AI and Physics: Hydrodynamics and the Riddle of Turbulence
Вставка
- Опубліковано 12 вер 2024
- The recent development of AI presents challenges, but also great opportunities. In this clip I will discuss how hydrodynamics could benefit from AI.
Mind also my backup channel:
odysee.com/@Th...
My books: www.amazon.com/Alexander-Unzicker/e/B00DQCRYYY/
Discovering non-linear oscillators, chaos and all that in my Junior Level Classical Mechanics course (1983) put the "WOW!" back into Physics for me. (Thanks, Professor Scott!) I took Fluid Dynamics the following quarter; it revealed that ordinary, everyday physical phenomena are intractable. I find all of this more fascinating than the "strangeness" of Quantum Mechanics, Relativity and Cosmology, etc..
Prior to that, it seemed that any physical problem could be boiled-down to either an initial-value or boundary-value problem with the solution composed of series of eigenfunctions; the hard part was computing the coefficients, which didn't seem very interesting.
Buy a canoe, buy a kayak, take a sailboat out on the ocean. The brain can be trained to see what math can't do.
I used to do what are called "big jumps" on snowboard and mountain bike, physics be damned.
My channel is dedicated to plasma physics and it all came about from sailing. One of the best fluid simulations you will see is a dolphin swimming through Bio-illuminesants
@@nonlinearplasma1370 You had your comments turned off on the double helix. I had never seen one slip like that before. Are there any math formulas for that?
@@martinsoos I haven't seen any, I tend to discourage the use of Math to understand the 3D Helical wave form because everyone was taught to use 2D mathematics to model the 3D. For example if we look at Schroeder's equations it will tell you that the positive phase value is above the axis for a 2D waveform and negative is below the same axis. It will then implement that system into a 3D point source wave and claim the negative phase is now above the axis but separate from the positive waves value. The geometry change of the reference point shouldn't mess with our definition of positive and negative, yet it does and this messes with the resultant values when you construct or deconstruct multiple waves in a simulation. A good way to look at the situation is, if we say the wave peak is traveling around the axis that is our reference point equalling zero, at what point did the value of the wave we are measuring cross the axis? It only crossed the axis due to our perspective of the 2D wave, it physically remained at the same height above the axis of rotation we have said is our zero reference point that determined if the wave was positive or negative in the first place. Then the question becomes, if I'm looking at a charged particle from one perspective and you have the opposite perspective and I claim it is negative, and we both agree on definitions that resulted in that assumption, would you also claim it is negative or would you claim it is positive? My point is above or below the axis does not exist in 3D space but it is ingrained in us to use math that says it does exist.
@@martinsoos ah, comments were off as I said the video was made for kids which turns of the comments. I have changed that now. I'm sure APS who made the inserted video have the formula they used for it. It's a channel I follow as they do some amazing videos of models they built.
I know you apologised for the audio quality but seriously, just re-record it.
Vortex modeling may be the key to quantifying the ground-state properties of nucleons. See "Unbound low-energy nucleons as semiclassical quantum networks"
As one who no doubt can read and understand complex German science articles, I'm wondering if you've read Ott Christoff Hilgenberg's papers? Such as:
"Über Gravitation, Tromben und Wellen in bewegten Medien"
and,
"Über Strömungsversuche mit Senken und Quellen, die das Wesen der Schwerkraft grundlegend erklären"
or,
"Über den Magnus-Effekt, die experimentelle Bestätigung seiner Umkehrung und den Zusammenhang dieser Strömungseffekte mit meteorologischen, ballistischen und elektrischen Vorgängen"
He studied hydrodynamics extensively and the mechanics of tornadoes and hurricanes.
I would very much appreciate your reading and evaluation of his papers.
The main issue that we currently have is that we don't understand that static water is heavier than turbulent water and laminar water is heavier than static water. Think of a water jet in a static body, why does the water bubble up? Why does it not go down when it is supposed to be heavier? What happens at the boundary layer between the Jet and the Static medium?
This problem seems so obvious to me. Magnetic field lines would appear to be the place where nonlinear dynamics manifest, whereas between fields lines flow should be mechanical and predictable. This goes back to the molecular structure of water being determined by cis-trans hydrogen bonds being ordered and sequenced by fluid flow around the inertial plane of a magnetic field. When I first saw one of Ken Wheeler's (Theoria Apophasis) structured water devices, I knew that it "worked," but I had to devise a proof of concept experiment to prove the mechanism of action. Line the tube with PH sensors and you will see the PH perturbations in the water as it flows around the inertial plane of the magnet. I really hope Nick Lane sees this. It is the source of the motive force of biology. Also the basis of a gene drive. Lane's lab didn't even know they needed PH gradients for the electrodynamics of the system to work. Structuring water and sequencing RNA are fundamentally the same electrodynamic process. I'll wait for AI to have my insight into basic phenomena.
