[Gate-users] GeomTest Errors in SPECT Benchmark - GATEv.6.2

[Giuseppe Magro]

Hello GATEers.

In GATEv.6.2 I've added the command for geometry recursive test after the
initialization of the SPECT benchmark example (benchSPECT.mac), i.e.
...
/gate/run/initialize
/geometry/test/recursive_test
...
and surprisingly found lots of GeomTest Errors. By visual checking, in
fact, holes in collimator overlap to each other, for example. Mantaining
the same geometry but redefining holes with the following lines

...
/gate/collimator/daughters/name hole
/gate/collimator/daughters/insert hexagone
/gate/hole/geometry/setHeight 3. cm
/gate/hole/geometry/setRadius .15 cm
/gate/hole/placement/setRotationAxis 0 1 0
/gate/hole/placement/setRotationAngle 90 deg
/gate/hole/setMaterial Air
/gate/hole/describe

/gate/hole/repeaters/insert cubicArray
/gate/hole/cubicArray/setRepeatNumberX 1
/gate/hole/cubicArray/setRepeatNumberY 9
/gate/hole/cubicArray/setRepeatNumberZ 9
/gate/hole/cubicArray/setRepeatVector 0. 1. 1. cm
/gate/hole/cubicArray/autoCenter false

/gate/hole/repeaters/insert linear
/gate/hole/linear/setRepeatNumber 2
/gate/hole/linear/setRepeatVector 0. 0.5 0.5 cm
/gate/hole/linear/autoCenter false
...

GeomTest Errors disappear as you expect by visual checking. Obviously the
lines

/gate/hole/cubicArray/setRepeatNumberX 1
/gate/hole/cubicArray/setRepeatNumberY 9
/gate/hole/cubicArray/setRepeatNumberZ 9

don't allow a complete coverage of the collimator, but if you type a number
greater than 9, error messages appear again although no mistakes have been
done by visual check. Another strange thing is that number 9 (for example)
does not work any more if you switch from autoCenter false to autoCenter
true option. It's a non-sense, since all volumes do not overlap and are
well defined inside their mothers.

Is this a bug or am I making any mistake?

Thanks in advance.

Giuseppe.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.opengatecollaboration.org/mailman/private/gate-users/attachments/20130307/f018da10/attachment.html>