Unable to run lid-driven cavity flow with 1M elements

dougalb · October 17, 2013, 21:19

I am trying to recreate a set of tests done for the hpcadvisorycouncil - http://www.hpcadvisorycouncil.com/pd...l_2680_FCA.pdf

It calls for "Lid Driven Cavity Flow - 1 Million elements, 2D, icoFoam solver for laminar, isothermal, incompressible flow", so I have taken the OpenFOAM tutorial and modified the blockMesh to be 1000 x 1000 versus 20 x 20. When I run, it gives me a "floating point error".

Any tips or ideas how I can configure the case to match the example cited? My goal is to run it on 1,2,4,8,16,32,64 cores.

haakon · October 18, 2013, 03:31

I think it would be useful to give us some more information, for example a more detailed error message.

My first suspicion is that you have a too large timestep tough... A 1M mesh in 2D would imply 1000x1000 cells (if square), and that require a very small timestep, in special because your top corners will behave as singular points with high velocities.

puneet336 · April 7, 2019, 00:34

Hi hakon, dougalb
I am trying the same experiment,

Post is here -
Setting up Lid driven Cavity Benchmark with 1M cells for multiple cores

input files for cavity experiment is here
This is the error message i am getting , could you please advice.

October 17, 2013, 21:19	Unable to run lid-driven cavity flow with 1M elements	#1
dougalb New Member dougal Join Date: Oct 2013 Posts: 1 Rep Power: 0	I am trying to recreate a set of tests done for the hpcadvisorycouncil - http://www.hpcadvisorycouncil.com/pd...l_2680_FCA.pdf It calls for "Lid Driven Cavity Flow - 1 Million elements, 2D, icoFoam solver for laminar, isothermal, incompressible flow", so I have taken the OpenFOAM tutorial and modified the blockMesh to be 1000 x 1000 versus 20 x 20. When I run, it gives me a "floating point error". Any tips or ideas how I can configure the case to match the example cited? My goal is to run it on 1,2,4,8,16,32,64 cores.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
lid driven cavity varying results	yasmil	OpenFOAM Running, Solving & CFD	2	October 6, 2016 22:42
Standard k-e model in Lid driven cavity	rajeshkunwar	Main CFD Forum	4	February 11, 2011 03:18
2D LID driven cavity	reeshav	Main CFD Forum	1	April 12, 2010 02:51
lid driven cavity	ani	Main CFD Forum	3	April 2, 2006 14:27
Benchmark (Lid driven cavity, ...)	rt	Main CFD Forum	3	April 1, 2006 10:27

October 18, 2013, 03:31		#2
haakon Senior Member Join Date: Dec 2011 Posts: 111 Rep Power: 19	I think it would be useful to give us some more information, for example a more detailed error message. My first suspicion is that you have a too large timestep tough... A 1M mesh in 2D would imply 1000x1000 cells (if square), and that require a very small timestep, in special because your top corners will behave as singular points with high velocities.

April 7, 2019, 00:34		#3
puneet336 New Member puneet Join Date: Nov 2015 Posts: 19 Rep Power: 11	Hi hakon, dougalb I am trying the same experiment, Post is here - Setting up Lid driven Cavity Benchmark with 1M cells for multiple cores input files for cavity experiment is here This is the error message i am getting , could you please advice.