Cexmc User`s Manual - Geant4
Contents
1. Energy deposit ED hits are implemented in class CexmcSimpleEnergyDeposit collects ED in the monitor counter and its descenders CexmcEnergyDepositInLeftRightSet ED in veto counters and CexmcEnergyDepositInCalorimeter ED in calorimeters All classes use event map of type G4ThitsMap lt G4double gt the main difference between them is how parts of detectors are indexed in the event map Indices of G4THitsMap lt gt have type G4int CexmcSimpleEnergyDeposit collects all events in index 0 CexmcEnergyDepositinLeftRightSet collects events in the left detector in index 0 and events in the right detector in index 1 lt lt 16 i e an integer value with 16 bit set CexmcEnergyDepositInCalorimeter marks that an event has occurred in the right calorimeter setting 16 bit number of row occupies bits 8 15 number of column occupies bits 0 7 Track points 7P hits are implemented in a similar manner in classes CexmcTrackPoints CexmcTrackPointsInLeftRightSet and CexmcTrackPointsInCalorimeter The main difference is that ED hits store integral value of all energy deposited in a detector or one of its part but TP hits store information about multiple tracks using specific id which contains type of a track and its copy number Therefore CexmcTrackPoints collects events in indices that correspond to the value of the id CexmcTrackPointsInLeftRightSet marks that an event has occurred in the right detector setting 24 bit and the id occupies bits 02. They must correspond to absolute positions and rotations of the target and the calorimeters in the world volume Their values are used in reconstruction procedures A logical volume of the calorimeter must be tagged by tag auxiliary with auxtype Special Volume and auxvalue 2 The calorimeters must be replicated horizontally and vertically which makes rows and columns of separate crystals Exact number of replicated volumes is not important Crystals may be wrapped in paper or other materials Following constraints exist but they can be removed in future currently they are mostly referenced in CexmcHistoManger solids of logical volumes of calorimeters crystals and the monitor counter must be boxes solid of logical volume of the target must be a tube a logical volume of the calorimeter must be tagged as SensitiveRegion with auxvalue 0 logical volumes of the monitor counter and veto counters must be tagged as Special Volume with auxvalue 0 and 1 respectively Figure 1 the physical setup 3 Run modes Cexmc has 3 major operational modes 1 Straight mode or Monte Carlo mode The program will read preinit and init macros then calculate acceptances and optionally save data in project files Project files are saved in a directory defined by environment variable CEXMC_PROJECTS_DIR or in the current directory if it is not defined name of the project is specified by option w Preinit and init macros are set by options p and m resp
3. 23 CexmcTrackPointsInCalorimeter marks that an event has occurred in the right detector setting 24 bit number of row occupies bits 16 23 number of column occupies bits 8 15 and the id occupies bits 0 7 Energy deposit hits are used in method CexmcEnergyDepositDigitizer Digitize to detect if the setup has triggered First of all Digitize converts ED data from calorimeters if user has parameterized resolution of calorimeters in init macro with commands cexmc detector addCrystalResolutionRange Then Digitize applies the trigger logic The trigger logic takes information from the monitor counter veto counters and calorimeters and tests if the monitor ED is bigger than related monitor threshold it means that a charged particle 9 supposedly 27 has passed the monitor ED in each veto counter is smaller than veto counter thresholds it means that no charged particle has passed the veto counters and full ED in inner crystals of each calorimeter is bigger than calorimeter thresholds it means that calorimeters have supposedly caught the decay products of output particles i e the 2y The energy thresholds are specified by user in init macro There are several variations of basic trigger logic related mostly to filtering out events when most energy deposit spot is located in outer crystals of a calorimeter all variations of the trigger logic are controlled by user in init macro In CexmcEventAction EndOfEventAction al
