[CFD] The k-omega Turbulence Model
Vložit
- čas přidán 22. 06. 2020
- An introduction to the k - omega turbulence model that is used by all mainstream CFD codes (OpenFOAM, Fluent, CFX, Star CCM+, COMSOL, Saturne). The following topics are covered:
1) 5:10 When was the k-omega model developed?
2) 12:28 What is omega?
3) 15:20 Why is k-omega better for aerodynamics than k-epsilon?
4) 16:43 What is the freestream dependency of the k-omega model?
#kOmega #fluidmechanics101 #cfd
====================================================
Some useful references:
1) NASA Turbulence Modelling Resource - The Wilcox k Omega Model
turbmodels.larc.nasa.gov/wilc...
2) J. C. Kok 'Resolving the dependence on free-stream values for the k-omega turbulence model', NLR-TP-99295, July 1999.
reports.nlr.nl/xmlui/bitstrea...
3) F. Menter, 'Influence of freestream values on k-omega turbulence model predictions', AIAA Journal, 30(6), doi.org/10.2514/3.11115
============================================
Want to learn more?
============================================
Grab a copy of my CFD Fundamentals Course (for beginners):
www.udemy.com/course/computat...
Learn how to write your own CFD code in MATLAB and python (for intermediates):
dr-aidan-wimshurst-s-school.t...
Learn how I draw my figures and diagrams in Inkscape (for everyone):
dr-aidan-wimshurst-s-school.t...
============================================
Did you like the video?
============================================
Download the lecture slides from my website:
www.fluidmechanics101.com/pag...
Buy me a coffee to say thanks:
www.buymeacoffee.com/NKEZrhvg0
Support the channel on Patreon (and get useful extras for your CFD studies):
/ fluidmechanics101
============================================
Donations:
============================================
1) PayPal
www.paypal.me/fluidmechanics101
==================================
Disclaimer
==================================
The methods, algorithms, equations, formulae, diagrams and explanations in this talk are for educational and demonstrative purposes only. They should never be used to analyse, design, accredit or validate real scientific / engineering / mathematical structures and flow systems. For such applications, appropriate trained, qualified and accredited (SQEP) engineers / scientists should be consulted along with the appropriate documentation, procedures and engineering standards. Furthermore, the information contained within this talk has not been verified, peer reviewed or checked in any way and is likely to contain several errors. It is therefore not appropriate to use this talk itself (or any of the algorithms, equations, formulae, diagrams and explanations contained within this talk) as an academic or technical reference. The reader should consult the original references and follow the verification and validation processes adopted by your company / institution when carrying out engineering calculations and analyses. Fluid Mechanics 101 and Dr. Aidan Wimshurst are not accountable or liable in any form for the use or misuse of the information contained in this talk beyond the specific educational and demonstrative purposes for which it was intended. - Věda a technologie
![[CFD] The Spalart-Allmaras Turbulence Model](http://i.ytimg.com/vi/Xivc0EIGFQw/mqdefault.jpg)
![[CFD] The Spalart-Allmaras Turbulence Model](/img/tr.png)
![[CFD] The k - epsilon Turbulence Model](http://i.ytimg.com/vi/fOB91zQ7HJU/mqdefault.jpg)
![[CFD] Eddy Viscosity Models for RANS and LES](/img/n.gif)
Oh my GOD! I cannot even think how on earth I would have understood these materials if I had not found your channel!
THANK YOU SO MUCH 🙏🙏🙏🙏🙏
The power of explanation. I would love to have this guy as my course instructor
When he will be your course intructor may be you wont like him🤣😛😛
@@vineettiwari5027 I will never not like him! He explains things like a king ☑
@@killua9369 I know. He teaches great . I have watched almost all his videos. Just kidding buddy😁
it looks good .. check this also
czcams.com/video/Uo1A5hChjes/video.html
He has an udemy course on the same topic. CFD
My thesis defense is next week and this is immensely helpful for prep work. THANKSSSSS !
it looks good .. check this also
czcams.com/video/Uo1A5hChjes/video.html
Thanks a lot . I waited a long for this and finally its here. Love u Aidan.
What a clear and brief explanation, thank you!
This channel is a gem!
