EIC code displayed by LXR master/​include/​celeritas/​phys/​PhysicsStepUtils.hh	Source navigation Diff markup Identifier search General search
	Version: master test Revert   Change

File indexing completed on 2026-01-06 10:05:28

 //------------------------------- -*- C++ -*- -------------------------------//
 // Copyright Celeritas contributors: see top-level COPYRIGHT file for details
 // SPDX-License-Identifier: (Apache-2.0 OR MIT)
 //---------------------------------------------------------------------------//
 //! \file celeritas/phys/PhysicsStepUtils.hh
 //! \brief Helper functions for physics step processing
 //---------------------------------------------------------------------------//
 #pragma once
 
 #include "corecel/Assert.hh"
 #include "corecel/Types.hh"
 #include "corecel/cont/Range.hh"
 #include "corecel/math/Algorithms.hh"
 #include "corecel/math/NumericLimits.hh"
 #include "corecel/random/distribution/GenerateCanonical.hh"
 #include "corecel/random/distribution/Selector.hh"
 #include "celeritas/Types.hh"
 #include "celeritas/grid/EnergyLossCalculator.hh"
 #include "celeritas/grid/InverseRangeCalculator.hh"
 #include "celeritas/grid/RangeCalculator.hh"
 #include "celeritas/grid/SplineCalculator.hh"
 #include "celeritas/grid/XsCalculator.hh"
 #include "celeritas/mat/MaterialTrackView.hh"
 
 #include "ParticleTrackView.hh"
 #include "PhysicsStepView.hh"
 #include "PhysicsTrackView.hh"
 
 namespace celeritas
 {
 //---------------------------------------------------------------------------//
 /*!
  * Sample the process for the discrete interaction.
  *
  * - If the particle is at rest and has an at-rest process, that is returned.
  * - Sample a process by selecting from probabilities proportional to the cross
  *   sections calculated at the beginning of the step.
  * - If the particle has changed energy over the step by continuous energy
  *   loss, and the process supports integral cross section rejection, then the
  *   cross section is recalculated with the new energy. The collision is
  *   sampled based by rejection sampling with the updated cross section.
  *
  * \todo Support competing at-rest processes?
  *
  * \return Sampled process, or null if integral xs rejection cancels
  * interaction
  */
 template<class Engine>
 inline CELER_FUNCTION ParticleProcessId
 find_ppid(MaterialView const& material,
           ParticleTrackView const& particle,
           PhysicsTrackView const& physics,
           PhysicsStepView& pstep,
           Engine& rng)
 {
     if (physics.at_rest_process() && particle.is_stopped())
     {
         // If the particle is stopped and has an at-rest process, select it
         return physics.at_rest_process();
     }
 
     // Sample the process from the pre-calculated per-process cross section
     ParticleProcessId ppid = celeritas::make_selector(
         [&pstep](ParticleProcessId ppid) { return pstep.per_process_xs(ppid); },
         ParticleProcessId{physics.num_particle_processes()},
         pstep.macro_xs())(rng);
     CELER_ASSERT(ppid);
 
     // Determine if the discrete interaction occurs for particles with energy
     // loss processes
     if (physics.integral_xs_process(ppid))
     {
         // Recalculate the cross section at the post-step energy \f$ E_1 \f$
         real_type xs = physics.calc_xs(ppid, material, particle.energy());
 
         // The discrete interaction occurs with probability \f$ \sigma(E_1) /
         // \sigma_{\max} \f$. Note that it's possible for \f$ \sigma(E_1) \f$
         // to be larger than the estimate of the maximum cross section over the
         // step \f$ \sigma_{\max} \f$.
         if (generate_canonical(rng) * pstep.per_process_xs(ppid) > xs)
         {
             // No interaction occurs; reset the physics state and continue
             // tracking
             return {};
         }
     }
     return ppid;
 }
 
 //---------------------------------------------------------------------------//
 /*!
  * Calculate physics step limits based on cross sections and range limiters.
  *
  * Processes with integral cross section rejection calculate an estimated
  * "maximum" over the step: see PhysicsTrackView.calc_max_xs .
  *
  * \todo For particles with decay, macro XS calculation will incorporate
  * decay probability, dividing decay constant by speed to become 1/len to
  * compete with interactions.
  *
  * \todo For neutral particles that haven't changed material since the last
  * step, we can reuse the previously calculated cross section (but may not want
  * to if it's expensive).
  */
 inline CELER_FUNCTION StepLimit
 calc_physics_step_limit(MaterialTrackView const& material,
                         ParticleTrackView const& particle,
                         PhysicsTrackView& physics,
                         PhysicsStepView& pstep)
 {
     CELER_EXPECT(physics.has_interaction_mfp());
 
