yade-users team mailing list archive

[Andreas Plesch <andreasplesch@xxxxxxxxx>]

Thanks, your explanations are extremely helpful. You
may want to reformat slightly and put this on the web
page, probably after "first steps" or so.

Unfortunately, I am not an experienced c++ coder and
the extensive usage of external libraries such as
boost or the math libraries does not make it easier to
understand the code (what is se3?). But I can find my
way around and may be able to slightly modify
exisiting code to make it do what I want.

I am very willing to give feedback and be generally
helpful with things I can do.

I think it is would be great to have a FileGenerator
which can read from a simpler format which encodes all
of the interesting existing objects. But the xml is
nice, too.

One may even think about a little script language
which can control a simulation dynamically, e.g. while
running.

I will look at the commandline interface to see what
states it stores and if it is possible to reload.

If is easy to do, I think a recording button would be
a great enhancement to the qtgui simulation.

--- Janek Kozicki <janek_listy@xxxxx> wrote:

> Andreas Plesch said:     (by the date of Fri, 21 Oct
> 2005 16:47:47 -0700 (PDT))
> 
> > Hi Janek,
> > 
> > yes, of course, my question was hard to address. I
> > should mention that we are using PFC2d right now
> for
> > geodynamic problems, and are looking into 3d. So I
> 
> > am familiar with DEM, and concepts such as shear
> and
> > normal contact stiffness, walls and friction, or
> > particle bonding and bond breaking.
> 
> in yade you will find them in
> persistentInteractions,
> volatileInteractions.

volatile only exists until there is no more contact ?
And persistant can exist without contact ?

> Also we currently have
> 'modified mohr coulomb'
> criterion but it is curretnly implemented in dirty
> manner and soon
> should be separated from ElasticContactLaw to
> CoulombCriterion. This
> will allow to add and mix with other criterions.
> 
> > The most similar
> > model for us to start expanding on would
> > be the "spheres settling on a plane" model. 
> 
> so - I suggest you to modify SDECSpheresPlane.cpp,
> by simply adding
> another vertical plane, and attach it to a
> TranslationEngine (so it will
> apply a kinetic movement to that plane). And observe
> the results. If you
> want to record a force or anything during the
> simulation - have a look
> at how classes ForceRecorder, VelocityRecorder,
> AveragePositionRecorder
> are used in some filegenerators (eg.
> SDECImpactTest).

ok, I will look at this.  

> 
> > For
> > example we would like to explore sand which very
> > slowly piles up in front of a moving vertical
> wall.
> > In particular we would interest in the formation
> > of cracks and faults in the sand structure.
> 
> without bonds and bond breaking this simulation is
> currently possible,
> simply by modifying SDECSpheresPlane.
> 
> > 
> > As a practical matter, one element which would be
> > essential would be the option to save the state of
> > a simulation, 
> 
> you should have a look at NullGUI - it's a command
> line interface to
> yade. You can start a simulation without opening a
> graphical window, and
> without losing CPU time for display. You can also
> configure it to save
> simulation when it is running.
> 
> In fact QtGUI is unable to save simulation when it's
> running - it's easy
> to add, but there were lot of more important things
> to do ;) And also
> because we were using NullGUI when we needed
> simulation snapshots. Run:
> 
> ./yade -n -H
> 
> to see help for NullGUI. btw, I just noticed that
> ./yade -n doesn't
> print help, which is bad, because no one has a way
> to discover -H to see
> the help. I should fix it quick.
> 
> > modify or add/substract elements and
> > then restart. 
> 
> currently the only way to do that is by modifying
> .xml file by hand.
> When you change the number of elements in the
> container, remember to
> change the container size (on the top). In the
> future I hope there will
> be a way to do such modifications using graphical
> interface, by clicking
> on the elements.
> 
> Also when you save and later load the simulation you
> should check if the
> results (eg. force on the bottom plane, or whatever)
> are simular if
> simulation runs uninterrupted. That is because I
> don't remember now if
> 'save' actually saves all the data required later to
> restore the
> simulation. It's possible that shearing force may be
> accidentally reset
> to zero. Maybe that problem was fixed - I don't
> remember now, so better
> you will check that anyway.
> 
> > I suppose this would be
> > serialization/deserialization in the yade language
> ?
> 
> yes, exactly. At minimum we will add binary
> frontend, then probably yaml.
> 
> > 
> > Does the DEM allow for bonding and bond breaking
> > between elements currently ?
> 
> I replied in other post to that matter ;)
> 
>
https://lists.berlios.de/pipermail/yade-users/2005q4/000042.html
> 
> 
> > In the end, I would like to use other tools in
> > addition to the FileGenerators to generate input
> files
> > to the actual simulator and also be able to look
> at
> > the state of the model with other tools at any
> stage.
> 
> you can add another class similar to ForceRecorder
> that will record to
> file any value you need, when the simulation is
> processing. In such a
> way you can easily prepare any input needed by other
> programs.
> 
> You can also write another file generator, that
> instead of asking the
> user to input data - will load a given file, and
> from data inside it -
> create .xml for simulation. In fact this is how FEM
> works - there is a
> file that contains FEM mesh inside, and this mesh is
> loaded
> (in yade-examples)
> 
> -- 
> Janek Kozicki                                       
>                  |
> _______________________________________________
> Yade-users mailing list
> Yade-users@xxxxxxxxxxxxxxxx
> http://lists.berlios.de/mailman/listinfo/yade-users
> 



		
__________________________________ 
Yahoo! FareChase: Search multiple travel sites in one click.
http://farechase.yahoo.com
_______________________________________________
Yade-users mailing list
Yade-users@xxxxxxxxxxxxxxxx
http://lists.berlios.de/mailman/listinfo/yade-users