Back to the programme printer.gif Print




Delegates are invited to meet and discuss with the poster presenters in this topic directly after the session 'Advanced rotor technologies' taking place on Tuesday, 11 March 2014 at 11:15-12:45. The meet-the-authors will take place in the poster area.

Bernhard Stoevesandt Fraunhofer IWES, Germany
Co-authors:

(1) Fraunhofer IWES, Oldenburg, Germany

Printer friendly version: printer.gif Print

Abstract

Dynamic stall simulations with OpenFOAM

Introduction

Numerical simulations of airfoils are of great importance for the aerodynamics of wind turbine blades. Since most load calculation code rely on polars of specific airfoils, the data needs to be known. However, due to manufacturing reasons, often slight changes in the geometry of the airfoils are needed. Also, the inflow at the turbines is not always as smooth as in wind tunnels. Fluctuations in the angle of attack at smaller time scales are quite frequent, especially in strong wind shear conditions. Polar data for such conditions is nevertheless rare. Measurements for every airfoil type are very expensive.

Approach

Here we will present the results of steady and unsteady computational fluid dynamics (CFD) simulations using the open source code OpenFOAM. The results are compared with measurement data from Bak et al. [Bak 2000] using a NACA63-415 airfoil and an identical setup.
The simulations were done using a 2D-computation with periodic period boundary conditions in the spanwise direction. The structured mesh of


Main body of abstract

57000 cells consisted of three different parts: A C-grid around the airfoil, surrounded by an inner O-grid. This was surrounded by an outer O-grid. The reason for this setup was, the possibility to use the AMI functionality of OpenFOAM and rotate the two O-grids against another for dynamic stall simulations. .
The simulations were run fully turbulent using the k-omega-SST model by Menter [Menter1994].
Simulations were performed for angles of attack between -6 to 18 degrees angle of attack. Since steady state results using the simpleFoam solver for angle of attacks larger than 10° show strong deviations, all simulations with higher angles of attack were calculated using the unsteady solver pimpleFoam. The later has also been used for the simulations with fluctuating angles of attack at 9.6 and 15.6 degrees.
Comparing the measurement results and the simulations, the lift curve for the simulations at steady angles of attack show an overshoot around stall predicting a two high maximum lift (Cl-max = 1.4 instead of Cl-max=1.25).
Further simulations were done using a changing angle of attack according to the experiments by Bak et al. using a mean angle of attack of 9.6 and 15.6 degrees respectively. The amplitude of the fluctuation was set to +/- 1.5 degrees for the 9.6° case and +/-1.8 for the 15.6° case using a reduced frequency of k=0.092 as given by Bak. The results were quite satisfying, showing in general the same structure of the lift. However, the lift and the effect of the change in angle of attack were a bit overpredicted.

Conclusion

2D-CFD simulations have been done for changing angles of attack on an airfoil using the open source tool OpenFOAM. Although the lift curve for steady anlges of attacks predicted a two high maximum lift, the simulations of the periodic changing angle of attack showed good agreement with measurements. Further research will be needed to generalized the the setup for airfoil simulations.


Learning objectives
The use of CFD for aerodynamic purposes becomes more and more reasonable and affordable for industrial applications. The accuracy depends on the setup and can be quite well.