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.
Validation of VAWT with Sandia 34
Hello David,
I’m experimenting with QBlade and tried to get an aerodynamic test LLFVW-model for my further researches.
Unfortunately I got results, where QBlade overshoot the CP-Values in comparison to experimental results for different VAWT-setups.
Then I found your paper “Validation and comparison of a newly developed aeroelasticdesign code for VAWT” and your validation data
for the Sandia 34m VAWT and so I tried to recalculate the power and power coefficient curve to understand the impact of different modelling parameters.
The results of my recalculation matches good to the power curve, but overshoots the cp-values significant.
Do you have any ideas why there is a difference? I tried to set the same parameters as in the paper.
I also tested to change some parameters (dyn. stall models, without structural model, wind inputs) but ended up, that there is nearly no differences in
the results. As startup, I used the Sandia file from the download section.
I uploaded both the results as diagrams (RECALC are my results) and the QBlade-setup file.
Best regards and thanks in advance
Martin
- You need to login to have access to uploads.
Hello Martin,
I had a look at your project file and the issues that I am seeing are:
- The wake is not discretized fine enough, especially for a high rotor loading the evaluated Cp value is very sensitive to this as there is a lot of wake interaction
- The tower shadow model is not “active”. When simulating with a structural model the tower drag coefficient is defined in an additional column in the tower structural data table
- The simulation time is not sufficiently long for the Cp value to “settle” to a final value
- The “2 point lift drag correction” is active, which did not exist back when I wrote the paper
The “old” project files that I used during the validation are not compatible with the current version of QBlade, so I cannot easily find out what exact settings were used back then, but I did a quick simulation with settings that are better suited. You find this project file attached to this post. In this new calculation the Cp at the TSR of 7.5 that you simulated is already slightly below 0.4.
High rotor loading is always quite tricky to simulate, especially for VAWT’s, where there is a lot of “interaction” between the wake and the blades. Furthermore, the SANDIA34 demo project also uses polars with a moment coefficient (Cm) of zero. In reality, the structural twist induced by the blade pitching moment might also slightly change the rotor performance, specially when the blades are as soft as the SANDIA blades.
BR,
David
Uploaded files:- You need to login to have access to uploads.
Hello David,
thank you very much for your help.
I thought the default settings in the wake discretization are a good starting point and after some simulation time the cp value seems “stable”.
Best regards
Martin
Hi Martin,
regarding the wake settings: the required wake length to accurately reproduce the flow around the rotor changes with rotor loading (or TSR). As the TSR increases, a longer wake is required due to the larger induced velocities and the slower convection of wake filaments away from the rotor.
The “default” length in QBlade is set to work with TSR’s up to rated conditions for most rotors, but not beyond that. The main reason is simply the computational cost, that goes up the more wake elements are included in the simulation.
BR,
David
Hello David,
I would like to ask why the angle of attack varies between Linear and Cosine blade interfaces in aerodynamic discretization under the same conditions (H-type vertical axis). The graph shows that the angle of attack is larger for Linear. Which one should I choose?
Best regards
Uploaded files:- You need to login to have access to uploads.
Hello,
the top graph you are looking at shows the AoA at a particular blade position. I would guess that the position you chose is located at the blade tip. The cosine distribution resolves the area around the blade tip much finer, and better captures the steep drop in the spanwise AoA distribution towards the tip, caused by the blade tip vortices. The cosine distribution in this case generates more accurate results.
BR,
David
Hello David,
Thank you very much for your help. I have another question. Is it feasible to simulate a unit by overlaying the H-type and V-type rotors through multiple wind turbine simulations (with the V-type and H-type on the same main shaft but with different azimuth angles)?
Best regards
Hi,
this could be realized using the multi-rotor simulation feature in QBlade-EE, where both rotors (H- and V-type) could be attached to the same shaft and then be simulated in a single simulation.
BR,
David
