Turbulence is Everywhere! Examples of Turbulence and Canonical Flows

[@chronicnerd8337]

If I ever get to do a Ph.D., I sincerely hope that I get you as my Ph.D. advisor. These are really great videos!!

[@Music_Engineering]

You can always apply to Washington University :)

[@Eigensteve]

Wow, thank you!

[@Nehekar1]

Glad and happy to see our work (3:55) finds it's way into amazing content like yours! Thanks!

[@Eigensteve]

Thanks, I really love that pipe flow video -- super cool!

[@mmountainbiker2563]

I can’t believe how useful and practical the knowledge of this professor is. I have an exam in fluid mechanics and my prof just writes shity formulas without explaining during lectures. Reading my profs notes looks like wtf is this , when will I use this in real life, but Prof. Brunton gave me a hell of lot motivation

[@rezakhosrozadeh5101]

It was such a fantastic lecture. As a mechanical engineer in heat and fluid that had done some research in this field, I learned a lot after watching this video. I wish you talk about turbulence in heat transfer in the upcoming videos. Thank you.

[@Eigensteve]

Thanks! I'll put it on the list. I'm not a heat transfer expert, but there are definitely some really interesting connections there.

[@pierregrand2609]

You enthusiasm for the topic (in this, and in previous videos) is contagious. Thank you for doing this so that those of us outside of the classroom (in my case, a few thousand miles away) can benefit too.

[@JousefM]

Great presentation as always Steve!

[@Eigensteve]

Thanks so much!!

[@saeedtk]

I'm a Postdoc at University of Queensland, Australia and I've found your videos very helpful. Keep up the great work Steve!

[@manuelo1738]

Dr. Brunton, Thank you very much for making these great videos!

[@Eigensteve]

My pleasure!

[@ahmedkamiss7643]

One of my life goals is to watch and understand each amd every lecture of yours!

[@engineerdo6568]

Steve, very good explanation with great examples. You brought it to the point. If someone wants to get started with the simulation, we are recommending openFoam. It is relatively easy to use but can be sometimes frustrating to learn. The entry hurdle is not easy to overcome...something we are working on in the moment.

[@MuhammadUsman-ny5rh]

I am an undergraduate student trying to enter the field of CFD and turbulence modeling, I because of this entry hurdle unable to get grip on OpenFOAM but loved you instructions

[@engineerdo6568]

@@MuhammadUsman-ny5rh Thanks. I hope they were helpful and you were able to solve your problem.

[@ilpreterosso]

I majored in math and was originally interested in AI. Though I was looking for further study in that field, but then got carried away with CFD. Now I'm working on DG/HDG/SLDG schemes. Excellent introduction to this fascinating field!

[@unverozkol]

Excellent summary. Very engaging and fun to watch.

[@Eigensteve]

I appreciate it -- it was fun to make too! :)

[@KillianDefaoite]

What a fantastic video this is. I am starting my MSc in CFD this October and I am excited to learn more.

[@DerekWoolverton]

I was once flying into LAX and the right wing of the jet hit the wake flow from the jet in front of us, causing it to lose lift. The plane tilted right and started falling out of the sky, and for the next 15 seconds we all thought that was it. Remarkably the pilot regained control in time and brought us in for landing. So yes, understanding these things is quite important.

[@Eigensteve]

Yikes, very scary. Glad you made it out safely. Wake vortices are no joke.

[@dsdy1205]

Pilot probably had a long word with ATC after that one.

[@ramiedirisinghe3141]

LOVED this video! I liked that you made great connections to the sometimes abstract research that's done in fluids research and its applications to real world problems. This is sometimes one of the stumbling blocks I have when explaining fluids research to my non- engineering friends. I also loved that you talked about the importance of experiments and simulations and how they complement each other.

[@mattkafker8400]

Great series so far!

[@Eigensteve]

Thanks Matt!

[@MichaelEvans-yq7xj]

My first reaction to the Feynman quote about seeing a hurricane from Navier Stokes, was to think about the hairy ball theorem. We want to know where the flow "lives".

