Structure optimization Problem
Posted: Thu Jun 19, 2014 1:17 pm
Hello
I want to optimize the FCC structure of 58 atoms using GGA (PBE) pseudopotential. But the structure is not relaxed after 43 iteration of geometry optimization. The input file for kpoint convergence is attached. I am confused when i check log file. I checked :
at Broyd/MD step 11, gradients have not converged yet.
max grad (force/stress) = 4.4142E-03 > tolmxf= 5.0000E-05 ha/bohr (free atoms)
at Broyd/MD step 12, gradients have not converged yet.
max grad (force/stress) = 4.0370E-03 > tolmxf= 5.0000E-05 ha/bohr (free atoms)
at Broyd/MD step 13, gradients have not converged yet.
max grad (force/stress) = 3.3132E-03 > tolmxf= 5.0000E-05 ha/bohr (free atoms)
at Broyd/MD step 16, gradients have not converged yet.
max grad (force/stress) = 2.9676E-03 > tolmxf= 5.0000E-05 ha/bohr (free atoms)
at Broyd/MD step 17, gradients have not converged yet.
max grad (force/stress) = 5.0281E-03 > tolmxf= 5.0000E-05 ha/bohr (free atoms)
at Broyd/MD step 18, gradients have not converged yet.
max grad (force/stress) = 5.9417E-03 > tolmxf= 5.0000E-05 ha/bohr (free atoms)
at Broyd/MD step 22, gradients have not converged yet.
max grad (force/stress) = 4.3647E-03 > tolmxf= 5.0000E-05 ha/bohr (free atoms)
at Broyd/MD step 25, gradients have not converged yet.
max grad (force/stress) = 2.8561E-03 > tolmxf= 5.0000E-05 ha/bohr (free atoms)
at Broyd/MD step 27, gradients have not converged yet.
max grad (force/stress) = 2.9564E-03 > tolmxf= 5.0000E-05 ha/bohr (free atoms)
at Broyd/MD step 29, gradients have not converged yet.
max grad (force/stress) = 5.2835E-03 > tolmxf= 5.0000E-05 ha/bohr (free atoms)
at Broyd/MD step 32, gradients have not converged yet.
max grad (force/stress) = 4.5316E-03 > tolmxf= 5.0000E-05 ha/bohr (free atoms)
at Broyd/MD step 34, gradients have not converged yet.
max grad (force/stress) = 4.8938E-03 > tolmxf= 5.0000E-05 ha/bohr (free atoms)
at Broyd/MD step 35, gradients have not converged yet.
max grad (force/stress) = 4.5135E-03 > tolmxf= 5.0000E-05 ha/bohr (free atoms)
at Broyd/MD step 36, gradients have not converged yet.
max grad (force/stress) = 4.2217E-03 > tolmxf= 5.0000E-05 ha/bohr (free atoms)
at Broyd/MD step 38, gradients have not converged yet.
max grad (force/stress) = 3.2240E-03 > tolmxf= 5.0000E-05 ha/bohr (free atoms)
at Broyd/MD step 39, gradients have not converged yet.
max grad (force/stress) = 2.3726E-03 > tolmxf= 5.0000E-05 ha/bohr (free atoms)
at Broyd/MD step 40, gradients have not converged yet.
max grad (force/stress) = 1.9986E-03 > tolmxf= 5.0000E-05 ha/bohr (free atoms)
at Broyd/MD step 41, gradients have not converged yet.
max grad (force/stress) = 2.2147E-03 > tolmxf= 5.0000E-05 ha/bohr (free atoms)
at Broyd/MD step 42, gradients have not converged yet.
max grad (force/stress) = 1.9428E-03 > tolmxf= 5.0000E-05 ha/bohr (free atoms)
at Broyd/MD step 43, gradients have not converged yet.
max grad (force/stress) = 2.3150E-03 > tolmxf= 5.0000E-05 ha/bohr (free atoms)
After seeing oscillations in the value of max grad(force/stress), i stopped this calculation. Is there any mistake in input file or any other fault. I am using only 8 cores for this calculation and these 43 iterations take approximately 10 days. Should i continue ???
