Geant4 9.3 Release Notes

December 18th, 2009

The code and binary libraries for the supported systems are available through our Source Code Web page.

We are grateful for the efforts of Geant4 users who have provided detailed feedback or comprehensive reports of issues. We thank in particular those who have contributed corrections, improvements or developments included in this release.

Please refer to the Geant4 User Documentation for further information about using Geant4.

Contents

1. Supported and Tested Platforms
2. CLHEP and AIDA
3. Items for migration of the user code
4. New Developments and Capabilities
5. Expected effects on physics and performance
6. Known Run-Time Problems and Limitations
7. Compilation Warnings
8. Known Run-Time Warnings
9. Geant4 Software License
10. Detailed list of changes and fixes

1. Supported and Tested Platforms

• Linux, gcc-4.1.2 and gcc-4.3.2 (SLC5).
  Tested on 32 and 64 bit architectures (Intel or AMD) with Scientific Linux CERN 4 (SLC4) (based on RedHat Linux Enterprise 4) and Scientific Linux CERN 5 (SLC5) (based on RedHat Linux Enterprise 5). Geant4 has also been successfully compiled on other RedHat versions as well as Debian and Suse.
• MacOSX 10.5 with gcc-4.0.1 and 10.6 with gcc-4.2.1
• Windows/XP and CygWin Tools with: Visual C++ 9.0 (Visual Studio 2008)

• Linux, gcc-4.4.2, gcc-4.3.2, gcc-3.4.6 (SLC4)
• Linux, Intel-icc 11.1

• AIX 4.3.2, xlC 6.0
• DEC V4.0, cxx C++ V6.1-027
• HP 10.20, aCC C++ B3910B A.01.23
• MacOSX 10.4, gcc-3.3
• SGI V6.5.5, CC 7.2.1
• SUN Solaris 5.8, C++ CC-5.5.

2. CLHEP and AIDA

3. Items for migration of the user code

• A new method CheckParticleList() has been added to G4VUserPhysicsList to check the consistency of the list of defined particles before instantiating the physics processes. For example, e+, e- and gamma should exist if any electromagnetic process is simulated... A fatal exception will be thrown if any particle type mandatory for the simulation is missing.
  It is possible to skip this check, by making use of the method DisableCheckParticleList() from G4VUserPhysicsList.

• All Low Energy Electromagnetic Physics process classes have been entirely redesigned (migrated) in the Geant4 toolkit in order to offer to users a coherent approach for the modeling of electromagnetic physical interactions. The software design is now common to both 'standard' and 'low energy' electromagnetic modules. These changes affect Livermore, Penelope, ICRU'73 ion and Geant4-DNA Physics. Users are invited to modify their own physics-list(s) in order to use these new processes and models. The web page indicated below describe how to use these new processes and models, with examples of physics list implementations and new features of the adopted common design:
  https://twiki.cern.ch/twiki/bin/view/Geant4/LoweMigratedProcesses Please note that it is still possible to use previous implementation of the low energy processes for backward compatibility, however, these models are no longer maintained and will be removed in the near future. A warning message will also be displayed when running older low energy model versions.

• It is recommended to migrate user custom physics-lists by replacing the G4MultipleScattering physics process with the following:
  • G4eMultipleScattering for e+-;
  • G4MuMultipleScattering for muons;
  • G4hMultipleScattering for hadrons and ions.
  The former G4MultipleScattering still exists in the Geant4 distribution but will be deleted in a future release.
  Nuclear stopping power is no longer part of ionization processes, and is now responsible only for electronic stopping power; for simulation of nuclear stopping, a new process G4NuclearStopping should be added to custom physics-lists.

• Qualification for several physics-lists have changed. Several are now declared obsolete. A new category of 'experimental' is used for newer physics-lists or ones with new physics options. Some physics-lists have been removed. See in Section 10 for the details.

• The OpenGL Print EPS facility now always depends on the compression library, zlib. An internal version of this library is provided for users whose system does not provide the library libz. To use this internal version, set the environment variable G4LIB_USE_ZLIB. Users who build with the standard Geant4 Configure script should not have to worry about this as the environment will automatically be set up correctly.

• New data for measured optical surface reflectance, RealSurface.1.0.
• New low-energy data set, G4EMLOW.6.9.

4. New Developments and Capabilities

• A new stepper G4NystromRK4 offers better computing performance in integrating the trajectories of charged particles in a magnetic field. It uses the standard Nystrom method and a novel analytical estimation of the integration error. This reduces greatly the number of field calls per integration step (from 10 to 3). It achieves comparable accuracy with G4ClassicalRK4 in test cases. It can be used only for pure magnetic fields which are not time-dependent.
• A new class G4CachedMagneticField caches the value of a magnetic field class, in order to reduce the number of calls to an expensive field calculation or interpolation method. For positions within the chosen radius of the previous position where it evaluated the field, it returns the last value of the field is automatically. It can be used only for pure magnetic fields which are not time-dependent.

• GDML: it is now possible to customise the reader as well as the writer to treat ad-hoc user extensions to the GDML schema. The GDML writer now also supports dumping of optical surface properties associated with surfaces and materials.

• Software migration: as explained in section 3, all low energy electromagnetic processes and models are now adopting the same design as for the Standard Electromagnetic Physics, for a coherent approach of the modeling of electromagnetic physical interactions.
• Livermore Physics: new G4LivermorePolarizedPhotoElectricModel and G4LivermorePolarizedGammaConversionModel are available for the simulation of polarized photon interactions.
• ICRU'73 ion model : the upper energy limit of ICRU'73 stopping power tables was extended to 1 GeV/nucleon (compared to 10 MeV/nucleon in the previous release). Revised tables were included for G4_WATER, G4_WATER_VAPOR, G4_Cu, G4_NYLON-6-6 (all tables provided by P. Sigmund/Univ. of Southern Denmark).
• Geant4-DNA Physics:
  • Energy range coverage of several Geant4-DNA Physics models has been extended: electron elastic Champion's model from 7.4 eV up to 10 MeV; electron ionization Born's model from 11 eV up to 1 MeV; proton excitation Born's model up to 100 MeV; proton ionization Born's model up to 100 MeV.
  • Geant4-DNA models are valid for liquid water only and use now the G4_WATER NIST material as liquid water material definition. Users should define liquid water as G4_WATER in their application(s).
• Data set: G4EMLOW version 6.9 is required.

• The set of multiple-scattering models has been expanded:
  • G4UrbanMscModel92 is the default model for e+/e- equivalent to the Urban model included in the previous release 9.2. It is used as the default and in the Option1 electromagnetic physics constructors for physics-lists.
  • G4UrbanMscModel93 is a tuned version of the Urban model used for e+/e- in all other physics constructors of physics-lists (Option2, Option3, Livermore, Penelope).
  • G4GoudsmitSaundersonMscModel is a new model for e+/e- based on formalism developed for Penelope and EGSnrc.
  A new relativistic model G4PairProductionRelModel is available for gamma conversion into e+e- pair. In this model, the LPM effect is taken into account. The model is suggested for use in high-energy astrophysics applications.

• Improved and extended the FTF model. Improved its pion absorption on nucleons, introduced Reggeon cascade stage, and added excitation energy calculation. Tuned the model parameters of nuclear de-excitation and implemented charge-exchange for pn->np in elastic and inelastic interactions.
• The validity of the pion-nuclear cross-sections in G4PiNuclearCrossSections, and the nucleon-nuclear cross-sections in G4NeutronInelasticCrossSections and G4ProtonInelasticCrossSections has been extended to 100 TeV, by assuming a constant cross-section at very high energies.
• Glauber-Gribov cross-sections for ion-ion collisions are now available in the class G4GGNuclNuclCrossSection, spanning from 0.1 MeV to 100 TeV.
• The nucleon-nucleon and pion-nucleon partial cross-sections used in the Bertini cascade were reviewed and the incorrect cross-sections found below 1 GeV were fixed. This results in slightly improved agreement with data below 1 GeV.
• The CHIPS model has been extended with hadronic interactions covering all energies for all hadronic particles; in addition hadron- and lepton-nuclear reactions are extended to high energies (~TeV). It can be used by choosing the new CHIPS physics-list. CHIPS ion-ion elastic scattering model is also now available.
• An option was added to allow models for elastic scattering to be treated in a similar way to inelastic models. Models can now be assigned to the process class G4WHadronElasticProcess.
• G4NeutronRadCapture is a simplified new model which produces a single photon after neutron capture. 4-momentum is conserved in the reaction.
• Improved probabilities and channels in the G4GEM model bring it into closer agreement with data.
• Unit conversion errors were corrected in fission and Fermi break-up models.

• Measured angular reflectivity distributions for common surface treatments and reflectors have been incorporated as look-up tables (LUT).

