MSC ADAMS adm file 1

[josefl9]

Good evening, I am new to this, so please forgive any irregularities in the way i post this.

I have exported an .adm file from an MSC ADAMS .bin model, but for some reason the bearings from the .bin model don't get passed on to the .adm file as bearings, but rather as simple geometries,constrains and bearing forces.

I also tried using the same .adm file on linux, but things were even worse, with the model beeing completely missaligned from the bearrings (as if it were offset).

Does anyone know how to I could fix this, or what might be causing this problem?

Thank you in advnace

 

[GregLocock]

It's not a problem, it's a feature. The adm file is written in the language that the Solver understands, and the Solver doesn't know anything about bearings. It knows about constraints and forces and torques and masses and accelerations and so on. So a bearing will typically be a constraint with all but one degree of freedom locked.

What are you trying to do with the exported file?


Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[josefl9]

Good morning,

I am trying to run an optimization script (using Matlab) to determine the model's contact parameters (virtually using brute force).

I do get similar results between the ADAMS VIEW model (.bin) and the ADAMS SOLVER model (.adm). That beeing said, my main problem is that while the .adm file's bearings (or forces and constrains as you said), are recognized properly when I run the ADAMS SOLVER on a computer with Windows OS, and so the model runs properly, when i try the same thing with the exact same model on a computer with Linux, the bearings are offset, and the model excecutes wrong (everything just kind of vibrates in position instead of rotating, because of the bearing misallignment I presume).

Here is my methodology, in hopes of me explaing my problem better:

1) I initially export an .adm file from the original .bin fine model. This automatically exports and xmt file with my geometry as well.

2) I run the ADAMS SOLVER using a proper .acf script (basically just run for N seconds with M timesteps) from my command line. I have tried doing this on
two different computers, one running on Windows and one on Linux.

3) In the case of the Windos run, the results are acceptable, and by the graphics I can determine that the model runs properly. On the other hand, on
Linux, the model doesn't run properly, with everything just vibrating in place (as I explained above).

My initial guess was that this traslation of the bearings to just constrains and forces as you said was the problem, but based on what you told me, I begin to suspect that the problem is just one of compatibility between the two operationg systems. So I guess that either my .adm or my .xmt file are not beeing "read" properly by the Linux system.

I hope I've explained everything adequeatly,

Have a nice day,

J. Kou

 

[GregLocock]

I used to run adams on Unix, it worked fine. I really have a hard time believing that an adm file cares what the OS is.

We also swap adm files between different computers, and they work just fine.

Silly question, are you using the C++ solver in both windows and linux? If you are using different solvers then the error settings etc may need to be different.

Examination of the msg file, after enabling DEBUG/EPRINT, may help.

Other than that I can only suggest asking MSC support about this, it is too curly for me.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[josefl9]

Yes I've checked to make sure that both run on the C++ solver. I haven't tried checking the DEBUG/EPRINT command though. Guess I'll try that.

Thanks a lot! You have been very helpfull.

Have a nice day,

J. KOU