Spatially varying displacement boundary condition-Analytical field

[VaggelisKas]

Hello community,

I have the following simulation problem.
I want to simulate the motion of a rigid body along another deformable body.
The rigid body moves along a curve that is given by a function.
I need a way to implement this curve to Abaqus.
I insert the function via the analytical field toolset but it does not work.
Plus if I use a rigid body constraint no analytical field can be used because I
get the message that the displacement equals to 0 due to the analytical field.

Any ideas?
Any help would be most appreciated.


Thanks in advance.

 

[Mustaine3]

See my post here

http://www.eng-tips.com/viewthread.cfm?qid=397728

 

[VaggelisKas]

Thank you very much for your answer.
Problem is that this time I can not simulate the curve via many lines because my tool is a lot smaller than the
previous one and I also need the tool to be vertical to the curve at the same time.
This is why I need a new method to apply my spatially varying displacement BC and the tool I need is analytical fields.


Thanks in advance.

 

[Mustaine3]

Sorry, I'm not getting what you want to do. Try to explain it better and maybe with an image.

 

[VaggelisKas]

If I use two different amplitudes to describe the displacement in the two axis, as you have advised, I virtually give many (as much as the lines of the amplitudes) linear displacements along the two axis.
My goal is however, to insert to the program a displacement boundary condition that is not uniformly distributed but rather follows a curve that can not be simulated accurately by many linear displacements.
So, I am now looking into the analytical field toolset to insert the curve's equation right into the the boundary condition.


Again, any kind of help would be highly appreciated.
Thanks in advance.

 

[Mustaine3]

Use Amplitudes of type Smooth Step. It creates a 5th order polynom. See Manual for more details.

A user subroutine DISP or VDISP would also be an option.

 

[VaggelisKas]

I am aware of the smooth step amplitude but in the manual it clearly states that the smooth step definition cannot be used to interpolate smoothly between a set of data points.
Also I was looking for a way to avoid using subroutines but it seems it is unavoidable.

Thank you very much for all your answers.

 

[Mustaine3]

I didn't say that a interpolation is happening.

When you create a table with many time points and use a smooth step amplitude with that, I cannot believe that this is not good enough. What procedure are you using, what is the time period and what's the typical increment size in there?

 

[VaggelisKas]

I created a table with 100 time points for each displacement axis along with a smooth step amplitude and I get the error :
'The ratio of deformation speed to wave speed exceeds 1.0000 in at least one element'
If I use less points the job is working but not at the required curve precision.

My step time period is 0.01 seconds and the maximum size increment is 10^(-6)