[@Eigensteve]

Interesting observation -- I like that.

[@LucasDimoveo]

Your communication style is fantastic. Have you seen the mathematics in computational modeling used to model physical phenomena like fracture, corrosion, or another Materials Science topics?

[@Eigensteve]

Thanks! I am not an expert in these material science topics, but I believe people are using computational modeling extensively for these fields too. I've seen some work on crack propagation, which is really challenging because it is multiscale, from the molecular level up to the macroscale. Characterizing advanced materials computationally is one of the biggest modern challenges, since the performance of the material will depend on the composition of the material at all scales (from very small to very large). We don't even know the performance envelope for these materials; biological materials, like a sea-shell, are often heterogeneous and non-isotropic at many different scales, which makes them incredibly challenging to simulate (a lot like turbulence!)... I'm really fascinated by this topic.

[@LucasDimoveo]

@@Eigensteve Thanks for the response! I'm interested in if surface chemistry and transport phenomena have an effect on crack propagation in devices under mechanical load in some chemical solution. Examples of this could be a beam under stress in a river or a neural probe within an organism. What's interesting here is that mathematics more or less exists to explain transport phenomena, thermodynamics, and material failure - but attempting to model all of these on a multiscale level would be a computational nightmare. I think that such devices could have sensors on or in them that could let us know what their 'state' is. That in combination with the algorithms being pioneered by yourself and Prof Kutz could then give us the most important equations at play with regard to what stresses these devices are under. I could type more, but I should probably be doing homework! I look forward to learning from you directly once I conclude my undergraduate degree.

[@Eigensteve]

@@LucasDimoveo Wow, that sounds like a very interesting (and tough) problem! I agree, some combination of first principles and data-driven modeling would be interesting here. Keep in touch, as I'd love to hear how things go after you graduate!

[@weert7812]

Awesome video thanks! Got a new SCUBA term for describing current. Cavity flow! Cavity flows are regularly experienced when diving nearshore in tidally driven ecosystems!

[@smoothcortex]

You guys might think I'm a bit mad. I study human behaviour, development and neuroscience - highly complex systems, just badly understood. Some of these models could possibly be very useful to describe many things in the brain. Students in psychology are not taught how complex systems evolve, nor the math to understand it. Yet that's exactly what we're studying.... Thank you Steve as always 👍great lecture

[@GabrieleNunnari]

Your idea is not wrong, in fact I would suggest you to give a look at the works of Theo Geisel and peoples that worked with him at the MPIDS. I do believe you can find some interesting information about your idea.

[@smoothcortex]

@@GabrieleNunnari Thank you. Definitely an interesting institute. I'll get reading

[@Eigensteve]

Not at all. That is a very good connection. For a long time there have been connections between turbulence research and neuroscience, as they both involve high-dimensional, nonlinear, multiscale systems, where we only have partial measurements. Thanks for the comment!

[@Eigensteve]

@@GabrieleNunnari Thanks for this info!

[@smoothcortex]

@@Eigensteve 'Rethinking Innateness - A connectionist perspective on development' a great book exploring the non-linearity of the brain. I'm still reading it. Some of the computational concepts and maths are a little tough for me but I'm sure you guys would have no problem there.

[@ziqiangyang281]

Amazing work!!!

[@dekonildo]

Hi! Great video, as always! The link to the Lisbon group is wrong in the video. It is laseF.ist.utl.pt, not laseR (which I assume was an auto correction or something like this). And I also think their youtube video is no longer available. Cheers!

[@Eigensteve]

Thanks for fixing the link! Yes, must have been autocorrect. I really love their videos!

[@firstcommenter202]

i was wondering whether any of these has applications in medical and biophysics

[@ahmedelkhateeb1869]

Dr Steve, Is there some kind of map where I can follow your material in order? would appreciate something like that

[@rcv3208]

Have you done video on airfoil boundary Layers, aerofoil design and mathematical computation

[@sciWithSaj]

Thankewe for this great content

[@Eigensteve]

Thank you!

