Heat transfer between solid and fluid domain

gartz89 · February 24, 2018, 04:03

Hello guys, I would like to simulate transient heat transfer between solid and fluid domain. I've got simple geometry one pipe (solid and fluid domain). In fluent, the model work with energy equation, k - epsilon turbulent model, the materials is ceramics and air. My BC is velocity inlet and temperature, outlet pressure outlet, I've got shadow wall interface (meshing interface coupled) for CHT, for outside walls I would like to use adiabatic wall (symmetry or zero flux W/m^2), because I would like to know my temperature field in the solid body and what happens. For example air flows in the pipe for 5 seconds, and the temperature of the body is lower from the air.The heat transfer between the fluid domain and solid domain for many seconds becoming steady state and the outlet temperature become equal to the inlet temperature. My question is what BC I have to specify for this problem. I don't know how to specify vary temperature in the solid domain, because I don't want to fixed values. Could you help me guys, please?

gartz89 · February 25, 2018, 02:15

I try to explain one regenerator (without porous media). When I specify temperature inlet and velocity inlet and walls are adiabatic. I have to generate some heat source in the solid domain, because this is my heat storage (accumulate heat from other process). Everything is alright, but I expect different results. In Fluent I have transient heat transfer, outlet temperature of the fluid decrease when I decrease time steps (seconds). I mean in many seconds the temperature have decreasing and this is our steady state condition. In my experimental results the temperature in the solid domain gives own energy source to the fluid domain (heat transfer). The outlet temperature in many seconds is approximately equal to the inlet temperature this is our steady state model and the temperature on the solid (wall) is decreasing in many seconds. In Ansys Fluent this is vice versa, how can I specify this experimental results in my BC.Should I have to specify some profiles or something else. Could you tell me, please.

gartz89 · February 28, 2018, 11:27

Must I write UDF? Please recommend something, give me some advice.

obscureed · March 2, 2018, 11:17

Hi gartz89,

Your set-up sounds OK to me, so I am not sure what changes to propose and I am not sure what you are asking. I do not see any obvious need for profiles or UDF for what you have described.

Well, you asked about how to get time-varying temperatures in the solid domain. This should arrive naturally in a Fluent simulation with transient and energy effects. Obviously you should check that the correct materials are assigned to each zone, and check the material properties -- but if you neglect this, you will get aluminum as the default solid material and you should see plenty of conduction.

One worry I have about your set-up is the non-conformal interface. These can be a bit tricky (and I avoid them whenever possible for this reason). Obviously you need to be sure to use the Coupled Wall option. There are some limited troubleshooting tools available: in particular, go to the "Create/Edit Mesh Interfaces" dialogue box and press List for a specific interface. This will show you the percentage of each zone that has been included in the interface, and the percentage that has reverted to adiabatic wall: check that these are as expected. If large fractions of the intended interface have not connected, this might explain your results. Go and read about the Matching and Mapped options in interface set-up, and check the geometry. (By the way, I have had very little success in editing an interface when it has been set up. I usually delete and start again.)

You might find that the time-step that is suitable for fluid flow simulation is small compared to your conduction-based timescales. If this is the case, you may need to experiment with "freezing the flow": run transient with flow+conduction equations and some small timesteps; then disable the flow equations (in Solution...Controls...Equations) and run some longer timesteps; then repeat these two steps.

The other issue you might need to think about is initialization. It probably makes sense to start from a converged steady-state flow-field, but you can change parts of this starting-point by Solution...Initialization...Patch. For example, the steady-state solution presumably has the solid zones at equilibrium with the gas (and this is a small test of your mesh interface), but you can patch the solid zones to a cold temperature to start the transient.

Good luck!
Ed

gartz89 · March 3, 2018, 04:30

God bless you obscureed for your reply. My main problem was Solution...Initialization and Patch the temperature of the solid body. When I did that I solved my model. But I meant to specify some profiles or UDF for my temperature field of the solid body for reverse flow. I mean the fluid flows for 10 second for example and reverse for 13 second for example, but if I want to reverse the flow the next step I think is to specify the temperature field like Profile or UDF I don't know. I'll look what is happening in the heating and cooling periods for the regenerator. Is there some other options, to add this temperature field, because I had done now. I want to specify the temperature field in the solid body in the reverse flow (the next model). Thank you again !!!

February 24, 2018, 04:03	Heat transfer between solid and fluid domain	#1
gartz89 Member Angel Penev Join Date: Apr 2016 Location: Bulgaria Posts: 47 Rep Power: 10	Hello guys, I would like to simulate transient heat transfer between solid and fluid domain. I've got simple geometry one pipe (solid and fluid domain). In fluent, the model work with energy equation, k - epsilon turbulent model, the materials is ceramics and air. My BC is velocity inlet and temperature, outlet pressure outlet, I've got shadow wall interface (meshing interface coupled) for CHT, for outside walls I would like to use adiabatic wall (symmetry or zero flux W/m^2), because I would like to know my temperature field in the solid body and what happens. For example air flows in the pipe for 5 seconds, and the temperature of the body is lower from the air.The heat transfer between the fluid domain and solid domain for many seconds becoming steady state and the outlet temperature become equal to the inlet temperature. My question is what BC I have to specify for this problem. I don't know how to specify vary temperature in the solid domain, because I don't want to fixed values. Could you help me guys, please?

