Additive manufacturing sees initial uptake for serial production
Laser additive manufacturing (AM) is becoming increasingly established as a serial production tool in automotive and aerospace.
This is due to a number of ongoing developments – such as advancing metrology systems, better data handling and optimised multi-laser scanning strategies – dramatically ramping up its repeatability and productivity.
Such was shared in a recent Laser Systems Europe webinar by Zach Murphree, VP of Global Sales and Business Development at Velo3D, and Jasmin Saewe, Head of Laser Powder Bed Fusion at Fraunhofer ILT.
The two experts were interviewed by our Editor, Matthew Dale, who sought to establish the degree to which laser AM has been making its way into these major industries for serial production.
Aerospace: Rapid uptake in space and defence
Murphree began by explaining that Velo3D has focused on the goal of developing AM systems for serial production since its founding, and that demands for such systems are now increasing in industry. “This is particularly the case for aerospace, where we’ve noticed quite a rapid uptake of AM for serial production here in the US,” he said.
However, as one might expect, this fast adoption hasn’t been in heavily regulated commercial aviation, but rather in the specialist sub-sectors of space and defence.
“This is partly because there is a lower bar here when it comes to verification – a lot of the qualification is done by testing individual parts, rather than needing to obtain an FAA certificate for both a part and an entire manufacturing process,” Murphree explained. “So these sectors have been really fast to adopt AM, which has also partly been driven by them having the highest demands in terms of component performance – for example with regards to lightweighting and durability.”
He noted that flagship AM components in space and defence include rocket engines and other propulsion components, which is where Velo3D has seen the most value and fastest adoption of AM technology.
“This is a great example of taking a lot of the universal advantages of AM and incorporating them in real components,” he said. “So in both these sectors, lightweighting is an obvious advantage enabled by AM, but it has also enabled parts such as cooling jackets, injectors and other complex cooling components to be consolidated into single parts, in addition to unlocking other dramatic performance improvements.”
Left: A 1m tall, one-piece rocket nozzle made on a Velo3D Sapphire 1MZ system out of IN718 with 20 injectors and optimised internal thin wall regenerative cooling | Right: A rocket thrust chamber printed out of GRCop-42 on a Velo3D Sapphire system – both on display at Formnext 2022
Asked whether having this now-established foothold in space and defence could potentially accelerate AM’s uptake in commercial aviation, Murphree remarked: “I definitely think so. I’d draw comparisons with the space race in the 1960s, where systems and components that had been rapidly qualified for manned space flights started to make their way into other fields and industries, including aviation. We’re seeing a similar effect happen now, with NASA once again starting to work towards qualifying AM systems and components for manned space flight, which will be the first step towards increased qualification of components for commercial aviation in the future.”
Enhancing metrology and calibration
Key to the success of AM’s uptake in serial production will be the ongoing developments in metrology and calibration being made by system developers such as Velo3D.
“When we started out, it was all about enabling our customers to print parts with geometries they otherwise wouldn’t have been able to produce,” said Murphree. “This was a key step in drawing more people into the world and capabilities of additive manufacturing. But as you scale into production there are other things you also need to deliver, such as cost-effectiveness, throughput, quality and overall: consistency. These are all aspects we’ve been able to improve through advancing our metrology and calibration systems.”
An example of these advancements is Velo3D’s ‘One Click’ calibration capability, which enables a user to quickly calibrate the optical system of their printer without requiring external measurement tools or a field service engineer. This means that the calibration can be performed prior to each build, which means that the system is in a known good state when a print starts. Typically this sort of calibration is time-consuming and requires the attendance of service personnel from the system manufacturer, and so is usually performed every 90 days or so. This leaves periods of time where the system may or may not still be in spec for certain builds.
Advancing Velo3D’s metrology and calibration has been crucial for facilitating what Murphree says are the two frameworks currently taking shape for AM serial production.
“One is what many people might envisage already – a factory floor filled with numerous AM systems, all printing parts simultaneously – and we do have customers like this,” he explained. “The other is where our customers use just one AM system for product development, and then rely on a network of contract manufacturers to produce their parts at scale. This second framework in particular relies immensely on each AM system in the network being able to deliver consistent parts that match exactly the specifications sent to them by the original customer. This is where the optimised metrology and calibration of our machines becomes of critical value.”
