CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   CFX (http://www.cfd-online.com/Forums/cfx/)
-   -   The formula to compute the mass flow averaged value in a surface (http://www.cfd-online.com/Forums/cfx/122278-formula-compute-mass-flow-averaged-value-surface.html)

Anna Tian August 16, 2013 06:23

The formula to compute the mass flow averaged value in a surface
 
Hi

May I ask where could I find the formula that CFX-post use to compute the mass flow averaged value in a surface? I checked the CFX tutorial but didn't find it. If tutorial does include this formula, please let me which tutorial and which page. Thanks.

RicochetJ August 16, 2013 06:35

Someone correct me if I'm wrong, but isn't just areaAvg(massflow)@location?

oj.bulmer August 16, 2013 07:33

Mass weighted average of quantity \phi on surface: \frac{\int \phi \rho \phi dA}{\int \rho \phi dA} = massFlowAve(\phi)@<Surface name>

Area weighted average of quantity \phi on surface: \frac{1}{A} \int \phi dA=areaAve(\phi)@<Surface name>

OJ

Anna Tian August 17, 2013 06:08

Quote:

Originally Posted by oj.bulmer (Post 446083)
Mass weighted average of quantity \phi on surface: \frac{\int \phi \rho \phi dA}{\int \rho \phi dA} = massFlowAve(\phi)@<Surface name>

Area weighted average of quantity \phi on surface: \frac{1}{A} \int \phi dA=areaAve(\phi)@<Surface name>

OJ

Are you sure about the mass flow averaging formula? There is no velocity term in it?

Where did you find it? Any CFX tutorial?

RicochetJ August 17, 2013 06:37

..........

Anna Tian August 17, 2013 09:32

Quote:

Originally Posted by RicochetJ (Post 446267)
..........

I'm asking about the math averaging algorithm that CFX use to compute. Not the command.

oj.bulmer August 17, 2013 09:54

Quote:

Are you sure about the mass flow averaging formula? There is no velocity term in it?
Touche! That is what I would call lost in translation (in LaTex) :)

Please read the second \phi in the formula as v. Thus, more formally, the formula is


\frac{\int \phi \rho | \overline{v}.\overline{dA}|}{\int \rho | \overline{v}.\overline{dA}|}=\frac{\sum \phi m }{\sum m}

where \overline{v}, \; \rho, \;  \overline{dA} are the local velocity vector, local density and the area vector at facet of the local cell respectively.

OJ

Anna Tian August 17, 2013 09:59

Quote:

Originally Posted by oj.bulmer (Post 446296)
Touche! That is what I would call lost in translation (in LaTex) :)

Please read the second \phi in the formula as v. Thus, more formally, the formula is


\frac{\int \phi \rho | \overline{v}.\overline{dA}|}{\int \rho | \overline{v}.\overline{dA}|}=\frac{\sum \phi m }{\sum m}

where \overline{v}, \; \rho, \;  \overline{dA} are the local velocity vector, local density and the area vector at facet of the local cell respectively.

OJ


Where did you find this formula? How do you know CFX use this formula?

oj.bulmer August 17, 2013 10:09

Have a look at documentation, stitch the information together and apply judgement...

CFX:
http://www.sharcnet.ca/Software/Flue.../i1308570.html

FLUENT:
http://www.sharcnet.ca/Software/Flue...th/node416.htm

OJ

Anna Tian September 1, 2013 10:11

Quote:

Originally Posted by oj.bulmer (Post 446298)
Have a look at documentation, stitch the information together and apply judgement...

CFX:
http://www.sharcnet.ca/Software/Flue.../i1308570.html

FLUENT:
http://www.sharcnet.ca/Software/Flue...th/node416.htm

OJ

By the way, how does CFX calculate the volume of the fluid domain? Just sum up all the cells size in the fluid domain?

ghorrocks September 1, 2013 19:39

Yes, sum up the cell volumes.

Anna Tian September 3, 2013 08:05

Quote:

Originally Posted by ghorrocks (Post 449261)
Yes, sum up the cell volumes.

How do you know that? Any reference?

ghorrocks September 3, 2013 20:08

Isn't it obvious? If you break up a volume into zillions of little volumes then the sum of those little volumes is the total volume. It is hardly rocket science.

Anna Tian September 6, 2013 14:43

Quote:

Originally Posted by ghorrocks (Post 449686)
Isn't it obvious? If you break up a volume into zillions of little volumes then the sum of those little volumes is the total volume. It is hardly rocket science.


Yes. It is obvious that we can compute it in this way. But it is not obvious that CFX choose this way to compute the volume. Any CFX reference about it?

singer1812 September 6, 2013 18:06

If you are worried about how CFX calculates the total volume and need a reference for it, then I think you have your work cut out for you.

You are heading towards questioning every single equation, which is quite a task.

Are you getting a huge discrepency in the CFX calculated volume as opposed to the CAD? If so, I would question your mesh resolution first prior to blaming the software.

RicochetJ September 6, 2013 20:07

Quote:

Originally Posted by singer181250281
If you are worried about how CFX calculates the total volume and need a reference for it, then I think you have your work cut out for you.

You are heading towards questioning every single equation, which is quite a task.

Are you getting a huge discrepency in the CFX calculated volume as opposed to the CAD? If so, I would question your mesh resolution first prior to blaming the software.

Absolutely this - 100%. Most cfd issues I have witnessed have been either:
1. Mesh issues
2. A lack of understanding
3. A combination of 1 and 2

mat_cfd September 7, 2013 06:03

Remember only mesh goes into solver not real geometry. It must be cell volume or calculation through cell edges but if ur mesh is close to geometry solver volume will match the real geometry.
Anyhow domain volume in not used in discretized eqn so it wont effect CFX solution. But If you have compromised on geometry to improve mesh quality then you should use volume from preprocessing for calculation or expresion in CFX

Anna Tian September 7, 2013 06:17

Quote:

Originally Posted by mat_cfd (Post 450331)
Remember only mesh goes into solver not real geometry. It must be cell volume or calculation through cell edges but if ur mesh is close to geometry solver volume will match the real geometry.
Anyhow domain volume in not used in discretized eqn so it wont effect CFX solution. But If you have compromised on geometry to improve mesh quality then you should use volume from preprocessing for calculation or expresion in CFX

I ask for the reference because I use CFX-post to compute volume and I need to refer it and put the volume calculating formula on my thesis. Where could I find it?

And what is the formula to compute the surface averaged heat flux?

Anna Tian September 7, 2013 06:37

Quote:

Originally Posted by oj.bulmer (Post 446298)
Have a look at documentation, stitch the information together and apply judgement...

CFX:
http://www.sharcnet.ca/Software/Flue.../i1308570.html

FLUENT:
http://www.sharcnet.ca/Software/Flue...th/node416.htm

OJ

These links do not work now. Where could I find the documentation?

RicochetJ September 7, 2013 12:40

Quote:

Originally Posted by Anna Tian (Post 450335)
These links do not work now. Where could I find the documentation?

In the Ansys theory or solver modelling guide !


All times are GMT -4. The time now is 01:21.