10 Replies Latest reply on Feb 15, 2012 7:03 PM by Anthony Botting

    Cables on Tension Structures

    Mauricio Martinez-Saez
      Ho can cables of a tension structure can be modeled and analyzed with COSMOS?

      Cosmos can perform non-linear analysis (tension structures experiment very large displacements), however I do not know how to build a model to analyze the behavior of a cable (with a catenary) working on tension inside the pocket at the edge of a tension structure membrane (the membrane is laminated PVC) when the membrane receive a wind load (PSF). The cable need to show his characteristic stretch which as you know is not just the stretch of the material since it also have mechanical effect as the strands of the cable compress again each other and the cable "rotates" under tension.

      I used before a FEA system that have especific fuctionality for cables (used to analyze suspension bridges) but it is to expensive for what I need know... I believe that COSMOS can help on the analysis but I do not have a clear idea of how to do the model.

      One of the simple structure shapes that I need to analize is the clasical "Hypar" structure. Attached are somo pictures of small structures, what I need to analyze are the cables tensioning the edge of the membranes, etc.

      Any clues?

      Thanks.
        • Cables on Tension Structures
          Use surface mesh for membrane and add "cable" connectors .
          Search for Cable under COSMOSWorks help.

            • Cables on Tension Structures
              Mauricio Martinez-Saez
              Shaodun,

              The problem is not the way to model the membrane (surface or solid, that only afffect the process time and RAM required to create the mesh and anaylis), or to place "cable connectors". The problem is that the "CABLES" are inside pockets at the edge of the membranes and as the membranes receive wind load they transfer load to the cables which creeate a "tension load" on the cables which then "elongate" and therefore change the shape of the "catenary" of the cable which in turn change the value of the axial tension load acting on the cables resulting from the lateral load transferred by the in-plane tension of the membrane.... the problem is on how to model the cable in order to simulate the mechanical strech of a cable under tension (not the strech of the material, but the strech due to the mechanical action of the force on the cable).

              Modeling cable catenaries receiving a lateral load is a very complex problem... as mentioned before there are several systems that do that but they are very expensive and they do not work with SolidWorks. I want to see if I can do the analysis with COSMOS (since COSMOS Advanced Professional can perform non-linear analysis).

              Do someone have done this before?
            • Cables on Tension Structures
              Bill McEachern
              cable tensioned membrane structures are a bit tricky no matter what code you use. Well, I can't say I have done one but at the very least you need non-linear capability as you need to load the cables and the membranes and then you can resist the applied wind loads and what not. I would say this can not be done with any degree of relism in CosmosWorks though you could probably get it done in geostar. Even still, in my experience, if the loads are sufficient to cause any sort of loss of stiffness in the membranes then you would need to be able to predict which ones and then take them out and try it again till you got to the load you need to get to.

              good luck as you will need it. This is a non-trivial problem at least in my view.
                • Cables on Tension Structures
                  Mauricio Martinez-Saez
                  Bill,

                  I appreciate your coments, I see that you have been in this inferno before... yes to model the membrane of a tension membrane structure is a little complicated and tricky... I have been doign that for some years and designed many of them (in fact the structures in the pictures of the first post are examples of some of the small structures I design in the pass).

                  For the design of the PVC membrane I use an specialty Software System (ForTen) which do the form finding, the membrane calculation and the patterning of the PVC pannels required to fabricate the membrane. This program was developed just to calculate and model tensile fabric membrane structures and do a nice job on the modeling and calculation of the membrane. I also have another software (EASY) that was also developen for this type of analysis...

                  The problem is that I will like to use COSMOS to analize the cables, connectors and the compression members of Tensegrity structures as well as the columns and other steel mebers (which I am modeling on SW ans since I have COSMOS Advantage Professional as well as Geostar I what to model those elements (cables) on a way that can be analyzed with COSMOS (using lienear or non-linear as required).