Dr. Aidan, I found your videos very resourceful and interesting. Thanks for sharing :)
You are the genius. Great to learn lectures from you 🙂
You are an excellent teacher. If you ever end up in Academia your future students will thank you. To round out your talks on turbulence models, I would suggest discussing: DNS, LES and DES. Of these DES is the most practical but understanding all three really adds to a person's fundamental understanding of turbulence.
Thank you so much. I've been trying to get into CFD and these videos really help me to understand how everything fits together.
Fantastic talks - really appreciate the time that you obviously put into these. I've been using them to get up to speed as a neophyte CFD analyst, and I appreciate the way that you stress what things mean over just the math!
Thanks for your kind sharing ! Take care at this tough period of time :)
Thanks a lot for sharing Aidan. You're the best
Once again, fantastic. Thank you so much for these extremely insightful and well laid out slides and videos.
Very nice and smooth explanation.
Thank you for your efforts.
As always an awesome explanation Prof.
Your explanation was easy to understand, and catch the key points. Thank you very much!
You're going to be big. Keep this up!!
How did I miss you all those years in CZcams!!!!?????
Excellent man! thank you so much, is very useful and easy to understand.
Excellent explanation! Thank you!
Thanks a lot for the lecture! The information will help me to perform well at the presentation.
thanks for the crystal clear explanation!
Excellent lecture Aiden, thanks very much !
Thanks from Colombia men, it helped me
That was really fantastic and amazing explanation, and I used it for my present in Turbulence course.
it looks good .. check this also
czcams.com/video/Uo1A5hChjes/video.html
Very powerful speaking, I learned very much, thanks!
Hello mr Aidan,
Thank you so much for your videos.
I have been curious for CFD since I was an undergrad student. Its amazing how elegant and simple are your explanations. Thank you for your videos and keep on the good work!
it looks good .. check this also
czcams.com/video/Uo1A5hChjes/video.html
Love your explanation!
Hi Aidan, Many thanks for the wonderful videos on the basics of CFD. I successfully defended my PhD thesis and your videos were immensely helpful. Keep up the good work. Best regards
it looks good .. check this also
czcams.com/video/Uo1A5hChjes/video.html
Amazing! Thank you!
Fantastic and amazing explanation
Thanks again for the brilliant lecture🤩🙏👍
Very good explanation, thank you!
Great explanation. Being in a preliminary stage of learning about CFD, these videos clarifies the concepts very well and quite intuitively. Great work. Thank you very much.
it looks good .. check this also
czcams.com/video/Uo1A5hChjes/video.html
Thanks a lot. You're a great teacher, in the true sense of the word.
it looks good .. check this also
czcams.com/video/Uo1A5hChjes/video.html
Great video and brilliant explanations. I really enjoyed it. Thanks!
it looks good .. check this also
czcams.com/video/Uo1A5hChjes/video.html
Thanks Aiden for all these great videos with excellent clarity on explanations. The only thing I would say maybe to add more value is annotating the slides using a digital pen while you explain. This may help to draw the attention of the audience even more. Thanks again for your great contribution towards teaching the most complicated concepts in the field of CFD.
That's actually a really good suggestion. Thanks Venkata!
thanks for the video. It's really good!
The best channel to know the magic behind CFD
For an application engineer like me this channel is safe hevaen
extremely helpful video. i've been only able to get convergence in my model using k w SST. i needed to add turbulence and energy dissipation.
I found this talk very usefull!
These videos are great !!!!
Love you Aidan!!
Very helpful, thnks a lot!
Great explanation 👍
Thank you
Thank you very much dear Aidan 😍
i rarely leave comments but i just wanted to say these videos are amazing. You have a way of explaining things that is rare in people. Your style and methods are something i'm going to use when explaining CFD concepts.
Thanks Austin, that really means a lot to me
@@fluidmechanics101 i'm using a 12mm tetrahedral mesh with k omega model with an inlet turbulence scale of 20mm and energy dissipation of 2% in my inlet velocity field. i'm restricted in my total mesh size by my memory constraint of 8GB so my total volume is not quite as large as it should be for the duct of 250mm radius the overall space is 1000mm radius cylinder. with a depth of 5000mm. Will i get results that will yield insights for tuning the duct profile? At 300 iterations i am seeing the emergence of detached flow in the trailing section so i see that as a positive sign. I think i need another 300 or 400 iterations before all residuals are below 0.001
As you are limited by memory, you can only do the best that you can. I would just accept that your mesh is under resolved and look for qualitative flow features and stick to comparisons between different cases. You can still get good understanding from CFD, even if your results aren't accurate. This is really what CFD is good for: understanding, not prediction 👍
thanks mate, your explanations are amazing.
it looks good .. check this also
czcams.com/video/Uo1A5hChjes/video.html
Good explanation of the K-w model. Please provide an explanation of the different K-w models' respective applications. Thank you
You are a God! Thanks a lot!
very easy and useful.
