Comparison of 2D model and 3d model with periodic boundary condition
My Lab Book:
Feb 27
Tasks and log:
1. try to duplicate long energy dissipation rate
to compare with(control / reference case): S:\AguaClara\Single Bend Simulation\k-e realizable 1_5 W\Clearance 1.5W.cas
try to duplicate with (with different partitions):
- run simulation from Yong Sheng's msh = S:\AguaClara\Single Bend Simulation\k-e realizable 1_5 W\Clearance 1.5W.msh
- duplicate ALL KNOWN parameters
case 1: directly run FLUENT
case 2: parallel -t1
case 3: parallel -t2
(more: do 2 control cases with water vapor and air)
LOG:
case 1:
- redefine material to water liquid in REFERENCE case and iterate again convergence:
converge after 600 iterations
similar shape with that of before redefining - need to save the older contour plot next time
- case 1 steps:
define model-viscous-rke;
define material-water liquid;
define boundary conditions: inlet-check-ti 1% hd 0.04, outlet-check-default;
Solve-directly second order, residual-directly e-6;
START iteration - converge after 1276 steps;
- case 1 observations:
difference shape and sizes of energy dissipation region;
recall the shape for from 1st order e-3 to 2nd order e-6: minor difference in the shape of energy dissipation
reagion, but similar size, not sure about the scale;
- case 1 conclusions, questions and further test based on the above observations:
check the sensitivity to iteration routes;
can we go ahead and skip the problem? - probably no, energy dissipation rate is all that this is about!
...? can't remember;
POPUP ideas:
more differences to look into:
number of partitions;
iteration steps;
how important is the color scale of the contour;
use spreadsheet as my lab book;
how to design a more effective test that tell us the answer faster ?????????
Findings:
partition is related to parallel computing - the number of cores being used
List of Case files:
case 1: run simulation from Yong Sheng's msh in :S:\AguaClara\Single Bend Simulation\k-e realizable 1_5 W