bodies, the technology offers larger build envelopes than otherwise available with metal AM.
Most companies experimenting with TVF’s Filamet have been doing so on a small scale – testing individual parts or using the technology as a design tool rather than a production method. Woods says he expects that to be the case for some time.
“At this point, they’re looking for methods to manufacture intricate parts – multiple component parts – and they are unable to make the entire piece subtractively,” Woods says. “In aerospace stators and rotors for example, manufacturers are working to combine elements, consolidate as many features as possible, and 3D-fabricate fins.”
TVF’s President, Tricia Suess, adds that major manufacturers have billions of dollars invested in traditional manufacturing technologies that work well, so AM will be in more of a design/support role for some time.
“The technology is advancing but won’t replace mass production today. That’s not what we’re focused on,” Suess says. “3D printing is shining in engineering and prototype shops.”
Specialty AM applications could also be a growth area. Suess notes that several companies have discussed custom filament materials for low-volume, oddball products that are hard to produce economically.
TVF also has tungsten and copper filaments in its repertoire, with both providing radiation shielding while still in a green state. An early test sample with the tungsten showed great radiation shielding properties while being lightweight, Suess notes. This development had the team thinking about other applications, and by the end of February 2019, TVF signed a joint venture (JV) with Vulcan Global Manufacturing Solutions for tungsten Filamet’s use in radiation shielding for medical and industrial purposes.
“We still have the original test sample and have since printed a collimator – it guides the radiation – but those were for proof of concept,” says Mike Daniels, TVF’s global sales leader. “Since then, we have done proof-of-concept printing of apertures for cancer treatment and held test runs with a local hospital.”
Woods adds that production volumes should increase quickly as manufacturers complete initial tests with new technologies and get more comfortable specifying 3D-printed components in various finished products. Automotive powertrain supplier ZF North America Inc. has test-printed in copper, and other motor vehicle manufacturers are playing with the technology.
“Once they’ve vetted our materials, we can move along as they prove our product on their products,” Woods says. “Being vetted in many areas and seeing a fair amount of prototyping going on means we’re heading in the right direction.”
The Virtual Foundry
Laser powder bed fusion (PBF) metal additive manufacturing (AM) produces solid and porous geometries in the same process, saving time and material. Advantageous for orthopedic implants, PBF creates complex structures that simulate the mesh-like porous properties of bone while delivering strength and durability. Porous textures can be built into implants of any shape or size – from acetabular cups to lumbar cages – allowing serial production of a portfolio of implants.
Maximizing machine use
Optimizing porosity size and distribution in orthopedic implant designs can generate high volumes of data that often slow down build processors. Betatype’s Engine build processor provides supercomputing power to overcome this, rapidly creating scan data and enabling serial production. Its build generation scalability can optimize build data.
Betatype recently worked with a company to create serial production build data that produced files in excess of 50GB. The Engine was able to scale up to 640 virtual CPUs with 4.88TB of RAM, generating build data in a few hours. Special Engine algorithms for converting complex geometry enabled implant designers to work in file formats up to 96% lighter than traditional STL files, such as Betatype’s ARCH format or nTopology’s LTCX data. For example, a spinal cage model was only 8MB as an LTCX file compared to 235MB as an STL file. These conversions simplify and shorten the process, letting designers innovate without dealing with mesh data.
Improved use of overall build volume in AM lowers cost per part, unlocking serial production of orthopedic implants. Betatype can stack implant parts by designing lattice node matched supports, using the entire build envelope to produce multiple, complex implants in a single build. Engineered supports can be removed using standard media blasting.
Reducing build times
Three major components of PBF build time need to be addressed to speed up the process:
- Dosing – Applying powder to the machine bed
- Fusion – Applying energy to the powder bed
- Motion – Movement between fusion
To optimize laser firing times and reduce delay times, with or without multiple lasers, Betatype technologies can reduce overall build times by up to 40%.
One orthopedic manufacturer using this technology portfolio decreased implant build to 15.4 hours from 25.8 hours. Betatype technologies optimized the laser scan paths, reducing firing and movement time required for complex lattice structures. Galvo-driven path optimization reduces delay times to 3 hours from 13 hours by optimizing the delays on an exposure-to-exposure level, also reducing laser travel distances to 100km from 170km.
