1.6 Shared inputs, conventions and evidence
1.6.1 Complete shared inputs and analysis templates
1.6.1.1 Input block 1
&GLOBAL
PROJECT water_A
RUN_TYPE ENERGY_FORCE
PRINT_LEVEL MEDIUM
PREFERRED_FFT_LIBRARY FFTSG
&END GLOBAL
&FORCE_EVAL
METHOD Quickstep
&DFT
BASIS_SET_FILE_NAME BASIS_MOLOPT
POTENTIAL_FILE_NAME GTH_POTENTIALS
CHARGE 0
MULTIPLICITY 1
&QS
METHOD GPW
EPS_DEFAULT 1.0E-12
&END QS
&MGRID
CUTOFF [Ry] 500
REL_CUTOFF [Ry] 60
NGRIDS 4
&END MGRID
&SCF
SCF_GUESS ATOMIC
EPS_SCF 1.0E-7
MAX_SCF 150
&OT
MINIMIZER CG
PRECONDITIONER FULL_ALL
&END OT
&END SCF
&POISSON
PERIODIC NONE
POISSON_SOLVER WAVELET
&END POISSON
&XC
&XC_FUNCTIONAL PBE
&END XC_FUNCTIONAL
&END XC
&END DFT
&SUBSYS
&CELL
ABC [angstrom] 16.0 16.0 16.0
PERIODIC NONE
&END CELL
&COORD
UNIT angstrom
O 8.000000 8.000000 8.000000
H 8.980000 8.000000 8.000000
H 7.750000 8.950000 8.000000
&END COORD
&KIND O
ELEMENT O
BASIS_SET DZVP-MOLOPT-GTH
POTENTIAL GTH-PBE-q6
&END KIND
&KIND H
ELEMENT H
BASIS_SET DZVP-MOLOPT-GTH
POTENTIAL GTH-PBE-q1
&END KIND
&END SUBSYS
&END FORCE_EVAL
1.6.1.2 Input block 2
&SUBSYS
&CELL
ABC [angstrom] 5.43 5.43 5.43
PERIODIC XYZ
&END CELL
&COORD
SCALED T
Si 0.00 0.00 0.00
Si 0.00 0.50 0.50
Si 0.50 0.00 0.50
Si 0.50 0.50 0.00
Si 0.25 0.25 0.25
Si 0.25 0.75 0.75
Si 0.75 0.25 0.75
Si 0.75 0.75 0.25
&END COORD
&KIND Si
ELEMENT Si
BASIS_SET DZVP-MOLOPT-SR-GTH
POTENTIAL GTH-PBE-q4
&END KIND
&END SUBSYS
1.6.2 Conventions and learning evidence
1.6.2.1 CP2K
Original bilingual course for quantumchemistry.net. Teaching baseline: CP2K 2025.2, with links to that release's input reference wherever available. This is a designed course, not a report of calculations performed. Input values are starting points for numerical experiments; no energy, spectrum, optimized geometry, diffusion coefficient or free-energy barrier is asserted as an obtained result. Read the case-specific applicability and validation requirements before using an input in research.
1.6.2.2 How to use the course
Every case contains an intuition, prerequisites, an input or precisely located delta, a workflow, interpretation, checks, an exercise and an original diagram specification. A delta is not a standalone input: replace the indicated section of a previously validated baseline rather than appending duplicate sections. Blocks labeled schematic identify scientific decisions or missing model files and must not be submitted unchanged. The original figure specifications are for newly drawn vector graphics, not reproductions of source figures. In bilingual rendering, use the EN or ZH label set, not both squeezed into the same small panel.
Use a directory for each calculation and retain input, coordinates, data-file checksums, executable version, stdout, stderr and analysis script. Specify the precise observable before choosing a tolerance: a cutoff adequate for a rough energy ranking need not be adequate for stress or a numerical Hessian. Do not confuse convergence of a numerical representation with adequacy of the chosen exchange-correlation functional. Atomic units in output require deliberate conversion; common energy and force outputs are Hartree and Hartree/bohr, while coordinates in our explicit blocks are angstrom unless noted. Check output headings instead of assuming units.
1.6.2.3 Baseline A: isolated water
This complete teaching input uses a centered, deliberately non-optimized water geometry and a cubic nonperiodic box. The WAVELET Poisson solver requires a supporting build and compatible cell; if unavailable, choose a documented isolated solver for your build and repeat the box-convergence test rather than silently using periodic electrostatics. For this WAVELET baseline, GLOBAL explicitly selects PREFERRED_FFT_LIBRARY FFTSG as required by the 2025.2 reference. The executable must find the installed BASIS_MOLOPT and GTH_POTENTIALS files. Their names are files, not automatically downloaded resources.
See input block 1 above.
1.6.2.4 Baseline B: conventional silicon cell
Create B by replacing A's SUBSYS with the block below, replacing POISSON with PERIODIC XYZ and POISSON_SOLVER PERIODIC, and changing PROJECT to silicon_B. Keep PBE, GPW and the starting grid. For this small cell, Gamma-only is an educational starting point, not a converged Brillouin-zone integration. Case 12 supplies the diagonalization and k-point route. The chosen lattice parameter is an input guess, not this course's calculated equilibrium lattice constant.
See input block 2 above.
1.6.2.5 Course assessment and reproducibility checklist
1.6.2.6 A capstone that connects the chapters
Choose one bounded question: a water-dimer interaction curve, a silicon equilibrium volume, or a structural observable of liquid water. Write the physical model before writing the input. State charge/spin, composition, boundary conditions, exchange-correlation model, dispersion choice, orbital basis, pseudopotentials and the reference convention. Declare the observable and a useful accuracy target in its natural units. Build a numerical error budget that separates basis, grid, SCF and sampling error; for dynamics add integration and statistical error. A successful submission may conclude that the available trajectory is too short or the chosen model is inadequate. That is a stronger scientific result than a precise-looking number unsupported by tests.
Submit the actual input and model files, software/data provenance, a table of performed calculations with failed attempts retained, analysis code, and an uncertainty-aware conclusion. Figures must distinguish measured data from conceptual diagrams. Each figure needs axis units, legends and the definition of any normalization. For a comparison, state what was held fixed and what was changed. Include one independent validation: a finite-difference force, an equation-of-state cross-check, a restart continuity test, or an analysis tested against a known synthetic configuration. Do not use an unperformed test as a checkmark.
1.6.2.7 A practical preflight
- Count atoms, elements, charge and explicit electrons before the first SCF.
1.6.2.8 Terms that must not be conflated
SCF convergence solves an electronic stationary problem to a threshold; basis convergence approaches a more complete representation; k-point convergence improves periodic integration; geometry convergence finds a stationary nuclear structure; equilibration removes initial-condition transients; sampling convergence quantifies ensemble averages with adequate independent information. None automatically proves the others. The same distinction applies to a total electronic energy, an interaction energy, a binding energy with deformation, an enthalpy and a free energy. Write the equation rather than relying only on the label.