statistical modeling joints in adams/insight

[istoria]

Good morning!
I am using adams/insight software for the thesis and should do a study on how cinematic outputs vary if statistically vary the position and orientation of the joints (specifically, position and orientation of a cylindrical joint).
in adams/view I can define the variables by going into the "move" of the bond and parameterizing the coordinates; I carry everything in adams/insight, add these variables to the factors and launch the experiment. adams/insight leaves and does all the simulations. . .simplely by not changing the parameters, i.e. even if in the doe matrix it shows me all the variables that take different values when then simulates the system always has the same values as these variables (equal to the initial values). ..I would be very grateful to those who know how to give me some indication on how to solve the problem...also because the time tightens!! Thank you.

 

[MacGG]

Hello history,

I don't know if you parameterized the model or not.

or did you create design variables and use these to do the doe?

Hi.

♪

 

[istoria]

Hello, macgg,
I created variables (joint position) and goals, then I exported everything to adams/insight within which I created the experiment. . .

 

[MacGG]

Hello history,

So in the factors, did you promote candidates interested in factors to consider in simulation (i.e. appear in the inclusions branch, with the tie factors command)?

then what kind of simulation do you do?
I suggest you start with a study - discipline, to understand how things go.

Hi.

♪

 

[istoria]

thank you first.. .
So, my variables appear in inclusions but I didn't use the comendo tie factors (which I don't even know what it does...can the error be here??), I just used "promotes".
each variable has a uniform distribution between a minimum and a max...
for what concerns the type of experiment before I tried with variation-latin hipercube, then, since it did not come, with doe screening (2 level). . .

Thank you! ! ! ! !

 

[MacGG]

Hello history,

but in adams/view, if you change the variables (use build -> design_variable -> modify), the model changes?

Hi.

♪

 

[istoria]

Good morning macgg,
I found that in reality the problem is that you can't parameterize markers while the points yes...so to change the position of the joint I have to "attach" a point to the marker of the joint and the coordinates x,y, and z are my variables (which change both if the modifier in adams/view is if I launch the experiment with adams/insight)...so it works but now I have to be able to parameterize the orientation of the right now. .
of course we accept ideas and suggestions. . .
Thank you.
History

 

[MacGG]

Hello history,

Okay, now I know what you were doing.

It is not true that you cannot parameterize markers, but these, contrary to points, do not automatically parameterize.

In fact, to parameterize a marker (but also any other object), you have to create the variable design and associate it with the value you want to vary. so to vary the orientation, you create the angular variables design you need to change l0orientation (it is not said that you have to change all 3 angles, but it could only take you 1 angle and then just create only one variable.

with the points, the variables are already defined and if you use the points to create something else, automatically this is parameterized to you on the point.

to parameterize orientation, then, or use a variable that directly commands the same orietation of the marker, or you can use those orientation functions that allow you to continue using points as objects that also allow you to orient (functions like ori_along_axis, etc.).

use the view help and see: improving your model designs > parameterization basics > .

cation

macgg

 

[istoria]

First of all thank you thank you macgg and excuse if at first my explanation was not clear.
I tell you how I tried to create marker variables:
"move" table of the marker compared to the ground====================================================================================================================================================================================================================================================
In this way I can define variables that however it is as if they were "fittizie" in the sense that when I make them change in reality do not change (even if I can set max and min, distribution in adams insights etc).

As for the use of functions I had also thought about it but I have not very clear how to do and which to use: the idea should be to define a variable that recalls the expression of the function builder right?? However the variable is a climber while almost all orientation functions return me a carrier...