**From:** Kyriakos Kachrimanis (*kgk_at_pharm.auth.gr*)

**Date:** Mon Dec 22 2008 - 06:40:44 CST

**Next message:**Luca Muccioli: "Re: xsc file format"**Previous message:**DimitryASuplatov: "NAMD for Intel Xeon Cluster with MPICH"**In reply to:**Luca Muccioli: "Re: xsc file format"**Next in thread:**Luca Muccioli: "Re: xsc file format"**Reply:**Luca Muccioli: "Re: xsc file format"**Messages sorted by:**[ date ] [ thread ] [ subject ] [ author ] [ attachment ]

Dear Luca,

thank you for your reply, but I still can't understand what those numbers

represent (this is probably because I am thinking in terms of unit cells and

not simulation boxes). Isn't the cell represented by axis lengths and

angles? Is it represented by a lattice vector matrix?

In the practical case of a crystal with the following unit cell parameters

(monoclinic) :

length a =10.066, length b =28.719, length c =4.831

angle alpha =90, angle beta =103.45, angle gamma =90

how should I represent the lattice lengths and angles in an xsc file? If

lattice vectors are needed, I can easily obtain the lattice vector matrix

for the unit cell using crystallographic software, but maybe I should have

asked first, is it possible to define a non-orthorhombic unit cell in namd?

I looked it up under "periodic boundary conditions" in the user manual but

it is not clear whether namd supports non-orthorhombic simulation boxes.

Regards,

Kyriakos.

----- Original Message -----

From: "Luca Muccioli" <luca_at_ms.fci.unibo.it>

To: "Kyriakos Kachrimanis" <kgk_at_pharm.auth.gr>

Cc: <namd-l_at_ks.uiuc.edu>

Sent: Monday, December 22, 2008 1:22 PM

Subject: Re: namd-l: xsc file format

*> dear Kyriakos,
*

*>
*

*> in the most general case you have nine figures for the simulation cell in
*

*> the xst/xsc file, e.g.
*

*>
*

*> # NAMD extended system configuration restart file
*

*> #$LABELS step a_x a_y a_z b_x b_y b_z c_x c_y c_z o_x o_y o_z
*

*> 20000000 45.8525 0 0 0.144183 41.6781 0 0.362906 0.23746 43.5339 0 0 0
*

*> 1 2 3 4 5 6 7 8 9
*

*>
*

*> they represent the cell axes (a,b,c), and they are the columns of the
*

*> matrix called CO in the fortran routines below. I adapted them, maybe they
*

*> need some changes to work.
*

*>
*

*> to compute the distances, you need also the inverse of the CO matrix=OC,
*

*> and do like this:
*

*>
*

*> r12=Mol2%m_center-Mol1%m_center
*

*> r12=matmul(OC/2._rk,r12)
*

*> Delta=2._rk*NINT(0.5_rk*r12)
*

*> r12=r12-Delta
*

*> r12=matmul(CO/2._rk,r12)
*

*> True_r=r12
*

*>
*

*> cheers
*

*>
*

*> Luca
*

*>
*

*> -------------------------------------------
*

*>
*

*> Luca Muccioli, Ph. D.
*

*> Dipartimento di Chimica Fisica e Inorganica, INSTM
*

*> Universita` di Bologna
*

*> Via Risorgimento 4
*

*> 40136 Bologna (Italia)
*

*>
*

*> E-mail: luca_at_fci.unibo.it
*

*> Phone: +39-051-6446992
*

*> Fax: +39-051-2093690
*

*> Web: http://www2.fci.unibo.it/~luca/
*

*>
*

*>
*

*>
*

*> function HMatrix(sides,angles) result(CO)
*

*> real(rk),intent(in):: sides(3),angles(3)
*

*> real(rk):: CO(3,3)
*

*> real(rk):: ax,bx,by,cx,cy,cz,cang(3),qt
*

*>
*

*> co(2,1)=0._rk
*

*> co(3,1)=0._rk
*

*> co(3,2)=0._rk
*

*>
*

*> ax=sides(1)
*

*>
*

*> cang(1)=COS(deg_to_rad*angles(1))
*

*> cang(2)=COS(deg_to_rad*angles(2))
*

*> cang(3)=COS(deg_to_rad*angles(3))
*

*>
*

*> qt=SIN(deg_to_rad*angles(3))
*

*>
*

*> bx=sides(2)*cang(3)
*

*> by=sides(2)*qt
*

*> cx=sides(3)*cang(2)
*

*> cy=sides(3)*(cang(1)-cang(2)*cang(3))/qt
*

*> cz=sqrt(sides(3)**2-cx*cx-cy*cy)
*

*>
*

*> co(1,1)=ax
*

*> co(1,2)=bx
*

*> co(1,3)=cx
*

*> co(2,2)=by
*

*> co(2,3)=cy
*

*> co(3,3)=cz
*

*>
*

*>
*

*> end function HMatrix
*

*>
*

*> function BoxSides(co,alpha,beta,gamma) result(v)
*

*> IMPLICIT none
*

*> integer, intent(in) :: cfg
*

*> real(rk):: CO(3,3)
*

*> real(rk) :: V(3)
*

*> real(rk),intent(out),optional:: alpha,beta,gamma
*

*>
*

*> real(rk):: ax,bx,cx,by,cy,cz,qt
*

*>
*

*> ax = CO(1,1)
*

*> bx = CO(1,2)
*

*> cx = CO(1,3)
*

*> by = CO(2,2)
*

*> cy = CO(2,3)
*

*> cz = CO(3,3)
*

*>
*

*> v(1) = ax
*

*> v(2) = sqrt(bx**2+by**2)
*

*> v(3) = sqrt(cx**2+cy**2+cz**2)
*

*>
*

*> if (present(alpha).and.present(beta).and.present(gamma)) then
*

*> gamma = bx/v(2)
*

*> beta = cx/v(3)
*

*> qt = by/v(2)
*

*> alpha = cy*qt/v(3)+beta*gamma
*

*>
*

*> alpha = rad_to_deg*acos(alpha)
*

*> beta = rad_to_deg*acos(beta)
*

*> gamma = rad_to_deg*acos(gamma)
*

*> endif
*

*>
*

*> end function BoxSides
*

*>
*

*>
*

*> On Mon, 22 Dec 2008, Kyriakos Kachrimanis wrote:
*

*>
*

*>> Date: Mon, 22 Dec 2008 11:24:46 +0200
*

*>> From: Kyriakos Kachrimanis <kgk_at_pharm.auth.gr>
*

*>> To: namd-l_at_ks.uiuc.edu
*

*>> Subject: namd-l: xsc file format
*

*>>
*

*>> Dear NAMD users,
*

*>> I am new to NAMD and I would like to simulate an organic crystal and I
*

*>> could not find a detailed description of the .xsc file format. Could
*

*>> someone please provide an example of an xsc file for a crystalline
*

*>> system? How are the unit cell axes lengths and angles defined?
*

*>> Thanks in advance.
*

*>
*

**Next message:**Luca Muccioli: "Re: xsc file format"**Previous message:**DimitryASuplatov: "NAMD for Intel Xeon Cluster with MPICH"**In reply to:**Luca Muccioli: "Re: xsc file format"**Next in thread:**Luca Muccioli: "Re: xsc file format"**Reply:**Luca Muccioli: "Re: xsc file format"**Messages sorted by:**[ date ] [ thread ] [ subject ] [ author ] [ attachment ]

*
This archive was generated by hypermail 2.1.6
: Wed Feb 29 2012 - 15:50:18 CST
*