first_post-first_question

[Filippo Monari]

Hi, It's my first post and I premise I'm totally new to cfd.
I need to investigate the air flow path around a little building and inside it.
Now I'd like to know how and if is it possible set up temperatures of walls as boundary conditions.
This is not much important if I run an external simulation but for the interior space could be.
I have read something about but (coupling with syrthes, use the enthalpy to define the temperature) but nothing specific and clear (maybe is my fault), so if someone could give some advice about where I have to look I would be glad.
Furthermore any kind of advice about how to set up my simulation environment are welcome.
Thank you in advance,
FM

[César Vecchio]

Ciao, Filippo:
 
I've made analysis similar to yours, but instead of a building it was about a commercial kitchen with make-up air systems. You certainly can impose temperatures and heat fluxes as boundary conditions, and there's no need of Syrthes to do that. Syrthes (on which I completely lack experience) is, I think, useful when you need to determine the interaction between heat fluxes across solids and fluids; but if you're dealing only with fluids Saturne alone is quite powerful.
Have you already made some turorials? If so, imposing wall temperatures is no big deal. In the GUI, go to "Thermophysical models", then expand the options tree and select "Thermal model". Once there, I suppose you'll do fine with choosing "Temperature (Celsius)" or  "Temperature (Kelvin)". Choosing one or the other will determine which scale you'll use to define the temperatures.
Once you've done that, you'll see that in the rest of the options tree there will appear new fields related to the temperature (e.g. "initializations temperature"). Under the "Boundary conditions" section a new table is shown for you to impose a temperature or a flux. Take a look at this and see if it's what you need.
Regards,
 
César Vecchio

[Filippo Monari]

Ciao Caèsar,
yeah I've made some tutorials and you give me confirm I'm looking in the right place but there is something i miss.
What I don't figure out (or maybe I misunderstand) is the following:
1) I set the thermal model as Temperature (Celsius);
2) I initialize the domain temperature as a value that I know to be quite near the final temperature;
3) I set the the boundaries
4) (here we are) I go to boundary conditions section:
------> I select, for example, a wall of my domain
------> than a scalar tab appears with the scalar: TempC, condition type: prescribed flux, value: 0
------> (maybe the error lies here) I change the condition type to prescribed value (because I now the temperature value) and I set my boundary value for the temperature
5) I save the xml file and run the simulation, which immediately gives to me an error.
Looking at the list file I find a series of errors like:
 
@                                                           
@ INCOHERENCY BOUNDARY CONDITIONS VELOCITY-SCALAR           
@   FACE          1; PROPERTY 1:        -1; VARIABLE TempC  
@     SCALAR NUMBER          1; PHASE          1
@     ICODCL SCALAR          5; ICODCL VELOCITY          6
@                                                           
 
I think maybe there's some wall on which I can't impose a condition about a fix temperature (I don't know It's my first approach to the subjects...).
If you (or someone else) could have a look and help to figure out where is the problem...I would be very glad.
Thank you in advance

[Alexandre Douce]

Could you attach your file of parameters (cells2-1) ?

[Filippo Monari]

yeah sure, sorry.
here it is.
thx for your help

[Alexandre Douce]

For the "bottom" boundary condition you should not set the roughness options. Try to switched off in the GUI, or delete the markup  <roughness>0.03</roughness> in the file of parameters.
 

[Filippo Monari]

Hi Alexandre and thank you.
It' seems to work but can ask you why?
I means where was the error?
If I have to represent a particular outside ground which could have high temperature and some degree of roughness what do I have to do?
Thank you again

[César Vecchio]

Filippo, although I cannot answer your last question, maybe you can see if by specifying viscosity and/or turbulence parameters you may obtain your desired behavior. However, as I suppose you're working with external flows, and not veru high speed ones, perhaps working with a laminar flow approach is not that bad (at least for initial runs). Given the scale of a building, the roughness of its walls and/or surrounding floor could be neglected. For example, when it comes to concrete piping of big diameters (such as those for waste waters), despite you feel them as sandpaper, for estimations with a Moody diagram you can safely say they are smooth. Technically, a surface can be considered smooth as long as there are no protuberances exceeding the laminar sublayer height.
I'm afraid that's all I can guess.

[Filippo Monari]

ok, thank you both guys.
I'll make some test simulation and I'll see what I'll get.

[Filippo Monari]

hi guys, just a last question...
I have a domain of about 43x103x21m...what could be a reasonable time, along which run the simulation?
thank you