March 2, 2018, 11:17		#4
obscureed Senior Member Join Date: Sep 2017 Posts: 246 Rep Power: 11	Hi gartz89, Your set-up sounds OK to me, so I am not sure what changes to propose and I am not sure what you are asking. I do not see any obvious need for profiles or UDF for what you have described. Well, you asked about how to get time-varying temperatures in the solid domain. This should arrive naturally in a Fluent simulation with transient and energy effects. Obviously you should check that the correct materials are assigned to each zone, and check the material properties -- but if you neglect this, you will get aluminum as the default solid material and you should see plenty of conduction. One worry I have about your set-up is the non-conformal interface. These can be a bit tricky (and I avoid them whenever possible for this reason). Obviously you need to be sure to use the Coupled Wall option. There are some limited troubleshooting tools available: in particular, go to the "Create/Edit Mesh Interfaces" dialogue box and press List for a specific interface. This will show you the percentage of each zone that has been included in the interface, and the percentage that has reverted to adiabatic wall: check that these are as expected. If large fractions of the intended interface have not connected, this might explain your results. Go and read about the Matching and Mapped options in interface set-up, and check the geometry. (By the way, I have had very little success in editing an interface when it has been set up. I usually delete and start again.) You might find that the time-step that is suitable for fluid flow simulation is small compared to your conduction-based timescales. If this is the case, you may need to experiment with "freezing the flow": run transient with flow+conduction equations and some small timesteps; then disable the flow equations (in Solution...Controls...Equations) and run some longer timesteps; then repeat these two steps. The other issue you might need to think about is initialization. It probably makes sense to start from a converged steady-state flow-field, but you can change parts of this starting-point by Solution...Initialization...Patch. For example, the steady-state solution presumably has the solid zones at equilibrium with the gas (and this is a small test of your mesh interface), but you can patch the solid zones to a cold temperature to start the transient. Good luck! Ed gartz89 likes this.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Question about adaptive timestepping	Guille1811	CFX	25	November 12, 2017 17:38
Error - Solar absorber - Solar Thermal Radiation	MichaelK	CFX	12	September 1, 2016 05:15
Question about heat transfer coefficient setting for CFX	Anna Tian	CFX	1	June 16, 2013 06:28
Error finding variable "THERMX"	sunilpatil	CFX	8	April 26, 2013 07:00
Convective Heat Transfer - Heat Exchanger	Mark	CFX	6	November 15, 2004 15:55

February 25, 2018, 02:15		#2
gartz89 Member Angel Penev Join Date: Apr 2016 Location: Bulgaria Posts: 47 Rep Power: 10	I try to explain one regenerator (without porous media). When I specify temperature inlet and velocity inlet and walls are adiabatic. I have to generate some heat source in the solid domain, because this is my heat storage (accumulate heat from other process). Everything is alright, but I expect different results. In Fluent I have transient heat transfer, outlet temperature of the fluid decrease when I decrease time steps (seconds). I mean in many seconds the temperature have decreasing and this is our steady state condition. In my experimental results the temperature in the solid domain gives own energy source to the fluid domain (heat transfer). The outlet temperature in many seconds is approximately equal to the inlet temperature this is our steady state model and the temperature on the solid (wall) is decreasing in many seconds. In Ansys Fluent this is vice versa, how can I specify this experimental results in my BC.Should I have to specify some profiles or something else. Could you tell me, please.

February 28, 2018, 11:27		#3
gartz89 Member Angel Penev Join Date: Apr 2016 Location: Bulgaria Posts: 47 Rep Power: 10	Must I write UDF? Please recommend something, give me some advice.

March 3, 2018, 04:30		#5
gartz89 Member Angel Penev Join Date: Apr 2016 Location: Bulgaria Posts: 47 Rep Power: 10	God bless you obscureed for your reply. My main problem was Solution...Initialization and Patch the temperature of the solid body. When I did that I solved my model. But I meant to specify some profiles or UDF for my temperature field of the solid body for reverse flow. I mean the fluid flows for 10 second for example and reverse for 13 second for example, but if I want to reverse the flow the next step I think is to specify the temperature field like Profile or UDF I don't know. I'll look what is happening in the heating and cooling periods for the regenerator. Is there some other options, to add this temperature field, because I had done now. I want to specify the temperature field in the solid body in the reverse flow (the next model). Thank you again !!!