Identification of modal parameters of machine tools during milling

机床的振动限制了机床的工作精度和生产率。它们对工具和机器零件的磨损有很大的影响。此外，它们还通过噪音和地面振动影响环境。实验模态分析（EMA）是研究机床动态特性的一种成熟方法。由于EMA假设被调查对象的行为是线性的、互反的、时不变的、因果的和稳定的，因此在进行EMA时，机床是在静止状态下评估的。因此，在加工过程中发生的影响，如由陀螺效应引起的模式变化，由不同的静态预载荷引起的非线性行为部件的刚度变化，由工件附加质量、工件固定、刀具等引起的惯性质量差异以及包括工艺阻尼在内的不同阻尼特性被忽略。原则上，运行模态分析（OMA）提供了在测量分析中考虑这些影响的可能性，因为它捕获了机床在与实际切削条件相关的工作点的动态行为。在机床上执行OMA需要新的方法来通过切削过程进行宽带激励，考虑到机床的时变行为，以及处理机床的操作负荷，特别是切削过程，与OMA中假设的激励不同这一事实。虽然宽带激励和时变行为已经在自己的初步工作中得到解决，但该项目侧重于开发两种处理激励的新方法，当从机床运行期间捕获的振动信号中识别模态参数时，这些方法集中在几个测量点上。在此背景下，将为这两种方法开发新的数学工具，并对激励特性（统计性和确定性）以及干扰的影响进行研究。这两种新方法将通过仿真模型和实验加以发展。最后，在机床上进行了实验验证。除了考虑与机床运行相关的影响外，OMA在机床中的成功和可靠应用可以显著地有助于使用模态参数等系统描述值来监测机床状态和加工过程。此外，OMA可用于数字孪生体的周期性更新，以反映机床的真实结构力学性能。

Vibrations in machine tools limit the working accuracy and the productivity. They significantly contribute to wear of tools and machine parts. Furthermore, they affect the environment by noise and ground vibrations. The experimental modal analysis (EMA) represents an established method for the investigation of the dynamic behaviour of machine tools. As the EMA assumes linear, reciprocal, time-invariant, causal and stable behaviour of the investigated subject, machine tools are assessed at standstill, when conducting EMA. Therefore, the effects occurring during machining like changed modes due to gyroscopic effects, changing stiffness of parts with non-linear behaviour caused by the different static preloads, differences in inertial mass resulting from the additional masses of the work piece, the work piece fixation, the tool etc. as well as different damping properties inclusively process damping are ignored.In principal, the Operational Modal Analysis (OMA) offers a possibility for consideration of such effects during a measurement analysis due to capturing the dynamic behaviour of a machine tool in the operating point related to the real cutting conditions. Performing OMA on machine tools requires new approaches for broadband excitation by the cutting process, for consideration of the time-variant behaviour of a machine tool and for the treatment of the fact, that the operating load of a machine tool, particularly the cutting process, is distinct from the excitation being assumed in OMA. While the broadband excitation and the time-variant behaviour have been addressed within own preliminary work, this project focusses on the development of two new methods for treating an excitation, that is concentrated on a few measuring points when modal parameters are identified from vibration signals captured during operation of a machine tool. In this context, new mathematical tools for the two methods are going to be developed and the excitation character (statistic and deterministic) as well as the effect of disturbances are investigated.The two new methods are going to be developed by using simulation models and experiments. Afterwards, the methods are validated in experiments on a machine tool.Beside the consideration of the effects related to the operation of a machine tool, the successfully and reliably application of OMA in machine tools can significantly contribute to monitoring of machine tools condition and machining processes using values for system description like modal parameters. Moreover, the OMA can be used for a periodic updating of digital twins in order to reflect the real structural-mechanical properties of a machine tool.