CFD Online Logo CFD Online URL
www.cfd-online.com
[Sponsors]
Home > Forums > FloEFD, FloWorks & FloTHERM

Semi-transparent window glass model

Register Blogs Members List Search Today's Posts Mark Forums Read

Reply
 
LinkBack Thread Tools Display Modes
Old   June 15, 2014, 05:04
Default Semi-transparent window glass model
  #1
New Member
 
sarawut
Join Date: Jul 2013
Posts: 2
Rep Power: 0
sagumngen is on a distinguished road
I have some questions about CFD simulation because I am a beginner of the CFD field and also FloEFD. I have tried it to solve several problems in the text books, which I found, in most cases, quite accurate in comparing with hand calculation(e.g. conduction, convection, opaque surface radiation) to make sure prior to do a full DSF model. But when use the radiation heat model for the semi-transparent model(simple window glass).

1.I just want to consider only radiation, so I did not put the convection coeff. as a boundary. Is it correct?

2. When I activate the semi-transparent option. Only absorption and refrac coeff. can be specify or there may be a relationship between solid absorption and emissive, but Kirchhoff’s Law.

3.How can the leaving radiant flux becomes invalid or 0 if the solid temperature is not 0.

4.when I check the heat transfers via convection and radiation modes inside floEFD, I've noticed that for convection I can specify the fluid temperature in different domain seperately. In case of radiation, how can I model the heat exchange if the environmental temperature is different between outside and inside the room as in the general setting box allow me to specify only one value as global? Do I have to create my own equation goals?
Attached Images
File Type: jpg 1.jpg (81.0 KB, 21 views)
File Type: jpg 2.jpg (91.4 KB, 21 views)
File Type: jpg 3.jpg (72.3 KB, 12 views)
File Type: jpg 4.jpg (75.1 KB, 20 views)
File Type: jpg 5.jpg (60.1 KB, 12 views)
sagumngen is offline   Reply With Quote

Old   June 23, 2014, 02:19
Default
  #2
Senior Member
 
Boris Marovic
Join Date: Jul 2009
Posts: 378
Rep Power: 12
Boris_M is on a distinguished road
Hi Sarawut,

1. The way you set up the project is with heat conduction only so the glass window will also conduct heat but there is no fluid to have any convection. It is like the model is set into a vacuum. Is this the reason you set the environment temerature to 0K as if it is somewhere in space with absolute 0 temperature?

2. From your task description you have the transmissivity coefficient. This coefficient is basically the rest that didn't get absorbed. So idealized (or simplified) the transmissivity = 1 - absorbtion, meaning that in your graph of the task definition the transmissivity between 0.5 and 1.5µm is 90% so the absorption is 10%. This is, if the properties are the actual material properties. Usually this is measured by basically shooting a light ray through the glass and measure what reached the other side. But what you also have to consider in reality is that a fraction is also reflected on the surface of the glass, then another part is absorbed in the glass which is heating it up and the rest is leaving the glass. So the more accurate definition or way to get to the absorption coefficient is by the following equations:

transmissivity = I_out/I_in

So the ratio between the outcoming and the incoming Intensity of the radiation or light which is what I mentioned above which is usually measured.

The absoroption however is:

a = -1/L*ln(T/(1-r)^2)

with r = ((n-1/(n+1))^2

and L being the thickness of the material and n the refraction index. T is the transmissivity and ln the logaritmic with the base e so no other variable such as T or n :-)

That then will give you the accurate absorption coefficient taking into account the reflection on the incoming surface which will be calculated by FloEFD solver on its own due to the laws applied. This is necessary as the "solid material" definition is only for what happens inside the solid, not the surface, for that a surface condition is required.
In general the absorption is equal to the emissivity because everything the material absorbes is also again emitted by its increased temperature. So the glass doesn't need an additional emissivity coeficient to be defined.
Is there a reason you want to define a surface radiation boundary condition?

3. I cannot tell why the goal becomes invalid, I will check if it appears also in my test.

4. The environmental settings in the general settings are defined for the radiation coming from outside the computational domain into your simulation. For example if you place an object on a table, the table and the object is radiating with their temperature. In reality also the walls of the room radiate onto the object and the table but you don't necessarily want to simulate the whole room with it so you set the enviroment temperature which is radiating from the outside. For any internal radiation the actual body temperature of the interior of a box for example is used.

In case of your radiation task, you should define a 3 bands for the given transmissivity. One from 0-0.5µm, then from 0.5-1.5µm and above. So select 3 bands and define the value 500nm and 1500nm. Then also your averaging of the absorbtion in these bands is fully correct. if you use 300nm and 2500nm from the plain glass curve for the tainted one, you will have a lower average for the absorbtion. You can try it by going to your material settings and you will see the absorbtion value is not 0.9 anymore but lower as you have a lot between 0.3 and 0.5µm and between 1.5 and 2.5nm that has 0 transmissivity.

I hope this helps and I will test your goal problem and get back to you soon.
Boris
Boris_M is offline   Reply With Quote

Old   June 23, 2014, 02:51
Default
  #3
Senior Member
 
Boris Marovic
Join Date: Jul 2009
Posts: 378
Rep Power: 12
Boris_M is on a distinguished road
I found what was wrong in your goals. You need to use the "Net volume radiant flow" as goals.
If you go in the help and search for that you will find it in the "List of parameters and their definition" on what they mean.

