|
Title:
|
|
Drag coefficient distribution on a wing at 90 degrees to the wind [ECN-C--95-061]
|
|
|
|
Author(s):
|
|
|
|
|
|
Published by:
|
Publication date:
|
|
ECN
|
1996
|
|
|
|
ECN report number:
|
Document type:
|
|
ECN-C--95-061
|
ECN publication
|
|
|
|
Number of pages:
|
Full text:
|
|
38
|
Download PDF
|
Abstract:
The drag coefficient along a horizontal axis wind turbine blade or anairplane wing, for about 90 degrees angle of attack, is the studied quantity
(CD90) in this report. The arguments to assess CD90 are (1) that a typical
design feature for the rotor is to have blades which are unpitchable, the
method to avoid too high loading being handled by natural stall (stall
control). Ultimate loadings for such turbines may come from the situation
when a storm wind, parallel to the main shaft, hits the parked rotor. Flat
wise bending occurs, creating potentially very high stress levels along the
blade; (2) a similar need as in 1 may arise for a pitched rotor for which
there is always a combination of angles for pitch, azimuth and yaw for which
the blade presents its flat area to the wind; (3) the need to assess the
coefficients of both lift (CL) and drag (CD) for high angles of attack while
the turbine is operating. The connection between CD90 and CL and CD, at
arbitrary angles of attack, is briefly described in this report; and (4) some
rotors have aerodynamic brakes in the form of pivoting tips. Their braking
capability is calculated for very high angles of attack typically around 90
degrees. In order to find a method for the CD90 assessment two different
hypotheses are investigated. These hypotheses are basically concerned with
the distribution of CD90 along the blade. This distribution can be constant
(Method 1) or variable (Method 2). The basis for finding CD90 comes from
literature covering statistics from many examples of flat plates without
taper and four sets of data from one small and three full scale wind turbine
blades. Following Method 1 an approximate value for CD90 of 1.4 constant
along the blade is used. It is good for the majority of today's wind turbine
blades. However, for more slender blades the number is higher. It is
concluded that the amount of data from real blades is too small for an
ascertain assessment of the quality of Method 2. Yet, Method 2 should be
useful whenever alternative designs are compared. Finally, a recommended path
of calculation is presented in the form of a stepwise algorithm. 20 figs., 3
tabs., 14 refs.
Back to List