CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   STAR-CCM+ (http://www.cfd-online.com/Forums/star-ccm/)
-   -   localized temperatures (http://www.cfd-online.com/Forums/star-ccm/80482-localized-temperatures.html)

mobius September 27, 2010 17:11

localized temperatures
 
hopefully this will be an easy question for the more experienced users. So I have modeled a cavity that will be filled with multi-component gas. I have successively wrapped, remeshed and volume meshed the interior of the cavity. Now at various locations, which I already have designated the boundaries for, I need to simulate a hot wall. So, along this wall I will have a number of heat sources. I was hoping that I could just assign it as a local boundary condition, but apparently I don't have the option to input a local wall temperature. Is this possible, if so how do I do it? I was hoping to stay away from volume meshing the fluid region and solid region and doing interfaces or something complex like that.

vishyaroon October 4, 2010 16:59

You can create a field function that has localized temperatures.

For eg.

($$Position[1]=1.0)&&($$Position[2]=4.3)?473.15:273.15

which is the equivalent of

if (y==1.0)&&(z==4.3) then Fieldfn = 473.15 else Fieldfn = 273.15

Once you define this you can assign this field function as your Temperature Profile. You may have to go for nested loops if you have several temperature sources.

mobius October 4, 2010 18:19

Thanks for the suggestion vishyaroon. It just seems that with a complex piece of software like this you could assign a temperature boundary condition to a section of a wall that was previously defined as it's own region It's frustrating, kinda like it's just not right to not have an undo button. What we are trying right now is that we made all of our hot sections a velocity boundary with no velocity and temperatures specified by field functions. Not sure if this is going to cause problems or not, but we have a test run crunching away. If this doesn't work, we'll look into your suggestion. Thanks

RobertB October 5, 2010 07:35

Had you considered using a Table?

mobius October 5, 2010 10:02

yeah, we have temperature profiles for each heat source so we were going to input those as tables, though we still couldn't figure out how to assign that to a wall region. since we have a limited number of points and the curves are not linear, we decided to do a curve fit for each source and input the profiles as field functions instead of using tables.

RobertB October 7, 2010 08:12

I am not sure I understand what you are saying.

If you want to apply a spatially varying temperature condition wouldn't you do the following


Define a table with the right spatial variation
Read it in as a table
Define the region as a wall
Define the thermal specification as temperature
Define the method as table(x,y,z)

If you need finer discretization you could add more points to the table.

mobius October 7, 2010 22:40

Maybe I don't understand your method. When we specify something as a wall we can only assign a global temperature value, all the wall regions seem to want to use this global temperature value. The components on the wall that are heat sources are complex shapes and I think it would be difficult to create a thermal table for each point on the components. We would also have create tables for every box, but also the temperatures of each component changes over time. The rest of the wall is assumed to be adiabatic. Our current method seems to be working well and is much easier/quicker to set up though it just doesn't make sense that we have to do it that way. Thus far we are seeing good results with no anomalies around the heat sources.


All times are GMT -4. The time now is 19:52.