Non linear Cable analysis

Hi there, I’m currently trying to design cables which aim to support net for golf driving range. This cables are supposed to carry the net, resist to wind load and snow load, and show acceptable displacement when loaded.

I modeled this situation with the Analyse NonLin Component with this setup for the load cases :

LC0 : Gravity

LC1 : Dead load of the net supported by the cable

LC2 : Pretension of the cable (I did that because I anticipate the fact that the displacement of the cable will be important, thus I want to see which amount of pretension I will recommand to apply)

LC3 : Wind Load

For the moment I didn’t modeled the snow because it’s probably way smaller than the wind load.

At the beginning my cable was modeled with a line to truss component (in LineElem) but it appeared that this kind of modelisation did not converge. It eventually converged when I used a Line to beam component, and the convergence is pretty good. I’ve checked the resultant moment in the cable, it is really low so it appears to me that this model is physically correct.

But, because there is a but, it also appeared that the component Initial strain that I used, makes no difference to the final displacement at midspan of the cable, even though I’m moving down the epsilon0 value (which is negative here).

I must confess that I’m quite a newbee to non linear analysis in karamba3D so I might I’ve done an error in my way of modelling this situation so I ask for the help of the community.

ThX

Cable simple_V2_envoi.gh (49.5 KB)

Hi @Xavier7,

in case you want to use cables for carrying transverse loads, those need to be prestressed first - otherwise the system is kinematic.
Here is the example with ThII calculation:

Cable simple_V2_envoi_cp.gh (59.2 KB)

I placed all loads in Lc0 and added an initial NII for stabilization.For small displacements this gives reliable results. In case of largere transverse displacements the deformation has an impact on the normal force which is not captured by small displacement ThII analysis.

Sadly, I could not make the large displacement non-linear calculation converge. This is one of the features which currently get improved but are still work in progress - sorry.

-- Clemens

Thanks Clemens for your answer.

I really appreciate that you take time help me out with this, 2 other questions though.

Understood what you said about Analyse ThII, and it was my first choice but I was not really confident because I couldn’t find some kind of ratio between the displacement returned through the Analyse THII component and the span of the beam you’re working on that makes the hypothesis of small displacement and thus the utilization of Analyse THII reliable. From my point of a view, here it’s not the case that’s why I moved to Non Linear Analysis. What would be the maximum ratio acceptable to use confidently the Analyse ThII ? I can find a value for the Eps0 input in Initial strain component that makes the cable’s displacement within the range of acceptable displacement, but how can I be sure that it is reliable if the hypothesis of small displacement isn’t really ?

When I tried your set up It turned out that it is very sensitive to the value of Eps0 which makes sense, but my second question is, how you set the value of NII in the modify element component and does it have something to do with the Eps0 input or you can put a random value here?

Thanks for your help

Xavier.

Dear @Xavier7,

maybe you find this example helpful: Analyze_ThII_Truss.gh (76.9 KB).It comes with the installer examples under “06_Algorithms\” and demonstrates the application of the ThII analysis on a small example that involves two truss elements. The example also contains the formulas that can be used to verify the numerical results.

The ThII small displacement analysis remains accurate as long as the axial elongation of the elements in their deformed state is relatively small. If that is not the case then the axial forces in the elements get underestimated and the displacments overestimated.

As a verification step, you could compare the element lengths in the initial and deformed states, determine the corresponding elongations, and use them to calculate the additional force increment.

When you use the ThII component any positive value of NII will do, since it gets updated during the second iteration step based on the loads present in the system (for example, frome eps0).

– Clemens