• New electromagnetic physics constructors combining standard and low-energy electromagnetic models are available: G4EmLivermorePhysics, G4EmPenelopePhysics, G4EmDNAPhysics, G4EmLivermorePolarizedPhysics.
• New default builder for optical photon physics has been added: G4OpticalPhysics together with its messenger, G4OpticalPhysicsMessenger.
• All lists based on FTF and CHIPS modelling use the improved versions of FTF and CHIPS.
• Newly added physics lists:
  • CHIPS, a physics-lists which uses the CHIPS model for nuclear interactions of all energies and particles..
  • QGSP_INCL_ABLA, a list using INCL/ABLA. It is mainly intended for use with energies less than 3 GeV. This is useful for e.g. spallation studies and Accelerator Driven Systems (ADS) applications.
• Other new physics-lists, already present in 9.3-Beta:
  • Some are based on existing lists, with different transitions between models:
    FTFP_BERT_TRV, lower the transition energy from BERTini to FTFP in the range from 6-8 GeV.
  • Others offer options for EM physics:
    FTFP_BERT_EMV, FTFP_BERT_EMX, QGSP_BERT_EMX, QGSP_BIC_EMY.
• Attempts to reduce usage of LEP in the intermediate energy range:
  • QGSP_BERT_NOLEP, using QGSP starting from 8.5 GeV;
  • QGSP_FTFP_BERT, using FTFP in the energy range between BERT and QGSP.

• Improved ability of visualization system to render Boolean solids. Certain cases involving very near co-planar faces used to result in incorrect visualization (though not incorrect tracking). An existing mechanism to deal with this has been improved; this mechanism used to shift slightly the position of one solid in a chosen direction. Now it attempts shifts in different directions by varying small amounts until one is found that does not give an error. The system still can not handle every situation. However it results in far fewer visualization inaccuracies than were seen previously in such difficult cases.
• Visualization attributes can now be individually set for each volume within a nested parameterization.
• Rich Trajectories and Rich Trajectory Points now include additional information such as the complete volume path of the starting and ending volume including all replica and copy numbers, etc.
• A new command, /vis/viewer/set/targetPoint, sets a specific point to be the center of the view. Helps with doing extreme zoom in to a specific point.
• gMocrenFile: First Beta release of this new visualization driver to output data in the gdd format used by the gMocren volume visualization tool. gMocrenFile can include volume data from sensitive detector volumes, or from command-based scoring meshes, in addition to detector geometry and trajectories. Use new command /vis/scene/add/psHits to add scorer hits to gMocren output. For details, see the Application Developers Guide and the gMocren web site.
• G4UIExecutive: a new utility class for automatic instantiation of a user session selected according to defined environment variables G4UI_USE_*.
• Qt: Continued improvements to this new driver. The Qt visualization driver and the associated Qt User Interface now work well on almost all platforms and with both Qt3 and Qt4 (but still considered a Beta release due to a few rare combinations of commands on certain platforms that can cause crashes).
• PostScript printing: vector and bitmapped printing through the UI command /vis/ogl/printEPS is now fully functional for all OpenGL viewers (previously only for OGLIX and OGLSX). Related new commands are /vis/ogl/set/printSize and /vis/ogl/set/printFilename.
• HepRepFile: new UI commands of the form /vis/heprep allow easy control of output file name and output directory, whether a file name should be reused (rather than having a sequential suffix appended at each new file) and whether culled geometry volumes should be included. Also fixed a bug in representation of G4Tubs which had caused wrong visualization for cases in which the tube was at an unusual angle.

5. Expected effects on physics and performance

• The new ad-hoc memory allocation system for touchables in navigation has shown an average CPU speed improvement of ~5% and a reduction of overall memory allocation and fragmentation for reasonably complex detector setups.

• Computing performance of low energy models has been significantly improved thanks to new implementations for data loading and interpolation methods, transparent to the user's application.
  No significant changes on the physics outcomes are expected due to the new software design of process and model classes, or from these new implementations of data loading and interpolation methods.

• Initialization of the physics-tables is now considerably faster, being measured up to a factor 3, depending on the application complexity.
• A more precise Spline interpolation for the physics-tables, provides accuracy of the order of 0.1%.
• A new method ResetConverters() has been added to G4ProductionCutTable. This method, if called once initialised the run, allows to reduce the memory footprint by clearing temporary loss tables and range vectors used at initialisation.

• Improvements in the de-excitation code (G4CompetitiveFission, G4FermiBreakup, etc.) produce better agreement with IAEA benchmark data below 200 MeV.
• Bug-fixes in G4GEM evaporation code remove odd-even A (in fragment mass) discrepancies with triple differential benchmark data.
• Bertini cascade now generates 10% less memory churn (creation and deletion of objects) and is about 6% faster as a result.
• Improved low energy partial cross-sections in Bertini cascade give slightly better agreement with low energy benchmark data (< 1 GeV).
• Improvements in INCL/ABLA code allow QGSP_INCL_ABLA physics list to run much longer (several thousand events) without crashing.
• Improvements in the FTF high energy model, allow the model to be used successfully down to 3-5 GeV incident particle energy.

6. Known Run-Time Problems and Limitations

• Reports have been made of particle tracks stuck at a volume boundary or at an interface between volumes. This occurs infrequently, on the order of one step per 10^11, but can effect large productions. The behaviour involved particles which do not progress but move in very small steps, and as a result the simulation of this track never finishes.
  • A partial solution is implemented in this release, which uses an existing mechanism to identify better these stuck tracks. It is able to find all such tracks in the setups reported, using a large multiple of the value of the geometrical tolerance for the setup. Users should pay attention that the value is chosen well for their setup in case they encounter a similar problem (using the existing functionality for setting this value).
  • The treatment for these stuck tracks is to abandon the relevant track. This results in the track being killed by the G4Transportation process. A warning is issued (if compiled in verbose mode, the default) in the case that the energy of the track is above a threshold energy. Its default value is 100 MeV. This value should be chosen by a user in order to get warning for all particles of energy important to your simulation. A new value can be assigned using the SetThresholdWarningEnergy method of G4Transportation.

7. Compilation Warnings

8. Known Run-Time Warnings

    G4Transportation: Statistics for looping particles 
       Sum of energy of loopers killed: 
       Max energy of loopers killed: 

9. Geant4 Software License

10. Detailed list of changes and fixes

• Configure script:
  • New simplified layout for queries and updated some defaults.
  • Fixed detection of QtCore library; addressing problem report #1079. Removed of -s switch to sed in Qt.U to fix problem with Qt setup.
  • Added flag to allow running with query options on non-interactive terminal. Resolves issue reported in Hypernews Install-and-Config #1213.
  • Added new data set for G4REALSURFACEDATA variable and updated G4LEDATA data set version.
  • Corrected text for installation of g3tog4 module; cernlib is not required.
  • Removed query on analysis; now G4ANALYSIS_USE flag has to be set explicitely by the user, if necessary.
  • Added force build of zlib when WIN32 GL or Inventor vis drivers are built.
• Added G4OPTDEBUG option for hybrid optimized-debug build also to Darwin-g++ and WIN32-g++ configurations.
• Corrected inclusion of -lz for zlib in architecture.gmk; corrected also inclusion of the G4zlib library in link list in binmake.gmk for the case of global libraries.
• binmake.gmk: added include paths to new CHIPS modules.
• G4VIS_USE.gmk: fixed dependency issues on WIN32-VC for gl2ps utility library in visualization. Added new gMocren include path. Removed references to obsolete GAGTREE driver.
• Darwin-g++.gmk: added paths to $QTHOME/include and $QTHOME/lib to setup for Qt4 configuration.
• WIN32-VC.gmk: added gd32.lib use32.lib to OGLLIBS to fix problems of linking when G4LIB_BUILD/USE_WIN32_SESSION is not set.

• Corrected counting of flux and current passing through curved surfaces (affecting cylinder and sphere surface scorers).
• Added missing virtual keyword for qualification of CheckROVolume() and FindROTouchable() methods in G4VReadOutGeometry.
• Added a few accessors for visualization to G4VScoringMesh.

