Differentiable Function - Differential Geometry - Solved Exam, Exams for Computational Geometry. Aligarh Muslim University
sanjog
sanjog18 February 2013

Differentiable Function - Differential Geometry - Solved Exam, Exams for Computational Geometry. Aligarh Muslim University

PDF (84 KB)
6 pages
1000+Number of visits
Description
This is the Solved Exam of Differential Geometry which includes Normal Vector, Normal Vector, Binormal Vector, Curvature, Torsion, Binormal Vector, Speed Space Curve etc. Key important points are: Differentiable Functio...
20points
Download points needed to download
this document
Download the document
Preview3 pages / 6
This is only a preview
3 shown on 6 pages
Download the document
This is only a preview
3 shown on 6 pages
Download the document
This is only a preview
3 shown on 6 pages
Download the document
This is only a preview
3 shown on 6 pages
Download the document

SOLUTIONS TO FINAL EXAM

QUESTIONS FROM PREVIOUS YEARS.

Curves.

1. Let s be the arclength parameter.

γ′(t) = (1, f ′). ds

dt = |γ′(t)| =

√ 1 + f ′2.

ds =

1√ 1 + f ′2

(1, f ′)

d

dt

ds =

1√ 1 + f ′2

(0, f ′′) − f ′f ′′

(1 + f ′2)3/2 (1, f ′)

= 1

(1 + f ′2)3/2 (−f ′f ′′, f ′′)

d2γ

ds2 =

f ′′

(1 + f ′2)2 (−f ′, 1)

curvature = ∣∣∣∣ d2γ

ds2

∣∣∣∣ = f ′′

(1 + f ′2)3/2

Curvature of the parabola y = kx2 is

2k (1 + 4k2x2)3/2

.

2. Let t be the arclength parameter for α and let s be the arclength parameter for β(t) =

Typeset by AMS-TEX 1

2 SOLUTIONS TO FINAL EXAM QUESTIONS FROM PREVIOUS YEARS.

α′(t). Let t,n,b be the Frenet frame for α.

dt =

d2α

dt2 = kn.

ds

dt =

∣∣∣∣

dt

∣∣∣∣ = k. dβ

ds = n.

d

dt

ds =

dn dt

= −kt− τb. d2β

ds2 = −kt− τb

k = t− τ

k b.

curvature of β = ∣∣∣∣ d2β

ds2

∣∣∣∣ = √

1 + τ2

k2 .

Tangent space and derivatives of maps.

1. There are misprints in this question. Let us assume that the cylinder is x2 + y2 = 1 and that the tangent vector in T(1,0,0) is (0, 1, 1). In the given coordinates (a coordinate map is just the inverse of a parameterization), dϕn has matrix

( ∂nθ ∂θ

∂nθ ∂z

∂z ∂θ

∂z ∂z

) =

( n 0 0 1

) .

Now (1, 0, 0) corresponds to (θ, z) = (0, 0). We see that ϕn(1, 0, 0) = (1, 0, 0). The basis for the tangent space associated to the coordinates is

(x, y, z) ∂θ

= (sin θ, cos θ, 0), ∂(x, y, z) ∂z

= (0, 0, 1).

At (θ, z) = (0, 0) this basis is (0, 1, 0), (0, 0, 1).

Then at (1, 0, 0),

dϕn(0, 1, 0) = (0, n, 0), dϕ(0, 0, 1) = (0, 0, 1).

Hence dϕn(0, 1, 1) = (0, n.1).

General second fundamental form questions.

SOLUTIONS TO FINAL EXAM QUESTIONS FROM PREVIOUS YEARS. 3

1. Following Do Carmo, for the graph z = f(x, y), the matrix of the second fundamental form with respect to the basis

(1, 0, fx), (0, 1, fy),

is ( fxx fxy fxy fyy

) .

In our case, the basis at (0, 0, 0) is

(1, 0, 0), (0, 1, 0)

and the matrix of the second fundamental form is (

e f f g

) =

( 2A B B 2C

) .

The matrix of the derivative of the Gauss map is

(

E F F G

)1 ( e f f g

) =

( 1 + f2x fxfy fxfy 1 + f2y

)1 ( fxx fxy fxy fyy

) ,

which gives

K = fxxfyy − f2xy 1 + f2x + f2y

, H = (1 + f2y )fxx + (1 + f

2 x)fyy − 2fxfyfxy

2(1 + f2x + f2y ) .

In our case, K = 4AC −B2, H = 2A + 2C.

