Use of (conservative) velocities at the wall in CEL
 

Hello everybody,

I am trying to set a heat transfer coefficient which is a function of the tangential velocities at the wall.

In order to try on an easy example, i made a simple duct with a given temperature at the wall and a heat transfer coefficient given by

100[W/m^2/K] + Velocity u * 10 [W/m^3/K * s]

The fluid is water with a velocity (Velocity u) of 1 [m/s]

When I postprocess the results I see that the HTC is equal to 100 [W/m^2/K] everywhere although Velocity u (conservative) isn't. I've read everywhere that CEL expressions make use of conservative values. Is this wrong or am I missing anything?

Thank's in advance!

You said you set up an expression for HTC, and I assume you meant an external heat transfer coefficient since there is no parameter name Heat Transfer Coefficient for the internal (as in heat transfer between the fluid and the wall). Correct ?

You also indicated the results for the "External" HTC on the boundary remained 100 [W m^-2 K] ; therefore, you assume the code used a Velocity u of 0 [m s^-1] in your expression instead of the velocity associated to control volume at the boundary. correct ?

If both of the above are correct, the CEL is using the hybrid value of Velocity u, i.e. 0 [m s^-1] for a stationary wall, and not the "conservative value" you seem to be expecting.

Do you need the expression to the conservative value instead ?

You are right. I'm talking about the Wall External Heat Transfer Coefficient.

It is also right that the code uses a Velocity u of 0 [m s^-1].

I was expecting the code to use the conservative value. I'd be very thankful if anybody would provide me the expression.

The default behavior of CEL at boundaries may be open to debate. For example, if I set a boundary to a fixed velocity of 0 [m s^-1] and temperature of 300 [K], I would expect that expression evaluated at such boundary use those numbers.

However, I may also want to access the control volume value in some circumstances (depending of numerics details). How could you access such value ? First, you must understand that for most cases, the evaluation at boundary faces are evaluation at the integration points of those faces (read theory doc for i.p and faces details), and values can be associated to those location by interpolating control volume values using the element shape functions, or assigning it the closest (topologically speaking) control volume value.

For shape function interpolation, you can use Velocity u.trilin

For nearest control volume (topologically speaking), you can use Velocity u.linlin.

Hope the above helps,

Thanks for your help. It is exactly the control volume values that I want to Access, but both the .linlin and the .trilin options make the solver crash.


----------------------------------
Error in subroutine GETCORE :
LIN-LIN is an operator on boundary conditions.
GETVAR originally called by subroutine CAL_FLX_BCS

+----------------------------------------------------------+
| ERROR #001100279 has occurred in subroutine ErrAction. |
| Message: |
| Stopped in routine GV_ERROR |
| |
| |
| |
| |
| |
+----------------------------------------------------------+
+----------------------------------------------------------+
| An error has occurred in cfx5solve: |
| |
| The ANSYS CFX solver exited with return code 1. |
| No results file has been created. |
+----------------------------------------------------------+

The error occurs when I use the expression

Alpha = 100[W/m^2/K] + Velocity u * 10 [W/m^3/K * s]

and set Alpha as value for "external" Heat Transfer coefficient at the Wall

but also when I set 100[W/m^2/K] + Velocity u * 10 [W/m^3/K * s] inside the field "Heat Trans. Coeff.".

Any idea?

...of course i use Velocity u.trilin and Velocity u.linlin instead of Velocity u