Sudden change of inflow turbulence intensity Tu

airsupply · August 30, 2016, 02:23

I proposed one turbulence model that rely on the evolution of turbulent kinetic energy equation to capture transition. So far in several classic test cases it can reproduce bypass, natural and separation induced transition. However the paper reviewer said that the transition prediction was too fragile. One reason he claimed was multiple solutions for the same set-up could happen. More specifically, a solution started from a laminar initial condition would converge to a different solution than the same simulation staring from a fully turbulent set-up. With T3B flat plate boundary layer, I tried that and found that a fully turbulent set-up as the initial condition could lead to a transition location backwards, but the new result was different from both fully turbulent condition and the experiment.

I begin to wonder how DNS or experiments behave. As known, T3B was almost an in-compressible case so the fully developed turbulent boundary layer as the initial field would certainly propagate to the upstream and alter the transition location. Has anyone ever studied this?

duri · August 31, 2016, 23:43

This is interesting. Why do you expect transition location to be same with clear initialization (laminar) and contaminated initialization (turbulent). It doesn't happen in reality. How turbulent boundary layer become laminar before transition point in the second one.

airsupply · September 1, 2016, 06:35

Quote:

Originally Posted by duri

This is interesting. Why do you expect transition location to be same with clear initialization (laminar) and contaminated initialization (turbulent). It doesn't happen in reality. How turbulent boundary layer become laminar before transition point in the second one.

Thanks for your reply. Would you please tell me what you major in?

It is not me but the reviewer of my manuscript believes that a proper transition approach would surely yield the same results regardless of the initialization. I don't really know what you mean "It doesn't happen in reality", for we can reduce the inflow intensity during the wind tunnel test.

duri · September 1, 2016, 07:38

Laminar flow can become turbulent. But for the same condition relaminarization would not occur in nature. So when flow is initiallly turbulent it should continue to be turbulent. In your case it appears boundary condition drive the initial turbulent flow to laminar flow. which is not real.
In wind tunnel transition doesn't occur like full scale model. Hence tripping is used to create turbulent in the boundary layers.
Transition location is quite sensitive to external disturbance and turbulent intensity.
In your case, probably you need to figure out which terms in your equation or bounary condition forces laminar condition in a turbulent flow.

August 30, 2016, 02:23	Sudden change of inflow turbulence intensity Tu	#1
airsupply New Member Alvin Join Date: Jun 2011 Posts: 7 Rep Power: 15	I proposed one turbulence model that rely on the evolution of turbulent kinetic energy equation to capture transition. So far in several classic test cases it can reproduce bypass, natural and separation induced transition. However the paper reviewer said that the transition prediction was too fragile. One reason he claimed was multiple solutions for the same set-up could happen. More specifically, a solution started from a laminar initial condition would converge to a different solution than the same simulation staring from a fully turbulent set-up. With T3B flat plate boundary layer, I tried that and found that a fully turbulent set-up as the initial condition could lead to a transition location backwards, but the new result was different from both fully turbulent condition and the experiment. I begin to wonder how DNS or experiments behave. As known, T3B was almost an in-compressible case so the fully developed turbulent boundary layer as the initial field would certainly propagate to the upstream and alter the transition location. Has anyone ever studied this?

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Displaying Turbulence Intensity in LES and creating a constant Turbulence Intensity	Anwer	FLUENT	1	August 8, 2016 17:59
Question about turbulence intensity	Anna Tian	Main CFD Forum	8	July 15, 2013 19:19
specify the turbulence intensity and the turbulence length	bennoman75	Main CFD Forum	0	May 15, 2013 06:36
How to asses turbulence intensity	Kwiaci	FLUENT	0	September 3, 2012 06:48
Turbulence Intensity Vs CFD Simulation	Apple L S Chan	Main CFD Forum	3	December 15, 1998 18:28

August 31, 2016, 23:43		#2
duri Senior Member duri Join Date: May 2010 Posts: 245 Rep Power: 17	This is interesting. Why do you expect transition location to be same with clear initialization (laminar) and contaminated initialization (turbulent). It doesn't happen in reality. How turbulent boundary layer become laminar before transition point in the second one.

September 1, 2016, 07:38		#4
duri Senior Member duri Join Date: May 2010 Posts: 245 Rep Power: 17	Laminar flow can become turbulent. But for the same condition relaminarization would not occur in nature. So when flow is initiallly turbulent it should continue to be turbulent. In your case it appears boundary condition drive the initial turbulent flow to laminar flow. which is not real. In wind tunnel transition doesn't occur like full scale model. Hence tripping is used to create turbulent in the boundary layers. Transition location is quite sensitive to external disturbance and turbulent intensity. In your case, probably you need to figure out which terms in your equation or bounary condition forces laminar condition in a turbulent flow.