chrono software calculation accuracy problems
FSI-in opened this issue · 4 comments
Dear all,
I am using QBlade software for aeroelastic simulation of wind turbines and I learned that QBlade uses chrono software for its structural calculation module. Until then I have been using openfast to calculate aeroelasticity and there is a module in openfast dedicated to calculate structural deformation, it is BeamDyn, which uses geometrically accurate beam model. I learned that chrono software uses what seems to be the Euler Bernoulli beam model. So I would like to know which of these two modules is more accurate.
Best regards,
It is a good question, but not an issue of code development.
And it has recovered to an open question again when the blade of wind turbine becomes extremely long and slender, up to >100m.
In theory and in general, the accuracy of different beam models is: GEB > Corotational Timoshenko beam > Corotational Euler. But it also depends on how you implement the codes.
The Corotational Euler beam element used in QBlade seems work well for some 'strong' blades. It is suggested to refer to the papers from QBlade team.
Hi @pchaoWT ,
Thank you for your answer, you mean that OpenFAST's BeamDyn is more suitable for structural calculations of large wind turbine blades. However, in QBlade software the aerodynamic part can be calculated using the lift line free vortex wake method, which is more accurate than the BEM method in OpenFAST. We can only say that each of them has its own advantages. I would like to know if chrono module has geometrically accurate beam model to provide calculations? It would be great if it does.
Best regards,
I didn't look into the source code of QBlade, but in Chrono we have IGA beam element which is a kind of GEB. Thus in principal, you have a chance to extend QBlade to involve the feature of GEB(IGA).
Regarding the accuracy, well, accuracy is always expensive! GEB is expected to be slower than the corotational formulation. The details in the code implementation would also affect the final accuracy of the beam element, for instance, the integration method to obtain the tangent stiffness matrix, the model to deal with the offset of mass, elastic, shear centers, moreover, whether the 6*6 fully-populated-stiffness matrix is considered, whether the local buckling is considered, etc.
Two years ago, I also tried to search for the most accurate blade model, but now I would prefer a solution with a good trade-off between accuracy and computational efficiency and numerical robustness. Simpler is better, sometimes.
Hi @pchaoWT
Yes, I am thinking like you did two years ago, hoping to find the perfect model, and thank you for giving me such a valuable idea, in fact the structure is very slow in OpenFAST using the BeamDyn module and sometimes it scatters, making me reduce the time step to prevent it from scattering. Especially when calculating 15MW wind turbines, it is even slower. I think I should choose the right tool for this purpose. Thank you for your advice and have a good day at work. Good night and good dreams.