CosmosMotion- wrap force around object, i.e. not straight line 1

[kevinVCU]

Hello everyone,
I've searched google for help on solidworks many times, and I always end up at this site...there is some great help here. This is my first post, hopefully you guys can give me some tips.

I am modeling a shoulder joint in Solidworks w/CosmosMotion. My model includes a humerus (ball joint) and scapula (socket). The muscles that life the arm can be modeled as forces, which attach on the opposite side of the contact area--meaning they have to stretch over the top of the humerus. Some of the other muscles wrap around the sides as well, and I cannot figure out how to model a force other than a straight line!

I found some info on making ropes/cables that may be able to reroute the force, but I figure I'd ask you guys first before I wasted another few days shooting in the dark.

At this point, the forces I do have go from their origin to insertion, straight through the bone, which is obviously anatomically incorrect. I've attached some pictures to help illustrate this...I'd like to get some fresh ideas from some pros, as I just started using SW less than a year ago.

Thanks,
Kevin

 

[TheTick]

There is no such thing as a "force that wraps". You need to create something that conducts force, like a cable.

 

[kevinVCU]

Hi, thanks for the response...once I posted, I realized I screwed up the title, but couldn't figure out how to edit it. (If there's a way, show me bc I'd like to fix it)

Am I correct in assuming that the cable can be created in Solidworks Routing? Once created, it can then be used in Cosmosmotion to transmit force?

Can you point me to a tutorial or provide basic instructions for creating a cable that can bend during my simulation?

Really appreciate the help guys,
Kevin

 

[TheTick]

I have not done anything with cables in C/Motion.

There are ways to combine forces and motion generators. There might be other ways.

 

[CorBlimeyLimey]

kevinVCU,

Use the Inappropriate post? If so, Red Flag it! link (at bottom centre of your first post) to submit a request to Site Management to edit the title.

 

[GregLocock]

I haven't used COSMOS, but have modelled ropes running over things.

Essentially a rope is a series of flexible tension only links, joining a series of massy rigid discs. Then you need to model Contact between the discs and the humerus underneath.

All of this is computationally very expensive, and may not be possible in a cut down mechanisms package. Packages that will definitely allow it are ADAMS and Working Model 3D.

I would estimate that an experienced analyst could get that working in a couple of days. WM3D is probably the best bet as it has a built in rope constraint (ie tension only), and has some very weird but handy properties for 'solid' elements to model elasticity and friction.



Cheers

Greg Locock

SIGlease see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[josephv]

Hi Kevin,

Have you tried modeling a cable as a part, mating it to your assembly and then attaching a linear motor to the cable in COSMOSMotion?

Joseph

 

[kevinVCU]

Joseph, that's an interesting thought that I haven't considered yet. How would you go about making a cable, and will it flex within Motion as the components move?

I tried creating a spline, along which I swept a circle to create a rope like structure. I could not get this structure to bend as the simulation progressed...I did find the rigid/flexible option but that's only for a subassembly.

Thanks for the help,
Kevin

 

[josephv]

Hi Kevin,

As far as I know, you can't get bodies to bend in COSMOSmotion. You can do this is COSMOSworks, however.

Here is an idea, instead of modeling a cable, how about if you use a Spring (you can add a spring in COSMOSmotion, just like you can add a motor or gravity)?

Hope this helps,

Joseph

 

[kevinVCU]

If the bodies can't bend, then I guess the only way is to make a string of smaller bodies that will glide along the surface. I had been using a distance mate to keep these "beads" together, but was running into a problem with them bouncing all over the place and not staying in contact with the surface during motion.

I have tried using springs as well, but same problem, the spring is only defined by a start and end point, with a straight line acting between the two. Additionally, I haven't found a way to make the spring tension only, without instead doing a custom force expression.

 

[josephv]

Hi Kevin,

The "beads" method with its limitations might still be the best solution. You may need to add more mates to simplify the problem.

Just out of curiosity, why do you need it to bend? Are you simulating a tendon or a muscle?

 

[GregLocock]

If you try to use FEA for this you'll need to use a solver that can handle non linearities in material properties, contact, and large displacements.

That seems like a big ask for an add-in to a CAD program.

Cheers

Greg Locock

SIGlease see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[josephv]

Hi Kevin,

Greg makes an excellent point, COSMOSmotion might not be the right tool for this analysis.

I am just trying to understand the problem. Why did you originally choose COSMOSmotion for this analysis? Where you interested in a kinematic analysis (i.e. motion, speed, acceleration)? Why did you not use COSMOSworks and do a large displacement analysis (i.e. non-linear)?

cheers,

Joseph

 

[kevinVCU]

Yes, I am trying to simulate body tissue that wraps around a bone. The muscles should stay in contact with the surface as the arm moves around over a wide range of motion.

I will have to work with the beads some more...though the model works fine now, it's just not anatomically correct.

My master's degree program has me limited to just this program, without any FEA analysis. Again, at this point, I'm just limited to a computational model.

Thanks, I'll try to let everyone know how the bead approach works out.
Kevin

 

[josephv]

Sounds interesting, good luck, Kevin!