4. current event Tpt track points trigger events are related to events with studied interaction triggered and file fdb and edt energy deposit trigger events to events with the setup trigger and file edb As such using most of tpt commands requires that event data were saved with the greedy event data verbosity Only a few tpt commands are safe when events data have been saved in an economy mode they are listed in the example cf Here is the content of the example cf to feel the thing I EDT examples EDT delete events when sum of absorbed energy in left calorimeter is more than specified value delete edt if Sum clEDcol gt 350 MeV EDT keep events if sum of absorbed energy in two specified crystals in left calorimeter is more than sum of absorbed energy in outer crystals of left calorimeter keep edt if clEDcol 2 2 clEDcol 2 3 gt Sum Outer clEDcol EDT delete events if sum of absorbed energy in inner crystals of right calorimeter is more than specified value delete edt if Sum Inner crEDcol gt 200 Mev II TPT examples Can be safely used only for rich event data sets TPT make diagonal cut in monitor counter delete tpt if bp_mon_posl 1 lt bp mon posl 2 TPT make target radius smaller delete tpt 1f Sqrt Sqr ep tgt_pesll1 Sqr ep_tgt posl 2 Sqr ep gt posl 31 gt lt om III TPT examples Can be safely used for any ev
5. eventDataVerboseLevel and skipInteractions WithoutEDT that can be set in init macro in cexmc run directory There are 4 levels of event data verbosity 1 Do not save event data cexmc run eventDataVerboseLevel is nosave 2 Save minimum event data economy mode In this case eventDataVerboseLevel is trigger All events with triggered interaction or the setup trigger are marked in the fdb file edb file has verbose information only on events with the setup trigger Data can be used in replay mode but if this mode was used in rep ay run mode then next rep ay run will give wrong acceptances 3 Save minimum event data with skipInteractions WithoutEDT equal false This mode is effective for the replay run mode and must be used in long chain runs 1 e multiple sequential replay runs 4 Save all event data greedy mode In this case eventDataVerboseLevel is interaction Every event entry available in fdb file has a counterpart in edb file This mode is useful if user wants to use the custom filter for simulated data in future but size of data will be much larger than in economy modes 10 Custom filter The custom filter is an optional feature Compilation of the custom filter is controlled by macro CEXMC_USE_CUSTOM_ FILTER It can be used only in the rep ay run mode The idea of the custom filter is to make it possible to remove specific events from fdb and edb files in a very flexible way This can be useful for example when the ex
6. on Fig 1 The setup is completely described in file ht gdml Besides the geometry this file contains information about sensitive detectors using tag auxiliary with attributes auxtype equal to EnergyDepositDetector or TrackPointsDetector Tag auxiliary with attribute auxtype equal to SpecialVolume marks logical volumes that play important roles in the program the monitor counter the target and the calorimeters Tag auxiliary with attribute auxtype equal to SensitiveRegion marks logical volumes where user can define custom production cuts particularly setting custom cuts is supported in calorimeters Using ht gdml is not mandatory In fact geometry file is set by command cexmc geometry gdmlFile in preinit phase In principle user can completely change geometry description of the setup but any description must meet following requirements l A logical volume of the target must be tagged by tag auxiliary with auxtype SpecialVolume and auxvalue 3 The corresponding physical volume is referenced in user stepping and tracking actions The setup must have 2 arms left and right The arms contain detectors that collect energies of particles produced in studied interactions Also names of physical volumes in the right arm that have been made sensitive detectors must contain Right Three pairs of positions and rotations must be defined TargetPos TargetRot CalorimeterLeftPos CalorimeterLeftRot CalorimeterRightPos and CalorimeterRightRot
7. reconstruction model we can reconstruct physics processes from experimental data obtained from the physical setup To be sure that calculated acceptances correspond to the experimental data we have to apply same reconstruction model to the true Monte Carlo data The main target of the experiment is to find differential cross sections of the studied interactions The reconstruction model has to retrieve the angle of the studied interaction output particle in the Center of Mass system both for experimental and true Monte Carlo data Knowing angles of output particles let us parameterize experimental and Monte Carlo data by these angles and thus find the cross sections Cexmc finds two kinds of setup acceptances The first kind is the real acceptances they are parameterized by angles from the true Monte Carlo data the second kind is the reconstructed acceptances which are parameterized by reconstructed angles of the output particle The reconstruction model is applied only when the setup has triggered and includes following parts 1 Reconstruction of energies of the incident beam particle and of the output particle s decay products the 2y having entered the calorimeters vertices of the interaction in the target and of entries of the 2y in the calorimeters 2 Reconstruction of the studied interaction and the angle of the output particle in the Center of Mass system 3 Optionally reconstruction model may include different energy cuts and othe
