| Summary: | "Expected normal-global-frame to be valid, i.e. a unit vector!" warning in G4Navigator for G4Trap solid | ||
|---|---|---|---|
| Product: | Geant4 | Reporter: | emanuele.leonardi |
| Component: | geometry/solids | Assignee: | Evgueni.Tcherniaev |
| Status: | RESOLVED WORKSFORME | ||
| Severity: | minor | CC: | Koichi.Murakami |
| Priority: | P4 | ||
| Version: | 10.3 | ||
| Hardware: | All | ||
| OS: | All | ||
Thank you for reporting the issue. Unfortunately we are unable reproduce the problem, the methods that compute the normal, G4Trap::SurfaceNormal() and G4Trap::DistanceToOut(), work normally in our tests. Anyway, next week it will be released new Geant4 10.4. It will contain revised G4Trap, which is much more performant than G4Trap of Geant4 10.03. Please note that new G4Trap will be more strict relative to planarity of lateral faces of a G4Trap object. The G4Trap solid from your report will not pass the planarity test at the construction time. The discrepancy with planarity is 0.0003 mm, that is much more than the surface tolerance kCarTolerance = 10e-9 mm and potentially can be the source of the issue. I suggest you to try the following parameters for that G4Trap solid: (-55.133,-51.75,-185) (-54.253, 58.25,-185) ( 54.253, 58.25,-185) (-48.423,-51.75, 185) ( 48.423,-51.75, 185) (-47.647, 45.25, 185) ( 47.647, 45.25, 185) Possibly the problem will go away. |
Inside tha NA62 simulation I am getting a warning from the G4Navigator while tracking particles crossing a G4Trap solid. The warning is: -------- WWWW ------- G4Exception-START -------- WWWW ------- *** G4Exception : GeomNav0003 issued by : G4Navigator::GetGlobalExitNormal() WARNING> Expected normal-global-frame to be valid, i.e. a unit vector! - but |normal| = 0 - and |normal|^2 = 0 which differs from 1.0 by -1 n = (0,0,0) Global point: (13.68420223,-982.8710887,203142.968) Volume: LAV_PbGlBlock Solid: pbglblock, Type: G4Trap ----------------------------------------------------------- *** Dump for solid - UTrap *** =================================================== Solid type: Trapezoid Parameters: half lengths X1-X4: 55.133mm, 54.253mm, 48.423mm, 47.647mm half lengths Y1,Y2: 54.976mm, 48.41mm half length Z: 185mm Solid axis angles: Theta=1.016661093302284deg, Phi=-90deg Face axis angles: TanAlpha1=0deg, TanAlpha2=0deg ----------------------------------------------------------- ============================================================ State of Navigator: The current state of G4Navigator is: ValidExitNormal= 1 ExitNormal = (0,0,0) Exiting = 1 Entering = 0 BlockedPhysicalVolume= None BlockedReplicaNo = -1 LastStepWasZero = 0 Current Localpoint = (-14.53172,-76.107898,184.70006) PreviousSftOrigin = (13.910411,-983.15151,203142.47) PreviousSafety = 0 Current History: History depth=5 Level=[0]: Phys Name=[World] Type=[N] Level=[1]: Phys Name=[LAV_RR2] Type=[N] Level=[2]: Phys Name=[LAV_Station_10] Type=[N] Level=[3]: Phys Name=[LAV_Banana] Type=[N] Level=[4]: Phys Name=[LAV_Block] Type=[N] Level=[5]: Phys Name=[LAV_PbGlBlock] Type=[N] ============================================================ Value obtained from stored global-normal is not a unit vector. *** This is just a warning message. *** -------- WWWW -------- G4Exception-END --------- WWWW ------- The G4Trap in question is created with the following call: new G4Trap(name,vtx); where vtx contains the 8 vertexes of the trapezoid which are: (-55.133,-51.693,-185) ( 55.133,-51.693,-185) (-54.253, 58.259,-185) ( 54.253, 58.259,-185) (-48.423,-51.693, 185) ( 48.423,-51.693, 185) (-47.647, 45.127, 185) ( 47.647, 45.127, 185) I verified the trapezoid geometry and it looks correct. On the other hand, given the vertexes used, the "Current Localpoint = (-14.53172,-76.107898,184.70006)" reported by the Navigator seems to be well outside the trapezoid. I searched the web for this warning and found several occurences which were due to some problem in the solid implementation. See, e.g., https://bugzilla-geant4.kek.jp/show_bug.cgi?id=1736 Could this be some similar problem? P.S.I hope this Description is readable: the Preview tab does not work for me...