sympy -> physics -> continuum mechanics -> ?fluids?

[Rahul Manavalan]

Hello All 

Would be realistic at all, to extend the CM library to include fluids as well. 

1. The obvious candidate would be to write a viscous NS equation solver to begin with. I couldn't help wonder why it was not included in the library in the first place. 

Could there have been a design decision made to not include it in the applications library or was it simply oversight. I would really like to know if such a thing would be welcome to the library.  

Any advice on this matter is highly appreciated. 

With best regards 

Rahul 

[Oscar Benjamin]

On Sun, 21 Mar 2021 at 20:53, Rahul Manavalan
<admissions.ra...@gmail.com> wrote:
>
> Hello All
>
> Would be realistic at all, to extend the CM library to include fluids as well.
>
> 1. The obvious candidate would be to write a viscous NS equation solver to begin with.

Don't assume that others will understand what you mean with these
abbreviations. By CM you mean Continuum Mechanics and by NS I am
guessing you mean the Navier-Stokes equation.

> I couldn't help wonder why it was not included in the library in the first place.
> Could there have been a design decision made to not include it in the applications library or was it simply oversight. I would really like to know if such a thing would be welcome to the library.

SymPy is a symbolic library and should only include topics where some
significant symbolic computation can be done. When you refer to a
viscous NS solver do you mean some kind of symbolic algorithm or a
numerical solver?

Personally I think that the bar to extending the feature set of sympy
by adding a new module should be high. A strong case would need to be
made for why this should be in sympy. You say you couldn't help wonder
why it wasn't included. I've never wondered why there was no solver
for the viscous Navier Stokes equation in sympy. Apparently to you it
seems like an obvious addition but it doesn't to me. If you think that
it is something that sympy should have then you need to give more
explanation and justification.

Oscar

[Rahul Manavalan]

I apologize, Oscar, for not making a better case for the statements I made. 

I believe that such an extension could be useful to those engineers who are looking to secure a ball park figure for the flow parameters for simple geometries without resorting to expensive mesh based solvers. 

Usually (atleast where I used to work) the design pipeline for a product is tedious and more often than not people are forced to make design decisions (for flow geometries) based on empirical data. This results in numerous design iterations that are not warranted. 

This is something I had in mind when I wrote the previous message. I think it may have other didactic uses as well.

With best regards

Rahul

[Oscar Benjamin]

On Sun, 21 Mar 2021 at 21:31, Rahul Manavalan
<admissions.ra...@gmail.com> wrote:
>
> I believe that such an extension could be useful to those engineers who are looking to secure a ball park figure for the flow parameters for simple geometries without resorting to expensive mesh based solvers.
> Usually (atleast where I used to work) the design pipeline for a product is tedious and more often than not people are forced to make design decisions (for flow geometries) based on empirical data. This results in numerous design iterations that are not warranted.
>
> This is something I had in mind when I wrote the previous message. I think it may have other didactic uses as well.

Can you give more detail and also some examples of what this
Navier-Stokes solver would do?

So far I have no clear idea what it is you are referring to. I presume
that you want to be able to use sympy to develop analytic
approximations of something but I'm not sure what sort of thing you
want to approximate or what type of approximation you are referring
to.


Oscar