Forum
Important Notice for New User Registrations
To combat an increasing number of spam and bot registrations, we now manually approve all new user registrations. While this may cause a delay until your account is approved, this step is essential to ensure the quality and security of this forum.
To help us verify your registration as legitimate, please use a clear name as user name or an official email address (such as a work, university, or similar address). If you’re concerned that we may not recognize your registration as non-spam, feel free to email us at with a request to approve your username.
F Function
Hi Dr. Marten,
When extrapolating a Re150k NACA0018 airfoil polar to 360° via the Montgomerie model, I get the results as shown in the images. Figure 1 shows the F function dipping to zero in the center, which is unphysical because it would mean that the flow is detached at that angle-of-attack (α). Uploading an experimental polar* also presents the same issue. An idea is that it wouldn’t matter if F is zero in the center, because at that α the lift is close to zero, but it’s difficult to confirm this from my side. Figure 2 presents additional data for you to look over. Figure 3 presents the F function of an extrapolated Re3k NACA0018 airfoil via the Viterna model. Regardless of the Re, I keep getting these shapes. I’ve looked over your SANDIA example file and it seems that your polar F functions are very nice. However, I’ve tried all the options to obtain what you have thus far for a few hours but no success.
Furthermore, the QBlade documentation describes the polar decomposition requiring the dynamic polar set option, but I don’t see/understand how it’s related to polar extrapolation and decomposition. I’m unsure what ‘state’ is (degrees of α, state number, etc).
Additionally, the F function should be a value of 1 at the reverse linear regimes of: > ± 170° to ensure attached reverse flow. Yet there doesn’t seem to be an option to adjust these such as provided in the previous version of QBlade
Thank you for your time, hope we get to resolve these issues so other users in the future can benefit from this thread
*An Experimental and Numerical Assessment of Airfoil Polars for use in Darrieus Wind Turbines — Part II- Post Stall Data Extrapolation Methods, Bianchini et al., 2015
Additional Info: 1. Slope of 0.110 is selected because it corresponds to the thin airfoil theory slope of 2π.
2. A± and B± were tuned such that the lift, attached lift, separated lift, and drag polars are as smooth as possible
3. St± were also tuned to make the aforementioned polars as smooth as possible, but changing them to realistic values (±15°) also presents the same issue of the F function dip in the center
- You need to login to have access to uploads.
Hi,
I just checked the section in the Documentation and it is misleading. One of our writers mixed things up. The polar decomposition does not have anything to do with the “dynamic polar set” option.
In general, the polar decomposition functionality is, similar to the extrapolation functionality, a helper tool rather than a method. It cant be very well automated due to the large difference in data values, resolution and quality that polars can have. Meaning that it always requires user intervention and checks. To get good results you need to “tune” or “play around” with all the available parameters in the dialog such as CD90, slope, range, etc…
In detail the decomposition in QBlade is carried out as explained in the DTU publication “ATEFlap Aerodynamic Model, a dynamic stall model including the effects of trailing edge flap deflection”. If for a certain polar the decompsition is not producing good results you always have the option to export the pola, then modify the data and reimport it again – to get rid of discontinuities and such.
On another note on the ATEFlap model: From a few discussion among colleagues in the research community I heard that the ATEFlap DS model is not really valid for Vertical Axis Wind turbines with large AoA changes – such as LE flying first when operating at 180° AoA. For VAWT I recommend not to use the ATEFlap model but the Gormont-Berg model – and this model, similar to the Oye model, does not require decomposed polars but works solely on the lift, draf, moment coefficients . This should also address the point on the linear reverse regime that you mentioned: if the model is not valid there it also is not required that the decomposition works in this area…
Best,
David
Alright, allow me to try this again with what you’ve suggested. I’ll get back to you on this thread after attempting it again
Hi Dr. Marten,
Coming back to this after a few days, I am still attempting to model the self-starting process of the Rainbird VAWT [1]. Thus, I need the virtual camber to change gradually. However, because virtual camber transformation can only be applied to a single airfoil, I am attempting to do this:
Define a single uncambered NACA0018 airfoil, with two polars
Polar1: Re3k extrapolated in QBlade via the Montgomerie model
Polar2 [2]: Import experimental polar at Re150k virtually transformed at approximately a similar c/R ratio
Only two polars are input so that I can interpolate and “include” all the effects of the virtual camber as the rotor ranges from 3k <Re < 150k; it is a workaround the issue of only being able to apply one virtual camber transformation per airfoil. For example, by interpolating between the two ranges, a virtual camber effect might at a Re of 75k (Corresponding to a tip speed ratio of roughly 2) may be achieved approximately.
