Hi Dear Friends,

The question is this one:

Why one needs to use turbulence modelling?

Best Regards

VG Ferreira

Practicality.

Try doing DNS for high Reynolds number flow over a complete aircraft.

Dear PC,

I am in the same positon

Ferreira

My response was somewhat tongue-in-cheek.

Seriously, though: these models are used to represent the effects of what we cannot resolve. You introduce terms for the creation/destruction/transport of turbulent kinetic energy, rather than compute the fluctuations at all scales directly and pull out the statistics from a long, expensive transient run. Modeling is currently the only practical option for most real problems and geometries of interest.

Many of these models do rather well for various problem classes, but of course they're not the whole solution. Many a degree has been awarded to people exploring turbulence modeling...

Dear PC,

I am simulating a 3D problem using my code, and it is given segmentation fault. Is it possible to avoid this interruption by implementing the K and Epsilon equations?

Ferreira

A segmentation fault is caused by an error in programming - a divide by 0 or going beyond memory limits. It has nothing whatsoever to do with implementing a turbulence model. You have a programming error - not a physics error.

Hi, If you are getting segmentation fault, I would suggest you to check the code throughly and try handle all the exceptions with proper error message so that each time you can know where your code is blowing.

Some times wrong physics my spoil the quality of solution matrix which may cause segmentation fault. Modeling turbulence would enhance the dissipation in your code, this will try to smoothout some sharp gradient which can give stability to your solver.