• Adjoint
  • New revised electromagnetic processes for "Reverse MonteCarlo" technique. All e- processes have been improved and adjoint proton and ion ionisation have been added. The use of adjoint cross section matrices can be now limited only to e- Ionisation and Ion ionisation.
• High Energy
  • Simplified initialisation in G4mplIonisationModel and G4eeToHadronsMultiModel.
  • G4AnnihiToMuPair: added method CrossSectionPerVolume(); added integral option to PostStepDoIt(), providing more precise simulation near cross-section threshold.
  • G4eeTo3PiModel: fixed vector product in matrix element.
  • Some code cleanup: improved comments, avoid inlined virtual methods.
• Low Energy
  • Added G4PenelopeBremsstrahlungModel, namely Bremsstrahlung according to Penelope model, now inheriting from G4VEmModel. Completed migration to standard-EM design for all Penelope processes.
  • Added G4RayleighScattering process (inheriting from G4VEmProcess) describing Rayleigh scattering. Removed G4Rayleigh to avoid duplication.
  • Added G4LivermoreIonisationModel, first implementation of low-energy ionisation in the new design. Updated Livermore photon processes classes for handling of compound materials; added gamma particle and energy range test in cross section computation.
    Updated public interface of G4eIonisationCrossSectionHandler in order to facilitate migration from G4LowEnergyIonisation to G4LivermoreIonisationModel. Set default angular generator in G4LivermorePhotoElectric as Sauter-Gavrila.
  • First fully tested version of G4LivermoreBremsstrahlungModel, completing migration of Livermore processes to new design.
  • General cleanup of new model classes for Livermore and Penelope processes.
  • Fixed bug in G4PenelopeRayleigh (calculation of mfp for compounds), and in momentum normalization in G4PenelopeBremsstrahlungModel. Fix in G4PenelopePhotoElectricModel to avoid misleading warnings.
  • Added proton case in G4CrossSectionExcitationMillerGreenPartial.
  • Design update, bug fixes and models update for atomic shell hadronic cross-sections calculation.
  • Cleared annoying printout from G4AugerData transition table.
  • Added ActivateAuger() method to Penelope Compton and Ionisation, so that all Penelope models have now the same interface.
  • Added protections in G4eIonisationSpectrum and G4eIonisationParameters, addressing problem report #1042.
  • Revision of G4IonParametrisedLossModel. Added G4IonDEDXHandler, a new handler class used by the revised model class. Added G4VIonDEDXScalingAlgorithm, used by the handler; removed G4IonParametrisedLossTable. Moved from original ICRU'73 classes to new class G4IonStoppingData, capable of reading stopping power data files stored in the data set; with the new tables the upper energy limit of ICRU'73 stopping powers is increased to 1 GeV/nucleon. Added modifier for energy loss limit parameter. Changed default energy loss limit to smaller value (from 0.15 to 0.01) to avoid approximate loss description in case of large steps. Removed nuclear stopping from model since dedicated process is available in standard EM package. Minor fixes applied in energy loss computation. Bug fix avoiding wrong ion ranges in case different cuts are used for a given material in different regions. Added UpdateRangeCache() method, removed GetRange() method.
  • Added new scaling algorithm G4IonDEDXScalingICRU73, to obtain heavy ion stopping powers for ions not covered by ICRU'73 report (scaled from existing ICRU'73 data). Included algorithm in ion model, extending the applicability of the model to very heavy ions.
  • Geant4-DNA processes and models:
    • New migrated processes and models.
    • Added ability for DNA charge change processes to kill particle without total energy deposit.
    • Increased energy range validity of e- elastic Champion model up to 10 MeV.
    • Extended high energy coverage of proton ionisation (G4DNABornIonisationModel) and excitation (G4DNABornExcitationModel); decreased temporarily high energy coverage of electron ionisation from 30 keV to 20 keV.
    • Added G4_WATER NIST material and verbose condition to all Geant4-DNA models.
    • Bug fix in G4DNABornIonisationModel for energy deposit.
    • Improved energy limit display.
    • Modified low energy limit of e- elastic Champion model to be used with new data set. Extended coverage of e- ionisation model as well.
    • Fixed cross section computation bug in G4DNAScreenedRutherfordElasticModel and G4DNAChampionElasticModel.
    • Added vapor water shell constants to G4DNARuddIonisationModel.
    • Added message to obsolete non-migrated Geant4-DNA cross section and final state classes, and non-migrated Livermore and Penelope classes.
  • Modified classes G4VeLowEnergyLoss, G4eLowEnergyLoss and G4hLowEnergyLoss according to updates in G4PhysicsVector.
  • New implementation for data loading and interpolation methods; LoadData() method now loads the data from the G4EMLOW dataset and, then calculates the respective logarithmic values during initialization. Both data formats (original and logarithmic) are loaded to EM datasets and used later by interpolation methods to perform faster calculations. LoadNonLogData() method implements the old version for data loading and interpolation.
  • Requires new data set G4EMLOW-6.9.
• Muons
  • G4MuIonisation: fixed initialisation for allowing to configure external model of fluctuations. Simplified initialisation in G4MuBetheBlochModel, G4MuBremsstrahlungModel and G4MuPairProductionModel.
  • G4EnergyLossForExtrapolator: removed hidden bin to use full scale of of G4PhysicsVector; set linLossLimit to 0.01.
  • G4MuBetheBlochModel: compute density effect correction using new method of G4IonisParamMat class, allowing to apply revised parameterization of the density effect.
  • G4MuMultipleScattering: use G4UrbanMscModel90 by default.
  • Some code cleanup: improved comments, avoid inlined virtual methods.
• Polarisation
  • Bug fix in G4PolarizationHelper::GetSpinInPRF() for wrong normalization of Y component.
  • Bug fix in G4PolarizedAnnihilationModel::PostStepDoIt() for wrong treatment in error check.
  • Bug fix in G4StokesVector::RotateAz() for wrong usage of vector product.
  • Simplified initialisation in G4PolarizedAnnihilationModel.
• Standard
  • Simplified initialisation of all models and code cleanup.
  • G4PEEffectModel: substituted ComputeMeanFreePath() with CrossSectionPerVolume() resulting in minor CPU speedup for compound materials.
  • Added new class G4HeatedKleinNishinaCompton, prototype model for plasma.
  • New class G4PairProductionRelModel, relativistic model for gamma conversion.
  • Energy loss models:
    • G4UniversalFluctuation: modification in width correction, the dependence of the correction on energy deposition at previous steps is removed, width correction function is used instead. Addresses report from T2K setup.
    • G4IonFluctuations: removed extra phenomenological factor in fluctuation width.
    • G4BohrFluctuations: changed internal parameter to provide for smoother transition to Landau regime in the case of small cuts.
    • Set linLossLimit to 0.02 in G4ionIonisation allowing to smooth computations of mean energy loss of ions.
    • G4BetheBlochModel and G4MollerBhabhaModel: compute density effect correction using method of G4IonisParamMat class, applying revised density effect parameterisation.
    • G4BetheBlochModel: fixed and simplified initialisation for ions needed for the new G4IonParametrisedLossModel of low-energy package. Minor speedup in CorrectionsAlongStep() method.
    • G4BetheBlochModel, G4BraggModel, G4BraggIonModel: removed computations of nuclear stopping power, which will is provided by the dedicated process, some CPU speedup.
    • G4WaterStopping: fix to Ar data; added Fe ion data.
  • Added new classes G4NuclearStopping and G4ICRU49NuclearStoppingModel, a new model providing computation of nuclear stopping.
  • G4PAIModel, G4PAIPhotonModel: removed usage of random numbers at initialisation, to avoid potential non-reproducibility issues.
  • Added new G4alphaIonisation process to be used in experimental physics-lists.
  • Single and multiple scattering:
    • Added G4UrbanMscModel92 and G4UrbanMscModel93 classes, corresponding to 9.2 and 9.3 versions of the Urban multiple scattering model.
    • Set default model G4UrbanMscModel92 for G4eMultipleScattering.
    • G4UrbanMscModel93: new tuning for the central part and for the tail of the angular distribution using the old e- scattering data only (Phys. Rev. 84 (1951) 634; Phys. Rev. 61 (1942) 254). New version of theta0 and correction in tail tuning using MuScat data.
    • G4WentzelVIModel and all G4UrbanMscModel*: use generic methods of G4VMscModel to access safety and other information with the geometry modeller.
    • Added new alternative multiple-scattering model for e+-, G4GoudsmitSaundersonMscModel.
    • Added warning in obsolete multiple-scattering classes G4MultipleScattering, G4UrbanMscModel, G4UrbanMscModel2, G4MscModel71.
    • G4hMultipleScattering: removed unused obsolete methods.
    • G4eCoulombScatteringModel: added relativistic factor to Rutherford cross section; added cut per recoil using proton production threshold (1 mm cut in range corresponding to 100 KeV); Reduced low-limit from 1 KeV to 0.1 KeV to provide smoother transport cross-section table. Added protection for very low energies and changed screening parameters; fixed kinematic for ion/ion scattering.
    • G4CoulombScatteringModel: fixed energy in the center of mass reference system; added cut per recoil using proton production threshold.
• Utils
  • G4VEmModel:
    • Moved multiple-scattering methods to derived class G4VMscModel.
    • Added new methods SampleDeexcitationAlongStep(), GetParticleChangeForLoss, GetParticleChangeForGamma.
    • Insure definition of pointers to the current G4Element and G4MaterialCutsCouple objects both at initialisation and at run time.
  • G4VMscModel:
    • Added generic methods used by all multiple-scattering models InitialiseSafetyHelper(), ComputeSafety(), ComputeGeomLimit() and ComputeDisplacement().
    • Set default facsafety to 0.3 and range-factor to 0.04 to be the same as in G4UrbanMscModel92.
  • G4VEnergyLossProcess, G4VEmProcess, G4VMultipleScattering:
    • Set high energy limit to 10 TeV and number of bins to 77. Adopt updated G4PhysicsVector with no longer hidden bin.
    • Added initialisation of polarAngleLimit and highEnergyLimit parameters for all models.
    • Fixed bug in method RetrieveTables(); check that each its physics-vector is retrieved before setting the SplineFlag.
  • G4VEnergyLossProcess, G4VEmProcess:
    • Added calls to de-excitation post-step; added de-excitation along-step per region.
    • Added ActivateDeexcitation() method in G4EmProcessOptions and related UI command to messenger.
  • G4VMultipleScattering:
    • Added initialisation of generic multiple-scattering model parameters
    • Added SetModel() method, allowing to overwrite default.
    • Removed obsolete method GetMscContinuesStepLimit().
  • G4LossTableManager, G4EmProcessOptions, G4EnergyLossMessenger:
    • Added parameter FactorForAngleLimit with accessors, modifiers and UI command, to be used in the computation of limit on -t (invariant momentum transfer) inside single scattering and G4WentzelVI models.
  • G4EmCalculator:
    • Added possibility to be used by DNA low-energy processes taking into account special DNA particles.
    • Added method FindIon()
    • Improved computation for electronic stopping power of ions using CorrectionAlongStep method of G4vEnergyLossProcess allowing to be applied for any model including the model based on ICRU'73 data.
    • Fixed bug in selection of models.
  • G4EmConfigurator: fix for the case when only fluctuation model is set and main model is the default.
  • Added G4EmConfigurator in G4LossTableManager providing for easier addition of models per region.
  • G4EmSaturation: use PDG encoding instead of pointer to G4ParticleDefinition.
  • G4EmCorrections: fixed bug reported in physics-list forum #532.
  • G4LossTableBuilder: added protection in BuildRangeTable() against zero dEdx. Use copy constructors to create physics vectors in order to reduce number of calls to exp() function.
  • G4EmModelManager:
    • Fixed energy range selection algorithm for the case of a small intersection of model energy intervals.
    • Fixed energy range selection algorithm for the case when there is an intersection of energy regions of standard and low-energy models.
    • Reduced size and number of internal vectors if no model per region are initialized.
    • Avoid to initilise unused models; simplified initialisation by smoothing procedure. The size of executable should be reduced.
  • G4EmElementSelector: added protection for zero cross section at first and last bins of physics vector and disabled Spline interpolation.
  • Added G4AtomicShell class from use by the low-energy module.
  • Added new interface class G4VAtomDeexcitation.
• Xrays
  • Exclude short-lived particle from IsApplicable() for G4Cerenkov and G4Scintillation processes.

