If you're seeing this message, it means we're having trouble loading external resources on our website.

If you're behind a web filter, please make sure that the domains *.kastatic.org and *.kasandbox.org are unblocked.

Main content

2d curl nuance

The meaning of positive curl in a fluid flow can sometimes look a bit different from the clear cut rotation-around-a-point examples discussed in previous videos. Created by Grant Sanderson.

Want to join the conversation?

  • hopper cool style avatar for user Iron Programming
    The video's title is nuance but Grant never says what Nuance is. Can I get an exact definition of 2D Curl Nuance please?
    (4 votes)
    Default Khan Academy avatar avatar for user
  • leaf green style avatar for user Adrian
    in the video, where does the forumla [-y x] come from?
    (4 votes)
    Default Khan Academy avatar avatar for user
  • male robot hal style avatar for user Julian
    Although the paddle wheel idea is a great educational tool for curl, I can't help but feel that it doesn't really hold true in a real-world setting (hence, Grant's comment of counterintuitiveness). Just curious. Maybe I'm wrong. Any thoughts?

    (Thanks for the videos Grant!)
    (2 votes)
    Default Khan Academy avatar avatar for user
    • blobby green style avatar for user Igor Medvedyev
      Introduction of the paddle wheel would change the local flow significantly in a real life example. (Thus changing the problem definition) And that change to the flow would be different for different points in the flow field. However, assuming that the flow is exactly as given by the formula in the example, and unaffected by the introduction of the paddle wheel, the paddle in the center would rotate with the same angular velocity anywhere in the flow field.
      (7 votes)
  • blobby green style avatar for user rickytan
    How does one find the "center" of curl (the point the fluid is rotating about)? For example, the fluid in the video rotates about the origin, but the curl is the same regardless of what point it's evaluated at. Is there another way to determine that the origin is the "center"? Also, what if there are multiple "centers"?
    (4 votes)
    Default Khan Academy avatar avatar for user
    • blobby green style avatar for user asad.qzr
      My best guess would be to take that function which we can use to find the rotation:

      2d-curl(v) = ∂Q/∂x - ∂P/∂y

      And that is but a function of x,y itself. And the defining property of the 'centre' you describe (If I am not mistaken) is the local maximum of that multivariable function, the place where we have that x,y function most fitting to the perfect example!

      This must mean that you just have to find the point on that x,y function [ let's call it U(x,y) ] where

      ∂U/∂x = 0,
      ∂U/∂y = 0.

      Which implies
      ∂U/∂x + ∂U/∂y = 0.

      The expression on the right is again a multivariable function, find where it is equal to zero, and you have all your centres.

      P.S. this was all on the spot typing and thinking, so if I am wrong, please correct me.
      (0 votes)
  • blobby green style avatar for user Qinggao Hong
    I don't understand why the rotations of the two regions are the same visually.
    (2 votes)
    Default Khan Academy avatar avatar for user
  • leaf green style avatar for user apurba12107
    Up till now the 2D curl has given us an intuitive concept about the direction of rotation. But I want to know how curl might give us a quantitative approach in measuring the magnitude of rotation?
    (2 votes)
    Default Khan Academy avatar avatar for user
  • male robot johnny style avatar for user MustardManExtremeTurboEdition
    got a bit of a question regarding how the curl would be shaped if both dq/dx and dp/dy were positive but dp/dy was lesser in magnitude than dq/dx. Would this not imply a positive curl but result in a shape different from that typical counterclockwise rotation? As x goes up, the flow of the field will push things up because of the dq/dx being positive, but then would that increase in vertical distance push things to the right and result in just an infinitely long, not very curly part of the vector field?

    I'm mostly asking cuz I have a pretty hard time visualizing these things.

    Also kinda curious how zero dp/dy would work, as it should just mean the function does not move to the right or left at all with a change in height which kinda just... means no curling at all and just straight lines. How would you still have positive curl with that?

