Know your Surfaces

Geometry Level 4

x 2 + 4 x y + 2 x z + z 2 = 5 x^2+4xy+2xz+z^2=5

What kind of a surface is this?

Hyperboloid of two Sheets Cone Elliptic Cylinder Hyperboloid of one Sheet Hyperbolic Cylinder Ellipsoid

This section requires Javascript.
You are seeing this because something didn't load right. We suggest you, (a) try refreshing the page, (b) enabling javascript if it is disabled on your browser and, finally, (c) loading the non-javascript version of this page . We're sorry about the hassle.

2 solutions

Otto Bretscher
Mar 5, 2016

Completing the squares gives ( x + 2 y + z ) 2 + z 2 4 ( y + z 2 ) 2 = 5 (x+2y+z)^2+z^2-4(y+\frac{z}{2})^2=5 . Since we have two positive and a negative coefficient, this is a hyperboloid of one sheet .

Background: Sylvester's Law of Inertia guarantees that any diagonal matrix congruent to the symmetric matrix A A of the quadratic form has the same number of positive, negative and zero diagonal entries... we need not go through the trouble of diagonalizing A A orthogonally (via eigenvalues).

So what you're saying is that we just need to find the coefficients of A , B , C A,B,C that satisfy the quadratic form

A ( f ( x , y , z ) ) 2 + B ( g ( x , y , z ) ) 2 + C ( h ( x , y , z ) ) 2 = D A( f(x,y,z))^2 + B (g(x,y,z))^2 + C (h(x,y,z))^2 = D

for some D > 0 D> 0 , where f , g , h f,g,h are linear combinations of x , y , z x,y,z . Then use the table I constructed below, but we have "signs of the 3 coefficients of A , B , C A,B,C " instead of "signs of the 3 eigenvalues"?

Pi Han Goh - 5 years, 3 months ago

Log in to reply

Exactly! It's all based on Sylvester's "Law of Inertia"... see my previous post.

To put it differently: We don't need to express the form with respect to an orthonormal basis ... any old basis will do.

Otto Bretscher - 5 years, 3 months ago

Log in to reply

This is brand new information for me! Thank you, Comrade Otto!

Pi Han Goh - 5 years, 3 months ago

Log in to reply

@Pi Han Goh My pleasure, Comrade Pi.

Otto Bretscher - 5 years, 3 months ago
Pi Han Goh
Mar 5, 2016

We want to classify the signature of a quadratic form.

We have an equation of the form x T A x = c x^T Ax = c , where A A is a symmetric real matrix and c > 0 c> 0 . The classification will depend on the signs of the real eigenvalues of A A .

Type of surface Signs of the 3 eigenvalues
Ellipsoid 3 positive eigenvalues
Hyperboloid of 1 sheet 2 positive and 1 negative eigenvalues
Hyperboloid of 2 sheets 1 positive and 2 negative eigenvalues
Empty locus 3 negative eigenvalues
Elliptic (circular) cylinder 2 positive eigenvalues, 1 zero eigenvalue
Hyperbolic cylinder 1 positive eigenvalue, 1 zero eigenvalue, 1 negative eigenvalue
Two parallel planes 1 positive eigenvalue, 2 zero eigenvalues

In this case, the matrix is A = [ 1 2 2 2 0 0 1 0 1 ] A = \begin{bmatrix}{1} && {2} && {2} \\ {2} && {0} && {0} \\ {1} && {0} && {1}\end{bmatrix} . By Cayley-Hamilton theorem,

det ( A λ I 3 ) = 0 1 λ 2 2 2 λ 0 1 0 1 λ = 0 λ 3 2 λ 2 4 λ + 4 = 0 \det(A - \lambda I_3) = 0 \Rightarrow \begin{vmatrix}{1-\lambda} && {2} && {2} \\ {2} && {-\lambda} && {0} \\ {1} && {0} && {1-\lambda}\end{vmatrix} = 0\Rightarrow \lambda^3 - 2\lambda^2 - 4\lambda+ 4 = 0

Let f ( λ ) = λ 3 2 λ 2 4 λ + 4 f(\lambda) = \lambda^3 - 2\lambda^2 - 4\lambda+ 4 . By trial and error, f ( 2 ) < 0 < f ( 1 ) , f ( 0 ) > 0 > f ( 1 ) , f ( 2 ) < 0 < f ( 3 ) . f(-2) < 0 < f(-1), \quad f(0) > 0 > f(1), \quad f(2) < 0 < f(3) \; .

Because f ( λ ) = 0 f(\lambda) = 0 is cubic equation, then by bisection method , there exists exactly one root in each of the following intervals: ( 2 , 1 ) , ( 0 , 1 ) , ( 2 , 3 ) (-2,-1), (0,1), (2,3) . In other words, there are 3 eigenvalues: 1 of which is negative, 2 of which is positive. Thus the surface in question is a Hyperboloid of 1 sheet \boxed{\text{Hyperboloid of 1 sheet}} .

This is a great systematic solution (+1), but it really isn't necessary to do so much work. If the goal is merely to find the type of the surface, we can just complete the squares (see my solution).

Otto Bretscher - 5 years, 3 months ago

Log in to reply

Where's your solution?

Pi Han Goh - 5 years, 3 months ago

Log in to reply

Coming up! Give me 3 minutes...

Otto Bretscher - 5 years, 3 months ago

0 pending reports

×

Problem Loading...

Note Loading...

Set Loading...