Improving Robotic Drilling Accuracy in CFRP Structures through RTM-Embedded Reference Markers and Parallel-Kinematic End-Effectors

acm7 Tracking Number 42

Fiber-reinforced plastics (FRP) are widely established in the production of large structural components due to their lightweight properties. In the transition to CO₂ neutrality, priorities are shifting towards highly efficient and energy-optimized production methods. The demand for innovative solutions is particularly high in the aerospace industry, such as for drilling functional holes in wing structures made of carbon fiber reinforced plastics (CFRP). Characterized by small batch sizes and tight tolerance requirements, these processes are difficult to operate economically with conventional CNC gantry machines, as the small machining volume is opposed by poor utilization. This work presents an innovative manufacturing approach combining a standard industrial robot with a specialized drilling end-effector. The end-effector integrates a parallel-kinematic compensation unit (hexapod) with a milling spindle and a laser light sectioning sensor. Its novelty lies in the use of integrated reference markers embedded directly into the component by Resin Transfer Molding (RTM), enabling repeatable and precise localization. The drilling sequence follows a two-step positioning procedure: (1) Coarse positioning - the robot places the end-effector within millimeter accuracy above the marker; (2) Fine positioning - the hexapod performs a 3D scan of the marker with the integrated sensor. A best-fit algorithm computes the marker’s center from the point cloud and determines the transformation to the tool center point (TCP). This transformation is applied through the hexapod unit, compensating for translational and rotational deviations. The subsequent drilling is executed entirely within the hexapod, ensuring superior stiffness and precision compared to the robot’s serial kinematics. Preliminary tests demonstrate a sub-millimeter positional accuracy improvement of around 50% and a reduction in standard deviation exceeding 70% compared to conventional robot drilling. Final verification will extend these findings under varied conditions and advanced strategies. The presented method decouples positional accuracy from typical influencing sources such as insufficient robot rigidity, calibration inaccuracies, thermal effects, and part deformation by utilizing local referencing. It simplifies setup, reduces manual calibration, and increases system productivity. Using standard robots instead of CNC gantries reduces investment costs and moving masses, enhancing both economic and energy efficiency - representing a promising approach for automated CFRP machining.