The Delft University of Technology’s collaborative R&D center develops sensor- and robot-based automation for composite material 4.0 manufacturing, including the STUNNING project for welding and assembly of thermoplastic composite body demonstrators. #cleansky #composites4_0

Photo source for all images: SAM|XL, and amazing project: GKN Fokker.

Intelligent Advanced Manufacturing XL or SAM|XL (Delft, The Netherlands) is a 2,000 square meter collaborative research center located on the campus of Delft University of Technology, which uses composite materials and other advanced materials to develop intelligent manufacturing automation. It was established in 2018 by two departments of Delft University of Technology-Aerospace Engineering and Cognitive Robotics-in collaboration with partners in the aerospace industry in the region, including GKN Fokker Assembly Automation Competence Center (Papendrecht), Aerial Bus defense and Dutch Aerospace (Leiden), KVE Composites Group (The Hague) and Airborne (The Hague), as well as the regional development agency EFRD. Kjelt van Rijswijk, CEO of SAM|XL, said: “There are many centers that use robots for composite material processing, but most of them are based on offline programming. The industry needs a closed-loop manufacturing center.”

SAM|XL has an inter-academic and interdisciplinary team from Delft University of Technology, as well as a dedicated central team composed of composite materials, aerospace robots, software and electromechanical engineers. It focuses on fiber placement and winding processes, 3D printer-based additive manufacturing, quality assurance (QA) and inspection, machining, surface treatment and assembly processes, such as thermoplastic composite welding. It pursues three main types of projects:

SAM|XL is a non-profit organization supported by companies that formally join the participants. They made recommendations on the center's strategic direction and technical roadmap, and shared investments in its infrastructure. SAM|XL owns the infrastructure and provides it together with relevant expertise to projects in industry and academia. Suppliers that support SAM|XL in establishing and advancing the central infrastructure are technology partners.

We will continue to build our own team of engineers," van Rijswijk said. "We believe that it is very important to connect contextual areas such as aerospace and composites with software, robotics, and mechatronics. The team is mainly composed of the latter, but several of us are mainly from aerospace and composite manufacturing. Our project team is always composed of product and manufacturing engineers from our partners, as well as machine integrators and software personnel from SAM|XL. We work together to refine the concepts, test and demonstrate them, and possibly support their final implementation into production. "

He explained that one of the main goals is to "have a software toolbox that we can share and an open infrastructure in which we can adjust these toolboxes and our infrastructure to realize the special use cases of our project." Van Rijswijk points out that this also establishes the knowledge and capabilities of the center, which can then be used to advance the manufacturing of other projects.

“In everything we do with robots,” van Rijswijk said, “we can listen to the sound of the sensor and implement a closed feedback loop. Although the way we do this will vary from use case to use, based on manufacturing and part requirements, we The idea is to work with our industry and academic partners to develop closed-loop control automation solutions that are flexible and can be expanded in the future.” He took automatic welding tools as an example. "You need to have a closed loop of temperature and speed to ensure that your welding is good, and it also needs to be related to the geometry of the components that the welder must follow."

A key aspect of this closed-loop control is zero programming of the robot. "This means you must have robotic devices and sensors that communicate with each other," van Rijswijk said. He showed a grinding robot and a metal workpiece. "The metal workpiece is unique, the robot has never seen it before," van Rijswijk said. "The robot is also not pre-programmed. Instead, it scans the workpiece, and the operator selects the specific surface the robot will grind. The robot recognizes the selection and performs the grind operation on its own. The operator is not programming the robot, but just clicking which ones Surface and proceed. Robots are using sensors to teach themselves what to do, rather than waiting for expensive programmers to program them."

This is reminiscent of the June 2020 sidebar "There is no business case for teaching robots." Dr. Michael Kupke, the head of the Lightweight Production Technology Center (ZLP, Augsburg, Germany), asserted that "Composite 4.0 is more than just using robots. . It’s this technology that ensures you don’t have to teach robots because there is no business case.”

Van Rijswijk agreed, "Because that will be programming again. Especially in a high-variation, low-volume manufacturing environment with large composite products, you have to retrain the robot for each change that occurs." He gave some examples: Product changes, changes in shape, changes in the orientation/location of the production workshop, and the surface of the product (such as polishing the entire surface or only rough patches?). "Even in unlikely circumstances, teaching is commercially viable," Van Rijswijk said. "The robot is only performing its task, not monitoring/controlling the material deposition and polymer conversion processes that ultimately give the composite product strength."

