I am reading this paper that has the following:

Suppose $M$ is an (n-1) dimensional closed hyper surface immersed in $\mathbb{R}^{n}$. Let $e_1, \cdots, e_n$ be orthonormal frame in $\mathbb{R}^n$ such that $e_1,\cdots, e_{n-1}$ are tangent to $M$ and $e_n$ is the outer normal. Let $\omega_i$ be the corresponding coframes and $\omega_{i,j}$ be the connection forms. And use the same notation for the pull back of the forms through immersion. Then the second fundamental form is defined by the symmetric matrix $\{h_{ij}\}$ with $$\omega_{i,n+1}=h_{ij}\omega_j.$$

I use Boothby's An introduction to differentiable manifolds and Riemannian geometry for reference on differential geometry. In this book, the notations are like this:

  1. $e_1,\cdots,e_n$ are the orthonormal frames.
  2. $\omega^i$ are the coframes. (so I figured this is equal to $\omega_i$ in the paper.)
  3. $\omega_{j}^{k}$ are defined so that $\nabla_{X} e_j = \Sigma_k \omega_{j}^{k}(X)e_k$, where $X$ is a vector field.
  4. $\omega_{i,j}=\Sigma_k \omega_{i}^kg_{kj}$. In this case, since this is an orthonormal frame, I guess $\omega_{i,j}=\omega_i^j$.

I assume the symmetric matrix is defined by $h_{ij}=\langle e_n,\nabla_{e_i}e_j\rangle$. (Correct me if I am wrong.) Using 3 and 4, I get $h_{ij}=\omega_{j,n+1}(e_i)$. To arrive at the equation in the paper, I will have to justify (note that $\omega_j=\omega^j$) $$\Sigma_j \omega_{j,n+1}(e_i)\omega_j = \omega_{i,n+1}.$$

I am stuck at this step. Can anyone help me?

  • $\begingroup$ for curiosity, which paper were you reading? $\endgroup$ – cesare borgia Nov 8 '17 at 17:40

Basically, you have to prove that $\omega_{i, n}(e_l) = \omega_{l, n}(e_i)$, which is proved from $\omega^k_l(e_i) = \omega^k_i(e_l)$. Now, see that $\nabla_{e_l}e_i = \sum_k\Gamma^k_{li}e_k$, the usual Christoffel symbols. Seeing that the Christoffel symbols are symmetric (see Christoffel symbols and fundamental forms, for example), what you want follows.

  • $\begingroup$ I guess what the author meant by "orthonormal frame" is orthonormal list of vector fields induced from local parametrization. $\endgroup$ – henryforever14 Jun 30 '15 at 14:31

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.