Understanding Hamiltonian mechanics: (1) The math

แชร์
ฝัง
  • เผยแพร่เมื่อ 17 ธ.ค. 2024

ความคิดเห็น • 67

  • @gcarcassi
    @gcarcassi  4 หลายเดือนก่อน +1

    These are results from our research project Assumptions of Physics. More details on this topic in our open access book: assumptionsofphysics.org/book/ . If you are interested in our active research, see our other channel www.youtube.com/@AssumptionsofPhysicsResearch .

  • @Rey2u
    @Rey2u ปีที่แล้ว +3

    One of the best educational videos I have ever seen. It gave me much better understanding and insight than anything else I have seen on the subject.

  • @stevewhitt9109
    @stevewhitt9109 2 ปีที่แล้ว

    After all these many many years I FINALLY get the Hamiltonian! Thanks

  • @SohailSiadat
    @SohailSiadat ปีที่แล้ว

    Finally I found what I was actively searching for. Thank you!

  • @Monllorf
    @Monllorf 10 ปีที่แล้ว +2

    Excellent presentation and formal but ,easy to understand.

  • @myyoutubeaccount4537
    @myyoutubeaccount4537 4 ปีที่แล้ว

    The brackets on the right of equation 1:51 and on the left of the equation at 2:23 should be parenthesis, not brackets, because in a vector field context brackets are usually Lie brackets.

    • @gcarcassi
      @gcarcassi  4 ปีที่แล้ว

      I think the problem is the coma. The square brackets are used for all linear algebra elements, including vectors and matrices. Unfortunately, there is no way to fix it at this point. FYI, there is no consistent notation that spans all the different areas touched by this video series. There are plenty of other cases where one would object. Remaining in the realm of differential geometry, IMHO, the use of grad/curl (instead of the exterior derivative) would probably be a bigger offender...

    • @myyoutubeaccount4537
      @myyoutubeaccount4537 4 ปีที่แล้ว

      Gabriele Carcassi Yes I agree on that, having a notation that fits everything is always a problem. Nevertheless, when it comes to Hamiltonian / symplectic math and physics the notation is already quite standard across books. Also, grad is standard in differential geometry, it is the Riemannian musicality of df. And one could define curl via Hodge duality.

    • @gcarcassi
      @gcarcassi  4 ปีที่แล้ว

      Unfortunately the coordinate free notation used in differential geometry is good for the math, but not good at highlighting the physics. So that's a no go.

  • @seanki98
    @seanki98 8 ปีที่แล้ว +6

    the hamiltonian equations seems to be a little bit reminiscent of the cauchy riemann equations. Is there a link?

    • @gcarcassi
      @gcarcassi  8 ปีที่แล้ว +6

      I wasn't familiar with the Cauchy Riemann equations... but, at a first glance, I think there is a link: they both seem to preserve a symplectic form. In the case of Hamilton's equations, it's the symplectic form that corresponds to the areas/volumes in phase space. In the case of Cauchy Riemann equations it is the preservation of the complex structure. See en.wikipedia.org/wiki/Linear_complex_structure and en.wikipedia.org/wiki/Hamiltonian_system. In both articles you'll essentially find the same block diagonal matrix, the preservation of which is what switches the order and adds the minus sign. I am not sure, though, whether the similarity ends here or there is something more...

    • @seanki98
      @seanki98 7 ปีที่แล้ว +1

      Thank you very much sir
      =)

  • @G11713
    @G11713 7 ปีที่แล้ว +15

    This would be more helpful if you had described the physical meaning of the symbols.

    • @gcarcassi
      @gcarcassi  7 ปีที่แล้ว +11

      Noted. Though that would unfortunately open a gigantic can of worms... Conjugate momentum is technically non-physical: it is not gauge invariant, and anything that we measure must be gauge invariant. What we actually measure is kinetic momentum, which is a different thing. Conjugate momentum is basically "the other variable we use to keep track of states". The Hamiltonian has similar status. My take is that going through these issues would only confuse things...

    • @CstriderNNS
      @CstriderNNS 6 ปีที่แล้ว +3

      he clearly says he does in part 2, ya have to finish the video,its at the end.

    • @MrSuvidh
      @MrSuvidh 4 ปีที่แล้ว

      I think he says that in this video he will stick to geometric perspective only.

