can i use any other CFD model for micro scale flows or the mesh?

munib · April 16, 2018, 23:52

Hi,
Actually, I am trying to calculate friction factor for flows in rectangular sinusoidal mini-channel, while the height, width length are .250, 1 and 1 mm respectively, I also tried to model the mesh by following a published research paper, the developed mesh sizes where 1million to 4 million cells.

But still, the results were far away from the theoretical and experimental values. like the expected friction value was expected around 0.2 but the values I got were more than 1.
For the analysis, I considered the viscous laminar model with velocity inlet and press outlet were B. conditions. while the considered Re number is 100,
so am I supposed to use any other CFD model? or the problem is with the mesh?
Thanks in advance

LuckyTran · April 20, 2018, 15:35

Try adding some prism layers. You probably don't need to worry about stretching but it really helps to have a few layers that are wall parallel because it really helps in computing the gradient near walls. Even as few as 3-5 layers can do wonders.

Did the referenced paper also use a tet mesh? You need a lot more tets than hexahedral cells to make up for the poorer gradients that you get with tets. Your geometry looks simple enough to where a hexahedral or polyhedral mesh could be easily generated. That could help, but try the prism layers first.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Keeping intermediate files	OVS	SU2	5	December 5, 2021 11:41
Icemcfd 11: Loss of mesh from surface mesh option?	Joe	CFX	2	March 26, 2007 18:10
Feasiblity of CFD model for Real-Time applications	la mimi	Main CFD Forum	2	January 28, 2007 04:49
ICEM CFD 5.1 Hex-Tet mesh merging failure	bogesz	CFX	1	January 29, 2005 06:46
Mesh generator and CFD solver	Gennady Kireyko	Main CFD Forum	0	May 6, 2001 11:13

April 20, 2018, 15:35		#2
LuckyTran Senior Member Lucky Join Date: Apr 2011 Location: Orlando, FL USA Posts: 5,676 Rep Power: 66	Try adding some prism layers. You probably don't need to worry about stretching but it really helps to have a few layers that are wall parallel because it really helps in computing the gradient near walls. Even as few as 3-5 layers can do wonders. Did the referenced paper also use a tet mesh? You need a lot more tets than hexahedral cells to make up for the poorer gradients that you get with tets. Your geometry looks simple enough to where a hexahedral or polyhedral mesh could be easily generated. That could help, but try the prism layers first.