(321b) Identification of High Energetic Flow Structures in the Separation Chamber of a Uniflow Cyclone with POD

For the optimization of the design criteria of a uniflow cyclone CFD simulation are carried out. However, numerical results require validation, e.g. due the sensitivity of the velocity distribution on the boundary conditions. Therefore reference PIV measurements are used to evaluate the flow field in the separation chamber of the uniflow cyclone [1]. The principal of the measurement technique is to record snapshots of tracer particles in the fluid flow to visualize the movement inside interrogation windows and calculate corresponding vectors [2]. Due to turbulence and local instabilities in the fluid flow pattern the cross-correlation analysis for each snapshot results in varying velocities per vector and snapshot. The goal is to identify the most energetic modes of the flow field to smooth the velocity field and to reach a better comparability with respect to the calculated velocity distribution over the examined region of interest. For the identification of the most corresponding fluid structures, proper orthogonal decomposition (POD) is introduced to extract these statistically coherent structures. The solution of the eigenvalue problem leads to the POD modes (eigenvectors) with the associated eigenvalues. These values represent the contribution of each POD mode to the overall power of the signal. The eigenvalues are ordered downwards. The modes with the highest contribution represent the most energetic structures [3]. This method leads to a better comparability of the CFD calculation and the measured PIV data. Moreover this leads to a more significant prediction of the flow pattern an assists the design criteria decision process.

[1] Kofler, T., Pillei, M., and Kraxner, M., â??Influence of the Inlet Vane Geometry On the Uniflow Cyclones Performance (639f)â?. AIChE 2015 Annual Meeting, 2015.
[2] Raffel, M., ed., Particle image velocimetry: A practical guide, 2nd ed. Experimental fluid mechanics. Berlin: Heidelberg, 2007.
[3] Chen, H., Reuss, D. L., Hung, D. L., and Sick, V., â??A practical guide for using proper orthogonal decomposition in engine research,â? International Journal of Engine Research, vol. 14, no. 4, pp. 307â??319, 2013.