LaunderSharmaKE diverges and return 'nan' error

vcvedant · July 26, 2017, 17:02

Hello,

I am simulating turbulent flow over flat plate using low Reynolds number model: LaunderSharmaKE.

Boundary conditions:
Inlet: U (30, 0 ,0), p 0, k 0.13, epsilon 2.56, nut calculated.
Outlet: zeroGradient
top: zeroGradient
plate: U (0 0 0), p - zeroGradient, k 1e-10, epsilon 1e-8, nut 0.

I am using simpleFoam solver but after running for 14 iterations my solution diverges and I receive an error:
wrong toke type - expected Scalar, found on line 0 the work 'nan'.
When I checked the log files I found that k and epsilon have very large min and max bounds.

Can someone help me resolve this or guide me in the write direction?

Thanks,
Vedant

fumiya · July 29, 2017, 10:00

You might want to check the following points:
1. [Mesh] Y+ value should be at most order of 1 on the plate boundary.
2. [Boundary Condition] Inlet: p zeroGradient, Outlet: p 0

Best regards,
Fumiya

sheaker · July 29, 2017, 11:29

Hello.
Have You tried to set relaxations factor?
Check my post:
Sudden divergence of simulation
Or it could be the mesh problem too.

EDIT:
'nan' means not a number.

vcvedant · July 29, 2017, 16:53

Thank you all for your reply.
I think it was a mesh problem. I created a mesh in pointwise for OpenFOAM and used that which was giving me errors. I created a new mesh (using the same parameters as pointwise) using blockMeshDict which was able to converge.
I checked the mesh imported from pointwise using checkMesh and found no errors. My y+ was 0.3. So, I don't understand what was the problem with the previous mesh?

July 26, 2017, 17:02	LaunderSharmaKE diverges and return 'nan' error	#1
vcvedant Member Vedamt Chittlangia Join Date: Feb 2016 Posts: 64 Rep Power: 9	Hello, I am simulating turbulent flow over flat plate using low Reynolds number model: LaunderSharmaKE. Boundary conditions: Inlet: U (30, 0 ,0), p 0, k 0.13, epsilon 2.56, nut calculated. Outlet: zeroGradient top: zeroGradient plate: U (0 0 0), p - zeroGradient, k 1e-10, epsilon 1e-8, nut 0. I am using simpleFoam solver but after running for 14 iterations my solution diverges and I receive an error: wrong toke type - expected Scalar, found on line 0 the work 'nan'. When I checked the log files I found that k and epsilon have very large min and max bounds. Can someone help me resolve this or guide me in the write direction? Thanks, Vedant atulkjoy likes this.

July 29, 2017, 10:00		#2
fumiya Senior Member Fumiya Nozaki Join Date: Jun 2010 Location: Yokohama, Japan Posts: 266 Blog Entries: 1 Rep Power: 19	You might want to check the following points: 1. [Mesh] Y+ value should be at most order of 1 on the plate boundary. 2. [Boundary Condition] Inlet: p zeroGradient, Outlet: p 0 Best regards, Fumiya __________________ [Personal] Blog: https://caefn.com/ Twitter: https://twitter.com/FumiyaNozaki SlideShare: http://www.slideshare.net/fumiyanozaki96

July 29, 2017, 11:29		#3
sheaker Senior Member Oskar Join Date: Nov 2015 Location: Poland Posts: 184 Rep Power: 11	Hello. Have You tried to set relaxations factor? Check my post: Sudden divergence of simulation Or it could be the mesh problem too. EDIT: 'nan' means not a number. __________________ How to set two stroke engine with scavenging? Salome, sliding interface and "Zero length edge detected." Last edited by sheaker; July 29, 2017 at 16:50. Reason: new thought

July 29, 2017, 16:53		#4
vcvedant Member Vedamt Chittlangia Join Date: Feb 2016 Posts: 64 Rep Power: 9	Thank you all for your reply. I think it was a mesh problem. I created a mesh in pointwise for OpenFOAM and used that which was giving me errors. I created a new mesh (using the same parameters as pointwise) using blockMeshDict which was able to converge. I checked the mesh imported from pointwise using checkMesh and found no errors. My y+ was 0.3. So, I don't understand what was the problem with the previous mesh?