pitzdaily error

[Dcn303]

hi i ran the pitzdaily case then i get this error in the log.rhoPimpleFoam
i ran the case as

1. blockMesh (using normal openfoam)
2. (time rhoPimpleFoam) 2>&1 | tee openfoam log.rhoPimpleFoam (using the rapidCFD)
   then i got the error given below
   
   is this error comming because rhoPimpleFoam read an errorneous data ,?? if so from which file could it be comming from ?
   thanks

[TonkomoLLC]

Hi, The error comes from the residuals being "nan" instead of a real number.

I re-started the same case with my RapidCFD installation, and here is the screen output at a later time step:

Courant Number mean: 0.0471995 max: 0.289798
Time = 0.00886

PIMPLE: iteration 1
smoothSolver:  Solving for Ux, Initial residual = 0.00251807, Final residual = 3.23725e-06, No Iterations 3
smoothSolver:  Solving for Uy, Initial residual = 0.007397, Final residual = 9.32929e-06, No Iterations 3
smoothSolver:  Solving for e, Initial residual = 0.0269257, Final residual = 3.52643e-05, No Iterations 3
AINVPCG:  Solving for p, Initial residual = 0.133418, Final residual = 0.00106084, No Iterations 12
diagonal:  Solving for rho, Initial residual = 0, Final residual = 0, No Iterations 0
time step continuity errors : sum local = 6.39626e-07, global = 7.24221e-08, cumulative = -9.93557e-06
rho max/min : 1.17125 1.15959
PIMPLE: iteration 2
smoothSolver:  Solving for Ux, Initial residual = 7.5693e-05, Final residual = 9.41498e-07, No Iterations 2
smoothSolver:  Solving for Uy, Initial residual = 3.01746e-05, Final residual = 3.9318e-07, No Iterations 2
smoothSolver:  Solving for e, Initial residual = 0.0211342, Final residual = 2.82121e-05, No Iterations 3
AINVPCG:  Solving for p, Initial residual = 0.0597856, Final residual = 0.000467141, No Iterations 12
diagonal:  Solving for rho, Initial residual = 0, Final residual = 0, No Iterations 0
time step continuity errors : sum local = 2.73647e-07, global = -2.64104e-08, cumulative = -9.96198e-06
rho max/min : 1.17124 1.15959
PIMPLE: iteration 3
smoothSolver:  Solving for Ux, Initial residual = 9.11402e-06, Final residual = 1.22537e-07, No Iterations 2
smoothSolver:  Solving for Uy, Initial residual = 1.38894e-05, Final residual = 1.85428e-07, No Iterations 2
smoothSolver:  Solving for e, Initial residual = 0.00143433, Final residual = 2.39084e-07, No Iterations 4
AINVPCG:  Solving for p, Initial residual = 0.00421198, Final residual = 9.20821e-07, No Iterations 22
diagonal:  Solving for rho, Initial residual = 0, Final residual = 0, No Iterations 0
time step continuity errors : sum local = 5.42648e-10, global = -1.49144e-11, cumulative = -9.962e-06
rho max/min : 1.17124 1.15959
smoothSolver:  Solving for k, Initial residual = 0.000321624, Final residual = 3.98415e-07, No Iterations 3
bounding k, min: 0 max: 9.7763 average: 0.0932602
ExecutionTime = 154.35 s  ClockTime = 155 s

fieldAverage fieldAverage1 output:
    Calculating averages

I cannot reproduce the error that you are showing nor have I come across this error before with this case with RapidCFD.

I am using Ubuntu 20.04 and CUDA 11.2. the code was executed with a simple rhoPimpleFoam. RapidCFD was compiled using the instructions on issue number 93 from the RapidCFD-dev GitHub repo.

The most common mistake I have seen people make over the past few years of helping the RapidCFD community is that the sm_# is not set correctly in the wmake rules. The next most common mistake I see is that OpenFOAM was not compiled fresh from the given system state. For example, the CUDA version changed but RapidCFD was not recompiled from scratch and so on. If you have any doubts about your installation, do a completely new fresh clean install starting with git clone from the RapidCFD repo.

