While simulating the response of a 5x5 PWO Matrix Geant (4.3.0) enters an infinite loop when tracking e- (so far only seen with these particles) with a somewhat lower energy (within the first 1000 events). The 25 modules are made of segments of G4Sphere. When changing to G4Box the problem disappears (>100000 events). The physics list used is from ExN04. The verbose tracking output is: ********************************************************************************************************* * G4Track Information: Particle = gamma, Track ID = 21, Parent ID = 3 ********************************************************************************************************* Step# X(mm) Y(mm) Z(mm) KinE(MeV) dE(MeV) StepLeng TrackLeng NextVolume ProcName 0 1.43e+03 -0.763 -4.05 16.8 0 0 0 PD_physical initStep 1 1.46e+03 1.66 -10.2 16.8 0 36.7 36.7 expHall Transportation 2 1.47e+03 1.86 -10.7 16.8 0 3.03 39.8 PD_physical Transportation 3 1.55e+03 7.4 -24.9 0 0 84 124 PD_physical conv :----- List of 2ndaries - #SpawnInStep= 2(Rest= 0,Along= 0,Post= 2), #SpawnTotal= 2 --------------- : 1.55e+03 7.4 -24.9 4.54 e- : 1.55e+03 7.4 -24.9 11.2 e+ :----------------------------------------------------------------- EndOf2ndaries Info --------------- [...] ********************************************************************************************************* * G4Track Information: Particle = e-, Track ID = 37, Parent ID = 21 ********************************************************************************************************* Step# X(mm) Y(mm) Z(mm) KinE(MeV) dE(MeV) StepLeng TrackLeng NextVolume ProcName 0 1.55e+03 7.4 -24.9 4.54 0 0 0 PD_physical initStep 1 1.55e+03 7.45 -25 3.63 0.916 0.99 0.99 PD_physical Transportation 2 1.55e+03 7.45 -25 3.63 0 0 0.99 PD_physical Transportation 3 1.55e+03 7.45 -25 3.63 0 0 0.99 PD_physical Transportation 4 1.55e+03 7.45 -25 3.63 0 0 0.99 PD_physical Transportation and so on ... It seems that the particle is on the surface of a volume. I've read that there is a problem report on BREPSolids, but as far as I see G4Sphere is no BREP. The detector is constructed like this: G4double dAngleC = 2.05/140 * radian; G4double dAngleS = 0.05/140 * radian; G4double dAngleAll = dAngleC + dAngleS; for (G4int i=0; i<5; i++) { for (G4int j=0; j<5; j++) { digit1 = G4int((5.*i+j+1)/10.); digit2 = (5*i+j+1) - digit1*10; crystal_name = "PD"+G4String(char('0'+digit1))+G4String(char('0'+digit2)); G4double phiS = -3.*dAngleS-2.5*dAngleC+i*dAngleAll; if (phiS<0) phiS += 2.*M_PI; G4double thetaS = M_PI/2.-3.*dAngleS-2.5*dAngleC+j*dAngleAll; G4Sphere *s1 = new G4Sphere("PWOPart1", 140*cm, // inner radius 155*cm, // outer radius phiS, // start phi dAngleC, // delta phi thetaS, // start theta dAngleC // delta theta ); PD_log_crystal = new G4LogicalVolume(s1,PWO,crystal_name); sprintf(text,"/mydet/ring%02d",5*i+j+1); PWOSD* pdsd = new PWOSD(G4String(text),1); SDMan->AddNewDetector(pdsd); PD_log_crystal->SetSensitiveDetector(pdsd); G4VisAttributes *pd_att = new G4VisAttributes(G4Color(0,1,1)); // magenta PD_log_crystal->SetVisAttributes(pd_att); PD_physical = new G4PVPlacement(0,G4ThreeVector(0,0,0), "PD_physical", PD_log_crystal, experimentalHall_phys,false,0); } } The code with G4Box that worked is this: G4double dlC = 2.05*cm; G4double dlS = 0.05*cm; G4double dlAll = dlS + dlC; for (G4int i=0; i<5; i++) { for (G4int j=0; j<5; j++) { digit1 = G4int((5.*i+j+1)/10.); digit2 = (5*i+j+1) - digit1*10; crystal_name = "PD"+G4String(char('0'+digit1))+G4String(char('0'+digit2)); G4Box *s1 = new G4Box("PWOPart1",7.5*cm, 1.025*cm, 1.025*cm); PD_log_crystal = new G4LogicalVolume(s1,PWO,crystal_name); sprintf(text,"/mydet/ring%02d",5*i+j+1); PWOSD* pdsd = new PWOSD(G4String(text),1); SDMan->AddNewDetector(pdsd); PD_log_crystal->SetSensitiveDetector(pdsd); G4VisAttributes *pd_att = new G4VisAttributes(G4Color(0,1,1)); // magenta PD_log_crystal->SetVisAttributes(pd_att); PD_physical = new G4PVPlacement(0, G4ThreeVector(147.5*cm,(i-2)*dlAll,(j-2)*dlAll), "PD_physical", PD_log_crystal, experimentalHall_phys,false,0); } } I observed the same with version 4.2!
Bug was fixed in G4Sphere::Inside. It was responsible for the problem
