"Smart" initial displacement and velocity

Renga@MFRC

Hi

I am trying to set up a simple simulation as a practice for a larger more complex simulation. I am attempting to move a body downward (-z direction) at a constant velocity, then apply a constant acceleration, So when I use the motion command I need to input the the initial displacement and velocity. I know I could just put in -250 in for the initial displacement but I am trying to have a 'smart system' as this object will be interacting with a powder using EDEM co simulation, and I wont know the initial displacement of the body. I am trying to use DZ function but its not working, when the motion command is used the body jumps back to its starting position, as if I had entered '0' for ic_disp

my script is below, if I enter -250 for ic_disp it work as I want, B_0 is the body being move, and is the only body

<Simulate

analysis_type = 'Transient'

end_time = '5'

print_interval = '0.01'

/>

<Motion_Joint

id = '301001'

joint_id = '301001'

motion_type = 'T'

val_type = 'A'

ic_disp = 'DZ({b_0.cm.idstring},{Global_Frame.idstring})'

ic_vel = '-50'

type = 'EXPRESSION'

expr = '10'

/>

<Simulate

analysis_type = 'Transient'

end_time = '15'

print_interval = '0.01'

/>

Praful

Hello Edward - Here is how I did it (see attached model). A displacement type motion is used and a sensor is defined to capture the value of the displacement at the needed time. The value of that evaluation is used using the SENVAL statement to modify the motion.

In my example, the motion is deactivated initially to let the body fall under gravity, but this is not necessary. The motion can be active if needed.

 

Hope this helps.

 

Regards

Praful

Unable to find an attachment - read this blog

Renga@MFRC

Hi Praful, thanks for your response

I have tried to use this with my simulation, it does not work. I think the issue is that the Initial displacement and initial velocity required for an acceleration motion, can only be numbers, they do not seem to accept functions.

I cannot just use displacements, I need to use accelerations to control the loading in the co-simulation

I have attached my model with the attempt to use sensors

Unable to find an attachment - read this blog

mpv

Hi,

Before finding out a better resolving, you may try to apply the motion to a dummy body and then try to active/deactive a coupler between the two body to conduct sequent simulation. It will avoid the conflict in initial condition of the motion.

 

I attached the model was fixed.

<?xml version="1.0" encoding="UTF-8"?>

Unable to find an attachment - read this blog

Renga@MFRC

Hi

I ended up using forces instead to control the acceleration and used the sensor method outlined by Praful. So that works well. I think the issue that I was having is that the initial displacement and velocity have to be 'real' values as apposed to 'expressions', they cannot have functions in them. 

Chris Coker_21312

This is probably the best method, since you mentioned earlier you would not know the exact position of the body, due to the external influences of the EDEM co-simulation. Relying on enforced displacements would likely always be  problematic, in this case.