Face numbering for Shell elements and CAX elements

disla

Hi Victor,

I’m finding an unexpected result, and I think I found the source of the discrepancy.
I’m validating an axisymmetric model in which there is a flow Network model thermally interacting with the CAX elements.

Something like this but way simpler so I can build an analytical solution to compare:

https://www.calculix.de/fw05.html

As the thermal analysis in Mecway do not consider Axisymmetric elements I had to set them with custom cards.
The face numbering for Shell elements and CAX elements is not the same and Mecway is assuming the Shell numbering convention for the CAX.

I guess that makes the forced convection to be assigned to the wrong faces.
¿Is it possible to fix this is the next version?. If Mecway finds a Custom element definition the face numbering should be adjusted to the custom element.

disla

I have written the surfaces by hand and the result now matches very well with the theoretical expected result. The small deviation is due to the idealization of zero tube wall thickness used by the formulas. I have had to use a minimum thickness to be able to build a convective support surface. All the actual elements work perfectly but I decided using CAX4. I think it could be interesting to implement it in MECWAY and could be used as reference.



Example based on:
Fundamentals of Heat and Mass Transfer - Incropera - 6th Edition
Example 11.1

disla

Hello Victor,

I'm considering including an amplitude for the mass flow. First thing I have done is to request the actual value with *NODE FILE MF and I have found that the one defined as negative (water in the attached file) is showing and alternating value in the result window. Final temperature is as expected but Mass flow is showing wrong for the water network. One can see the alternating color in the screen. ¿Is that a Mecway or ccx issue?

*NODE FILE
MF

disla

Here an example. Thermal Transient with varing negative mass flow.
Result shows an alternating MF. That do not happen when the mass flow is positive. (constant or varing). I think this is a ccx problem because frd is showing the alternating value too.

-3
1PSTEP 42 21 1
100CL 121 1.000000000 11 1 21 1
-4 MAFLOW 1 1
-5 MF 1 1 0 0
-1 1 1.00000E+00
-1 2-1.00000E+00
-1 3 1.00000E+00
-1 4-1.00000E+00
-1 5 1.00000E+00
-1 6 1.00000E+00
-1 7-1.00000E+00
-1 8-1.00000E+00
-1 9 1.00000E+00
-1 10 0.00000E+00
-1 11-1.00000E+00
-3
9999

Victor

Hello Disla

Regarding axisymmetric element faces, I don't want the rest of the GUI to interpret anything defined in the CCX branch because it'll become confusing as to what things it understand and what it doesn't.

For mass flow rate at end nodes, the CCX manual says the midside nodes have that value (which is -ve as you defined) but I don't really understand what it means about the end node values. Maybe where it says "mass flow leaving the node", it means it disregards the flow entering so we're seeing only the +ve exit flow which is balanced by an invisible -ve mass flow rate entering?