8. Cexmc User s Manual Index TAMOS AO a A AAA AA A AA 3 ZAS UN AM o e de o 3 A A STR RRA TOT te RPE A R A STORER TET Ot ORE PE ee 5 O 5 5 Fastsim lation modi ed ia 7 6 Charge exchange production Mode dni dado iaa 9 7 Sensitive detectors hits and ES SN AS 9 8 d AITE EOI TST eA EO D PAARE EEEE AAT E E 10 A Sanda ubisesteeacatulunonaesouesduaduazsatesnedSensaletnn E eE EOE EA a 11 10 C stoni Mier sre seca slg acs Dinas ast nese een cl eae cat EEE Mesh Sunes RA sedate A 11 LOA na loas e 13 12 User command la as 13 LI VA AS RS LAS ele A a 13 1 Motivation Program Cexme which stands for Charge EXchange Monte Carlo was used to find best geometry of a physical setup and calculate acceptances of this setup for 2 types of hadronic interactions tpn e Tp gt wmn The interactions were studied for different energies of 1 in momentum range 400 720 MeV c for n production and 700 730 MeV c for y production The physical setup can have different degrees of freedom calorimeters are freely moved combining different distances and angles in respect to the target see Fig and its best geometry is roughly such a configuration when acceptances have maximum values Setup acceptances can be calculated using Monte Carlo method One might define them as a set of numbers that give a fraction of triggered events from the whole number of simulated interactions when output particle is scattered in a given solid angle in the Center of Mass system
9. HadronicInteraction and CexmcProductionModel The only method it provides is ApplyYourself Apply Yourself uses a phase space generator to generate secondaries and add them to theParticleChange There are two phase space generators that user can choose original cernlib FORTRAN routine GENBODY called from CexmcGenbod Generate and same routine reimplemented in class CexmcReimplementedGenbod Implementation of GENBOD in CexmcReimplementedGenbod was mostly adopted from CERN ROOT class TGenPhaseSpace The first choice CexmcGenbod makes the program depend on FORTRAN libraries and cernlib The choice is controlled by macro CEXMC_USE_GENBOD by default CEXMC_USE_GENBOD is not defined 7 Sensitive detectors hits and digits Cexme collects two kinds of hits energy deposits in detectors and track points of various tracks of interest beam particle output particle decay products of output particle etc Related sensitive detectors are created in ht gdml with help of GDML tag auxiliary and attributes EnergyDepositDetector and TrackPointsDetector To avoid direct correlation between a logical volume name and a sensitive detector name in the program code a notion of detector role was introduced detector role is specified in the attribute auxvalue of the tag auxiliary There are 4 kinds of detector roles Monitor VetoCounter Calorimeter and Target Logical volumes may share same roles but one logical volume may not have 2 or more different roles
10. Thus setup acceptances are parameterized by angular ranges and strictly bound to physical setup trigger logic and studied interaction Knowing setup acceptances makes it possible to find differential cross sections of the studied interaction The value of differential cross section in a given solid angle can be deduced from experimental data and the value of acceptance for this angle da No LH dQ Ag where No is number of experimental events registered in a given solid angle and Aois calculated value of the setup acceptance for this solid angle Despite the program was developed for a concrete physical experiment its structure is flexible enough to support different setup geometries and interaction models This flexibility is a result of using GDML for description of setup geometry and a special fast simulation engine that provides easy replacement of studied interactions 2 Setup geometry The physical setup is constructed from a cylindrical liquid hydrogen target with radius 5cm two calorimeters that make 2 arms of the setup and contain 6x4 sets of CsI crystals a monitor counter before the target 2 veto counters before the calorimeters and various target constructive elements The calorimeters will detect decay products of the output particle 2 or y The most intensive decay mode for the both particles is 2y The 2 arms of the setup are responsible for catching the gammas and collect their energies The setup is depicted
11. ally it makes them occur almost in every event generated by Geant4 engine This almost arises from 2 major facts 1 Not all incident beam particles visit the target for example if simulated beam is very wide or a beam particle was scattered and missed the target 2 There is one important restriction on the distribution of studied interaction vertices it must be uniform along the beam path in the target From the point of view of an engine which samples vertices of studied interaction in the target it means that if geometry of the target makes path length of beam particles depend on position of the particle on the beam projection then probabilities of interaction sampling must differ for different positions of the particle on the beam projection For physical setup on Fig particles on left and right edges of a wide beam projection will only scratch left or right edges of the cylindrical target and an interaction sampling engine must miss most of interactions for such edgy particles to ensure uniform distribution of interaction vertices in the target Many methods related to fast simulation engine are located in classes CexmcPhysicsManager and CexmcStudiedProcess this class inherits G4WrapperProcess and registers a physical process supplied in template argument of class template CexmcStudiedPhysics in its method ConstructProcess However the most interesting parts occur in CexmcTrackingAction PreUserTrackingAction and CexmcSteppingA c
12. d by a reference physics list argument StudiedPhysics corresponds to user s physics and must be implemented as a class template parameterized by a G4VProcess subclass and derived from G4VPhysicsConstructor argument ProductionModel corresponds to user s interaction model this is a class template parameterized by an arbitrary type in case of Charge Exchange production model it is parameterized by output particle s type e g 2 or n Production model must inherit class CexmcProductionModel which provides methods to access production model data and methods to set and access triggered angular ranges Concrete production model factory is instantiated in CexmcBasicPhysicsSettings hh ifdef CEXMC USE QGSP BIC EMY typedef QGSP BIC EMY CexmcBasePhysics else typedef QGSP_ BERT CexmcBasePhysics endif typedef CexmcProductionModelFactory lt CexmcBasePhysics CexmcHadronicPhysics CexmcChargeExchangeProductionModel gt CexmcChargeExchangePMFactory typedef CexmcChargeExchangePMFactory CexmcPMFactoryInstance Two reference physics list are supported from the makefile QGSP_BERT and QGSP_BIC_EMY they are controlled by macro CEXMC_USE_QGSP _BIC_EMY If user wants to use a different reference physics list he can edit CexmcBasicPhysicsSettings hh CexmcProductionModelFactory has method Create which creates and returns an object of class template CexmcPhysicsList instantiated with template arguments sup
13. ectively In the straight mode preinit macro must be specified explicitly as far as desired production model can be instantiated only in preinit phase 2 Replay mode or Read project mode In this mode the program will not use common Geant4 s event loop Instead it will sequentially read event data from an existing project and pass them into CexmcEventAction EndOfEventAction The read project is specified by option r This mode is useful when user wants to recalculate data from an existing project with different conditions for example with different reconstruction parameters The results of run can be written again into another project In fact user can make long chain of write read write runs sequentially tuning different parameters but validity of results of a long chain run depends on event data verbose level specified in init macro in in between runs 3 Show results mode or Output mode The program will output various data from an existing project specified by option r Type s of data are specified in option o For example to show results of a run user can specify orun in command line To show events geometry and run results user can specify oevents geom run Replay and show results modes are available only if macro CEXMC_USE_PERSISTENCY was enabled at compilation time There are also 3 different run modes classified by user interaction type 1 Batch mode The program will run without any interaction with user 2 Int
14. ent data sets TPT delete all events in angular range 0 0000 1 0000 delete tpt if op cosTh SCM lt 0 TPT delete all events with id less than 500 delete tpt if event lt 500 TPT delete all events when monitor was not triggered TPT delete all events without TPT i e false triggered events 11 Histograming Histograming is an optional feature implemented using CERN ROOT libraries Compilation of the histograming manager is controlled by macro CEXMC_USE_ROOT this macro is set automatically in the makefile if ROOT libraries were found in the system Histograms are saved besides other project files in a file with extension root Available histograms can be listed with command cexmc histo list Live histograms are available in interactive Qt sessions they can be shown on screen even if no project is written 12 User commands There are many user commands available in cexmc directory Subdirectories include geometry physics gun detector event run reconstructor vis and histo All user commands are well documented 13 Utilities There are several utilities in directory util Utilities mkjob and jobprog are related to running Cexmc in the gridengine environment mkjob creates a job and jobprog shows all running Cexmc instances and related process information like completeness in percents and current size of written project files Utility mkacchist reads output
15. eractive mode The program provides an interactive shell To run in the interactive mode command line option i must be specified 3 Graphical Ot mode This mode is specified by command line option g In Cexme number of events specified by command run beamOn may have different meanings which depends on the value set by command cexmc run eventCountPolicy If eventCountPolicy is all or not specified then the run manager will generate the specified number of events as usual 1f 1t 1s interaction then the run manager will generate the specified number of the studied interactions and if it is trigger then the run manager will generate the specified number of the setup triggers 4 Physics list Cexmc physics list consists of two parts one of Geant4 s reference physics list and a production model provided by user for fast simulation in the target Both reference physics list and production model are instantiated as template arguments of class template CexmcProductionModelFactory which is defined as template lt typename BasePhysics template lt typename gt class StudiedPhysics template lt typename gt class ProductionModel gt class CexmcProductionModelFactory public static G4VUserPhysicsList Create CexmcProductionModelType productionModelType static CexmcBasePhysicsUsed GetBasePhysics void private CexmcProductionModelFactory a Template argument BasePhysics is to be substitute
16. l collected energy deposits data are converted into a new object of type CexmcEnergyDepositStore which is then passed to the reconstructor methods can be printed out on screen used for histograming and saved in event data project files Class CexmcEnergyDepositStore uses G4Allocator as far as its objects are created and deleted in every EndOfEventAction Track points hits are used in method CexmcTrackPoitntsDigitizer Digitize to detect if a studied interaction has occurred In CexmcEventAction EndOfEventAction track points data are converted into a new object of type CexmcTrackPointsStore which is then can be printed on screen marked by a visualizer the green dots on Fig 2a and 2b used for histograming and saved in event data project files Class CexmcTrackPointsStore uses G4Allocator as far as its objects are created and deleted in every EndOfEventAction 8 Physics reconstruction Energy deposit and track points hits are true or real Monte Carlo data One can compare them with the real physics that hides behind the experimental data In real world measured experimental data do not have to be the exact reflection of the underlying physics we cannot know exactly energy of the incident beam particles or coordinates of the studied interaction vertices in the target instead we have to suppose that the beam particle had some average energy and that the studied interaction vertex was exactly in the center of the target But using a specific
17. of cexmc launched in the show results mode and creates a C script which then can be used in the ROOT environment to produce a nice graphical image of the calculated acceptances like this Setup acceptances 730_60_65 100min_20100909 reccuts 0 0014 0 0012 0 001 0 0008 0 0006 1 0 8 0 6 0 4 0 2 0 02 04 06 08 1 Figure 4 setup acceptances from results of a run 13
18. perimental data have some known defect Imagine that you know that your left calorimeter was misconfigured in the experiment and all data with energy deposits more than a specific value are wrong With the custom filter you won t have to throw the experimental data or compile another algorithm in the Cexmc All you need is to remove defective events from the experimental data and create specific custom filter script to remove events that meet same dynamic requirements from Monte Carlo events The custom filter has even more power allowing to investigate changes in geometry of the setup after run for example user can 11 select to study an area within the target by removing events when the studied interactions occurred outside this area To achieve this dynamic flexibility the custom filter implements a specific arithmetic language using boost spirit library This language allows arbitrary arithmetic expressions with type safety in mind and is aware of variables related to events data and several functions like Sgr and Sqrt There is an example of a custom filter script example cf in the code repository A custom filter script consists of several expressions that are evaluated sequentially from top to bottom Expressions start from commands delete tpt keep tpt delete edt or keep edt appended by an arbitrary logical expression Evaluation stops as soon as current logical expression is evaluated to true and the related command is executed on the
