[ahmadbakri]

hi all,


I am simulating friction stir welding process, when we set a convection thermal conditions for the walls of some material, does the STAR CCM consider that there is a heat transfer (by convection) between that material and Air for example. Or do we need to put another material to have a heat transfer with.
i.e. when we select a convection for thermal condition, is it enough? or do we need to set a medium (another material like Air) close to that material to have heat transfer with ?

thanks

[charltonjames]

Hi,

Im looking at CFD to do similar studies. Im afraid i dont have an answer for you, but was wondering if you had made any progress in this area?

Regards

James

[fshak92]

I have also the same question.
Is there anybody knows the answer?

[abdul099]

It is enough to set the convection condition and set appropriate values.
Convection will need an ambient temperature and a heat transfer coefficient. The HTC depends on flow conditions and material properties near the outer side of the boundary. So you don't need to model the ambient, but you need reasonable values and should consider if you can model your physical problem with convection boundaries.

[fshak92]

Originally Posted by abdul099

It is enough to set the convection condition and set appropriate values.
Convection will need an ambient temperature and a heat transfer coefficient. The HTC depends on flow conditions and material properties near the outer side of the boundary. So you don't need to model the ambient, but you need reasonable values and should consider if you can model your physical problem with convection boundaries.

I've done a sample simulation to check it.
I Created a solid in which boundaries have convection without creating its surrounded air.
I set the ambient temperature to(e.g) 22 and the initial condition to 28 (and a little heat source more than 28) but its interesting that the temperature of the solid part never goes less than initial condition!!!
It means that the ambient temperature and convection boundary have NO influence on the simulation.
Do you know why?
Thank you again for your answers and helps.

[abdul099]

Originally Posted by omid88

It means that the ambient temperature and convection boundary have NO influence on the simulation.

Sorry, that's just rubbish. Or do you think, boundary conditions are just to drive newbies crazy? Put some reasonable numbers and you will get the right result.
Check your mesh, the energy source, boundary conditions etc, but the convection boundary condition DOES work.

Please describe your setup more detailed, maybe we can track down the reason for the unexpected behaviour. How does your domain look like? Which boundaries do you have and which condition did you set? What about the energy source?

[fshak92]

Originally Posted by abdul099

Sorry, that's just rubbish. Or do you think, boundary conditions are just to drive newbies crazy? Put some reasonable numbers and you will get the right result.
Check your mesh, the energy source, boundary conditions etc, but the convection boundary condition DOES work.

Please describe your setup more detailed, maybe we can track down the reason for the unexpected behaviour. How does your domain look like? Which boundaries do you have and which condition did you set? What about the energy source?

Thank you again for your help.

I attach the temperature distribution of the body.
I meshed it : surface remesher,prism layer,trimmer,Base size 12mm-
using one solid continuum with initial temperature :28 c.
Using convection for whole bodywith the ambient temp 22 C.(except one little part considered as a heat source) .
Using constant temperature for the little surface inside the body : 30C.
120 iteration with no oscilating residual about 10^-3.
(The meshing is not so good but its just for a test and im not looking for the exact result in this step)

The problem is that the body does not go less than 28C(initial condition),,and the influence of the convection cannot be seen,i know its not realistic but i just say the thing that software gives me.

Again thank you for the time you take to help.

[fshak92]

Sorry another question: (this question is about my general problem''not the simulation above'')
i have constant temperature of the machine(showed above) according to their coordinates.my goal is finding HTC of the body.
First I created the air around the machine(using 2regions) but there are some problems because with that situation a lots of parameter (like the distance of the wall and...) have influence on the HTC i get and i could not set just ambient temperature.

Just i want to have an ambient temperature around my body(the body has constant temperature).
Second,im trying to achieve my goal by creating 1region,,,In the simulation i showed you i need to have constant temperature on my body and have an ambient temperature around that!!! but i dont know how can i simulate this situation.

1)by creating 2regions i cannot set an ambient temperature for the air
2)by creating 1region(just machine) i cannot both assign constant temperature and ambient temperature around the machine.
((the goal is getting constant temperature of the machine of another simulation software, and giving the HTC by starccm))

Sorry for the lots of things i told but i would appreciate if you have any idea or opinion.

[Zeppo]

Originally Posted by omid88

Using convection for whole bodywith the ambient temp 22 C.(except one little part considered as a heat source) .

What if you try increasing HTC, say, 100 times more and do steady simulation.

[fshak92]

Originally Posted by Zeppo

What if you try increasing HTC, say, 100 times more and do steady simulation.

Thanks I will try it to check the influence.

But my simulation is steady therefore after the mentioned iteration i think we should expect the influence of the convection(even with the normal HTC=1)

I will check it and say the result here soon.