• Inclusion of pi+, pi- and proton in physics list.
• Added possibility to account for error deflation (needed to make 'smoothing').
• Corrected usage of G4ThreeVector constructor in G4ErrorMagFieldLimitProces.

• Introducing G4SmartTrackStack class. By using this new stack as the urgent stack of G4StackManager, the next track popped up from the stack is not necessarily the last track stored in the stack, but the track of same particle type as the previous one, as long as such a track exists in the urgent stack. This mechanism is expected to improve the performance for ultra-large scale simulation such as LHC, by increasing the cache hit rate of the physics tables.
  The use of G4SmartTrackStack is temporarily optional. To use it, one should uncomment the #define line in include/evmandefs.hh.
• Added UI comand /event/stack/verbose providing a short report of peak number of tracks in the urgent stack.
• New classes implementing the "Reverse MonteCarlo" technique: G4AdjointPosOnPhysVolGenerator, G4AdjointPrimaryGenerator and G4AdjointStackingAction.

• Cuts
  • G4VRangeToEnergyConverter: modified implementation of ConvertCutToKineticEnergy() to improve performance. Measured a factor 2 to 3 improvement in initialisation speed when building physics tables. Modified BuildRangeVector() where integration is perfomed in this method; extended for e+/e-. Changed TotBin to be constant (i.e. same for all converters) and set to 100 (200 for previous versions). Added MaxEnergyCut (10 GeV set in default). Added fRangeVectorStore to ensure that range-vector is created only once for each material. Do not apply converter for neutron, anti-proton, and anti-neutron.
  • A new method ResetConverters() has been added to G4ProductionCutTable. This method allows to reduce the memory footprint by clearing temporary loss tables and range vectors used at initialization, if invoked when starting the first run of a session.
  • Added possibility to handle cuts for protons.
  • Added new command /cuts/setMaxCutEnergy to messenger.
• Optical
  • Added capability of simulating surface reflections with Look-Up-Tables (LUT), containing measured optical reflectance for a variety of surface treatments. Requires new data set RealSurface-1.0 controlled by environment flag G4REALSURFACEDATA. Courtesy of M.Janecek and W.Moses (LBNL).
  • G4OpBoundaryProcess: added initialization of reflectivity and efficiency at the beginning of the PostStepDoIt(); added method BoundaryProcessVerbose() which will now print the process status for every call to PostStepDoIt(). Change in PostStepDoIt() so that all dielectric_dielectric surfaces may now have a reflectivity less than 1; not only frontpainted as was the case in the past.

• Divisions
  • Allow for generic divisions along Z for polyhedras and polycones; now divisions can happen along Z-axis with width and offset, which fulfils the condition that division does not span over more than one segment.
  • Fixed initialization of division in Phi for polyhedras to not take into account user defined width.
• Magnetic field
  • New stepper G4NystromRK4 for magnetic field with analytic estimation of integration error. Greatly reduces number of field value per step.
  • Added new virtual method CalculateRightHandSide() to G4MagIntegratorStepper for use in caching momentum (and field value) by G4NystromRK4; default implementation in G4MagIntegratorStepper calls RightHandSide().
  • New simple class G4CachedMagneticField to cache value of magnetic field. Speeds up code when calculation of field value is complex.
  • New G4EqEMFieldWithEDM class for RHS of EofM in a combined electric and magnetic field, with spin tracking for both MDM and EDM terms. Courtesy of Kevin Lynch, Phys. Dept. at Boston Univ.
  • Bug fix in G4EqEMFieldWithSpin, following report on Hypernews EM-field forum #161.
• Management
  • Fixed implementation of G4LogicalVolume::TotalVolumeEntities() to become invariant to call sequence. Addressing problem report #1082.
  • Clear possible excess capacity in node vectors in G4SmartVoxelHeader BuildNodes() method. May reduce excess memory allocated in optimisation in medium/large detector setups, at the price of a slight time increase in initialisation for geometry voxelisation.
  • G4VTouchable: moved inline vitual methods, constructor and destructor to source, to avoid problem of memory alignment in allocation of derived objects.
  • G4Region: added pointer to field-manager and associated accessor/modifier, to allow for association of a local field to a geometrical region.
• Navigation
  • Refinements in G4PropagatorInField:
    • Activated field per region. A field manager can now be assigned to a region, and it will be used for all logical volumes it contains, except for those which override it by having a field manager of their own.
    • Improved handling of repeated small steps at boundaries which are due to geometry imprecision, navigation optimisation, or the details of the algorithm used for tracking in field.
      • Added new member fZeroStepThreshold, to enable tuning of this existing threshold for identifying tiny or zero steps.
      • A new value for this threshold fZeroStepThreshold is a large multiple (10^4) of kCarTolerance; the old value was 0.5*kCarTolerance. Addresses a problem of tracks stuck at boundaries of volumes reported by ATLAS.
      • The condition for identifying a tiny step is changed in order to avoid misidentifying small proposed physical steps as problem tiny steps.
    • Changed parameters which determine the size of the next trial step. This is used only when trying to recover from consecutive tiny/zero steps.
      • The default decrease factor is now 0.25 (from 0.1)
      • When the step size is comparable to the value of fZeroStepThreshold a different decrease factor is used. The value given to the decrease factor was changed to 0.35 (from 0.25).
    • Added new method RefreshIntersectionLocator() to update the state of helper class.
    • Improved printing of diagnostics.
    • Deleted obsolete internal methods: SetAccuraciesWithDeltaOneStep(), SetDeltaIntersection(), SetDeltaOneStep().
  • G4VIntersectionLocator: fixed problem in ReEstimateEndPoint() for very small steps. Corrected constructor to ensure that it initialises all data members. Labelled methods in header to make noticable those that must change attributes at every step.
• Solids (CSG)
  • G4Sphere:
    • Implemented speed improvements and corrections from joint code review. Cached computation for half-tolerances and use of Boolean flag for identifying if full-sphere, shell or section.
    • Implemented caching of trigonometric values, now directly computed inside modifiers for Phi and Theta angles as required for parameterised cases.
    • Rationalised usage of relative radial tolerances.
    • Correction in DistanceToOut(p,v) for phi sections for rays passing through zero.
    • Fix for the calculation of the normal in DistanceToOut() to avoid cases of division by zero in specific configurations. Addresses problem report #977.
  • G4Tubs, G4Cons: rationalised usage of modifiers for Phi angles and simplified constructors.
  • G4Cons: fix to DistanceToIn(p,v), added a check on the direction in case of point on surface. Fixes a problem of stuck tracks observed in CMS, due to wrong reply from the solid for points on the inner radius surface base, with direction along the imaginary extension of the cone.
  • G4Torus: fix in SolveNumericJT() in order to take into account the difference in the value of theta for different intervals, [0:pi] or [-pi:0], and for SPhi in [0:twopi] or [-twopi:0]. Addresses problem report #1086.
  • G4Orb: moved debug warning in DistanceToIn(p,v) within G4CSGDEBUG flag.
• Solids (Specific)
  • G4Ellipsoid: refined fix in DistanceToIn(p,v) for points located on the curved surface, and correct treatment of geometrical tolerance. Addresses problem report #1076.
  • Avoid redundant assignment of value to 3-vector in G4PolyconeSide and in G4TwistTubsSide.
• Volumes
  • Fix in G4PVParameterised::CheckOverlaps() to not modify daughter volume transformation matrix for application to points in the daughter's frame. Adopt Inverse() instead of Invert() in call to G4AffineTransform. Addresses problem report #1078.
  • Added missing implementation for composing G4AssemblyVolume constructor.
  • Use G4Allocator for dynamic allocation of G4TouchableHistory objects. Should reduce the number of direct calls to malloc and memory fragmentation. Measured an average improvement of ~5% in CPU speed with same memory footprint. Moved inline constructors to source to avoid problems of memory alignment in allocation.

