Non-Linear Dynamics Homework Solutions

Week 6
Chris Small
March 6, 2007

Please email me at [email protected] with any questions or concerns reguarding these

solutions.

6.8.3 Locate annd find the index of the system given by

ẋ = y−x
ẏ = x2

First we find the x-nullcline to be y = x and the y-nullcline to be x = 0. Next plot the nullclines
and figure out what direction the field flows along each nullcline. Then we fill in the vector
field by remembering that the field inbetween any two nullclines points in a direction which is
inbetween the direction of the field at those nullclines. Now we use the hint and draw a closed
curve around the fixed point and arrive at Figure 1.
Now as we start at zero radians on the circle, and move counterclockwise around it, we see
that the arrows never make ar full revolution, implying that the index must be zero.
6.8.6 A closed orbit in the phase plane encircles S saddles, N nodes, F spirals and C centers, all of
the ususal type. Show that N + F + C = 1 + S.

Figure 1: Plot of Vector Field With Closed Curve

1
 Let IS be the index of a spiral, IN the index of a node, and so on. Closed orbits have an index
of 1, so by Theorem 6.8.1, from Strogatz,

1 = SIS + N IN + F IF + CIC
= −S + N + C + F.

Rearanging this equation we get the desired result.

6.8.9 Our claim is that the statement is false. We show this by providing the following counter-
example

ṙ = r(r − Ro )(r − Ri )
θ̇ = R − (Ro − Ri )/2

This system has the desired properties. It is smooth, has an outer stable orbit of Ro and an
inner one at Ri , and halfway between the sign of θ̇ switches, so that one of the stable orbits
flows one way and the other one flows the other. However, the only fixed point in this system
is at the origin.
6.8.12 Matter and Anitmatter Problem We explore the analogy between particle anti-particle collisions
and bifurcations of fixed points by studying a two dimensional version of the saddle node
bifurcation given by the system

ẋ = a + x2
ẏ = −y

a) Find and classify all fixed points of the system.

√
First we find x-nullclines at x = ± −a and y-nullclines at y = 0, Thus for a = 0 we have
precisely one fixed point and for a < 0 we have precisely two, indicating a saddle-node
(blue-sky) bifurcation. When we linearize (first finding the Jacobian and then evaluating
at the fixed points) we get the linearization matrices
 √ 
±2 −a 0
A=
0 −1
with the plus√sign for the positive
√ square root and the minus for the negative. These
have D√= ∓2 −a and T = ±2 −a − 1 with the plus and minus signs assigned as before.
For (+ −a, 0), we therefore always
√ have a negative determinant, implying that this fixed
point is always a saddle. For (− −a, 0), the determinant is always positive, implying that
the point is always either a spiral or a node. Furthermore, the trace is always negative,
so whatever it is, it’s always stable. To figure when it is what, we compute
√
T 2 − 4D = −4a + 4 −a + 1

and note that since our fixed point only exists for negative values of a, this is always
positive, implying that D < T 2 /4 and that consequently the point is always an unstable
node.

2
 Figure 2: From problem 6.8.12

When a = 0, linear analysis fails, since it predicts a whole line of nonisolated fixed points,
which we have already shown does not exist. For this case, we look at Figure 2, which
shows the vector field for a = 0.
This is not any of the kinds of fixed points that we have talked about, so if I may, I would
like to take the liberty to call it a Wal-Mart fixed point, since there are infinitely many
ways in but only one way out.
b) Show that the sum of indices is conserved as a is varied.
One of the two distinct fixed points is a saddle, and consequently always has an index of
-1, whereas the other is always a node and so always has an index of 1. When a = 0,
the two fixed points have collided, giving us the vector field of Figure 2, which we can
readily verify has an index of zero, by keeping track of the direction that vectors point
in as we move around a circle centered at the origin; doing this, the vectors never make
a full revolution, so the index must be zero. For a > 0, there are no fixed points, and
consequently there is a net index of zero. Thus we see that the sum of all indices is
conserved, and always equal to zero.
c) State and prove a generalization of the result for systems of the form ẋ = f (x, a), where
x ∈ R2 and a ∈ R.
Claim: The sum of all indices involved in any bifurcation for a system of the form above,
if it is a smooth vector field, is conserved through the bifurcation.

Proof. Consider a closed curve which is made to continuously deform as the parameter a
is varied in such a manner that the curve encloses only the fixed points involved in the
bifurcation. Then if one varies a continuously, the index will vary continuously, and since
it takes on only integer values, it must stay constant as a function of a. Thus, we have
completed the proof.