The question is: what do we wish to understand about turbulence? We know the equations that describe the laminar flow, we know that turbulence is random and perhaps chaotic, we know in which conditions turbulence sets in, we know the equations that statistically describe fully developed turbulence, we know that water wets. A machine should be asked a specific question, but obviously we have none. Extraordinary claims about computers have been issued since the 80's, even in 60's sci-fi they took over the control, yet none of all that happened, this opinion aged very badly, why? Making further extraordinary claims won't answer this question.
I love the ideas of the Ancient Greeks; they first thought water precedes everything. Later then went on the idea that fire precedes everything. I believe this idea can be refined with light or photons ∆E=hf preceding everything. In flowing water we hydrogen bonds continuously forming and breaking with the exchange of light photon energy. Everything is continuously radiating light of EM waves even water LOL.
I don’t see that AI and pattern recognition could possibly help. If turbulent flow is chaotic then it is inherently not predictable. Even if it were possible to identify some patterns in the water flow in the gully behind the prof then surely if you turn upside down just one stone in the bottom of the gully then you get a totally different set of output patterns. Is that not the case?
taylor curls lol
Any thoughts on Mike McCulloch's Quantised Inertia theory?
We need to make simulations for predictions.
Ai won't solve turbulence. I will. The future Nobel prize winner
It seems that you don't know much about AI because there are papers on this. The dream of Heisenberg is probably not realistic, unfortunately. It didn't turn out to be anything. One should not try to follow every hype if one doesn't understand things. As the German saying goes, 'Stick to what you know best.In life, there are always those who think they know everything better... They just believe so... Arrogance is the enemy of trust. 🙂
Last
If calculate energy losses of river where pebbles rocks in bottom of river and rough terrain of embankment (sharp corners). Reynolds number was temporary solution and became ingrained in science consciousness, viscosity is somehow internal friction, even Brownian motion for water has 10^14 collisions per seconds mixing rate. Only small dust (paint) particles flows in laminar pattern, or microscopic flower pollens in Brownian observation.
Couette laminar flow description is laughable. You take very fine dye potassium permanganate and it has laminar flow exist for short time but at same conditions take very very small dust particles, for example, use inkjet printer ink that same laminar layers due to Brownian motion mixes togethers immediately. Shear roughness of channel or pipe creates stooping. If you try present mathematical model of turbulence and laminar flow solutions to scientist they call you deletant oh my God Reynolds number how dear you, even he used as temporary measure in 19 century.
❤❤❤❤❤❤❤❤❤❤❤❤❤❤ Canadians and Japanese join forces on this one………huh eh
❤❤❤❤❤❤❤❤❤❤❤❤❤ I have loved unz for years ……… now he is going kookoo
May you specify what you dislike?
What did you find crazy? He made a lot of sense.
I've always been confused about what the problem here even is.
fundamentally the issue is that turbulence is random... which you can't predict by definition... so only an extremely stupid person would obsess over this question.
it's like being obsessed with finding an equation in sub-linear time that can identify whether a given number is prime
the issue is not that Navier-Stokes equations are non-linear. the Navier-Stokes equations are non-linear because turbulence cannot be described by a linear function, because it includes randomness. so it's not only non-linear, it's not even a function.
one of the most severe failures of modern mathematics is the insistence that everything should be describable with functions. but we know from trivial observation that in truth virtually nothing can be described with functions.
this means that the 'riddle' here is not a riddle, since the problem is simply that everyone working on the problem is just insisting on telling the universe how it should behave instead of letting the universe reveal how it actually behaves.
whenever a scientist or mathematician proclaims that something they've observed shouldn't happened, it means that that person is a solipsist who doesn't understand anything.
@@sumdumbmickYou are making a hypothesis here by saying ”Turbulence is Random”. What is your definition of random? Unpredictability? If yes, then you just defined deterministic chaos as random. One simple 1-dimensional example is the logistic map iterated at r=4 (see Wikipedia), which has an analytical solution but output is unpredictable. A more (financially interesting) daily example is the stock market behavior, but it is far more complicated than turbulence due to human interactions. Turbulence is more simple and is seen in (super)fluids. It is more accurate to say ”Turbulence is Chaotic (or Deterministic Chaos)”, which is supported by both observations and maths. Understanding turbulence not only can help us to understand physics in all scales from atomic size to galactic size, but also help to explain emergent phenomena like gravity, Casimir effect, dark energy, …, even human consciousness.
@@sumdumbmick as I understand turbulence can only occur because of the nonlinearity of the NS equations.
e.g. you can find a critical reynoldsnumber where small disturbances accumulate.
Mathematically speaking you can construct a solution with and represent it with eigenvectors, that solve the NS equations.
If the Reynoldsnumber is small enough the damping in the equations are large enough that the solution/flow remain stable laminar.
But if you increase the Reynoldsnumber the amplitudes of the eigenvectors are increasing, and through the non-linear term (convective term) the eigenfrequency multiplies... and you get chaotic behaviour.