     // Loop over all processes that apply to this track (based on particle
     // type) and calculate cross section and particle range.
     real_type total_macro_xs = 0;
     for (auto ppid : range(ParticleProcessId{physics.num_particle_processes()}))
     {
         real_type process_xs = 0;
         if (auto const& process = physics.integral_xs_process(ppid))
         {
             // If the integral approach is used and this particle has an energy
             // loss process, estimate the maximum cross section over the step
             process_xs = physics.calc_max_xs(
                 process, ppid, material.material_record(), particle.energy());
         }
         else
         {
             // Calculate the macroscopic cross section for this process
             process_xs = physics.calc_xs(
                 ppid, material.material_record(), particle.energy());
         }
         // Accumulate process cross section into the total cross section and
         // save it for later
         total_macro_xs += process_xs;
         pstep.per_process_xs(ppid) = process_xs;
     }
     pstep.macro_xs(total_macro_xs);
     CELER_ASSERT(total_macro_xs > 0 || !particle.is_stopped());
 
     // Determine limits from discrete interactions
     StepLimit limit;
     limit.action = physics.scalars().discrete_action();
     if (particle.is_stopped())
     {
         limit.step = 0;
     }
     else
     {
         limit.step = physics.interaction_mfp() / total_macro_xs;
         if (auto grid_id = physics.range_grid())
         {
             auto calc_range = physics.make_calculator<RangeCalculator>(grid_id);
             real_type range = calc_range(particle.energy());
             // Save range for the current step and reuse it elsewhere
             physics.dedx_range(range);
 
             // Convert to the scaled range
             real_type eloss_step = physics.range_to_step(range);
             if (eloss_step <= limit.step)
             {
                 limit.step = eloss_step;
                 limit.action = physics.scalars().range_action();
             }
 
             // Limit charged particle step size
             real_type fixed_limit = physics.scalars().fixed_step_limiter;
             if (fixed_limit > 0 && fixed_limit < limit.step)
             {
                 limit.step = fixed_limit;
                 limit.action = physics.scalars().fixed_step_action;
             }
         }
         else if (physics.num_particle_processes() == 0)
         {
             // Clear post-step action so that unknown particles don't interact
             limit.action = {};
         }
     }
 
     return limit;
 }
 
 //---------------------------------------------------------------------------//
 /*!
  * Calculate mean energy loss over the given "true" step length.
  *
  * Stopping power is an integral over low-exiting-energy
  * secondaries. Above some threshold energy \em T_c we treat exiting
  * secondaries discretely; below it, we lump them into this continuous loss
  * term that varies based on the energy, the atomic number density, and the
  * element number:
  * \f[
  *   \difd{E}{x} = N_Z \int_0^{T_c} \frac{d \sigma_Z(E, T)}{dT} T \dif T
  * \f]
  * Here, the cross section is a function of the primary's energy \em E and the
  * exiting secondary energy \em T.
  *
  * The stopping range \em R due to these low-energy processes is an integral
  * over the inverse of the stopping power: basically the distance that will
  * take a particle (without discrete processes at play) from its current energy
  * to an energy of zero.
  * \f[
  *   R = \int_0 ^{E_0} - \difd{E}{x} \dif E .
  * \f]
  *
  * \note The inverse range correction assumes range is always the integral of
  * the stopping power/energy loss.
  *
  * \todo The GEANT3 manual \cite{geant3-1993} makes the point that linear
  * interpolation of energy
  * loss rate results in a piecewise constant energy deposition curve, which is
  * why they use spline interpolation. Investigate higher-order reconstruction
  * of energy loss curve, e.g. through spline-based interpolation or log-log
  * interpolation.
  *
  * \note See section 7.2.4 Run Time Energy Loss Computation of the Geant4
  * physics manual \cite{g4prm}. See also the longer discussions in section 8
  * of PHYS010 of the Geant3 manual.
  *
  * Zero energy loss can occur in the following cases:
  * - The energy loss value at the given energy is zero (e.g. high energy
  *   particles)
  * - The Urban model is selected and samples zero collisions (possible in thin
  *   materials and/or small steps)
  */
 inline CELER_FUNCTION ParticleTrackView::Energy
 calc_mean_energy_loss(ParticleTrackView const& particle,
                       PhysicsTrackView const& physics,
                       real_type step)
 {
     CELER_EXPECT(step > 0);
     using Energy = ParticleTrackView::Energy;
 
