[Kratos] Nelson. Serialization

Janosch Stascheit janosch.stascheit en rub.de
Mie Feb 16 13:14:21 CET 2011


Hi,

yes, this answers my questions. It should indeed cause no problems.

Best regards
Janosch

Am Mittwoch, den 16.02.2011, 12:46 +0100 schrieb Riccardo Rossi:
> Imagine that you have a composite constitutive law, made by an elastic
> and a plastic constitutive law that cohesist in the same volume and are
> in parallel (or in series...)
> 
> the composite constitutive law will have a pointer to the two
> constitutive laws,
> 
> so upon save it should serialize the INTERNAL_VARIABLES for the two
> constitutive laws it contains + any data it is needed for itself. In
> this case it will serialize more than one vector, but this will still
> provide no problem at all with the method i propose
> 
> imagine that composite_law has the pointers p1 and p2 to two other
> constitutive laws.
> 
> composite.save
>    Vector a1 = p1->GetValue(INTERNAL_VARIABLES)
>    Vector a2 = p2->GetValue(INTERNAL_VARIABLES)
>    Vector a3 = this->GetValue(INTERNAL_VARIABLES)
>    serialize(a1)
>    serialize(a2)
>    serialize(a3)
> 
> composite.load
>     load(a1)
>     load(a2)
>     load(a3)
>     then set all needed
> 
> does this explain what i mean?
> 
> ciao
> Riccardo
> 
>     
> -- 
> ________________________________________________________________
> Riccardo Rossi, Ph.D, Civil Engineer
> member of the Kratos Group: kratos.cimne.upc.es
> Centro Internacional de Métodos Numéricos en Ingeniería (CIMNE)
> Universidad Politécnica de Cataluña (UPC)
> Edificio C-1, campus Norte UPC
> Gran Capitan, s/n
> 08034 Barcelona, España
> Tel. (+34) 93 401 73 99
> ________________________________________________________________
> 
> AVISO IMPORTANTE
> 
> Los datos de carácter personal contenidos en el mensaje, se registrarán
> en un fichero para facilitar la gestión de las comunicaciones de CIMNE.
> Se pueden ejercitar los derechos de acceso, rectificación, cancelación y
> oposición por escrito, dirigiéndose a nuestras oficinas de CIMNE, Gran
> Capitán s/n,  Edificio C1 - Campus Norte UPC, 08034 Barcelona, España.
> 
> AVÍS IMPORTANT
> 
> Les dades de caràcter personal contingudes en aquest missatge es
> registraran en un fitxer per facilitar la gestió de les comunicacions
> del CIMNE. Es poden exercir els drets d'accés, rectificació,
> cancel·lació i oposició, per escrit a les nostres oficines del CIMNE,
> Gran Capità s/n, Edifici C1, Campus Nord UPC, 08034 Barcelona, Espanya.
> 
> IMPORTANT NOTICE
> 
> All personal data contained in this mail will be processed
> confidentially and stored in a file property of CIMNE in order to manage
> corporate communications. You may exercise the right of access,
> rectification, deletion and objection by letter sent to CIMNE, Gran
> Capitán, Edificio C1 - Campus Norte UPC, 08034 Barcelona, Spain.
> 
> 
> 
> 
> 
> On Wed, 2011-02-16 at 12:36 +0100, Janosch Stascheit wrote:
> > Hi,
> > 
> > now I am a bit confused. With composite constitutive laws you mean that
> > it is possible to combine different return mapping schemes and yield
> > functions?
> > 
> > However, if these selections are stored as enums, it should be possible
> > to store their state as well, right?
> > 
> > Bests
> > Janosch
> > 
> > Am Mittwoch, den 16.02.2011, 11:38 +0100 schrieb Riccardo Rossi:
> > > Hi,
> > >    my idea is that we should reuse what was already done with
> > > INTERNAL_VARIABLES and take the occasion to extend it more.
> > > 
> > > the important thing is that the serialization should be able to handle
> > > "composite constitutive laws" the way nelson does. This will require
> > > some more handling...but it should still leverage the INTERNAL_VARIABLES
> > > stuff
> > > 
> > > Riccardo
> > > 
> > > -- 
> > > ________________________________________________________________
> > > Riccardo Rossi, Ph.D, Civil Engineer
> > > member of the Kratos Group: kratos.cimne.upc.es
> > > Centro Internacional de Métodos Numéricos en Ingeniería (CIMNE)
> > > Universidad Politécnica de Cataluña (UPC)
> > > Edificio C-1, campus Norte UPC
> > > Gran Capitan, s/n
> > > 08034 Barcelona, España
> > > Tel. (+34) 93 401 73 99
> > > ________________________________________________________________
> > > 
> > > AVISO IMPORTANTE
> > > 
> > > Los datos de carácter personal contenidos en el mensaje, se registrarán
