hi, what 's the difference between internal and interface boundaries? i'm working with different solids in contact!each one has a thermal conductivity!what should i impose as boundary for faces connecting solids? thanks!!

Hi,

It is contact resistance if you want to pick-up/predict heat transfer from one solid to other.... When you don't specify anything its assumed perfect contact between these two solids... How are these two solids in contact in reality? Are these in contact due to some pressure applied by bolt etc... then you have to go into next level of modeling to incorporate the thermal contact resistance between solids which is complex phenomena and depends on many factors: surface roughness of two surfaces, thermal conductivity, hardness, pressure etc....

Read few articles on this topic if these are relevant for you: http://meweb.ecn.purdue.edu/~CTRC/re...ts/TCC/tcc.htm

http://en.wikipedia.org/wiki/Thermal...ct_conductance

Type thermal contact resistance and perform google search and you will find many many articles on this...

There is one site which could help you a lot on contact thermal resistance is following:

http://mhtlab.uwaterloo.ca/MHTLarchive.html

Have fun...

AJ

hi AJ, in deed i wonna study heat transfer between the solids but i can't find "contact resistance" in boundary conditions list!!!??? any way thanks for help and for the information!!

Here it goes for contact resistance in Fluent:

You could define thickness of material and material which will offer the thermal resistance...

Select the wall which is between the two solids in contact, compute the contact thermal resistance externally using different methods ( CMY, Cooper etc...) and convert this resistance/impedance into a material which could offer this resistance.

Under the BCs for this wall --> thermal tab, it should be coupled thermal condition, set material to the effecive material property, set thickness to the required thickness ( calculate 1D resistance from the t/k like)....

You should be able to solve contact resistance problem this way... Icepak (another software) which uses Fluent has better GUI where in you could use a plate in the contact region which will have the contact resistance properties....

Just try with a material with fictious k=0.1 W/m-K and keep changing this value and see if the temperature is changing, with some finite thickness... You could also specify planar conduction by enabling Shell conduction ON for this interface material...

Hope this helps...

ENjoy...

AJ

[momemon]

I am simulating a test setup where a thermal interface material is in contact between two aluminum blocks. So in other words the TIM is sandwiched between two Al blocks. I defined the four contacting surfaces(lower of first block, upper and lower of TIM, upper of second block) as interfaces in Gambit.

The heat tranfer is taking place with the material properties (density, K and Cp) but it cannot be accurate since I do not have the heat coefficient value as an input. I do not understand how to input the heat coefficient h for an interface?

Do I have to changes the interfaces to walls? Would the heat transfer take place then?

Any help would be highly appreciated.

Omar

[Sai Krishna]

Originally Posted by AJ
;142226

Here it goes for contact resistance in Fluent:

You could define thickness of material and material which will offer the thermal resistance...

Select the wall which is between the two solids in contact, compute the contact thermal resistance externally using different methods ( CMY, Cooper etc...) and convert this resistance/impedance into a material which could offer this resistance.

Under the BCs for this wall --> thermal tab, it should be coupled thermal condition, set material to the effecive material property, set thickness to the required thickness ( calculate 1D resistance from the t/k like)....

You should be able to solve contact resistance problem this way... Icepak (another software) which uses Fluent has better GUI where in you could use a plate in the contact region which will have the contact resistance properties....

Just try with a material with fictious k=0.1 W/m-K and keep changing this value and see if the temperature is changing, with some finite thickness... You could also specify planar conduction by enabling Shell conduction ON for this interface material...

Hope this helps...

ENjoy...

AJ

Hi AJ,
I know this is old post to comment or ask a question?
but iam also simulating the same condition, in which i have to provide thermal resistance between 2 materials.
u r talking about assigning material to interface boundary. Can we already assign a material to a solid in cell-zone conditions right...? Again assigning different material to surface is OK with fluent ....?
one more thing, in shell conduction also one material tab is available. which one to change to provide proper material.
any help will be greatly appreciated in this regard.

Thanks in advance....

[MKuhn]

Concernig your screen shot, the left material selction is only valid when shell conduction is off. Assigning different material to surfaces is Ok, for example a layer of painting, a layser of insulating and so one. Heat generation can be a film heater between to solids or heat generation due to X-ray.


To define a thermal contact resistance, keep the shell conduction off. Choose an arbitrary solid (left side of your screen shot), check the thermal conductivy of this solid and calculate the wall thicknes according to your desired contact resistance:


[contact resistance] = [wall thicknes / thermal conductivity of the solid]

[Sai Krishna]

Originally Posted by MKuhn

Concernig your screen shot, the left material selction is only valid when shell conduction is off. Assigning different material to surfaces is Ok, for example a layer of painting, a layser of insulating and so one. Heat generation can be a film heater between to solids or heat generation due to X-ray.


To define a thermal contact resistance, keep the shell conduction off. Choose an arbitrary solid (left side of your screen shot), check the thermal conductivy of this solid and calculate the wall thicknes according to your desired contact resistance:


[contact resistance] = [wall thicknes / thermal conductivity of the solid]

Thanks MKuhn for your reply.

I have one more doubt.
For a face can we check the normal direction in fluent, means consider a plane normal to z axis. If I define velocity or heat flux normal to the boundary, will it flow in +z or -z direction.
If we know the face normal direction while defining the velocity itself we can give value with sign right....?

Thank you

[MKuhn]

Regarding the boundaries, it has nothing to do with the direction. A positive value means heat goes into your model, a negative value means heat goes out of your model. A good practice is also to check the energy balance over the boundaries via "Flux Reports" and "Total Heat Transfer Rate".

[Sai Krishna]

Originally Posted by MKuhn

Regarding the boundaries, it has nothing to do with the direction. A positive value means heat goes into your model, a negative value means heat goes out of your model. A good practice is also to check the energy balance over the boundaries via "Flux Reports" and "Total Heat Transfer Rate".

isn't it the converse, as usually for flux we define them towards outward normal direction right? so in that direction wont it be positive ....?

[MKuhn]

I think for the energy balance the heat flow must be defined as I noted before, independent from the direction vector of the boundary. This is different from defining the velocity, here it is like you wrote.

Maybe some one else can clarify this a little bit better.