[kirankarki]

Hi everyone,
can anyone tell me if openfoam is capable to handle the heat transfer mechanism through building envelope ? since heat transfer in building should include solar radiation due to outdoor temp, weather that varies with time, conduction through different layers of materials, convection due to natural ventilation air etc that determines the inside temp and heat flux. Which solver is good for this or new solver is to be created myself?

Thanks

[eugene]

You are pretty much in create new solver territory. To treat the wall heat transfer in a transient way, you pretty much have to do some form of conjugate heat transfer as well. If a steady approach to wall heat transfer is ok, then you just need to couple the inside and outside via a layer based resistance.

Let me know if you need a hand, I might have some appropriate code lying around.

[kirankarki]

Thanks eugene for response,
i m very new to this openfoam and i m construction management major thus no effecient background of programming ,i want to have transient heat transfer through wall and the outside weather data like temp, wind vel changes every hour, if u have some solvers for this case to send me, it will be highly appreciated, also i need to know how many kind of files should i create for system , constant and time directory ,
thanks again

[eugene]

Unfortunately I do not have a ready made solution for you.

There is a however a thermal boundary condition called "wallHeatTransfer" that couples an outside temperature to the inside wall via a fixed transfer coefficient. Unfortunately the BC has several limitations: 1 it wont help you with solar radiation, 2. it assume laminar flow on the inside.

You can use this T boundary in conjunction with the rhoSimpleFoam and other "compressible" solvers.

Good luck,

[val46]

Hi eugene,

what do you mean by layer based resistance?
Is there a way to set a thermal resistance between 2 regions?

Regards,
Toni

[eugene]

No, there is no way to currently set layer based resistance between two regions. It isn't that hard to implement if you only want a steady state solution though. If you want a transient solution in the solid, you will need some kind of mesh for the solid.

This is something I would like to to do at some point in the future. If I ever get round to it, I will let you know.

[petroo]

Hello,

Originally Posted by kirankarki

can anyone tell me if openfoam is capable to handle the heat transfer mechanism through building envelope ? since heat transfer in building should include solar radiation due to outdoor temp, weather that varies with time, conduction through different layers of materials, convection due to natural ventilation air etc that determines the inside temp and heat flux. Which solver is good for this or new solver is to be created myself?

Maybe you might take a look at the EnergyPlus system that is completely dedicated to problems like this. All your requirements are dealt with, including instationary calculation over time. EnergyPlus is capable of individual room-resolved heat load calculations etc., and there is a rather large library of construction materials with their thermodynamic properties.

Regards,

Peter

[maysmech]

hi,
One question:

Can buoyantBousinesqPimpleFoam solver handle a room ventilation simulation with inlet velocity(Fan, Cooler, Heater, ...) and pressure outlet (Door) with temperature (Energy equation)?

i asked this question because i didn't see any tutorial file case same as the case i told and all of them are same as hot room with a hot section without any forced flow like fan.
Best,
Maysam

[eugene]

Yes it can. Your boundary conditions will depend on your version of OpenFOAM though.

[maysmech]

Originally Posted by eugene

Yes it can. Your boundary conditions will depend on your version of OpenFOAM though.

Thanks Eugene,
i need a solver to simulate a room ventilation with inlet and outlet. the most similar solver i found was Heat Transfer solvers like buoyantBousinesqPimpleFoam. but i don't know why it doesn't understand the flow temperature for example i run it for a room with inlet fan with colder T than room but the flow effects on temperature was not real physically i think it set temperature of fan as a fixed wall instead of setting inlet patch flow temperature as desired T. Also velocity contour shows velocity to top in below of fan and it couldn't be true.

[eugene]

The solver you used should give ok answers for the problem you described. Not sure why its not working.

[maysmech]

Originally Posted by eugene

The solver you used should give ok answers for the problem you described. Not sure why its not working.

Here is geometry of the problem.
It is a room with 6 inlet fans at ceiling with T=290K and U=0.1 m/s. A window with 305 K is shown and walls and door have zeroGradient temperature.



After solving for 120 sec with time step of 0.1 the solution is:


--


and in real it is not acceptable to have T over all other temperatures. also contour of U_Z shows velocity to top instead of down of room in below of fans
I don't know where does the problem come form? the first thing i think is the solver may be not suitable for simulating flow with inlet and outlet with temperature like modeling of a room ventilation.

[eugene]

Are you sure your fans are blowing into and not out of the room?

[maysmech]

Yes, in boundary cells of fans u_z is -0.1 but below of them u_z is positive to up and shows a circulating flow to both sides of fan.

[Hasselhoff]

Did you check your gravity settings? Tell us if you solved that problem, I was dealing with quite a similar problem some time ago.

[maysmech]

----------

[maysmech]

hi,
Gravity is -9.81 m/s2 as default.
when i increase the fans inlet velocity from 0.1 to 1 the velocity contour along z is more acceptable. i mean in u_z=-0.1 velocity near ceiling fans is to up but for u_z=-1 is to down and i think it is because of changing dominant between buoyancy (U to up) and inlet fan velocity (to down). I am not sure about this. Any more comment will be appreciated.

[Hasselhoff]

In my case, the cold air from the radiators sinks down to the floor as it is supposed to be. Maybe the 0.1m/s is just too weak of a flow in your case. Or maybe your domain bounding box is scaled too big?
Good luck..
/M

[maysmech]

The other problem is how can we define T B.C for windows.

I think fixedValue is not a good assumption for windows temperature.

[Hasselhoff]

For windows I use heat flux boundaries. What kind of turbulence model are you using?