Re: [nek5000] Natural convection with stratified background

[Ashwin Vishnu Mohanan]

Hi Zilin,

To set the diffusion terms along with Reynolds number and the Prandtl number you can set the following properties in the par file.

[VELOCITY] 
viscosity = -5000 
density = 1.0
 
[TEMPERATURE] 
conductivity = -5000.0 
rhoCp = 1.0

I think then you can use the following userf

include 'SIZE'
include 'NEKUSE'

ffx = 0
ffy = 2 * temp / param(2)

And the 4th source term in the energy equation (34) should be specified in userq – although it is not clear to me from the documentation or NekExamples how to do that.

For more info:
https://nek5000.github.io/NekDoc/problem_setup/case_files.html#parameter-file-par
https://nek5000.github.io/NekDoc/appendix.html#internal-input-parameters-switches

Cheers,
Ashwin

On 9/19/21 14:15, Zilin Zong wrote:

Hi Neks,

I have been working on this set of dimensionless equation for a natural convection case.

I have put the RHS term into the ffx and ffy term in usrf() as follows,
ffx = 1/Re

ffy = 1/Re + 2*temp/Re

And also

ffx = mu/uL

ffy = mu/uL + 2*temp*mu/uL

after knowing that the forcing term gets multiplied by rho. However the result from both run looks similar and far different from the benchmark result. I am suspecting that 2/RePr and 1/RePr term in the temperature scalar is the reason for it. Is there anyway that I can fix this?

Thanks

Zilin

--
You received this message because you are subscribed to the Google Groups "Nek5000" group.
To unsubscribe from this group and stop receiving emails from it, send an email to nek5000+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/nek5000/6c150954-6f6a-4f0a-a421-ee4bee7369c8n%40googlegroups.com.

​

[Ashwin Vishnu Mohanan]

Some corrections in the userf I wrote above:

include 'SIZE'
include 'INPUT'
include 'NEKUSE'

ffx = 0
ffy = 2 * temp * param(2)

Note: by specifying negative values in par files we set Reynolds number and Prandtl number. Therefore param(2) would be 1. / 5000 in this example.

Ashwin

​

[Zilin Zong]

Thanks Ashwin ! Much appreciated