CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   OpenFOAM Running, Solving & CFD (http://www.cfd-online.com/Forums/openfoam-solving/)
-   -   nonNewtonianFluidFoam - no grad(nu) & grad(U) (http://www.cfd-online.com/Forums/openfoam-solving/109909-nonnewtonianfluidfoam-no-grad-nu-grad-u.html)

chilledkroete November 29, 2012 11:07

nonNewtonianFluidFoam - no grad(nu) & grad(U)
 
Hello folks,

does anybody know why in this solver no grad(nu) & grad (U) is taken into account in Ueqn?

In interFoam one finds
fvVectorMatrix UEqn
(
fvm::ddt(rho, U)
+ fvm::div(rhoPhi, U)
- fvm::laplacian(muEff, U)
- (fvc::grad(U) & fvc::grad(muEff))
);
due to the fact that muEff is variable in space.

In nonNewtonianIcoFoam
- (fvc::grad(U) & fvc::grad(muEff))
is ignored.

Thanks for replies.

JonW November 29, 2012 11:49

Answer: continuum mechanics
 
Hello there

the additional term "- (fvc::grad(U) & fvc::grad(muEff))" is correct and is now present in nonNewtoninanIcoFoam in the OpenFOAM 2.1.x. You can also see this term in the interFoam solver.

If you are only used to standard fluid mechanics, you will be puzzled by this term. If you want to understand the physical reason for this term, you have to go into continuum mechanics. My favourite book about the subject is...

G. E. Mase. Schaums Outline Series: Theory and Problems of Continuum Mechanics. McGraw-Hill Inc., USA, 1970.


Hope this helps
J.

chilledkroete November 29, 2012 12:16

Hello Jon,

thank you for your answer. I just wondered why this term was missing in the solver I used so far (v.1.6-ext).

I actually thought it was missing.

I recently found out that you reported this bug in April.

Thanks a lot,
janto

JonW November 29, 2012 13:30

sorry, apparantly I was reading your post too fast,...

good luck,...
J.

sharonyue August 26, 2013 20:01

Excuse me, where can I download this book?

Theory and Problems of Continuum Mechanics

JonW August 27, 2013 12:00

You have to buy the book (I dont know of any download). But it is cheep. :)
See for example:

http://www.amazon.com/Schaums-Outlin...mechanics+mase

Quote:

Originally Posted by sharonyue (Post 448212)
Excuse me, where can I download this book?

Theory and Problems of Continuum Mechanics


sharonyue September 20, 2013 19:50

Quote:

Originally Posted by JonW (Post 448339)
You have to buy the book (I dont know of any download). But it is cheep. :)
See for example:

http://www.amazon.com/Schaums-Outlin...mechanics+mase

Sorry, so late.

This book is not on sell in our country. Do you have another articles? Thanks.

JonW September 21, 2013 13:15

Quote:

Originally Posted by sharonyue (Post 452826)
Sorry, so late.

This book is not on sell in our country. Do you have another articles? Thanks.

Ohh, thatís bad.
There is another book I use :
L. E. Malvern. Introduction to the Mechanics of Continuous Medium. Prentice-Hall, Inc., USA, 1969.

...but that book is much more expensive :(

The only thing I can suggest if you have to download is search by using the string: "continuum mechanics pdf"
hope this help

Cheers
Jon

sharonyue September 21, 2013 21:48

Quote:

Originally Posted by JonW (Post 452897)
Ohh, thatís bad.
There is another book I use :
L. E. Malvern. Introduction to the Mechanics of Continuous Medium. Prentice-Hall, Inc., USA, 1969.

...but that book is much more expensive :(

The only thing I can suggest if you have to download is search by using the string: "continuum mechanics pdf"
hope this help

Cheers
Jon

Okay...Thanks, but this book is too long. My primary research is not this. So Jon, do you have any notes by yourself?

sharonyue November 4, 2013 09:09

I still cant get out why there is (fvc::grad(U) & fvc::gradfluid.nu()). It exists in nonnewtonianIcoFoam and interFoam. Who can give me a hint pls?

chilledkroete November 4, 2013 11:35

1 Attachment(s)
Hi,

maybe this picture helps: Attachment 26578

The first term is equivalent with eta*laplace(v).

Regards,
Janto

sharonyue November 4, 2013 11:54

2 Attachment(s)
Quote:

Originally Posted by chilledkroete (Post 460483)
Hi,

maybe this picture helps: Attachment 26578

The first term is equivalent with eta*laplace(v).

Regards,
Janto

Hi Janto,
You mean
[LaTeX Error: Syntax error]
Why shouldnt it be:
[LaTeX Error: Syntax error]

chilledkroete November 4, 2013 14:00

Hello Forrest,

you would be right if the viscosity was constant all over the domain. However, in interFoam and nonNewtonianIcoFoam the viscosity is a field variable (have a look in createFields where the viscosity is defined) and depends therefore on the spatial coordinates. Owing to that the gradient of the viscosity does not vanish.

JonW November 4, 2013 14:07

1 Attachment(s)
Hello Sharonyue

Here is something I did in haste (there might be some typo errors, like vector arrow symbol missing over the velocity U), but basically this is the approach.
You would not use array approach as you showed earlier, but rather directly work with dyads and vectors directly. Such is much easier. I hope you can use this document, but you need to know continuum mechanics,... there is no way around it, if you are going to do serious CFD work.

Cheers and good luck
Jon

sharonyue November 5, 2013 00:43

NEAT!!!
To Janto
Quote:

Originally Posted by chilledkroete (Post 460506)
Owing to that the gradient of the viscosity does not vanish.

Yeah, I think this is where I was wrong.

To Jon:

Wonderful, It helps me a lot!!

Thank you both very much!


All times are GMT -4. The time now is 13:43.