class rgpot::StillingerWeberSi

Overview

Stillinger-Weber silicon with the 1985 published parameters. More…

#include <SWSi.hpp>
 
class StillingerWeberSi: public rgpot::Potential< StillingerWeberSi > {
public:
    // fields
 
    static double kEpsilon = 3.4723e-12* 1.0e-7 / 1.602176634e-19;
    static double kSigma = 2.0951;
    static double kA = 7.049556277;
    static double kB = 0.6022245584;
    static double kP = 4.0;
    static double kCut = 1.80;
    static double kLambda = 21.0;
    static double kGamma = 1.20;
 
    // construction
 
    StillingerWeberSi();
 
    // methods
 
    virtual PotCaps caps() const;
    virtual uint64_t paramsKey() const;
    void forceImpl(const ForceInput& in, ForceOut* out) const;
    void setNeighbourMode(nlist::Mode mode);
    void energyAndGradient(const double* positions, const int* atomic_numbers, const double* box, std::size_t n_atoms, double* energy, double* gradient, nlist::NeighborList* neighbors = nullptr) const;
    void energyAndGradientBatch(std::size_t n_config, const double* positions, const int* atomic_numbers, const double* box, std::size_t n_atoms, double* energies, double* gradients) const;
};

Detailed Documentation

Stillinger-Weber silicon with the 1985 published parameters.

One call evaluates a whole configuration. energyAndGradientBatch evaluates many configurations that share a cell and a species list, so a caller crosses into the library once per batch rather than once per pair.

Fields

static double kEpsilon = 3.4723e-12* 1.0e-7 / 1.602176634e-19

Eq. (2.9) prints 3.4723e-12 erg per atom pair, 2.16723920 eV. 1 erg = 1e-7 J. The CODATA electron-volt is 1.602176634e-19 J. PotType::SWSi is 2.3149945 eV, larger by 1.068176741337.

static double kSigma = 2.0951

0.20951 nm, Eq. (2.9).

static double kCut = 1.80

Reduced cutoff a, Eq. (2.7).

Methods

virtual PotCaps caps() const

Concurrency and evaluation capabilities of this potential.

Multi-threaded callers (NEB images, dimer endpoints) derive their sharing and cloning policy from this instead of hard-coded lists.

virtual uint64_t paramsKey() const

Fingerprint of the potential’s parameter set.

Mixed into the result-cache key so results never cross parameter sets. Implementations hash their config field by field (see ParamHash.hpp) plus a kernel-version salt that changes whenever the numerics change. The default (0) suits parameter-free potentials whose numerics never changed.

void forceImpl(const ForceInput& in, ForceOut* out) const

Forces, energy and the Cauchy stress sigma = (1/V) dE/dε in eV/Angstrom^3, the convention of stress.hpp. The virial sums each folded bond vector times the energy gradient with respect to it.

void setNeighbourMode(nlist::Mode mode)

Auto, Stored and Stream all use the shared neighbour layer. One thread agrees with the old ordered walk to within 1e-12.

void energyAndGradient(const double* positions, const int* atomic_numbers, const double* box, std::size_t n_atoms, double* energy, double* gradient, nlist::NeighborList* neighbors = nullptr) const

Energy and dE/dx for one configuration.

Parameters:

positions

Row-major n_atoms by 3 coordinates, Angstrom.

atomic_numbers

Length n_atoms. Unused: the 1985 set is silicon.

box

Row-major 3 by 3 cell, Angstrom. Periodic in every direction.

n_atoms

Atom count.

energy

Receives the potential energy, eV.

gradient

Receives dE/dx, length 3 * n_atoms, eV/Angstrom. The library force is the negation of this gradient.

neighbors

Optional Verlet list. Null uses the internal list, with a skin of one fifth of the cutoff (none in Stream mode). A list is reused while each atom has moved by at most half its skin and the cell is unchanged.

void energyAndGradientBatch(std::size_t n_config, const double* positions, const int* atomic_numbers, const double* box, std::size_t n_atoms, double* energies, double* gradients) const

Energy and dE/dx for a batch of configurations.

Parameters:

n_config

Number of configurations.

positions

n_config blocks of n_atoms * 3 coordinates.

atomic_numbers

Species shared by every configuration.

box

Cell shared by every configuration.

n_atoms

Atom count of each configuration.

energies

Length n_config.

gradients

Length n_config * n_atoms * 3.