pyChrono: ANCF hollow cylinder

[RBergua]

Hi all,

I’m trying to model a wind turbine tower using hollow cylindrical ANCF beams.

Looking at the beamsANCF.py example (https://github.com/projectchrono/pychrono-examples/blob/main/fea/misc/beamsANCF.py), it appears that the sample uses rectangular beam sections via:
element.SetDimensions(dx, thickness, width)

I would like to use:
element = fea.ChElementBeamANCF_3243()

Is there a straightforward way to define a hollow circular (tubular) cross-section for ANCF beams?

For comparison, when using Euler–Bernoulli beams I can define the cross-sectional properties directly:

beam_section = fea.ChBeamSectionEulerAdvanced()
Area = math.pi*((D_o/2)**2-(D_i/2)**2)      # Area [m^2]
I = (math.pi/4)*((D_o/2)**4-(D_i/2)**4)     # Second moment of area [m^4]
J = (math.pi/2)*((D_o/2)**4-(D_i/2)**4)     # Polar moment of inertia J = Iyy + Izz
beam_section.SetArea(Area)
beam_section.SetIyy(I)
beam_section.SetIzz(I)
beam_section.SetJ(J)
builder = fea.ChBuilderBeamEuler()
builder.BuildBeam(mesh,
                  beam_section,
                  5, # Elements per section...

Is there an equivalent approach for ANCF beams, or a recommended way to model hollow cylindrical sections?

For reference, I use pyChrono v9.0.1.

Thanks for the support!

[AAC8020]

Hi RBergua, I have the same question. Have you found a way to do this?

[RBergua]

Hi @AAC8020,

Unfortunately, I did not receive any feedback regarding the ANCF.

I was able to define hollow Euler-Bernouilli beams defining the cross-sectional properties as follows:

beam_section = fea.ChBeamSectionEulerAdvanced()

Area = math.pi*((D_o[i]/2)**2-(D_i[i]/2)**2)      # Area [m^2]
I = (math.pi/4)*((D_o[i]/2)**4-(D_i[i]/2)**4)     # Second moment of area [m^4]
J = (math.pi/2)*((D_o[i]/2)**4-(D_i[i]/2)**4)     # Polar moment of inertia J = Iyy + Izz

beam_section.SetArea(Area)
beam_section.SetIyy(I)
beam_section.SetIzz(I)
beam_section.SetJ(J)

I also tried to do something similar with the Timoshenko beams, but I encountered quite some issues with the discretization. See for reference: https://groups.google.com/g/projectchrono/c/ztqTkJUT_0g

Btw, the Chrono forum seems more active than the GitHub repo.

I hope this helps!

[AAC8020]

Thanks, that helps a lot😁