Strange spikes in my 2nd Order Solutions
 

Hi all,

I have trouble with the second order schemes for the convective part at high reynolds numbers.

I am running a 2D cylinder at Re 27400 with a RAS turbulence model (kE or RNG) and get a solution like the one shown in the picture I linked:

http://www.harly.de.vu/www/spikes.png

The coefficients are calculated with the common controlDict entries and the plot is made with a simple gnuplot command. As you can see the solution is "decorated" with random spikes(these can be huge ~100000).

I am using the "Gauss linear" for the convective term.

Anyone any ideas?
-harly

Hi,Daniel

I could not answer your question.I recently also do some numerical analysis of the cylinder flow at hight Re number.I remember once I have tried a case similiar to yours.2*10^5,but get a differenct result.Could you share the files?I would like to do a comparative analysis of this case.

Sure,

it will take me a day or two, to come back to you, but just post your email address and I will send you the file.

- harly