     Energy const pre_step_energy = particle.energy();
 
     // Calculate the sum of energy loss rate over all processes.
     Energy eloss;
     {
         auto grid_id = physics.energy_loss_grid();
         CELER_ASSERT(grid_id);
 
         auto calc_eloss_rate
             = physics.make_calculator<EnergyLossCalculator>(grid_id);
         eloss = Energy{step * calc_eloss_rate(pre_step_energy)};
     }
 
     if (eloss >= pre_step_energy * physics.scalars().linear_loss_limit)
     {
         // Enough energy is lost over this step that the dE/dx linear
         // approximation is probably wrong. Use the definition of the range as
         // the integral of 1/loss to back-calculate the actual energy loss
         // along the curve given the actual step.
 
         // Use the range limit stored from calc_physics_step_limit
         real_type range = physics.dedx_range();
         if (step == range)
         {
             // NOTE: eloss should be pre_step_energy at this point only if the
             // range was the step limiter (step == range), and if the
             // range-to-step conversion was 1.
             return pre_step_energy;
         }
         CELER_ASSERT(range > step);
 
         // Calculate energy along the range curve corresponding to the actual
         // step taken: this gives the exact energy loss over the step due to
         // this process. If the step is very near the range (a few ULP off, for
         // example), then the post-step energy will be calculated as zero
         // without going through the condition above.
         auto calc_energy = physics.make_calculator<InverseRangeCalculator>(
             physics.inverse_range_grid());
         eloss = pre_step_energy - calc_energy(range - step);
 
         // Spline interpolation does not ensure roundtrip consistency between
         // the range and the inverse. This can lead to negative values for the
         // energy loss
         eloss = clamp_to_nonneg(eloss);
     }
 
     CELER_ENSURE(eloss >= zero_quantity());
     return eloss;
 }
 
 //---------------------------------------------------------------------------//
 /*!
  * Choose the physics model for a track's pending interaction.
  *
  * - Sample from the previously calculated per-process cross section/decay to
  *   determine the interacting process ID.
  * - From the process ID and (post-slowing-down) particle energy, we obtain the
  *   applicable model ID.
  * - For energy loss (continuous-discrete) processes, the post-step energy will
  *   be different from the pre-step energy, so the assumption that the cross
  *   section is constant along the step is no longer valid. Use the "integral
  *   approach" to sample the discrete interaction from the correct probability
  *   distribution (section 7.4 of the Geant4 Physics Reference release 10.6).
  */
 template<class Engine>
 CELER_FUNCTION ActionId
 select_discrete_interaction(MaterialView const& material,
                             ParticleTrackView const& particle,
                             PhysicsTrackView const& physics,
                             PhysicsStepView& pstep,
                             Engine& rng)
 {
     // Nonzero MFP to interaction -- no interaction model
     CELER_EXPECT(physics.interaction_mfp() <= 0);
     CELER_EXPECT(pstep.macro_xs() > 0);
 
     // Sample the process from the pre-calculated cross sections
     auto ppid = find_ppid(material, particle, physics, pstep, rng);
     if (!ppid)
     {
         return physics.scalars().integral_rejection_action();
     }
 
     // Find the model that applies at the particle energy
     auto find_model = physics.make_model_finder(ppid);
     auto pmid = find_model(particle.energy());
 
     ElementComponentId elcomp_id{};
     if (material.num_elements() == 1)
     {
         elcomp_id = ElementComponentId{0};
     }
     else if (auto table_id = physics.cdf_table(pmid))
     {
         // Sample an element for discrete interactions that require it and for
         // materials with more than one element
         auto select_element
             = physics.make_element_selector(table_id, particle.energy());
         elcomp_id = select_element(rng);
     }
     pstep.element(elcomp_id);
 
     return physics.model_to_action(physics.model_id(pmid));
 }
 
 //---------------------------------------------------------------------------//
 }  // namespace celeritas

This page was automatically generated by the 2.3.7 LXR engine. The LXR team