3
$\begingroup$

I was given this question in class and I assume it is a spin off of Green's theorem for finding the area of a closed curve $\lambda$ in 2D but expanded to 3D I believe. Anyways I am pretty confused about it so if anyone could help I would appreciate it,.

Question: Let $\lambda$ be a simple closed smooth space curve that lies in a plane with unit normal vector n = (a, b, c) and has positive orientation with re- spect to the normal vector n of the plane. Show that the plane area enclosed by $\lambda$ is $$\frac12 \int_{\lambda} (bz-cy)dx + (cx-az)dy + (ay-bx)dz$$

$\endgroup$
3
$\begingroup$

The Green's theorem we know takes the form for the area enclosed by a closed curve:

$$A = \iint_A dx \, dy = \frac{1}{2} \oint_C (-y\, dx + x\, dy) = \frac{1}{2} \oint_C \vec{F}\cdot d\vec{r}$$

where $\vec{F}=(-y,x,0)$ and $d\vec{r} = (dx,dy,dz)$. Note that this is for the $xy$ plane, in which the normal vector is $\hat{n}=(0,0,1)$. You can see that, for a position vector $\vec{r}=(x,y,z)$:

$$\hat{n} \times \vec{r} = \vec{F}$$

This is true in general. For an arbitrary plane in which $\hat{n}=(a,b,c)$,

$$\vec{F}=\hat{n} \times \vec{r} = (b z-c y,c x-a z,a y-b x)$$

Therefore

$$A = \frac{1}{2} \oint_C \vec{F}\cdot d\vec{r} = \frac{1}{2} \oint_C [(b z-c y) dx + (c x-a z) dy + (a y-b x) dz]$$

as was to be shown.

$\endgroup$
  • 1
    $\begingroup$ Ok, I understand that dr = (dx,dy,dz). Where I am confused is how n x r = F...? So n = (a,b,c) is normal to the plane curve. Is r = (x,y,z) just a position vector that is on the surface of the curve? I see how it all comes together but I just can't see where r x n = F comes from. $\endgroup$ – user68203 Apr 11 '13 at 6:32
  • $\begingroup$ I was trying to illustrate by analogy to the simple case you already know. $\hat{n} \times \vec{F} \cdot d\vec{r}$ is coordinate independent, so if it computes an area in the $xy$ plane, it does so for any plane. $\endgroup$ – Ron Gordon Apr 11 '13 at 6:51
  • 1
    $\begingroup$ where does the 1/2 come from? $\endgroup$ – user68203 Apr 12 '13 at 14:40
  • $\begingroup$ Each of the area elements of the integrand represents an infinitesimal triangular element with respect to a common 3rd point in the interior of the closed curve; the area of this element is one-half of a determinant formed by these points, which turns out to be the integrand. See, for example, this: people.richland.edu/james/lecture/m116/matrices/… $\endgroup$ – Ron Gordon Apr 12 '13 at 14:45
  • 1
    $\begingroup$ Ok, so it comes from taking small triangles from the surface and when you find the area you just take the 1/2 out of the integration...? $\endgroup$ – user68203 Apr 12 '13 at 14:49
0
$\begingroup$

This is immidiate using the stoke's theorem where $S$ being surface measure and $A$ be surface enclosed by the curve, you take $$ f(x,y,z) = (1/2)(bz-cy,cx-az,ay-bx) $$ And you find $ curl(f) = (a,b,c) = n $. and you have $n.n = 1 $ and $\partial A = \lambda $. so from stoke's theorem you get area of the surface S(A) as $$ S(A) = \int_A dS = \int_A curl(f).n dS = \int_{\partial A} f.dr = \frac{1}{2}\int_\lambda (bz-cy)dx + (cx-az)dy + (ay-bx) dz $$

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.