Steady state Thermal ANSYS. HELP!!! Suggestions needed!!!

[Aravindan280287]

Dear all,

TEST TYPE 1:
I have a metal housing with some electronics inside. This housing is kept inside a heating chamber where hot air is circulated at a desired temperature. The hot air is not directly blown into the housing which will make this as a forced convection problem. Instead, the hot air circulates slowly inside the chamber to make the ambient temperature to be the temperature of the hot air. There are some resistors in the electronics which will generate heat from inside the housing. After 3 hours, temperature will be measured on the outside surface of the Housing at different points and also on the PCB.

Test was done. The test results were close to ANSYS simulation results (~2°C variation).

ANSYS Simulation parameters:
Type: Steady state Thermal
Convective heat transfer coefficient: 10W/m2K (Sink)
All other materials were given Thermal Conductivity values
Bonded contact between all parts with program controlled values
Resistor was given Internal heat generation (Source)

TEST TYPE 2:
The same housing was tested outside in open atmosphere with room temperature conditions. And this time, the potting material inside housing was removed during the test (which means a hollow housing without potting). The test results vary with the simulation results (~10°C variation).

Simulation parameters:
Type: Steady state Thermal
Convective heat transfer coefficient: 10W/m2K (Sink)
All other materials were given Thermal Conductivity values, potting volume is suppressed and hollow space inside.
Bonded contact between all parts with program controlled values
Resistor was given Internal heat generation (Source)

WHAT WILL BE REASON FOR RESULTS MISMATCH???

Do I need to assign any properties for the void space inside or convection inside?

I would be happy to have valid suggestions.
Thanks all

 

[lamboram]

two questions

1. in both cases which was hotter (simulation or measurement)?
2. what exactly is the potting volume? if it was suppresed to have a hollow volume, why was the internal Heat generation still given in the resistors?

Regards,
Ram

 

[Aravindan280287]

Hello Ram,

Thanks for your reply.

1. In both cases, temperature values from simulation were higher than that of test.
2. The potting volume is a significant volume (~occupies nearly 60% of housing). It is supposed to act as a shock absorber. The resistors in the other hand, are heat generating sources where power loss occurs in most electronics.

Now that I have a hollow volume inside my model (in simulation), do I need to assign convection on those internal faces with a lower convective heat transfer coefficient value?

 

[Aravindan280287]

TEST TYPE 2: Only the potting material is removed. All other components stay (PCB board and resistors) inside the Housing.

 

[lamboram]

possible explanations, since I do not know exactly the setup.

where is the housing placed upon? it could be a metal plate or something which acts as a sink. in case of the chamber, because of the chamber's higher temperature, this contact between the housing and the placement where it is placed upon doesnt bring much difference. however in room temperature, with internal heat generation, the contact between the housing and the placement makes a difference. This would be a good place to start. This explains the higher temperature of Simulation compared to reality (absence of a sink).

regarding the void. Ansys treats the voids as adiabtic vaccum. Since the simulation is hotter, i dont think it has something to do with the convection inside the housing. maybe by fitting the convection coefficient value you might fit the simulation to the exact measured value (please consider the Temperature sensor's tolerance which was used to measure in the test lab, when doing so.)

Upon further info, i can give you more suggestions!

Ram

 

[Aravindan280287]

Will come back with more info on tests.
Thanks for the inputs