Ewald energy calculation error when using KERNEL_TRUNCATION
Posted: Sun Dec 21, 2025 4:38 am
Dear experts,
I am reporting an issue where the Ewald energy calculation fails (returns NaN) when using the KERNEL_TRUNCATION option for a long 2D slab system.
this error appears to be dependent on the system size (number of layers). I have tested the same setup with different layer thicknesses
2 and 3 layers: The calculation completes successfully without any issues.
4 or more layers: Ewald energy TEWEN = NaN error occurs consistently from the first iteration.
OUTCAR:
Code: Select all
--------------------------------------- Iteration 1( 1) ---------------------------------------
POTLOK: cpu time 0.9096: real time 0.9199
SETDIJ: cpu time 0.0627: real time 0.0630
EDDAV: cpu time 2.2556: real time 2.2956
DOS: cpu time 0.0038: real time 0.0039
--------------------------------------------
LOOP: cpu time 3.2321: real time 3.2844
eigenvalue-minimisations : 16128
total energy-change (2. order) : NaN (-0.9811510E+05)
number of electron 104.0000000 magnetization
augmentation part 104.0000000 magnetization
Free energy of the ion-electron system (eV)
---------------------------------------------------
alpha Z PSCENC = 0.00000000
Ewald energy TEWEN = NaN
-Hartree energ DENC = 518075.75668292
-exchange EXHF = 0.00000000
-V(xc)+E(xc) XCENC = 359.79731494
PAW double counting = 7734.02099078 -7647.93876759
entropy T*S EENTRO = -0.00165409
eigenvalues EBANDS = -96731.10811112
atomic energy EATOM = 9682.53318035
Solvation Ediel_sol = 0.00000000
---------------------------------------------------
free energy TOTEN = NaN eV
energy without entropy = NaN energy(sigma->0) = NaN
I have attached the INCAR and POSCAR files I used below. The structure has a thickness of 21.7 Å with a 50 Å vacuum, and the layers are positioned in the center of the cell.
INCAR:
Code: Select all
System = MoS2
PREC = Accurate
ENCUT = 500.0
EDIFF = 1E-8
ISYM = 0
NCORE = 2
KPAR = 4
ALGO = Normal
#------rVV10 of Sabatini et al. ------------------------
GGA = ML
LUSE_VDW = .TRUE.
IVDW_NL = 2
BPARAM = 6.3 # default but can be overwritten by this tag
CPARAM = 0.0093 # default but can be overwritten by this tag
LASPH = .TRUE.
ISPIN = 1 #does spin-polarized calc.
#--------OUTPUT FILE CONTROL---------
LORBIT = 11
NEDOS = 2000
LWAVE = .FALSE.
LCHARG = .FALSE.
ISMEAR = 0
SIGMA=0.1
KERNEL_TRUNCATION {
LTRUNCATE = T
IDIMENSIONALITY = 2
LCOARSEN = F
ISURFACE = 3
}POSCAR:
Code: Select all
Generated by VASPKIT code
1.00000000000000
3.2182657598431033 -0.0000010946757415 0.0000000188723630
-1.6091338279532288 2.7870993183750494 -0.0000000073360658
0.0000004151795089 0.0000000574878470 71.7187625893623846
Mo S
4 8
Direct
0.0000000245375861 0.0000000253101545 0.5430315299245932
0.6666666880175047 0.3333333364044524 0.3707663947433152
0.6666666650141739 0.3333333544272531 0.6292241052657448
0.3333333422475198 0.6666666930535939 0.4569429145887748
0.0000000220151665 0.0000000124768911 0.3487857909612084
0.0000000111490195 0.9999999829017471 0.6072500396082302
0.6666666792821385 0.3333333434751254 0.4349762429389977
0.3333333522353835 0.6666666890091242 0.5210655114016074
0.9999999952437634 0.0000000133540298 0.3927534447758242
0.9999999837961203 0.9999999923279267 0.6512141798867679
0.6666666732158412 0.3333333270948957 0.4789323123252539
0.3333333402457930 0.6666666631648132 0.5650213565796909bset and regards