In-process prediction of surface delamination during circumferential milling of carbon fibre-reinforced components

• Applicants: Professor Dr.-Ing. Jan Hendrik Dege; Professor Dr.-Ing. Andreas Fischer
• Subject Area: Production Automation and Assembly Technology ; Plastics Engineering ; Measurement Systems ; Metal-Cutting and Abrasive Manufacturing Engineering
• Term: since 2026
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 574307962

Lightweight components made of carbon fiber reinforced plastic (CFRP) are manufactured close to their final contours, but almost always require machining to finish the edges, which often leads to component damage. Because CFRP machining differs fundamentally from metal machining due to the highly orthotropic material properties, machining models for metallic materials cannot be adequately transferred. Despite extensive research into individual influencing factors and damage mechanisms, there is a lack of holistic CFRP machining models that take micro-scale mechanical effects into account, which is why component damage cannot be reliably predicted. The project´s goal is therefore to create a 3D process model for a CFRP contour milling process that can predict top layer delamination based on surface deformation data collected during the process. The model will examine in particular the (interactive) effects of (wear-related) tool geometry, fiber orientation, and process parameters. The surface deformation of the component is determined using the laser speckle photography laser optical measurement method and converted into strains, which are compared with limit strain values derived from the process model, from which a delamination prediction can be derived. The process model and the measurement technology are designed and validated using a well-defined reference milling process, which is first used to collect and compare deformation data and damage information for various process parameters (e.g., cutting speed) for an ideally rigid workpiece clamping. In addition, the project investigates how non-ideal component clamping during the normal machining process affects the measurement uncertainty of the deformation measurement, what influence the lateral deformations to be expected in the process have on the measurement uncertainty, and how these can be separated from the deformations of interest here in the direction of the surface normals. Since the 3D process model is far too slow for use in delamination prediction during the process, additional research is being conducted into how and with what deviations the process model can be converted into a faster replacement model. Finally, in-process validation of delamination detection and prediction is to be carried out in a commercial milling machine. The aim is to use the results of the first phase to clarify the uncertainties involved in using in-process deformation measurement and a fast process model to achieve early detection of delamination formation in parallel with the process, which influencing factors contribute significantly to the uncertainty budget, and which process variables have a decisive influence on component damage. Based on these fundamentals, a process control system for preventing delamination damage will be developed in the second phase.

• DFG Programme: Research Grants