ABACUS learning pathways
ABACUS · PW · LCAO
Learn how plane waves and numerical atomic orbitals turn a physical question into an auditable calculation. Follow eight chapters from the three input files to converged structures, electronic properties and reproducible workflows.
All example inputs are unexecuted teaching templates. Pin your ABACUS release, validate pseudopotentials and orbitals, and establish convergence for the requested observable. Advanced features and external tools require the version and build checks described in each lesson.
1 Files, bases and reproducible setup
1.1 Understand INPUT, STRU and KPT
Understand INPUT, STRU and KPT
Study the workflow →1.2 Match pseudopotentials and numerical orbitals
Match pseudopotentials and numerical orbitals
Study the workflow →1.3 First silicon calculation with plane waves
First silicon calculation with plane waves
Study the workflow →1.4 First silicon calculation with LCAO
First silicon calculation with LCAO
Study the workflow →2 Convergence and electronic stability
2.1 Converge the PW energy cutoff
Converge the PW energy cutoff
Study the workflow →2.2 Converge numerical orbital quality
Converge numerical orbital quality
Study the workflow →2.3 Metal k-points and smearing
Metal k-points and smearing
Study the workflow →2.4 Diagnose SCF mixing and charge sloshing
Diagnose SCF mixing and charge sloshing
Study the workflow →3 Molecules, forces and structural response
3.1 Water in a periodic vacuum box
Water in a periodic vacuum box
Study the workflow →3.2 Relax atomic coordinates
Relax atomic coordinates
Study the workflow →3.3 Relax the cell and validate stress
Relax the cell and validate stress
Study the workflow →3.4 Build a controlled equation of state
Build a controlled equation of state
Study the workflow →4 Bands, density and magnetism
4.1 Bands from a converged charge density
Bands from a converged charge density
Study the workflow →4.2 DOS and orbital projections
DOS and orbital projections
Study the workflow →4.3 Charge-density differences on matched grids
Charge-density differences on matched grids
Study the workflow →4.4 Test collinear magnetic states
Test collinear magnetic states
Study the workflow →5 Surfaces, defects and adsorption
5.1 Converge slab thickness and vacuum
When is a repeated slab a model of an isolated surface rather than a stack of interacting films? Separate the number of atomic layers from the empty distance between periodic images. The two controls address different errors and cannot be replaced by one large total cell height.
Study the workflow →5.2 Balance bulk and slab surface energies
How much energy is required to create two equivalent surfaces from a bulk crystal? The calculation is a subtraction between large energies; a mismatch in bulk and slab settings can dominate the small excess being sought.
Study the workflow →5.3 Neutral vacancy formation and finite-size checks
What is the energetic cost of removing one atom from bulk silicon and returning it to a bulk reservoir? This case concerns a neutral supercell only. It teaches atom counting, symmetry breaking and image interactions without adding unsupported charged-defect corrections.
Study the workflow →5.4 Adsorption energies with consistent references
Does a molecule bind to a chosen surface in the selected electronic-structure model? Establish a balanced energy cycle before assigning a sign or comparing sites. An adsorption energy is not an activation barrier, a finite-temperature free energy or a measured sticking probability.
Study the workflow →6 Phonons, molecular dynamics and analysis
6.1 Finite-displacement phonons with Phonopy
Can a relaxed crystal resist a small atomic displacement? Phonons connect the curvature of the potential-energy surface to collective vibrations. A stable structure has restoring forces near its local minimum. An imaginary harmonic frequency can indicate a structural instability, but numerical force noise, inadequate supercell size or a poorly relaxed reference can produce a similar symptom. This lesson teaches how to distinguish these possibilities before interpreting a dispersion plot. It uses bulk silicon and the matched data family from the first LCAO case. The example supercell is a proposed starting point, not a converged physical result.
Study the workflow →6.2 NVE trajectories and energy drift
Does the numerical trajectory approximately conserve the energy of an isolated model system? NVE fixes particle count, volume and total energy. Potential and kinetic energy exchange as atoms vibrate; neither part must remain individually constant. The diagnostic is the sum. This lesson uses a relaxed silicon supercell with all relevant atoms mobile, fixed cell vectors and the electronic model established earlier. It is a short integrator diagnostic rather than a production thermal-property calculation. No trajectory or drift value has been computed for this course.
Study the workflow →6.3 NVT sampling and thermostat choices
How can a fixed-volume simulation sample a chosen temperature without confusing temperature control with physical correctness? NVT allows energy exchange with a heat bath while holding particle number and cell volume fixed. The instantaneous kinetic temperature fluctuates even in equilibrium. A perfectly flat temperature trace is not the goal. This lesson compares a Nose–Hoover chain protocol with the conservation diagnostic in the preceding NVE lesson. It uses the same silicon supercell and force-converged electronic model, and provides an unexecuted setup rather than an equilibrium claim.
Study the workflow →6.4 Analyze trajectories with units and uncertainty
Which conclusions can be justified by a finite atomic trajectory? A simulation contains many frames, but adjacent frames are correlated and may not constitute independent samples. This lesson turns the validated silicon AIMD workflow into a reproducible analysis pipeline. It separates coordinate handling, physical observables and statistical uncertainty. A short solid-state trajectory can diagnose stability and vibrations; it cannot automatically measure diffusion or a reaction rate. The displayed definitions and analysis template do not contain simulated data or claimed results.
Study the workflow →7 Version-gated electronic methods
7.1 DFT+U and correlated-subspace choices
DFT+U and correlated-subspace choices
Study the workflow →7.2 Hybrid functionals and exact-exchange controls
Hybrid functionals and exact-exchange controls
Study the workflow →7.3 Spin–orbit coupling and noncollinear states
Spin–orbit coupling and noncollinear states
Study the workflow →7.4 Dispersion corrections as a controlled comparison
Dispersion corrections as a controlled comparison
Study the workflow →8 Performance, restarts and machine learning
8.1 Benchmark MPI, threads and solvers
Benchmark MPI, threads and solvers
Study the workflow →8.2 Restart safely and archive provenance
Restart safely and archive provenance
Study the workflow →8.3 DeePKS inference and domain validation
DeePKS inference and domain validation
Study the workflow →8.4 Export first-principles labels for ML workflows
Export first-principles labels for ML workflows
Study the workflow →Completion criteria
Preserve inputs, raw outputs, executable/build information and the provenance of pseudopotentials and numerical orbitals. Demonstrate target-property convergence and physical consistency separately from successful process exit. Report measured results only after your own validated calculations.
Official and university-linked teaching references
The common keyword baseline is ABACUS 3.9.0, not a claim that this is the newest release. Pin the installed executable; newer releases, advanced features and build-dependent workflows need their own compatibility checks. The USTC index links the community teaching archive below; teaching examples supplement the version-specific manual.