Is it right when two dimension of a 3 dimensional analysis is chosen? Does it always mean the gradient of variation in the omitted dimension is negligible in comparison to other criterion that may be applied?
Yes what you said is exactly true, but remembering that in most of engineering application the 3D approach should be considered. Worth mentioning that even in laminar boundary layer analysis the 3D approach is needed, eg, Taylor vortices and ...
Professor you've been my teacher in fluid mechanics 1 and 2. I remember, the sessions full of informative sentences. I always wanted to record your voice.
Any flow, in general, is unsteady and non-uniform As you know, the unsteady flow means, the flow parameters (velocity, pressure, mass-flow rate, viscosity, surface tension, etc. ) at any location are variable with time. A steady flow means the flow parameters do not vary with time. Also, the flow parameters vary in spatial planes ( x, y and z directions). In practice, we consider the flow to be steady and uniform. That is, the time-dependence at any plane is neglected and also the variation in one plane is considered negligible .
But , in steady-flow cases too, the flow parameters could be varying in all the three spatial directions, x,y and z, simultaneously, as in turbomachines .In such cases, if we form two dimensional grids of x-y and x-z planes ( for example), with the flow parameters in the x-direction being common, we can solve the problem in all the three-dimensional planes simultaneously. Then, it will not be necessary to make presumption of variations in the third plane being negligible.
it is right that Any flow, in general, is unsteady and non-uniform, but relatively we can negligee one dimension. Any way it depends on the physical application.