| Accurate dynamic identification of mechanical components is key for current and future Digital Twins of mechanical systems. However, current state-of-the-art methods do not allow non-intrusive and high-spatial density based parameter identification for components under operational conditions. The focus of this project is to develop a framework for identifying these operational parameters for detailed nonlinear finite element component models from non-contact and high spatial density optical measurements. |
In order to obtain this framework, a fundamentally new approach will be developed which tackles the image measurements and efficient inverse model evaluation in a strongly integrated setting. This approach will revolve around three key contributions. First a methodology will be developed for extracting a detailed deformation field over a wide frequency range from (relatively low frequency) image data. In an original contribution, this deformation field data will be exploited for setting up parameterized reduced order models in order to circumvent the high setup cost typically associated to these approaches. The image based data and reduced order models are then combined in a parameter identification process in the frequency or time domain, depending on the component of interest. These developments will be validated on a range of experimental setups, ranging from academic to industrial complexity.
