yade-users team mailing list archive

[Janek Kozicki <janek_listy@xxxxx>]

Omer Ardic said:     (by the date of Tue, 18 Jul 2006 21:39:12 +0300)

> Before asking questions, I want to introduce myself. After 10 years of
> professional career as civil engineer, I decided to turn back to my academic
> career. Now I am a master student writing thesis in geotechnics.

Hello :)
 
> Instead of preparing an ordinary thesis, I wanted to select a challenging
> subject. So I decided to study on the use of DEM in geomechanical problems.

You have to decide what kind of geomechanical experiment you want to
perform numerically. It is not possible to do this study without an
example simulation.

You have numerous choices, like triaxial test, biaxial test, wall
foundation settling down in the soil, many, many others. Better pick one
and do it thoroughly, than trying to make all of possible variants.

> In my master thesis, I will  just introduce DEM, make some literature search
> and prepare a simple problem to show the effectiveness of DEM in
> geomechanics.

What is the measure of "effectiveness"?
- is is by comparing with with real world experiments?
- is it by comparing with other simulation methods, like with
  hypoplastic FEM soil simulation, or any other simulation method?
- is it just the speed of computation?
- or maybe something else?

> I am planning to prepare my own program on this subject in PhD thesis but at
> this moment I have to use a package program. I searched the web and I
> realized that there are no much DEM program. I have found PFC2D/PFC3D and
> EDEM as commercial and YADE as a open-source program.  So I am here to
> investigate it.

YADE uses SDEC code to perform DEM simulations. SDEC has been thoroughly
verified thorough 10 years of continuous scientific work. But, it's
*very* important - YADE has not yet been verified with comparison to
SDEC with respect to the results. Until this comparison is done I'm not
100% sure that YADE gives exactly similar results as SDEC.

I hope that Bruno will make this comparison between SDEC and YADE with
triaxial test.

With this in mind you are welcome to try out YADE :)

> I have 6 months to complete my thesis. So do you think that I can use this
> program in my thesis?

For master thesis YADE should be sufficient, even if you have just 6
months. Several people are doing PhD with with YADE. but it takes years,
not 6 months :)

When you decide what kind of test you want to make, you will only have
to write your own FileGenerator. You can start by looking at the
existing ones.

The way how FileGenerator works and is used in YADE will be totally
changed, and will be much easier to use. But I don't know how much time
it will take me to code this. So if you have just 6 months, you cannot
wait - and you will have to use the current version, just as other
people are using it now.

> Also what do you recommend me to read and investigate
> as reference to my study? I have to understand the theories and philosophy
> behind YADE.

Ok, so the documentation of class design, and how to extend yade with
new simulations is still not finished. But even without it some people
were able to write their own FileGenerators - with the help from
yade-users mailing list.

So instead of documentation, you should read following posts, where
people ask (and got answers) - about how to write a FileGenerator:

http://lists.berlios.de/pipermail/yade-users/2005q4/000049.html
http://lists.berlios.de/pipermail/yade-users/2005q4/000052.html
http://lists.berlios.de/pipermail/yade-users/2005q4/000073.html

http://lists.berlios.de/pipermail/yade-users/2005q4/000097.html
http://lists.berlios.de/pipermail/yade-users/2005q4/000098.html
http://lists.berlios.de/pipermail/yade-users/2005q4/000100.html
http://lists.berlios.de/pipermail/yade-users/2005q4/000101.html
http://lists.berlios.de/pipermail/yade-users/2005q4/000105.html
http://lists.berlios.de/pipermail/yade-users/2005q4/000107.html

http://lists.berlios.de/pipermail/yade-users/2006q1/000170.html
http://lists.berlios.de/pipermail/yade-users/2006q2/000176.html


less important threads (including all answers to them...):

http://lists.berlios.de/pipermail/yade-users/2005q4/000076.html
http://lists.berlios.de/pipermail/yade-users/2005q4/000102.html
http://lists.berlios.de/pipermail/yade-users/2006q1/000147.html
http://lists.berlios.de/pipermail/yade-users/2006q2/000171.html


Huh, I didn't thought there are so many of them :)

When you start writing a FileGenerator and got any problems, just ask,
and here on yade-users you will get help :)

Also read the current documentation from the website, about how to debug
yade, because I'm sure that you will need this :)

-- 
Janek Kozicki                                                         |
_______________________________________________
Yade-users mailing list
Yade-users@xxxxxxxxxxxxxxxx
http://bat.berlios.de/mailman/listinfo/yade-users