[ghost82]

Hello,
I have some code to compile, which is not written by me.
The code compiled good in fluent 17.2, not with the newest 18.0.

I have error in this line:
init_tracked_particle(tp, pp, dpm_par.unsteady_tracking, FALSE, FALSE);

The error when compiling is too few arguments.

I cannot find any usefull info about init_tracked_particle.

Can anybody point me in the right direction?

Thank you

[Bruno Machado]

Originally Posted by ghost82

Hello,
I have some code to compile, which is not written by me.
The code compiled good in fluent 17.2, not with the newest 18.0.

I have error in this line:
init_tracked_particle(tp, pp, dpm_par.unsteady_tracking, FALSE, FALSE);

The error when compiling is too few arguments.

I cannot find any usefull info about init_tracked_particle.

Can anybody point me in the right direction?

Thank you

My best (and only) advise is have a look in the patch notes from version 17.2 to 18.0. Maybe there is a information related to this.

[ghost82]

Thank you Bruno, I can see there are several changes in the dpm calculations, however no specific information about the init function.
Thank you anyaway!

[rw511]

Did you ever find the reason behind this? I've just discovered the exact same issue.

The odd thing is my UDF still compiles fine for Fluent 17.2 on both Windows and Linux, it's OK for fluent 18.0 on Windows, it's only on Linux with Fluent 18.0 that I get the init_tracked_particle error.

Richard

[ghost82]

No, but it seems in the newest fluent 19.0 init_tracked_particle is back to 5 arguments...

Did you find the meaning (or some documentation) of these arguments?

[obscureed]

Hi ghost 82,

In any installation, Windows or Linux, you will be provided with header files containing macro definitions and prototypes for the available built-in functions. In Windows, they are typically somewhere like "C:\Program Files\ANSYS Inc\v180\fluent\fluent18.0.0\src" (then subdirectories such as "dpm"). So, you can grep for the function you want. Sometimes the prototype will show you only the types of the arguments, but in this case the developers have done the kind and rational thing and supplied typical variable names as well: for 18.0 on Windows:

FLUENT_EXPORT void init_tracked_particle(Tracked_Particle *tp, Particle *p, cxboolean unsteady_tracking,int tracking_migrants, cxboolean update, cxboolean need_cphase);

So it looks like the new argument comes before the final ",FALSE,FALSE)"; it's an integer; and it's something to do with migrants (probably from partition to partition in parallel).

Now comes the guesswork. Is it a numerical parameter ("maximum number of transitions per particle")? Or is it a selection ("0 = default model, 1 = new model, 2 = bad model")? I have no idea.

This is the beginning and the end of the available "documentation" at this level of UDF programming. (You can search the ANSYS Customer Portal -- I did in this case, and was not surprised when I found nothing.) If ANSYS have a justification for this very low level of documentation, I guess it might be that the expense of maintaining full documentation would be considerable, and would not be of any value to most users. However, I think they might (and certainly should) give you an answer if you send in a support query. If you could share the answer here, that would be helpful. You might as well ask about 18.x and 19.x simultaneously.

I don't understand why Windows and Linux would be different in rw511's case (18.0, in 2017), unless one of the compilers is dangerously flexible.

Cheers,
Ed

[rw511]

I never did find much information on this but I've just got new errors with too many arguments in my call now, so as you say Bruno, back to five we go.

If I find anything, I'll let you know.

[rw511]

Hi All,

I was able to get the following information about init_tracked_particle:

the parameters for init_tracked_particle are as follows:

Tracked_Particle *tp, /* tracked particle struct to be initialized */
Particle *pp, /* pointer to particle which should be used to fill the tracked particle */
cxboolean unsteady_tracking, /* is the simulation using unsteady particle tracking? */
cxboolean update, /* shall the particle positions be updated as at the end of a flow time step? */
cxboolean need_cphase, /* shall the continuous phase property information filled to compute drag, etc? */

And the parameter that appeared in 18.0 was removed in 18.1 related to parallel tracking.

Hope it's of some use.