[cyrusIII]

I am totally new in CFD. I have worked in Solid mechanics and Material science so I have little bit of experience with FEM of Solids but not Fluids. In FEM software (at least the commercial ones) we use certain steps for example geometrical modeling then applying material then boundary conditions and forces and meshing and solving etc. Is it the same in CFD and FSI ?
What would be the diffreces of FSI and CFD? Is this forum good for FSI too? if not what are your recomendations?
Sorry for too many questions
Thank you

[FMDenaro]

Originally Posted by cyrusIII

I am totally new in CFD. I have worked in Solid mechanics and Material science so I have little bit of experience with FEM of Solids but not Fluids. In FEM software (at least the commercial ones) we use certain steps for example geometrical modeling then applying material then boundary conditions and forces and meshing and solving etc. Is it the same in CFD and FSI ?
What would be the diffreces of FSI and CFD? Is this forum good for FSI too? if not what are your recomendations?
Sorry for too many questions
Thank you

Yes, modelling a flow problem is based practically on the same steps and if you think about the elastic equations they are somehow resembling.
The main differences that I can say:
1) the relevance of the non-linear terms
2) the relevance of unstedy problems.

[cyrusIII]

Originally Posted by FMDenaro

Yes, modelling a flow problem is based practically on the same steps and if you think about the elastic equations they are somehow resembling.
The main differences that I can say:
1) the relevance of the non-linear terms
2) the relevance of unstedy problems.

Thank you , one of the reasons I asked that question is that in some solid dynamic simulations I have seen they start geometries with mesh rather than a solid. They basically start mesh as soon as they open the software and start modeling whereas in Abaqus we define type of element and then the geometry etc.
In navier stokes formula I remember we were canceling some terms for simplicity (depend on the problem type). How could you do that in CFD softwares especially the commercial ones to save some computational time?

[FMDenaro]

the mesh generation is a fundamental issue also in CFD and is the first step, you need to do... obviously the material can be fluid or solid (a wall for example).

To simplify the NS equations, the classical procedure is to perform a non-dimensional analysis of the various terms. If it appears that some term is less relevant (looking at the corresponding non-dimensional number), then you can simplify that term. For example, small Mach number suggest to use the incompressible flow model.
Unfortunately, commercial softwares work with dimensional equations.

[cyrusIII]

Originally Posted by FMDenaro

the mesh generation is a fundamental issue also in CFD and is the first step, you need to do... obviously the material can be fluid or solid (a wall for example).

To simplify the NS equations, the classical procedure is to perform a non-dimensional analysis of the various terms. If it appears that some term is less relevant (looking at the corresponding non-dimensional number), then you can simplify that term. For example, small Mach number suggest to use the incompressible flow model.
Unfortunately, commercial softwares work with dimensional equations.

Thank you . Just to make sure, So basically you mean in commercial software it considers/calculate the whole terms of NS equation even if some of the terms are negligible? Is that what you mean by " commercial softwares work with dimensional equations"

[FMDenaro]

Originally Posted by cyrusIII

Thank you . Just to make sure, So basically you mean in commercial software it considers/calculate the whole terms of NS equation even if some of the terms are negligible? Is that what you mean by " commercial softwares work with dimensional equations"

you can find specific codes for simplified problems...for example, if you assume no-viscosity you can get softwares that solves the Euler equations instead of the Navier-Stokes equations...

[cyrusIII]

Originally Posted by FMDenaro

you can find specific codes for simplified problems...for example, if you assume no-viscosity you can get softwares that solves the Euler equations instead of the Navier-Stokes equations...

Where do you usually find the reliable codes? Are they available in this forum or there are specific websites?
P.S these codes are usually for free software not commercial, right?

[FMDenaro]

Originally Posted by cyrusIII

Where do you usually find the reliable codes? Are they available in this forum or there are specific websites?
P.S these codes are usually for free software not commercial, right?

Many softwares are offered.... just a couple of examples:
- OpenFOAM is a free software you can use
- Fluent is a commercial and expensive code

[cyrusIII]

Originally Posted by FMDenaro

Many softwares are offered.... just a couple of examples:
- OpenFOAM is a free software you can use
- Fluent is a commercial and expensive code

So within these softwares there are options for users that they can choose which equation or what type of code or solver (Euler, NS , compressible, heat etc) they can use? What is your personal preference on the software? (I know it depends on the type of problem but I would glad to hear peoples experiences and opinions)

[FMDenaro]

I had experience using both OF and Fluent, but I spent the most of my work using own-made codes.

[cyrusIII]

Originally Posted by FMDenaro

I had experience using both OF and Fluent, but I spent the most of my work using own-made codes.

Wow, would please tell me what text books are good (or other resources with hands on examples) if I want to start writing my own codes? Do you write and run your codes in Fortran or MATLAB or it is importable or executable by those software you mentioned?