492
J. Apostolakis
Most tools also account for the effects of external electromagnetic fields on charged
particles, provide visualisation of the geometry and simulated events. They provide
for tallies, output of key physical quantities calculated during the simulation, such
as totals of energy deposition, dose, particle flux and fluence. They also provide the
opportunity for the user’s code to filter and record track quantities at each step.
The geometry module provides the ability to describe the material composition
and the geometry of the setup in terms of volumes. The tool must be able to navigate
inside this volume description, identify the volume in which a point is located and
calculate the distance to the next boundary in a given direction. The capabilities of
the geometry modeler determine the type of volumes and their relative placement:
whether volumes are generated directly as finite shapes, or whether they are the
result of the intersection of surfaces; whether all volumes must be placed within a
single ‘world’ volume, or a hierarchy of volumes can be created. In order to simulate
a large, complex detector with hundreds of thousands to millions of volumes, the
geometry modeler needs to support hierarchical geometry definitions.
To ensure good performance the key geometry operations must be computationally efficient; in particular, the computation of the distance to intersecting a
boundary is critical. Optimisation methods which rely on data precomputed at
initialization inspired by ray-tracing are used to greatly reduce computation time
wherever many candidate sub-volumes exist.
Some experiments have chosen to use a geometry modeler external to the
simulation tool. They use the same geometry description and modeler inside a
Virtual Monte Carlo framework [34]. This interfaces to different simulation tools
for modeling interactions: GEANT, FLUKA and GEANT4; they are labeled ‘physics
engines’ and can be selected at runtime.
11.3.1 External Fields
The effects of external electromagnetic fields on the trajectory and energy of a
charged particle track are modeled utilizing the Lorentz equation. The equations
of motion for the position, the momentum and optionally the polarisation of the
particle are integrated to obtain the position and state of a particle after a distance
s. In special cases, such as a constant magnetic field, an analytical solution can be
used. In the general case, numerical integration is used, typically with a Runge-Kutta
method.
After integration the idealized curved path of a particle track in a magnetic field
is propagated through the geometry of a detector. The curved trajectory is split
into linear chord segments, which are used to navigate in the model geometry.
The intersection of a chord segment is progressively refined to identify the location
where the curved path crosses a geometry boundary.
J. Apostolakis
Most tools also account for the effects of external electromagnetic fields on charged
particles, provide visualisation of the geometry and simulated events. They provide
for tallies, output of key physical quantities calculated during the simulation, such
as totals of energy deposition, dose, particle flux and fluence. They also provide the
opportunity for the user’s code to filter and record track quantities at each step.
The geometry module provides the ability to describe the material composition
and the geometry of the setup in terms of volumes. The tool must be able to navigate
inside this volume description, identify the volume in which a point is located and
calculate the distance to the next boundary in a given direction. The capabilities of
the geometry modeler determine the type of volumes and their relative placement:
whether volumes are generated directly as finite shapes, or whether they are the
result of the intersection of surfaces; whether all volumes must be placed within a
single ‘world’ volume, or a hierarchy of volumes can be created. In order to simulate
a large, complex detector with hundreds of thousands to millions of volumes, the
geometry modeler needs to support hierarchical geometry definitions.
To ensure good performance the key geometry operations must be computationally efficient; in particular, the computation of the distance to intersecting a
boundary is critical. Optimisation methods which rely on data precomputed at
initialization inspired by ray-tracing are used to greatly reduce computation time
wherever many candidate sub-volumes exist.
Some experiments have chosen to use a geometry modeler external to the
simulation tool. They use the same geometry description and modeler inside a
Virtual Monte Carlo framework [34]. This interfaces to different simulation tools
for modeling interactions: GEANT, FLUKA and GEANT4; they are labeled ‘physics
engines’ and can be selected at runtime.
11.3.1 External Fields
The effects of external electromagnetic fields on the trajectory and energy of a
charged particle track are modeled utilizing the Lorentz equation. The equations
of motion for the position, the momentum and optionally the polarisation of the
particle are integrated to obtain the position and state of a particle after a distance
s. In special cases, such as a constant magnetic field, an analytical solution can be
used. In the general case, numerical integration is used, typically with a Runge-Kutta
method.
After integration the idealized curved path of a particle track in a magnetic field
is propagated through the geometry of a detector. The curved trajectory is split
into linear chord segments, which are used to navigate in the model geometry.
The intersection of a chord segment is progressively refined to identify the location
where the curved path crosses a geometry boundary.