Bending Active Timber strips with Kiwi3D

(Guillaume Caussarieu) #1

Hi,
For a project, we are developing a timber bending active structure and are trying to modelize it through Kiwi3D.
We are aiming at bringing the timber strips into a bent shape and calculate the stresses implied. From this bent geometry, we will further load the elements using the load history capacity of Kiwi.

I have been trying to achieve a simple geometry totry the methodology based on the papers : _Form-finding bending-active structures with temporary ultra-elastic contraction elements) _ and Exploring Software Approaches for the Design and Simulation of Bending Active Systems and this post about bending active with Kiwi.
I am using contracting cables and uplift forces to create the initial geometrical imperfection in order to control the direction of the buckling.
As a result, I get the following bending moments, shear Forces and normal forces. The normal forces look odd but it would seems that the sum of it is actually equal to zero? :
1 2 3
I was wondering if the methodology presented in the definition is the proper way to simulate this behaviour and what should be the criteria for the pre-stress in the cable and for the forces to create the imperfections in order to get a convergence in the calculation?

Bending_Active_Kiwi.gh (35.4 KB)

Thank you for you help!
Guillaume

(Anna) #2

The definition itself is almost fine. You have to use polynomial degree 1 in the refinement of contracting cables:

In order to avoid these unsmooth shear forces, you have to use a higher polynomial degree for the beam as the shear force is the derivative of the moment.
grafik

The oscillating of the normal force can unfortunately not completely be eliminated but it becomes less with more elements ( U in the refinement). This is due to the higher order basis functions.

Furthermore, the system is kinematic. In this case, it works but this may also cause problems in other simulations.

(Guillaume Caussarieu) #3

Thank you for your clear and useful answer.