Thank you for the video
These videos are great :)
Man, I'll pay you my tuition fee and you come teach me. I wish all our lecturers could pass on the knowledge like you do. Wonderful work! Plus, I appreciate that ,thanks to you, I can now work on my dissertation with confidence
I have 1 question as it is hard to find exact data. Ansys 2022 has the Transitional k-kl-omega model. Is it the same as k-omega sst? There is also Tansition SSt one as well. A bit confusing. Mind giving some clarification on that? Thanks :)
I haven't checked out the manual but it sounds like they are different models (but either could be used for modelling transition to turbulence)
@@martita505 No they are not the same. You can find the k-omega SST model as one of the k-omega model options in Ansys.
Realizable k-epsilon model with enhanced wall treateament is also useful for aerodynamics applications. Great video.
would also love to see a video on that one
Hey man, I have been watching of lately a lot of your videos, you really inspired me to try out a bunch of CFD little problems. I want to pursuit a career working as an aerodynamicist, I'm currently studying mechanical engineering :)
Awesome!
@@fluidmechanics101 thank you so much! By the way I would like to know if you could make a video on the algorithm that XFoil uses for 2D flow over airfoils
XFOIL is awesome! I am a big fan. It uses a classic 'panel Method' which is a bit outdated for fully turbulent flows. But it's treatment of natural and forced transition on transitional aerofoils is really good! I might make a video one day, it is a bit niche to aerodynamics though 😄
Thanks for the video
Thank you very much ! you are such an inspiration for me the way you explain things so easily ! It really made me to give you a feedback. I would just like to know a bit more on tricks or tips to select the ideal turbulence model for different geometries.
it looks good .. check this also
czcams.com/video/Uo1A5hChjes/video.html
dr aidan ı wish you were a teacher in my school you are the best
Thank you Aidan for the nice explanation!
Another rather popular RANS turbulence model is the KSKL (or k-sqrt(k*L)) model. It is quite popular for maritime applications. The advantage is a less sensitivity to the value of y+ at the wall and an easier iterative convergence because you don't need to specify a very large value of omega at the wall (especially useful at high Re when the cell size must be extremely small and so omega at the wall becomes huge). It could be worth a video about it. :)
An interesting thing would be to see how these turbulence quantities (k, epsilon, omega, mu_t, production) are distributed in the flow (for instance in a boundary layer) to yield the correct mean flow solution. In this way one could develop a sort of "feeling" about where a particular quantity should be large or small, so one could better evaluate whether the CFD solution makes sense or not.
Thank you again!
it looks good .. check this also
czcams.com/video/Uo1A5hChjes/video.html
Thank you!
awesome mate
Well explained gud 👍
you're the MAN
Great lecture sir please make a good series on large eddy simulation as no one talks about it so frequently it would be great
it looks good .. check this also
czcams.com/video/Uo1A5hChjes/video.html
Just finished my degree at Imperial in Aeronautical engineering. I really wished I'd seen these videos earlier!!! Absolutely fantastic, and very clear and consistent. One very minor suggestion could be adding a box to your slides and then overlaying your video into the box post-production? Avoids you having to cover any information on your slides... Thank you!
Do you have a twitter account? I have some inquires about Imperial; my twitter is @Killua_xy please contact me
it looks good .. check this also
czcams.com/video/Uo1A5hChjes/video.html
Thank you Aidan!! This was the video I was waiting for so much :D
What do you think on a video about turbulent mass transfer and wall functions?
it looks good .. check this also
czcams.com/video/Uo1A5hChjes/video.html
Hi Adian, I would like to point you one correction at 10:30. It should be Pk.. not Pe.
Remaining all is good and wonderful material.
Well spotted! Thanks 😄
It would be more interesting if you could give a lecture about the v2-f turbulence model.