• G4PhysicsVector:
  • Added ComputeSecDerivative() and SplinePossible() methods as a computation of Spline coefficients for small number of bins. Removed old obsolete hidden bin approach.
  • Added method ComputeSecondDerivatives() for the case when user provides first derivative at end-points; use "Not-a-Knot" algorithm for the computation of second derivatives in default method FillSecondDerivatives().
  • Cleaned up initialisation of cached values; simplified Value() method, now not looking into previous bin; added method ScaleVector() needed for ICRU'73 data.
• New class G4Pow for fast computation of log() and pow() for integer and double arguments in the interval [1-256].
• Added G4ErrorStage enum to G4ErrorPropagatorData class.
• Cleared warning for unused argument in G4Allocator internal method.
• Changed date for release 9.3.

• Improved ability of visualization system to render Boolean solids. Certain cases involving very near co-planar faces had resulted in incorrect visualization (though not incorrect tracking). A previously included mechanism to deal with this by shifting the position of one solid in a chosen direction has been expanded by attempting shifts in different directions by varying small amounts until one is found that does not give an error. The system still can not handle every situation, but results in far fewer visualization inaccuracies than were previously seen in such difficult cases.
• Improved infrastructure for scoring (supporting new UI command /vis/scene/add/psHits).
• G4VVisManager: added notify scene handlers (G4VGraphicsScene objects) taking care to notify that the scene has changed so that they may rebuild their graphics database, if any, and redraw all views.
• G4VMarker: added enum SizeType {none, world, screen}; added GetSizeType() and SetSize/Diameter/Radius() methods.
• G4VGraphicsScene: added method AddCompound(HitsMap).

• Cross sections
  • Added new Glauber-Gribov cross section for ions, G4GGNuclNuclCrossSection.
  • G4IonsKoxCrossSection, G4IonsShenCrossSection, G4TripathiCrossSection, G4TripathiLightCrossSection: added protections against numerical FPE exceptions.
  • Use as a default hadron-nucleon cross section parameterisation in G4GlauberGribovCrossSection, G4GGNuclNuclCrossSection, G4NeutronElasticXS, G4NeutronInelasticXS, G4BGGNucleonElasticXS, G4BGGNucleonInelasticXS, G4BGGPionElasticXS, G4BGGPionInelasticXS.
  • Bug fixes for G4BGG* classes and added new classes G4NeutronElasticXS, G4NeutronInelasticXS, G4NeutronCaptureXS.
  • Modified upper limit of validity to 100 TeV in G4PiNuclearCrossSection, G4NeutronInelasticCrossSection and G4ProtonInelasticCrossSection. Use constant cross-section above original range.
• Management
  • G4HadronicInteraction: renamed SampleT as SampleInvariantT; added recoilEnergyThreshold parameter and accessor methods; use const pointers to G4Element and G4Material; moved G4ReactionDynamics instance to parameterised model class G4InelasticInteraction.
  • G4HadronicInteraction: added SetModelName() method.
  • G4V3DNucleus: added two functions to sort nucleons along +/- Z.
• Abrasion
  • Fixed problems of negative sqrt() and made code more robust to potential loop problems.
• Binary Cascade
  • Added interface to set de-excitation handler in G4BinaryLightIonReaction used to de-excite lighter initial nucleus; precompound handling low energy interactions can be set too.
  • Removed dependency on unused class G4NucleiPropertiesTable.
  • Made capture-threshold a tunable parameter in G4GeneratorPrecompoundInterface.
  • Changes to G4BinaryCascade to protects against energy non-conservation: shift of particle's formation times; inclusion of captured particles; correction in GetExcitationEnergy() for the case when all nucleons are ejected; correction for the nuclear field for baryons.
  • Bug fix in G4BinaryCascade::ApplyCollision(): decay products outside nucleus were nevertheless counted as if these were within nucleus.
  • Corrected assignment to 3-vector, avoid use of 3vector=0.
  • Added protection against void nucleus after cascade.
• Bertini Cascade
  • New class G4PionSampler: samples partial cross sections, multiplicities and final state particle types required for pi+, pi- and pi0 scattering within a nucleus. This class uses re-evaluated pion partial cross-sections and includes production of strange particle pairs.
  • Added class G4NucleonSampler which replaces old pp, pn and nn partial and total cross-sections; also modified other classes which use them.
  • Added class G4FinalStateSampler.
  • G4CascadeSigmaPlusNChannel: corrected sigma plus cross-sections.
  • G4CascadSpecialFunctions: moved pion, kaon, hyperon total cross-sections to this class. Put copy of total cross-section method in G4NucleiModel, also removed unneeded energy scale arrays, now replaced with a new one.
  • G4ElementaryParticleCollider: re-written to use G4PionSampler class for cross-sections and multiplicities; removed method reChargering(), now redundant.
  • Moved large arrays which are instantiated at every call of frequently used methods. They are now initialized once in constructors.
  • G4CascadeInterface: moved instantiation of various collider classes outside of loop, and put on stack instead of heap.
  • G4ElementaryParticleCollider: redesigned to use G4NucleonSampler and remove instantiation of large arrays from within loops in code.
  • G4CascadParticle, G4InuclElementaryParticle: added integer argument to constructor to indicate which generation cascade particle belongs to. Changed G4NucleiModel and G4BertiniNucleiModel to use this.
  • G4NucleiModel: removed instantiation of large arrays from with loops; added new total cross section sampling, removed use of all but one energy scale. Updated pi- and pi0 total cross-sections. Clean up of dead code.
  • Added assignment operator to class G4CollisionOutput.
• Chiral Invariant Phase Space (CHIPS)
  • First implementation of CHIPS ion-ion elastic.
  • Extended model for all energies for all particles. Removed obsolete classes.
  • Added photon and lepton-nuclear CHIPS inelastic; improved model to work for all targets, including Hydrogen. Added diffraction to CHIPS inelastic.
  • Added 8 new cross-section classes and bug fixes in pA cross-sections. Tuning of the G4QCollision for better description of the pA interactions (E<290 MeV).
  • Reorganised CHIPS classes in sub-modules: body, cross_sections, fragmentation, interface and processes. Better readability of the CHIPS code and code improvements.
  • Improvements in hadron+proton, heavy fragments in low-energy, various bug fixes in G4QNucleus, G4QEnvironment, G4QCollision.
  • Improvement of G4QuasiFreeRatios for application in QGSC from E=0.
  • MuCaptureEMCascade excluded from capture at rest for mu-; all the atomic de-excitation energy of capture is concentrated in the energy deposition.
  • Bug fix against division by zero in G4QEnvironment; improvement for the case when the Hypernucleus cannot decay in the residual nucleus and pion; now it is decayed in the residual nucleus and gamma.
  • Fixed non-initialised variables in G4QContent and fix in quasi-elastic for photons and leptons.
  • Change of GetNuclearMass(A-1.,Z-1.) (and similar) to GetNuclearMass(A-1,Z-1) to be consistent with changes in particles.
  • Replaced internal RndmDir() with general G4RandomDirection() in the code.
  • Bug fix in G4QNeutronCaptureRatio (Cu65): accelerate CuSc by 1.5.
  • Cleared dependency of now removed G4NucleiPropertiesTable class.
• De-excitation
  • GEM code:
    • Fixes in G4GEMProbability and G4GEMChannel to provide same probability as proposed in original papers. Fixed old problem in G4GEMChannel, computation of Coulomb barrier was not consistent with probability. Added protection on unphysical initial state in G4GEMChannel. Fixed units (hbar_Plank -> hbarc) in G4GEMProbability.
    • G4EvaporationDefaultGEMFactory: new class where light meson probabilities are computed from default classes and other fragments from GEM.
    • Introduction of new classes G4GEMCoulombBarrier, G4AlphaGEMCoulombBarrier, G4DeuteronGEMCoulombBarrier, G4He3GEMCoulombBarrier, G4ProtonGEMCoulombBarrier, G4TritonGEMCoulombBarrier and corresponding changes in other classes of the sub-package.
  • Added a G4NeutronRadCapture model, a simple 1-gamma emission with 4-momentum balance.
  • Fixed 4-momentum balance by using direct Lorentz computations in G4CompetitiveFission. Make use of internal units.
  • G4Evaporation: added method SetCombinedCannels() to define new evaporation factory G4EvaporationDefaultGEMFactory.
  • G4FissionLevelDensityParameter, G4FissionParameters: tuned against IAEA benchmark spallation data.
  • G4FermiFragmentsPool: fixed units (keV -> MeV), removed wrong levels, added new levels according to Broohaven data (see chart of nuclides).
  • Fixed G4ExitationHandler for the case when Fermi-BreakUp and multi-fragmentation model are activated.
  • Improved implementation of Gamma function computation in configuration weight inside G4FermiConfiguration.
  • Updated ablation largely to comply with updates in G4Evaporation and handling of G4Fragment; fixed a bug in fragment identification when evaporation isn't used.
• Elastic scattering
  • Model precise sampling and 'on the fly' table preparation in G4DiffuseElastic.
  • Migrated to the hadronic model design of elastic scattering. Modified G4ElasticHadrNucleusHE to be now capable to work with old and new interface.
  • Added new elastic process G4WHadronElasticProcess.
  • Added new class G4VHadronElastic, base class for elastic models with default GHEISHA model
  • New class G4CHIPSElastic, model interface to CHIPS parameterization.
  • Removed had-hoc NaN finder and printout in destructor of G4HadronElastic.
• High Energy Theoretical
  • Quasi Elastic part is made more used in QGSC at low energies to improve the performance of QGSC at low energies.
• Im_r_matrix
  • Fix in G4CollisionComposite reproducibility. Removed dependence of the energy grid of the cross-section set on the order of the particles by forcing heavier particle to be at rest when caching the cross section.
  • Reduced usage of strings; maps of cross-sections use G4Particledefinition in place of particle name.
• INCL/ABLA
  • Bug fixes to INCL nuclear potential handling; the interpolation function used to produce discontinuous results due to an array index off-by-one. This bug fix corrects the impact parameter distribution.
  • Fixed several variable initialization and F77 to C++ translation issues in INCL.
  • Added several safeguards for division by zero or negative sqrt().
  • Use centralised random number generator instead of built-in one.
  • Minor fix in the INCL particle reflection time calculation.
• Parton-String
  • Improved FTF model for pion obsorption on nuclei: introduced status of nuclear nucleon involved in an interaction (Status 0 - spectator, 1 - involved nucleon, 2 - absorbed nucleon); introduced connection between a participant nucleon and a nuclear nucleon in G4InteractionContent.
  • Completed FTF model with Reggeon cascade and added excitation energy calculation.
  • Tuned FTFP with parameters of nuclear de-excitation.
  • Implemented charge-exchange for pn->np in elastic and inelastic interactions simulated by FTF. Implemented also pion absorption by a nucleon.
  • New Boolean parameter added in G4VSplitableHadron class for activation and corresponding methods to operate with it; needed for absorption of meson in nuclear collision generated by FTF.
  • Fixed bug in G4VLongitudinalStringDecay for the calculation of formation/creation time, c_light was inserted. Due to this string tension parameter was set to the usual value (1 GeV/fm) in G4LundStringFragmentation.
  • G4VParticipants: use modified name of method to sort nucleons in G4V3DNucleus.
• Photolepton-hadron
  • Code cleanup.
• Pre-equilibrium
  • G4PreCompoundModel: CEM transition probabilities are set by default. In G4PreCompoundTransitions, CEM transition probabilities have been renormalized (transitions with Delta N=+2 are increased by factor 5).
  • G4PrecompoundEmission: modified AngularDistribution() to avoid floating point exceptions. Removed calls to log() in rho() and replaced by use of new logfactorial().
• QMD Reaction
  • Fixed bug for incidence of neutral particles in G4QMDReaction and G4QMDGroundStateNucleus; verified for nucleon-nucleus interactions.
  • Beta version of code for enabling meson incidence.
• Radioactive Decay
  • Added dependence on EM 'util' module for usage of the G4AtomicShell class.
• Re-Parameterized Gheisha
  • G4RPGInelastic: added temporary message stating the RPG model is under development and should not be used for production.
• Stopping
  • Removed dependency on unused class G4NucleiPropertiesTable.
  • G4PiMinusAbsorptionAtRest, G4KaonMinusAbsorptionAtRest: fixed usage of 3vector constructor.
• Utils
  • Restored methods for storing/retrieving G4SplitableHadron, now in use in FTF model.
  • Corrected assignment to 3-vector, avoid use of 3vector=0.
  • G4WilsonRadius: fixed data vector type qualification.
  • G4NuclearFermiDensity: added protection to GetDeriv() for positions far outside the nucleus causing an FP overflow in exp().

