How to automate the production of sandwich panels

In one of our last blog articles, we’ve already presented the grant-supported project ACoSaLUS (automated composite sandwich lay-up system) and its latest developments. In the past months, the project partners have made further progress on which we like to report. Cevotec has focused on the improvement of the laminate design and patch placement process as well as the robot-based, automated handling of honeycomb cores. Additionally, the optical post-placement control of patches using AI has been further improved.

In order to implement an end-to-end automated lay-up process, we have further developed our general Fiber Patch Placement technology during this joint project and successfully placed the first skin patches directly onto the strongly curved tool (see video below). The placement features our Rolling Motion movement of the robot. Rolling Motion is a movement of the 6-axis robot to place patches precisely on highly complex, convex shaped surfaces. The movement of the placement robot follows a curve with a different normal for each point of the curve. To support this accurate placement, our gripper has been improved as well and now features enhanced vacuum & blow off controls. Currently, we are optimizing the placement parameters in preparation for the final demonstrator production.

We also improved our patch vision inspection and are developing a post-placement control utilizing AI based image recognition to match the strict manufacturing specifications and quality requirements needed by our aerospace customers. After teaching the AI’s neural network, the system can now detect edges and map different materials correctly. This enables our vision inspection to identify a successful placement as well as FOD and other defects. As illustrated in the picture below, the system can differentiate between various materials, e.g. honeycomb (incl. “hard points” for assembly), UD and fabric patches. The new functionality of the vision inspection will be also tested in the production process of the final demonstrator.

Last year we’ve already presented our innovative honeycomb gripper (patent pending), which can handle honeycomb cores automatically. After successful tests, we’ve now integrated the gripper prototype into our SAMBA production system. The 6-axis robot is able to pick and precisely place the honeycomb core fully automated on the mold (see video below). The additional process step now automated with FPP technology represent a major achievement of this project, as honeycomb placement could not be automated until now. A very critical, manual process in the production of composite sandwich parts can thus be eliminated, increasing reproducible quality and production economics.

The project has entered its final phases of execution with completion until year’s end. The demonstrator produced as part of this project will be showcased at upcoming trade shows in the second half of this year – we’ll keep you updated!

This research and development project was funded by the Bavarian State Ministry of Economic Affairs, Regional Development and Energy (funding code LABAY113C). The responsibility for the content of this publication remains with the authors.