CFD Online Discussion Forums

CFD Online Discussion Forums (
-   OpenFOAM (
-   -   good convergence but bad display (

zaynah04 January 8, 2013 16:37

good convergence but bad display
Dear members,

my solution iterate ..
when i display on paraFoam i get bad results.
example for Pressure i obtain only RED everywhere and for velocity i get the Blue everywhere..

Can any one please guide me..

norkistar January 8, 2013 22:45

Can you give some details about your problem and the results that we can have a look?

zaynah04 January 9, 2013 03:14

hi thanks for your reply.

I have achieved iteration and when i display in paraFoam, i get bad results. by bad results i mean that for the pressure display i get red color in the whole region! and in velocity contour i get Blue colour in the whole domain region..:(
i have used the Glymph filter as instructed in the manual before display my velocity contour.

I think i have not been able to use the paraview settings as per regulations.
can you please guide me?


norkistar January 9, 2013 03:28

Hi I think you might need to refer to the user guide of the paraview and then become familiar with how to use it since you can display pressure and velocity using Glymph filter


zaynah04 January 9, 2013 03:31

I had actually read the user manual and follow the steps..:(

this is why i am posting my problem..:(

norkistar January 9, 2013 03:36

Can you list the step that you did here that we can have a look in detail?

zaynah04 January 9, 2013 03:58

1- in display panel i choose U for display of velocity
2- in glymph filter i choose glymph arrow, scale factor to 0.005.

then i display it all...

and for pressure..
i choose p for display pressure and rescale to data range and also i choose surface,,,

norkistar January 9, 2013 04:01

then i think after you choose U, you should choose magnitude just at the rightside of U

zaynah04 January 9, 2013 04:05

i did choose that..:(

norkistar January 9, 2013 04:15

you choose surface at the rightside of magnitude?

zaynah04 January 9, 2013 04:18

surface was chosen for style presentation!

Traib January 9, 2013 08:13

Hi zaynah,

Its difficult to guess from your description coz you have tried most of the stpes already but just to check
1. are you running the simulation in parallel? If yes, did you do 'reconstructPar'?
2. did you try 'rescale to data range' at following time steps? sometimes for slower systems the data takes some time to load so the calculated data appear after some time. I get this problem when I am using my laptop which is quite slow, so I go back and forward in time steps and rescale.
3. did you check the time text files to see the calculated data values? Just in case :D
4. or like user norkistar said, you could attach more details so that we could have a look.


zaynah04 January 9, 2013 10:07

1 Attachment(s)
Hi traib,
no i not run in parallel and yeah i rescale to data range ..:(

i have attached the picture..
below is part of the result of iteration..

Time = 10

Courant Number mean: 4.89461e-06 max: 9.43364e-05
DILUPBiCG: Solving for Ux, Initial residual = 0.000645717, Final residual = 3.21458e-06, No Iterations 1
DILUPBiCG: Solving for Uy, Initial residual = 0.00104037, Final residual = 3.59143e-06, No Iterations 1
DICPCG: Solving for p, Initial residual = 0.00184126, Final residual = 9.98725e-07, No Iterations 293
DICPCG: Solving for p, Initial residual = 9.24984e-05, Final residual = 9.73594e-07, No Iterations 100
DICPCG: Solving for p, Initial residual = 5.70579e-06, Final residual = 9.16013e-07, No Iterations 4
time step continuity errors : sum local = 2.79268e-16, global = 2.18109e-17, cumulative = 5.8943e-13
DICPCG: Solving for p, Initial residual = 0.00019026, Final residual = 9.77781e-07, No Iterations 274
DICPCG: Solving for p, Initial residual = 2.72308e-05, Final residual = 9.93397e-07, No Iterations 33
DICPCG: Solving for p, Initial residual = 2.03678e-06, Final residual = 8.69365e-07, No Iterations 1
time step continuity errors : sum local = 2.65075e-16, global = -4.99172e-20, cumulative = 5.8943e-13
ExecutionTime = 1881.44 s ClockTime = 1884 s


Traib January 9, 2013 10:51

Hi zaynah,

It appears the calculated values are infact loaded into paraview as they are in your calculation results. Check the U and p files in time directory 10 to go through the numbers and see if there are values of pressure calculated as you desire. If you don't see the expected values there, check the problem set up to see if you are missing anything. From the iteration results you have attached the calculation is not yet close to convergence seen from the nonidentical residual values.


zaynah04 January 9, 2013 11:11

i had make the iteration up to 1e-4..

to acheive convergence what mst i do?

Traib January 9, 2013 15:14

Firstly, you need to make sure that your case has been set up correctly. Then you can run the calculation to a time where the residual errors are acceptably small, which is depending on the tolerance values for variables you define in fvSolution dictionary for your case. However, like I said, first you need to make sure your case set up is correct. I also see that your maximum courant number has dropped to about 1e-4 which is very low, values around 0.5 should be good enough for your calculation. You can add 'runTimeModifiable' and 'adjustTimeStep' in your controlDict to automatically adjust the time steps to more appropriate values and see how the calculation works for a try.

zaynah04 January 9, 2013 15:32

i did as u suggested and my courant number remained low..:(
and same result..:(

zaynah04 January 10, 2013 02:00

hi i have been able to do it.
dear Traib and Bin xu

i had used the cell centres incorrectly its all fine..
thank you all for your help..


All times are GMT -4. The time now is 11:20.