Thank You Very Much, I felt very comfortable the time I switched on your lecture, My entire work of research is dependent on these models, But at this point in time, literally, I don't have any idea of these models and their usage. I am way too late for my research proposal. I am trying to connect things together. So I am trying to reconstruct everything to get a better understanding, I just found your channel, don't know about other videos yet, have to watch them all, In case if I don't understand then you can help me out, One thing I have to ask you to make is a video on simulations in CFD Ansys on a sample of models
Great video, please do one on DES, thanks
Now everything falls into place :)
Wow! So useful :) Could you please consider making a video on types of separation and which model is best suited for which type of separation?
it looks good .. check this also
czcams.com/video/Uo1A5hChjes/video.html
Thanks
Big thanks for video, this is awesome. Will you make video about Reynolds Stress Model (RSM) in future?
Yes, hopefully if I get time 😅
MVP
Thank you very much Aidan! Could you do a video about the v2f model sometime in the future? Greetings from Brazil
Yep, its on my list!
Thanks! very informative, it would be great to go through the non-linear eddy viscosity models since you are creating videos around the RANS models, it's just a suggestion. It might be also worth looking at the weaknesses of linear EVMs, in particular within turbomachinery applications when high curvature and rotation exist in the flow filed, hence Curvature Corrections, etc. again just another suggestion.
Yep, there is lots to cover. I want to get through all of the turbulence models eventually 😊i am trying to decide whether to go for LES (and its variants) first or go for NLEVM and maybe Reynolds stress. I will bare this in mind! Thanks for your suggestion
@@fluidmechanics101 That would be a very useful archive to have. Also, you are right, since the main idea of the NLEVM is taken from the RSMs, I think it is better to start from the RSMs and build upon that.
it looks good .. check this also
czcams.com/video/Uo1A5hChjes/video.html
THANKS SIR
Firstly, Thank you Aidan for the videos. It is very beneficial. Keep going. Best wishes.
I have a question, Is Y+ value immaterial for SST-kw model? Can you explain. Thanks in advance.
I LOVE YOU METROMAN
Thank you so much for such a nice presentation! Would you mind to clarify with regard to Free Stream Turbulence, please?
Aiden brilliant explanation!!...one doubt though ..you mentioned that this model is good when having adverse pressure gradient/ mild seperation...what if we have a very large seperation??? What turbulence model do we go for???
very nice talk. In practice Eq 9 is what is used in the k-omega model but I think the omega transport equation as derived from (or equivalent to) the epsilon transport equation should have another term that is proportional to (rho/omega * grad k times * omega) which is zero only for homogeneous flows.
Thanks for the pointer. Do you have a good reference for the additional term? I would love to have a read and check it out
@@fluidmechanics101 Pope "Turbulent Flows", chapter 10. The exact transport equation for omega (implied by epsilon) is Eq. 10.99, which reduces to 10.94 with very minor assumptions and finally to 10.93 (i.e. your Eq.9) fro homogenous turbulence
Yes that's the one. Thanks Roberto
@@fluidmechanics101 A pleasure. Again, your lecture is very clear
Hi Aiden! I was comparing the equations writed down in the slides with the ones on the NASA website, it seems that the SIGMA_k and SIGMA_omega coefficients are placed in a slightly different way; in particular, they are at the numerator and SIGMA_omega is used instead of SIGMA_k within the omega equation. I think that the same thing also happens in the k - omega SST video. Please, let us know if there is a little typo or the adopted coefficients have simply a different value. Big thanks for these beautiful resources!
I would always go with the NASA page. The page exists exactly for this reason ... to check for typos in CFD user manuals and CZcams videos 😅
Excellant video.Would be really great to understandd eulerian multiphase model.
I think i have a video on multiphase flows 🙃 check it out!
@@fluidmechanics101 I'm really sry.Did notice it later.Was awesome!
Thank you so much for your video!!!
You mentioned in the video (time 15:30) that k-epsilon model requires damping function for solving the boundary layer while k-omega doesn't require any damping function. This difference explains why k-omega is better for cases where resolving boundary layer is essential to account for the adverse pressure gradient.
Some time ago, I tried to study why k-omega can resolve the boundary layer without using damping function. However, it is still unclear to me since I couldn't find any reference with a good explanation on it. May I learn from you about that?