> > > en un fichero para facilitar la gestión de las comunicaciones de CIMNE.
> > > Se pueden ejercitar los derechos de acceso, rectificación, cancelación y
> > > oposición por escrito, dirigiéndose a nuestras oficinas de CIMNE, Gran
> > > Capitán s/n,  Edificio C1 - Campus Norte UPC, 08034 Barcelona, España.
> > > 
> > > AVÍS IMPORTANT
> > > 
> > > Les dades de caràcter personal contingudes en aquest missatge es
> > > registraran en un fitxer per facilitar la gestió de les comunicacions
> > > del CIMNE. Es poden exercir els drets d'accés, rectificació,
> > > cancel·lació i oposició, per escrit a les nostres oficines del CIMNE,
> > > Gran Capità s/n, Edifici C1, Campus Nord UPC, 08034 Barcelona, Espanya.
> > > 
> > > IMPORTANT NOTICE
> > > 
> > > All personal data contained in this mail will be processed
> > > confidentially and stored in a file property of CIMNE in order to manage
> > > corporate communications. You may exercise the right of access,
> > > rectification, deletion and objection by letter sent to CIMNE, Gran
> > > Capitán, Edificio C1 - Campus Norte UPC, 08034 Barcelona, Spain.
> > > 
> > > 
> > > 
> > > 
> > > 
> > > On Wed, 2011-02-16 at 10:18 +0100, Nelson Lafontaine wrote:
> > > > Hi Janosch Riccardo
> > > > Janosch, i dont understand  when you say: it should be enough to just
> > > > store the Vector to be
> > > > restored on an unchanged geometrical configuration. What about the
> > > > changed configuration??
> > > > 
> > > > The Idea of save the internal variables is just to know the state of
> > > > the constituive law. Serialization 
> > > > of this vector is enough. no???
> > > > 
> > > > Best regards
> > > > 
> > > > Nelson
> > > >     
> > > > 
> > > > 2011/2/16 Janosch Stascheit <janosch.stascheit en rub.de>
> > > >         Hi Riccardo, Nelson,
> > > >         
> > > >         I think that this is a proper way of dealing with the issue.
> > > >         In any
> > > >         case, the use of one single Vector of internal variables is
> > > >         not a nice
> > > >         but a much more performing way than implementing an own
> > > >         variable for
> > > >         each distinct internal variable.
> > > >         
> > > >         However, the interpolation does not necessarily need to be
> > > >         combined with
> > > >         the serialization, although this would facilitate the transfer
> > > >         of
> > > >         internal variables between dissimilar meshes. For the
> > > >         serialization as
> > > >         such, though, it should be enough to just store the Vector to
> > > >         be
> > > >         restored on an unchanged geometrical configuration.
> > > >         
> > > >         What is your opinion about that?
> > > >         
> > > >         Bests
> > > >         Janosch
> > > >         
> > > >         Am Mittwoch, den 16.02.2011, 09:48 +0100 schrieb Riccardo
> > > >         Rossi:
> > > >         
> > > >         > Hi guys,
> > > >         >     i sneak into the discussion as I believe that the issue
> > > >         of the
> > > >         > constitutive laws is rather important.
> > > >         >
> > > >         >    I believe that a clean way of dealing with the
> > > >         serialization of the
> > > >         > constitutive_laws is implementing it so, that we can later
> > > >         leverage the
> > > >         > work for interpolation.
> > > >         >
> > > >         > that is, up to now we decided that interpolation should be
> > > >         done by
> > > >         > writing all what is needed to a vector of INTERNAL_VARIABLES
> > > >         (or
> > > >         > something of the sort). A clean way to do this would be that
> > > >         upon save
> > > >         > we call the same function that writes the internal variables
> > > >         to a vector
> > > >         > and then we serialize that vector (which should be
> > > >         automatic)
> > > >         >
> > > >         > Nelson, Janosch,
> > > >         >
> > > >         > what do you think about this?
> > > >         >
> > > >         > Riccardo
> > > >         >
> > > >         >
> > > >         
> > > >         
> > > >         
> > > > 
> > > > 
> > > > 
> > > > -- 
> > > > Ing. Nelson M. Lafontaine
> > > > 
> > > > 
> > > 
> > 
> > 
> 




Más información sobre la lista de distribución Kratos