Parallel collision algorithm

stephan · October 31, 2006, 10:43

Dear All,

i want to write a parallel version of a collisionmodel in libdieselspray.
Unfortunatly i cant find the point (in liblagrangian) where the global particlelist is splitted into smaller ones for each processor.
can anybody give me some hints how the parallel-tracking works.
any help would be appreciated
stephan

gschaider · November 1, 2006, 16:44

Hi Stephan!

As I understand it each processor administrates the particles in his mesh regions. If the particles leaves his sub-mesh he hands it over to the other processor (who's mesh he enters). So there is no need for splitting. But you might have a look at decomposePar - that should split the list.

stephan · November 2, 2006, 07:14

Hi Bernhard,

at first : thanks for your help!
i already looked through the decompose-files and i think i found what i need. (of course understanding is something different..)
i want to implement an soft sphere modell with a spring-dashpot-slider element. so i look for opportunities to have particle-collisions in the near or at a processor-boundary, which is not possibel right now.

another thing: i am working in berlin and would be interested in some knowledge exchange if you are interested too?
what kind of problems your are working on?

stephan

October 31, 2006, 10:43	Dear All, i want to write a	#1
stephan Senior Member Stephan Gerber Join Date: Mar 2009 Location: Germany Posts: 118 Rep Power: 17	Dear All, i want to write a parallel version of a collisionmodel in libdieselspray. Unfortunatly i cant find the point (in liblagrangian) where the global particlelist is splitted into smaller ones for each processor. can anybody give me some hints how the parallel-tracking works. any help would be appreciated stephan

November 1, 2006, 16:44	Hi Stephan! As I understand	#2
gschaider Assistant Moderator Bernhard Gschaider Join Date: Mar 2009 Posts: 4,225 Rep Power: 51	Hi Stephan! As I understand it each processor administrates the particles in his mesh regions. If the particles leaves his sub-mesh he hands it over to the other processor (who's mesh he enters). So there is no need for splitting. But you might have a look at decomposePar - that should split the list. __________________ Note: I don't use "Friend"-feature on this forum out of principle. Ah. And by the way: I'm not on Facebook either. So don't be offended if I don't accept your invitation/friend request

November 2, 2006, 07:14	Hi Bernhard, at first : tha	#3
stephan Senior Member Stephan Gerber Join Date: Mar 2009 Location: Germany Posts: 118 Rep Power: 17	Hi Bernhard, at first : thanks for your help! i already looked through the decompose-files and i think i found what i need. (of course understanding is something different..) i want to implement an soft sphere modell with a spring-dashpot-slider element. so i look for opportunities to have particle-collisions in the near or at a processor-boundary, which is not possibel right now. another thing: i am working in berlin and would be interested in some knowledge exchange if you are interested too? what kind of problems your are working on? stephan

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
About Collision Model	taksue	OpenFOAM Post-Processing	6	January 14, 2009 02:10
particle collision	Sri	FLUENT	0	November 28, 2006 03:07
DPM Collision UDF	Dave	FLUENT	0	October 23, 2006 04:14
DPM drop collision	kiao	FLUENT	0	May 31, 2005 04:17
droplet collision	kiao	FLUENT	0	May 29, 2005 23:01