4

I am recently taking the undergraduate version of differential geometry. This week we finished the Compatibility Equations. Actually the contents given in the lecture are quite different from the book I read, Differential Geometry of Curves and Surfaces, by De Carmo.

In this book, the author takes the following equations as the Gauss Formula and Mainardi-Codazzi Equations. $$(\Gamma_{12}^2)_u - (\Gamma_{11}^2)_v + \Gamma_{12}^1\Gamma_{11}^2 + \Gamma_{12}^2\Gamma_{12}^2 -\Gamma_{11}^2\Gamma_{22}^2 - \Gamma_{11}^1\Gamma_{12}^2 = -EK,$$ $$e_v - f_u = e\Gamma_{12}^1 + f(\Gamma_{12}^2 - \Gamma_{11}^1) + g\Gamma_{11}^2,$$ $$f_v - g_u = e\Gamma_{22}^1 + f(\Gamma_{22}^2 - \Gamma_{12}^1) + g\Gamma_{12}^2.$$ where $\Gamma_{ij}^K$ is the Christoffel Symbols and $E,F,G,e,f,g$ are the coefficients of the first fundamental form and the second fundamental form respectively.

The version given in the lecture: $$\langle X_u,RX_v \rangle = K(EG-F^2),$$ $$\nabla_{\partial u}(\zeta X_v) = \nabla_{\partial v} (\zeta X_u),$$ where $K$ is the Guassian curvature, $\nabla V = (dV)^T = dV - \langle dV, N \rangle N$ and $\zeta$ is the weingarten operator. $X$ is the corresponding parametrization compatible with the orientation. Moreover, $R$ is the curvature tensor.

In fact, I do not know what is the meaning of tensor, so let me just give the definition the lecturer presented.

Definition. $$RV = \nabla_{\partial u}\nabla_{\partial v}V - \nabla_{\partial v}\nabla_{\partial u}V.$$

Could anyone explain the connection between the two sets of equations? How to show they are equivalent? Thanks.

Siamese
  • 197
  • Are you sure you typed the first equation correctly? I think the first one might be wrong (I did some calculations - which, granted, might be wrong - and it seems to imply $eg - f^2 = 0$ always, which is false) – Matheus Andrade Jul 17 '21 at 04:31
  • @MatheusAndrade You mean the first equation of the first set or the first equation of the second set? For the one in the first set, I checked the book again, I am sure there is no problem with it. For the one in the second set, honestly, I am not sure if it is correct (maybe I copied it wrong...) – Siamese Jul 17 '21 at 05:52
  • I mean the equation whose righthand side has $K(EG-F^2)$. But the definition for $R$ is weird too, it should take as input three vector fields and not just one. – Matheus Andrade Jul 17 '21 at 06:07
  • @MatheusAndrade Well, I checked the lecture note, the equation is indeed correct. I think the symbol $\nabla$ does not stand for "gradient" but the Levi-Civita connection along the surface $X$. By the way, I've already checked those equations listed above by myself and all of them are correct. – Siamese Jul 17 '21 at 06:31
  • I see, I must have made some mistake then. I'll try again with a fresher mind later. – Matheus Andrade Jul 17 '21 at 06:35
  • 2
    This is a reasonable question on its face, but offhand I don't see how to answer it without effectively developing the geometry of surfaces in some detail (e.g., moving frames) or otherwise building a dictionary between Do Carmo's and the lecturer's respective formalisms, and then either doing a lengthy computation or showing the two sets of equations manifest the same geometric phenomena. If this assessment is right, it may be hard to get a good answer here. – Andrew D. Hwang Jul 17 '21 at 16:13

0 Answers0