I have proposal for making CFD more interesting when haveing to make lots of repetitive simulations and pass the time. An in build Sudoku solver attached to code. I am thinking of this in C++ - has anyone tried this - will the Sudoku solvers numeric overhead impact upon the CFD solver - or should I write it in parallel and use one chip for CFD and one for Sudoku?

Do yopu not need more in your day to do. You should think not only CFD but is the answer real. If you are a GUI monkey who just sets off runs and loafs about tht is disgrace to the porffesion. Comnputers work overnight - we work what 6 - 7 hrs day so there planty of time to think is what you do right and see if all works OK.

If that bored why notr do wank and be done with, rahther than clutter up PC with paperwork games that could otherwise use CFD on processor?

I agree with you Rambo, CFD needs motivation.

I agree, but the motivation should be the solution of a problem or creating a new innovation for that is what engineering is all about. For me it's all in understanding the physics of the problem.

I've been trying to understand how we might adapt the Latin Hper-Cube sampling technique from design of experiments to solve Suduko puzzles - perhaps theres a thesis in it ?

so is a GUI monkey someone who actually uses CFD to solve usefull industrial problems as opposed to accademic monkeys who like to solve flow over a flat plate to fantastic orders of magnitude whilst not actually achieving anything usefull?

I enjoy being GUI-Monkey, it gives me lot of time for myself. I have even watched movies while my work is been done.

What movies? King Kong?