Tensioning cables in Karamba

Replies to This Discussion

Permalink Reply by Rafael Pastrana on November 20, 2016 at 7:40pm

Hi,

Have you tried with the "Pretension" in the "Loads" Component? Youo basically have to input an elongation rate and apply it to your structural elements.

Best,

Rafael

Permalink Reply by Merijn de Leur on November 21, 2016 at 2:25am

Hi Rafael,

Thanks for your reply! After I posted this I tried this indeed. It worked for straight forward models like this: 

In this case, I gave the cables a strain of dL[mm]/L[m]. So subtracting original cable length from the deformed cable length, multiplying by 1000 (m to mm) and deviding this by the original length (in m). This gave me a deformation of approximately zero.

Note: I gave the pylon in the top example an infinite stiffness.

However, when I try this in a more complicated model, as mentioned before, I get this:

Here, again, I took the original cable length from the input for the assembly, subtracted it from the deformed cable lengths (derived from the deformed model), multiplied by 1000 (to get to mm from m) and devided it by the original length (in m).

Again, for the sake of the test, the pylon is infinitely stiff, so the deformation in the pylon doesn't influence the cable strain.

It seems to be overcompensating somehow.

Kind regards,

Merijn

Permalink Reply by Karamba3D on November 21, 2016 at 12:54pm

Hi Merijn,

did you consider the fact, that the cables add to the stiffnes of the system? If you prestress one of the cables it influences the state of all others.

In case your system is symmetric with respect to the longitudinal axis in terms of loading, geometry and boundary conditions you should get a symmetric deformation pattern.

Best,

Clemens

Permalink Reply by Merijn de Leur on November 23, 2016 at 2:54am

Ah, of course! Didn't take that into account.

Thanls!

Best,

Merijn

Permalink Reply by David Andersson on November 2, 2017 at 2:49am

Hi Merijn,

I am wondering how you actually modelled the cable elements to begin with? 

Regards,

David

Permalink Reply by Merijn de Leur on November 2, 2017 at 12:14pm

Hi David,

It's a while ago, but I modelled the cables as beams and then modified the elements to have no bending stiffness and not to take buckling into account (which I guess isn't really nescessary).

Side note for those interested, I "cheated" my way past extensive calculations for cable-tensioning by optimising the elongation in each cable with Galapagos towards a minimal deformation.

Permalink Reply by MT on May 15, 2019 at 12:53am

Hello Merijn, 

I know this is from a while ago but I was just wondering how you connected the cable elements to the beams of the deck. Was it using the Joint-Agent component from Karamba? Or is there another way?

Thanks for any help.

Mohamad 

Permalink Reply by Merijn de Leur on May 15, 2019 at 1:04am

Hi Mohamad,

It definitely has been a while ;)

At the time, I connected the cables the regular way you would connect beams, but by using the modify element node as shown above and disregarding bending and buckling, the connection will always be a hinged connection in all directions.

Not sure what you are trying to accomplish, you could achieve the same connection using either the Joint-Agent or the Beam-Joints component, however, for cables it's important to indicate that they don't have any bending resistance. 

Merijn

Permalink Reply by MT on May 15, 2019 at 1:15am

Hello again, 

Thanks for you quick reply :)

I'm trying to connect a shell deck to cables, so it's slightly different (not beam to beam). I will try with joint-agent in the hopes something can work out. 

Thanks for your note on the ModifyElement, it's very helpful in setting up the cables. 

Best, 

Mohamad 

Permalink Reply by Merijn de Leur on May 15, 2019 at 4:01am

https://www.grasshopper3d.com/xn/detail/2985220:Comment:1985991

Alright, that's possible. You will however get peak stresses in your shell element where the cables are connected.

You could fix that by modeling an interface where you introduce the cable tension onto a larger surface using dummy-beams. 

Good luck!