Is there any way of analyzing a welded assembly with COSMOS express? It disappears from my pull down menu in an assy.....
Cosmos express/solidworks Q
- Thread starter MMike
- Start date
Go ask this question over on www.fsae.com at their forums. I'm pretty sure you can do an fea on the piece as an assembly... At least we have been able to do that with the software here at wsu on our formula cars' frame
I don't think you can but, if you looking to analze the welds even the full blown cosmos isn't much help. The most accurate way that I know of is to calculate it the old fashion way. Look up your welding rod properties and calculate your weld area, thats what I've done in the past.
Well I went old-school on its ass and did it like I would have in CATIA v4.... I just redid the assy as a single part. (It's a simple thingy, so no big deal).
Worked well too......
Worked well too......
probably cut down on processor time anyways.MMike said:
Have you been fooling around with any surface modelling? i'm only midly impressed so far
No...and to be honest I've never really been a surface guy. I got the CATIA surface training at Pratt back in 96 or 97, but that's been the extent of it. I've never had a job where I did any surfacing. At Boeing I was 99.99999% E3D wire routing, and Pratt was pretty basic stuff too. So I'm kind of a surfacing void......zedro said:
Right now, I need to get good with the welded assy's made with the structural steel shapes...... (pretty basic)
I've never been able to get any useful information out of cosmos express, the one time I really needed FEA I couldn't get cosmos to work with the assembly so like you I had to rebuild the whole thing as a single part, unfortunately for me I had several complex pieces.trialsmasta said:
geez, i'm surprised Catia actually has a leg up in that respect.Kornphlake said:I've never been able to get any useful information out of cosmos express, the one time I really needed FEA I couldn't get cosmos to work with the assembly so like you I had to rebuild the whole thing as a single part, unfortunately for me I had several complex pieces.
Cosmos express is nothing more than a teaser to get users to go for Structural Research & Analysis Corp's full blown products. It's fairly useless.
Well we were just doing some "tests".... to see how close the results are. The stuff we're curious about, has never really been analyzed. Ya eye ball it....looks good? Ok make it a little bigger, and call it good.
So I did a quick analysis of our part and it looks like I can remove a gusset from the part. (Two gussets instead of three). But now we're like..."Do we trust this?". So I did a simple cantilever beam and calculated by hand. We got a 3% difference between what COSMOS says, and what we calculated.
For what we do, I THINK it's not too bad.
So I did a quick analysis of our part and it looks like I can remove a gusset from the part. (Two gussets instead of three). But now we're like..."Do we trust this?". So I did a simple cantilever beam and calculated by hand. We got a 3% difference between what COSMOS says, and what we calculated.
For what we do, I THINK it's not too bad.
How are you guys rebuilding as a single part? Your assembles must be pretty simple because any time I've tried to cheat and save an assembly as a part file it only comes in a surfaces and nothing useful for FEA.
They are quite simple. Just flat steel plates welded together.trialsmasta said:
But now I was just checking another assy. COSMOS gives me a safety factor of 1.25. But the other engineer who designed the thing calculated a safety factor of almost 2. And his reasoning seems sound to me. and his conditions we worse. So now I'm skeptical of COSMOS xpress....
i dont exactly know what your calculating, but even different hand calculation techniques have their own accuracies, some which are very much on the conservative side, and others where you really miss the boat.MMike said:
This is all it is. Well....there's a little more to it, but I've dumbed it down a little for my purposes. BUt the big plate is fixed and there is a vertical load on the horizontal plate with the gussets.
Attachments
But I'm not building a boatzedro said:
i really doubt you could get the hand calcs close to the FEA with that part.MMike said:
and why such a low safety factor?
Well we've cleared up the confusion... The other guy wasn't taking something into account and I was jsut blindly taking his word for it. IT looks like the COSMOS model is quite valid..... We've actually managed to improve our little doodad somewhat......
Gotta love technology...
Speaking of, I gots a new computer for work sitting on the floor beside me. Gonna hook it up tomorrow.....
Gotta love technology...
Speaking of, I gots a new computer for work sitting on the floor beside me. Gonna hook it up tomorrow.....
I don't see solidworks as being a real industrial strength engineering tool, but more of a mechanical fit and form design tool, which it does well. Still the FEA capabilities are pretty lackluster in my opinion.zedro said:
Have you seen the "improvements" for the '05 version? I think I saw one useful improvement, the rest of the stuff was features that nobody would ever use, mostly giving sheetmetal functions to solids, but without showing where material would be stretched, compressed, or otherwise misshapen in real life. Some programmer thought it would be a cool function and spent half a year writting the code before ever asking a user if the function would actually be usefull. Supposedly the '05 version of cosmos is more powerful but for whatever reason they didn't give a demo. Don't get me wrong compared to other 3d CAD programs I've used Solidworks is much more comprehensive and intuitive, but some of the features are either too buggy to be useful or just straight useless.
I would tend to agree with you. SW does fine for most of our design stuff with the exception of some radius issues but doesn't do well for some of our actual production. We end up exporting iges files to use for creating flat patterns, complex contour form blocks with spring back, etc. ProE Wildwire is the tool of choice for the more complex parts and for FEA we tried Cosmos but ended up with ProMechanica.Kornphlake said:
On the other hand, I haven't had any trouble with SW and Cosmos when working on my side projects from home.
Solidworks wasn't intended to be an FEA tool. It's a modeling tool, and a good one at that. You can download a working demo of Cosmos designstar. It's pretty fun to play with and can open SW files directly. As for Cosmosexpress, I never found it very useful. However, I'm not a real engineer.
