From: Chris Theis (Christian.Theis@cern.ch)
Date: Tue Dec 19 2006 - 11:36:30 CET
Hello Martin,
I've quickly checked your input and you've spotted the problematic
regions.
The expansion and optimized rewriting of the regions V19 & V33 results
in a huge amount of concatenated subregions.
However, you might be able to simplify those regions as there seems to
be some redundant information there. For example in V33 you subtract a
union, but parts of this union are not enclosed by the bounding box of
the resulting region and thus, they don't contribute. For example only
the part ( -BP4___IN +BP4__OUT -GENP__03 +GENP__04) of the
subtracted union in V33 should matter. But please verify this as I did a
very quick check only due to a lack of time.
Similar "manual optimizations" are likely to be applicable to V19.
The geometry kernels implemented in FLUKA and the one which I wrote for
SimpleGeo are actually totally different and based on different
architectures for CSG trees. Thus, I'm not surprised that there could be
a difference in the handling as SimpleGeo does not require any
parentheses expansion. However, I'm pretty sure that there are also some
tricky and non-obvious issues associated to SimpleGeo's approach, which
are for the moment still lurking in the dark.
Cheers
Chris
------------------------------------------------------------------------
-- Chris Theis CERN/SC-RP - European Organization for Nuclear Research 1211 Geneva 23, Switzerland Phone: +41 22 767 8069 Office: 892-2A-015 e-mail: Christian.Theis@cern.ch www: http://www.cern.ch/theis ------------------------------------------------------------------------ -- > -----Original Message----- > From: Stefan Roesler > Sent: Tuesday, December 19, 2006 9:28 AM > To: Alfredo Ferrari > Cc: Vasilis Vlachoudis; Francesco Cerutti; Sebastien WURTH > (E-mail); Chris Theis; m.p.holbourn@dl.ac.uk; fluka-discuss@fluka.org > Subject: RE: Struggling to find error > > Hi > > Attached is *out and *log. Yes, there is a subscript out of > range in log. > > Ciao, stefan > > On Tue, 19 Dec 2006, Alfredo Ferrari wrote: > > > Please keep all discussions on the fluka-discuss list, not private. > > > > Martin, did you check what is in the log file? Very likely > you got a > > message there detailing what went wrong (if Stefan is correct, you > > should have a message like "subscript out of range.... "). > > > > I never got the input file so I cannot help more. > > > > Alfredo > > > > On Tue, 19 Dec 2006, Stefan Roesler wrote: > > > > > Hello > > > > > > I pass this on to Alfredo/Vasilis. RPNORM is called with > > > LASTEL=114705, which obviously exceeds MXEXEL.... > > > > > > Thus, it could also be a problem in the code. > > > > > > Cheers > > > Stefan > > > > > > > > > On Tue, 19 Dec 2006, Holbourn, MP (Martin) wrote: > > > > > > > Stefan, Francesco, Sebastien > > > > Thanks very much for replying > > > > attached input file. > > > > I am really confused aas the geometry seeems to build ok in > > > > simplegeo. I am suinf free format with parentheses and > I suspect > > > > it is my use of parentheses that is causing the > problems. I have a > > > > 4 sections of beampipe made up of intersecting cyliders > cut by planes at the bends in the pipe. > > > > Region CU2 shows this. I also have a mirror and a collimator in > > > > the same vacuum region, V33. I specified V33 as : Total > Volume - > > > > (Baeampipe) - > > > > (Mirror) - (Collimator). I am guessing it is the expansion of > > > > these parentheses that is causing the problem because > if I reduce > > > > the pipe to just the first 2 sections Fluka runs OK. > Issues start > > > > appearing when I add the third section. Fluka runs ok > but seems to > > > > take a lot longer.( Itested this by only running 20 > particles and > > > > the more complicated the parentheses the longer it > takes). Loading > > > > the geometry file in FLUKAGUI seems to take several > minutes also. > > > > I have also attached the build from simplegeo (rrof removed) so > > > > you have an idea of what my geometry looks like> Any commments? > > > > > > > > Thanks very much for your interest. > > > > > > > > Regards > > > > Martin > > > > > > > > PS > > > > Chris hope you don't mind but I have included you on > this e-mail > > > > as you have very kindly helped me in the past and it > does appear > > > > at this stage to be a geometry issue. > > > > Sorry for troubling you > > > > Regards > > > > MArtin > > > > > > > > -----Original Message----- > > > > From: Stefan Roesler [mailto:sroesler@mail.cern.ch] > > > > Sent: 18 December 2006 22:08 > > > > To: Holbourn, MP (Martin) > > > > Subject: Re: Struggling to find error > > > > > > > > > > > > Hi Martin > > > > > > > > Please send the input. Most likely a format error, hard to tell > > > > without input file... > > > > > > > > Regards > > > > Stefan > > > > > > > > On Mon, 18 Dec 2006, Holbourn, MP (Martin) wrote: > > > > > > > > > Dear all, > > > > > Fluka aborts with the message > > > > > ..../myFlukaArea/flutil/rfluka: line 311: 20178 Aborted > > > > > > > > > > moving inside temporary directory there is no .err file. The > > > > > *.out file echos my input cards up to the end of the body > > > > > definitions with the last few lines > > > > > > > > > > Number of bodies 62 > > > > > Length of FPD-Array 641 > > > > > > > > > > There is nothing after this. > > > > > > > > > > The geometry builds fine in SimpleGeo and using the > debugger in > > > > > SimpleGeo shows no problems. > > > > > > > > > > Using gdb on the core.nnnn file provides no clue as > gdb reports > > > > > core.nnnnn not in executable format File format not > recognised. > > > > > > > > > > Anybody any ideas on how to find the error? > > > > > > > > > > > > > > > Martin Holbourn > > > > > Radiation Protection Adviser > > > > > > > > > > Daresbury Laboratory > > > > > Daresbury > > > > > Warrington > > > > > Cheshire > > > > > WA4 4AD > > > > > > > > > > Tel: 01925 603266 > > > > > Fax: 01925 603381 > > > > > mailto:m.p.holbourn@dl.ac.uk > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > -- > ___________________________________ > > Stefan Roesler > CERN, SC/RP > CH-1211 Geneva 23 > Switzerland > > Phone: +41-22-7679891 > Fax: +41-22-7669639 > E-mail: Stefan.Roesler@cern.ch >
This archive was generated by hypermail 2.1.6 : Tue Dec 19 2006 - 16:25:20 CET