From: John Clem (clem@bartol.udel.edu)
Date: Mon Feb 11 2008 - 06:00:30 CET
Giuseppe,
A preliminary analysis of a few test runs over the weekend look very
promising and clearly the problem has been fixed. Currently we're building
statisitics to better understand the fluctuations with scaled-back
thresholds (both EMF-CUT blank and PROD-CUT are now being used).. Paola
suggestions changing the MGFIELD inputs and removing STEPSIZE were also
implemented. The combination of changes has signifcantly reduce run-time
(i.e. cpu time/event ratio)... Thanks again for the help...
With Best Regards,
John
> John Clem
> Research Associate Professor
> Bartol Research Institute
> Department of Physics and Astronomy
> University of Delaware
> Newark, DE 19716 USA
> 302-831-4354
----- Original Message -----
From: "Giuseppe Battistoni" <Giuseppe.Battistoni@mi.infn.it>
To: "John Clem" <clem@bartol.udel.edu>
Cc: <fluka-discuss@fluka.org>
Sent: Thursday, January 31, 2008 4:54 AM
Subject: Re: problem in running large gamma showers
> Hi John
> we hope that using the last FLUKA respin your problem is definitively
> fixed, and we thank you for having pointed out that hidden bug.
> In addition to this I wish to recommend 2 other issues for your
> case:
> a) in addition to yout EMF-CUT cards with blank SDUM to control transport
> cutoff of e+,e-,gamma, you should also introduce EMF-CUT cards with
> SDUM=PROD-CUT, in order to take full control of production thresholds.
>
> b) it has been tested (Paola Sala) that if you take away the STEPSIZE
> cards and set the first 3 WHAT of the magnetic field card
> MGFIELD to values that we have used before for atmospheric shower
> simulations ( MGFIELD 20. 100. 30. ..... ), then the messages in the logs
> concerning problems in precision will disappear. This should speed up a
> little your problem.
>
> Best regards
> Giuseppe
>
> On Fri, 30 Nov 2007, John Clem wrote:
>
>> All,
>>
>> While attempting to characterize VHE gamma ray showers in the atmosphere,
>> we
>> ran into the below problem that occured for every other primary. For this
>> case the primary energies are fixed at 20TeV.
>>
>> "Subscript out of range on file line 357, procedure
>> kaskad.f/kaskad.Attempt
>> to access the 5501-th element of variable agesec.equests:"
>>
>> It became less frequent after maximizing our energy thresholds, however
>> any
>> thing above these levels would shroud our analysis goals. Incrementing
>> the
>> dimensions would seem like the next obvious step. Lack of access to this
>> part of the code prevents such a test. Any suggestions ??
>>
>> Thanks, John
>>
>> > John Clem
>> > Research Associate Professor
>> > Bartol Research Institute
>> > Department of Physics and Astronomy
>> > University of Delaware
>> > Newark, DE 19716 USA
>> > 302-831-4354
>>
>
>
>
>
This archive was generated by hypermail 2.1.6 : Mon Feb 11 2008 - 09:18:47 CET