tubes and cosmos analysis

Forget CosmosExpress for something like this. Although the analysis of frames is typically given as a college level programming exercise because beam elements are among the simplest to program, this functionality has been missing from SW from the beginning. Cosmos/M can handle beam elements which makes it far quicker than SW when doing this sort of problem.

TOP

I second the last comment, in fact Cosmos Express is a total POS. It was introduced as a teaser to get you interested in Cosmosworks, a much more expensive POS. Cosmosworks will do what you want but I wouldn't bet the farm on the results it gives. Might be enought though to tell you where the weak spots in the design is.

--

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

----- ---- -- -

TOP is correct, you'd need to use beam elements -- which Xpress does not have.

For a quick/dirty analysis with the tools you do have, you might model your frame with solid rounds -- and not tubes.

The stress results from this would of course be totally meaningless but you may gain some insight from the displacements -- ie where your structure might benefit from gusseting, etc.

Be careful of sharp inside corners -- which stinks because

1) a weldment will have lots of these and 2) thats where your gussets would be. Use some "engineering sense" and try to guess where the critical points are likely to be. Fillet these joints with a radius at least equal to your intended wall thickness.

Try to run as fine a mesh as your machine will handle (don't know how much control Xpress gives you here either).

Finally, you'll have no buckling information -- so keep an eye out where tubes meet at a "T" or where the displacement analysis bends them significantly from compressive forces.

Far from ideal -- but again, may give you some insight.

-Tony

Thanks for all input.

So what I hear is that COSMOSXpress can't handle my structure. =CD have no problem using solid rounds for my model - but as I'm sort of a beginner I'm not sure how much extra work that will mean (and how do I practically do it - sounds like little bit more work than using a simple 3D sketch).

Don't know if COSMOSWorks is available for me, will check.

Any possibility to export the endpoints of the 3D sketch to another program where simple analysis can be carried out? I find it easy to design and change the 3D sketch (and later stage add other elements) in SolidWorks but what I hear is there is no simple way to do any kind of load analysis (not even the very simplest ones).

Tony,

Actually, I hadn't thought of this solution. If Hendrick is after deflections what he might do is this:

1. Calculate the moment of inertia of the tube.
2. Back calculate the diameter of solid bar that has the same moment of inertia.
3. Use SW weldment functionality to build the space frame from solid bars.
4. It just might be possible to run this in CosmosExpress for global deflections and hot spots if Hendrik can get the boundary conditions right. St. Venant and some long thin members might come in handy there. There will be a lot of elements in any event. The stresses won't be right because c/I will be too low. You might be able to factor that up to get the stress in your tube though.

TOP

Top, I'd have to give that some thought but working backwards may get him (her) into the 20-30% range (error, that is). I guess it boils down to just how big his model is and how "approximate" is "approximate".

Hendrik/BMW -- the easiest way to model this is to still use the weldments feature. You can edit the 'profile templates' (don't remember what they're really called). Find the template for the tubing size you want -- open & edit -- delete the inside circle (tube ID) and save it as a new template (Solid_Round_XXmm) in the same directory as the other templates.

When you pull up the weldment feature, you'll have a new profile in there -- the solid round you just saved.

Be sure to clean up the ends where 3 or more members meet up as you may have some sharp corners that Xpress definintely won't like.

-Tony

It Hendrik is doing the space frame then I wouldn't be too concerned about the corners except qualitatively. If it is a vehicular space frame it is going to be plenty strong. The concern will be where the plastic hinges will form when in a crash and in rigidity which is calculated from loads and deflections. If a particular corner looks to be a problem then model just that connection with appropriate loads. Unfortunately pulling those loads off of CosmosXpress will be a problem.

TOP