fix move INTEGRATION

Submitted by rob1 on Fri, 12/04/2015 - 17:49

Hello everybody,

I am inserting some atoms in a mesh/tet region and then fixing the body as rigid.
I would like to move this rigid body but, when using fix move, I get the warning:
"One or more atoms are time integrated more than once"
that gives unexpected results on the outputs I am calculating (e.g. forces on a mesh).

fix motion bottom_atoms move linear 0.01 0.0 0.0

THIS WARNING APPEARS EVEN IF I DO NOT CARRY OUT OTHER INTEGRATION (e.g. fix nve on atoms NOT BELONGING to the bottom_atoms group).
Do you have any suggestion? Maybe the erroneous integration is inside fix move?!

Thanks a lot,
rob1

Forums:

rob1 | Thu, 12/10/2015 - 17:51

I have discovered that FIX RIGID already does a time integration, so I cannot use FIX MOVE to move the rigid body.
Is it ok simply to use VELOCITY?

But now I have another problem: since I am pushing some particles with a servo wall towards this rigid body created above, the servo wall never stops its motion, pushing all the particles and the rigid body out of the domain!
Why the rigid body does not stay fixed in the space and is moved by the approaching wall?!

Thank you again,
rob1

ckloss's picture

ckloss | Thu, 12/10/2015 - 22:09

Hi rob1,

fix move and fix rigid are simply not made for each other :-)
Just use one or the other, depending on if the velocity should be fixed or if the body should move freely subject to interaction forces
You can also use one after the other

Christoph