EIC code displayed by LXR |
|
|
|
|||
Warning, file /include/celeritas/em/distribution/EnergyLossGaussianDistribution.hh was not indexed or was modified since last indexation (in which case cross-reference links may be missing, inaccurate or erroneous).
0001 //------------------------------- -*- C++ -*- -------------------------------// 0002 // Copyright Celeritas contributors: see top-level COPYRIGHT file for details 0003 // SPDX-License-Identifier: (Apache-2.0 OR MIT) 0004 //---------------------------------------------------------------------------// 0005 //! \file celeritas/em/distribution/EnergyLossGaussianDistribution.hh 0006 //---------------------------------------------------------------------------// 0007 #pragma once 0008 0009 #include <cmath> 0010 0011 #include "corecel/Assert.hh" 0012 #include "corecel/Macros.hh" 0013 #include "corecel/Types.hh" 0014 #include "corecel/math/Algorithms.hh" 0015 #include "corecel/random/distribution/NormalDistribution.hh" 0016 0017 #include "EnergyLossHelper.hh" 0018 0019 namespace celeritas 0020 { 0021 //---------------------------------------------------------------------------// 0022 /*! 0023 * Sample energy loss from a Gaussian distribution. 0024 * 0025 * In a thick absorber, the total energy transfer is a result of many small 0026 * energy losses from a large number of collisions. The central limit theorem 0027 * applies, and the energy loss fluctuations can be described by a Gaussian 0028 * distribution. See section 7.3.1 of the Geant4 Physics Reference Manual and 0029 * GEANT3 PHYS332 section 2.3. 0030 * 0031 * The Gaussian approximation is valid for heavy particles and in the 0032 * regime \f$ \kappa = \xi / T_\textrm{max} > 10 \f$. 0033 * Fluctuations of the unrestricted energy loss 0034 * follow a Gaussian distribution if \f$ \Delta E > \kappa T_{max} \f$, 0035 * where \f$ T_{max} \f$ is the maximum energy transfer (PHYS332 section 0036 * 2). For fluctuations of the \em restricted energy loss, the condition is 0037 * modified to \f$ \Delta E > \kappa T_{c} \f$ and \f$ T_{max} \le 2 T_c 0038 * \f$, where \f$ T_c \f$ is the delta ray cutoff energy (PRM Eq. 7.6-7.7). 0039 */ 0040 class EnergyLossGaussianDistribution 0041 { 0042 public: 0043 //!@{ 0044 //! \name Type aliases 0045 using Energy = units::MevEnergy; 0046 using EnergySq = RealQuantity<UnitProduct<units::Mev, units::Mev>>; 0047 //!@} 0048 0049 public: 0050 // Construct from distribution parameters 0051 inline CELER_FUNCTION 0052 EnergyLossGaussianDistribution(Energy mean_loss, Energy bohr_stddev); 0053 0054 // Construct from helper distribution parameters 0055 inline CELER_FUNCTION 0056 EnergyLossGaussianDistribution(Energy mean_loss, EnergySq bohr_var); 0057 0058 // Construct from helper-calculated data 0059 explicit inline CELER_FUNCTION 0060 EnergyLossGaussianDistribution(EnergyLossHelper const& helper); 0061 0062 // Sample energy loss according to the distribution 0063 template<class Generator> 0064 inline CELER_FUNCTION Energy operator()(Generator& rng); 0065 0066 private: 0067 real_type const max_loss_; 0068 NormalDistribution<real_type> sample_normal_; 0069 }; 0070 0071 //---------------------------------------------------------------------------// 0072 // INLINE DEFINITIONS 0073 //---------------------------------------------------------------------------// 0074 /*! 0075 * Construct from mean/stddev. 0076 * 0077 * This formulation is used internally by the Urban distribution. 0078 */ 0079 CELER_FUNCTION EnergyLossGaussianDistribution::EnergyLossGaussianDistribution( 0080 Energy mean_loss, Energy bohr_stddev) 0081 : max_loss_(2 * mean_loss.value()) 0082 , sample_normal_(mean_loss.value(), bohr_stddev.value()) 0083 0084 { 0085 CELER_EXPECT(mean_loss > zero_quantity()); 0086 CELER_EXPECT(bohr_stddev > zero_quantity()); 0087 } 0088 0089 //---------------------------------------------------------------------------// 0090 /*! 0091 * Construct from distribution parameters. 0092 * 0093 * The mean loss is the energy lost over the step, and the standard deviation 0094 * is the square root of Bohr's variance (PRM Eq. 7.8). For thick absorbers, 0095 * the straggling function approaches a Gaussian distribution with this 0096 * standard deviation. 0097 */ 0098 CELER_FUNCTION EnergyLossGaussianDistribution::EnergyLossGaussianDistribution( 0099 Energy mean_loss, EnergySq bohr_var) 0100 : EnergyLossGaussianDistribution{mean_loss, 0101 Energy{std::sqrt(bohr_var.value())}} 0102 { 0103 } 0104 0105 //---------------------------------------------------------------------------// 0106 /*! 0107 * Construct from helper-calculated data. 0108 */ 0109 CELER_FUNCTION EnergyLossGaussianDistribution::EnergyLossGaussianDistribution( 0110 EnergyLossHelper const& helper) 0111 : EnergyLossGaussianDistribution{helper.mean_loss(), helper.bohr_variance()} 0112 { 0113 } 0114 0115 //---------------------------------------------------------------------------// 0116 /*! 0117 * Sample energy loss according to the distribution. 0118 */ 0119 template<class Generator> 0120 CELER_FUNCTION auto 0121 EnergyLossGaussianDistribution::operator()(Generator& rng) -> Energy 0122 { 0123 real_type result; 0124 do 0125 { 0126 result = sample_normal_(rng); 0127 } while (result <= 0 || result > max_loss_); 0128 return Energy{result}; 0129 } 0130 0131 //---------------------------------------------------------------------------// 0132 } // namespace celeritas
| [ Source navigation ] | [ Diff markup ] | [ Identifier search ] | [ general search ] |
|
This page was automatically generated by the 2.3.7 LXR engine. The LXR team |
|