Establishing trust through data ‘pedigree’
Another important factor that will influence AM’s uptake in serial production is quite simply, trust.
“For a traditional process such as casting, there is literally decades of data available on both the process itself and the parts it produces, and so there’s this established pedigree of data,” said Murphree. “We need to get to the point where AM parts have that same level of data and backing that enables people to have confidence in the mechanical properties, quality and consistency of the process.”
While headway is slowly being made here in commercial aviation, according to Murphree, it’s only the biggest names in the sector that currently have the resources to get AM processes and parts certified, due the extensive amount of testing and validation required. This limits the amount of data being generated surrounding the deployed processes and parts.
“We’re currently working on overcoming this barrier at Velo3D by providing data around the quality of printed parts in a format that our customers can rely on as part of their qualification process,” Murphree said. “This reduces the amount of redundant work that has to be done for AM to really grow at scale.”
Velo3D has developed and defined numerous processes using a range of materials for its systems, and distributes that information – as well as the mechanical properties of printed parts – to its customer base. “By doing this we’re gradually establishing that required pedigree of data for a range of qualified, fixed processes that can then be used across different manufacturers to produce consistent results,” remarked Murphree. “With that you really start to get the scale of data and statistics required to hit the critical mass that will accelerate AM adoption and drive it into other industries as well.”
Automotive: Cost-per part needs to come down
Despite the numerous regulations surrounding aerospace – especially commercial aviation – acting as a barrier to widespread AM adoption in series production, it does have certain leniencies that aren’t shared with other industries.
“In aerospace there exists a higher freedom in terms of the amount a part is allowed to cost, and so it’s more the quality and performance of a manufacturing process and the parts it produces that leads it to being chosen – as opposed to the expense,” said Jasmin Saewe of Fraunhofer ILT. “Automotive on the other hand is a very cost-driven industry. To introduce a new manufacturing technology, a business case must first be clearly defined. Even if a part performs better due to it being produced by AM, if overall the process is more cost-intensive per part, it will not as likely be invested in by automotive manufacturers.”
However, similar to space and defence in the aerospace industry being more lenient with regards to cost in favour of performance, there have also been certain sub-sectors of the automotive industry acting as early adopters of AM for the same reason. “The racing sector in particular has been keen to adopt AM technology to produce high-performance parts,” confirmed Saewe, “We’re also starting to see more and more additively manufactured structural parts such as brackets or brake calipers being used in the premium car market as well.”
In order to penetrate the consumer-vehicle automotive sector however, AM will have to lower the overall cost per part in order to create a business case for the technology. This will involve increasing the productivity of both the AM process itself as well as the entire automotive process chain.
“But importantly, the key to getting AM into the main automotive sector will be to implement it without changing the overall process chain too much, as each new process and part has to be qualified and standardised – although not quite to the same extent as aerospace,” said Saewe. “So while you certainly could apply AM across the whole process chain, that actually wouldn’t help the implementation at all due to the sheer volume of qualification and standardisation work that would have to take place.”
Saewe noted that part of the solution in reducing cost-per-part in automotive could be to get the most out of existing AM technologies using what’s called an ‘adaptive process strategy’. This refers to, rather than using one set of parameters for an entire part, instead using parameters that deliver the exact level of quality required for each segment of the part. “For example, there are certain areas of automotive parts that have larger forces acting on them than others, and it is here where the parameters delivering the highest-quality – and consequently the longest print times – will be required. For the other areas of the part, less time could be taken, which while delivering ‘less’ quality, would actually increase productivity while delivering the exact level of quality required. This adaptive strategy is not yet being used enough amongst developers and users of AM technology.”
Implementing such a strategy was one of Fraunhofer ILT’s main goals within the recent three-year collaborative project IDAM (Industrialisation and Digitalisation of Additive Manufacturing), which saw the development and installation of two fully autonomous serial AM production lines for automotive parts at partners BMW and GKN Additive.
