- #1
InbredDummy
- 85
- 0
(1) The gradient of a differential function f: S --> R is a differentiable map grad f: S --> R^3 which assigns to each point p in S a vector grad f(p) in the tangent space of p s.t.
<grad f(p), v> at p = dfp(v) for every v in the tangent space Tp(S)
(a) If E, F, G are the coefficients of the first fundamental form in a parameterization X: U a subset o R^2 --> S, then grad f on X(U) is given by
grad f = [(df/du *G - df/dv *F)/(EG - F^2)]*dX/du + [(df/dv *E - df/du *F)/(EG - F^2)]*dX/dv
- Oh man, I'm really lost on this one. My book didn't really cover this too much, and I'm not sure what all the notation means.
(b) If you let p in S be fixed and v vary in the unit circle, ie |v| = 1 in Tp(S), then dfp(v) is maximum iff v = grad f/|grad f|
(c) if grad f is nonzero at all points of the level curve C = {q in S | f(q) = constant}, then C is a regular curve on S and grad f is normal to C at all pojnts of C.
Notation: S is a regular surface, ie. a manifold. X is a mapping from the uv plane to the surface S such that it is differentiable and a homeomorphism. Tp(S) is the tangent plane of p for a given p in S.
(2) Show that at a hyperbolic point (a point who Gaussian curvature is less than zero), the principal directions bisect the asymptotic directions.
Pf.
I have that the asymptotes are indeed the asymptotic directions and via the Dupin indicatrix, i see that principal directions e1 and e2 which are eigenvectors for our curvatures k1 and k2 which determine K do indeed bisect the asymptotes, but I don't have no godly reason why they do.
(3) Let C be a subset of a regular surface S with Gaussian curvature K > 0. Show that the curvature k of C at p satisfies:
k is greater than or equal to min (|k1|, |k2|)
where k1 and k2 are principal curvatures of S at p.
Pf.
Again, kinda stuck like the previous problem
(4) Show that the mean curvature H at p in S is given by:
H = 1/pi * int from 0 to pi (kn(theta)) d(theta) where kn(theta) is the normal curvature at p along a direction making an angle theta with a fixed direction.
- What exactly is normal curvature? My book uses the term at will and I don't understand what it means in this context?
(5) Show that the sum of normal curvatures for any pair of orthogonal directions, at a point p in S, is constant.
- ??
(6) Prove that the absolute value of the torsion T at a point of an asymptotic curve, whose curvature is nowhere zero, is given by:
|T| = sqrt(-K) where K is the Gaussian curvature of the surface at a given point.
Sorry for not showing any work I've done, but I'm really really stumped on these problems.
<grad f(p), v> at p = dfp(v) for every v in the tangent space Tp(S)
(a) If E, F, G are the coefficients of the first fundamental form in a parameterization X: U a subset o R^2 --> S, then grad f on X(U) is given by
grad f = [(df/du *G - df/dv *F)/(EG - F^2)]*dX/du + [(df/dv *E - df/du *F)/(EG - F^2)]*dX/dv
- Oh man, I'm really lost on this one. My book didn't really cover this too much, and I'm not sure what all the notation means.
(b) If you let p in S be fixed and v vary in the unit circle, ie |v| = 1 in Tp(S), then dfp(v) is maximum iff v = grad f/|grad f|
(c) if grad f is nonzero at all points of the level curve C = {q in S | f(q) = constant}, then C is a regular curve on S and grad f is normal to C at all pojnts of C.
Notation: S is a regular surface, ie. a manifold. X is a mapping from the uv plane to the surface S such that it is differentiable and a homeomorphism. Tp(S) is the tangent plane of p for a given p in S.
(2) Show that at a hyperbolic point (a point who Gaussian curvature is less than zero), the principal directions bisect the asymptotic directions.
Pf.
I have that the asymptotes are indeed the asymptotic directions and via the Dupin indicatrix, i see that principal directions e1 and e2 which are eigenvectors for our curvatures k1 and k2 which determine K do indeed bisect the asymptotes, but I don't have no godly reason why they do.
(3) Let C be a subset of a regular surface S with Gaussian curvature K > 0. Show that the curvature k of C at p satisfies:
k is greater than or equal to min (|k1|, |k2|)
where k1 and k2 are principal curvatures of S at p.
Pf.
Again, kinda stuck like the previous problem
(4) Show that the mean curvature H at p in S is given by:
H = 1/pi * int from 0 to pi (kn(theta)) d(theta) where kn(theta) is the normal curvature at p along a direction making an angle theta with a fixed direction.
- What exactly is normal curvature? My book uses the term at will and I don't understand what it means in this context?
(5) Show that the sum of normal curvatures for any pair of orthogonal directions, at a point p in S, is constant.
- ??
(6) Prove that the absolute value of the torsion T at a point of an asymptotic curve, whose curvature is nowhere zero, is given by:
|T| = sqrt(-K) where K is the Gaussian curvature of the surface at a given point.
Sorry for not showing any work I've done, but I'm really really stumped on these problems.