We're planning a major overhaul of AeroDyn this year.  Below is a list of things we currently plan to do.  If you have additions to the list or opinions on priorities, please send email to Marshall Buhl.

CY2005 Plans:

• Instead of using 1 sec of BEM to initialize GDW, use the non-iterative, no-swirl BEM equation to initialize GDW's history.
• Add new turbulent wake state model for generalized dynamic wake, as recommended by Dave Peters.  Add state calculation routine for generalized dynamic wake and include the number of states (in radial and azimuthal directions) as input variables.
• Add state calculation routine for generalized dynamic wake and include the number of states (in radial and azimuthal directions) as input variables
• Add upwind tower wake effects.
• Add ability to read new TurbSim binary wind files.

CY2006 Plans:

• Latest proposed AeroDyn input file.
• Use NWTC Subroutine Library.
• Convert the input file to the NWTC style.
• Totally rewrite the code from scratch.
• Use the IEC coordinate system.
• Improve readability of code.
• Change input of blade properties and analysis elements.  Properties will be defined from root to tip in non-dimensional locations.  Analysis elements will be specified in a separate table.  Element locations will be defined in non-dimensional element lengths.  Element print flags will be in this second table.  At startup time, AeroDyn will interpolate airfoil data tables to be located at the centers of the analysis elements.  Airfoil data will be interpolated for the union of all alphas and Reynolds numbers.  The interpolated data will be added to the end of the optional <rootname>.opt file for inspection.
• Replace AeroDyn's BEM algorithm with WT_Perf's algorithm.  Use the no-swirl closed-form solution for BEM for the initial guess.
• Eliminate RENUM.  Add "EOF" line to end of airfoil files to catch error of having multiple tables, but only telling the code there is one.
• Streamline the interface with the structural codes so that modules are no longer used to pass information.  Modify FAST, ADAMS2AD, and YawDyn to use the new interface.
• Loop through all elements of all blades at each time step; but on the first element call, send AeroDyn all of the element information, let AeroDyn do its calculations, and return that element's loads; on each additional element's call, simply return the computed load.
• Use non-dimensional DRNodes instead of blade stations to define the blade properties.
• Gaussian quadrature within an element.
• Convert routines to DLL form.
• Modify how AeroDyn incorporates the effects of blade/turbine velocity on the induced velocity calculation. That is, high frequency vibrations/turbulence should have little effect on the wake, whereas low frequency vibrations/turbulence should have a direct effect on the wake.
• Add feature to allow FAST/Linear to use the frozen wake assumption (i.e., maintain fixed induction while perturbing states). Do this by allowing the structural code (i.e., FAST) to pass AeroDyn a flag which indicates whether or not to compute new induced velocities or keep the old induced velocities.
• Make code more modular so that new aerodynamic modules can be incorporated.
• Add aerodynamic tower loading and the ability to specify variably-spaced tower nodes/elements.
• Add nacelle aerodynamics.
• Add hub aerodynamics.
• Add additional calculation for skewed wake correction based on local induced velocity (for each annulus) as opposed to rotor averaged induced velocity.  Should we eliminate skewed-wake corrections for BEM?
• Allow CLCD() to be called without knowledge of specific blade element--so that the tail fin can use CLCD() without a workaround.
• With rotor-furl, the rotor plane can translate in any direction due to yaw rates and rotor-furl rates; thus, we should really specify hub velocities in the plane of the disk in two directions (horizontal and vertical) and add a hub velocity normal to the disk.  The only velocity currently taken into account is the horizontal component within the disk (through variable HubVDue2Yaw).
• Allow users to be able to change the default value of various constants, like the dynamic stall routine constants.
• Allow the structural codes to get the output channel information that is sent to the AeroDyn output file so that AeroDyn's output channels can be included with the output channels in the output file of the structural codes.
• Update user's guide.  Document all undocumented features.  Add guidance for option selection.

Long-Range Plans:

• Make the inertial reference frame origin unrelated to the undeflected tower centerline, since this is meaningless when platform DOFs are around.
• Add new tip loss model from Shen, Mikkelsen, Sorensen, and Bak (Wind Energy).
• Add the ability to couple dynamic stall and the generalized dynamic wake, although we will have to explore how stable this is when implemented.
• Add rotational augmentation correction (Selig/Du for Cl, Eggers for Cd?).
• Incorporate Morten Hansen's (from Riso) modified Leishman-Beddoes dynamic stall model that can be used during linearization analyses, which is described in: http://www.risoe.dk/rispubl/VEA/ris-r-1354.htm.
• Advanced vortex methods.

This page was last updated by Marshall Buhl on 5-January-2006.