[P_P]

Hello,

I am currently trying to determine the turbulent velocity profile with matlab for a flow between two parallel plates. In order to do so, I am using the universal velocity profiles of Karman and Taylor/Prandtl. The veloityprofile is considered to be symmetric.
In both cases the velocity gradients close to the walls show realistic behaviour. Still I have my doubts concerning the values on the centerline where the velocity profile shows a peak. Does anyone know models which represent a smoother transition on the centerline or rather a better approximation for high y+ numbers?
A plot of the velocity profile and the corresponding Matlabcode can be found in the appendix.
Every input is welcome

Greetings

P_P

[FMDenaro]

Originally Posted by P_P

Hello,

I am currently trying to determine the turbulent velocity profile with matlab for a flow between two parallel plates. In order to do so, I am using the universal velocity profiles of Karman and Taylor/Prandtl. The veloityprofile is considered to be symmetric.
In both cases the velocity gradients close to the walls show realistic behaviour. Still I have my doubts concerning the values on the centerline where the velocity profile shows a peak. Does anyone know models which represent a smoother transition on the centerline or rather a better approximation for high y+ numbers?
A plot of the velocity profile and the corresponding Matlabcode can be found in the appendix.
Every input is welcome

Greetings

P_P

Hello,
I suggest plotting the velocity u+ versus y+ so that you can better check for viscous sub-layer and log region

[P_P]

Hello FMDenaro,

thanks a lot for your fast reply. I have plotted u+ versus y+. The linear behaviour of the laminar sublayer can be seen for y+<11.5.
You can see the the plot in the appendix.

Greetings P_P

[FMDenaro]

Originally Posted by P_P

Hello FMDenaro,

thanks a lot for your fast reply. I have plotted u+ versus y+. The linear behaviour of the laminar sublayer can be seen for y+<11.5.
You can see the the plot in the appendix.

Greetings P_P

to check further, you can superimpose the law u+=y+ and the log law using a logarithmic scale

[P_P]

I don't know if I unterstood you well. I have created a semilogarithmic plot u+ versus y+. What is the idea behind checking the laminar sublayer for my problem?

[flotus1]

Looks reasonable.
The "peak" in your initial figure is due to the fact that the equations you used are for the flow over a flat plate with an infinite fluid domain. Not for a fluid between parallel plates.

[P_P]

Hello flotus1,
thanks for your reply. Do you know any models which offer a better desciription of the velocity profile between two parallel plates? Or is it unevitable to excute a CFD simulation with common CFD software in order to export the velocity profile.


P_P

[flotus1]

I havent come across any analytical formula for this case.
However there is plenty of DNS data available for turbulent flow between 2 parallel flat plates. Here is just one example...