I read confusing answers... laminar flow is NOT a model but is defined as a state (a regime) of the flow. Laminar, transitional and turbulence regimes obey to the SAME governing equations, the Navier-Stokes equations. Models in such equations are the constitutive laws (Newton, Fourier) and the model for a gas. The turbulence models are required only if you tranform the original NSE by applying averaging or filtering.

Therefore the answer is the if one assumes the flow problem as laminar, no transformation of the NSE is required and no closure model must be introduced. Using RANS for laminar flows is wrong. The turbulence model in genera does not vanish and can affect the solution.

Hi,

If the flow is laminar, then mass,momentum, energy and species conservation equations are sufficient to solve the problem.

Reynolds number

https://en.wikipedia.org/wiki/Laminar_flow

I strongly suggest to read a textbook of fluid mechanics before using Fluent. You need a minimum background in this field and understand the Navier-Stokes equations and some basic classical exact solution for laminar flows.

Regards

It is hard to simulate Navier-Stokes (NS) equations(Flow governing equations) directly so we approximately solve NS equations by means of Reynolds averaging by splitting any flow variable to Mean and Fluctuating Parts.

For laminar flow the solver considers only the mean values.

For turbulent flow the solver considers both parts and approximate the fluctuating part through the so-called turbulence models.

For more information, check this Re..

J. Blazek, Computational Fluid Dynamics: Principles and Applications, 2nd Edition, 2005

Hi!, the equations used in the solver are the classical equation systems (Continuity, Momentum and Energy), using as Omer said it, the mean values. If is still unclear I highly recommend consulting the book: "Fluid Mechanics" by Pritchard-Fox-McDonalds, in the section of fundamental equations

Laminar flow strictly means that there is no fluctuation of velocity. For the fluctuations we use RANS modeling, and for closer of that equation we use turbulence models.

If there wont be any fluctuation then there is no need to solve the turbulence equation. So simply select the laminar option.

Many thanks for your kind reply ...

Mr. Vigneshwaran Sankar , Mushtaq Ahmad , Filippo Maria Denaro , Omer Musa , Oscar Alcocer , and Vikash Kumar Singh Chauhan.

Regards

With certain simplification, Euler equations can give you some approximate solution. Blasius solution is best laminat flow solution. But first you should check flow Reynolds number.

what are your flow assumptions? viscose or inviscose also, the geometry for your problem 2d or 3d, if you specify your case you can write your govern equations for your problem. regards.

I read confusing answers... laminar flow is NOT a model but is defined as a state (a regime) of the flow. Laminar, transitional and turbulence regimes obey to the SAME governing equations, the Navier-Stokes equations. Models in such equations are the constitutive laws (Newton, Fourier) and the model for a gas. The turbulence models are required only if you tranform the original NSE by applying averaging or filtering.

Therefore the answer is the if one assumes the flow problem as laminar, no transformation of the NSE is required and no closure model must be introduced. Using RANS for laminar flows is wrong. The turbulence model in genera does not vanish and can affect the solution.

Since laminar flow is special case of the flow depending on Reynolds number, so RANS approach could be used to simulate it. How?

Simply, laminar flow uses Mu_t=0 (Turbulent Eddy Viscosity) then no turbulence model needed. So, the solver employs ONLY mean values and ignores the fluctuating parts.

If you are writing your own code (as I did) you will see that JUST putting Mu_t=0 is enough to transfer your code from simulating turbulent flow to LAMINAR flow.

Even in Ansys-Fluent, I don't think that they're so stupid to write NS equations TWICE , one for laminar flow and repeat them for Turbulent flow.

Mohsen,

Euler equations are used ONLY for inviscid flow (viscosity = 0).

Omer,

I don't want to debate if people working in Fluent are stupid or not.

The approach of setting mu_t=0 in RANS to solve laminar flow can be stupid or not, depends on what you want to solve.

For example, if you want to solve the laminar vortex shedding that is unsteady in its nature and one uses the steady RANS equations just setting mu_t=0 the results will be wrong.

And in general, the eddy viscosity assumption models the anysotropic part of the velocity gradient while the isotropic part is just enclosed into the pressure. For incompressible flows that has small relevance, changing only the meaning of the field obtained from the Poisson equation. But for compressible flows, it could be considered that a model for the isotropic part could be added. Therefore setting only mu_t=0 can be not sufficient.

Finally, I want to stress that the numerical discretization for the RANS equation is generally of low order of accuracy. Therefore, a code written for RANS can poorly work on some laminar cases where accuracy can be relevant.

Thus, I let you to decide what is stupid.

Regards

Well done Sankar.

Why anyone would try to solve unsteady laminar vortex shedding using steady RANS?

Omer's answer, making mu_t = 0 for laminar case, is self explanatory that if one need to solve unsteady case then making mu_t = 0 in URANS is meant by him.

(This same answer was accidentally deleted since earlier I answered using my mobile. so I am adding again)

As I wrote before, some solutions for laminar flows require to have an accuracy order higher than that one used in the RANS code. If I want to write a code for RANS I don't take much care in write high order discretization (still nowaday you can find RANS with first or second order upwind in space and low order in time).

And if you have a RANS code modelling also the isotropic part of the tensor, you should also eliminate it, non only set mu_t=0.

Dear Experts,

• Case is 2D in nature.
• Calculated Reynold Number is between 5-30.
• Physics is - one air bubble is generated from very small needle in static solution of glycerol (highly viscous medium). Some vortex shedding is seen both side of bubble.

Regards

Muhummad, I recommend you look for the low Reynolds solutions, it could help you to understand the physics a little more.

In the first to differ between laminar and turbulence we regime we calculate the Raynolds NUmber . For Laminar flow strictly means that there is no fluctuation of velocity.

In RANS approch to modelite the fluctuations we close equation by using turbulence models.

Dear Filippo Maria Denaro

Thank you.

Yes, it is the right idea to read a textbook of fluid mechanics before using Fluent,Because there are different phenomenas to learn and to know .

There seems to be a lot of confusing answers here.

Laminar flow is a not a MODEL, but a state of flow, as mentioned by Filippo earlier.

There are many states of flows, depending on the Reynolds number. They are Stokes flow, laminar, transitional to turbulent and fully turbulent.

All of them are governed by the same equation, i.e. Navier Stokes equation (under suitable assumptions).

For solving different states of flow, we change the form of the NS equations. (Boundary layer equations for laminar, RANS for turbulent etc ).

In RANS also, there are two types of averaging : time-averaging for incompressible and Favre averaging for compressible flow.

You should read a good Fluid Mechanics textbook to understand these things. Boundary Layer Theory by H. Schlichting is a very good reference.

Hope this helps.