The unit vector in direction (x, y, 0) is (cos θ, sin θ, 0) = cos θ(1, 0, 0) + sin θ(0, 1, 0). The normal curvature in this direction is

(cos θ, sin θ) (

2A B B 2C

)( cos θ sin θ

) = 2A cos2 θ + 2B cos θ sin θ + 2C sin2 θ.

2. elliptic points = {q : K(q) (0,∞)}.

4 SOLUTIONS TO FINAL EXAM QUESTIONS FROM PREVIOUS YEARS.

Since K is a continuous (indeed a smooth) function on S and (0,∞) is open, the set of elliptic points of S is open, and is hence an open neighborhood of each of its points. The same is true for the hyperbolic points, but not for parabolic or planar points. For example, the graph

z = x4 + y2

is parabolic at (0, 0, 0) and elliptic elsewhere, while the graph

z = x4 + y4

is planar at (0, 0, 0) and elliptic elsewhere.

Gauss curvature questions.

1. (a).

yu = xu + cNu ⇒ 〈yu,N= xu,N+ c〈Nu,N= Nu,N= 12

∂u 〈N,N= 0.

Similarly, N is perpendicular to yv, and hence is normal to the surface parameterized by y. (b). Since xu and xv are lines of curvature, we have

Nu = µ1xu, Nv = µ2xv

for smooth scalar functions µ1, µ1 (we don’t specify which is bigger.)

yu × yv = (xu + cNu)× (xv + cNv) = xu × xv + cxu ×Nv + cNu × xv + c2Nu ×Nv = xu × xv − c(µ1 + µ2)xu × xv + c2µ1µ2xu × xv = (12cH + c2K)xu + xv.

(c). We think of N as being a function of (u, v). Since x and y have the same normal, N gives the Gauss map for either surface. Consider (u0, v0) and set x(u0, v0) = x0 and y(u0, v0) = y0. Consider a small disc R around (u0, v0). Then

curvature of x at x0 = lim diam R→0

areaN(R) areax(R)

,

curvature of y at y0 = lim diam R→0

areaN(R) areay(R)

.

Hence

curvature of y curvature of x

= lim diam R→0

areax(R) areay(R)

= lim diam R→0

R |xu × xv| dudv

R |yu × yv| dudv =

1 12cH + c2K .

SOLUTIONS TO FINAL EXAM QUESTIONS FROM PREVIOUS YEARS. 5

2. K = 1: the sphere x2 + y2 + z2 = 1. K = 0: the plane z = 0. K = 1: the surface of revolution obtained by revolving (e−s, 0,

s 0

1− e−2s ds), for s > 0, about the z-axis.

We derived the formula for the curvature surface obtained by revolving the curve (φ(s), 0, ψ(s)) around the z-axis in the case when s is the arclength - that is φ′2 + ψ′2 = 1. The Gauss curvature is

−φ ′′

φ .

In particular it is zero precisely when φ′′ = 0. This implies that

φ(s) = as + b, ψ(s) = √

1− a2s + c.

This means that the curve is a straight line and the surface of revolution is a cone or cylinder.

3. xs = α′(s) + ub(s) = t(s) + n(s), xu = b(s),

so

E = 1 + u2τ2

F = 0 G = 1

Furthermore

xs × xu = (t + n)× b = n + t, N = n + t1 + u2τ2

.

g = xuu,N= 0,N= 0. f = xuv,N= 〈τn, −n + t

1 + u2τ2 = −τ√

1 + u2τ2 .

K = eg − f2

EG− F 2 = − τ2

(1 + u2τ2)2 .

Isometric surfaces.

1. Done in class.

2. Consider the parameterizations x(u, v) and

x̄(u, v) = r(u, v + d) = (f(v + d) cos u, f(v + d) sin u, g(v + d)).

6 SOLUTIONS TO FINAL EXAM QUESTIONS FROM PREVIOUS YEARS.

Then r̄ r1

is the map (u, v) (u, v + d). This is an isometry if and only if E = Ē, F = , G = Ḡ. But

( E F F G

) =

( f2(v) 0

0 (f ′(v))2 + (g′(v))2

) ,

( Ē F̄ F̄ Ḡ

) =

( f2(v + d) 0

0 (f ′(v + d))2 + (g′(v + d))2

) .

Equating E and we get f(v + d) = ±f(v) for all d. Assuming f is not identically zero, this gives that f is constant and the surface is a cylinder.

comments (0)
no comments were posted
be the one to write the first!
This is only a preview
3 shown on 6 pages
Download the document