7.1.2 Sketch the phase portrait (in polar coordinates) of the following system

ṙ = r(1 − r2 )(9 − r2 )
θ̇ = 1

See Figure 3 for the phase portrait.

3
 Figure 3: Phase Portrait for Problem 7.1.2

7.2.7 Consider the system

ẋ = y + 2xy
ẏ = x + x2 − y 2

a) Show that ∂f /∂y = ∂g/∂x.

We assume that f and g are defined as in 7.2.5. Given this, we find that

∂f /∂y = 1 + 2x = 1 + 2x = ∂g/∂x.

b) Find V .
Recall that V is the function which the system is the gradient of. Thus
∂V
= y + 2xy
∂x
∂V
= x + x2 − y 2 .
∂y

From these equations we obtain two integral expressions for V by seperation of variables.
This leads to the equalities

V = yx + x2 y + c(y)
V = xy + x2 y − y 3 /3 + d(x)

We can see that the two equations agree iff we set c(y) = −y 3 /3 and d(x) = 0. Then

V = xy + x2 y − y 3 /3.

c) Sketch the phase portrait.

4
 Figure 4: Nullclines, Vector Field and Phase Portrait

7.2.10 By constructing a Liapunov function of the form V = ax2 + by 2 , for suitable a and b, show
that there exist no closed orbits in the following system

ẋ = y − x3
ẏ = −x − y 3

First notice that there exists only one fixed point. This follows from the fact that our nullclines
are y = x3 and x = −y 3 . We could resort to a graphical argument to see that the only
intersection is at the origin, or we could plugg the first equation into the latter to get the
equality x = −x9 . So x = 0 is clearly a solution, but if x 6= 0 then we can divide through so
that x8 = −1, which has no real solutions since no real number raised to the power of an even
number is negative. Thus, as claimed, the origin is the only fixed point.
Let V = ax2 + by 2 . Now so long as a, b > 0, V is positive definite (see Strogatz, pg. 201).
To see what further restrictions must be placed on V so that V̇ < 0 for all ~x 6= ~x∗ (the other
condition on pg. 201 that must be satisfied), we find that

V̇ = 2axẋ + 2by ẏ
= 2axy − 2bxy − 2ax4 − 2by 4 .

So if we let a = b we get V = −2a(x4 + y 4 ) which for any positive a is always negative, so

long as we are not evaluating at the origin. Thus, V = x2 + y 2 is a Liapunov function for this
system. Consequently, there exist no closed orbits in this system.
7.3.1 Consider the system

ẋ = x − y − x(x2 + 5y 2 )
ẏ = x + y − y(x2 + y 2 ).

5
a) Classify the fixed point at the origin.
We first compute the Linearization matrix using the Jacobian
   
1 − 3x2 − 5y 2 −1 − 10yx 1 −1
J|O = =
1 − 2yx 1 − x2 − 3y 2 O
1 1

which has T = 2 and D = 2, implying that the origin is an unstable spiral since T 2 −4D =
−4 < 0.
b) Rewrite the system in polar coordinates using rṙ = xẋ + y ẏ and θ̇ = (xẏ − y ẋ)/r2 .
From the first equation we get

rṙ = x2 − xy − x2 (x2 + 5y 2 ) + yx + y 2 − y 2 (x2 + y 2 )

= x2 + y 2 − x4 − y 4 − 6x2 y 2
= r2 − ((x2 + y 2 )2 − 2x2 y 2 ) − 6x2 y 2
= r2 − r4 − 4r4 cos2 θ sin2 θ

Note that in this simplification we have used the facts that x2 + y 2 = r2 , x = r cos θ and
y = r sin θ. Dividing this final equation through by r2 we find that

ṙ = r − r3 (1 + 4 cos2 θ sin2 θ).

We also find in a similar manner that

r2 θ̇ = (x2 + yx − yx(x2 + y 2 )) − (yx − y 2 − yx(x2 + 5y 2 ))

