[CIG-SEISMO] PS

[Dimitri Komatitsch]

PS: that new version of SPECFEM3D_GLOBE has several other bug fixes and 
improvements (in particular in the case of viscoelasticity); therefore 
we encourage users to upgrade to version 5.1.5.

Best regards,
Dimitri.

On 08/29/2012 01:59 PM, Dimitri Komatitsch wrote:
>
> Dear all,
>
> This summer several users of SPECFEM3D_GLOBE had informed us that the
> latest version of the code was unstable. Daniel Peter has just found and
> fixed the problem, therefore please get the new version and you should
> be all set.
>
> Please inform us if you still get stability issues.
>
> Thank you,
> Best regards,
>
> Dimitri.
>
> On 08/18/2012 03:27 PM, Dimitri Komatitsch wrote:
>>
>> Dear Arjun,
>>
>> Thank you for your bug report. This bug has also been noticed by other
>> users before the summer. Two of the GLOBE developers (Daniel Peter and
>> Joseph Charles; I cc them) are currently investigating it and fixing it.
>> Hopefully we should be able to release a new version in the middle of
>> September. We will then send an email to the CIG mailing list.
>>
>> Sorry about the problem. It seems someone committed a modification that
>> made the package unstable.
>>
>> (you can thus temporarily revert to an older version (which you can get
>> from the SVN server using "svn checkout -r YYYY", where YYYY is an old
>> release number (which you can get using "svn log" and selecting a
>> version from a few months ago, or maybe even 2011 or 2010 to be safe).
>>
>> Thank you,
>> Dimitri.
>>
>> On 07/31/2012 01:49 PM, A. Datta wrote:
>>> Hi,
>>>
>>> I have been using SPECFEM3D GLOBE for some forward global simulations
>>> but
>>> my simulations often terminate with the following error message:
>>>
>>> "forward simulation became unstable and blew up in the fluid"
>>>
>>>> From a glance at the code I know this corresponds to the maximum
>>>> potential
>>> in fluid in all slices becoming greater than a threshold value (or
>>> negative).
>>>
>>> Can someone please explain what this means, i.e. when this happens ?
>>> Could
>>> it have anything to do with the meshing ? I use NEX_XI =192 and
>>> NPROC_XI =
>>> 3, so as to engage 54 processors.
>>>
>>> Thanks,
>>> Arjun
>>>
>>>
>>
>

-- 
Dimitri Komatitsch - komatitsch aT lma.cnrs-mrs.fr
CNRS Research Director (DR CNRS), Laboratory of Mechanics and Acoustics,
UPR 7051, Marseille, France    http://komatitsch.free.fr