After trying to draw surfaces with Mastercam today, I have a new appriciation for SW. I can't wait for Mastercam X!
After trying to draw surfaces with Mastercam today, I have a new appriciation for SW. I can't wait for Mastercam X!
I know I might get lynched for this, but I prefer Autodesk Inventor. It has a real file management system and I don;t have to deal with that sucky text box when I want to write an equation. And for the love of god when will they let you put negative numbers in SW? And why the F*** don't my model dimensions show up where I want them?
trialsmasta said:
You will get lynched...I hate every product Autodesk makes. We are currently fighting a part that was modeled in Inventor. Because the geo is crap, the parts are coming out all F'ed up.
Neg numbers in SW? I do it all the time.
I can't do negatives in SW either...
I hate Autodesk.
But ok...as long as we're sharing, I've still not yet figured out how to REALLY design in SW (and Mechanical Desktop...similar philosphy). I haven't learned how to do "top down" modeling. Right now, I have to know exactly what my part is....almost fully defined...before I start. I mean, you have to your parts done in order to build your assy.
In the "old days", with CATIA v4, you could overlay other models, from other pople working on different aspects of the plane or engine, and then you just drew on top of them. Then you would break you indivdual part out from there. The down side of that is that the assy did not update with a change to the part...or vice versa. And it makes for big, heavy models.
I haven't had a chance to play with the "Large assy mode" and all that yet....
that may answer some of my questions....
I hate Autodesk.
But ok...as long as we're sharing, I've still not yet figured out how to REALLY design in SW (and Mechanical Desktop...similar philosphy). I haven't learned how to do "top down" modeling. Right now, I have to know exactly what my part is....almost fully defined...before I start. I mean, you have to your parts done in order to build your assy.
In the "old days", with CATIA v4, you could overlay other models, from other pople working on different aspects of the plane or engine, and then you just drew on top of them. Then you would break you indivdual part out from there. The down side of that is that the assy did not update with a change to the part...or vice versa. And it makes for big, heavy models.
I haven't had a chance to play with the "Large assy mode" and all that yet....
that may answer some of my questions....
i'm still here....unemployed....MMike said:
and up early?zedro said:
yeah, it's the new me, quit smoking, 3 meals a day, early to bed and early to rise....my life rulesMMike said:
jogging?zedro said:
bahahhahahhahaaa(cough)hahaha......good oneMMike said:
I love Solidworks, but I cant stand the things you mentioned actually. But I spend probably 60% of my awake life working on Solidworks so I have had to cope. Im going to pass your gripes onto the local rep, maybe they can revise/trialsmasta said:
Mike, give me a call, I have some methods for you
Dave
MMike said:
Could that be because you are a native SW user? The only guy that dosn't cuss like the rest of use heathen engineers is a native SW user and he dosen't know what hes missing. I hear that sad song from SW all the time "that will be in service pack X or rev X" A perfect example that SW has there heads up there ass. V2004 Hmmm view normal to plane...wrong side, click it once more, right side. Damn this SW thing is kick ass. V2005 they take it out WTF!!!! And do you think anything from my bitch list got upgraded in 2005 :nuts:. SW isn;t all bad, but I've had too many bad experiances to rave about it.dw said:
its easy, just think of how the part would be created in real life. Also i may make a 'primary' sketch or blueprint of the piece and reference elements off it later. This way you can make changes directly from the primary sketch only and not necessarily off of seperate secondary sketch elements.MMike said:
For example with a full sus bike design, the entire skeleton assembly of the bike may be created in the primary, essentially the elementary bike design blueprint (headtube/bb/seatube points, linkage vectors, etc) and all part creation is initially referenced from it. So changing the geometry or linkage progression or layout is all done from a single sketch and everything auto regenerated. It may seem like more work, but it keeps everything organized and in control. Usually.
Yeah I did more or less that in mechanical desktop.... To be honest I've not really done any heavy duty design at this job. ....after all I'm DIRECTOR of engineering (cough cough...)... But any of the stuff I've done has just been little bits and pieces that the shop has needed to get a truck out the door. The most complicated thing I've done is a set of steps to hang off the back of a truck. Not exactly rocket science...
I'm by far the most experienced with 3D modeling in the company...(not hard to do). So really what I did was learned Mechanical Desktop, then taught the "real" design engineer how to use it, and now he's doing his stuff with it.
However we're essnetially duplicating the work. Because he's doing all his conceptual stuff in 2D still... He's more comfortable doing that. Then once he's defined it pretty well, then he starts modeling. And really, whatever works right? Out stuff is not terribly complicated. Thing goes up, thing goes down. But I'd be happier if he could do his concept work in 3D. and he is getting there....and remarkably quickly I must say...
But MDT is just so archaic...it's frustrating when you KNOW that better tools exist out there...
I'm by far the most experienced with 3D modeling in the company...(not hard to do). So really what I did was learned Mechanical Desktop, then taught the "real" design engineer how to use it, and now he's doing his stuff with it.
However we're essnetially duplicating the work. Because he's doing all his conceptual stuff in 2D still... He's more comfortable doing that. Then once he's defined it pretty well, then he starts modeling. And really, whatever works right? Out stuff is not terribly complicated. Thing goes up, thing goes down. But I'd be happier if he could do his concept work in 3D. and he is getting there....and remarkably quickly I must say...
But MDT is just so archaic...it's frustrating when you KNOW that better tools exist out there...
zedro said: