| Geant4 Cross Reference |
1 ----------------------------------------------
2 ----------------------------------------------
3
4 =========================================
5 Geant4 - Composite calor
6 =========================================
7
8 README
9 ---------------------
10
11 CompositeCalorimeter is an example of a test-
12 by the CMS Collaboration to validate Geant4 a
13 (in 1996) in a CMS Hadron calorimeter test-be
14 The name "Composite" for this example emphasi
15 test-beam had the goal of studying the hadron
16 part of the data was taken with the presence
17 crystal calorimeter in front of the hadronic
18 reproduce the situation as in the real CMS ex
19 The geometry of the simulation has been setup
20 very easily, at run time (therefore without n
21 see below for the details) the inclusion or e
22 electromagnetic calorimeter part.
23 Although some important aspects, for a detail
24 test-beam data and simulation, like beam prof
25 have been omitted here (to avoid too many tec
26 nevertheless, this example is able to reprodu
27 most of the relevant observables as measured
28 The output of this example consists of a set
29 and one ntuple which are stored on a ROOT fil
30 In our opinion, the most original "lesson" wh
31 advanced example for the Geant4 user is to sh
32 the Sensitive/Hit part of the simulation is t
33 Although the details of how this is done vary
34 experiment (it is worth, for instance, to com
35 advanced example lAr_calorimeter), the main d
36 are quite general: to have consistency, but a
37 and couplings as much as possibile, between S
38 and Visualization. Notice that the solution o
39 by CMS could appear "overdone" for the sake o
40 relatively simple test-beam setup; but it sho
41 that the same approach is used also for the f
42 simulation, as well as for any subdetector.
43
44
45 1. Setting up the environment variables
46 ---------------------------------------
47
48 The user should first setup, as "usual", the
49 variables (e.g. the script produced by cmake)
50 Then the specific setup for this example shou
51
52 > source envExample.csh in the case
53 or
54 > . envExample.sh in the case
55
56 The analysis part is based on the native g4an
57 the output is a ROOT file. This can be change
58
59
60 2. Sample run
61 -------------
62
63 Once the environmental variables are setup, y
64 CompositeCalorimeter
65 by configuring with cmake and then running th
66 Then, you can execute it using the Geant4 mac
67 as follows:
68
69 > ./CompositeCalorimeter test.g4mac
70
71 which simulate a few events, each being a 100
72 electromagnetic crystal calorimeter followed
73 without magnetic field.
74 The output is the ROOT file "ccal.root" .
75 See part "8. Analysis / Histogramming" below
76 content of that file.
77 If you run instead:
78
79 > ./CompositeCalorimeter
80
81 after having setup the Geant4 visualization v
82 you can visualize the geometry of the apparat
83 events. Similarly, you can get a very simple
84 that allows to select the particle type, its
85 of events (between a limited number of possib
86 For more details, see part "9. Visualization
87
88
89 3. Detector description
90 -----------------------
91
92 Let's start with a brief description of the t
93
94 There are two possible configurations:
95 i) HCAL only, that is only the hadronic ca
96 ii) ECAL+HCAL, that is the electromagnetic
97 is placed in front of the ha
98 ECAL is made of 23 cm long PbWO4 crystals (co
99 25.8 radiation lengths, and 1.1 interaction l
100 test beam a 7 x 7 = 49 matrix of crystals i
101 HCAL is a sampling calorimeter, with plastic
102 part and copper as absorber. 28 scintillator
103 absorber of varying thickness in between, and
104 and type of scintillator. More precisely:
105 --- layer 1: 2 cm of Copper
106 --- layer 2 to 7: 4 cm of Copper
107 --- layer 8 to 21: 6 cm of Copper
108 --- layer 22 to 27: 8 cm of Copper
109 For the scintillators: 2 mm passive Plastic;
110 1 mm passive Plastic.
111 The total length of HCAL consists of 152 cm o
112 The dimension orthogonal to the beam directio
113 The ECAL and HCAL considered here are prototy
114 Endcap calorimeters of the CMS detector (whic
115 region |eta| < 3.0 ; CMS has also a Forward c
116 the region 3.0 < |eta| < 5.0, but this part w
117 this test-beam setup). Notice, however, that
118 (28 instead of 19 in the Barrel or 18 in the
119 test-beam setup than in the real CMS detector
120 energy containment. Therefore, the ECAL+HCAL
121 to more than 11 radiation lengths as for the
122 19 layers of the Barrel have each 6 cm of abs
123 18 layers of the Endcap have each 6.6 cm of a
124 number of interaction lengths are rougly the
125 Five values of the magnetic field (parallel t
126 have been considered in the test-beam: 0.0 ,
127
128 In order to set the magnetic field, you have
129 dataglobal/fmap.tb96
130 and change the first number (which appears in
131 that file, on the first column; the unit bein
132 #. Field map
133 *DO FLDM
134 0.0 9 652.0
135 for example, if you want a magnetic field of
136 line must be set as follows (the magnetic fie
137 30.0 9 652.0
138
139 In order to deactivate either the ECAL or the
140 to comment out the corresponding line in the
141 using "#" as the comment character. For insta
142 without ECAL:
143 "HcalTB96" "tbhcal96"
144 #"CrystalMatrixModule" "tbhcal96xta
145
146
147 In this test-beam setup, at the back of ECAL,
148 material for support and readout, which has b
149 simulation. For the HCAL, only the fibres are
150 and because they have the same composition as
151 they are adequately represented in the simula
152 of the readout, including the photomultiplier
153 far away from the HCAL, and hence are not pre
154
155 Let's summarizes now the geometry description
156 As said in the introduction, this part is the
157 important of this example, but it is quite co
158 appreciated only in the context of the CMS so
159 particular in the relation between Simulation
160 Visualization. Therefore we limit ourself to
161 pointing to the internal CMS documentation fo
162
163 --- In order to share the same geometrical an
164 about CMS between Simulation, Reconstruct
165 avoiding inconsistencies, duplications, a
166 all these information is store, once for
167 (typically in XML format), instead of put
168 as usually done in simpler detector descr
169 the Geant4 examples, novice or advanced,
170 is kept inside the concrete class which i
171 G4VUserDetectorConstruction). For simplic
172 these "databases" are nothing more than A
173
174 datageom/ : tbhcal96.geom, tbhcal96hca
175 store the information abou
176 the HCAL, and the ECAL, re
177
178 dataconf/ : g4testbeamhcal96.conf, tes
179 store the information abou
180 (HCAL only, or ECAL+HACL)
181 Simulation and Reconstruct
182
183 dataglobal/ : fmap.tb96, material.cms,
184 The first one is the magne
185 intensity of the magnetic
186 orthogonal to the beam dir
187 the beam axis). The second
188 keeps the full collection
189 the CMS detector (not only
190 although we are simulating
191 The third one, rotation.cm
192 rotation parameters (angle
193
194 datavis/ : tbhcal96.vis, tbhcal96hcal.
195 visualization information for, re
196 experimental Hall, HCAL, an
197
198 --- In order to allow an high degree of flexi
199 level the user can choose which subsystem
200 should be simulated and can activate or d
201 parts, subsystem by subsystem. This can b
202 by modifying one of the above database in
203 of putting the hands on the code, recompi
204
205 --- There are two "parallel geometry factorie
206 classes, which are independent from the S
207 can be used, for instance, by the Reconst
208 is specific of the Simulation. In the lat
209 the geometry model), all the geometry fac
210 base class CCalG4Albe. Furthermore, using
211 of them derives also from the counterpart
212 The design of the CCalG4Able class helps
213 interchanging at the level of subdetector
214 transition from the simulation of the ent
215 just a part of it, or to a test-beam geom
216 Of course this modular, flexible, and gen
217 for free: the price to pay here is its co
218 otherwise unjustified if we limited ourse
219 of a relatively simple test-beam setup.
220
221 --- See "10. Classes Overview" below for a sc
222 various classes involved in the Geometry
223
224
225 4. Physics processes
226 --------------------
227
228 The factory physics list is used, therefore t
229 is steered by the environmental variable PHYS
230 (Note: if this environmental variable is not
231 which is used is FTFP_BERT).
232
233
234 5. Particle Generator
235 ---------------------
236
237 The 1996 test-beam has been taken with the fo
238 --- 225 GeV muons (for calibration)
239 --- 10 to 300 GeV pions
240 --- 10 to 300 GeV electrons
241 therefore the standard Geant4 Particle Gun ha
242 generator. Notice that, for the sake of keepi
243 complicated, the proper simulation of the bea
244 beam contamination have been neglected.
245
246
247 6. Hits
248 -------
249
250 In CMS there are two groups of hits: Tracker-
251 Only the latter one appears in this example.
252 For the same reasons, as seen for the Geometr
253 duplication of information and unnecessary co
254 Reconstruction, and Visualization, the simula
255 CCalG4Hit, doubly inherits from the common Ge
256 all hits, G4VHit, and from the "core" (i.e. s
257 CMS calorimeter hit class, CCalHit.
258 A new Hit object is created
259 - for each new particle entering the calori
260 - for each detector unit (i.e cristal or fi
261 - for each nanosecond of the shower develop
262 The information stored in each CCalHit object
263 - Entry : local coordinates of the entranc
264 in the unit where the shower sta
265 - the TrackID : Identification number of t
266 - the IncidentEnergy : kinetic energy of t
267 - the UnitID : the identification number of
268 (crystal, or fiber, or scint
269 - the TimeSlice : the time interval, in nan
270 hit has been created;
271 - the EnergyDeposit : the energy deposit in
272 Notice that all hit objects created for a giv
273 values for the first three pieces of informat
274
275
276 No Noise and Digitization
277 --------------------------
278
279 In order to keep the complexity of this examp
280 level, both noise and digitization effects ha
281
282
283 7. User Actions
284 ----------------
285
286 In this example. there have been used the fol
287
288 --- G4UserRunAction (the derived, concrete cl
289 it is used to invoke the Analysis object
290 the Run, to instantiate it and passing it
291 at the end of the Run, to inform it that
292 and therefore the histograms, ntuples, et
293
294 --- G4UserEventAction (the derived, concrete
295 it is used to examine, at the end of the
296 (calorimeter) hits, extract the various o
297 interesting (to the goal of understanding
298 of magnetic field on scintiallator; choic
299 thickness by optimizing resolution versus
300 the absorber depth in the energy caontain
301 calorimeter contribution in the electron
302 and finally call the analysis object to s
303 information on histograms and/or in the n
304 The name of the class "CCalEndOfEventActi
305 fact that at the beginning of the Event n
306
307 --- G4UserSteppingAction (the derived, concre
308 it is used to extract some "unphysical" i
309 experimentally measurable, although inter
310 understanding of the shower development),
311 and the deposit as a function of the time
312 for more details"), which is available on
313 at the end of Event, the analysis object
314 information on histograms.
315 Please notice that the stepping action is
316
317 --- G4UserStackingAction (the derived, concre
318 it is used to ensure that the same track
319 originating a shower appears in all hits
320 of class CCalHit) of the shower, in any c
321
322
323 8. Analysis / Histogramming
324 ----------------------------
325
326 The analysis part of CompositeCalorimeter is
327 and is based on the g4tool interfaces.
328 Both the histograms and the ntuple are saved
329 ROOT file "ccal.root" (default: this can be c
330 formats supported by the g4analysis tools).
331 Please note that in a multiple run session, t
332 the output file.
333 What the histograms and the variables of the
334 explained below:
335
336 Histograms h100 - h127 : energy deposit (in
337 (plastic scintillato
338 calorimeter module (
339 from 0 to 27, and th
340 ID = 100 + number of
341 Ntuple variables hcal0 - hcal27 : provide t
342
343 Histograms h200 - h248 : energy deposit (in
344 electromagnetic towe
345 7 x 7 = 49 towers, n
346 the corresponding hi
347 ID = 200 + number of
348 Ntuple variables ecal0 - ecal48 : provide t
349
350 Histograms h300 - h339 : total energy depos
351 electromagnetic crys
352 (either in a sensiti
353 in one of the 40 nan
354 words, histogram 30
355 contains the total d
356 I and I+1 nanosecond
357 (Notice that the tim
358 40 nanoseconds, is
359 bunch-crossing of 2
360
361 Histograms h400 - h469 : energy profile (in
362 sensitive (plastic s
363 (copper absorber), a
364 distance (in centime
365 ( ID histo = 400 + r
366
367 Histogram h4000 : total energy deposit (in
368 of either the electromagne
369 Ntuple variable edep provides the same info
370
371 Other ntuple variables are the following:
372 --- elab : energy (in GeV) of the inci
373 --- xpos, ypos, zpos : position (in mm
374 has been shot.
