Why are CFD programs called 'codes?'

[phonix]

I have a CS degree and have been programming in C/unix since 1986. I have never seen anyone use the word 'codes' until I started playing with CFD. Why do they use that term? How is a 'code' different from a 'program?' Is it a Fortran term from the 60s? (I never did Fortran).

 

[btrueblood]

Hm. Dunno. I've heard the term "source code" used for everything from Fortran and C/C++/Java to PLC and microcontroller assembly language files.

 

[SylvestreW]

I'd hazzard a guess it has to do with the solver being used. There are many different ways to approximate the PDEs, each with it's own associated error and best application etc.
Each particular way wasn't really a program (as it's more of a mathematical implementation). I guess that's why each is called a "code" as it's not a program per se.

 

[phonix]

We call those algorithms.

 

[swampyankee]

I suspect that most CFD programs were, at least originally, written by engineers. Many were distributed as source code (and almost always written in Fortran, because the Fortran community has always placed a great deal of importance on code portability).

Solving PDE's is non-trivial; different conditions require different solution algorithms, which may be implemented in quite different ways.