= r2 + 4xy 3
θ̇ = 1 + 4r2 cos θ sin3 θ

c) Determine the circle of maximum radius, r1 , centered on the origin such that all trajec-
tories have a radially outward component on it.
This task is equivalent to finding the maximum radius for which ṙ > 0 for all θ. If we
look at our equation for ṙ, we can see that the bigger the periodic term (1 + 4 cos2 θ sin2 θ)
is, the harder it will be to make ṙ > 0, so we want to find the maximum of the term
inside the parenthesis, since we’ll our r1 to make ṙ > 0 for that maximum, but will also
be guarenteed that ṙ > 0 for all other θ values, since they would only make the periodic
term less, and hence, make ṙ even larger. To find this maximal value, we could resort to
analytic techniques, ie. find the derivitive of the periodic term with respect to θ, set that
equal to zero and then figure out which of the resulting critical points give us maxima and
which give us minima. There may even be an easier analytic method involving geometric
of trigonometric aproaches, but a graphic argument is really easy and quite convincing.
Taking a look at Figure 4, we can se that the maximum value of this term is 2. We
can therefore set that term equal to 2 in our equation and see what values of r makes ṙ
positive. This brings us to solving the inequality

0 < ṙ = r − 2r3
= r(1 − 2r2 )

6
 2

1.8

1.6

1.4

1.2

1 2 3 4 5 6

Figure 5: Plot of the periodic term from 7.3.1 vs. θ

0.2

0.1

0.2 0.4 0.6 0.8 1

-0.1

-0.2

-0.3

Figure 6: Plot of ṙ versus r

p
Since this expression for ṙ is a downwardp facing cubic with x intercepts at r = ± 1/2
and r = 0 we can deduce that for all r < 1/2 but greater than zero, ṙ > 0. See Figure
5 for p
graphical clearification on this arguement. Thus we get the desired value by setting
r1 = 1/2 − ǫ for any positive but small as we like ǫ.
d) Determine the circle of minimum radius, r2 , centered on the origin such that all trajec-
tories have a radially inward component on it.
We proceed here as we did in part (c), only here we want the smallest r for which ṙ is
negative for all θ. Consequently, we want to find the minimum of (1 + 4 cos2 θ sin2 θ)
instead of the maximum. By applying analytic methods such as those described in part
(c), or looking at Figure 4 again, we can see that the minimum of the term is 1. We use
this in the equation for ṙ as we did above, and find that if ṙ = r(1 − r2 ) then we will have
a negative value of ṙ so long as r > 1. Thus we may set r2 = 1 + ǫ for some small but
positive ǫ.
e) Prove that the system has a limit cycle somewhere in the trapping region r1 < r < r2 .

Proof. By our construction we have a trapping region that satisfies the conditions of the
Poincaré-Bendixson Theorem. Consequently, there exists a limit cycle in the region in
question.

7.3.7 Consider the system

ẋ = y + ax(1 − 2b − r2 )
ẏ = −x + ay(1 − r2 )

7
where a and b are parameters (0 < a ≤ 1, 0 ≤ b < 1/2) and r2 = x2 + y 2 .

a) Rewrite this system in polar coordinates.

We proceed as we did in problem 7.3.1.

rṙ = xẋ + y ẏ
= (ax2 + ay 2 )(1 − r) − 2bax2
ṙ = ar(1 − r2 − 2b cos2 θ)

r2 θ̇ = −x2 + ayx(1 − r2 ) − y 2 − ayx(1 − 2b − r2 )

θ̇ = −1 + 2ba cos θ sin θ

b) Prove that there is at least one limit cycle, and that if there are several, they all have the
same period T (a, b).
As in problem 7.3.1, we find r1 and r2 with the properties we need to have a trapping
region. We want to maximaize the periodic term 2b cos2 θ, which happens when θ = kπ
for some integer k. At these points the term in question equals 2b, so we substitute this
2 2
into our
√ equation for ṙ and find that ṙ = (1 − r − 2b) = ar((1 − 2b) − r ), so if we set
r1 = 1 − 2b − ǫ, for some small but positive ǫ we get what we need.
Since the minimum of the periodic term is 0, we can use this to find a suitable r2 . We
get ṙ = ar(1 − r2 ) so if we let r2 = 1 + ǫ for some small but positive ǫ, then we have our
outer trapping region boundary.
By the Poincaré-Bendixson Theorem, and the existence of a trapping region for this
system, the system must have at least one limit cycle in the region inbetween r1 and r2 .
Reguarding the period of such trajectories, note that θ̇ depends only on θ and not on r.
Consequently, for any two initial conditions with the same initial θ value, the trajectories
will make a full revolution in the same amount of time, since their angular positions are
governed by the same dynamical equations.
c) Prove that for b = 0 there is only one limit cycle.

Proof. When b = 0 our inner trapping radius will be r1 = 1 − ǫ and our outer will be
r2 = 1 + ǫ. For two limit cycles to exist, there must be a radial distance dr between them
for any given value of θ, but we can make the values of ǫ as low as we like, so low that
2ǫ < dr so that there is no way both could fall inside the trapping region.