Re: [ESPResSo-users] BOND_ANGLE, restarting simulation, dynamic bonding

[Arash Azari]

Dear Rudolf,

Thank you very much for your reply. I
checked “myconfig.h” file (location and features) and the “code_info” output and
everything is fine; all the required features are available in
the ESPResSo, but still I have the same problems: unknown interaction for
bond angle and “invalid command name "collision_detection"
“ for dynamic bonding. 
I highly appreciate any suggestion or
recommendation.
Thank you very much,


Best regards,
Arash
 
Arash Azari


----- Original Message -----
From: Rudolf Weeber <address@hidden>
To: Arash Azari <address@hidden>
Cc: "address@hidden" <address@hidden>
Sent: Monday, April 1, 2013 6:08 PM
Subject: Re: [ESPResSo-users] BOND_ANGLE, restarting simulation, dynamic bonding

Hi,
On Mon, Apr 01, 2013 at 08:26:19AM -0700, Arash Azari wrote:
>  
> I am using the nightly build of
> - I compiled it with the BOND_ANGLE
>     option, but inter command output recognizes all the bond angle
>     potentials as an “unknown”. I thought it is not very important
>     because simulation looks fine in terms of the bond angle potential
>     behaviour. The problem is that when I am using the blockfile to
>     create an output file, ESPResSo cannot recognize the interaction
>     type (defined as an unknown in output file). Is it a bug?
> 
> -I compiled the ESPResSo with the
>     required features (COLLISION_DETECTION and VIRTUAL_SITES_RELATIVE)
>     to use the dynamic bonding, but I receive the “invalid command
>     name "collision_detection" “ error message? I used
>     single CPU to run this dynamic bonding simulation. Could you please
>     point out my mistake?
One possible reason might be, that your myconfig.h was not in the directory 
from which you executed the configure script, and you did not specify the path 
to myconfig.h explicitly when calling configure.
In that case, the features mentioned above might not actually be compiled in.
To find out, what really was compiled in, call
code_info
from within Espresso.


> 
> - Is it possible to manually delete
>     all or some of the the bonds which are created during the dynamic
>     bonding simulation? I mean when we restart the simulation not during
>     the same simulation.
That's a bit tricky, because you don't only need to remove the bond (via the 
part command), but also the two virtual sites.
To do it, you would have to write a loop which goes over all particles and 
locates the virtual sites related to the particles you want to separate.

1. Delete the bond between the two non-virtual particles you want to separate 
2. Loop over all particles and for each particle
  check whether it is a virtual site and it is related to one of the particles 
you want to separate.
  If this is the case, delete the virtual site (using the "part delete" command)

When you do this separation, make sure that the collision_detection is turned 
off. Otherwise, the particles will be bound again in the next time step.

Hope that helps!

Regards, Rudolf