MEchanics GAthering -MEGA- Seminar: A cluster-based network analysis of post stall load fluctuations
Event details
| Date | 07.04.2022 |
| Hour | 16:15 › 17:30 |
| Speaker | Fatma Ayancik (UNFoLD, EPFL) |
| Location | Online |
| Category | Conferences - Seminars |
| Event Language | English |
Abstract The cycle-to-cycle variability of the vortex shedding and the associated fluctuations of the aerodynamic force coefficients during the dynamic stall are physical phenomena. This phenomenon always drops beneath the researcher’s radar due to phase-averaged presentation of velocity fields and lift histories. A correct representation of the maximal and minimal bounds of aerodynamic force fluctuations is important for the accurate prediction of the aerodynamic efficiency and the dynamic stall hysteresis of wind turbines or helicopter blades. We used cluster-based analysis to identify the chain of events leading to flow separation and associated post-stall lift fluctuations and their bounds. A feature space is spanned by the aerodynamic forces. The feature space of trajectories is then partitioned into discrete clusters with similar aerodynamic characteristics using unsupervised clustering. Each cluster corresponds to a characteristic phase of the flow in the feature space, and it is represented with a cluster centroid. Statistical properties of cycle-to-cycle variations are determined at each cluster centroid. We propose an inverse model to rebuild the experimental data from the information obtained from clustered statistical properties of cycle-to-cycle variations. This model would be used to introduce post-stall fluctuations to dynamic stall models for the accurate prediction of dynamic stall hysteresis without the requirement of additional second-degree terms.
Bio Fatma Ayancik is a postdoctoral researcher at UNFoLD, EPFL. She received his B.S. and M.Sc. in mechanical engineering from TOBB ETU and Ph.D. in mechanical engineering from Lehigh University in 2019. She is working on modeling of dynamic stall with data driven methods.
Bio Fatma Ayancik is a postdoctoral researcher at UNFoLD, EPFL. She received his B.S. and M.Sc. in mechanical engineering from TOBB ETU and Ph.D. in mechanical engineering from Lehigh University in 2019. She is working on modeling of dynamic stall with data driven methods.
Practical information
- General public
- Free
Organizer
- MEGA.Seminar Organizing Committee