scaling the mesh and model validation
 

If you validate your model by comparing to experimental data. Then you make a model which is 2 times the original model, is this ok to use a mesh two times the original mesh?
obviously you don't have any experimental results for a bigger model.

Did you ever hear "Reynolds number"? It might be okay, but it also might be not okay. It depends on your geometry, on the Reynolds number, on the mesh resolution, on the Courant number etc...

That's like you would say, a glass of an alcoholic drink is okay, is a glass with twice the size also okay? That depends not just on the size ratio of the glasses, but also on the initial size of the glass (2x 0.02l is different to 2x 1l), the drink (whiskey is different to beer), your constitution (60kg body mass or 260kg?), the time (are you having this two glasses at the same time or one at an evening?) etc...

I hope, you got the point. It is impossible to say whether it's okay or not, since nobody here is a visionary and can use any magic methods to see your problem. It MIGHT be okay, but I will NEVER give you any guarantee!

I think what Mehrdad might be implying is if CFD can capture a scale effect, e.g., the difference between simulating a toy boat and a full-scale tanker. The short answer is that it cannot. So, if your nondimensional flow parameters, like Reynolds number, are consistent, then the solution should be, as well.

So, yes, it is okay to double the model size as long as the flow constraints are identical.