0% found this document useful (0 votes)
84 views

Math 521 Lecture #33 4.3.2: Solved Examples 4.3.3: First Integrals

This document summarizes key concepts from a math lecture on the Calculus of Variations, including: 1) The Euler equation is a necessary condition for an extremum of a functional that is satisfied by the solutions of a differential equation. 2) Two examples are provided to demonstrate finding the extremals of functionals by solving the resulting Euler equations. 3) When the Lagrangian does not explicitly depend on one of the variables, the form of the Euler equation is simplified, such as becoming an algebraic equation or involving a "first integral".

Uploaded by

Anjali Priya
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
84 views

Math 521 Lecture #33 4.3.2: Solved Examples 4.3.3: First Integrals

This document summarizes key concepts from a math lecture on the Calculus of Variations, including: 1) The Euler equation is a necessary condition for an extremum of a functional that is satisfied by the solutions of a differential equation. 2) Two examples are provided to demonstrate finding the extremals of functionals by solving the resulting Euler equations. 3) When the Lagrangian does not explicitly depend on one of the variables, the form of the Euler equation is simplified, such as becoming an algebraic equation or involving a "first integral".

Uploaded by

Anjali Priya
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 4

Math 521 Lecture #33

§4.3.2: Solved Examples; 4.3.3: First Integrals


Recall that the Euler equation Ly −(d/dx)Ly0 = 0 is a necessary condition that is satisfied
by a minimizer (or extremum) y0 of a functional
Z b
J(y) = L(x, y, y 0 ) dt, y ∈ C 2 [a, b], y(a) = ya , y(b) = yb .
a

Showing that a solution of the Euler equation is a minimizer requires analysis beyond
the scope of this course.
However, the search for a minimizer is reduced through the Euler equation to solutions
of a differential equation.
Example 4.15. Find the extremals of the functional
Z 1
(y 0 )2 + 3y + 2x dx, y(0) = 0, y(1) = 1.

J(y) =
0

We tacitly assume that y ∈ C 2 [0, 1] so that we can obtain the Euler equation.
With
L(x, y, y 0 ) = (y 0 )2 + 3y + 2x,
the Euler equation for J is
d
2y 0 = 0.

3+
dx
This is the second-order linear non homogeneous equation
3
y 00 = .
2
We solve this by integrating twice to obtain
3x2
y = C1 x + C2 + .
4
From the boundary conditions y(0) = 0 and y(1) = 1 we obtain the linear system of
equations,

C2 = 0
3
C1 + C2 = 1 − .
4
We have a unique solution for C1 and C2 , and hence a unique solution of Euler equation
subject to the boundary conditions, namely,
x 3x2
y0 = + .
4 4
Unfortunately, we do not know if this is a minimizer or maximizer of J(y), or if it is
neither, because all we know is that δJ(y0 , h) = 0 for all admissible variations h.
Example 4.16. Recall that we found the Euler equation for the arc length functional
Z 1 p
J(y) = 1 + [y 0 (x)]2 dx, y(0) = 0, y(1) = 1.
0

We illustrated the proof of the Theorem for the Euler equation and got
!
0
d y
p = 0.
dx 1 + [y 0 ]2

We can verify it this now using the Lagrangian


p
L(x, y, y 0 ) = 1 + [y 0 ]2 .

For then the Euler equation is


!
d y0
0+ p = 0.
dx 1 + [y 0 ]2

Solving this gives for an arbitrary constant C that


y0
p = C.
1 + [y 0 ]2

We can solve this for y 0 :


p
y 0 = C 1 + [y 0 ]2
[y 0 ]2 = C 2 (1 + [y 0 ]2 )
[y 0 ]2 − C 2 [y 0 ] = C2
C2
[y 0 ]2 = .
1 − C2
This implies that y 0 is a constant, so that

y(x) = Kx + M.

From the boundary conditions y(0) = 0 and y(1) = 0 we get that K = 1 and M = 0,
so that y0 (x) = x is the only solution of the Euler equation subject to the boundary
conditions.
On geometric grounds we can argue that y0 (x) = x is a global minimizer of the arc length
functional.
When the Lagrangian L does not explicitly depend on one of the three variables x, y,
and y 0 , we can make a simplification in the Euler equation.
When L does not depend on y 0 , then the Euler equation is not a differential equation,
but an algebraic equation,
Ly (x, y) = 0.
Generally (through the implicit function theorem), this implicitly (and sometimes explic-
itly) defines y as a function of x.
When L does not depend on y, then the Euler equation is
d
Ly0 (x, y 0 ) = 0, or Ly0 (x, y 0 ) = C
dx
for an arbitrary constant C (as it did for the arc length functional).
When L does not depend on x, then multiplying the Euler equation through by y 0 gives
 
0 d
0 = y Ly − Ly0
dx
d
= y 0 Ly − y 0 Ly0 + Ly0 y 00 − Ly0 y 00
dx
d
= Ly y 0 + Ly0 y 00 − y 0 Ly0 − y 00 Ly0
dx
dL d
= − y 0 Ly0 − y 00 Ly0
dx dx
d
= [L − y 0 Ly0 ] .
dx
This says that the quantity L − y 0 Ly0 is a conserved quantity or a first integral.
Formally, a first integral of a second-order ordinary differential equation F (x, y, y 0 , y 00 ) =
0 is a function g(x, y, y 0 ) which is constant along each solution of F (x, y, y 0 , y 00 ) = 0.
In the Calculus of Variations, when L is independent of x, a first integral or conservation
law of the Euler equations of the extremals is

g(y, y 0 ) = L(y, y 0 ) − y 0 Ly0 (y, y 0 ).

We can use a first integral to determine the extremals of a functional.


Example 4.17. The functional for the brachistochrone problem is
a
p
1 + [y 0 ]2
Z
J(y) = p dx, y(0) = b, y(1) = 0.
0 2g(b − y)

The Lagrangian for this functional is independent of x, so a first integral of the Euler
equation is p
0 1 + [y 0 ]2 (1 + [y 0 ]2 )−1/2
C = L − y Ly0 = p − (y 0 )2 p .
2g(b − y) 2g(b − y)

We can absorb the common number 2g into C.
p
Multiplying the resulting equation through by 1 + [y 0 ]2 gives
p 1 + (y 0 )2 (y 0 )2 1
C 1 + [y 0 ]2 = √ −√ =√ .
b−y b−y b−y
Squaring both sides gives
1
C 2 (1 + [y 0 ]2 ) = .
b−y
Simplifying this gives

1 1 − C 2 (b − y)
(y 0 )2 = − 1 = .
C 2 (b − y) C 2 (b − y)

From physical considerations (the bead is rolling downward), we know that dy/dx < 0.
Thus we have obtain the first order equation
s
dy 1 − C 2 (b − y)
=−
dx C 2 (b − y)

in addition to the second-order Euler equation that an extremal of the brachistochrone


function must satisfy.
If we write C1 = C −2 and separate variables, the first-order equation becomes

b−y
dx = − p dy.
C1 − (b − y)

Through the substitution

b − y = C1 sin2 (φ/2), −dy = C1 sin(φ/2) cos(φ/2)dφ,

we obtain

C 1 sin φ/2
dx = p C1 sin(φ/2) cos(φ/2) dφ
C1 (1 − sin2 φ/2)
= C1 sin2 (φ/2) dφ
C1 
= 1 − cos φ dφ.
2
Integration gives
C1 
φ − sin φ + C2 .
x=
2
The functions x and y of φ are parametric equations for a cycloid.

You might also like