“You can teach robots,” he continued, “but this does not mean that you will get a product of reliable quality, just as the product is made by certified composite workers. Take wind turbine blades as an example, two half-moulds The bonding layer between may sometimes be 1 cm thick, sometimes 5 cm thick. And these wind turbine blades obviously do not always match the CAD model of the blade with centimeter accuracy when manufacturing. However, the skill of sanding the bonded blade Workers can make up for these differences. But if you just send a robot based on a common program, it will become a mess. You need direct feedback between the 3D cameras to tell the robot what the shape of the turbine blade is at that time, And the robot can be recalibrated. If you don’t do this, when you finish teaching a robot, you might as well use cheaper labor to sharpen the blade manually. I totally agree that teaching is just another way of programming, so in our use case type There is no business case. On the other hand, robot teaching is a very powerful tool for setting up research experiments and robotics education."

Part of what makes this coordination between robots and sensors possible is that SAM|XL uses ROS Industrial as the robot operating system. "In fact, it is more like an environment than a software platform," van Rijswijk said. "It is a middleware that allows nodes to be connected together, where each node represents a piece of software that can read sensors or mobile actuators." ROS Industrial is specially developed for connecting robots with external sensors and is in the global robot community Used to develop this intelligence. "For example, if you want to get real-time feedback between a 3D camera or a pressure sensor, then you need something that can send updates to the robot controller in real time. Normally, your robot controller can’t do this because it’s just from The programmer receives the fixed program and executes it blindly."

Another advantage of using ROS Industrial is that it has nothing to do with the brand. "Whether you use Kuka, Fanuc or ABB robots, it works the same way," van Rijswijk points out. "Normally, if you start programming with Kuka, you will either be locked in Kuka by the vendor, or if you want to switch to a new programming language, you must start over from scratch. However, with ROS Industrial, you only You need to replace the Kuka configuration file with a different configuration file to reuse the entire intelligence on different robots. For us as a research institution, it makes more sense to use middleware, because not all of our participants and technical partners are Use the same type of robot."

van Rijswijk said another factor in additive manufacturing is that “design must also be part of the closed-loop manufacturing process, because digital design is the benchmark for the verification process and manufacturing parts. For me, composite materials are additive manufacturing; therefore, 3D printing is no different. Therefore, if you want to do closed-loop manufacturing, your design needs to be part of it. Therefore, what we are doing is ensuring that your component design is directly related to your manufacturing process."

"If we are moving from manual to closed-loop digital processes, then your design must change," he continued. "Products designed for manual production use the reference of human sensors, for example, that can be seen by the human eye. The design of automated manufacturing takes into account different aspects."

Van Rijswijk regards the quality assurance and certification of safety-critical and load-bearing 3D printed composite parts as the "elephant in the room", but it has not attracted enough attention. "Everyone is doing 3D printing, but if you cannot follow the information provided by the printer in carefully monitoring/controlling the process conditions, how can you ensure that the mechanical properties produced meet the design values ​​used? The only way is to print more The quality of a product is verified through destructive testing. This is not only very expensive, but also defeats the purpose of single-piece manufacturing through 3D printing. This is why I think this kind of data-accessible closed-loop manufacturing is necessary for affordable certification The premise. Then you have the data to show that the product is good and qualified. It links the sensor to your automated process, and this is what we do."

Van Rijswijk sees new, artificial intelligence-based authentication methods for additive manufacturing and composite materials being developed—although still in its infancy—and manufacturing that is more interactive with the production environment (including production workers). "All sensors can generate a lot of data, which can be projected onto a screen or augmented reality glasses," he pointed out. "This way at least the operator responsible for the process steps can be notified. It may be just a green light instead of a red light, so he or she At least you can safely press the robot’s play button and know that it is correct. There will also be a combination of humans performing human tasks, robots performing robotic tasks, and humans and robots working together. But both need to have a good interface in order to Understand what each other is doing from a practical and safety perspective."

