[jeanpinto24|]

Dear all, in the classroom simulation with people I must assign a warmth of people. I have seen the BC heat transfer solver called solidWallHeatFluxTemperature. My idea is to create a new BC that allows me to take into account the heat delivered by people to the environment. Do you recommend that I perform this task? Thank you very much in advance.

Simulation done:

https://drive.google.com/open?id=0B89sG2XRPciIcjNYMVVnUTVPQVU

[piu58]

As you might know already, every person is calculated heating with 100 W. For a classroom is may be sufficient to apply the total power using the ground. If you need to calculate the air stream more accurately it may be necessary to model every person in form of a simple body like s sphere or a cube.

[jeanpinto24|]

The heat flow must be in W / m2, so I have to divide that power by the area of the person. My doubt is that I have no knowledge in programming in C ++, I need to include the BC wallHeatTransfer to solver buoyanBoussinesqPimpleFoam and I do not know very well the steps that I have to perform. The BC in question is in the compressible folder. This BC of compressible fluids can be included in an incompressible fluid solver?.

Thank you very much for your answers.

[piu58]

I don't see a problem here. Divide the total power by the are of the floor and use the ready available boundary conditions. In a similar way you may apply the power to a replacement body of your choice.

[jeanpinto24|]

The BCs available for assigning a heat flux have as input variables thermophysical properties such as density of the wall and thermal conductivity. Clearly, in this case I do not need those properties, I just need to allocate a constant heat flow. On the other hand, in the books of CFD comment that a wall can be applied a temperature or heat flow (not the two BC), because the system is overstocked. It occurs to me to create a condition BC that assigns a heat flow, where the temperature will be assigned with the condition "zeroGradient". The new BC comes to mind with the BC of U, because in this last condition of BC can be assigned fixed values. I do not have the current knowledge to perform these. In this part it is costing me for not having the programming knowledge.

[piu58]

Dear Jean,

I did not work with heat transfer until know, but the b.c. you need is an elementar one. So I thought it should be readily available. I looked a bit in this forum and fount that you may use

externalWallHeatFluxTemperature

for your problem. May be you find this text helpful, which vovers this b.c.

Unfortulately, I had not the time to try it for myself.

[akidess]

externalWallHeatFluxTemperature is more complicated than necessary. If you don't need to compute the temperature within the people, and you know the heat generation, and you know the people's surface, and you assume that the heat release is distributed evenly over the peoples skin, all you need is a fixedGradient. Alternatively, since people's body temperature is more or less constant, you could apply a fixedValue condition on the skin. By the way Jean, this is not at all a programming problem (since you've mentioned that repeatedly). What you are missing is spending more thought on how you want to mathematically describe your problem and what you want to compute in the end.

[jeanpinto24|]

Dear Anton:

Today in the morning investigate a way that finally did not work consisting of: add a term related to the heat delivered by people (104 W / m2) to the energy equation, create a scalar field with initial condition and BC called Q (Term added to the energy equation) and add this field to the solver createFields.H file. The first error was related to the units, make a change from kg / s3 (W / m2) to K / s for the solver to run. I ran the solver, visualized the results in Paraview and were unrealistic results. The idea I have is to simulate people in the first instance as regular hexahedrons (6 faces), on one side assign a constant value of heat flow and the other 5 assign the body temperature (298ºK). Now what you recommend to me to use the fixedGradient condition I do not know how to adapt it to my problem. On the other hand, in the classroom there are 20 people, that is, I have a heat of 2080 W / m2 that the fluid air will win. Could you give me some more guidance, please?

[akidess]

First, please explain how you went from 104 W/m2 to kg / s3 (W / m2) and finally to K/s. I'm curious.

Second, are you familiar with Fourier's law? Can you use it to determine the temperature gradient for a given heat flux?

[jeanpinto24|]

The fundamental units of W (watt) are (kg * m2) / s3. If I divide the unit W in m2 (W / m2) it turns out that the units are kg / s3. When you enter these units into the created scalar field (Q), the solver throws an error and recommends changing the units to ºK / s. In that case, you recommend me to use the law of Fourier, the doubt that arises is with respect to the value of thermal conductivity, is used the vallor k of the fluid or person. On the other hand, I have another question regarding the types of heat transfer. I read in a paper that inside the human body exists the conductivity of the heat, and in the surroundings and the outer surface there is convection and radiation. Now, why do you recommend me to use Fourier's law (conductivity) and does not take into account the other heat mechanisms?

[akidess]

Originally Posted by jeanpinto24|

The fundamental units of W (watt) are (kg * m2) / s3. If I divide the unit W in m2 (W / m2) it turns out that the units are kg / s3. When you enter these units into the created scalar field (Q), the solver throws an error and recommends changing the units to ºK / s.

Have you thought about the equations you are trying to solve, and that a quantity with units kg/s3 is not equal to an amount of heat?

In that case, you recommend me to use the law of Fourier, the doubt that arises is with respect to the value of thermal conductivity, is used the vallor k of the fluid or person. On the other hand, I have another question regarding the types of heat transfer. I read in a paper that inside the human body exists the conductivity of the heat, and in the surroundings and the outer surface there is convection and radiation. Now, why do you recommend me to use Fourier's law (conductivity) and does not take into account the other heat mechanisms?

You ask about convection: How large is the velocity at the boundary?

You ask about the thermal conductivity to use: Which side of the boundary are you interested in?

I recommend you get away from your computer, grab a piece of paper and a pen, and write down the heat balance you wish to investigate (1D). In fact, you can probably get a reasonable estimate with pen&paper to verify your simulation results later.

[jeanpinto24|]

http://editorial.dca.ulpgc.es/ftp/am...%20ambient.pdf

[jeanpinto24|]

Apply the Fourier law in this case, to obtain the FixedGradient of surface temperature of the human body. Inevitably his appreciations left me more doubts. I'm thinking after calculating the FixedGradient, How I add the heat gain in the fluid product of the heat emitted by the people?.

[akidess]

The surface heat flux *IS* the heat gain.

[jeanpinto24|]

What do you mean by * IS * ?. I was thinking of adding the heat gain in the energy equation. I'm not sure if it's according to the problem I'm dealing with.

[jeanpinto24|]

I have found the possible solution to my problem, in the solver chtMultiRegionSimpleFoam when using a radiation model, a new BC Qr with units of W / m2 is defined. My idea is to change the name of that BC from Qr to Qpeople and then simulate people in the first instance as a regular 6-sided hexahedron. To the top face include the BC Qpeople and the others to set them with a fixed temperature. I await your comments regarding this possible solution.

[piu58]

Dear Jean, this solution is what I mentioned in my first answer: replacing the people by simple geometric bodies. I wish you much success.

[jeanpinto24|]

I have searched without getting results, the point I am researching is the last of my degree thesis. I do not really know what to do. A foamer told me that the BC to allocate a heat flow is already available.

[jeanpinto24|]

Dear all:

I'm really impatient. I'm simulating heat transfer in a classroom with people. I simulate people like a six-sided hexahedron. I need the top face to have an outgoing heat flux of 104 W / m2. I have searched the forum and I think wallHeatFluxIncompressible is indicated, as I am using buoyantBoussinesqPimpleFoam. What do you recommend? I look forward to your help.

[Moderator note: Moved from the thread wallHeatFlux utility for an incompressible case]

[piu58]

Dear Jean, if you don't bring the b.c. to work, but need a result:

Why don't you use the fixed temperature b.c., calculate the total power transfer and adjust the temperature a few times until your power is in the region you want to have it? Not an elegant solution, but at least a solution-