Skip to content

Bug: ABACUS cannot get SCF convergence with combined use of ultrasoft and norm-conserving pseudopotential #5596

@kirk0830

Description

@kirk0830

Describe the bug

                                                                                     
                              ABACUS v3.8.1

               Atomic-orbital Based Ab-initio Computation at UStc                    

                     Website: http://abacus.ustc.edu.cn/                             
               Documentation: https://abacus.deepmodeling.com/                       
                  Repository: https://github.com/abacusmodeling/abacus-develop       
                              https://github.com/deepmodeling/abacus-develop         
                      Commit: 146a50931 (Mon Oct 21 16:41:35 2024 +0800)

 Mon Nov 25 11:00:06 2024
 MAKE THE DIR         : OUT.ABACUS/
 RUNNING WITH DEVICE  : CPU / Intel(R) Xeon(R) Platinum 8255C CPU @ 2.50GHz
 UNIFORM GRID DIM        : 48 * 48 * 48
 UNIFORM GRID DIM(BIG)   : 48 * 48 * 48
 DONE(0.150851   SEC) : SETUP UNITCELL
 DONE(0.187362   SEC) : SYMMETRY
 DONE(5.36323    SEC) : INIT K-POINTS
 ---------------------------------------------------------
 Ion relaxation calculations
 ---------------------------------------------------------
 SPIN    KPOINTS         PROCESSORS  
 1       1140            16          
 ---------------------------------------------------------
 Use plane wave basis
 ---------------------------------------------------------
 ELEMENT NATOM       XC          
 B       1           
 O       1           
 ---------------------------------------------------------
 Initial plane wave basis and FFT box
 ---------------------------------------------------------
 DONE(5.39719    SEC) : INIT PLANEWAVE
 DONE(5.4094     SEC) : LOCAL POTENTIAL
 DONE(5.76873    SEC) : NON-LOCAL POTENTIAL
 MEMORY FOR PSI (MB)  : 82.9742
 DONE(5.82108    SEC) : INIT BASIS
 -------------------------------------------
 STEP OF RELAXATION : 1
 -------------------------------------------
 START CHARGE      : atomic
 DONE(5.87049    SEC) : INIT SCF
 * * * * * *
 << Start SCF iteration.
 ITER       ETOT/eV          EDIFF/eV         DRHO     TIME/s
 DA1     -5.15988845e+02   0.00000000e+00   1.4746e-01  27.96
 DA2     -1.20301889e+04  -1.15142001e+04   5.8264e+01  30.88
 DA3     -2.46284030e+05  -2.34253841e+05   6.4471e+01  43.78
 DA4     -4.00052587e+06  -3.75424184e+06   6.6798e+01  35.23
 DA5     -6.40785857e+07  -6.00780598e+07   1.4915e+01  31.69
 DA6     -1.02532576e+09  -9.61247171e+08   2.4406e+01  23.65
 DA7     -1.64052839e+10  -1.53799582e+10   2.3530e+00  28.55
 DA8     -2.62484615e+11  -2.46079331e+11   5.4242e-02  21.42
 DA9     -4.19975391e+12  -3.93726930e+12   1.1759e-03  22.42

 notconv = 12
 DiagoDavid::diag', too many bands are not converged! 

 notconv = 11
 DiagoDavid::diag', too many bands are not converged! 

 notconv = 12
 DiagoDavid::diag', too many bands are not converged! 

 notconv = 14
 DiagoDavid::diag', too many bands are not converged! 

 notconv = 11
 DiagoDavid::diag', too many bands are not converged! 

 notconv = 15
 DiagoDavid::diag', too many bands are not converged! 

 notconv = 14
 DiagoDavid::diag', too many bands are not converged! 

 notconv = 11
 DiagoDavid::diag', too many bands are not converged! 

Expected behavior

Ultrasoft pseudopotential can be used together with norm-conserving, this is at least supported by QE.

To Reproduce

703c4655-fddc-4a9d-a47b-b499a1414d1c.zip

** Now ALL my calculations show this behavior, the above is just the one I randomly pick **

Environment

No response

Additional Context

No response

Task list for Issue attackers (only for developers)

  • Verify the issue is not a duplicate.
  • Describe the bug.
  • Steps to reproduce.
  • Expected behavior.
  • Error message.
  • Environment details.
  • Additional context.
  • Assign a priority level (low, medium, high, urgent).
  • Assign the issue to a team member.
  • Label the issue with relevant tags.
  • Identify possible related issues.
  • Create a unit test or automated test to reproduce the bug (if applicable).
  • Fix the bug.
  • Test the fix.
  • Update documentation (if necessary).
  • Close the issue and inform the reporter (if applicable).

Metadata

Metadata

Assignees

Labels

BugsBugs that only solvable with sufficient knowledge of DFT

Type

No type

Projects

No projects

Milestone

No milestone

Relationships

None yet

Development

No branches or pull requests

Issue actions