How far can additive manufacturing processes be automated? Ever since industrial 3D printing arrived in the industry, this question has been passionately discussed. Especially during the Covid 19 crisis with the associated supply bottlenecks, calls were made for a realignment of the supply chain.
Additive manufacturing also often came into play here, although it requires a higher degree of automation for higher volumes. This was already a hotly discussed topic at Formnext 2021 - but how widespread is automation already in the industry and what solutions are there here?
Leading AM companies have been running pilot projects to explore technologies to automate plastic 3D printing and overcome current technological challenges and barriers.
For example, the Polyline project was launched in March 2020 with a planned duration of three years to automate individual processes in the AM production line and connect them from part design to the final product. The technology is to be generally qualified for larger series and the automation is to further increase the quality of the products. The goal is also to bring this to market as an end-to-end solution, as part of a so-called holistic view of the automated AM workflow. "There's no point in just developing individual things," says Martin Friedrich, former Project Lead Series Projects Additive Manufacturing at BMW Group, "everything should be connected and fully automated."
For the consortium, which is funded by the BMBF (German Federal Ministry of Education and Research), EOS has brought together 15 industrial and research partners from Germany to each contribute their individual expertise to cover all aspects of software and hardware know-how to overcome the many technological hurdles in automation.
"So far, everything is labour-intensive"
"From a supplier perspective, we see a lot of interest in scaling polymer 3D printing technology," says Marcus Andrä, Product Line Manager Polymer Periphery and Automation at EOS. "The demand for a fully integrated, automated production line is very high."
The reasons why a large OEM like the BMW Group should tackle the automation of AM processes are obvious. "So far, everything is labour-intensive. A large part of the production of AM parts today is invested in repetitive processes such as unpacking the parts," says Friedrich. "As an automotive manufacturer, we want to be as efficient as possible and make the best use of our employees' capacities."
"It's often also about cost," adds Paul Osswald, R&D project manager for non-metallic additive manufacturing at BMW Group. "If AM wants to take the next step and increase production volume, costs have to come down. Automation is one way to achieve this. There is downtime, you have to rely on people being there. If you are able to run the machines productively without anyone being present, efficiency increases, which in turn lowers the final cost per part. Lower costs are also critical to scaling such manufacturing."
Automation is implemented on the production line through a combination of software, hardware, AI and robotics. In particular, automating an entire production line means automating the individual machines and the communication between them. Machines such as the EOS P 500 are fully automated polymer systems that can perform a range of operations such as loading, unloading and cooling parts without human intervention. Automated powder handling, unpacking, cleaning and finishing stations all operate autonomously and are interconnected by automated transport systems.
Algorithms for a more efficient overall process
The software is another important aspect. Real-time monitoring collects data from the workstations and uses it to develop algorithms to make the overall process more efficient. The product is tracked at every stage and can be checked at any time thanks to the thermal imaging cameras and sensor data from the machines. As a result, quality is controlled and monitored throughout the process, not just at the end.
At the same time, recording the data of all processes also improves traceability enormously. "If a BMW customer wants to know where all the information on a produced part is, they can simply press the button and get the information immediately," says Andrä.
However, there are limits to how far the technology can work independently of human operators, especially in the post-processing phase where the machines need to be fed many different structural and design requirements. "Currently, we can't automatically respond to complex geometries and new materials," says Julius Legenmajer, senior product engineer at DyeMansion. "This means that the process has to be adapted by our experts to the almost limitless variety of geometries and specifications of 3D printed parts. What we need as a next step are intelligent machines that can respond dynamically to changing input parameters and adapt to the limitlessness of AM designs."
Implementing this intelligent flexibility is a particular challenge because it actually contradicts the nature of automation. "Automation is about repeatability," explains Oliver Elbert, Head of Additive Manufacturing at Grenzebach. "AM is also so attractive as a manufacturing technology because the parts can have a very specific, individual layout. That is very difficult to automate."
An ideal set-up could produce series parts, but at the same time also be used in a prototyping environment. "Our main goal is to establish standards for AM in the industry to integrate AM as a reproducible and reliable process," Friedrich says. "For parts manufacturing in general, AM is a small brick in the big wall that has yet to find its place." The goal of the Polyline project is therefore to overcome these limitations and create standardised and repeatable processes for the entire production line.
Increasing efficiency and reliability of individual work steps
While pilot projects like Polyline aim for end-to-end automation, automation technology is already being used in the industry to increase the efficiency and reliability of individual steps. As Andrea Landoni, Product Manager 3D Printing EMEA at Protolabs, explains, as a 3D printing service provider, Protolabs has automated post-processing steps to serve a number of different sectors. For this reason, the company's main interest in automation is the ability to apply the same processes to many products simultaneously.
"Certain finishing steps are necessary on every product," Landoni explains. For parts made using the SLA or Polyjet processes, processes that are usually used for prototyping, all supports have to be removed. On the other hand, for SLS or Multijet Fusion, which are used for final products or small batches, powder removal is required. On the other hand, some products require more specific parameters and post-processing steps. "For us, it would not make sense to spend hours setting the parameters for a specific job," says Landoni. "What we need is to automate processes that can be applied to any product to improve reproducibility, times and costs."
Currently, Protolabs has automated machines for de-powdering and smoothing. Previously, both were done manually and the process was therefore more prone to errors. With an automated machine, efficiency can be greatly increased, especially in terms of consistent times and reproducible quality that does not depend on operator accuracy.
Protolab's R&D teams are always on the lookout for new automation technologies for their equipment. In Munich, they recently moved to a new, larger building to redesign their production processes and integrate new automation technologies step by step. "Although automation is most valuable to us at the moment where we can use standardised steps in post-processing," says Landoni, "there is no reason not to upgrade our automation technology to include a wider range of applications in the future."