[@value8035]

Please include the links given in he video in the notes as well, so we can check them out too without typing hyperlinks? Thank you for nice content!

[@Pavan_Gaonkar_abc]

thank you man love it!

[@clairezhao2395]

Instantly subscribed

[@Eigensteve]

Awesome, welcome to the channel!

[@nelsonphillips]

I run experiments on my gaming pc that run 500 core-hours, which is slightly less than 35million core-hours.

[@Eigensteve]

Definitely -- There is a lot of work in putting multiscale physics simulations on GPUs since they are getting so powerful. There are issues, like shared memory, communications, and how much of the core computation can be written as a matrix operation.

[@pradyumnchiwhane5730]

Steve Brunton I request also suggest you to launch a distance learning PHD course for this subject from the university of Washington for person like me, which can learn from India on their own.

[@jms547]

Steve I know you're going to go into some machine learning aspects of turbulence, but will you also cover some numerical methods for DNS? If so then going through pseudo-spectral methods for turbulence simulation would make a great connection with your Fourier Analysis course! (Other than that no particular reason why I suggest it, it's not like I'm working with a pseudo-spectral code at the moment no wait it's exactly like that ;)

[@Eigensteve]

Thanks for the suggestion -- I will likely cover some of this, but it might take a while to get there. Numerical methods for DNS is not my core area, but definitely very interesting.

[@jms547]

@@Eigensteve in that case, guest lecture from Nathan Kutz? ;) Actually I see that he's got some lectures on DNS from back in the Before Times. But it'd be nice to see similar material from a different point of view. Whatever the case I really look forward to seeing what you have to say about ML in fluid dynamics. This definitely seems to be a hot new area!

[@mariuszprzybylski8435]

Very inspiring!

[@Eigensteve]

Thank you!

[@Josephprouse1234]

What if they use an mri machine on a water current, is that something they have tried yet?

[@subhanbasha8813]

Hello Steve, Is it recommend to code all the machine learning algorithms from scratch so that I can learn math behind it or just understand and start to code?

[@bakrybsata3114]

Amazing

[@Eigensteve]

Thanks!

[@pipertripp]

Really enjoying this. Super accessible for people without a strong background. Does anybody have some good examples of toy simulations that an enthusiastic layman could attempt in python? I'm teaching myself numerical methods (numerical differentiation/integration and solving ODEs) and would love some real systems to crudely approximate with simple models. If there are some simple turbulence scenarios that could be modeled on a laptop, I'd love to know what they are.

[@Eigensteve]

Thanks! There are increasingly some great open source codes out there. nek5000 and openfoam are both really popular. The IBPM code is great to get started with simulations, although it won't go up to turbulence (github.com/cwrowley/ibpm)

[@pipertripp]

@@Eigensteve many thanks! I'll be sure to have a peek.

[@ammoniumhydroxide9739]

This is the video I should watch during my lunch to make my soup turbulent and tasty. : )

[@Eigensteve]

Sounds great!

[@ram_c]

very informative but there is no hands-on tutorial!

[@drskelebone]

Jurassic Park font?

[@Eigensteve]

Indeed, good eye!

[@aleph285]

I hope you’re teaching in my university

[@mdabdullahalmamun5067]

Sir, Turbulence is beauty, isn't it?

[@Eigensteve]

I agree!

[@tjcambre5236]

Where is magnetic flow

[@davidyates4857]

Would you be able to point me to a reproducible code on github which uses deep neural networks to predict turbulence?

[@Eigensteve]

I'm going to have a video on this soon, with some examples. A few key groups/papers you can look into: review paper by Duraisamy, Iaccarino and Xiao (www.annualreviews.org/doi/abs/10.1146/annurev-fluid-010518-040547) has a lot of good refs in it; Ling et al JFM 2016; Maulik et al JFM 2019; more recently github.com/lululxvi/deepxde and arxiv.org/abs/2010.08895

[@oedesse]

Braunschweig rocks!

[@Eigensteve]

I agree! First place I visited in Germany. Rich history of great scientists and engineers there.