[sudeshpowar]

I Sudesh Powar working as assistant professor in Mechanical Department at MIT Academy of Engineering, Alandi. Currently we associated with university of pune but now we plan to go for autonomous institute. For that we want to design a syllabus structure for Computational Fluid Dynamics (CFD). We came to know that many expert in this area are visiting this website. So it's great pleasure for us if you guide us. Some how we have formulated the syllabus of CFD but your views on present syllabus structure are beneficial for our institute. So we are requested to give your views in detail on present syllabus structure so that we can modify as per industrial requirements.
We hope for your positive reply.

[hariswch2]

Originally Posted by sudeshpowar

I Sudesh Powar working as assistant professor in Mechanical Department at MIT Academy of Engineering, Alandi. Currently we associated with university of pune but now we plan to go for autonomous institute. For that we want to design a syllabus structure for Computational Fluid Dynamics (CFD). We came to know that many expert in this area are visiting this website. So it's great pleasure for us if you guide us. Some how we have formulated the syllabus of CFD but your views on present syllabus structure are beneficial for our institute. So we are requested to give your views in detail on present syllabus structure so that we can modify as per industrial requirements.
We hope for your positive reply.

Above mentioned syllabus is good enough for getting basic knowledge in CFD for new students

[LuckyTran]

This seems to be an introductory course on the finite volume method with applications. I think in general it is a good layout but I suggest to:

Maybe drop finite element method or only discuss it in limited scope, it's not really needed to discuss finite element in FVM.

In the introduction, along with classification of pde's, it would be useful to also discuss initial value problems, boundary value problems, etc.

I tend to think that 3. numerical grid generation is inappropriate for an introductory course. Under-the-hood knowledge of grid generation doesn't really facilitate a better understanding of FVM or CFD. However, the qualities that make great grids great are definitely useful. So I would instead cover topics like grid requirements, grid quality, and grid convergence or merge this topic with 2. Discretization. This is just my opinion, but grid generation is clearly listed as a program objective.

I also recommend inclusion of heat/diffusion equations, advection equation, and advection-diffusion equations, as these are natural precursors of the transport equation and they are excellent for demonstrating discretization schemes in FVM. On the topic of spatial discretization, both spatial and temporal discretization could be discussed.

One very important topic that I do not see in the syllabus is the solution of the equations. Maybe it is included implicitly but I think it should be clearer in the syllabus. For example: what is the purpose of discussing all these discretization schemes? Is it to discuss CFD for the sake of discussing CFD or is it to solve actual problems? Why was discretization needed in the first place and what is the result of discretizing everything? That should be integral part of the entire course. At some point, the structure of the system of equations that needs to be solved should be discussed, which thankfully is a very sparse system. Otherwise there is no benefit as the FVM technique wouldn't scale well to industrial problems because of computational expense. In other words, what is different (mechanically) between solving the discretizied FVM equations numerically versus solving the original pde's themselves analytically?

[sudeshpowar]

Thank you for your valuable reply...It is very helpful