    • @andrewfetterolf7042
      @andrewfetterolf7042 2 ปีที่แล้ว

      @@gcarcassi right thats what we want, we want to open the whole can of worms and know what each one tastes like, so math can be a USEFUL TOOL and not a secret code. If the viewer of the math doesnt have and definitions for the symbols than the symbols are arbitrary and i believe they are not arbitrary but are the constitution of the terms in the equations

    • @gcarcassi
      @gcarcassi  2 ปีที่แล้ว +1

      @@andrewfetterolf7042 :-) Part of my research is indeed going back to the foundations of mathematics and trace all the mathematical definitions from the very beginning, so that we have a clear physical motivation for all of them, and so that we only have mathematical objects with physical motivations (see assumptionsofphysics.org/book/ ). It's a long journey... What is a real number? What is a manifold? What does a differentiable structure represent? All of these things need to be answered to do it properly. But it can be done and gives so much insight in both math and physics. The problem is having the time and resources to create other videos... :-/

  • @hfkssadfrew
    @hfkssadfrew 4 ปีที่แล้ว

    Incompressibility will give you a nice transport equation of density, which further leads to liouville equation.

  • @Doat876
    @Doat876 8 ปีที่แล้ว +1

    If you define the metric of phase space differently, you do not need to rotate the S.

  • @geoffrygifari3377
    @geoffrygifari3377 3 ปีที่แล้ว

    There's something i noticed about that last part, the flux across any closed boundary is zero == any region evolving in time preserves area (i believe this is "Liouville's theorem"?). would the divergence of [dH/dp, -dH/dx] have any special consequence if the phase space trajectory/constant H contour line *is not* an enclosed area?

    • @gcarcassi
      @gcarcassi  3 ปีที่แล้ว

      @Geoffry Gifari The divergence does not change if the H contour lines are not closed, it will be zero as well. The difference is that open lines represent "non-bound states". If you imagine you have a potential well, the "bound states" will keep circling inside the well, while the others will escape the well and fly off to infinity. That will correspond to closed and open contour lines. Does that help?

    • @geoffrygifari3377
      @geoffrygifari3377 3 ปีที่แล้ว +1

      @@gcarcassi oh i got the picture now. Area enclosed is something that's defined for bound states only

  • @solsticetwo3476
    @solsticetwo3476 5 ปีที่แล้ว +2

    Thanks. You are the man!. The series is the best.

  • @kaursingh637
    @kaursingh637 4 ปีที่แล้ว

    best on the internet - excellent -pls solve examples

  • @GaryWilbourn1
    @GaryWilbourn1 10 ปีที่แล้ว

    Good starter series on the subject of Hamiltonian mechanics.

  • @pablomartinezazcona8450
    @pablomartinezazcona8450 6 ปีที่แล้ว

    Great series of videos!! I've just studied a little bit of Lagrangian & Hamiltonian formalism (though the last one was merely an introduction) and this series of videos have made me understand it in more depth. Thanks!!

    • @epaulk1969
      @epaulk1969 2 ปีที่แล้ว

      Can you recommend any good books? I downloaded this free book on classical field theory written by some guy from the university of Utah. It wasn't bad until he got to the chapter on Hamilton mechanics. That's why I'm here.

  • @pennyl.8799
    @pennyl.8799 8 ปีที่แล้ว

    Super series! I look forward to going through all of them. I came across a paper of yours this week.

  • @Mikey-mike
    @Mikey-mike 4 ปีที่แล้ว +1

    Is your rot90degrees the same as curl? Or i?

    • @gcarcassi
      @gcarcassi  4 ปีที่แล้ว +1

      Not exactly. The gradient rotated 90 degrees kind of looks like a curl in 2D, but it does not generalize correctly... If you imagine each d.o.f. represented as q + i p, then multiplying by i is like rotating 90 degrees... but I think that complicates things in the end.

    • @Mikey-mike
      @Mikey-mike 4 ปีที่แล้ว

      @@gcarcassi
      Yes, thank you.
      Then it is a rotation matrix, or covector.
      I have never seen that notation before.
      Thanks for your brilliant idea and analysis.

  • @geoffrygifari3377
    @geoffrygifari3377 3 ปีที่แล้ว

    Woah i've never seen hamiltonian mechanics explained like this!

  • @rnoro
    @rnoro 2 ปีที่แล้ว +1

    very nice explanation! Just find your videos recently:)

    • @gcarcassi
      @gcarcassi  2 ปีที่แล้ว

      Thanks and welcome to the channel! :-)

  • @GodBlade132
    @GodBlade132 3 ปีที่แล้ว

    Excellent explanation and visualisation, thank you!

  • @abcxyz4207
    @abcxyz4207 4 ปีที่แล้ว

    I need Lagrange and Hamilton for Advanced control theory courses and never learned this

  • @jimkarkaloutsos4267
    @jimkarkaloutsos4267 3 ปีที่แล้ว +1

    Well explained Sir

  • @pir0kunn
    @pir0kunn 4 ปีที่แล้ว

    Hello, thank you for the informative explanation. Could you perhaps suggest any literature related to the topic? Thank you!

