Looking for insight of COSMOSmotion

[MadMango]

Searching through the forums, I can only find 2 threads that discuss COSMOSmotion:
thread559-43937
thread559-40074

I would like to get some user input on the software, as my company is investigating the possible ROI if we get a seat or two. I would also like to hear from Help Desk people (Scott Baugh you readin' this?) or others that might field COSMOSmotion questions and problems that might be able to shed some more realistic light on the subject. My brain has a hard time processing Sale-speak.

What's the learning curve? Is it going to be as easy to pick up and use as SW?

Is it going to be just as dangerous as COSMOSworks, in that it will answer the questions you ask of the software, even if the question you are asking it is wrong?

What can't the software do that you thought it could, or feel that it should? Ray Reynolds
Senior Designer
Read: faq731-376
"Probable impossibilities are to be preferred to improbable possibilities."

 

[SBaugh]

I wish I could help you on this Ray, but my knowledge is only on SW at this time.

They are a Gold partner to SW, that's a plus to me.

Have you had a demo done of it yet? That's what I would do first and Quiz the crap out of the AE and salesman when they come out there. After all that's there job to answer your complex questions.

Also check out the Support page at

http://www.cosmosm.com/

There you might find some answers to your questions. Knowledge base is a good place to look.

I hope that helps...? probably not moree than you already knew ;-)

Best Regards, Scott Baugh, CSWP
3DVision Technologies

http://www.3dvisiontech.com

http://www.3dmca.com

faq731-376
When in doubt, always check the help

 

[EdDanzer]

Although I haven’t used CosmosMotion, I have used Dynamic Designer which is what it was called before SRAC and SolidWorks took over distribution. We did a competitive up grade from Working Model because we couldn’t get correct results.
The problem we had with Working Model is that we could not assign acceleration and deceleration values to motors or actuators. We were modeling high pressure hydraulic cylinders in a grapple, the acceleration values were so high that the contact forces were extreme. We got better results with Dynamic Designer, and the user interface seemed better.
It would help to know what you trying to do to know if it will work. As with all software, garbage in garbage out.

 

[MadMango]

We work with hydraulic and electromechanical devices. We will need to set acceleration and deceleration values to accurately model the assemblies. The simulation capabilities of SW03 are nice for eye-candy, but not for validation of design. Ray Reynolds
Senior Designer
Read: faq731-376
"Probable impossibilities are to be preferred to improbable possibilities."

 

[macPT]

I run Dynamic Designer also, but still have not enough experience. But one thing is for sure: it as nothing to do with simulation on SW. The reason for the lack of experience is That I am focussed with other matters for the time being (I have tryed the tutorials and it seems to be user friendly and very well integrated with SW and CW).

I use simulation in SW only to check the the positions and trajectories of mechanisms and for interference detection. And thats all.

DDM (now CosmosMotion) can do it also, but it does a lot more, as you probably alredy know. You can input acceleration in joints in several ways (constant value, harmonic, user expression and others).

By the way, MadMango, what do you mean when you say "Is it going to be just as dangerous as COSMOSworks, in that it will answer the questions you ask of the software, even if the question you are asking it is wrong?"?

Regards

 

[MadMango]

If someone is inexperienced with FEA, they will apply restraints and forces incorrectly in COSMOSworks, and then solve. These restraints and forces represent the question you are asking the software to answer. It can only answer those questions you ask of it, and if the question is wrong, you will get a correct answer to your incorrect question. Does that make sense?

It's like asking, "Can I invade Iraq?" instead of asking, "May I invade Iraq?" They are both good questions, but only one is going to give you the correct answer you are looking for. Ray Reynolds
Senior Designer
Read: faq731-376
"Probable impossibilities are to be preferred to improbable possibilities."

 

[EdDanzer]

Ray,
Defining the constraints you need to get valid results will take some level of competence. If you can do the basic calculations manually, then using the software will make your life easy, and give great results.
When I say basic calculations, this includes accurately describing the force and time for an actuator or motor. If you use S curve acceleration and deceleration that must be defined as you intend to program it. The simple stresses created from the loads should be in line with manual calculations.
When we first started using CosmosWorks and Working Model we did simple project and did manual calculations to see if we were using the software correctly. We then did tests to validate the results. This worked ok until we did a project where the loads were not defined correctly and the fabrication should have been done with shell elements for parts, this was a $4,000.00 mistake.
We did a design verification for a customer that was having problems with a part of a Cat feller buncher. Our results were modeled after the failure, where the bolts were breaking, and the plate was bending. Unlike Cat we applied the load to break the bolt at the washer area and fixed the pin hole. Our model showed the area around the bolts was too thin, and would yield. If you fixed the washers and applied the load to the pin the design would not yield.

Some things can only be learned by making mistakes, even for large companies with smart people.

 

[macPT]

I agree MadMango. FEA is not for beginners (that's why I am a skeptic about CosmosExpress). But the same problem is in virtually every engineering problem. Even when you apply the classic formula for bending stress in a profile subjected to flexural forces, the result can only be trusted if you have made a good assumption for these forces. I have seen some designs fail, not because of the manual calculations them selves, but for wrong initial assumptions.

Of course this problem is worse in FEA (as in any other numeric simulation software). One of the reasons is that you can loose touch for some critical analysis of the results. That's why an experience designer always validate the model building other models, other constraints, other meshes, compare with other successful designs, make simple manual calculations, check the convergence.

CosmosMotion is less triky than CosmosWorks. This as nothing to do with the quality of the software. It as to do with the physical problems that we are trying to solve, which are easier to manage and criticize in the case of kinematic and dynamic than in the case of strength of materials. But even in this case, wrong assumptions will produce wrong results.

Regards

 

[JNR]

The way I try to explain the pitfalls of FEA, is this. Stress analysi of a simple rod is easy. Once you get to complex shapes it is very difficult. You have to break it up into a "Lego brick" representation of the real model. then you analysis each "simple" brick and iterate the result between all the adjoining bricks. now it should be pretty obvious to anyone who has seen a Lego brick maodel of anything that it is NOT a perfect representation it the real thing. The trick is to get it as close as possible. Great trides have been made in auto-meshing, etc. However it is difficult for the old inorganic logic box to completely replicate what the organic brain can acheive. Thus as you quite rightly say, if I may paraphrase: If your Lego brick model does not closely represent the real thing, you will get a very accurate anaylsis of something does not represent the real thing!! Short of exhuming Merlin, unfortunately we have to live with this for the forseable future.

Dynamics packages on the other hand have a better chance of representing the real world. So they are a bit easier to have confidence in.

I believe the ancestry of CosmosMotion dates back to the original ADAMS and DADS programs, (for which I was once a VAR!! - oh my, am I that old?) so I would be pretty much OK with that choice.

Also the Gold Partener integration (and same ownership under Cosmos-thus-SolidWorks) means a lot. You can make do and wait a rev or so for some special feature, but you can't beat tight integration for headache relief. Integration/conversion problems cause more delay, frustration, cost and wasted effort than anything else.