Janpreet
I want to optimize the FCC structure of 58 atoms using GGA (PBE) pseudopotential. But the structure is not relaxed after 43 iteration of geometry optimization. The input file for kpoint convergence is attached. I am confused when i check log file. I checked :
at Broyd/MD step 11, gradients have not converged yet.
max grad (force/stress) = 4.4142E-03 > tolmxf= 5.0000E-05 ha/bohr (free atoms)
at Broyd/MD step 12, gradients have not converged yet.
max grad (force/stress) = 4.0370E-03 > tolmxf= 5.0000E-05 ha/bohr (free atoms)
at Broyd/MD step 13, gradients have not converged yet.
max grad (force/stress) = 3.3132E-03 > tolmxf= 5.0000E-05 ha/bohr (free atoms)
at Broyd/MD step 16, gradients have not converged yet.
max grad (force/stress) = 2.9676E-03 > tolmxf= 5.0000E-05 ha/bohr (free atoms)
at Broyd/MD step 17, gradients have not converged yet.
max grad (force/stress) = 5.0281E-03 > tolmxf= 5.0000E-05 ha/bohr (free atoms)
at Broyd/MD step 18, gradients have not converged yet.
max grad (force/stress) = 5.9417E-03 > tolmxf= 5.0000E-05 ha/bohr (free atoms)
at Broyd/MD step 22, gradients have not converged yet.
max grad (force/stress) = 4.3647E-03 > tolmxf= 5.0000E-05 ha/bohr (free atoms)
at Broyd/MD step 25, gradients have not converged yet.
max grad (force/stress) = 2.8561E-03 > tolmxf= 5.0000E-05 ha/bohr (free atoms)
at Broyd/MD step 27, gradients have not converged yet.
max grad (force/stress) = 2.9564E-03 > tolmxf= 5.0000E-05 ha/bohr (free atoms)
at Broyd/MD step 29, gradients have not converged yet.
max grad (force/stress) = 5.2835E-03 > tolmxf= 5.0000E-05 ha/bohr (free atoms)
at Broyd/MD step 32, gradients have not converged yet.
max grad (force/stress) = 4.5316E-03 > tolmxf= 5.0000E-05 ha/bohr (free atoms)
at Broyd/MD step 34, gradients have not converged yet.
max grad (force/stress) = 4.8938E-03 > tolmxf= 5.0000E-05 ha/bohr (free atoms)
at Broyd/MD step 35, gradients have not converged yet.
max grad (force/stress) = 4.5135E-03 > tolmxf= 5.0000E-05 ha/bohr (free atoms)
at Broyd/MD step 36, gradients have not converged yet.
max grad (force/stress) = 4.2217E-03 > tolmxf= 5.0000E-05 ha/bohr (free atoms)
at Broyd/MD step 38, gradients have not converged yet.
max grad (force/stress) = 3.2240E-03 > tolmxf= 5.0000E-05 ha/bohr (free atoms)
at Broyd/MD step 39, gradients have not converged yet.
max grad (force/stress) = 2.3726E-03 > tolmxf= 5.0000E-05 ha/bohr (free atoms)
at Broyd/MD step 40, gradients have not converged yet.
max grad (force/stress) = 1.9986E-03 > tolmxf= 5.0000E-05 ha/bohr (free atoms)
at Broyd/MD step 41, gradients have not converged yet.
max grad (force/stress) = 2.2147E-03 > tolmxf= 5.0000E-05 ha/bohr (free atoms)
at Broyd/MD step 42, gradients have not converged yet.
max grad (force/stress) = 1.9428E-03 > tolmxf= 5.0000E-05 ha/bohr (free atoms)
at Broyd/MD step 43, gradients have not converged yet.
max grad (force/stress) = 2.3150E-03 > tolmxf= 5.0000E-05 ha/bohr (free atoms)
After seeing oscillations in the value of max grad(force/stress), i stopped this calculation. Is there any mistake in input file or any other fault. I am using only 8 cores for this calculation and these 43 iterations take approximately 10 days. Should i continue ???
Janpreet