    • @gcarcassi
      @gcarcassi  4 ปีที่แล้ว

      Hi Hory! The literature is vast... what type level/particular issue are you interested in?

  • @janlang8605
    @janlang8605 3 ปีที่แล้ว

    Really nice videos! Well done.

  • @fabricio.ferrari
    @fabricio.ferrari 2 ปีที่แล้ว

    great video series. Thanks

  • @MrPythonnn
    @MrPythonnn 5 ปีที่แล้ว

    really good series
    thanks your presentation.
    I love it.and waiting more❤️❤️❤️

  • @raulotero9706
    @raulotero9706 7 ปีที่แล้ว

    That was a great explanation! Thanks!

  • @death0intj
    @death0intj 11 ปีที่แล้ว +2

    great material!

  • @stephenanastasi748
    @stephenanastasi748 3 ปีที่แล้ว

    I guess I was wanting an explanation for physicists. Reference to S and whatever, just doesn't work for me.

    • @gcarcassi
      @gcarcassi  3 ปีที่แล้ว

      Well, this is the first video of a series, and is titled "the math"... then you have other videos: "measurements", "thermodynamics", ...

    • @stephenanastasi748
      @stephenanastasi748 3 ปีที่แล้ว

      @@gcarcassi Thanks. I wasn't complaining, just identifying that the focus on the maths without the physical context lost me. Nothing to hang my hat on.

  • @NovaWarrior77
    @NovaWarrior77 4 ปีที่แล้ว +2

    Excellent.

  • @vinitchauhan973
    @vinitchauhan973 6 ปีที่แล้ว

    I love this video so much thank you man.

  • @seungsoolee1949
    @seungsoolee1949 7 ปีที่แล้ว

    Hi, I think your videos saved my life. But I think it's rot-90 (grad(H(x,p))). Please correct me if I am wrong. Thank you so much

    • @gcarcassi
      @gcarcassi  7 ปีที่แล้ว +1

      Glad they helped!!! As for the sign of the angle, it's 90 degrees clockwise. If you follow the right-hand rule then it's -90. See en.wikipedia.org/wiki/Right-hand_rule . If you follow the left-hand rule then it's +90. Makes sense?

  • @lucasdevries5131
    @lucasdevries5131 3 ปีที่แล้ว

    Normal kids my age: social media, friends
    Me: maths

  • @pampasupas
    @pampasupas 10 ปีที่แล้ว +1

    Excellent
    Thank you very much

  • @the_infinite_lagrangian
    @the_infinite_lagrangian 9 ปีที่แล้ว

    seriously? why using x instead of q?

    • @gcarcassi
      @gcarcassi  9 ปีที่แล้ว +11

      Both x and q are used in the literature, with different frequency in different context. For example, in classical mathematical physics it usually is q, while in quantum mechanics and intro classical physics is usually x. The greek letter xi is also used in symplectic geometry for both q and p. The intended audience of the videos is intro level physics, so I went for what I believe is most common in that case.

    • @angelmendez-rivera351
      @angelmendez-rivera351 5 ปีที่แล้ว

      Victor Ovidiu Slupic Because q refers to any generalized coordinate system that applies to keep track of the state of the object of study. The implication behind using x is that those coordinates are Cartesian, but this is not necessarily the case, so using x would be misleading. It is for the same reason r is used instead of x to denote position in multiple dimensions.

  • @anoopsrana
    @anoopsrana 10 ปีที่แล้ว

    Good video to start

  • @gcarcassi
    @gcarcassi  11 ปีที่แล้ว +6

    Thanks! :-)

  • @jamesmonteroso824
    @jamesmonteroso824 4 ปีที่แล้ว

    excellent! thank you

  • @sarojpanday6236
    @sarojpanday6236 5 ปีที่แล้ว +1

    Amazing

  • @sheblywallyd8845
    @sheblywallyd8845 5 ปีที่แล้ว +1

    99 percent of teachers dont know what is Hamiltonian mechanics haha

  • @scissorhold1485
    @scissorhold1485 6 ปีที่แล้ว

    This is phisics

  • @andrewfetterolf7042
    @andrewfetterolf7042 2 ปีที่แล้ว

    You MUST define ALL of your terms or you dont have math you have a secret code which is only useful to writer

  • @natjimoEU
    @natjimoEU 7 ปีที่แล้ว

    you disurve a cookie sir !