Also, I have just access to one GPU that is newer than sm_52 (i.e., old!) and I haven't had time to compile RapidCFD on this machine. Thus I cannot make any guarantees about how RapidCFD works on modern GPU's. I have no idea what may happen... Although there was a post in the Rapid CFD issues here of an A100 being used. Modern Gpu's require more recent versions of CUDA, but above 11.1 I know there are issues that start to crop up that I do not know the solution to (see issue92)

sorry for a long reply without direct help. I do not know the root cause of your issue so can only supply ideas to consider.

[Dcn303]

Hi can anyone provide a test case that uses the incompressible/simplefoam solver that runs on the rapidCFD
any link would be helpful
Thanks

[TonkomoLLC]

I do not have a test case handy. Can you modify the OpenFOAM 2.3.x pitzDaly tutorial for your purposes? It should be similar to the compressible pitzDaly example I have on the RapidCFD-Ttests repo.

[Dcn303]

sir please configure this test case that i have attached to work with the RapidCFD
i shall be very thankful to you
simpleFoam_testcase_one.zip

[TonkomoLLC]

simpleFoam_testcase_one.tar.gz
Test case returned; it runs with RapidCFD. Good luck!

[Dcn303]

thanks a lot sir it did run on RapidCFD , this below is my log.simpleFoam

but sir how do we display the

its not printing this information in my log.simpleFoam how do we display it sir

[TonkomoLLC]

Hi,
I apologize, but I do not follow your question.
If you want to create a log file, it should work if you type:
simpleFoam > log.simpleFoam &

Then you can use nano, gedit or any other editor to view the output.

If you use the redirection > then the screen output that I think you want with the residual information will be in the text file.

Hope this helps. If not, please try to re-explain your question.

[Dcn303]

yes sir i did
(time simpleFoam ) 2>&1 | tee log.simpleFoam
but yes it does not print the redisual information it only prints

so sir how do we print the residual information as well

[Dcn303]

as this sir

[TonkomoLLC]

Residuals OF 2.3.0.pdf

Please see the above procedure. The residual plotting procedure in OpenFOAM has changed substantially since OpenFOAM 2.3.x. This is not the modern way to plot residuals with current versions of OpenFOAM (E.g., v10 or v2212), but it should work with RapidCFD.

[Dcn303]

ok sir thank you so much i will check , sir if you don't mind can i get the pdf version of the manual that you are using
😅😅

[TonkomoLLC]

I am not sure I can comply with your request; this was paid training material. My apologies. A lot of this information can be found publicly. For example, another residual plotting tutorial for older versions of OpenFOAM is here.

[Dcn303]

ok sir i understand , thank you sir

[Dcn303]

sir can you tell me some good paid training material where i can also get trained and get such material , suggest me sir

[TonkomoLLC]

Please check out Wolf Dynamics

You can start with the introduction training that Joel posted here. There are slides, example cases, and videos to watch.

Other training sets cover more advanced topics. There are a ton of other options for OpenFOAM training. If you are able to pay, Joel offers in-person training.

I personally took the training offered by ESI, albeit back in 2014 when OpenFOAM 2.3.0 was the latest version. That's where the training pages I sent came from. Overall I thought ESI's course was really good and set me up for success with OpenFOAM.

However, RapidCFD is stuck on OpenFOAM 2.3.0, so any modern training class will be more centered on OpenFOAM 9, 10 or maybe 2106+ of the ESI class. Finding training relevant to OpenFOAM 2.3.0 takes a little looking. A lot of the same principles of newer versions of OpenFOAM apply to RapidCFD, though. Just some of the mechanics are different, like the way of plotting residuals in recent versions of OpenFOAM is just different from how residuals were plotted with OpenFOAM 2.3.0.

Hope this helps. Good luck!

[Dcn303]

its very helpful sir
thank you