Boris
Boris_M is offline   Reply With Quote

Old   June 26, 2014, 16:34
Default
  #4
New Member
 
sarawut
Join Date: Jul 2013
Posts: 2
Rep Power: 0
sagumngen is on a distinguished road
Dear Boris,

Thank you very much for your reply! And it is really helpful and I have some ideas to solve my issues. However, some information I gave were not so clear from the beginning, these are;

1. Like you said, the actual behavior is the combinations of three modes. The reason I did was to make sure I can understand (manually checking) the effects of each mode (this case only radiation, get rid of others) before conjugation. Otherwise, it is hard for me to check the results after running if reasonable or not.

2. For glass window applications, there are coatings on clear/tinted glass surfaces (e.g. low-E, solar reflection) for several purposes. That is why I am looking for this boundary condition. So, can you please suggest me what should I do? Perhaps there are some techniques for modeling directly/indirectly.

3. One thing I do not understand is that as long as the object temperature is not 0K, it should radiate the heat wave at the surface either opaque or semi-transparent. But floEFD considers it invalid and I have to use”Net volume radiant flow” as you suggested. I also have read the meaning of each variable, but there are no formulae mentioned how to obtain the answers. It is hard for me to cross check unlike convection and conduction.

4. If I want to simplify the model without creating the interior wall for putting the temperature as boundaries (only glass pane to be modelled), what is the environmental temperature I should set? Regardless of other modes, if the external temperature is Te and the internal temperature is Ti, the glass pane sits between outside and inside of the room, as shown in the attached pics. Or I have no choices, but have to establish a box as an interior boundary.

In case of the radiation task, I have tried as per your comments as shown the attached pic as well (I don’t know if what I put is 100%correct) another thing, I put only 10000nm for you to noticed. Actually it should be 100000nm? The average value of each bandwidth has been shown in the “red box” my problem is if I want to know the overall average value. How can I get it?


Thank you again!
Attached Images
File Type: jpg G1.jpg (79.1 KB, 9 views)
File Type: jpg G2.jpg (25.0 KB, 17 views)
sagumngen is offline   Reply With Quote

Old   June 30, 2014, 08:20
Default
  #5
Senior Member
 
Boris Marovic
Join Date: Jul 2009
Posts: 378
Rep Power: 12
Boris_M is on a distinguished road
Hi,

2. No it is not possible to define a coating in the way you want to do it with a radiation surface. You can however create a thin layer with the corresponding absoption coefficient for the thickness of the layer and FloEFD can handle that without resolving the thichkness of the layer. However I wouldn't use nm thickness as there are limitations to the CAD dimensions and tollerances in general.

3. I don't know what formula's you need specifically as at least for me the parameter description is explaining it. A glass does not have the radiant flux or rate parameter it has the volume radiant flux or flow respectively. Solar always considers only the radiation from a solar source, Thermal due to any other temperature such as a component or environment without solar. So the "Net" stands for the difference between the emitted and the absorbed radiation => Net ...= Rad_emmitted-Rad_absobed.
With the "Absorbed volume radiant flux" (which you saw works) you can tell how much is the emitted part as you have the "Net volume radiant flux and the absorbed and then it is a simple calculation to get to the emitted value.

4. No, the "environment radiation" comes from outside the computational domain. You can show the influence by right click on the project name and I think something like "show environment radiation" or so. It will show a wavy arrow on each computational domain side. So it cannot be split into left and right side of the domain. You can create a different radiation source on a wall with a fixed temperature but you will have to model that wall.

Yes, the error in the average comes from the hard cut-off in the definition of the material properties. If you use 299 for 1 and 300 for 0.1, then 2499 for 0.1 and 2500 for 1 it works much better. You also don't have to define the 100 and 10000 as the last and first value will be kept constant for the values after the last and before the first. The error then is only minor (less than 1%).

I hope this helps.
Boris
Boris_M is offline   Reply With Quote

Old   July 1, 2014, 02:46
Default
  #6
Senior Member
 
Boris Marovic
Join Date: Jul 2009
Posts: 378
Rep Power: 12
Boris_M is on a distinguished road
By the way, you can create like a room/box around the one side of the glass, define the box as insulator so it doesn't conduct any heat and apply to interior walls a pressure boundary condition so the fluid can leave the box like in a natural convection case if you want to use fluid at some point and then apply a radiation source on the inside of the box with your ambient temperature.

Boris
Boris_M is offline   Reply With Quote

Reply

Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are On
Pingbacks are On
Refbacks are On


Similar Threads
Thread Thread Starter Forum Replies Last Post
Is it possible to model natural convection in a 2D horizontal model in fluent caitoc FLUENT 1 May 5, 2014 13:32
Error finding variable "THERMX" sunilpatil CFX 8 April 26, 2013 07:00
2 stage axial turbine model convergence issues sherifkadry CFX 2 September 7, 2009 20:51
multi fluid mixture model issue rystokes CFX 3 August 9, 2009 19:13
Kato-Launder model sam Main CFD Forum 13 September 21, 2006 10:15


All times are GMT -4. The time now is 22:48.