Convergence/residuals with different turbulence models

MMatt · May 23, 2013, 13:32

Hello everyone,

I have an external aerodynamics case which I am simulating. However I am unsure which model to use. First, my mesh is solid, I have at least 95% of the cells being of perfect quality and all the regions have skewness angles lower than 85°.

Using an all y+ wall treatment, I had a look at the scalar scene to see that most of the surface is between 0 and 30, while the maximum that I have is below 200. Is that enough to say that the y+ assessment of the model is correct?

Secondly, the residuals that I have are relatively high (and I have other simulations with even higher ones) with quite a lot of oscillations as you can see below:

Is that acceptable? I am using a k-e model, explicit steady, segregated flow, but I am unsure if I am doing things right. If I need to provide more explanation please let me know.

Thanks for your help!

Paulh · May 23, 2013, 14:46

If you're using a Low Reynolds number wall model, you should probably keep your Y+ between 0.5 & 2.5. If you're using a High Reynolds number wall model, 30 to 200. The all Y+ model give you the most flexibility, but I'd stay away from values between 2.5 & 30. Info from CD-adapco.

For my external aero models, I create reports & monitors for vehicle Cd and 'wind tunnel' Inlet pressure. These are what I use to judge convergence. You might also look at the max velocity in the model - that also needs to settle down.

MMatt · May 24, 2013, 04:49

Quote:

Originally Posted by Paulh

If you're using a Low Reynolds number wall model, you should probably keep your Y+ between 0.5 & 2.5. If you're using a High Reynolds number wall model, 30 to 200. The all Y+ model give you the most flexibility, but I'd stay away from values between 2.5 & 30. Info from CD-adapco.

For my external aero models, I create reports & monitors for vehicle Cd and 'wind tunnel' Inlet pressure. These are what I use to judge convergence. You might also look at the max velocity in the model - that also needs to settle down.

Thank you for your answer!

My Reynolds number is about 1.3E7 so a high one but I am using an all y+ treatment. Ironically most of my values are between 2.5 and 30. Using 5 layers for the BL could change the y+ value?

CD is also what I am using for convergence, but never heard of the inlet pressure, on what do you base this? Your personal experience or do you have documentation? Finally to assess this pressure, are you using a "maximum" report on the inlet and plotting the pressure (sorry for the dumb question)? If so then the pressure varies only between 40 and 120Pa.

Cheers

sampathevs · May 25, 2013, 02:34

Hello,

For external aerodynamics, prefer using k-omega turbulence model. and you did not mention your Reynolds number. if the case if for supersonic or hypersonic, change the solver to coupled. else, continue with segregated.

and yea, as its an external aerodynamics problem, you should keep y+ ( al wall y+ treatment) . y+ maybe more than 30 as well for external aerodynamics. depends on your flow.

if your Reynolds number is high, change to coupled solver from segregated, density based from pressure based.!!

I hope this reply helped you .!

sampathevs · May 25, 2013, 02:35

and dont worry with the residual plot. try checking the parameters which you are looking for.

example: if your are solving to determine the pressure drop, then wait until the net pressure becomes constant. residuals alone cannot decide your convergence.

MMatt · May 25, 2013, 08:13

Thank you for your answer!

I have specified the Reynolds number in my previous message (1.3E7). What would be the benefit to change from segregated flow to coupled flow (always thought for external vehicle aero it should be segregated).

Also you are talking about density and pressure based. I am using a constant density physic model, is that what are you talking about? Sorry for my ignorance.

Cheers for your help!

May 23, 2013, 13:32	Convergence/residuals with different turbulence models	#1
MMatt Member MMatt Join Date: May 2013 Posts: 59 Rep Power: 12	Hello everyone, I have an external aerodynamics case which I am simulating. However I am unsure which model to use. First, my mesh is solid, I have at least 95% of the cells being of perfect quality and all the regions have skewness angles lower than 85°. Using an all y+ wall treatment, I had a look at the scalar scene to see that most of the surface is between 0 and 30, while the maximum that I have is below 200. Is that enough to say that the y+ assessment of the model is correct? Secondly, the residuals that I have are relatively high (and I have other simulations with even higher ones) with quite a lot of oscillations as you can see below: Is that acceptable? I am using a k-e model, explicit steady, segregated flow, but I am unsure if I am doing things right. If I need to provide more explanation please let me know. Thanks for your help!

Thread Tools	Search this Thread
Show Printable Version Email this Page	Search this Thread: Advanced Search
Display Modes
Linear Mode Switch to Hybrid Mode Switch to Threaded Mode

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
different zones, different turbulence models	gmwsy	FLUENT	5	June 17, 2020 15:42
Y plus for various turbulence models	taram	CFX	8	December 16, 2013 11:44
Zero Equation Turbulence models	stefan.gracik	OpenFOAM Programming & Development	3	April 17, 2013 14:12
KOmega Turbulence model from wwwopenFOAMWikinet	philippose	OpenFOAM Running, Solving & CFD	30	August 4, 2010 10:26
Discussion: Reason of Turbulence!!	Wen Long	Main CFD Forum	3	May 15, 2009 09:52

May 23, 2013, 14:46		#2
Paulh Member Paul Hancock Join Date: Mar 2009 Location: Bellingham, WA Posts: 31 Rep Power: 17	If you're using a Low Reynolds number wall model, you should probably keep your Y+ between 0.5 & 2.5. If you're using a High Reynolds number wall model, 30 to 200. The all Y+ model give you the most flexibility, but I'd stay away from values between 2.5 & 30. Info from CD-adapco. For my external aero models, I create reports & monitors for vehicle Cd and 'wind tunnel' Inlet pressure. These are what I use to judge convergence. You might also look at the max velocity in the model - that also needs to settle down.

May 25, 2013, 02:34		#4
sampathevs Member EMAVI V SURYA SAMPATH Join Date: Aug 2011 Location: Malaysia Posts: 85 Rep Power: 14	Hello, For external aerodynamics, prefer using k-omega turbulence model. and you did not mention your Reynolds number. if the case if for supersonic or hypersonic, change the solver to coupled. else, continue with segregated. and yea, as its an external aerodynamics problem, you should keep y+ ( al wall y+ treatment) . y+ maybe more than 30 as well for external aerodynamics. depends on your flow. if your Reynolds number is high, change to coupled solver from segregated, density based from pressure based.!! I hope this reply helped you .!

May 25, 2013, 02:35		#5
sampathevs Member EMAVI V SURYA SAMPATH Join Date: Aug 2011 Location: Malaysia Posts: 85 Rep Power: 14	and dont worry with the residual plot. try checking the parameters which you are looking for. example: if your are solving to determine the pressure drop, then wait until the net pressure becomes constant. residuals alone cannot decide your convergence.

May 25, 2013, 08:13		#6
MMatt Member MMatt Join Date: May 2013 Posts: 59 Rep Power: 12	Thank you for your answer! I have specified the Reynolds number in my previous message (1.3E7). What would be the benefit to change from segregated flow to coupled flow (always thought for external vehicle aero it should be segregated). Also you are talking about density and pressure based. I am using a constant density physic model, is that what are you talking about? Sorry for my ignorance. Cheers for your help!