I am calculating about some phenomena around submarine in CFD, but I have never found out the comparable experiment data with my calculation. If you ,CFD friends,have experiment data, can you share data? it is the secret of the military in our country.

you see. submarine is a very classified research area. if you really want to know sub. data. you need to describe your submarine's shape.

It is really a very "strange" question about getting experimental data, especially one is running the CFD calculation. It is like counting number of fingers while using an electronic calculator. Are you having problems with your CFD calculations? Are you using a commercial code ? or your own code ?

Dear Jong-Hyun,

Depending on specifically what phenomena you're interested in, you may not need a comprehensive experimental data around fully appended, complete (real) submarines.

If you're interested in appendage-body juncture flows, I believe there are a considerable volume of experimental data in the literature (ASME JFE data bank probably). If you're interested in high-incidence angle maneuvers and want to compute the flow structure on leeside of the body for instance, you may want to look for experimental data around elongated bodies such as spheroids. Finally, if you're interested in turbulent boundary layers past underwater bodies (idealized submarine hull form), I can share one with you.

Wing-body juncture flows, turbulent boundary layers, tip vortices emanating from appendages, and flow around elongated bodies at incidence, these are the phenomena that are all important in the hydrodynamics for submarines. They are building blocks of the flow you're interested in, and studying these flows individually often give us deeper insights to the flow.

CFD calculation for complete (fully-appended) submarines is still a challenge not only because of the sheer complexity of the physics and the lack of our understanding of how to model it (e.g. turbulence) but because of the mesh resolution required to resolve the details of the flow all at the same time. I've once used almost 2 million cells, but even with such a large mesh, the flow was still under-resolved by all accounts. And it wasn't fully appended.

Please let me know what you're specifically interested in.

validation

I think if your code has been validated for some basic test-cases (shapes, angles, curves, cylinder,...the more the better), you theorically do not have to worry so much about your calculation, isn't it?

jy