Think that manually manufacturing composites is hard? Think again. I recently tried it with my daughter in our very own kitchen. All you need are two hands, a baking mold and an average oven. This simple manufacturing process has always been one of the main attractions of composite. But it has now become its biggest problem. Let me show you why.
Making composite is like making lasagna. Layer by layer, we placed some leftover prepreg in a mold and let it bake in a 180°C oven for a couple of hours. There, the magic takes place: the soft and flexible prepreg turns into an incredibly hard and stiff structure. And even though our do-it-yourself creation might not meet aerospace conditions, the result was still impressive. This is one of the reasons a lot of people started to make their own surfboards or boat hulls.
Getting all the parameters right
But manufacturing composites is not just a piece of cake: it is a conversion process. The material changes from liquid to solid form during the production process. If you wait too long, this will happen automatically so there is a limited time you can process the material. Also, keeping the fibres well aligned when they are still in a liquid or molten resin is not easy. So in order to create high quality composites you need to get all the parameters exactly right. Indeed, just like cooking!
Now add the fact that you are dealing with two components, fibres and the resin, that have very different properties. Fibres are tiny, stiff and strong but only in one direction. They are very vulnerable and can break easily when you bend them. Resin is soft during manufacturing but brittle after curing and can crack easily. Trying to bring them together in one product suddenly gives you a whole lot more variable parameters.
Automating human craftsmanship
But the biggest challenge we face today when manufacturing composites aren’t the variables though. It is the fact that we humans are so damn good at it. We might not be strong enough to work with metals without tools and machines. But we do have delicate fingers, perfectly suited to work with composites. We can accurately build up layers and place tapes next to each other without gaps or overlap. We are good in recognizing local defects and the more we work with composites the more we understand the materials and its habits and can adjust. The reason manual labour is still used in advanced composite manufacturing such as for aircraft structures is not only because the technology is relatively new and lagging behind other industries. It is also because of our human craftsmanship.
But if we want to advance composites into more industries and larger volume applications, scaling up manual work will not work. By getting more efficient and bringing down the cost of manufacturing, the intrinsic superior properties in lightweight and durability can be used in much more applications and we can make the world more sustainable. Making this process industrial and automated is a big challenge.
What we need
So what kind of technologies do we need to make this happen? Traditionally, the applications of composites in lightweight structures such as in aerospace or automotive, has been about replacing metal structures. That is why you see similar approaches in those industries: large and heavy presses for fast processing in automotive similar to metal forming presses, or large heavy gantries in aerospace that are derived from metal milling machines. These solutions do work and have brought the cost of manufacturing down and increased the output. But we can take it further and look into smarter solutions that work specifically for handling composite materials.
Similarities across industries
Thankfully, there are a lot of industries to learn from. For example, at Airborne we have looked at the paper handling industry during the development of a high volume thermoplastic composite line, aimed at mass production. It will be presented for the first time at the JEC World 2018 and was developed with our partners SABIC and Siemens: do check it out. It’s incredible to see at which high speed a delicate material like paper can be handled. There are interesting similarities with composite material handling.
Another interesting industry is food handling. Lacquey, a spin-off of the Robotic Institute of TU Delft, developed advanced robotic technology to grip vegetables. The challenge is the variation in products and the fragile nature of vegetables, which is very similar to the composites challenge. Also vision and machine learning is advancing rapidly. Fizyr, a spin-off of the same TU Delft Robotics Institute, is a specialist in 3D vision and deep learning and won the Amazon Picking Challenge in 2016. Technologies like these are very much needed to be implemented in composite automation, to recognize the defects and variations in the material and adapt the processes on-the-fly.
Let’s get on it!
So yes, composites are easy for humans, but difficult for machines. The industry has already done a lot on automation but I would regard that as the first wave. Let´s ride the second wave by using the rapid developments in advanced robotics, vision and machine learning much smarter. But also by thinking conceptually from the soft and fragile aspects of the materials, instead of mimicking metal manufacturing.
What do you think are smarter solutions to automate composite manufacturing? Or are there concepts from other industries that handle fragile materials we can use for composites?
This blog was also posted on LinkedIn.
