I feel very headache to solve the below questions:

i. I always ask my self is there any standard to simulate room or building, for safety purposes of course.For fire size, what is the value should I put?

ii. If we put the radiation model, how good are the results would be?

iii. For fire, should we use steady fire?

iv. KE model, how good is the results by using this model?

Hi Loh, to answer you somebody must know the main question you have to the simulations.

If you want to simulate the smoke, there are different levels of modeling (zone models or field models).

I know there some free programs with zone models with different optional appromixations for heat and smoke release of the fire in the internet.

In germany (where i come from), there is a commercial package (kobra3d) on the market, which uses the ke-modell to simulate the spread smoke. So i think the ke-modell is applicable to this case. (I know not all things commercial packets make are right )

Greets

PS: please apologize my bad english

Let's go to the specific questions anyway. In fire simulations, a lot of factors will affect the results. KE model for example, how good will the results as compared to Chen-Kim KE-EP model and RNG-derived KE-EP and etc. Normally the engineering software like PHOENICS and Fluent are using FV method to solve these type of difference equations. One of the doubt might occurs, should we add radiation model inside the simulation if only to simulate the room? I am not very sure about these things. If we switch ON the radiation model, six flux model can be used. But another problem again, TIME. We have to solve three more variables and an entra boundaries. Time may not be the serious one, the most severe can be the absorption and scattering coefficients and etc.. What values should I put?

Now, go into more detail, what are the differences for each variable if we are now using radiation model? Let's say, I'm not concerning the all the variables except temperature? I can test to get a set of results to investigate these dependencies. For difference runs in the future, what are the things should I take as a standard?

Does anyone know how good is the simple(simplex) method? By assuming the walls as adiabatic, and velocities at the openings to be zeros, do these make sense?(using Simple method)

More interesting subject, auto-tuning to minimizing the errors of the equations. We talk about optimizing the convergence. What are the effects if we use lower and higher initial relaxations.

May I know is it a new trend to simulate solid stresses by using FV method? For my knowledge, normally FEM will be used in this area.

What kind of fire and what kind of room ????

CFD is not only switching between some features.

If you mean your living room with a burning candle you may switch off the radiation.

If you want to burn coal in a big power plant you have to turn on the radiation.

What fire size can we use? Woods fire size? If the fire size is 2Mw/m3(peak), can we switch on radiation if the room(Lab.) size equal to 18x9x2.8m? How big is the fire is consider as big a fire? Any limit or boundary to apply radiation model? Any suggestions for scattering and absorption coefficients for air in this case?

In the fire simulation, how can we consider ceiling failure? At what time should we swicth off the ceiling?(purpose of this simulation is to investigate solid stressess of steel structure above the ceiling.

Seems that your ultimate objective is to simulate stresses in steel structures above the ceiling in some "room" which has a fire burning in it.

I don't think you need to simulate a full fledged fire using CFD for that! If you look in a few references (Fire Technology journal; find more references in any article therein) you should be able to get heat release rates from fires. Use freely available zone models to determine the temperatures near the ceiling. Use these temperature values as input to calculate the thermal stresses in your structures. Zone models are more conservative than field models which means that they overpredict the temperatures anywhere from 5% to 15%, sometimes more. However if you use the model carefully, you can limit the error to an overprediction of 5%. This will be good 'cos it will give you an inbuilt "factor of safety" from an engineering standpoint, for the thermal stress calculation. Zone models do account for the radiant heat transfer from fires.

-Jay.

pleaseclarify more.thanks