COSMOS: Simulating Cable on a tower - design exercise

I am not that familiar with Cosmos works. But, had some ideas about simulating cable.

Essentially the idea is to make a series of links connected with pivot joints. In your case, you may be able to get by with only 3 or 4 lengths along the length of the cable. It might be best to make the joints between the segments not pivot joints, to reduce the amount of freedom, just a plane pivot joint would work.

Joe Dunfee

Hi, I asked our FEA gurus what they would do in Nastran and I believe you could also do this. Run the analysis and determine whick rope elements are in compression and which are in tension. Remove the elements in compression from the model and rerun the analysis.

I don't use Cosmos, but some codes have cable (tension only) elements.

Jerry

Depending on the mass of the cable, would it be a good idea to replace the cable with some sort of dead weight to simulate the mass of the cable acting on the tower?

Stephen.

If the cable mass is to make a big difference to the results then I guess you could resolve that by adding the cable mass pulling the tower to the force applied pushing the tower?

--

--------------------------------- --- -- - Posted with NewsLeecher v3.8 Final Web @

Remember to pin joint the cables to the ground and the tower to allow them to rotate when under tension or you will induce bending into the cable elements.

--

--------------------------------- --- -- - Posted with NewsLeecher v3.8 Final Web @

If you don't need to account for the weight of the cables, you could replace them with extension only springs. Make your spring constant equal to the stiffness of the cable. Actually, if you include a preload you should be able to do a reasonable job of simulating the weight of the cables as well.

Obviously the cables will be pretensioned and have a significant weight. They will always be in tension. Cosmos/M can handle this with truss elements which can only take an axial force. You might get a beam element to almost behave that way by giving it a real small moment of inertia. That coupled with a non-linear elastic material definition would certainly come close. I don't know if you can get there in CosmosWorks. I use NE Nastran and this kind of thing would be a piece of cake.

One of the things that is important in tower design is the aero- elastic behavior and none of the above will catch this easily without coupling them with a CFD program.

TOP

We use NE Nastran as well, a really nice relatively cheap (compared to MSC) piece of kit. I especially like the dual dongles for both pre and post processors allowing multiple workstation installs.

--

--------------------------------- --- -- - Posted with NewsLeecher v3.8 Final Web @