[Cal]

Hi all,
I am new to openfoam and have been given code for my thesis to edit. It was written for openfoam v5 and I am running the code using openfoam v5. I am using the solver "sprayfoam" and after successful execution of the code, my solutions contain a lot of unreadable or unrecognized characters in the solution (see attached for examples)


Note: this issues is only present when there is a nonuniform list of scalars in the solution. The random characters shown in the picture attached continue for hundreds of lines before the code ends resulting in files from 700 kb - 2.2 mb.



Let me know if there are any other files you wish to see (controldict, 0 folder, etc), I haven't uploaded any because I have no idea what would be relevant to this issue


Thanks so much,


Cal

[Antimony]

Hi,

Is your writeFormat set to binary in the controlDict? If you want it to be human readable, then you should set the writeFormat to ascii.

Cheers,
Antimony

[Cal]

Hi Antimony,


Thanks for the reply, yes the format is set to ascii in the controldict do you know of any other possible reasons?


Thanks


Cal

[AndreasPe]

If you want to change the format of already existing files, you need to execute the "foamFormatConvert" utility.

[Cal]

Hi Andreas,


Thanks for the advice, I will try that on files I already have.

But I would like to create new solutions which aren't in this format, do you know hopw I would possibly do this?


or do you know what format my solutions are currently in?


Thanks

[AndreasPe]

If your controlDict says "ascii" as writeFormat, the solver should create only ascii format files. I tried the aachenBomb tutorial with "writeFormat ascii;" in the controldict and all files that were created are full ascii, so there should be no problem with sprayFoam there.

The header of the file you posted says "binary", so i would assume the writeFormat was set to binary (as it is originaly in the aachenBomb tutorial too).

[adhiraj]

Just throwing this out there: is there any particular reason you want to avoid using binary files for writing the solution fields?

Binary files seem to speed up my runs (anecdotal, did not do a benchmark), and take less space on disk. Also, when writing in ASCII some information is lost depending on the write precision, which can impact restarts.

If you need to look at a particular field, you can always post-process the binary files. For anyone working with OpenFOAM I would recommend using binary files for output.

[GerhardHolzinger]

A few points come to mind.

When to use ASCII write format in OpenFOAM?

• Trouble-shooting: ASCII has the great benefit of being human-readable
• Working with multiple tools: binary formats might get you into compatibility trouble, e.g. when importing the mesh into another tool to perform some operations on the mesh.
• Working with multiple OpenFOAM variants/versions: foam-extend might/will write a slightly different binary format than the Foundation release or the ESI/OpenCFD release

When not to use ASCII format in OpenFOAM?

• When you do actual computation runs
• When you do mesh generation, and there is no need to use ASCII

[Cal]

Hi all,


Thanks for all the help. I have been able to implement all your suggestions and have discovered that my issue isn't binary/ascii related after all.


Thanks so much for sharing your knowledge, it helped a lot!


Cal

[AdrianaPinto2]

Originally Posted by Cal

Hi all,


Thanks for all the help. I have been able to implement all your suggestions and have discovered that my issue isn't binary/ascii related after all.


Thanks so much for sharing your knowledge, it helped a lot!


Cal

Hello Cal, I know this is kinda of an old thread, but I have a similar problem and would like to know how you resolved yours.

Thank you for your attention.

[unilord]

Originally Posted by AdrianaPinto2

Hello Cal, I know this is kinda of an old thread, but I have a similar problem and would like to know how you resolved yours.

Thank you for your attention.

Hello Adriana and Cal. I am having a similar problem. For context, my university lab bought a desktop with a intel 9 13 serioes CPU, and openfoam, when using the decomposePar command, inserts unreadable characters in random files. For instance, sometimes it creates unreadable characters in the processorX/constant/ files (can be in faces, points, etc), sometimes it creates in the processorX/timestep/p,U,k,omega,nut files, and I can not reconstructPar. It is sad because we get the most out of the CPU when using all the cores, and not running in a single core.


If someone has any idea, please let me know. It literally writes the code in ascii but in some lines it inserts characters such as &, and unreadable characters as you can see in the attached image.