Fraunhofer Institute for Production Technology IPTResearch project "dynaTWIN"
Digital twin predicts vibration dynamics during milling of turbomachinery components
In milling, the vibrations of workpieces during the process often lead to high tool wear and, at the same time, to costly damage to the surface of the part. Especially tight manufacturing tolerances, such as those for thin-walled components for turbomachinery, can often no longer be maintained due to such vibrations of the workpiece.
During the milling process, the stiffness and mass of the workpiece change continuously due to the material removal, the tool position and various other influences, so the dynamics also vary during milling: This can lead to high local vibration amplitudes. Such strong vibrations are then the cause of increased wear of the tool and surface damage to the workpiece. Costly rework of the part then quickly leads to higher production costs.
In order to determine stable process parameters, time-consuming trial-and-error procedures are currently required in production. However, it would be better to be able to predict the milling dynamics during process design and avoid instabilities in 5-axis milling. One solution would be to identify and set the optimum spindle speed prior to milling and actively control it during the process when necessary.
Simulation of workpiece and tool dynamics in the milling process
The aim of the "dynaTWIN" research project is to develop a data-driven simulation module in which the variable dynamic properties of the manufacturing system are taken into account so that a highly productive milling process for engine components is enabled. The CAx module simulates the dynamics of the coupled mechanical system, consisting of tool, workpiece and fixture, and also takes material removal into account. Sensor data is acquired and processed during the milling process so that finally all dynamic data can be collected and visualized on the basis of a digital twin.
Self-optimization through continuous data-driven adaptation
The dynamic process model optimizes itself with the help of the collected and processed vibration data by continuous adaptation. The new procedure is intended to achieve significant improvements in the milling process:
- Reduction of vibrations in the machining process by up to 70 percent
- Reduction of machining time by up to 50 percent for finish milling
- Increase in surface quality by up to 35 percent
- Reduction of mechanical tool wear by up to 30 percent
- Reduction in workpiece reject rate ("First Time Right")
Optimization of spindle speed with the digital twin
In the first stage of the "dynaTWIN" project, the Fraunhofer IPT, together with the project partners, will simulate vibrations of the workpiece using material removal, force simulation and FE modal analyses.
In the next phase of the project, various sensors will be used to acquire real vibration data during the milling process. The processed data will be transformed into a digital twin that represents the actual milling dynamics. Subsequently, the milling dynamics are simulated using digital models to identify deviations between the real and the predicted vibrations.
In the final step, the spindle speeds are adjusted in practical trials wherever deviations exist between measured and simulated vibrations that are caused by varying external influences. These include, for example, different material batches, manufacturing tolerances of the tool, overhang length of the tool or profile deviation of the component.
Funding
The "dynaTWIN" research project is funded by the German Federal Ministry of Education and Research (BMBF) as part of the Eurostars (Eureka) program.
Project Management Agency
German Aerospace Center e. V. (DLR)
Project Partner
- Fraunhofer IPT (project coordination)
- AICRUM IT; Murcia, Spain
- ModuleWorks GmbH, Aachen, Germany
- TUSAS Engine Industrie Inc.; Eskişehir, Turkey