One of SAM|XL’s current projects is to assemble the lower half of the Clean Sky 2 Multi-Function Airframe Demonstrator (MFFD) as part of the STUNNING project (see "Proof of MFFD's LM-PAEK Welding" and "Making the upper part of the airframe" Half part ") MFFD"). The lower half of the MFFD is 8 meters long, 4 meters wide, and 2 meters high. Leo Muijs, chief technical expert of the GKN Fokker assembly system, explained that assembly began in early 2021. NLR (Netherlands Aerospace Center, Amsterdam) The skin was made using AFP and then cured in an autoclave of Airbus in Stade, Germany. (https://www.nlr.org/news/nlrs-stunning-project-departs-for-next-generation -composite-planes/). After trimming to net size and NLR non-destructive testing, the finished skin will be delivered to SAM|XL at the end of October. Clean Sky CfP09 consortium TCTOOL has developed several end effectors for assembly and fuselage welding fixtures These will also be in place on SAM|XL at the end of October.

Large welding fixture designed by TCTOOL partner TWI. Image source: GKN Fork

GKN Fokker will use a pick-and-place system manufactured by TCTOOL's partner FADA-CATEC (Seville, Spain), which will be operated by the SAM|XL gantry system to place the stringers and ultrasonically locate and weld them in place. The laser projection system will be able to accurately position the stringers, which will be welded by continuous conduction welding. This will be done using welding fixtures designed by TCTOOL Alliance leader TWI (Cambridge, UK). The fixture is equipped with a 1-meter conductive welding head developed by GKN Fokker.

Automated welding in STUNNING will include (1) Conductive welding of stringers using welding beams (GKN Fokker); (2) Ultrasonic welding of clamps/skins and clamps/frames (TU Delft/SAM|XL); (3) Equipped Conductive welding of the floor grille and frame (GKN Fokker). Image source: GKN Fork

Next, the pre-assembled frame made by GKN Fokker will be welded in place using ultrasonic welding and injection clips in the EcoClip program. The end effector for this ultrasonic welding operation was jointly developed by SAM|XL and TU Delft. After the system is installed, the pre-installed floor grid will be connected to the fuselage by conductive welding through an end effector, which is mounted on a collaborative robot that hangs upside down on the SAM|XL gantry system. The end effector is designed by TCTOOL partner LSBU (London South Bank University, UK), and is equipped with a conductive welding device developed by GKN Fokker for short welds.

SAM|XL will work with partners in the PENELOPE project to integrate closed-loop manufacturing into welding components to achieve industry-ready processes. Image source: Penelope website

van Rijswijk explains: "Our work will focus on how to make this type of assembly an industry-ready continuous process." "There is no coupling between metrology and robotics. This will be done in the PENELOPE follow-up project that has already started. We cannot implement all the automated steps in a single one-time demonstrator. The lessons learned from STUNNING will be used to develop how to integrate closed-loop manufacturing into a welding-based assembly process. The goal is to account for thickness variations and without manual programming. Environmental conditions change."

SAM|XL will also participate in ENLIGHTEN is another plan. Under the leadership of Delft University of Technology professor Clemens Dransfeld, it will develop multi-scale modeling and physical testing. "The plan will include the formation of crystallinity at the microscopic level, which determines the strength of the polymer at that location. It will also study the fibers, their movement during the welding process and the subsequent stress development. Only when these microscopic phenomena are taken into account, In a large coherent process, from the micro to the macro to the micro, after such a multi-stage method, you can hope to get a correct prediction of the intensity of development."

"Similarly, STUNNING has not reached the level of fully closed-loop manufacturing, but it makes sense that we have to go so far. If I am welding, then I need more information to evaluate the quality of the joint. Therefore, this conference is a step towards The first step of this goal. It is about mechanization and mobile robots, as well as the development of the ability to connect sensors directly to the robot for continuous process monitoring and adjustment. The next step is to develop through the PENELOPE project, basically from these robot-based Get quality-proven components in the process."

Low-density SMC is leading the way because fiber-reinforced polymers are making progress in automotive parts that must be pleasing to the eye.

JEC World will be held in Paris from March 14th to 16th. CW previewed some products and technologies that will be displayed at the exhibition.

Magnetic microwires are capable of non-contact measurement of temperature and pressure during curing and use.

© 2021 Gardner Business Media, Inc. Privacy Policy [Login]