19. plied by the factory CexmcPhysicsList inherits reference physics list and CexmcPhysicsManager template lt typename BasePhysics template lt typename gt class StudiedPhysics typename ProductionModel gt class CexmcPhysicsList public BasePhysics public CexmcPhysicsManager public CexmcPhysicsList public CexmcProductionModel GetProductionModel void G4bool IsStudiedProcessAllowed void const void ResampleTrackLengthInTarget const G4Track track const G4StepPoint stepPoint private StudiedPhysics lt ProductionModel gt studiedPhysics G4bool proposedMaxILInitialized G4double proposedMaxIL Class CexmcPhysicsManager implements a number of methods related to fast simulation model and has interface method GetProductionModel Created object is passed as physics manager into constructors of a number of management objects like CexmcEventAction CexmcRunAction CexmcTrackingAction and CexmcSteppingAction which enables them to access the production model object Figure 2a electro magnetic interactions Figure 2b studied interactions in the target caused by reference physics list caused by instantiated charge exchange production model 5 Fast simulation model Fast simulation model implements production model simulation in the target The idea of fast simulation is to make studied interaction occur in the target and only in the target as often as possible Actu
20. r parameters that 10 will increase quality of the calculated data but possibly make statistics smaller The first part of the reconstruction is implemented in class CexmcReconstructor The second part is implemented in class CexmcChargeExchangeReconstructor Both parts have plenty of settings that user can choose to adjust the reconstruction model for his requirements All these settings are configured in init macro by the cexmc reconstructor subcommands and may be freely customized in the replay run mode Reconstructed vertices of the studied interaction and the entries of the output particle s decay products in the calorimeters are visualized with red dots see Fig 2a and 2b 9 Persistency Cexmce uses concept of a project Project files are saved in a directory specified by environment variable CEXMC_PROJECTS_DIR and then can be retrieved by Cexmc in show results mode or reused in replay mode Name of the project is specified by option w Project files include files with extensions rdb run data base fdb fast events data base edb events data base root and gdml File with extension root stores histograms with extension gdm geometry description actual for this project with extension rdb results of the run fdb and edb events data of the run Files rdb fdb and edb have binary format and are created using boost serialize library Files fdb and edb are complementary their contents depend on values of
21. tion UserSteppingAction The PreUserTrackingAction tags particles which definition matches the incident particle of instantiated production model with an object of type CexmclncidentParticleTrackInfo derived from G4VUserTrackInformation This object will hold information related to fast simulation engine for this particle for example current track length in the target and sampled track length Entries 9 439429e 07 Mean x 0 1037 Mean y 0 1194 Mean z 0 0008057 RMS x 20 37 RMS y 22 88 RMS z 15 23 Figure 3 uniform distribution of studied interaction vertices in the cylindrical target Track length is sampled in CexmcPhysicsManager ResampleTrackLengthInTarget in PreUserTrackingAction This method can be called once again for the same particle in UserSteppingAction for example if the particle visits the target second time CexmcStudiedProcess watches to sample the studied interaction in its method PostStepGetPhysicalInteractionLength It returns difference between sampled for this particle track length in the target and current track length the particle already passed Geant4 uses this value to know when to simulate a physical process that was registered in CexmcStudiedProcess 6 Charge exchange production model Class template CexmcChargeExchangeProductionModel implements the two hadronic models mentioned in the beginning of this document It inherits G4
Download Pdf Manuals
Related Search