                  Can you give me your ideas on how to model a membrane perimeter cable as decrived on my first a d second post? I see that you know what you about this type os structures (cables, membranes, etc.) and you may give me some good ideas... as you know a steel cable under axial load behave completelly different that a solid rod, since the cable is a mechanism and strech not just by the elongantionj of the steel strands but also due to the mechanical actions inside the cable (untwisting uder load and "penetration" of the strands on the inter- strand air space of the relased cable. There are softwares that have fucntionality to model cables (like Robot) I do not beliebe that cosmos have that build-in fucntinality, however I believe that the analysis of cables can be done in COSMOS if model with some tricks (the more dificult ones will be the membrane perimeter cables).

                  Can you give me some ideas to start playing with this? I know it will be a lot of fun.
                    • Cables on Tension Structures
                      Bill McEachern
                      I would ignore the internal cable effects and just stick with the bulk stiffness. I just can't see them being significant but hey I barely know what I am talking about on this one. I am pretty sure that the cables are likely to be similar to bridge strand with reduced stiffnesses over say solid steel. The untwisting of the cables seems like a second or third order effect and the rotations have got to be small or you would design in mechanism to isolate thsoe effects from the membranes. So either way I doubt very much they matter but hey I really don't know that much about them. In any event you need get a decent first order approximation of the shape in geostar and then sequentially in a non-linear analysis tension up the cables, apply gravity and then apply the loads and hope the thing solves. I would build the idealized geometry in Geostar from scratch and good luck.
                        • Cables on Tension Structures
                          Mauricio Martinez-Saez
                          Bill,

                          The main problem that we what to model is the follwing:

                          Membrane perimeter cables are in catenary and receive lateral load (transfer from the membrane in plane trensions). this forces create axial tension on the cables which is inversely related to the curvature of the catenary, then since cable endsd are fixed to structural members (columns, etc.) they transfer load to them crating a lateral load or moment which bends the column.... what hapen is that as in-plane tension on the membrane increases due to external loads, the cable become longer (due to the combination of strain on the material but also to the mechanical actions of the untwisting and strand penetration into air spaces, this machanical efect of the cables is the cause of the mayority of the increase in lenght). Then as the cable become longer and the ends closer (due to the under-load-bending of the columns) the catenary increases the curvature and the axial tension reduce creating a "spring" efect until the entire "mechanism" reach a state of equlibrium.

                          The trick on this structures is to desing the entire "mechanical efect" of the assembly on a way that the axial tension on the cables (the one creating deformation of "rigid" structural members (columns, etc.) of the structure) do not increase substantialy when the membrane receive external loads (due to the increase in curvature of the cables), in fact the objective is to have that energy partially absorbed by the "mechanical" efect on the complete structure.

                          I do not know if I explain what I mean correctly....

                          In summary, what I need is to analyze the behavior of a cable "beam" with pinned connections at both ends at the top of two columns, with the cable receiving a uniform load over his length and calculate the reactions at the ends as the cable stretch under load and the columns bend, and therefore the cable change lenght and curvature changing axial tension. One of the problems is to define the cable as a material (to define the properties), Maybe we need to do some S/S test of the cable to define the properties, since no manufacturer gives that...

                          We have a system that analyzes cable structures and cable nets, but it uses "internal" data (you just select a cable size and configuration) but the "properties" are hiden...

                          We designed the structures on SW, that is why we will like to analyze it with the cables on COSMOS (to avoid doing the entire model on the other application), the membranes and the cables are analyzed on another software, but there we do not have the structure. Usig remote loads is not a solution since the load will change as the structure deform and the geometry of the cable change.

                    • Re: Cables on Tension Structures
                      Joe Rodriguez

                      I just found this topic. My comment is use another software for Analysis purposes like NDN, amybe ANSYS for example. For design purposes I do not hink it will be complicated with some tricks. I have been working on Tensile Structures since 5 years

                      ago.

                      • Re: Cables on Tension Structures
                        Paul Kellner

                        I believe Cosmos/M has an example problem that solves this problem. Cosmos/M has a tension only shell element and can have tension only axial trusses (used to model the cable).

                        • Re: Cables on Tension Structures
                          Anthony Botting

                          HI Mauricio: I do know there is a cable model (or element type that you can make behave as a cable element) in the COSMOS/M GeoStar Interface (as you know, that comes with your SW Simulation Premium). I did read through and run that tutorial many years ago, so it should still be there (in the tutorials of GeoStar). - Tony