Using Betatype’s technologies with PBF maximizes machine use, optimizes data file sizes, and reduces build times, enabling faster, more cost-effective serial production of orthopedic implants.
With “eight figures with a long way to go” of his personal wealth invested into Keselowski Advanced Manufacturing (KAM), NASCAR driver Brad Keselowski is making a huge bet on additive manufacturing (AM). The Statesville, North Carolina, shop has two massive General Electric (GE) metal additive manufacturing (AM) machines, four Mazak CNC machining centers, a GF Machining Solutions wire EDM machine, Pinnacle X-Ray Solutions computer tomography (CT) scanning equipment, a coordinate measuring machine (CMM), a Nikon 3D scanner, and advanced software systems link the machines together.
“We’re all in with this. You have to jump into this sort of thing with confidence,” Keselowski says. “It’s part of the racing culture. You don’t find too many drivers who aren’t risk takers. If you’re not willing to take blind leaps, you’re not going to win.”
Racing will still be Keselowksi’s day job, but KAM is his future, he says. It’s where he hopes to greatly impact racing technology, commercial vehicle manufacturing, military contract jobs, aerospace work, and manufacturing for any industry that can benefit from AM technologies and Industrial Internet of Things (IIoT)/Industry 4.0 data-driven process technologies.
The immediate opportunities, situated in NASCAR country an hour north of Charlotte, will be custom parts for race teams. Drivers and teams often talk about the race before the race – the design and build time when crew chiefs swap out vehicle components in hopes of gaining a technical edge in the next week’s race. It’s a pressure-driven environment in which crews have to produce flawless parts (that will be subjected to harsh racing environments) with minimal turnaround times. Keselowski says it’s an environment where AM technologies are already showing their value.
In 2000, he remembers seeing his first 3D-printed racecar part, an air intake for a stock car. He immediately wanted to put it on a car and race with it before being told that the plastic piece would melt on the track. Keselowski says he’s been waiting for the technology to mature and move into metals ever since.
With GE’s Concept Laser metal AM machines, the equipment is ready, he says, and his engineering team is ready to make parts for various industries.
“Race teams are really engineering teams that occasionally get to race. I’ve seen the successes that we’ve had and watched other teams, and I’ve seen that the teams with the best engineers always seem to have the fastest cars,” Keselowski says. “That’s the spirit we’re going to take to other industries – that fast turnaround time, that part you can depend on the first time it’s made.”
Key to the shop’s strategy is attracting other industries to hybrid manufacturing’s potential – 3D printing some parts, machining others, starting some on printers and finishing them on machines – picking the right technology for the part.
“If this only serves motor sports, it will be a tremendous failure as a business,” Keselowski says. “This business has the potential to be bigger than anything I’ve done in racing.”
Spending more on equipment at launch than most 30-person shops will spend in a decade is critical to showing potential customers that KAM is a serious operation, not a vanity project for a professional athlete, says General Manager Steve Fetch.
“Metal AM is getting to that point where you can go from prototyping to volume production, but a lot of customers need to be convinced that it’s capable and that we’re capable of producing with it,” Fetch says. “Coming from racing, we know speed, and that’s our initial pitch. If you can eliminate tooling for new parts, speed becomes a big advantage in time to market.”
A lot of companies that want to outsource parts to shops such as KAM are excited about the potential for hybrid manufacturing technologies, Fetch says, so advocates for new ways of doing things and new production partners already exist. To take advantage of that, KAM had to support those people who enable and endorse new technologies.
So, proving quality is critical, he says. KAM engineers must show a customer that the shop can produce a usable metal part, built to stringent quality specifications. Customers can’t afford to take Fetch’s word that everything will work.
That realization led to one of the shop’s largest investments – the Pinnacle X-Ray solutions CT scanner. The machine stacks 2D X-ray images on top of each other to generate 3D images of parts – showing the internal cavities that often define 3D-printed components.
Pinnacle Managing Director and CEO Rod Meyer says aerospace companies have led metal AM use, and they’re his biggest customer base for CT systems.
