Different max residual values in post and the solver?

hilde · August 25, 2014, 06:41

Dear all,
I am running CFX 14.0 and am trying to see where in my mesh my maximal residuals are located.

According to the solver/The output file my 'V-Mom-Bulk' is my highest momentum residual and its final value is said to be around 1.7*10^-4.

When I want to see it in POST (Using Insert Volume/Isovolume/Variable:V Mom Bulk.Residual) it does not show any elements over 1*10^4. Going for a lower value and coloring the volume according to the variable value using a local scale, POST also suggests that my highest residual for 'V Mom Bulk.Residual' is ca 7.6*10^-5.

I am then not sure which maximal residual value what is true, 1.7*10^-4 or 7.6*10^-5 or if something is fundamentally wrong somewhere.

So, my question is if anyone have experiences this or can see a mistake in my reasoning/method described above and could please give me a hint about why this is?

Thank you in advance

ghorrocks · August 25, 2014, 07:51

If you use the function calculator and use the max() function does that also give you 7.6E-5?

hilde · August 25, 2014, 08:21

Thank you very much for fast reply!

I checked "maxVal(V Mom Bulk rotor.Residual)@rotor" where "rotor" is my problematic area and it was said to be 9.2*10^-5 which then was a third value.

But then something interesting happened:

Changing the definitions of my Isovolume from "Above Value: 5*10^-5" to "Above Value: 1*10^-5" the interval suggested by the local range changed from (Min:5*10^-5, Max:7.67*10^-5) to (Min:1*10^-5, Max:9.2*10^-5).

I could not explain this, but changed to "Above Value: -10" just to cover everything sort of say. The local range then became (Min:-1.7*10^-4, Max:9.2*10^-5) where the minimum value should have been the one I saw in the output file and the maximum one the one detected by using the max function you suggested.

These (or the absolute value of these) should then be the "real" max residual values and the 7.67*10^-5-one just a more visual one explaining the color if the Isosurface which might have surrounded "worse but non-colored" elements inside.

Do you this explanation sounds reasonable?

mvoss · August 25, 2014, 09:37

i´ve always got the iso-volumes and values as expected (values as in solver) when working with

Code:

maxResid = max(max(abs(U Mom.Residual),abs(V Mom.Residual)),abs(W Mom.Residual))

and creating an additional variable from the expression above to work with in cfd-post.
Post shows the Residuum as it is for each cell, the solver shows the RMS for all cells. So, ask for the max(abs(...)) should give you the max-Res. from the solver output.

August 25, 2014, 06:41	Different max residual values in post and the solver?	#1
hilde New Member Hilde Join Date: May 2014 Location: Copenhagen, Denmark Posts: 18 Rep Power: 12	Dear all, I am running CFX 14.0 and am trying to see where in my mesh my maximal residuals are located. According to the solver/The output file my 'V-Mom-Bulk' is my highest momentum residual and its final value is said to be around 1.710^-4. When I want to see it in POST (Using Insert Volume/Isovolume/Variable:V Mom Bulk.Residual) it does not show any elements over 110^4. Going for a lower value and coloring the volume according to the variable value using a local scale, POST also suggests that my highest residual for 'V Mom Bulk.Residual' is ca 7.610^-5. I am then not sure which maximal residual value what is true, 1.710^-4 or 7.6*10^-5 or if something is fundamentally wrong somewhere. So, my question is if anyone have experiences this or can see a mistake in my reasoning/method described above and could please give me a hint about why this is? Thank you in advance

August 25, 2014, 09:37		#4
mvoss Senior Member Matthias Voß Join Date: Mar 2009 Location: Berlin, Germany Posts: 449 Rep Power: 20	i´ve always got the iso-volumes and values as expected (values as in solver) when working with Code: maxResid = max(max(abs(U Mom.Residual),abs(V Mom.Residual)),abs(W Mom.Residual)) and creating an additional variable from the expression above to work with in cfd-post. Post shows the Residuum as it is for each cell, the solver shows the RMS for all cells. So, ask for the max(abs(...)) should give you the max-Res. from the solver output. Last edited by mvoss; August 25, 2014 at 11:04.

August 25, 2014, 07:51		#2
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,848 Rep Power: 144	If you use the function calculator and use the max() function does that also give you 7.6E-5?

August 25, 2014, 08:21		#3
hilde New Member Hilde Join Date: May 2014 Location: Copenhagen, Denmark Posts: 18 Rep Power: 12	Thank you very much for fast reply! I checked "maxVal(V Mom Bulk rotor.Residual)@rotor" where "rotor" is my problematic area and it was said to be 9.210^-5 which then was a third value. But then something interesting happened: Changing the definitions of my Isovolume from "Above Value: 510^-5" to "Above Value: 110^-5" the interval suggested by the local range changed from (Min:510^-5, Max:7.6710^-5) to (Min:110^-5, Max:9.210^-5). I could not explain this, but changed to "Above Value: -10" just to cover everything sort of say. The local range then became (Min:-1.710^-4, Max:9.210^-5) where the minimum value should have been the one I saw in the output file and the maximum one the one detected by using the max function you suggested. These (or the absolute value of these) should then be the "real" max residual values and the 7.6710^-5-one just a more visual one explaining the color if the Isosurface which might have surrounded "worse but non-colored" elements inside. Do you this explanation sounds reasonable?