    This could just be a massive misinterpretation of what curl really is, but it doesn't feel right to call it a positive curl if it doesn't curl. Does the formula for curl only apply as a means to measure that curl and not identify if there is any curling?
    (1 vote)
    Default Khan Academy avatar avatar for user
  • blobby green style avatar for user saurabh.vaastav
    Maybe a dumb question but wanted to know what is the meaning of the blue dots moving in the animation?
    Does it mean coordinates increasing?
    (1 vote)
    Default Khan Academy avatar avatar for user

Video transcript

- [Voiceover] In the last couple videos I've been talking about curl where if we have a two dimensional vector field V defined with component functions P and Q, I've said that the 2D curl of that function V gives you a new function that also takes in x and y's inputs and it's formula is the partial derivative of Q with respect to x minus the partial derivative of P with respect to y. And my hope is that this is more than just a formula and that you can understand how this represents fluid rotation in two dimensions but what I want to do here is show how the original intuition I gave for this formula might be a little oversimplified because for example, if we look at this, the partial Q, partial x component, I said that you can imagine that Q at some point starting off a little bit negative so the y component of the output is a little negative, then as you move positively in the x direction, it goes to being zero and then it goes to being a little bit positive, and with this particular picture, it's hopefully a little bit clear why this can correspond to counter-clockwise rotation in the fluid, but this is only a very specific circumstance for what partial Q, partial x being positive could look like. It might also look like Q's starting a little bit positive and then as you move in the x direction, it becomes even more positive and then even more positive. And according to the formula, this should contribute as much to positive curl as this very clear-cut whirlpool example, and to illustrate what this might look like, if we take a look at this vector field here, if we look at the center, this is kinda the clear-cut whirlpool counter-clockwise rotation example. And if we play the fluid flow, the fluid does indeed rotate counter-clockwise in the region. But contrast that with what goes on over here on the right. This doesn't look like rotation in that sense at all. Instead the fluid particles are just kind of rushing up through it. But in fact, the curl in this region is going to be just as strong as it is over here, and I'll show that with a formula and kind of computing it through in just a moment, but the image that you might have in your mind is to imagine a paddle wheel of sorts where let's say it's got arms kind of like that and then you hold down with your thumb that middle portion, so even though the paddle wheel left to its own devices would just fly up, I want to say, let's say you're holding that down with your thumb but it's free to rotate. Then the vectors on its left are pointing up but less strongly than the vectors on its right which are even greater, so if you imagine that setup and you have your paddle wheel there, then when you play the fluid rotation, holding your thumb down but letting the paddle wheel rotate freely, it's also going to rotate just as it would over here in the easier to see whirlpool example. And in terms of the formula, this is because a situation like this one here, where Q goes from being negative to zero to positive, should be treated just the same as a situation like this as far as 2d curls is concerned, because this term in the 2d curl formula is going to come out the same for either one of these. And it's worth pointing out, by the way, curl isn't something that mathematicians and physicists came across trying to understand fluid flow. Instead, they found this term as being significant and various other formulas and circumstances and I think electromagnetism might be where it originally came about. But then in trying to understand this formula, they realized that you can give a fluid flow interpretation that gives a very deep understanding of what's going on beyond just the symbols themselves. So let me go ahead and walk through this example in terms of the formula representing the vector field. It's a relatively straightforward formula actually. So P and Q, that x component is going to be negative y and the y component, Q, is equal to x. So when we apply our 2d curl formula and apply the partial Q with respect to x, so partial of this second component with respect to x is just one, and then we subtract off the partial of P with respect to y which up here is negative one because P is just equal to negative y. So the 2d curl is equal to two and in particular, it's a constant two that doesn't depend on x and y, which is pretty unusual, most times that you apply 2d curl to a vector field you're going to get some kind of function of x and y. But the fact that this is constant, tells us that when we look over at this fluid flow, the sense in which curl, the formula for curl wants to say that rotation happens around the center is just as strong as it's supposed to happen over here on the right, or anywhere on the plane for that matter. So if we're playing this and if you imagine you have the paddle wheel in the center, evidently it would be rotating just as quickly as the paddle on the right even though it might, I don't know, to me, that feels a little unintuitive cause the one on the right I'm thinking, okay, you know, maybe there's a little bit more torque on the right side than there is on the left and that's kind of a counter-balancing act, but the idea that that's actually the same as the very clear-cut, I see the counter-clockwise rotation with my eyes, example in the center, does seem a little unusual. But I think it's important to understand what else two dimensional curl can look like and what else this formula might be representing. So with that, I'll see you next video.