[rere]

Hiya,

i'm trying to determine Reynolds number for flow over a 2d airfoil.
The airfoil needs to be at low-speed with no shockwaves so I've concluded it's transonic.

The medium will be air, and the length is 1m.

Transonic flow = 0.8 -1.2 mach, so ive chosen 0.8 so there is no shockwave.
The flow speed i have selected as 10m/s, just as a low value.

do i use this equation to find re's number: Re = ul/v*
u = flow speed, l = chord, v = kinematic viscosity
*I'm not sure if I'm meant to calculate the flow speed or assume it, therefore unsure if this is the correct formula to use.
how do i then calc the boundary layer thickness for the meshing?

[DarylMusashi]

You assume transsonic flow with Ma = 0.8. This is your freestream velocity u you use to calculate Re number.

Ma * SQRT (kappa* R_air * T) = u.

kappa = 1.4 [], R_air=287 [J/(kg*K)], T [K] is the temperature you assume at freestream conditions.

The formular for Re= ul/v is correct.

If you are not very familiar with the theory maybe one of the online y+ calculators may help you to approximate you desired mesh size at the wall. (for example http://www.pointwise.com/yplus/).

Keep in mind that it depends on your turbulence model, which y+ value is the right one for you. If your turbulence model uses wall-functions (e.g. k-epsilon) y+ must not be smaller than 30.
If you want to use a low-reynolds-approach (k-omega, SST-model) to calculate the flow within the boundary layer without using wall-functions the mesh must be much finer: y+=1 is your target value. Additionally the mesh expansion factor in the boundary layer should not exceed 1.2.

[rere]

Originally Posted by DarylMusashi

You assume transsonic flow with Ma = 0.8. This is your freestream velocity u you use to calculate Re number.

Ma * SQRT (kappa* R_air * T) = u.

kappa = 1.4 [], R_air=287 [J/(kg*K)], T [K] is the temperature you assume at freestream conditions.

The formular for Re= ul/v is correct.

If you are not very familiar with the theory maybe one of the online y+ calculators may help you to approximate you desired mesh size at the wall. (for example http://www.pointwise.com/yplus/).

Keep in mind that it depends on your turbulence model, which y+ value is the right one for you. If your turbulence model uses wall-functions (e.g. k-epsilon) y+ must not be smaller than 30.
If you want to use a low-reynolds-approach (k-omega, SST-model) to calculate the flow within the boundary layer without using wall-functions the mesh must be much finer: y+=1 is your target value. Additionally the mesh expansion factor in the boundary layer should not exceed 1.2.

hi DarylMusashi,

Thank you for replying.

my tasks requires that i show some really good meshing for a 2d airfoil at low speed with NO SHOCKWAVES.

----Please tell me if anything is incorrect from this point on:----
So, i have gone for:

Flow is transonic, mach 0.8.
The fluid medium is air.
The chord length is 1m.
The flow speed is
Ma * SQRT (kappa* R_air * T) = u
= 0.8 *SQRT (1.4* 287 *288)
= 272.139

Reynolds number is:
= Re= ul/v
= (272.139 *1) / 1.5111E-5
= 18,009,331

Q1. How do i Determine the flow status (laminar or turbulent flow?

Q2. sO, i choose the right boundary layer thickness calculation formula by establishing which one of the turbulence models i should use?
i read up on the applications of k-omega, k-epsilon, SST etc and it seems SST is recommended for flat plat airflow modelling.
How do I go about implementing this in my mesh?
Thanks

[DarylMusashi]

Hi rere,
Your Re calculation is correct.

Answer Q1:
The transition from laminar to turbulent flows depends on different factors, one is the geometrical shape of the body. But as you have Re = 18 Mio you can definitely be sure to have turbulent flow (thats why you need to use a turbulence model).

Q2:
Very plain and simple, but good to remember:
k-epsilon: good results far away from the body (but not near the body)
k-omega: good results near the body (but not far away)
SST: combines k-epsilon and k-omega model with a blending function. It switches from k-omega near the body to k-epsilon far away from the body. So it combines the best of these two turbulence models.

If your mesh is very large or if you encounter stability problems you could also try to use a less advanced one-equation model, for example the Spalart-Allmaras model. It was created especially for turbulent flows over airfoils and blades of turbomachinery.

How do you implement this in your mesh - you must select the turbulence model of your choice in your solver settings of your software-package. Which software-package do you use? Or do you write your own code...?

Greetings,
Holger

[rere]

Originally Posted by DarylMusashi

Hi rere,
Your Re calculation is correct.

Answer Q1:
The transition from laminar to turbulent flows depends on different factors, one is the geometrical shape of the body. But as you have Re = 18 Mio you can definitely be sure to have turbulent flow (thats why you need to use a turbulence model).

Q2:
Very plain and simple, but good to remember:
k-epsilon: good results far away from the body (but not near the body)
k-omega: good results near the body (but not far away)
SST: combines k-epsilon and k-omega model with a blending function. It switches from k-omega near the body to k-epsilon far away from the body. So it combines the best of these two turbulence models.

If your mesh is very large or if you encounter stability problems you could also try to use a less advanced one-equation model, for example the Spalart-Allmaras model. It was created especially for turbulent flows over airfoils and blades of turbomachinery.

How do you implement this in your mesh - you must select the turbulence model of your choice in your solver settings of your software-package. Which software-package do you use? Or do you write your own code...?

Greetings,
Holger

NS equation question