In-Process Virtual Quality Monitoring

Currently, industrial metal cutting requires expert knowledge of the operator about the used machine tool, workpiece material, and the involved cutting tools to synthesise a part program. Once a part programme is created and run on the machine, only safety related real-time checks, interrupts, and interventions are programmed through the state-machine. This type of approach does not provide sufficient indication on the process efficiency and the production quality. And although non-cutting operations can be integrated within this program to check process efficiency and production quality, they either do not give accurate information or cost valuable machining time. Consequently, it is inevitable that most of the parts need to pass quality control afterwards, especially when it comes to safety-critical parts in aerospace. For some of the larger components, the post-process metrology cost can be up to 25-50 % of the total cost of the part. With the backdrop of Industry 4.0 and increased importance of in-process quality estimation—especially for the manufacturing in aerospace— the present work investigates an approach for virtual real-time quality estimation in machining titanium alloy components. A digital-model of the machine tool is created to realize this. Apart from the kinematic properties of the machine tool, this digital-model incorporates a dynamic compliance map of the machine tool augmented with dynamic models of various tools and tool holders. By simulating this digital-model in real-time with data obtained from the machine controller, and additional sensors, it estimates the 3D geometry of the part being produced on the machine. The real-time data consists of compensated tool position from machine tool controller, cutting forces, and tool table from the machine controller. Collectively such data is called the ‘digital fingerprint’ or ‘digital shadow’ of the process. The generated 3D faceted body in realtime—called as ‘in-progress digital workpiece’—can be easily compared to the target CAD of the part to estimate the quality of machining. This type of workflow provides better nformation on the process quality in real-time and thereby eliminates or reduces quality control time and effort in subsequent steps.