375 --- edec, edhc : total energy deposit
376 parts of, respectivel
377 and hadronic calorime
378 sum edec+edhc coinc
379
380 Notice that lateral profile (400-469) and ti
381 histograms show purely Monte Carlo quantitie
382 experimentally measured.
383 Please be careful that the range of the hist
384 in such a way to contain most of the entries
385 fill a large fraction of that range, whereas
386 fill only the first few bins (corresponding
387 therefore, when plotted they look almost emp
388 and the results are sensible. We suggest to
389 rather than the histograms, when the same in
390 from the ntuple.
391
392
393 9. Visualization / GUI
394 -----------------------
395
396 If you setup one of the following Geant4 envi
397 G4VIS_USE_DAWN
398 G4VIS_USE_VRML
399 G4VIS_USE_OPENGLX
400 which correspond to the use of DAWN, VRML, an
401 as visualization engine of Geant4, and set pr
402 PATH as well, it is then possible to visual
403 some events.
404 To do so, you have to run
405 > ./CompositeCalorimeter
406 without input file: you then see the detector
407 you can select the particle gun and its energ
408 case you want something different from the th
409 (which is a 100 GeV pi-), for example:
410 Idle> /gun/particle e-
411 Idle> /gun/energy 200 GeV
412 and then run some events, for example:
413 Idle> /run/beamOn 3
414
415 The tracks that are shown include both charge
416 of any momentum: if you want instead only cha
417 then you have simply to edit src/CCalEndOfEv
418 at the end of the method EndOfEventAction a
419 where the condition on the charge is made (it
420 straighforward to eventual add some other con
421 if you want to see only those particles that
422 kinematic conditions).
423
424 Rather than to specify "by hand" the type of
425 its energy, and the number of events, it is p
426 a very simple GUI (graphical user interface)
427 such choices, between a limited set of possib
428 Such GUI is based on Motif XmCommand widget,
429 straightforward, eventually, to make the nece
430 in order to use a different one.
431 The only thing you need to do to get the GUI
432 the following Geant4 environmental variables:
433 G4UI_BUILD_XM_SESSION=1
434 G4UI_USE_XM=1
435 Then, if you run the executable without speci
436 (like test.g4mac):
437 > $G4WORKDIR/bin/$G4SYSTEM/CompositeCalo
438 a window automatically pops out, with the men
439 can make your selection of particle type, ene
440 of events to be run.
441
442
443 10. Classes Overview
444 ---------------------
445
446 This is a schematic overview of the classes d
447
448 CCalPrimaryGeneratorAction
449 CCalPrimaryGeneratorMessenger
450 User action for primaries generator.
451
452 CCalDetectorConstruction
453 CCalAMaterial
454 CCalDataSet
455 CCalDetector
456 CCalEcal
457 CCalEcalOrganization
458 CCalG4Able
459 CCalG4Ecal
460 CCalG4Hall
461 CCalG4Hcal
462 CCalGeometryConfiguration
463 CCalHall
464 CCalHcal
465 CCalHcalOrganization
466 CCalMagneticField
467 CCalMaterial
468 CCalMaterialFactory
469 CCalRotationMatrixFactory
470 CCalVOrganization
471 CCalVisManager
472 CCalVisualisable
473 CCaloOrganization
474 CCalutils
475 Geometry and material definitions for the de
476 Notice in particular that:
477 CCalHall, CCalEcal, CCalHcal derive
478 CCalG4Hall, CCalG4Ecal, CCalG4Hcal d
479 corresponding classes and from CC
480 CCalEcalOrganization, CCalHcalOrgani
481 CCalVOrganization : each sensitiv
482 number for detector organization
483 ID not an hardware/electronic one
484
485 CCalHit
486 CCalG4Hit
487 CCalG4HitCollection
488 CCalSDList
489 CCalSensAssign
490 CCalSensitiveConfiguration
491 CCalSensitiveDetectors
492 CCaloSD
493 Hit and Sensitive Detectors.
494 Notice in particular that:
495 CCalG4Hit derives from G4VHit and CC
496 CCaloSD derives from G4VSensitiveDet
497
498 CCalActionInitializer
499 User-action initialization.
500
501 CCalAnalysis
502 Analysis manager class.
503
504 CCalRunAction
505 User run action class.
506
507 CCalEndOfEventAction
508 User event action class.
509
510 CCalStackingAction
511 User Stacking action class.
512
513 CCalSteppingAction
514 User Stepping action class.
515