“With AM, most of the problems are internal, so they’re not going to show up on a CMM or a 3D scan. You need to see inside the workpiece,” Meyer says. “We can detect porosity, unintended fusions, areas where the (metal) powders didn’t melt properly, and dimensional flaws.”
Fetch says with CT reports showing interior dimensions, coupled with CMM and 3D-scan data for part exteriors, KAM demonstrates part quality without time consuming, expensive destructive testing. The company also uses Mazak’s Smartbox IIoT systems to record every action machines take, generating process reports that KAM engineers can also share with customers.
“Everybody knows that additive is in the manufacturing world, but most people don’t really understand the challenges,” Fetch says. “They understand the concept, but they don’t know the five or six things you have to do after printing to get a finished part. Expectations are either too high or too low, so we have to do a lot of explaining.”
KAM is pursuing ISO standards for manufacturing quality and aerospace productivity. Keselowski says the company is also undergoing International Traffic in Arms Regulations (ITAR) certification to work on military contracts.
With North Carolina’s racing industry, heavy truck manufacturing, aerospace sectors, and military contractors close by, he expects the shop to be busy as soon as the ink dries on its various certificates.
Keselowski says he plans to keep racing, so he won’t run day-to-day operations at his shop. He’ll visit frequently and provide strategic vision, but Fetch and his team will be in charge.
“I’ll still be racing for a long time. A lot of drivers compete into their late 40s, so I have at least another 10 years in me,” Keselowski says. “I need a job to pay for all of this, especially for the next few years.”
Keselowksi Advanced Manufacturing
Additively manufactured (AM) lattice structures are popular because they’re lighter than solid materials yet equally strong. However, the simple honeycomb shapes favored by designers have an inherent weakness. When such structures are compressed, once the force is sufficient to cause permanent deformation, the lattice can shear along one or more node planes. With nothing to inhibit this shearing, the collapse becomes catastrophic.
Working in collaboration with colleagues from Imperial College London, Professor Iain Todd from the Department of Materials Science and Engineering at the University of Sheffield has been taking a novel approach to address the catastrophic failure issue by looking at the basics of metallurgy.
Lattice structures, also known as architected materials, typically have a uniform layout with nodes conforming to a regular array with the struts between the nodes, all following common planes. These uniform lattices replicate the structure of a metallic single crystal. The nodes in the AM lattice are equivalent to the atoms in the single crystal and the struts are equivalent to the atomic bonds. In each structure, the atomic planes, or nodes, are all perfectly aligned.
In some applications, such as the high-temperature section of a jet engine, single-crystal materials are ideal because of their ability to withstand deformation at extreme temperatures.
However, Todd explains that the mechanical failure possibility remains if structures undergo deformation – shearing along their planes.
In polycrystalline materials, atomic plane alignment is random. A crack will slow down or stop when it meets a crystal where the atoms are aligned differently. In addition, it is possible to introduce different materials in phases, precipitates, or inclusions to strengthen the materials; these materials also help inhibit crack propagation.
This fundamental metallurgical understanding inspired the scientists to mimic polycrystalline micro-structures in AM lattices with the aim of developing robust, damage-tolerant architected materials.
Through computer modelling of atomic structures – scaling them up and creating meso-structures based on polycrystalline materials – engineers are transforming the way that materials are designed and coined the term meta-crystals.
Experimental testing of components made from meta-crystals demonstrates they are highly energy absorbent, with the polycrystal-like material withstanding almost 7x the energy before failure compared to the materials that mimic the single-crystal structure.
While the basic metallurgical concepts are inspiring the development of architected materials, researchers see opportunities to study complex metallurgical phenomena.
“This approach to materials development has potentially far-reaching implications for the AM sector. The fusion of physical metallurgy with architected meta-materials will allow engineers to create damage-tolerant architectured materials with desired strength and toughness, while also improving the performance of architectured materials in response to external loads,” Todd says. “And while these materials can be used as stand-alone structures, they can also be infiltrated with other materials to create composites for a wide variety of applications.”
Dr. Minh-Son Pham of Imperial College London adds, “This meta-crystal approach could be combined with recent advances in multi-material 3D printing to open up a new frontier of research in developing new advanced materials that are lightweight and mechanically robust, with the potential to advance future low-carbon technologies.”
Imperial College London
University of Sheffield