File Units.hpp

Defines

ACTS_DEFINE_UNIT_LITERAL(name)
namespace Acts

Note

This file is foreseen for the Geometry module to replace Extent

namespace PhysicalConstants

Physical constants in native units.

Unit constants are intentionally not listed.

Variables

constexpr double hbar = 6.582119569509066e-25 * UnitConstants::GeV * UnitConstants::s

Reduced Planck constant h/2*pi.

Computed from CODATA 2018 constants to double precision.

namespace UnitConstants

All physical quantities have both a numerical value and a unit. For the computations we always choose a particular unit for a given physical quantity so we only need to consider the numerical values as such. The chosen base unit for a particular physical dimension, e.g. length, time, or energy, within this code base is called the native unit. The base units should be chosen such that they are internally consistent, require the least amount of explicit conversion factors (ideally none at all), and have typical numerical values close to unity to reduce numerical issues.

Here, the following native units are used:

  • Length is expressed in mm.

  • Time is expressed in [speed-of-light * time] == mm. A consequence of this choice is that the speed-of-light expressed in native units is 1.

  • Angles are expressed in radian.

  • Energy, mass, and momentum are all expressed in GeV (consistent with speed-of-light == 1).

  • Electric charge is expressed in e, i.e. units of the elementary charge.

  • The magnetic field is expressed in GeV/(e*mm). The magnetic field connects momentum to length, e.g. in SI units the radius of a charged particle trajectory in a constant magnetic field is given by

    \[radius = - (momentum / charge) / field\]

    With the chosen magnetic field unit the expression above stays the same and no additional conversion factors are necessary.

  • Amount of substance is expressed in mol.

Depending on the context a physical quantity might not be given in the native units. In this case we need to convert to the native unit first before the value can be used. The necessary conversion factors are defined in Acts::UnitConstants. Multiplying a value with the unit constant produces the equivalent value in the native unit, e.g.

double length = 1 * Acts::UnitConstants::m;       // length == 1000.0
double momentum = 100 * Acts::UnitConstants::MeV; // momentum == 0.1

The conversion of a value in native units into the equivalent value in a specific other unit is computed by dividing with the relevant unit, e.g.

double length = 10.0;                               // native units mm
double lengthInM = length / Acts::UnitConstants::m; // == 0.01;

To further simplify the usage, physical quantities can also be expressed via C++ user literals defined in Acts::UnitLiterals. This allows us to express quantities in a concise way:

using namespace Acts::UnitLiterals;

double length = 1_m;                     // == 1000.0
double momentum = 1.25_TeV;              // == 1250.0
double lengthInUm = length / 1_um;       // == 1000000.0
double momentumInMeV = momentum / 1_MeV; // == 1250000.0

To ensure consistent computations and results the following guidelines must be followed when handling physical quantities with units:

  • All unqualified numerical values, i.e. without a unit, are assumed to be expressed in the relevant native unit, e.g. mm for lengths or GeV for energy/momentum.

  • If a variable stores a physical quantity in a specific unit that is not the native unit, clearly mark this in the variable, i.e.

    double momentum = 100.0; // momentum is stored as native unit GeV
    double momentumInMeV = 10.0; // would be 0.01 in native units
    
  • All input values must be given as numerical_value * unit_constant or equivalently using the unit literals as value_unit. The resulting unqualified numerical value will be automatically converted to the native unit.

  • To output an unqualified numerical value in the native units as a numerical value in a specific unit divide by the unit constants as numerical_value / unit_constant or using the unit literals as value / 1_unit.

Examples:

#include <Acts/include/Definitions/Units.hpp>
using namespace Acts::UnitLiterals;

// define input values w/ units (via unit constants)
double width    = 12 * Acts::UnitConstants::mm;
double mmuon    = 105.7 * Acts::UnitConstants::MeV;
// define input values w/ units (via unit user literals)
double length   = 23_cm;
double time     = 1214.2_ns;
double angle    = 123_degree;
double momentum = 2.5_TeV;
double mass     = 511_keV;
double velocity = 345_m / 1_s;
double bfield   = 3.9_T;

// convert output values (via unit constants)
doube t_in_ns    = trackPars.time() / Acts::UnitConstants::ns;
// convert output values (via unit user literals)
double x_in_mm   = trackPars.position()[ePos0] / 1_mm;
double p_in_TeV = trackPars.absoluteMomentum() / 1_TeV;

Note

A helper script is available in Core/scripts/print_units_physical_constants.py to validate some of the numerical values.

Warning

Since using user-defined literals requires a namespace import of Acts::UnitLiterals it should not be used in headers since it would (accidentally) modify the namespace wherever the header is included.

Variables

constexpr double C = J / eV
constexpr double cm = 10.0
constexpr double cm2 = cm * cm
constexpr double cm3 = cm * cm * cm
constexpr double degree = 0.017453292519943295
constexpr double e = 1.0
constexpr double eV = 1e-9
constexpr double fm = 1e-12
constexpr double fs = 1e-15 * s
constexpr double g = 1.0 / 1.782662e-24
constexpr double Gauss = 1e-4 * T
constexpr double GeV = 1.0
constexpr double h = 3600.0 * s
constexpr double J = 6241509074.460763 * GeV
constexpr double keV = 1e-6
constexpr double kg = 1.0 / 1.782662e-27
constexpr double kGauss = 1e-1 * T
constexpr double km = 1e6
constexpr double m = 1e3
constexpr double m2 = m * m
constexpr double m3 = m * m * m
constexpr double MeV = 1e-3
constexpr double min = 60.0 * s
constexpr double mm = 1.0
constexpr double mm2 = mm * mm
constexpr double mm3 = mm * mm * mm
constexpr double mol = 1.0
constexpr double mrad = 1e-3
constexpr double ms = 1e-3 * s
constexpr double nm = 1e-6
constexpr double ns = 1e-9 * s
constexpr double pm = 1e-9
constexpr double ps = 1e-12 * s
constexpr double rad = 1.0
constexpr double s = 299792458000.0
constexpr double T = 0.000299792458
constexpr double TeV = 1e3
constexpr double u = 0.93149410242
constexpr double um = 1e-3
constexpr double us = 1e-6 * s
namespace UnitLiterals