
[Sponsors] 
March 20, 2012, 07:18 
DPM parcel mass content and convergence

#1 
Member
Join Date: Mar 2009
Location: Istanbul, Turkiye
Posts: 45
Rep Power: 9 
Hi,
I am using FLUENT DPM model for simulation of dilute gassolid flow. Now, I am looking for an information about the optimal, if exists, (regarding the optimal convergence) mass of particles in a parcel (Definition is included in FLUENT manual). However, I could not reach any info. So, does anybody has any recommendation on this topic? Thanks. 

March 20, 2012, 11:39 

#2 
Member
Domenico
Join Date: Sep 2011
Location: Atlanta
Posts: 39
Rep Power: 6 
hi
if i'm understanding well, you mean how much mass would be optimal a fluent particle (to be intended as cluster of particles) represents, correct? if yes, i think it's very important to consider the particle dimension with respect to the mesh resolution, particularly at the inlet. one cluster would tend to represent one real particle (in terms of its mass) if the mesh is ideally refined to infinite, or it generally would represent a more realistic mass if the mesh is highly refined, particularly for finer particles. this would help to a better convergence regards 

March 20, 2012, 16:13 

#3  
Member
Join Date: Mar 2009
Location: Istanbul, Turkiye
Posts: 45
Rep Power: 9 
Hi doronzo
Yes, you are correct. That is exactly what I tried to point. Thank you for your explanations. I am currently trying to optimize the number of parcels of several injections that have different particle diameters and mass flowrates for a specific gassolid conveying problem. So, do you have any recommendation on improving the convergence and/or any solution strategy for this type of problem. thanks. Quote:


March 20, 2012, 18:58 

#4 
Member
Domenico
Join Date: Sep 2011
Location: Atlanta
Posts: 39
Rep Power: 6 
the mass represented by each cluster/injection will depend on the inlet mesh, which is more like a choice or design than a method, and i haven't any specific suggestion.
what i can guess is a relatively high momentum exchange between the poorlysorted population of particles and the fluid, which in terms of convergence could be managed by using few iterations for the fluid/iteration for the solid, particularly when the mixture is relatively dense. also, the twoway turbulence coupling could physically help. how big are the smallest particles? how concentrated is the most dilute injection? 

March 21, 2012, 07:56 

#5 
Member
Join Date: Mar 2009
Location: Istanbul, Turkiye
Posts: 45
Rep Power: 9 
Hi doronzo,
Thank you for your comments. Representative average diameters are as follows: The smallest particle is approx 50 micron while the largest one is 600 micron. If you ask DPM concentration, the max volume fraction inside the comp domain is aprox. 3e4. regards. 

March 21, 2012, 08:16 

#6 
Member
Domenico
Join Date: Sep 2011
Location: Atlanta
Posts: 39
Rep Power: 6 
my experience would say that if you are using an implicit discretization, along with a relatively big time step, then you could use one iteration for the fluid/iteration for the solid, particularly when the fluid is compressible, and the solid is multiparticle
regards 

Thread Tools  
Display Modes  


Similar Threads  
Thread  Thread Starter  Forum  Replies  Last Post 
Species mass flow inlet  lorenz  FLUENT  3  March 15, 2012 08:26 