I would like to know that too, excelent question and fantastic video from Aidan again
I think this is explained in the original wilcox paper. It is because the behaviour of omega is relatively smooth all the way through the log law and viscous sub layer, while epsilon has a peak in the buffer layer. This makes it tricky to define a consistent function for epsilon that is smooth all the way to the wall. I would have another look at my video for ‘epsilon wall functions’ and you can see the variation of epsilon close to the wall. Good question though! It is definitely not very clear
Thank you for the reply, and I have saw these videos so many time kkkk.
@@fluidmechanics101 Thanks for your reply.
I revisited the Wilcox's book "Turbulence modelling for CFD". He tries to analyze the asymptotic behavior of omega.
According to my understanding, he analytically finds out the asymptotic values of k and omega at wall surface (in viscous sub-layer), and then analyze towards into the log-law sub-layer. Through fitting into the standard law-of-the-wall formula for log-law sub-layer, the analytical values of the formula constants (B and n)are evaluated. He compared the formula constants for different sets of k-omega models against the measured formula constants (B=5 and n=2). He points out that his k-omega model gives the fairly closed values, which argues that damping function is not required for k-omega model, and also the good performance of his model.
Hope my understanding is correct! But, frankly speaking, I am not in a good context with the procedures in his proof.
Pls kindly share your view, if any.
it looks good .. check this also
czcams.com/video/Uo1A5hChjes/video.html
I heard about Lattice Boltzmann instead of Navier Stokes but dont really get it. If you want an idea for a next video ^^
Muchas gracias por tu video, dejó mi comentario en español para que veas que te seguimos desde muchas partes. ¿Donde encuentro el video sobre el modelo k-e que mencionas?
Hi, Dr.Aidan, thank you for your interesting introduction to these RANS models, could you please share some of your ideas about RSM kinds of the second-moment closure model like this video. Thank you for your work.
it looks good .. check this also
czcams.com/video/Uo1A5hChjes/video.html
Phase change model, in particular Lee's Model for evaporation & condensation
amazing lecture!!!
I just have one question: why does the k-omega formulation does not need wall functions?
if we can convert between omega and epsilon freely, why this formulation does not use damping functions?
Hi, I was going through the account of k-omega model given in Versteeg Malalasekara and the cross diffusion term given there(page 91) looks a little bit different from what you’ve shown( which agrees with the NASA website). Please let me know which is the right representation.
I’m a big fan of your channel and have also endorsed it recently. Thanks a lot for these videos :)
The k omega model has changed quite a bit since the Versteeg and Malalasekeera book was written. So I would probably go with the NASA website, taking care to note which version you are referring to. This area is notoriously tricky, with changes in notation all over the place! You might also want to check the manual / source code of the CFD code you are using. I normally go with something like 'the version of the k omega model described in version of . That way you can be sure that the version you state is the version you use 👍
Is it possible to use your presentation in the form of screenshots with mentioning your name?
Thank you very much! Remark please:( May be )there is an error in equation 3, min: 2:47! you have the turbulent viscosity which is divided by (sigma epsilon which represents the diffusion coeff of the dissipation epsilon, not sigma k ) ! ( your equation must in any way contain this constant, which is not the case ) tanks you again for the excellent video!
Yep, probably a typo 😂
it looks good .. check this also
czcams.com/video/Uo1A5hChjes/video.html
Why not start making videos on LES simulations? Since, they are the state of the art now. In addition to that, I must say, your videos have just the appropriate mixture of math and physics. Keep making more :)
Thank you for your kind explanation! Also, I have a one question. Video said that k-w model does not need damping function. Then, does it mean that k-w model use empirical coefficients instead of damping function in the viscous sub-layer?
Essentially yes ... The empirical coefficients tune the model so that it has the correct asymptotic behaviour as it approaches the wall
Thank u so much for your video. But what happens if we use log-layer wall function with k-w model
Should still be fine if your flow is similar to a pipe flow 🙂 if you have adverse pressure gradients / strong buoyancy etc. then it will probably be in accurate and better to go with y+ ~ 1
@@fluidmechanics101 thank u for sharing, u're very kind.
Hi! How and what software do you use to create your images ? Thank you Aidan!
Hi Rodolfo, i make all my images in inkscape. Its free and cross-platform so you can use it yourself if you like 👍
@@fluidmechanics101 Thank you!