ESyS-Particle - How to intall / operate?

Asked by peterson magro

I'd like to know if is it too complicated to install (like i think), if it´s a interface to help us, and if all of the problem is resumed at "put the code in" the software?

I´m a windows' usuary, i didn´t know Linux but I liked a lot.

There is a lot of DEM (Discrete Element Method) for simulations of particles in the windows platafform, but they are payed and expensives, like http://www.dem-solutions.com/

Is the ESyS-Particle compatible with that solution? Do ESyS-Particle the same work?

sds
peterson magro

Question information

Language:
English Edit question
Status:
Solved
For:
ESyS-Particle Edit question
Assignee:
No assignee Edit question
Solved by:
Dion Weatherley
Solved:
Last query:
Last reply:
Revision history for this message
Dion Weatherley (d-weatherley) said :
#1

Hi Peterson Magro,

Thank you for your questions about ESyS-Particle.

At this stage, ESyS-Particle is designed for Linux operating systems only. There are no plans to port ESyS-Particle to Windows. ESyS-Particle has similar functionality to many of the the commercially available DEM software packages however it does not have a graphical user interface and may be lacking some of the features of particular software packages like DEM Solutions. The primary advantage of ESyS-Particle is that it is designed for multi-processor PCs and cluster supercomputers, allowing significantly larger models to be simulated than commercial software designed for desktop PCs. This is particularly important when conducting research on multi-scale phenomena like rock breakage.

It is difficult to say whether ESyS-Particle is suitable for your purposes without knowing more about the research you wish to conduct using DEM. If you would like to discuss this, please provide some more information about your proposed research. If you prefer, you can email me directly. My email address is <email address hidden>

Whilst ESyS-Particle may appear complicated to install for those unfamiliar with Linux, I have provided a number of installation methods to suite users with different levels of experience. The simplest is to use the ESyS-Particle Ubuntu LiveDVD. The LiveDVD allows you to use ESyS-Particle without having to install Linux on your PC hard drive. For more information about the LiveDVD, see https://launchpad.net/esys-particle/+announcement/1505

I have also provided step-by-step instructions on how to install ESyS-Particle on the latest version of Ubuntu Linux here:
https://answers.launchpad.net/esys-particle/+faq/261
This installation method is relatively simple and you are unlikely to encounter many problems if you choose this method.

I hope this answers your questions and thank you for your interest in ESyS-Particle.

Cheers,

Dion.

Revision history for this message
peterson magro (petersonmagro) said :
#2

Hi Dion,

Thanks for your answer.

I didn't said the purpose of my research at that time, but first of all I´ll show your the context.

I always thought about particles simulations utilizing computer solutions, but just one week ago I knew DEM calculation. After this I´m searching softwares, comercial or not, to forward with my dream.

I'm mechanic engineer, I work with machines for sugar and etanol plants, and I´m trying to simulate some events that I usually make an prototype, like a rotary dryer with lifters for sugar, cane bagace rotary feeder and others, like complex discharge chute for bagace conveyors, feeders for conveyors, that there is some literatures but that will be easyer if there were a DEM tool.

I've tried some alternative solutions utilizing the software Working Model 2D and 3D, for dynamic simulations but the process time is too big for not enough "particles" (in fact solid parts).

I´ve seen that DEM Solutions software do dynamic solutions.
http://www.dem-solutions.com/videos.php

I´d like know your oppinion about ESyS, if is it able to do the same?

I´ve seen ESyS's tutorial, and about the particle form I´ve seen that it is able to simulate spherical particles. Is it able to simulate other particles formats like prism and cylinder?

I´d like your opnion about the DEM Solutions software and, if is have you already done any thing like that.

best regards!

peterson.

Revision history for this message
Dion Weatherley (d-weatherley) said :
#3

Hi Peterson,

Thanks for your reply and explanation of your research interests. On first glance, it would appear that you are interested in studying the flow behaviour of granular media through an engineered system with a complex geometry. ESyS-Particle provides support for boundaries of complex geometry via the triangle-mesh walls so has some potential for application to your topic of interest. It would be remiss however not to mention that A) triangle mesh walls in ESyS-Particle are currently non-inertial so they do not react dynamically to forces imposed by particles and, B) the triangle mesh-wall implementation does not work for corners that subtend an angle less than 90 degrees.

