2
$\begingroup$

The equation I'm refering to was posed by Lagrange and states that if $S\subset \mathbb{R^3}$ is a surface with zero mean curvature, ie $H=0, \: \: \forall p \in S$ and S is given as the graph of a function $z(x,y)$ then $$ (1+z_x^2)z_{yy}-2z_xz_yz_{xy}+(1+z_y^2)z_{xx}=0 $$ I'm looking for a proof of this. I thought I would find it in Do Carmo's Differential Geometry of Curves and Surfaces but it seems to not be on there.

$\endgroup$
4
$\begingroup$

I'm going to change $x$ and $y$ to $u$ and $v$, respectively, so that I can call my patch $\textbf{x}$ without confusion.

Parameterize the surface via the patch $\textbf{x}$ given by $(u,v)\mapsto (u,v,z(u,v)).$ We compute $$\textbf{x}_u=(1,0,z_u),$$ $$\textbf{x}_v=(0,1,z_v),$$ and obtain the components of the first fundamental form as \begin{align*}E&=\textbf{x}_u\cdot\textbf{x}_u=1+z_u^2\\F&=\textbf{x}_u\cdot \textbf{x}_v=z_uz_v\\ G&=\textbf{x}_v\cdot \textbf{x}_v=1+z_v^2.\end{align*} The unit normal is given by $$U=\frac{\textbf{x}_u\times \textbf{x}_v}{\sqrt{EG-F^2}}=\frac{(-z_{u},-z_{v},1)}{\sqrt{1+z_u^2+z_v^2}}.$$ Next, we find that $$\textbf{x}_{uu}=(0,0,z_{uu}),$$ $$\textbf{x}_{uv}=(0,0,z_{uv}),$$ and $$\textbf{x}_{vv}=(0,0,z_{vv}).$$ We can now compute the quantities \begin{align*} L&=\textbf{x}_{uu}\cdot U=\frac{z_{uu}}{\sqrt{1+z_u^2+z_v^2}}\\ M&=\textbf{x}_{uv}\cdot U=\frac{z_{uv}}{\sqrt{1+z_u^2+z_v^2}}\\ N&=\textbf{x}_{vv}\cdot U=\frac{z_{vv}}{\sqrt{1+z_u^2+z_v^2}}. \end{align*} Finally, we can just use the formula $$H=\frac{GL+EN-2FM}{2(EG-F^2)}=\frac{(1+z_v^2)z_{uu}+(1+z_u^2)z_{vv}-2(z_uz_v)z_{uv}}{2(1+z_u^2+z_v^2)^{3/2}},$$ which equal zero if and only if $$(1+z_v^2)z_{uu}+(1+z_u^2)z_{vv}-2(z_uz_v)z_{uv}=0,$$ as desired.

Reference for formulas: Barrett O'Neil, Elementary Differential Geometry

$\endgroup$
  • $\begingroup$ This is perfect. Thanks! $\endgroup$ – D. Brito Jul 25 at 17:13
  • $\begingroup$ You're very welcome, I'm glad I could help! $\endgroup$ – cmk Jul 25 at 17:18
  • $\begingroup$ I'm sorry, I was checking it more carefully and there seems to be a minor mistake as $U$ should be given by: $U = (-z_u, -z_v, 1)/\sqrt{1+z_u^2+z_v^2}$. (the numerator was wrong) $\endgroup$ – D. Brito Jul 26 at 16:02
  • $\begingroup$ Yup, typo. Thanks! $\endgroup$ – cmk Jul 26 at 16:07

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.