The Re3k polar is extrapolated well in QBlade, however, when I try to upload the experimental Re150k polar, it seems Cm is not recalculated though upon import a prompt from QBlade mentions that it will. The original paper does not supply the Cm data. Since I’m trying to use the Gormont-Berg model as you recommended I need the Cm data, but as shown in the figure it is just a straight line intersecting with zero on the y-axis. However, running a simple steady state simulation seems fine, though I have not validated it yet. Could it be that Cm is actually recalculated, just not displayed properly?
[1] The Aerodynamic Development of a Vertical Axis Wind Turbine (Rainbird, MSc Thesis, University of Durham, 2007)
[2] An Experimental and Numerical Assessment of Airfoil Polars for use in Darrieus Wind Turbines — Part II- Post Stall Data Extrapolation Methods, Bianchini et al., 2015
- You need to login to have access to uploads.
Hello,
regarding the Cm recalculation functionality: The Cm coefficient is always defined for a reference point (such as 1/2 chord). The recalculation allows the user to reproject the Cm coefficient to the quarter chord (1/4 chord) point – since this is the reference point that is used in QBlade. If no Cm data is present in the imported polar it also cannot be reprojected (or recalculated) since the information about Cm cannot be obtained from Cl and Cd alone.
On the topic of the VC transformation. Its true that this is only based on a fixed TSR – usually one should use the optimum TSR. An alternative to using the VC correction is to use the “2 Point Lift Drag Evaluation” option in the turbine definition dialog. This correction includes the effect of the “pitch rate”, that VAWT experience due to their rotation kinematics, with the simulation through a slight shift in experienced AoA – in a similar way than the VS transformation. See this paper for more details:
- Li, A. and Gaunaa, M. and Pirrung, G. R. and Meyer Forsting, A. and Horcas, S. G., 2022, “How should the lift and drag forces be calculated from 2-D airfoil data for dihedral or coned wind turbine blades?”, Wind Energy Science, DOI:10.5194/wes-7-1341-2022
Best,
David
Quote from 23. August 2022, 17:54
regarding the Cm recalculation functionality: The Cm coefficient is always defined for a reference point (such as 1/2 chord). The recalculation allows the user to reproject the Cm coefficient to the quarter chord (1/4 chord) point – since this is the reference point that is used in QBlade. If no Cm data is present in the imported polar it also cannot be reprojected (or recalculated) since the information about Cm cannot be obtained from Cl and Cd alone.
I’ve just looked over the basic Gormont-Berg dynamic stall model equations [1] and it seems that it does not explicitly require Cm. Just to be sure, does this finding mean that if I use the imported experimental polar with no Cm it will still be the correct way to use this dynamic stall model?
Quote from 23. August 2022, 17:54
On the topic of the VC transformation. Its true that this is only based on a fixed TSR – usually one should use the optimum TSR. An alternative to using the VC correction is to use the “2 Point Lift Drag Evaluation” option in the turbine definition dialog. This correction includes the effect of the “pitch rate”, that VAWT experience due to their rotation kinematics, with the simulation through a slight shift in experienced AoA – in a similar way than the VS transformation. See this paper for more details:
- Li, A. and Gaunaa, M. and Pirrung, G. R. and Meyer Forsting, A. and Horcas, S. G., 2022, “How should the lift and drag forces be calculated from 2-D airfoil data for dihedral or coned wind turbine blades?”, Wind Energy Science, DOI:10.5194/wes-7-1341-2022
Regarding this model, I’ve looked at the paper and its application on a lifting line model is quite good for straight blades with no cone angle (exactly as I’m simulating). Just to be clear, implementing this model implies that I should remove any virtual camber polars and replace them with the untransformed polar, yes? Otherwise the effects of 2 Point Lift Drag Evaluation (2PLDE) and VC will compound
Hi,
in the DS models Cl, Cd and Cm are being handled independently. Thi means that whether or not you have a nonzero Cm coefficient does not affect the Cl and Cd coefficients. If Cm is zero it will also be zero after the DS model is applied.
Regarding the 2 point lift/drag evaluation you are correct and should not use this in combination with a VC airfoil treatment.
Best,
David
Quote on 24. August 2022, 17:36in the DS models Cl, Cd and Cm are being handled independently. Thi means that whether or not you have a nonzero Cm coefficient does not affect the Cl and Cd coefficients. If Cm is zero it will also be zero after the DS model is applied.
Alright. I’m curious to what extent does a zero Cm affect the solution accuracy? Especially considering the NLLFVW algorithm is strongly correlated with the input polars, and many publications do not report Cm. In that regard, is it not advisable to apply a polar with a zero Cm in conjunction with the Gormont-Berg dynamic stall model?
Hi,
the moment coefficient Cm is only really important when you also consider a structural model – as the moment acting around the blade axis can lead to a twisting of the blades. If you only look at purely aerodynamic rotor performance Cm does not play an important role.
Best,
David