• G4UIcommandTree: added a new function to complete a command path with maximum characters guess; FindCommandTree() will now return current tree when command path is "/".
  Fixed incorrect directory deletion of G4UIcommandTree::RemoveCommand(). Addresses problem report #1087.
• Added method GetPreviousSession() in G4UIbatch.

• G4UIExecutive: new class for automatic instantiation of user interactive sessions.
• G4VBasicShell: better command completion in command line. Fix for commands strings that match shorter commands.
• G4VUIshell: added SetLsColor() method.
• G4UIQt: added interactor to be able to add button on file and help menu; added public method to get the main window. Added CTRL+A and CTRL+E to move to begin/end of command line. Fixed a problem on macOSX 10.6 with Qt 4.6-Beta. Added debug flags for Qt. Fixed compilation warnings.

• ICRU'73 stopping power classes:
  • Migrated to the G4VIonDEDXTable interface: access methods of physics vectors have changed (vectors can also be identified via atomic number of material if material is pure). Classes now deliver mass stopping powers instead of stopping powers per unit length. Any dependency on ICRU'73 material densities was removed (no densities are included in the classes any longer). Stopping power vectors are now only built on request and not anymore in the constructor. Affected classes: G4IronStoppingICRU73, G4MaterialStoppingICRU73 and G4SimpleMaterialStoppingICRU73.
  • G4IronStoppingICRU73: extended upper limit of the data vector from 10 MeV to 1 GeV; 53 bins are now used. Added G4WATER data. Also fixed GetDEDX() methods in these classes for E=10 MeV (upper limit).
  • G4MaterialStoppingICRU73: extended energy range of revised ICRU'73 tables for water from 0.025-10 MeV/u to 0.025-1 GeV/u. Added new function AddPhysicsVectorFullRange() for creating corresponding physics vectors. Changed material name from G4_NYLON-6/6 to G4_NYLON-6-6.
• New class G4IonStoppingData, which can read ion stopping power vectors contained in G4LEDATA data set; allows for inclusion of larger physics vectors than those in the ICRU'73 classes, where vectors are restricted to the maximum energy 10 MeV/nuc.
• G4NistMaterialBuilder: mean ionisation potential for G4_WATER is set to 78 eV according to GSI Annual Report (2008) p.373, in order to be consistent with ICRU'73 tables. Fixed ICRU'73 data provided by Prof. P.Sigmund (Univ. Southern Denmark): G4WATER, G4WATER_VAPOR, G4_NYLON-6-6 (in G4MaterialStoppingICRU73); G4_Cu (in G4SimpleMaterialStoppingICRU73). Changed material names, G4_NYLON-6/6 to G4_NYLON-6-6, and G4_NYLON-6/10 to G4_NYLON-6-10, to allow matching with ion stopping power data files. Old names can still be used for backward compatibility. Changed density of G4_GRAPHITE according to PDG-2008, added new material G4_GRAPHITE_POROUS.
• G4NistMessenger: added UI command to print parameters of density effect data. Improved implementation of UI commands in G4NistManager, G4NistMessenger and G4NistElementBuilder.
• G4VIonDEDXTable: added interface class for electronic stopping power tables for ions. Added new method BuildPhysicsVector() which invokes the build of physics vectors on demand in the initialisation phase. Added method BuildPhysicsVector() according to base class for G4ExtDEDXTable, G4SimpleMaterialStoppingICRU73, G4MaterialStoppingICRU73 and G4IronStoppingICRU73.
• G4ExtDEDXTable: new utility class to handle external electronic stopping power tables for ions.
• Use 'const' references for strings as arguments/return-values in material properties tables and related classes. Inlined some methods and some minor cleanup.
• Added Entries() method to G4MaterialPropertyVector and accessors for maps in G4MaterialPropertiesTable, required for internal use in persistency of material properties. Make visible GetEntry() accessor in G4MaterialPropertyVector for use restricted to persistency.
• Increased flexibily of G4SurfaceProperty and G4OpticalSurface classes to allow for lookup-table boundary process simulations.
• G4Element: added method GetNbOfShellElectrons() returning the number of electrons on a shell.
• G4DensityEffectData: added data from Atom. Data Nucl. Data Tabl.30 (1984) 261, accessible via material name. added method GetIndex()
• G4IonisParamMat: added new members: Eplasma, AdjustmentFactor, Delta0 parameter of density effect parameterisation, and hook to G4DensityEffectData. Fixed copy constructor. Added new access method and method DensityCorrection(). Fixed density effect computation; added plasma energy computation and access to G4DensityEffectData by the atomic number.

• Added new classes for "Reverse MonteCarlo" technique: G4AdjointProton and light ions.
• Make use of new experimental class G4NucleiPropertiesTableAME03 including data updated to the Ame-2003 atomic mass evaluation (II) by G.Audi, A.H.Wapstra and C.Thibault, Nuclear Physics A729, 2003.
• Removed obsolete class G4NucleiPropertiesTable.
• Added method CreateAllIon() in G4IonTable and added /particle/createAllIon UI command. By this method and command, all stable ions are created.
• G4ParticleDefinition should be created in pre-init state; a warning is thrown if constructors of G4ParticleDefinition are called in a different application state.
• Added CalculateAnomaly() method to G4ParticleDefinition.
• Fixed small difference of mass of taus, mesons, and anti-baryons between PDG-2009.
• Changed G4ElectronOccupancies::MaxSizeOfOrbit to 10 to allow for higher excited atom.

• ASCII
  • Inclusion of extruded and tessellated solids.
  • Additions of accessors and constructors of G4tgbVolume and G4tgrVolume.
• GDML
  • Implemented virtual layer for allowing customisation of the writer for user-extended schemas.
  • Implemented ability to write surface properties associated with surfaces and material properties.
  • Allow for reading/writing of constant material properties.
  • Added support for the GDML assembly tag, implementing assemblies of simple placements through G4AssemblyVolume.
  • Added support for GDML expression tag, implementing generic expressions stored as literal constants.
  • Corrected handling of quantity tag; added missing caching of evaluation.
  • Some code cleanup: rationalised inclusion of headers, replacing them with forward declarations where possible.

