Dear all ALPS users
I am going to run a DMRG simulation on the following one dimensional spin-1 Hamiltonian with open boundary conditions, but with effective magnetic field at both edges :
H=-\sum_{j}^{N-1}S^z_{j}S^z_{j}S^z_{j+1} + D\sum_{j}^{N}(S^z_{j})^2 - h\sum_{j}^{N}S^z_{j} -B\sum_{j}^{N}S^x_{j} + h_{eff}(S_1+S_N),
Indeed in the absence of the effective magnetic field the procedure is straightforward, but I don't see how to modified the lattice or any other requirements to go further
Any help would be appreciated. Thanks,
Javad Vahedi,
Dept. of Physics, Sari Branch of Islamic Azad University, P.O.Box 48164-194, Sari, Iran. Tel: (+98) 911 1554504 Fax: (+98)151 2256679 E-mail: javahedi@gmail.com javahedi@iausari.ac.ir
Here is what you can do:
- create a graph from the lattice, either by hand or using e.g. the lattice/example4 program - edit the graph, assigning a different type to the boundary sites (e.g. type="1") - put the graph into your lattices.xml file - set the parameter h1 to your boundary field (h0 is the bulk field, for sites of type 0).
A procedure similar to this, for boundary spins in a S=1 chain is explained in the DMRG tutorials.
Matthias
On Jan 30, 2012, at 1:49 PM, Javad Vahedi wrote:
Dear all ALPS users
I am going to run a DMRG simulation on the following one dimensional spin-1 Hamiltonian with open boundary conditions, but with effective magnetic field at both edges :
H=-\sum_{j}^{N-1}S^z_{j}S^z_{j}S^z_{j+1} + D\sum_{j}^{N}(S^z_{j})^2 - h\sum_{j}^{N}S^z_{j} -B\sum_{j}^{N}S^x_{j} + h_{eff}(S_1+S_N),
Indeed in the absence of the effective magnetic field the procedure is straightforward, but I don't see how to modified the lattice or any other requirements to go further
Any help would be appreciated. Thanks,
Javad Vahedi,
Dept. of Physics, Sari Branch of Islamic Azad University, P.O.Box 48164-194, Sari, Iran. Tel: (+98) 911 1554504 Fax: (+98)151 2256679 E-mail: javahedi@gmail.com javahedi@iausari.ac.ir
comp-phys-alps-users@lists.phys.ethz.ch