Posts of 5.7.1 and 10 show noticeable difference
 

Hi all,

I accidently found something weird in post. I was using one pc (with CFX10) to conduct a simulation, got the results. I copied the result file to another pc which has cfx5.7.1. When I plot some results using the post, I found there are differences between cfx10 and cfx5.7.1 posts. I am using the same results file, ploting the same parameter at the location with the same amount of samples. I noticed this and examined the ploting parameters closely, to make sure I used the same for everything except two different posts. Wondered what caused the difference? Please check out this image for the difference two two posts produced. www.iesd.dmu.ac.uk/~yingchun/cfx/linegraphs.png

Any idea about this?

Thanks a lot

Forrest

Re: Posts of 5.7.1 and 10 show noticeable differen
 

Are you using exactly the same definition file that you created from 5.7.1 when you run it in CFX10 solver? It will be useful to give more details about the case you're solving.

Re: Posts of 5.7.1 and 10 show noticeable differen
 

Thank you for your reply, TB. Only one definition file created by 5.7.1 and then I run it in cfx10 to get a result file.

My case involves both solid and fluid domains. The difference shows at solid domains. Basically, it's modelling an external solar heat flux travel through a short distance (air domain), striking on a single glazing (glass, modelled as semi-transparent solid) then travel through an air cavity with Venetian blinds (small solid slats, only CHT modelled), back the air cavity is another glazing (only the thickness is bigger than previous one), and then the internal environment (air domain). It looks the differences all happen at the solid domains. It's a 2D modelling.

I am using the same results file from the same definition file, the posts doesn't plot consistent results, I noticed this accidently, wondered what happened.

Cheers, Forrest

Re: Posts of 5.7.1 and 10 show noticeable differen
 

Are you plotting hybrid or conservative variables? This might be the difference between post 10 and 5.7

Re: Posts of 5.7.1 and 10 show noticeable differen
 

Hi nico, thanks for your reply.

I creat a line first across the domain (middle height), and then export the 'temperature' and 'x' data on this line to a .csv file. Then I plot the data as shown in the graph using excel. It seems no where I can choose 'hybrid' ro 'conservative' option from this two procedure. But I assume it's conservative because the starting value is not my boundary setting (my opening with entrainment boundary condition at x=0 has a temperature 20C, the graph shows a slightly high value).

Cheers, Forrest