• New physics lists and builders:
  • QGSP_FTFP_BERT based on QGSP_BERT replacing the bridging LEP with FTFP, and transitioning from BERT to FTFP between 7 and 9 GeV.
  • FTFP_BERT_TRV list as variant of FTFP_BERT with transition between 6 and 8 GeV.
  • QGSP_BERT_TRV: with transition energies to use BERT ]0,9.9], LEP [9.5, 15] and QGSP [10, inf[.
  • New physics-lists with special EM options: FTFP_BERT_EMV, FTFP_BERT_EMX, QGSP_BERT_EMX, QGSP_BIC_EMY (opt3 EM physics).
  • CHIPS physics list, using CHIPS model for all energies. Added two builders G4QNeutrinoPhysics, G4QPhotoNuclearPhysics with messenger.
  • QGSC_QGSC: using QGS model with interface to CHIPS at all energies, starting from E=0; for neutrons it still uses LHEP models for the radioactive capture and fission.
  • QGSC_CHIPS: using QGS model with interface to CHIPS at all energies; for pA interactions the native CHIPS process G4QCollision is used below 290 MeV.
  • Added builders for DNA, Livermore and Penelope physics processes. Upper limit of Livermore models is set to 1 GeV. Adapted G4EmLivermorePolarizedPhysics builder as for the non polarized case. Added high energy limit to e- models in G4EmLivermorePhysics and G4EmLivermorePolarizedPhysics
  • Added G4OpticalPhysics and G4OpticalPhysicsMessenger classes.
• Physics lists qualification changes (obsolete/partially disabled/experimental/supported):
  • Obsolete and disabled (removed): FTF_EMV, replaced by FTFP_BERT_EMV; QGSC_EFLOW, replaced by QGSC_CHIPS; QGSP_EMX, replaced by QGSP_BERT_EMX.
  • Obsolete: FTFC, LHEP_BERT, LHEP_BERT_HP, LHEP_PRECO_HP QGSP_BERT_DIF, QGSP_BERT_NQE, QGSP_DIF, QGSP_EMV, QGSP_EMV_NQE, QGSP_NQE, QGSP_QEL, FTFP, QGSP, QGSC, QGSC_CHIPS, QGSC_EMV, QGSC_QGSC..
  • Explicitely marked experimental: QGSP_BERT_EMX, QGSP_BERT_TRV, QGSP_FTFP_BERT, QGSP_INCL_ABLA.
  • Marked as supported: FTFP_BERT, FTFP_BERT_EMV.
    G4PhysicsListFactory: updated choice of supported physics-lists.
• EM physics builders and options:
  • G4EmStandardPhysics, G4EmStandardPhysics_option1: use 9.2 physics configuration; use ApplyCuts=true.
  • G4EmStandardPhysics_option2: use G4eMultipleScattering for e+-. Use default StepFunction for ions.
  • G4EmStandardPhysics_option3: use updated G4UrbanMscModel2 and use G4RayleighScattering process.
  • Separated definition of processes for He3 and He4 ions; G4eMultipleScattering for e+, e-, G4MuMultipleScattering for muons.
  • G4UrbanMscModel93 for e+- and G4WentzelVIModel for muons for all builders except Opt0 and Opt1.
  • G4ParameterizedIonModel for Opt3 and all combined builders.
  • StepFunction (0.1, 0.02*mm) for ions inside Opt2.
  • G4EmStandardPhysics_option3, G4EmLivermorePhysics, G4EmPenelopePhysics, G4EmLivermorePolarizedPhysics: added G4NuclearStopping process for GenericIon, He3, alpha particles. Set finalRange=10um for ions.
• Updates to lists:
  • Added MaxEnergy() method in FTF builders.
  • QBBC list: removed templated interface and moved inline code to source; use new interface through G4VHadronPhysics class.
  • Change to G4HadroHElasticPhysics to make use of G4WHadronElasticProcess as part of the design review for elastic.
  • Change to QGSP_BERT_EMV to make use of G4HadroHElasticPhysics.
  • Updated QBBC physics-list and fixed cross-section for hadron elastic.
  • QGSP_BERT_NOLEP: use G4HadronQElasticPhysics.
  • Added initial version of G4IonQMDPhysics.

• G4VUserPhysicsList:
• Added method CheckParticleList() to probe consistency of list of particles before constructing processes; check being executed within call to BuildPhysicsTable(). Added DisableCheckParticleList() method as well and modified BuildPhysicsTable() to avoid problem when ions are created in the Pre-Init state.
• Adapted implementation for handling proton cuts.
• G4RunManagerKernel: now invokes CheckParticleList() from G4VUserPhysicsList just before invoking SetCuts(). Also checking G4VUserPhysicsList to confirm no particle is registered when G4RunManagerKernel is instantiated. Fixed two minor bugs for the dump printout of regions.
• Added UI command /run/setCutForAGivenParticle in G4UserPhysicsListMessenger.
• New classes implementing the "Reverse MonteCarlo" technique: G4AdjointSimManager, G4AdjointSimMessenger, G4AdjointPrimaryGeneratorAction.

• Modified implementation of GetVelocity() for particle-change for performance improvement.
• Set lowEnergyLimit to 1 eV in G4ParticleChangeForLoss. Added method SetLowEnergyLimit(), also to G4VUserTrackInformation; added a string property to indicate type of UserTrackInformation.
• Correction to G4SteppingManager::InvokeAtRestDoItProcs() in the handling of G4Scintillation process.
• New classes implementing the "Reverse MonteCarlo" technique: G4AdjointCrossSurfChecker, G4AdjointSteppingAction.
• G4Trajectory, G4SmoothTrajectory, G4RichTrajectory: added initial kinetic energy. G4RichTrajectory: added final kinetic energy. G4RichTrajectoryPoint: added "remaining energy" at each step and attributes.

• Overall
  • Rich Trajectories and Rich Trajectory Points now include additional information such as the complete volume path of the starting and ending volume including all replica and copy numbers, etc.
  • Fixed bug in screen location and screen ratio handling for X/Xm/Qt/Win32 viewers.
  • Vector and bitmapped printing through /vis/ogl/printEPS is now fully functional for all OpenGL viewers (previously only for OGLIX and OGLSX). Related new commands are /vis/ogl/set/printSize and /vis/ogl/set/printFilename.
• Management
  • G4VisManager: implemented handlers notification. A user may notify scene handlers that the scene has changed, so that they may rebuild their graphics database, if any, and redraw all views.
  • Introduced /vis/initialize UI command; instantiating /vis/verbose and /vis/initialize in constructor of G4VisManager so that a user may set verbosity before initialisation.
  • New UI command, /vis/viewer/set/targetPoint, sets a specific point to be the center of the view. Helps with doing extreme zoom in to a specific point.
  • New UI command, /vis/scene/add/psHits for visualization of scoring quantities to gMocrenFile.
  • Improved end of run event keeping information.
  • Allow for arbitrary viewer deletion (fixes some situations that had caused crashes).
  • Allow refresh in GeomClosed as well as Idle state.
  • G4VisExecutive: made methods inline.
• Modeling
  • Added warning if logical volume is not defined.
  • G4PhysicalVolumeModelTouchable: new class to allow visualization attributes to be individually set for each volume within a nested parameterization.
  • Added unit to setAux/StepPtsSizeType UI commands so that size can be specified either in pixels or relative to screen size.
  • Added G4PSHitsModel for gMocren driver.
  • G4ModelApplyCommands: added call to NotifyHandlers() to force update of all viewers after a change of model, e.g., trajectory colour, point size, etc.
• Externals
  • gl2ps: Added dependency on zlib (even if G4LIB_BUILD_ZLIB is not set).
  • Corrected setup for building DLLs on Windows.
• gMocren
  • First Beta release of this new visualization driver to output data in the gdd format used by the gMocren volume visualization tool. gMocrenFile can include volume data from sensitive detector volumes, or from command-based scoring meshes, in addition to detector geometry and trajectories. Use the new UI command /vis/scene/add/psHits to add scorer hits to gMocren output. For details, see the Application Developers Guide and: http://geant4.kek.jp/gMocren/.
• HepRep
  • New UI commands of the form /vis/heprep allow easy control of output file name and output directory, whether a file name should be reused (rather than having a sequential suffix appended at each new file) and whether culled geometry volumes should be included.
  • Changed messenger to a singleton to facilitate sharing by HepRepFile and HepRepXML.
  • Fixed a bug in representation of G4Tubs which had caused wrong visualization for cases in which the tube was at an unusual angle.
• OpenGL
  • G4OpenGLQtViewer: ensure the GL frame size will be correct by setting a zero size margin around.
  • G4OpenGLStoredQtViewer: removed picking to true at initialisation.
• OpenInventor
  • Created new "externals" module for gl2ps library, moving out code from OpenInventor module. Added ability to draw markers.
• Qt
  • Continued improvements to this new driver. The Qt visualization driver and the associated Qt User Interface now work well on almost all platforms and with both Qt3 and Qt4 (but still considered a Beta release due to a few rare combinations of commands on certain platforms that can cause crashes).
• RayTracer
  • Fixed a crash of RayTracerX on some platforms. Addressing problem report #1072.