I am not familiar with the capabilities of the DEM Solutions software but it would appear from their website that their software has been designed for the type of problems you wish to study. If you seek software designed for this type of problem, then perhaps ESyS-Particle is unsuitable. On the other hand, if you would prefer an Open Source solution, I would be interested to work with you to implement solutions to the two problems identified above.

Regarding your other question, currently ESyS-Particle is designed to use spherical particles or shapes constructed by clustering spheres together. We have just commenced a research project to develop new techniques for simulating non-spherical particles but I do not expect ESyS-Particle will support non-spherical particles in the near future.

Cheers,

Dion.

Revision history for this message
peterson magro (petersonmagro) said :
#4

Hi Dion,

Thanks for your answer. How I'm starting my research and i don't know anything about Linux (but i really like that) i don't know what can i do to make my simulatons.
In fact, i couldn't install ESyS completelly because of my LINUX inexperience, i followed ESyS tutorial but i had many errors.
Now i intend to change my mind and try a different way to start that. I know i need learn about Linux, Python and principally DEM theory.

Before finish this topic, i'll do one last question. I didn't understand very much if ESyS is able to simulate particles above a belt or inside a rotating drum, could you explain again, please?

Dion, i saw you have much experience, as in DEM simulation as to teaching others. I want learn more about DEM and i would like yours advices to start, to if possible, learn as faster as i can to (who knows?) work together some day. Today i don't know anything but i really want!

What do you think?

Cheers,
peterson

Revision history for this message
Best Dion Weatherley (d-weatherley) said :
#5

Hi Peterson,

I can certainly appreciate that ESyS-Particle is challenging to install and use without some prior experience with Linux and Python. To help you get started with Linux, I would recommend you contact a Linux User Group near you. It will make things a lot easier if you can find a more experienced Linux user willing to help you. The following website lists Linux User Groups by country:
http://www.linux.org/groups/

Regarding Python, I found the Python Tutorial on http://python.org/ to be very good. Like most programming languages, it can be a bit challenging to learn at first but I've found Python to be a very well designed programming language. More importantly, it is ideal for scientific programming tasks, particularly post-processing simulation data.

I haven't found a really good Beginner's Guide to DEM yet but I usually recommend the following book to my students:
"Computational Granular Dynamics: models and algorithms" by T. Poschel and T. Schwager, Springer, 2005.
http://www.springer.com/physics/mechanics/book/978-3-540-21485-4

Chapter 2 is particularly relevant but the other chapters are also very interesting. This book focuses more on the algorithms used in DEM software than the theory but is a reasonable starting point in my opinion.

ESyS-Particle has not been used for simulating particles inside a rotating drum as far as I know. I think it might be possible to do these kind of simulations with ESyS-Particle though. To model the rotating drum in ESyS-Particle simulations, we would need to construct a triangulated mesh of the drum geometry and write a subroutine to rotate that mesh at the desired speed. The particles inside the drum would interact with each other and the drum walls via elastic repulsive forces. Since ESyS-Particle has not been used to do these kind of simulations, I suspect it would take some time to perfect a model setup before it could be used for applied research.

I hope this answer helps you. I would be interested to discuss the possibility of research collaboration with you but it would be best to do so via private email communication rather than on this Launchpad forum.

Cheers,

Dion.

Revision history for this message
peterson magro (petersonmagro) said :
#6

Thanks Dion Weatherley, that solved my question.

Revision history for this message
Diego Peinado Martín (diego-peinado) said :
#7

Hello Peterson,

I'm thinking in using Esys for conducting simulations very similar than yours. I would like to see how aggregates goes through a rotary dryer, see mixing phenomena and other simulations concerning granular material.

I would like to interchange information with you as I progress on that issues.

Thanks in advance and best regards,

Revision history for this message
peterson magro (petersonmagro) said :
#8

Hello Diego,

This moment i'm not studying the ESyS because i haven't any time to spend in that issue, i've a lot of work in other areas.

If you are thinking about starting a study with particles, i recommend make contact with Dion Weatherley or Will, both of this lauch pad and ESyS specialists.
But before is better study LINUX (if you aren't a user), Phyton (to make the program) and the DEM theory like biography that Dion said.

I intend to study ESyS soon. If do you want to contact me for this or another issue, feel free to send me an email:
<email address hidden>

Regards!

Revision history for this message
Tory Moore (torymoore1) said (last edit ):
#9

The note at Step 1 applies to the the Python script published here (Launchpad.Internet) is in the Python layout and may be copied and pasted into an IDLE "New Window" then stored and run in Python. you can see script at http://treestandranger.com/.