Updated hubbard benchmarks

accuracyBenchmarks/Pt3Ni/README.md

@@ -1,11 +1,13 @@
 Description of the example
 ==========================
-This example demonstrates the inputs required for hubbard correction. This is a non-periodic spin polarized ground state calculation.
+This example demonstrates the inputs required for hubbard correction. This is an example for a non-periodic spin polarized calculation where all the atoms have hubbard corrections. 
 
 Studies performed
 =======================
 
-Ground-state calculation: convergence study of ground-state energy and ionic forces to reach chemical accuracy with respect to FE discretization, and validation against QUANTUM ESPRESSO (QE). 
+Ground-state calculation: convergence study of ground-state energy and ionic forces to reach chemical accuracy with respect to FE discretization, and validation against QUANTUM ESPRESSO (QE).  (Study1)
+
+Ionic relaxation: Relaxation of the atomic positions with spin polarization. (Study2)
 
 Discussion on the input parameters and the results
 ==================================================
@@ -29,12 +31,18 @@ corrections are applied on 1 set of orbitals. 9.0 is the initial occupancy on th
    78 2 - Atomic number (78) and hubbard species id (2).
 
 * QE:  
-        a) ecutwfc                  = 90 Ha 
-        b) U Pt-5d 3.0              = Hubbard correction applied to 5 d orbitals of Pt with U = 3.0 eV
-        c) U Ni-3d 1.0              = Hubbard correction applied to 3 d orbitals of Ni with U = 1.0 eV                                   
-Ground-state calculation results
+        a) U Pt-5d 3.0              = Hubbard correction applied to 5 d orbitals of Pt with U = 3.0 eV
+        b) U Ni-3d 1.0              = Hubbard correction applied to 3 d orbitals of Ni with U = 1.0 eV                                   
+Ground-state calculation results (Study 1)
 --------------------------------
 
 * Ground State Comparison with QE(90 Ha energy cut off):
-    a) Energy Difference = 1.863E-05 Ha/atom
-    b) Force Difference = 1.74E-05 Ha/bohr (max absolute error among all atoms and force components)
+    a) Energy Difference =  1.76E-05 Ha/atom
+    b) Force Difference =  3.89E-05 Ha/bohr (max absolute error among all atoms and force components)
+
+Ionic relaxation results (Study 2)
+--------------------------------
+
+* Ionic relaxation converged in 6 iterations. 
+
+* Comparison in with QE (60 Ha energy cut off) : The atomic positions differ by 1.74E-04 Bohrs. The energy of the final relaxed structure differed by 4.02E-06 Ha/atom.

accuracyBenchmarks/Pt3Ni/dftfe/parameters.prm → accuracyBenchmarks/Pt3Ni/dftfe/study1/parameters.prm

@@ -1,4 +1,5 @@
 set VERBOSITY=1
+set USE GPU        = false
 subsection Geometry
   set NATOMS=4
   set NATOM TYPES=2

accuracyBenchmarks/Pt3Ni/dftfe/study2/coordinates.inp

@@ -0,0 +1,4 @@
+28 18 0.0000   0.0000   0.0000 -0.2
+78 18 3.6000   3.6000   0.0000 -0.2
+78 18 0.0000   3.6000   3.6000 -0.2
+78 18 3.6000   0.0000   3.6000 -0.2

accuracyBenchmarks/Pt3Ni/dftfe/study2/domainVectors.inp

@@ -0,0 +1,3 @@
+80.0 0.0 0.0
+0.0 80.0 0.0
+0.0 0.0 80.0

accuracyBenchmarks/Pt3Ni/dftfe/study2/hubbard.inp

@@ -0,0 +1,10 @@
+3
+0 0
+1 28 0.0367492929 1 8.0
+3 2  
+2 78 0.1102478786 1 9.0
+5 2
+28 1
+78 2
+78 2
+78 2

accuracyBenchmarks/Pt3Ni/dftfe/study2/parameters_ionOpt.prm

@@ -0,0 +1,43 @@
+set VERBOSITY=1
+set USE GPU        = true
+set SOLVER MODE=GEOOPT
+subsection Geometry
+  set NATOMS=4
+  set NATOM TYPES=2
+  set ATOMIC COORDINATES FILE      = coordinates.inp 
+  set DOMAIN VECTORS FILE = domainVectors.inp
+  subsection Optimization
+    set OPTIMIZATION MODE         = ION
+    set ION FORCE = true
+    set FORCE TOL=2e-4
+  end
+end
+
+subsection Finite element mesh parameters
+  set POLYNOMIAL ORDER = 6
+  subsection Auto mesh generation parameters
+    set MESH SIZE AROUND ATOM = 0.8
+  end
+end
+
+subsection DFT functional parameters
+  set EXCHANGE CORRELATION TYPE   = GGA-PBE+U
+  set PSEUDOPOTENTIAL CALCULATION = true
+  set PSEUDOPOTENTIAL FILE NAMES LIST = pseudo.inp
+  set SPIN POLARIZATION               = 1
+  subsection Hubbard Parameters
+    set HUBBARD PARAMETERS FILE = hubbard.inp
+  end
+end
+
+subsection SCF parameters
+  set MIXING PARAMETER =0.1
+  set MIXING METHOD= ANDERSON
+  set STARTING WFC             = RANDOM
+  set TEMPERATURE              = 2000
+  set TOLERANCE                = 1e-5
+  subsection Eigen-solver parameters
+    set NUMBER OF KOHN-SHAM WAVEFUNCTIONS                    = 80
+    set CHEBYSHEV FILTER TOLERANCE                           = 1e-4
+  end
+end

accuracyBenchmarks/Pt3Ni/dftfe/study2/pseudo.inp

@@ -0,0 +1,2 @@
+28 Ni.upf
+78 Pt.upf