[lamboram]

Hello all,

this is the code I used for my source term heat generation. When I try to interpret it in Fluent. I get Parse error (line 3). But then I simulated it, the results are sensible and converged. What could be possibly wrong here.

#include "udf.h"

DEFINE_SOURCE(heat_source,cell,thread,dS,eqn)
{
real source;
real time;

source = 0.0;
time = CURRENT_TIME;

if (time >= 0 && time <= 10)
source = 129639.04;
else if (time >= 11 && time <= 50)
source = 56939.5;
else if (time >= 51 && time <= 1140)
source = 21606.5;
else if (time > 1140)
source = 14234.87;
else
source = 0.0;

return source;

}

Cheers,
Ram

[Bruno Machado]

If possible avoid the use of interpreted UDFs. The compiled UDF is more "safe".

[brunoc]

Originally Posted by Bruno Machado

If possible avoid the use of interpreted UDFs. The compiled UDF is more "safe".

Any example of when compiling a UDF is "safer" then using an interpreted version?

[Bruno Machado]

16 Interpreted vs. Compiled UDF’s
Functions can either be read in and interpreted at run time or compiled and grouped into a shared library that is linked with the standard FLUENT executable.

• Interpreted vs. compiled code
o Interpreter
 Interpreter is a large program that sits in the computer’s memory.
 Executes code on a “line by line” basis instantaneously
 Advantages – does not need a separate compiler
 Disadvantage – slow and takes up space in memory

o Compiler (refer to User’s Guide for instructions)
 Code is translated “once” into machine language (object modules).
 Efficient way to run UDF’s. Uses Makefiles
 Creates “shared libraries” which are linked with the rest of the solver
 Overcomes interpreter limitations e.g. mixed mode arithmetic, structure references etc.

Got this from internet (author: Saeed Sadeghi).

[brunoc]

Ok, but those just say that compiling a UDF is more efficient, both in speed and memory usage. There is nothing there about it being "safer".

Also, some of the points are a bit misleading. For instance, compiled UDFs also take up computer memory once they're loaded. And interpreted UDFs also create code that is shared with the solver.

Compiled UDFs are overall better, but interpreted are great for quick tests or when you don't have a compiler installed. One very good thing about interpreted UDFs in the past was that if wanted your code ready to run on every machine you had to re-compile it several times: win/linux, 32/64 bits, 2d/3d, single/double precision, serial/parallel. But with 16.0 compiling is as easy (and free!) in Windows as it is in Linux. Fluent will also auto-compile your UDF as needed, so no more of those multiple compilers.

Cheers.