• G4Py: updated configuration script and wrappers to physics-lists.
• Momo: updated documentation.

• New data for measured optical surface reflectance, RealSurface.1.0:
  • Path to data controlled by variable G4REALSURFACEDATA
• New low-energy data set, G4EMLOW.6.9:
  • Corrected corrupted data files: ioni/ion-sp-38.dat, ioni/ion-sp-56.dat and ioni/ion-sp-57.dat. Addresses problem report #1042.
  • Extended energy range of Champion e- elastic scattering model from 10 KeV up to 10 MeV.
  • Extended high energy coverage of proton excitation and ionization.
  • Improved precision of e- ionisation Born model.
  • Included ICRU 73 stopping powers and revised tables provided by P.Sigmund/Univ. of Southern Denmark.
  • Updated Geant4-DNA e- elastic Champion files and e- ionisation Born files.

• Various fixes, improvements, adoption of new features.
• Updated reference outputs.
• advanced
  • Retired advanced examples: cosmicray_charging, radiation_monitor, Tiara.
  • air_shower
    • Changed UltraPhysicsList to use new processes G4eMultipleScattering, G4MuMultipleScattering and G4hMultipleScattering.
  • amsEcal
    • New example implementing the AMS Ecal calorimeter.
  • brachytherapy
    • Updated physics-list and analysis setup.
  • composite_calorimeter
    • Use QGSP_BIC_EMY as default physics-list to test the low-energy EM.
  • gammaray_telescope
    • Updated physics-list setup.
  • hadrontherapy
    • Reviewed physics implementation. Now Hadrontherapy can be launched with physics-lists, packages and built-in physic models. All models can be activated via macro command.
    • Implemented new low-energy models Livermore and Penelope.
    • Modified and revised the README (now in PDF format) and macro files.
    • Added messengers to control the event number and to draw only particular tracks.
    • Added a method to retrieve stopping power values for protons, alphas and electrons.
    • Added local INCL/ABLA physics list for deuterons, tritons and alphas.
    • Added first version of the IAEA benchmark geometry based on E.Haettner's thesis. Added ability to remove the phantom by setting its thickness to zero. Added ability to select the geometry using macro commands.
    • Collect simulation metadata: number of events, distance of the detector (IAEA geometry), depth of the phantom (IAEA geometry), beam energy, energy error. Added ability to produce angular distribution plots.
    • Added ability to change the name of the output file between runs.
    • Added ability to use command based scoring; IAEA geometry: now can produce Bragg peak using command based scoring.
    • Added possibility to make a graphical user interface (GUI) using the Qt libraries.
    • Added some functionalities in order to change, via messengers, geometry, voxelization of the detector and disposition in the space of the detector/phantom.
  • lAr_calorimeter
    • Use QGSP_BIC_EMY as default physics-list to tests the low-energy EM.
  • medical_linac
    • New release of the example: updated physics models and deleted old and no longer supported processes; added possibility to activate the physics lists and reference physics-lists directly via UI command; added possibility to set cuts and max allowed step lenghts via UI commands.
  • microbeam
    • Updated physics-list to migrated Livermore processes.
    • Corrected plotting fuctions and switching field value.
  • microdosimetry
    • Updated physics-list to migrated Geant4-DNA processes. Modified SteppingAction accordingly. Updated README.
    • Replaced H2O material by NIST G4_WATER.
    • Added GenericIon definition to PhysicsList.
  • radioprotection
    • Updated physics-lists setup.
    • Changed analysis output format to XML.
  • Rich
    • Added possibility to use QGSP_BIC_EMY as physics-list; by default, the example's physics-list is used still.
  • xray_fluorescence
    • Migrated physics-list to new low-energy processes design
    • Migrated XrayFluoDataSet to meet new specifications in G4VEMDataSet.
  • xray_telescope
    • Updated and simplified physics-list.
• extended
  • analysis/A01
    • A01DetectorConstruction, A01MagneticField: improved handling of field. Addresses problem report #1012.
    • Removed A01Trajectory, to now use default G4Trajectory.
    • Update physics-list to use new multiple scattering classes.
    • Introduced use of G4UIExecutive.
    • General update to README.
  • biasing/ReverseMC01
    • New example demonstrating the use of Reverse MonteCarlo technique in Geant4.
  • electromagnetic
    • Physics list: renamed "standard" to "local"; "emstandard" to "emstandard_opt0". Substituted obsolete low-energy physics builders with builders from the reference physics_list library.
    • Replaced obsolete G4MultipleScattering in PhysListEmStandard.
    • Updated README and macro files.
  • electromagnetic/TestEm2
    • Substituted obsolete low-energy physics builders with builders from the 'physics_list' module.
    • Reduced number of hard-coded materials, added a possiblity to use NIST materials.
    • Some cleanup to macro files.
  • electromagnetic/TestEm3
    • Fix to HistoManager::PrintHisto(): avoid empty ascii file.
    • Removed G4EMConfigurator in physics-list.
  • electromagnetic/TestEm5
    • Added UI command /testem/phys/cutLimit allowing to change low limit for production threshold.
    • Added missing notification to G4RunManager when geometry or material are modified between runs.
    • Added PAI models configuration to Physics List.
    • Modified cuts in PhysicsList. Added blaha.mac input macro.
    • Removed UI command /testem/phys/cutLimit.
    • PhysicsList: removed obsolete low-energy EM builders, substituted by builders from the 'physics_list' module; added "standardGS" option: use GoudsmitSaunderson multiple-scattering model.
    • New local physics-list PhysListEmStandardGS.
    • Removed G4EMConfigurator in physics-list.
  • electromagnetic/TestEm6
    • Increased upper limit of energy in the tables to 1000 TeV.
    • Added StackingAction to allow for removal of secondary particles.
  • electromagnetic/TestEm9
    • Changed options to physics-list.
  • electromagnetic/TestEm18
    • Added UI command /testem/phys/cutLimit allowing to change low limit for production threshold.
    • Changed options to physics-list.
    • Removed UI command /testem/phys/cutLimit.
  • electromagnetic/TestEm13
    • Replacement with new Livermore and Penelope physics-lists.
  • electromagnetic/TestEm14
    • Replacement with new Livermore and Penelope physics-lists.
  • exoticphysics/monopole
    • Added messenger to define mass and charge of the monopole.
  • field
    • Fixed density of STP Carbon Dioxide in DetectorConstruction.
  • field/field04
    • Introduced /rndm/autoSeed UI command and changed default seeding of random generator.
    • Changed F04PhysicsList and F04PhysicsListMessenger to use G4PhysListFactory.
    • Introduced use of G4UIExecutive.
  • field/field05
    • New example to test the spin-frozen condition whereby the g-2 muon spin precession is canceled by an applied electric field. Courtesy of Hiromi Iinuma from KEK.
  • hadronic/Hadr00
    • Added possibility to define physics-list as a job parameter.
    • Added script for the test run of all physics-lists.
  • hadronic/Hadr01
    • Updated physics-list. Using FTFP_BERT_TRV.
    • Added histograms on energy depositions: total/electromagnetic, pion/proton.
  • medical/DICOM
    • Changed DICOM Handler to read CT files coming from CONQUEST DICOM server.
  • medical/electronScattering, electronScattering2
    • Two alternate approaches to simulating the published electron scattering benchmark experiment of C. K. Ross et al.
  • medical/fanoCavity[1,2]
    • Corrected computation of 'error' in EndOfRun.
    • Using GoudsmitSaunderson multiple-scattering model. Added cut for proton recoil in case of Single Scattering.
  • medical/GammaTherapy
    • Updated physics-list for EM.
  • optical/LXe
    • Exhibit the usage of a complex index of refraction for the photo cathode.
    • Use of G4OpticalPhysics builder/messenger, instead of LXeOpticalPhysics.
    • Introduced use of G4UIExecutive.
  • optical/wls
    • New example tracking optical photons in a realistic plastic scintillator bar geometry with a wave length shifting (WLS) fiber and propagating the WLS photons inside the fiber. Courtesy of Siu Pak Mok from TRIUMF.
  • persistency/gdml
    • G01: added new GDML samples for matrices+loops and optical-surfaces. Modified matricesandloops.gdml, following new specifications for loops. Added possibility to run application in batch mode giving a macro as input (e.g.: load_gdml solids.gdml g01.gdml g01.in); option to be used by system testing for batch test. Added assembly.gdml to demonstrate support for 'assembly' tags. Added usage of 'expression' and 'quantity' in solids.gdml.
    • G02: Added missing inclusion of headers after cleanup.
    • G03: Extended example to show also capability to write extended schema; added custom writer class. Renamed translation unit for main to "gdml_ext.cc".
  • persistency/P03
    • Use G4XMultipleScattering in physics-list.
  • polarisation/Pol01
    • Updated multiple-scattering physics.
  • radioactivedecay/exrdm
    • Fixed trivial compilation warning when compiled without G4ANALYSIS_USE.
• novice
  • Introduced use of G4UIExecutive.
  • N02
    • Use G4MuMultipleScattering for muons; separated alpha from GenericIon.
  • N03
    • Use G4MuMultipleScattering for muons; separated alpha from GenericIon.
  • N06
    • Use G4eMultipleScattering, G4MuMultipleScattering and G4hMultipleScattering instead of standard G4MultipleScattering.