about solid resident time distribution

Submitted by ngcw1986 on Sun, 04/14/2013 - 08:22

Dear Forumers:
I noticed a new feature form the announcement of the LIGGGHTS2.3 release, which can calculate solid resident time distribution in a region- "Added new fix property/atom/tracer and fix property/atom/tracer/stream commands along with a new compute nparticles/tracer/region command for the calculation of residence time distributions". Howerer, I can not find a test case or a tutorial to use this feature, and I also wonder how to ouptut the results as simulation processes. If anybody knows, please give me some clue. Thanks in advance.

ckloss's picture

ckloss | Fri, 04/19/2013 - 15:09

Hi,

just google for "residence time distribution", I'm sure this will help

Cheers
Christoph

ngcw1986 | Sun, 04/28/2013 - 15:17

Dear ckloss,
Following your suggestions, I google some information on the "residence time distribution" (RTD). Now, I still has some confusions how to get the useful information necessary for the calculation of RTD. The function of "fix tracer" can mark all particles in a certain region. Combined with the "compute tracer/region" function, liggggts can output the particle number variation in the specified region as simulaiton processes. However, only particle number can be obtained while the prerequisite concentration information (either a mass fraction or a number fraction of the marked particle in the specified region and especially at its outlet, E(t)) can not be obtained.
Is there any function to output or calculate these information? Can you give some more clues or instructions to figure out the RTD using liggghts? Thanks in advance.

ckloss's picture

ckloss | Thu, 05/02/2013 - 13:29

>>in the specified region and especially at its outlet, E(t)) can not be obtained.
sure it can - just use per-atom variables to calculate it

Christoph

ngcw1986 | Fri, 05/03/2013 - 09:41

I am still confused. As a matter of fact, the RTD of solid in a vessel can be figured out by dumping and mointoring the position of each solid at a small time interval. However, this is labourious. I originally thought the new function for RTD in liggghts is on a "group level", which may dump mass concentration of tracer particles at assigned regions to determining RTD. However, your reply gives me a impression that it is a per-atom variable (resident time for each tracer solid) which can be calculated and stored. Is that right?
If I want to figure out the RTD for a CFB riser, I first mark the solids in the region in the bottom part of the riser (fix property_atom_tracer), then I mointor the number variation of marked particle in the riser (computer_nparticles_tracer_region) as simulation proceeds. I dug into the "variable" document and still cannot figure out a way to separate the tracers form all the solids coming out of the outlet and calculate the corresponding tracer mass concentration. How could I figure out the tracer mass concentration using per-atom variables? Could you please explain a little bit more clear?
thank you for your reply!

ckloss's picture

ckloss | Fri, 05/03/2013 - 10:19

Hi ngcw1986,

the function was designed for batch systems and serves to mark representative particles when entering the vessel and you can measure when the marked particles exit the system

>>How could I figure out the tracer mass concentration using per-atom variables
see doc for details

>>which may dump mass concentration of tracer particles at assigned regions to determining RTD
you can do that easily using fix ave/spatial or the like.

Christoph

mhk | Mon, 09/16/2013 - 04:08

Hi, I am trying to use this commands, but when I try to dump the global scalar of this compute command, I keep receiving an error saying "Dump local compute does not compute local info", could you please let me know how I can dump it? or where is my mistake? the below is related command lines of my script.

region rtd block -0.5 -0.5 -1 1 -1 1 units box
fix rt all property/atom/tracer region_mark rtd mark_step 10000 marker_style heaviside check_mark_every 100000
compute nparticles all nparticles/tracer/region region_count rtd tracer rt
dump dumpRTD all local 100000 dump.dumpRTD c_nparticles

Cheers

ckloss's picture

ckloss | Thu, 09/19/2013 - 20:57

Hi mhk,

the fix computes a global value, not a local one. You can e.g. output this with fix print

Cheers
Christoph

kunzhang's picture

kunzhang | Tue, 02/17/2015 - 01:44

Hi everyone,

Just to follow this post, it is still unclear for me "how to measure when the marked particles exit the system". Or how to calculate the residence time of a particle staying in a defined domain? Could anyone please help me on this time output?

Thanks so much!

Kun

kunzhang