Near wake characteristics comparison of a tidal stream turbine with URANS-AL model
Wei Kang1*, David Apsley1, Tim Stallard1, Peter Stansby11 School of Engineering, University of Manchester, UK
Wake development has a great impact on performance of downstream turbines and ecological change when tidalturbines are installed in arrays. In this paper, velocity deficit profile and turbulent characteristics are studied using URANS Actuator Line (AL) model in comparison with experimental data in near-wake region. The results show that the URANS-AL model agrees well with experiment in wake velocity profile. However, there is great disparity for turbulent kineticenergy comparison in which tip vortex correction for turbulence should be considered in AL model.
Recent works have extensive studies on wake comparison and loading of tidal stream turbine but mostly in far wakeregion (x>8D). In this study, wake profile and turbulent characteristics in URANS-AL are compared with experimental data near wake (0<x<1D). The computation in this study is conducted using STREAM, an in-house finite-volume solverfor (RANS) equations on multi-block, structured, curvilinear meshes using SIMPLE pressure-correction algorithm. k-εturbulent model is used with standard wall function. The experiment is carried out in IFREMER flume (Payne, Stallard etal. 2017, Payne, Stallard et al. 2018)
The computational model is derived from the X-MED turbine (Payne, Stallard et al. 2017, Payne, Stallard et al. 2018)with a rotor diameter D=1200mm and mean flow velocity is 0.8m/s. Figure 1 shows the wake profile from CFD agreeswell with the experiment. The turbulent characteristics in near wake region in Fig 2 shows the turbulent development fromambient and cohere structures. It is seen from Fig.3 there exists a great disparity near wake where URANS-AL model canonly capture the ambient turbulence development but fail to describe the tip vortex structure in turbulent mixing
Fig.1 Comparison of velocity profile between CFD and experiment. Fig.2 Near wake turbulent characteristics at TSR=5.5. Fig.3 Comparison of k between CFD and experiment at x=0.1D
In this study, the velocity profile near wake agrees well with the experimental results, although the wake deficit with CFD method shows slightly wider. As for turbulent kinetic energy comparison, great disparity in near wake region shows URANS-AL model can only capture the ambient turbulence development but fail to describe the tip vortex structure inturbulent mixing. On-going work will focus on the tip vortex correction for turbulence prediction in AL model.
The authors would like to acknowledge the support of the EPSRC Supergen Offshore Renewable Energy Hub.References:. Apsley, D. D., T. Stallard and P. K. Stansby (2018). "Actuator-line CFD modelling of tidal-stream turbines in arrays." Journal of Ocean Engineeringand Marine Energy 4(4): 259-271.. Payne, G. S., T. Stallard and R. Martinez (2017). "Design and manufacture of a bed supported tidal turbine model for blade and shaft loadmeasurement in turbulent flow and waves." Renewable Energy 107: 312-326