Separation of printed components in an automated separation station on BMW's new serial production line. (Image: BMW Group)
“Together these two lines demonstrate exactly how AM in automotive can be approached,” remarked Saewe. “Fraunhofer ILT was heavily involved in optimising the digital process chain, including how to adapt and control the AM machines, how to prepare all the data and how to determine the optimal process parameters for each part. This involved fusing the output of multiple, high-resolution sensors – both visual and thermal – using machine learning algorithms, which allows us to begin to predict part quality for quality assurance purposes. This prevents the need to use destructive quality assurance processes, which are both time consuming and expensive for the automotive industry.”
The new lines will use laser powder bed fusion (LPBF) technology to produce around 50,000 components per year in common part production, as well as more than 10,000 individual and new parts. Using LPBF means certain tools are no longer required and new design possibilities can be realised to produce the parts – which greatly increases flexibility and cost-effectiveness for BMW and GKN Additive.
However, despite the great success of this project, Saewe believes the automotive industry is still a ways off from using AM to produce parts for consumer vehicles. “In addition to working with BMW, we’re also working together with other automotive companies to develop new AM parts and materials,” she said. “It’s through this work (in addition to project IDAM) that we’re really seeing the challenges automotive firms have to face in developing a business case for the technology – the sheer volume of parts that have to be made to even begin justifying AM usage is extremely high.”
Laser advances helping drive productivity
In addition to improvements along the digital process chain, Saewe noted that the recent introduction of multi-laser scanning strategies as well as beam shaping technologies are helping dramatically increase the productivity of AM.
“These advancements will all strongly benefit the automotive industry,” she remarked. “We’re seeing more and more system providers launch multi-laser systems that ramp up AM productivity.” The most extreme of these is currently SLM Solutions’ NXG XII 600 series, which utilises twelve 1kW lasers to offer build rates of up to 1,000cm3/hr.
In addition, lasers with in-built beam shaping functionality have the potential to dramatically improve productivity, Saewe continued. Such sources enable rapid switching from Gaussian beam shapes to ring-mode shapes and have been emerging for applications such as cutting and welding in recent years, and more lately for additive manufacturing.
Fraunhofer ILT displayed a part produced by a multi-laser gantry system at Formnext this year – an 800mm diameter turbine rear frame made of nickel-alloy 718.
“These different intensity distributions are able to induce more laser power into the parts without introducing defects, meaning the process speed can be increased,” Saewe explained. “And so that's something I think will be used more and more in additive manufacturing – some machine providers have already announced the integration of those lasers into their commercial systems.”
Fraunhofer ILT has recently been bringing all the latest AM advancements together in the development and optimisation of its own systems being developed and built in-house.
“For example, we’ve been exploring the idea of large scale additive manufacturing for applications in industries where larger, higher performance parts are in demand, for example in space and aviation” said Saewe. “And so we’ve developed a machine from scratch that can print larger parts up to 1,000 x 800 x 350mm in size. Every aspect of it, from the data preparation and the optimised shielding gas flow to the scanning strategy of the five laser systems, was completely designed and developed by ILT experts to deliver faster build times for large parts.” At the AM trade fair Formnext this year, the ILT displayed a part produced on the new system on its booth – an 800mm diameter turbine rear frame made of nickel-alloy 718.
New materials still required
In addition to the development of the technology itself, research into new materials is equally as important for further driving AM adoption in series production.
“While some say there are already plenty of materials to work with in LBPF, I would say that in conventional manufacturing, there are several factors more materials you are able to select from,” Saewe commented. “This is crucial for the widespread uptake of LBPF, as there will be many industry companies out there that would like to adopt AM technology while also being able to use a material similar to that which they already use with conventional manufacturing.”
At this year’s International Laser Technology Congress, AKL 22, hosted at Fraunhofer ILT, multiple papers were presented on the use of AI in the development of new materials. “That was an interesting approach to see,” Saewe remarked. “In addition, we also saw a definite increase in the number of papers reporting on increases to both the productivity and quality of LPBF. Standardisation was also a big topic – which of course generated a lot of interest in industry, as this will be key to their widespread adoption of the technology.”