Re: [ESPResSo-users] While using rigid body with virtual sites (position not inside a local plaquette dir 0...)

[Owen Hickey]

Hey Lena,

I am not sure exactly what the problem is but the first thing is to
compile with mass and give the non-virtual particle a mass and a
rotational inertia. I think the latter defaults to 1 1 1, but this is
probably far too small. This may also be why the simulation crashes. I
would also try lowering the time steps, which are rather large for an
LB simulation (although this of course depends on other parameters
too). How many steps does it integrate before crashing?
vs_auto_related fixes the position of the virtual particles relative
to the non-virtual central particle. I am not sure what you are
referring to as knots, but perhaps you mean the LB nodes. I have no
idea why it fails to fold the x-coordinate of one of the virtual
particles. having worked with LB and virtual sites there were some
problems when the particles weren't number from 0 to N with no breaks.
I think this is fixed, but it's worth a shot. Can you send us a copy
of your script? That might allow us to solve this issue.

Hope that helps!
Owen

On Mon, Jun 1, 2015 at 12:45 PM, Holst, Lena
<address@hidden> wrote:
> Hello everybody,
>
>
>
> I want to create a rigid object, that is in influence of a poiseuille flow.
> For that I use the feature VIRTUAL_SITES_RELATIVE. After one of the first
> integrations I get the error message: 0: map_position_to_lattice: position
> (79.24 , 23.5, 959.13) not inside a local plaquette in dir 0 ind[dir]=79
> rel=79.74 lpos=79.24.
>
>
>
> To create the virtual particles, I have create a text file with xyz
> coordinates. This coordinates are in distance of 2 relative to the
> non-virtual particle. In the appendix there is a picture, which shows the
> arrangement of the virtual particles.
>
>
>
> I´m using a dx=0.25, dt=0.1, agrid=1 and tau=1.
>
> The dimension of the box is: 60, 49, 1920
>
>
>
> The coordinates are not at grid knots. I think, that the feature
> vs_auto_relate arranges the particles to the nearest knots. Is that the
> case?
>
>
>
> At the last integration, before get the error, the x-position of the
> non-virtual particle is 2016995220.46 and the x-position of one of the
> virtual particles is 0.4644. How could it be? I thought at every time step
> the distance between virtual and non-virtual particle be the same.
>
>
>
> I have another question, too.
>
> I want to give the rigid object a mass. The manual tells,  that the inertia
> tensor of the non-virtual particle in the center of mass needs to be
> adapted. Does it mean, I have to add the feature mass when I create the
> non-virtual particle (like: part 0 pos x y z mass 4)?
>
>
>
> Thanks for help!
>
>
>
> Lena Holst
>
> _________________________________________________________
>
> Fraunhofer ICT-IMM
>
> Gruppe Modellierung / Modeling Group
>
> Abteilung Zukunftstechnologien / Future Technologies Department
>
> Carl-Zeiss-Straße 18-20, 55129 Mainz, Germany
>
> Phone +49 6131 990-396 / Fax +49 6131 990-205 / Mobile +49 1577-5327027
>
> address@hidden / www.imm.fraunhofer.de
>
>