Natural Convection - Evaluate Properties

Lemanes · May 11, 2015, 13:24

I am doing natural convection in a 2D square cavity, where one side is at T_hot=400K, the other side is at T_cold=380K, and up and bottom are adiabatic.

In my first try I used the full model (using the ideal gas from the library). In my second try I used the bousinnesq approximation, using air at 25C from the library. However, in both cases when I take the final average conductivity, or the final dynamic viscosity, I am getting the values for a gas at 25C. Is that correct? Therefore, should I enter a expression to change my temperature based on my expected average? In other word, should I create a new material with the properties of air at 290K? I think that everything should be automatic.

My goal is evaluate Rayleigh, therefore I have to get the correct properties at the air average temperature to use the following equation:

Rayleigh#=((density^2)*g*(T_hot-T_cold)*cavitysize^3)/(Conductivity*Dynamic Viscosity)

or in CEL (Ansys)

((ave(Density)@Fluido) *(ave(Density)@Fluido) *9.8[m* s^-2] *(ave(Specific Heat Capacity at Constant Pressure)@Fluido) *(1/Average)*(T_hot -T_cold) *cavitysize ^3)/((ave(Dynamic Viscosity)@Fluido)*(ave(Thermal Conductivity)@Fluido))

I really need your help guys, I already lost a month trying to solve it. The problem is that I gonna evaluated several cavities with different sizes and temperature gradiantes...

ghorrocks · May 11, 2015, 20:00

The physical properties of air ideal gas and air at 25C are all taken at 25C. If you want to use anything else you need to define what you want. You can easily define properties to be a function of temperature with CEL functions or 1D interpolation functions.

A comment on your CEL expression: You probably want to use volumeAve rather than ave. ave is just the arithmetic average of all values which can be distorted by changes in mesh size, volume average is the volume average which takes that into account.

Lemanes · May 11, 2015, 20:24

Thank you. Your answer was very helpful.

And just one more question: In future, I will change my t_hot boundary condition to heat being generated at the wall at an specific rate. Therefore, I would not know my average temperature. In which temperature would I define my fluid?

ghorrocks · May 12, 2015, 06:48

If you don't the temperature then use variable properties and it will take care of it automatically. It will also handle the physical properties difference from the hot side to the cold side as well.

Lemanes · May 12, 2015, 07:13

Thank you again.

I already set the function to change properties and it is working well. And in future, how can I use variable properties automatically? Is there any command?

ghorrocks · May 12, 2015, 07:15

Save it as a CCL snippet so you can easily add it to any future simulation when needed.

May 11, 2015, 13:24	Natural Convection - Evaluate Properties	#1
Lemanes New Member LManes Join Date: Oct 2014 Posts: 13 Rep Power: 11	I am doing natural convection in a 2D square cavity, where one side is at T_hot=400K, the other side is at T_cold=380K, and up and bottom are adiabatic. In my first try I used the full model (using the ideal gas from the library). In my second try I used the bousinnesq approximation, using air at 25C from the library. However, in both cases when I take the final average conductivity, or the final dynamic viscosity, I am getting the values for a gas at 25C. Is that correct? Therefore, should I enter a expression to change my temperature based on my expected average? In other word, should I create a new material with the properties of air at 290K? I think that everything should be automatic. My goal is evaluate Rayleigh, therefore I have to get the correct properties at the air average temperature to use the following equation: Rayleigh#=((density^2)g(T_hot-T_cold)cavitysize^3)/(ConductivityDynamic Viscosity) or in CEL (Ansys) ((ave(Density)@Fluido) (ave(Density)@Fluido) 9.8[m* s^-2] (ave(Specific Heat Capacity at Constant Pressure)@Fluido) (1/Average)(T_hot -T_cold) cavitysize ^3)/((ave(Dynamic Viscosity)@Fluido)*(ave(Thermal Conductivity)@Fluido)) I really need your help guys, I already lost a month trying to solve it. The problem is that I gonna evaluated several cavities with different sizes and temperature gradiantes...

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
natural convection in closed cavity with chtMultiRegionFoam and variable properties	FoamBoy	OpenFOAM Running, Solving & CFD	0	February 5, 2015 10:30
Thermophysical properties for natural convection	Ciefdi	OpenFOAM Running, Solving & CFD	0	November 7, 2013 11:44
properties in natural convection	ghassan77	Main CFD Forum	1	January 17, 2011 02:00
properties in natural convection	ghassan77	FLUENT	0	January 7, 2011 10:09
Natural convection. Air properties	Cristina	FLUENT	3	March 24, 2007 08:23

May 11, 2015, 20:00		#2
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,703 Rep Power: 143	The physical properties of air ideal gas and air at 25C are all taken at 25C. If you want to use anything else you need to define what you want. You can easily define properties to be a function of temperature with CEL functions or 1D interpolation functions. A comment on your CEL expression: You probably want to use volumeAve rather than ave. ave is just the arithmetic average of all values which can be distorted by changes in mesh size, volume average is the volume average which takes that into account.

May 11, 2015, 20:24		#3
Lemanes New Member LManes Join Date: Oct 2014 Posts: 13 Rep Power: 11	Thank you. Your answer was very helpful. And just one more question: In future, I will change my t_hot boundary condition to heat being generated at the wall at an specific rate. Therefore, I would not know my average temperature. In which temperature would I define my fluid?

May 12, 2015, 06:48		#4
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,703 Rep Power: 143	If you don't the temperature then use variable properties and it will take care of it automatically. It will also handle the physical properties difference from the hot side to the cold side as well.

May 12, 2015, 07:13		#5
Lemanes New Member LManes Join Date: Oct 2014 Posts: 13 Rep Power: 11	Thank you again. I already set the function to change properties and it is working well. And in future, how can I use variable properties automatically? Is there any command?

May 12, 2015, 07:15		#6
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,703 Rep Power: 143	Save it as a CCL snippet so you can easily add it to any future simulation when needed.