Note: This discussion is about an older version of the COMSOL Multiphysics® software. The information provided may be out of date.
Discussion Closed This discussion was created more than 6 months ago and has been closed. To start a new discussion with a link back to this one, click here.
surface to surface radiation without surface
Posted 01.03.2013, 05:28 GMT-5 Heat Transfer & Phase Change Version 4.3a 5 Replies
Please login with a confirmed email address before reporting spam
Hello,
I work on a molten charge and i have some issues with radiation. Usually i use a surface to ambient radiation condition but it is not enough precise for my problem, i have to take into account the reflection of the surfaces (emissivity of other surfaces emi(ext)=0.1) arround my charge.
I have a quite simple geometry and the ambiant view factor is close to 0.
Problem is that i can't add gaz and other surfaces in my model.
Can i introduce a simple heat flux on a boundary representing the reflection of my surface?
I have done lot of tests with simpler geometry but i can't find a way to put the good condition wich is closed to what i can measure with surface-to-surface condition
For example:
Radiation to ambiant condition is assumed to be:
q1=emi(ch)*sigma_const*(T_cr^4-T^4) (W/m^2)
T_cr is the temperature of the surface of other elements which i don't wan't to draw, T is the surface temperature of the charge.
The reflection is:
q2=-(1-emi(ext))*emi*(ch)*sigma_const*(T_cr^4-T^4)*(1-Famb) (W/m^2)
I have already measure Famb, is closed to 0. But when i add this heat flux, solver doesn't converge. It seems that i can't bring power to my surface. I have also tried boundary heat source without success.
I am in 2d-axy
Any ideas?
I need some help
Thank you very much
I work on a molten charge and i have some issues with radiation. Usually i use a surface to ambient radiation condition but it is not enough precise for my problem, i have to take into account the reflection of the surfaces (emissivity of other surfaces emi(ext)=0.1) arround my charge.
I have a quite simple geometry and the ambiant view factor is close to 0.
Problem is that i can't add gaz and other surfaces in my model.
Can i introduce a simple heat flux on a boundary representing the reflection of my surface?
I have done lot of tests with simpler geometry but i can't find a way to put the good condition wich is closed to what i can measure with surface-to-surface condition
For example:
Radiation to ambiant condition is assumed to be:
q1=emi(ch)*sigma_const*(T_cr^4-T^4) (W/m^2)
T_cr is the temperature of the surface of other elements which i don't wan't to draw, T is the surface temperature of the charge.
The reflection is:
q2=-(1-emi(ext))*emi*(ch)*sigma_const*(T_cr^4-T^4)*(1-Famb) (W/m^2)
I have already measure Famb, is closed to 0. But when i add this heat flux, solver doesn't converge. It seems that i can't bring power to my surface. I have also tried boundary heat source without success.
I am in 2d-axy
Any ideas?
I need some help
Thank you very much
5 Replies Last Post 05.03.2013, 02:38 GMT-5
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hi Julien
From what I understand the surface to ambient is assuming you give an integrated emissivity, and the solid angle is a hemi-sphere, if you now say you have lateral surfaces you nedd to use the View/ Form factors, or have COSMOL to estimate these. The fact you have no gas doesnt change the surface to surface radiation but you have somewhere an option of absorbing media inbetween, (by default off). So if you add surface to surface and sorface to ambient, (and COMSOL allows it without overriding, then you need to add some manual view factors on the surface to ambient, from my understanding)
Radiation is tricky, indeed, and the high non linearity makes it more complex to validate.
My wy is really to mae a simple model and then to dump all COMSOL variables and check to see if I get them to link to my manual equations, then to vary some to check that I understand the relations. But I admit I havnt done this the last year(s?) so I would ahve to start again ;) Multiphysics is great, but my head starts to become too small to remmeber all the details of all physics, I lso need regularly some training when starting on a new subject ;)
--
Good luck
Ivar
From what I understand the surface to ambient is assuming you give an integrated emissivity, and the solid angle is a hemi-sphere, if you now say you have lateral surfaces you nedd to use the View/ Form factors, or have COSMOL to estimate these. The fact you have no gas doesnt change the surface to surface radiation but you have somewhere an option of absorbing media inbetween, (by default off). So if you add surface to surface and sorface to ambient, (and COMSOL allows it without overriding, then you need to add some manual view factors on the surface to ambient, from my understanding)
Radiation is tricky, indeed, and the high non linearity makes it more complex to validate.
My wy is really to mae a simple model and then to dump all COMSOL variables and check to see if I get them to link to my manual equations, then to vary some to check that I understand the relations. But I admit I havnt done this the last year(s?) so I would ahve to start again ;) Multiphysics is great, but my head starts to become too small to remmeber all the details of all physics, I lso need regularly some training when starting on a new subject ;)
--
Good luck
Ivar
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Thank you again Ivar,
I discover that it is possible to have different domains with no contact defined in ht physic with surface to surface radiation! I was thinking that contact was an obligation on domains.
So i can draw other radiative parts of my model but i can't add a gas domain between radiative parts because i have a lot of boundary conditions, heat flux which i can't defined in internal boundaries.
So i don't know what to do with the absorption factor of transparent media. I 'm not able to find where it is defined in a classic case (with the transparent media), material properties ? surface to surface radiation interface ?
Julien
I discover that it is possible to have different domains with no contact defined in ht physic with surface to surface radiation! I was thinking that contact was an obligation on domains.
So i can draw other radiative parts of my model but i can't add a gas domain between radiative parts because i have a lot of boundary conditions, heat flux which i can't defined in internal boundaries.
So i don't know what to do with the absorption factor of transparent media. I 'm not able to find where it is defined in a classic case (with the transparent media), material properties ? surface to surface radiation interface ?
Julien
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hi
it's not "contact" but continuity, that is required for conduction or convection HT exchange,
But pure radiative is not considering the domain in between, except if you turn on the particpative media (check main HT physics), then you need to define the domain in between too, hence ada domain and continuity BCs
--
Good luck
Ivar
it's not "contact" but continuity, that is required for conduction or convection HT exchange,
But pure radiative is not considering the domain in between, except if you turn on the particpative media (check main HT physics), then you need to define the domain in between too, hence ada domain and continuity BCs
--
Good luck
Ivar
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Be carefull,
Comsol assistance ask me yesterday that surface to surface radiation in 2D-axy geometry doesn't give good results because of wrong view factor calculation in version 4.3a. In 3D and 2D it's ok but not in 2D-axy.
This error will be correct in 4.3b.
Comsol assistance ask me yesterday that surface to surface radiation in 2D-axy geometry doesn't give good results because of wrong view factor calculation in version 4.3a. In 3D and 2D it's ok but not in 2D-axy.
This error will be correct in 4.3b.
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hi
thanks for the info, that is an important note !
indeed its easy to forget the 2*pi*r in 2D-axi ;)
--
Good luck
Ivar
thanks for the info, that is an important note !
indeed its easy to forget the 2*pi*r in 2D-axi ;)
--
Good luck
Ivar