star cd convergence error

gagi · September 14, 2011, 04:16

hi im a chemical engineering post graduate student and i have to model and simulate a tray dryer using STAR CD and im somewhat new to CFD. i started my model as a .IGES file and created the surface mesh in pro-surf as a .dbs file n then as a .mdl file in pro amm. when i run the model there are convergence errors and when i check the residuals and browse the standard output file, after 2 iterations there is an error message saying,
"FLUID FLOW DETECTED AT SOME OUTFLOW BOUNDARIES MASS FLUX IS FORCED TO ZERO. IF THIS CONDITION PERSISTS, YOUR CALCULATIONS MAY BE WRONG "

the mesh didn't get converged after 100000 iterations. my mesh got around 1 500 000 fluid (air) trimmed cells, with 5ms-1 air speed at the inlet and flow is steady state and laminar. used SIMPLE method to solve it. and in the star guide panel, i make all the input parameters like i did in the STAR tutorial 2nd example, (flow in a partially blocked elbow).

im new to CFD and STAR CD and spend considerable time to figure out whats wrong with my model .. can you please give me some advice?

RobertB · September 14, 2011, 08:14

You probably have a recirculation zone hitting the exit boundary (from the error messages).

Either that or you have a problem that does not have a steady state solution, are the residuals moving cyclically?

gagi · September 14, 2011, 10:17

hi, thanks for replying.

i think you are correct. when i apply initial conditions, i used 1 atm as inlet in refernce pressure cell so flow may probably flows inward at the outlet as static pressure + dynamic pressure at outlet is greater than the inlet pressure. thanks a lot. il try it again for actual pressure.
can i select a cell form outlet boundary as a refernce pressure cell?

yaaa there were cyclic residuals and mass flux residual is circulating near 1 and other velocity u, v and w components are circulating near 0.01 and below.. i used 0.001 as the maximum residual tolarance.
initially im trying to model the air flow and later going to modify mesh to obtain uniform velocity distribution. as only air at 20 C temperature is involved i took it as steady state and SIMPLE mode (even PISO and SIMPSO did not give converged solution)

thanks again. tc

RobertB · September 14, 2011, 11:15

You can use any cell as your reference if you do not have pressure boundaries in the model. It only really has any real effect if your are running a compressible fluid (i.e. ideal gas).

Typically plotting up the flow field gives you a pretty good idea as to what is going wrong (even if unconverged).

September 14, 2011, 04:16	star cd convergence error	#1
gagi New Member Join Date: Mar 2011 Posts: 11 Rep Power: 15	hi im a chemical engineering post graduate student and i have to model and simulate a tray dryer using STAR CD and im somewhat new to CFD. i started my model as a .IGES file and created the surface mesh in pro-surf as a .dbs file n then as a .mdl file in pro amm. when i run the model there are convergence errors and when i check the residuals and browse the standard output file, after 2 iterations there is an error message saying, "FLUID FLOW DETECTED AT SOME OUTFLOW BOUNDARIES MASS FLUX IS FORCED TO ZERO. IF THIS CONDITION PERSISTS, YOUR CALCULATIONS MAY BE WRONG " the mesh didn't get converged after 100000 iterations. my mesh got around 1 500 000 fluid (air) trimmed cells, with 5ms-1 air speed at the inlet and flow is steady state and laminar. used SIMPLE method to solve it. and in the star guide panel, i make all the input parameters like i did in the STAR tutorial 2nd example, (flow in a partially blocked elbow). im new to CFD and STAR CD and spend considerable time to figure out whats wrong with my model .. can you please give me some advice?

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Compile calcMassFlowC	aurore	OpenFOAM Programming & Development	13	March 23, 2018 07:43
How to install CGNS under windows xp?	lzgwhy	Main CFD Forum	1	January 11, 2011 18:44
CGNS lib and Fortran compiler	manaliac	Main CFD Forum	2	November 29, 2010 06:25
compile errors of boundary condition "expDirectionMixed"	liying02ts	OpenFOAM Bugs	2	February 1, 2010 20:11
[Netgen] Installation of Netgen in SuSE Linux 92	edvardsenpriv	OpenFOAM Meshing & Mesh Conversion	23	January 16, 2009 06:12

September 14, 2011, 08:14		#2
RobertB Senior Member Robert Join Date: Jun 2010 Posts: 117 Rep Power: 16	You probably have a recirculation zone hitting the exit boundary (from the error messages). Either that or you have a problem that does not have a steady state solution, are the residuals moving cyclically?

September 14, 2011, 10:17		#3
gagi New Member Join Date: Mar 2011 Posts: 11 Rep Power: 15	hi, thanks for replying. i think you are correct. when i apply initial conditions, i used 1 atm as inlet in refernce pressure cell so flow may probably flows inward at the outlet as static pressure + dynamic pressure at outlet is greater than the inlet pressure. thanks a lot. il try it again for actual pressure. can i select a cell form outlet boundary as a refernce pressure cell? yaaa there were cyclic residuals and mass flux residual is circulating near 1 and other velocity u, v and w components are circulating near 0.01 and below.. i used 0.001 as the maximum residual tolarance. initially im trying to model the air flow and later going to modify mesh to obtain uniform velocity distribution. as only air at 20 C temperature is involved i took it as steady state and SIMPLE mode (even PISO and SIMPSO did not give converged solution) thanks again. tc

September 14, 2011, 11:15		#4
RobertB Senior Member Robert Join Date: Jun 2010 Posts: 117 Rep Power: 16	You can use any cell as your reference if you do not have pressure boundaries in the model. It only really has any real effect if your are running a compressible fluid (i.e. ideal gas). Typically plotting up the flow field gives you a pretty good idea as to what is going wrong (even if unconverged).