[jeanpinto24|]

dear all

I am working on my undergraduate project, simulating a classroom with two entrances and 4 exits, my question is related to how to define the number of reynolds in the entrance to define if it is turbulent or laminar flow. From what I've learned, OpenFoam is binary, that is, it simulates laminar or turbulent flows, not intermediate flows.

My teacher told me that in the first instance would not be valid the relation number reynolds for pipes or open channel.

Geometry:

https://drive.google.com/open?id=0B8...HQ5UTdqUVZhN2M

The solver used is buoyantBoussinesqPimpleFoam that for resolution of the fluid equations is the RANS method.

Best regards,
Jean.

[lukas.lebo]

Hey Jean

The Reynolds number is a qualitative estimation on how much momentum transfer you have by convection vs diffusion. In high Reynolds-number flows you have high momentum convection and comparably low momentum diffusion.

The Reynolds-number can be used to characterize when a flow becomes unstable/turbulent. In pipe-flows this is well explored to the point that we know that Re = f(u,D,nu) = ~2300 for turbulence to start.

The turnover point / critical Reynolds-number depends on the case respectively the geometry. If your geometry is not well explored you have to find it out for yourself and can take an arbitrary reference length-scale. Since all length-scales are in the same order of magnitude (height, width of inlets/outlets or the room itself) it doesn't matter at all. If someone else did a very similar geometry he then could use your results for the critical reynolds number.

In many cases a rough estimation of the magnitude is already enough to make an educate guess if the flow is turbulent. You will very likely find out that you have to consider turbulent behavior and depending on your goals you will then decide on a RANS or LES model. As far as I know there is nothing wrong in using a turbulent model for laminar or near laminar cases (but don't quote me on that).

I hope this helped you.

[jeanpinto24|]

Hi dear

You can recommend a book to me to calculate the critical reynolds number for unexplored geometries.

With the idea of defining if my flow is laminar or turbulent is the beginning of my undergraduate project.

[lukas.lebo]

I'm not sure you will find what you are looking for anywhere. The point of an unexplored geometry is as the name suggests that it is not well studied.

But you can estimate the reynolds number and if its not below 10^3 you better expect turbulent behavior. You can do a RANS simulation and investigate the turbulent kinetic energy (k variable) and thus evaluate how much turbulence you have in your domain. Though it would be based on some assumptions i.e. your turbulence intensity at the inlet. You could do such a study for different inlet intensities and then make a statement on wether you would expect turbulent or laminar behavior in your domain.

Cheers,

Lukas

[jeanpinto24|]

Perform the calculation of the reynolds number with the following estimated data:

Velocity = 0,1 m/s

Length characteristic = 0,8 m

Kinematic viscocity = 0,000015 m2/s

The number of reynolds gives me the following value: 5333,333

So, using the RANS model to solve the fluid equations would be valid considering its consideration of having a number of reynolds greater than 1000. Clearly, it is an estimate and first instance calculation.

[piu58]

There is not such a thing like a Reynold number for a classroom, Osborne Reynolds never considered classrooms . But his theory of similarity says that problems have can be transfered to another length scale as long as the the Reynolds number is equal. The characteristic length has to be selected by you.

You may use on of the sizes of the room or the space diagonal. The Reynolds number differs slightly if you change. In dependence of what you what you want to find out it may even be useful to use the size of the door or the windows for the Reynolds number.

[jeanpinto24|]

My question is that the buoyantBoussinesqPimpleFoam solver used the RANS model that is used for turbulent fluids, so I want to know the criteria for selecting the Reynolds number in a classroom. According to my calculations Reynolds number gives me values over 5000, therefore, in the first instance it would not be wrong to use the RANS model in this case.