How reliable a tool is WM2D for using to create a physical working model?

[JohnnyD]

I got to thinking, as WM2D has it's oddities, how much can it be relied upon to produce the goods?

[evgwheel]

Posted: Sun Aug 05, 2007 1:17 am Post subject: re: Proof of doing work
Bessler007
Why do people ask difficult questions? Have you used wmd2.
No, I haven’t, but it started an interesting debate.
I am here to make myself use my grey matter (what is left) and seek other people’s opinion, so I can learn

If you read through some of the post from 2 months ago (above link) you may find that everyone uses wmd2 as a tool, but do not trust it as an indication that it represents the real world EVG

[rlortie]

evgwheel,

you may find that everyone uses wmd2 as a tool

Sorry to say this, but not everyone of us use wm2d! I do not, nor does a reduced ratio of four out of five members that I work with use it.

As I recently stated in a private post, How can you expect an application to answer that which is outside the box of known physics. Even though the answer quite probable will fit in said box once found.

Ralph

[bluesgtr44]

I use WM2D as a tool....and a very good one it is if you keep it basic. Other than that, I would not put much emphasis on what you think works. Even scarier....that you missed because it said..."some items may be beyond the parameters..." or something like that.

Nothing beats a real build....nothing,,,


Steve

[murilo]

Nope... not yet... not yet!

[Bessler007]

Wit the arseistance of werkin model too dim-inshinal I has designed the world's first inertial levitation contraption. Also, I has perfected an intercuntinental balistic cat-o-pelt.

Werkin model is a fantastic product. I highly recommend everyone use it. They didn't pay me to say this either.

[Gregory]

Well, it is not quite reliable when it's compared to the real world, I can admit...
However it still depends on the way you use. It can offer a very good approximation in certain cases, and quickly fail in others.

By the way, the question cannot be answered correctly... :)

[Bessler007]

When I read your quote, evgwheel, it seems you're saying I asked, 'Why do people ask difficult questions? ...etc.' That was actually something you responded with when I asked, 'Have you ever used wm2d?'

Posted: Sun Aug 05, 2007 1:17 am Post subject: re: Proof of doing work
Bessler007
Why do people ask difficult questions? Have you used wmd2.
No, I haven’t, but it started an interesting debate.
I am here to make myself use my grey matter (what is left) and seek other people’s opinion, so I can learn

If you read through some of the post from 2 months ago (above link) you may find that everyone uses wmd2 as a tool, but do not trust it as an indication that it represents the real world EVG

The reason I wanted to know if you ever used wm2d was your statement in this quote below:

Hi
After reading all the post above and some others made me think about that wmd2 thingy.
Lots of members use it and I don’t know much about writing programs but somehow I feel that the program and others like them are written with all the possible references to known laws of physics and that all equations are references to those laws.

[Bessler007]

This assumes the application, wm2d, is inside the box of known physics. It isn't. Additionally it's a buggy program.

As I recently stated in a private post, How can you expect an application to answer that which is outside the box of known physics. Even though the answer quite probable will fit in said box once found.

[Bessler007]

One of the bugs in wm2d is it's inability to know where the system center of mass is. It is impossible to understand energy when you can't accurately track displacement of mass.

I've made very simple models and set pause on 0.1 sec. to let the program calculate the com. This should produce a constant result yet the program returns various results. By simple model I mean just masses not connected nor pinned.

It is a serious bug in the code that causes it to confuse where the system center of mass is.

This is one of several serious flaws in the code I've observed.

[P-Motion]

>>Nothing beats a real build....nothing,,,<<

Couldn't agree with you more.

I use WM2D as a tool....and a very good one it is if you keep it basic. Other than that, I would not put much emphasis on what you think works. Even scarier....that you missed because it said..."some items may be beyond the parameters..." or something like that.

Nothing beats a real build....nothing,,,


Steve

[Bessler007]

I began an elaborate explanation for P-motion about how trig is used in circles when I noticed a discrepancy either with wm2d or the microsoft and Linux calculators. Both calculators agree and both disagree with wm2d's placement of a mass in a unit circle. I looked at several angles. The only agreement I found was with the sin 30 degrees.

wm2d places the mass of a 75 degree rotation c. cw. from the horizon at (0.24897 , 0.96879). The calculators return the cos & sin at (0.25881 , 0.96592).

wm2d's x position is 3.802% below the calculator's value for the cos and y was 0.2971% above the calculator's sin value.

[jim_mich]

Explanation for discrepancy...

The numbers that you posted represent a weight location error of about 0.010257 inches. This is the approximate error for a model of this size when the accuracy is set at the default values. If you want more accurate results then reset the model accuracy. Of course this will slow down the animation as the program must do a lot more computations in order to find the more accurate object locations.

Also note that the center of mass will not be the center of the weight due to the mass of the rectangle.

Also note that you can turn on the center of mass display and it becomes a special object, which can be plotted on a graph.

[Bessler007]

hello Jim,

Thanks for the response.

I didn't think adjusting accuracy would change where wm2d began a simulation but I did make the changes. I changed the animation step from the default of 20 frames/sec to 90,000 frames/sec. (9.0000e+004). There was no change in the coordinates of the yellow mass. I then changed the integrator error from 3.9370e-001 to 3.9370e-003. No effect there either.

The yellow graphs indicate the COM of yellow circular mass(x,y) and the blue graph indicates the rotation the blue rectangle away from 3 o'clock. I didn't plot the system cog (10012).

I think I'm going to analytically solve for some functions to compare with wm to see what's going on.

Do you have any other ideas?