Solenne,

Based on the theory behind the ANCF cable element, it should provide good 
results in a pure axial test.  See: "Analysis of Thin Beams and Cables 
Using the Absolute Nodal Co-ordinate Formulation" by Gerstmayr and Shabana 
for the full technical detail on this element (as well as more details on 
bending).

When using ANCF elements, it is important to keep in mind the meaning 
behind the nodal coordinates.  Taking the ANCF cable element in particular, 
the nodal coordinates for each of the two nodes are the position of the 
node as well as the position vector gradient along the local element "x" 
axis at that node (6 coordinates per node).  The position vector gradient 
along the local element "x" axis defines the tangent to the beam axis in 
global coordinates at that node as well as the stretch along the beam axis 
at that node.

When you fix an ANCF node, you fix all of its coordinates.  So for your 
axial test, you do not get a constant state of stress in the meshed beam 
due to this imposed boundary condition and that is why adding elements will 
help you get closer to the analytical solution which assumes that constant 
state of stress.

That being said, I'm not sure if the boundary conditions explain all of the 
difference between the static solution and the analytic solution.  I 
haven't had time to investigate that further, but I wanted to at least pass 
on this information.

Best regards,

Mike

On Thursday, June 23, 2022 at 4:05:06 AM UTC-5 SoMdt wrote:

> Hi all,
> I am simulating a very simple cable element under traction (I also tested 
> pure flexion) : one node is fixed, and a force is applied to the end node.
>
> For the code attached, the elongation of the beam is 9.61 mm, while, 
> analytically, it should be 12.7 mm. I did the same simulation with 
> BeamEuler element, and the result is indeed 12.73 mm.
> With cable ANCF, even with lower force (divided by 10), the error is 
> significant (1.11 mm vs 1.27 mm).
>
> The problem is the same in flexion and the error does not seem to be 
> linear.
>
> When I increase the number of nodes in the builder, the error decreases 
> but it remains significant (10.88 mm vs 12.7 mm with 1000 nodes instead of 
> 1). 
>
> I am using these elements wrong ? I thought they were suitable for large 
> displacement and could replace Beam Euler if no twisting or shear were 
> present ? 
>
> Thanks a lot for the help,
> Solenne
>

-- 
You received this message because you are subscribed to the Google Groups 
"ProjectChrono" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To view this discussion on the web visit 
https://groups.google.com/d/msgid/projectchrono/8f137502-fd3f-4b75-af8d-50f5ab75145an%40googlegroups.com.

Reply via email to