Heat flow in aggregate (hybrid granular/molecular) particles

msbentley's picture
Submitted by msbentley on Mon, 08/29/2011 - 11:59

Hi all,

I have been playing with aggregated particles using a hybrid granular/molecular pair style, and am now introducing heat flow into the simulation. In the current implementation if I exclude the aggregates from neighbor calculations, there is no heat flow between member particles (this, of course, makes sense!). If I remove this exclusion, then heat does indeed flow, but at a rate determined by the thermal properties and the initial overlap when the aggregates are created (i.e. determining the contact area).

I would prefer either to be able to either:

  1. set a different conductivity within an aggregate (so that, for example, I can use a rather high value to allow the entire particle to equilibrate quickly), or
  2. treat the aggregate as a single particle, with a specified per-contact conductivity, but a lumped heat capacity and a constant temperature throughout.

Are either one of these currently possible? Perhaps something similar is already in the pipeline as part of the multi-sphere implementation?

Thanks!

Mark

Forums:

ckloss's picture

ckloss | Tue, 08/30/2011 - 09:16

Hi Mark, this is nothing that can be done out-of-the-box now. And we would like to insure that the methodology also works for an overlapping particle configuration, so the way to go for us would be option 2 (which could maybe later be enhanced by some more clever way to deal with internal heat transfer in the particle), but nobody has started working on that yet. A quick hack to establish that for a non-overlapping configuration would be a fix that loops through bodies and just sets the spheremass-averaged temperature throughout the body.

Christoph