I am first time using airfoil prep and i need to generate dat file to be later used in harp_opt for tidal turbine design. In the “3DStall” worksheetu, what values do we add for RPM i mean average rpm that we will have for our turbine ? Additionally what should i write in " chord" ? Should i put the upper bound chord value in this as we do in harp_opt ? Moreover what is r/R and how should we find this and same goes for Min alpha , max alpha and alpha end value how should we find this for part airfoil
Dear Ghayoor,
Normally we’d use the same rotational augmentation correction for all rotational speeds, so, I would normally set the rotor RPM in AirfoilPrep to the rated speed, but you could also look at the sensitivity to speed. The rotational augmentation (3DStall) model applies to a specific airfoil and station, so, the local chord and local nondimensional radial position (r/R) should be used. Min alpha and max alpha should be set such that the CL slope is well calculated (it is best to use a low alpha and high alpha near the ends of the linear region of the CL-alpha curve). You can likely leave Alpha End at 30 degrees, but you could also look at the sensitivity of this parameter; in the end, the accuracy of the rotational augmentation should be assessed (at least by visual inspection).
Best regards,
As you mentioned " The rotational augmentation (3DStall) model applies to a specific airfoil and station, so, the local chord and local nondimensional radial position (r/R) should be used" what do you mean that the model applied to specific station ? I am using one airfoil for entire blade and how to find local chord of an airfoil or r/R ? I am dividing blade into 10 segments so which r/R should i use and i don’t know how to find local chord
Dear Ghayoor,
The rotational augmentation will depend on the local chord and local radius even if you are using only one airfoil. So, you could use different AirfoilPrep workbooks for each of your 10 stations. It may be a reasonable simplification to use the same airfoil data for every other station, reducing the total number of unique airfoil tables to 5.
Best regards,
Thank you for your help. I understand it now but one more thing that is bothering me is that how will we find local chord at each radial position.Is this not my final objective to find local chord and twist at 10 stations through Harp_Opt so how will i know what is the local chord at each station of airfoil. Is there any formula or something to find that ?
Dear Ghayoor,
I agree that it would not be convenient to use AirfoilPrep within an optimization loop. You may need to simply guess at a reasonable, but fixed, chord if you wish to apply rotational augmentation with AirfoilPrep together with optimization using HARP_Opt.
Otherwise, if you wish to optimize with an automated rotational augmentation, it may be preferred to use the Python-based script AirfoilPrepPy: nwtc.nrel.gov/AirfoilPrepPy (although this still won’t work with HARP_Opt).
Best regards,
I think an ideal solution would be to modify the NWTC Library (or inside the WT_Perf code) to calculate the 3D correction, and then re-compile WT_Perf so that 3D corrections are calculated completely within the “blade element momentum” part of code. This would eliminate the burden of creating separate airfoil files as a “pre-processing” step.
An alternate solution, you could create a matrix of airfoil files with 3D/stall-delay corrections applied on range of tip-speed-ratio, chord/radius, etc (do this as a pre-processing step to running HARP_Opt). If you use a consistent file-naming convention then HARP_Opt can lookup which airfoil file to read depending on the current rotor iteration (as blade geometry is changing over the HARP_Opt iterations). As Jason suggests, it would be easier to use AirfoilPrepPy to create (maybe tens or hundreds) of airfoil files. HARP_Opt would need some further development to read a matrix of airfoil files (benefit of this solution is that it would only require Matlab coding and not touching any Fortran).
Best Regards,
Danny Sale
To address the question: how to find the local chord? … it is up to interpretation but when using HARP_Opt the chord is considered an unknown variable; however, you should be able to estimate the approximate chord length based on your upper/lower bounds … or run HARP_Opt to completion once and use those output chord values → create a new set of airfoil files for multiple blade radius (say only 5 or 10 radius values and use AirfoilPrep to create airfoil files at each radius) → run HARP_Opt again and see if the chord changes significantly … but I would recommend only 1-2 iteration of this method unless you understand the “airfoil correction” sensitivity very well. Preparation of the airfoil polars is a great source of uncertainty in blade-element methods.
Hello Jason,
I am working on implementing adaptive Trailing edge flaps on the NREL 5MW turbine. For this I need airfoil polars for multiple flap angles. As a verification case I obtained the airfoil polars for NACA 64(3)-618 airfoil for a Reynolds number of 6E6 and Ncrit 9 using XFOIL as specified in the report. although after applying 3d corrections using airfoilprep the data is not matching the one supplied with FAST. I used the rated values as input in the 3D stall sheet. Could you verify that the values I have used are correct in the attachment
Thanks in advance.
Dear Kenneth,
Please find attached the AirfoilPrep v2.0 file for the NACA 64 airfoil that was created when the NREL 5-MW baseline turbine model was first made.
AirfoilPrep_v2p0_NACA64_A17.xls.zip (107 KB)
Best regards,
Dear Jason,
Thank you for the sheet you provided with earlier.
I have a couple of doubts regarding what is to be entered in the dynamic stall worksheet.
Firstly, what is to be entered for:- Cn at stall value for negative angle of attack.
and second, how to determine the range of AoA for CN slope calculation.
Regards,
Kenneth
Dear Kenneth,
Regarding (1), this is the value of Cn at stall for negative angles of attack (just to the left of the linear region).
Regarding (2), you should choose a minimum and maximum angle of attack (alpha) that causes a good fit to the linear region–I would suggest choosing alphas at the extremities of the linear region just above negative stall and just below positive stall.
Best regards,
Dear Jason,
Thanks for the quick response. I needed another clarification as to what is the stall angle value to be used. Does it refer to the stall angle from airfoil CFD data or the stall angle viewed in the CN graph.
Regards,
Kenneth
Dear Kenneth,
I’m not really sure I understand your question; are you looking at multiple airfoil datasets? Regardless, the stall angle should be from the Cn graph, whether derived from CFD or something else.
Best regards,
Dear Jason,
I meant to ask whether stall angle is taken from the CN graph or not. Thanks for answering the question.
Regards,
Kenneth