[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [ESPResSo-users] Simulating partially immersed deformable objects

From: Ivan Cimrak
Subject: Re: [ESPResSo-users] Simulating partially immersed deformable objects
Date: Wed, 7 Sep 2016 22:24:44 +0200
User-agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.11; rv:45.0) Gecko/20100101 Thunderbird/45.2.0

Dear Edvin,

My apologies for so delayed  answer.

Object-in-fluid extension works almost independently of the fluid part, the only interaction is via the fluid-particle coupling. From what I know about Shan-Chen in Espresso, there is an analogical coupling. So in principle I think the OIF should perfectly work with Shan-Chen (although I have not tested it yet).

In any case, this is an interesting scenario and if you decide to try it, I will be glad to help you when you get into troubles with the OIF part, just let me know.

With best regards,


On 9/5/16 1:22 AM, Edvin Memet wrote:
Dear all,

I'm interested in simulating a system like the following: an array of nanopillars partially immersed in a fluid. Capillary, bending, and interpillar interaction forces are all of interest.

I've been trying to understand methods of coupling LB to deformable object in general, as well as the current capabilities of Espresso in particular.

From my limited understanding, the objects-in-fluid (OIF) extension seems like the closest starting point for what I want. However, I want the pillars to be partially immersed in the fluid, which I'm not sure is possible. Would it be possible to get that effect by using a two component Shan-Chen fluid (e.g. water and air)? If so, can you even use OIF with Shan Chen? And from my documentation, it seems Shan Chen only runs on GPU - is that still the case?

Or rather than using Shan-Chen, perhaps an easier way of simulating partial immersion is to set the friction coefficient to zero for part of the mesh (it seems it isn't currently possible to individually set the friction coefficient for each mesh node, but perhaps it shouldn't be too difficult for me to figure out a way to do so). Hope that makes sense.

Any guidance/thoughts appreciated.


reply via email to

[Prev in Thread] Current Thread [Next in Thread]