Hi I am studying minimal surfaces from Kuhnel's Curves - Surfaces Manifolds

The theorem states that: If $f$ is a conformal parametrization, $f$ is a minimal surface if and only if the functions $\phi_{1},\phi_{2}, \phi_{3}$ are holomorphic.

So he defines a surface element $f: U \rightarrow \mathbb{R}^3$ with components $f = (f_{1},f_{2},f_{3})$ and defines the map $\phi(u+iv) = \frac{\partial f}{\partial u}(u,v) - i\frac{\partial f}{\partial v}(u,v)$, which is in components for $j = 1,2,3$ \begin{align*} \phi_{j} (u +iv) = \frac{\partial f_{j}}{\partial u}(u,v) - i\frac{\partial f_{j}}{\partial v}(u,v) \ (*) \end{align*}

For showing the theorem I assume the result: $f$ defines a minimal surface $\iff$ the three component functions $f_{1}, f_{2}, f_{3}$ of $f$ have the relation

\begin{align} \frac{\partial^2 f_{i}}{\partial u_{1}^2} + \frac{\partial^2 f_{i}}{\partial u_{2}^2} = 0 \ (**) \end{align}

So all we have to do is take the second derivatives and show that their sum equals zero Now his proof is:

\begin{align*} \frac{\partial^2 f_{k}}{\partial u^2} &= \frac{\partial}{\partial u}(Re \ \phi_{k}), \ \frac{\partial^2 f_{k}}{\partial v^2} = - \frac{\partial}{\partial v}(Im \ \phi_{k}), \\ \frac{\partial^2 f_{k}}{\partial u \partial v} &= \frac{\partial}{\partial v}(Re \ \phi_{k}) = - \frac{\partial}{\partial u} (Im \ \phi_{k}) \end{align*} And from that he concludes that (**) is satisfied

Now firstly I am confused about if he is differentiating (*)? w.r.t. u and v But in that case I can't really make sense of the calculations, any help would be appreciated

  • 1
    $\begingroup$ Please do not vandalise posts, even if they are your own. $\endgroup$
    – postmortes
    Sep 30 '19 at 18:06
  • $\begingroup$ no reply, then what can I do $\endgroup$
    – Chengdu
    Sep 30 '19 at 21:06
  • 5
    $\begingroup$ $1$) Wait a little longer than $3$ days. If you were asking about elementary calculus then you could expect a near instantaneous response as there as a lot of people who want to answer those questions. Other questions might need some time for the right person to spot them -- not everyone is on the site all the time. $2)$ If you're not happy with the level of response you can add a bounty to the question to draw more attention. $3)$ you can try asking in some of the chat-rooms (there are general maths one) to draw more attention. Vandalising it guarantees you won't get help though. $\endgroup$
    – postmortes
    Oct 1 '19 at 6:20

Starting with the definition: $$\phi_k(u,v) = \frac{\partial f}{\partial u}(u,v) - i \frac{\partial f}{\partial v}(u,v)\qquad \mbox{for }k=1,2,3$$ we see that $\phi_k$ is being explicitly written as a sum of real and imaginary parts. So, $$ \mbox{Re}(\phi_k) = \frac{\partial f}{\partial u} $$ and $$ \mbox{Im}(\phi_k) = -\frac{\partial f}{\partial v} $$ So if we differentiate both sides with respect to $u$ for the real part and $v$ for the imaginary part, we get $$ \frac{\partial f}{\partial u}\mbox{Re}(\phi_k) = \frac{\partial^2 f}{\partial u^2} $$ and $$ \frac{\partial f}{\partial v}\mbox{Im}(\phi_k) = -\frac{\partial^2 f}{\partial v^2} $$ and for the mixed derivatives (which are equal thanks to us being $C^1$) $$ \frac{\partial f}{\partial u}\mbox{Im}(\phi_k) = -\frac{\partial^2 f}{\partial v \partial u} = -\frac{\partial^2 f}{\partial u \partial v} = -\frac{\partial f}{\partial v}\mbox{Re}(\phi_k) $$

Now write out the second derivative and use these relations to substitute in -- you'll find that everything cancels and leaves you with $0$. All without ever knowing what $f$ and $\phi$ explicitly look like!


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.