2
$\begingroup$

Let $P$ be a topologically embedded submanifold in a Riemannian manifold $M$. Then the tube $T(P, r)$ of radius $r \geq 0$ about $P$ is the set of all points $m \in M$ such that there exists a geodesic $\xi$ of length $\leq r$ from $m$ that meets $P$ orthogonally.

If $M = R^n$ or $M = S^n$, Weyl gave in "The volume of tubes", 1939 a formula for the volume of $T(P, r)$ for small $r$ (see also the book of Alfred Gray "Tubes" gives lots of background and formulas related to this problem). In particular, for $M = R^n$, the volume $V_{P} ( r )$ turns out to be a polynomial in $r$, whose coefficients carry some interesting geometric information related to intrinsic curvature (e.g. the coefficient of $r^{n}$ turns out to be of the form $f(n) \chi(P)$, where $\chi(P)$ is the Euler characteristic of $P$).

I am actually interested in $\bar{V}_P$ defined as the volume of the r-neighbourhood: $P + B_r$ in $R^n$, which is different from the $T(P, r)$ if P has a boundary. See this image for a visualization of the difference.

A rough approximation would suggest $$ \bar{V}_{P} (r) = V_{P} (r) + \frac{1}{2} V_{\partial P} (r) + \theta ( r ) $$ for some correction function $\theta$ that is small for small $r$.

For example, if $P$ is convex, then the Steiner formula gives $\theta ( r ) = \omega_n r^n$, where $\omega_n$ is the volume of the $n$-dimensional 1-ball. On the other hand, $\theta ( r ) = 0$ if $P$ is half a sphere.

Does there exist a formula for $\bar{V}_{P}$?

$\endgroup$

1 Answer 1

1
$\begingroup$

Apparently, the boundary corrections were computed by D. Q. Naiman - see this paper by Catherine Loader.

$\endgroup$
1
  • 1
    $\begingroup$ Here is a link to the Naiman paper. $\endgroup$
    – Eigentime
    Jul 24, 2015 at 4:35

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.