Integrating $\overline{v} = (-y/(y^2+x^2), x/(y^2+x^2))$ over the circle $x^2+y^2=1$?

49 Views Asked by At

I want to show that $I=\displaystyle \oint_C \overline{v} \cdot d \overline{r}$ is multivalued where $\overline{v} = (-y/(y^2+x^2), ~x/(y^2+x^2))$ and $C$ is the circle $x^2+y^2=1$. I know we can't use Green's theorem because $\overline{v}$ is not defined at the origin, which is the centre of the circle. But mindlessly applying Green's theorem gives $I=0$. Meanwhile, the parametrisation $x = \cos{\theta}, ~ y= \sin{\theta}$ gives $I= 2\pi$. But I can't provide an actual proof for this.

How do you prove that it's multivalued/it can take either of these values? I'm not sure I can say that when we take away the origin, Green's theorem is valid so $I=0$ and when we include the origin, the parametrisation gives the correc answer $I=2\pi$ since it avoids this problem. Thanks for your help!

2

There are 2 best solutions below

2
On BEST ANSWER

Green's Theorem doesn't apply, because its hypotheses include continuity of the partial derivatives of the integrand inside the entire region bounded by $C$. As for computing the integral directly, that looks relatively straight-forward. With the parametrization you've given, we have $$\mathbf{v}(\mathbf{r})=(-\sin(\theta),\cos(\theta)),\quad \mathbf{r}=(\cos(\theta),\sin(\theta)), \quad d\mathbf{r}=\mathbf{r}'\,d\theta=(-\sin(\theta),\cos(\theta))\,d\theta,$$ so that your integral becomes $$\oint\mathbf{v}\cdot d\mathbf{r}=\int_0^{2\pi}\mathbf{v}(\mathbf{r})\cdot\mathbf{r}'\,d\theta=\int_0^{2\pi}(-\sin(\theta),\cos(\theta))\cdot (-\sin(\theta),\cos(\theta))\,d\theta=\int_0^{2\pi}d\theta=2\pi,$$ as required.

0
On

Isn't multivalued in any reasonable sense of the word. Using Green isn't valid and $I = 2\pi$.

Quote from Wikipedia:

Let $C$ be a positively oriented, piecewise smooth, simple closed curve in a plane, and let $D$ be the region bounded by $C$. If $L$ and $M$ are functions of $(x,y)$ defined on an open region containing $D$ and have continuous partial derivatives there, then $$ \oint_C(Ldx + Mdy) = \iint_D(M_x - L_y)dxdy. $$

The circle minus the origin isn't "the region bounded by ..."