Problem in implementing cht

tilek · May 6, 2011, 12:29

Hi colleagues,

I made tests regarding implementing conjugate heat transfer problem.

Two heaters (5x5mm^2) are attached to an insulator (k=0.002W/m.k, Cp=100J/kg.K, 5X100mm^2), which is itself attached to a steel sheet metal. Only one of the heaters is now heated and is kept isothermal at 327K at its bottom, and the steel's top is maintained at 300K; solid-solid interface for insulator-heaters and another for insulator-sheet metal are created, the rest of the external surfaces are kept at default condition, i.e. insulated. Conservative heat flux is used. Fig 1 below shows the geometries.
fig1.jpg
a

The result is as expected, with the isotherms plot attched here (Fig 2).
fig2.jpg

Then, I tested another case. Here, I just changed the the material at the top to air, instead of steel in the earlier case. In the actual problem I am working on, there air has velocity (assume for instance a Blasius boundary layer is developing), a room heater. But, to simplify the problem the air has zero velocity now (i.e. it is only conducting heat, not convecting). Now, a fluid-solid inteface is used, and a symmetry boundary condition on the lateral sides is added to make it 2D flow problem. Steady state, laminar, thermal energy options are used for the liquid.

Unfortunately the result is odd. It shows as if both the heaters are heated, which is not the case (only one of the heaters is heated, as described above). The isotherm plot is attached here (Fig 3).
fig3.jpg

Can some one help me here. A problem creeps in when a liquid is used in the problem.

Thanking you in advance.

ghorrocks · May 7, 2011, 09:03

There is some error in your setup causing the second heater to get hot. Check your heat source definitions.

tilek · May 7, 2011, 17:18

Hi Glen,

As I have said this problem occured only when a fluid-solid interface is introduced. This problem never occured when all domains were all solid, and the second heater setting would have shown the error for this condition if it were the reason for the problem. I think I have explained in detail how I implemented the solid-liquid interface, all with the default settings except for the laminar, thermal energy (instead of isaothermal fluid), etc. options checked. Further, I have over simplified the geometry so that someone from this froum could easily reproduce the setup and let me know where I went wrong. I have tried differnt cases for the liquid motive source, with a forced and a natural convection case but the problem still remains there. Please also note that I haven't used a source option for the fluid, instead an isothermal boundary condition is applied at its far normal edge (with the rest of the fluid edges left with the defaults (insulated) conditions). I can also send the relevant files if someone want to run it.

Regards,

ghorrocks · May 8, 2011, 09:39

That may be the case, but the results show the second volume has a heat source applied to it and the solver is being told to do that somewhere. Either your setup is wrong or CFX has a bug in it. If you reckon your setup is correct then CFX has a bug, so you should a) report it to support to get it fixed and b) try to find a work-around. I bet splitting the interface into two interfaces (one for the left and right heaters) will work around the problem.

May 6, 2011, 12:29	Problem in implementing cht	#1
tilek New Member kelita Join Date: May 2011 Posts: 4 Rep Power: 15	Hi colleagues, I made tests regarding implementing conjugate heat transfer problem. Two heaters (5x5mm^2) are attached to an insulator (k=0.002W/m.k, Cp=100J/kg.K, 5X100mm^2), which is itself attached to a steel sheet metal. Only one of the heaters is now heated and is kept isothermal at 327K at its bottom, and the steel's top is maintained at 300K; solid-solid interface for insulator-heaters and another for insulator-sheet metal are created, the rest of the external surfaces are kept at default condition, i.e. insulated. Conservative heat flux is used. Fig 1 below shows the geometries. fig1.jpg a The result is as expected, with the isotherms plot attched here (Fig 2). fig2.jpg Then, I tested another case. Here, I just changed the the material at the top to air, instead of steel in the earlier case. In the actual problem I am working on, there air has velocity (assume for instance a Blasius boundary layer is developing), a room heater. But, to simplify the problem the air has zero velocity now (i.e. it is only conducting heat, not convecting). Now, a fluid-solid inteface is used, and a symmetry boundary condition on the lateral sides is added to make it 2D flow problem. Steady state, laminar, thermal energy options are used for the liquid. Unfortunately the result is odd. It shows as if both the heaters are heated, which is not the case (only one of the heaters is heated, as described above). The isotherm plot is attached here (Fig 3). fig3.jpg Can some one help me here. A problem creeps in when a liquid is used in the problem. Thanking you in advance.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Problem Importing Geometry ProE to CFX	fatb0y	CFX	3	January 14, 2012 20:42
Problem implementing DETAILED REACTION MECHANISM	Phil	FLUENT	0	May 5, 2007 12:38
Adiabatic and Rotating wall (Convection problem)	ParodDav	CFX	5	April 29, 2007 20:13
CHT strange problem	MatejK	CFX	3	September 1, 2005 06:32
CHT problem	sch	CFX	2	July 11, 2003 11:32

May 7, 2011, 09:03		#2
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,870 Rep Power: 144	There is some error in your setup causing the second heater to get hot. Check your heat source definitions.

May 7, 2011, 17:18		#3
tilek New Member kelita Join Date: May 2011 Posts: 4 Rep Power: 15	Hi Glen, As I have said this problem occured only when a fluid-solid interface is introduced. This problem never occured when all domains were all solid, and the second heater setting would have shown the error for this condition if it were the reason for the problem. I think I have explained in detail how I implemented the solid-liquid interface, all with the default settings except for the laminar, thermal energy (instead of isaothermal fluid), etc. options checked. Further, I have over simplified the geometry so that someone from this froum could easily reproduce the setup and let me know where I went wrong. I have tried differnt cases for the liquid motive source, with a forced and a natural convection case but the problem still remains there. Please also note that I haven't used a source option for the fluid, instead an isothermal boundary condition is applied at its far normal edge (with the rest of the fluid edges left with the defaults (insulated) conditions). I can also send the relevant files if someone want to run it. Regards,

May 8, 2011, 09:39		#4
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,870 Rep Power: 144	That may be the case, but the results show the second volume has a heat source applied to it and the solver is being told to do that somewhere. Either your setup is wrong or CFX has a bug in it. If you reckon your setup is correct then CFX has a bug, so you should a) report it to support to get it fixed and b) try to find a work-around. I bet splitting the interface into two interfaces (one for the left and right heaters) will work around the problem.