Is this differential form closed / exact?

1.2k Views Asked by At

Could you check if I calculated the exterior derivative of this differential form $\omega$ correctly?

$\omega \in \Omega_2 ^{\infty} (\mathbb{R}^3 \setminus \{0\})$

$\omega = (x^2 + y^2 + z^2)^{\frac{-3}{2}}(x \mbox{d}y \wedge \mbox{d}z + y \mbox{d}z \wedge \mbox{d}x + z \mbox{d}x \wedge \mbox{d}y)$

$\mbox{d} \omega = \mbox{d}( (x^2 + y^2 + z^2)^{\frac{-3}{2}}x (\mbox{d}y \wedge \mbox{d}z) + (x^2 + y^2 + z^2)^{\frac{-3}{2}}(y \mbox{d}z \wedge \mbox{d}x ) + (x^2 + y^2 + z^2)^{\frac{-3}{2}}( z \mbox{d}x \wedge \mbox{d}y))$

We differentiate the fist summand only by $x$, second only by $y$ and the third only by $z$, because in the other cases we get forms like $\mbox{d}x \wedge \mbox{d}x$, and due to the fact that exterior derivative is antisymmentric such forms are zero.

So:

$\mbox{d} \omega = \left( - \frac{3}{2} (x^2 + y^2 + z^2)^{\frac{-5}{2}} 2x \cdot x + (x^2 + y^2 + z^2)^{\frac{-3}{2}}\right) \wedge \mbox{d}x \wedge \mbox{d}y \wedge \mbox{d}z + \left( - \frac{3}{2} (x^2 + y^2 + z^2)^{\frac{-5}{2}} 2y \cdot y + (x^2 + y^2 + z^2)^{\frac{-3}{2}}\right) \wedge \mbox{d}y \wedge \mbox{d}z \wedge \mbox{d}x + \left( - \frac{3}{2} (x^2 + y^2 + z^2)^{\frac{-5}{2}} 2z \cdot z + (x^2 + y^2 + z^2)^{\frac{-3}{2}}\right) \wedge \mbox{d}z \wedge \mbox{d}x \wedge \mbox{d}y$

Because $\mbox{d}x \wedge \mbox{d}y \wedge \mbox{d}z = \mbox{d}y \wedge dz \wedge dx = \mbox{d}z \wedge \mbox{d}x \wedge \mbox{d}y$ (we need two transpositions), we have:

$\mbox{d} \omega = \left( -3(x^2 + y^2 + z^2) ^{\frac{-5}{2}} (x^2 + y^2 + z^2) + (x^2 + y^2 + z^2) ^{\frac{-3}{2}} \right) \mbox{d}x \wedge \mbox{d}y \wedge \mbox{d}z$ But this isn't equal to zero. Are my calculations correct?

I have one more question - how do determine if this form is exact? I know a differential form $\omega \in \Omega_n (U)$ is exact if there exists $\beta \in \Omega_{n+1} (U)$ s. t. $\omega = \mbox{d} \beta$.

But I don't know how to guess such $\beta$.

Could you help me?

$\mbox{d} \omega = \left( -3(x^2 + y^2 + z^2) ^{\frac{-5}{2}} (x^2 + y^2 + z^2) + (x^2 + y^2 + z^2) ^{\frac{-3}{2}} \right) \mbox{d}x \wedge \mbox{d}y \wedge \mbox{d}z$

Thank you.

EDIT: There is a mistake in my calculations:

There should be

$\mbox{d} \omega = \left( -3(x^2 + y^2 + z^2) ^{\frac{-5}{2}} (x^2 + y^2 + z^2) + 3(x^2 + y^2 + z^2) ^{\frac{-3}{2}} \right) \mbox{d}x \wedge \mbox{d}y \wedge \mbox{d}z$

and this is zero, so the form is closed.

1

There are 1 best solutions below

4
On BEST ANSWER

I did not find any errors in your calculations. So it looks like $\omega$ is not closed.

And how to figure out if you have exact form, given it is closed?

In this case you know that $H^2_{dR}(\mathbb{R}^3 \setminus \{0\}) \simeq \mathbb{R}$ There is just one closed but not exact form(up to scalar multiple). And it is(I'think) $$\alpha = \star d \frac1{\sqrt{x^2+y^2+z^2}}.$$

If you suspect that your form $\omega$ is exact, you know that $\omega + \gamma \alpha$ has to be exact for some $\gamma \in \mathbb{R}$. You can calculate integral $$\int_{S^2} \omega + \gamma \alpha$$ and find $\gamma_0$ for which is the integral zero, than $\omega + \gamma_0 \alpha$ is exact. If $\gamma_0$ happens to be zero than $\omega$ is exact.


edit: Finding $\gamma_0$ is simple thanks to linearity of integral. $$ \gamma_0 = - \frac{\int_{S^2} \omega}{\int_{S^2} \alpha} $$ In most cases you just want to show that $\int_{S^2} \omega = 0$. Thanks to that you do not need to know exactly what $\alpha$ is.


edit2: How to compute $\int_{S^2} \omega$

You can think of integrating 2-forms(in 3d) as integrating vector field $\vec F$. $$\int \omega = \int \vec F \cdot \vec n dA$$ $\vec n$ is outer normal.

I found this question which discuss correspondence of 1,2-forms and vector fields.

So we can apply this to $\int_{S^2} \omega$

$$ \int_{S^2} \omega = \int_{S^2} (x^2 + y^2 + z^2)^{\frac{-3}{2}}(x \mbox{d}y \wedge \mbox{d}z + y \mbox{d}z \wedge \mbox{d}x + z \mbox{d}x \wedge \mbox{d}y)= $$ $$ = \int_{S^2} 1 n\cdot n dA = 4 \pi $$

because $\vec n = (x,y,z)$ on unit sphere.