MECH1 Problem Sheet 1 and Solutions Due Friday 13 February

Kinematics
Do all problems
1. It was shown in class that if e1,2,3 and e01,2,3 are two orthonormal bases, then for a vector a which has
coordinates (a1 , a2 , a3 ) in the “old” basis, i.e., a = a1 e1 + a2 e2 + a3 e3 , its coordinates (a01 , a02 , a03 ) in the
“new” basis are  0   
a1 a1
 a02  = R  a2  ,
a03 a3
for a square matrix R, where the components rij of R are rij = e0i · ej .

(a) Explain why rij = cos(e[

0 , e ), where e[
i j
0 , e is the angle between the two unit vectors involved.
i j
Answer: Because e’s are unit vectors, and by definition of the dot product e0i · ej = |e0i ||ej | cos(e[
0 , e ).
i j
(b) Explain why RRT = RT R = I (such a matrix R is called orthogonal) where I denotes the 3 × 3 identity
matrix.
Answer: Because transposing R swaps the old and the new bases: rijT
= cos(e[ 0
i , ej ), so for any a one
T T
has RR a = R Ra = a.
(c) Explain why det R = ±1.
Answer: Because det R = det RT and determinant of the product is the product of determinants. So
(det R)2 = det I = 1.
(d) If R is an orthogonal matrix, what is the minimum number of its entries it suffices to know to determine
the whole matrix? Extend your answer to the case of n × n orthogonal matrices.
Answer: If rj are vectors – columns of R, then the condition RT R = I, by the matrix multiplication
rule, means ri · rj = 1 if i = j and 0 otherwise. This means n2 quantities rij satisfy n + n(n−1)
2 = n(n+1)
2
n(n+1) n(n−1)
equations. So one can recover them all by knowing some n2 − 2 = 2 . If n = 3, this equals 3.
(e) Consider a special case when e3 = e03 and both bases are right-hand. Show that in this case, for
α = e\
0 , e , one has
1 1  
cos α sin α 0
R =  − sin α cos α 0 
0 0 1
Answer: Zeroes and 1 in the third row/column is because the bases are orthonormal, so the dot product
of e3 with any other e, but itself is zero. Since the orientation is the same, the pair (e01 , e02 ) arises from
(e1 , e2 ) after turning it by α degrees around the e3 axis.

2. A point is moving along a straight line as x(t) = −t cos t + sin t, t ≥ 0. Identify all turning points. Find the
velocity and acceleration at the turning points and the average velocity between every two consecutive ones.
Plot the world-line x(t).
Answer: v(t) = ẋ(t) = t sin t. It is zero for t = πk, k = 0, 1, 2, . . . . Substitute these values of t into x(t) –
this yields the turning points xk = (−1)k+1 πk. At each turning point the particle changes its direction of
travel. The velocity at the turning points is zero, while the acceleration ẍ(t) = t cos t + sin t equals there
(−1)k πk. The time interval between each two consecutive turning points is π, while |xk+1 − xk | = π(2k + 1).
So, the average speed between xk and xk+1 equals 2k + 1, and it is positive for even k and negative for odd
k, i.e v̄[xk ,xk+1 ] = (−1)k (2k + 1).
3. A point is moving in a plane, so that its coordinates

x(t) = A cos ωt, y(t) = B sin ωt.

Describe the shape of the trajectory in the xy-plane as well as the world-line.

1
 Answer: x2 /A2 + y 2 /B 2 = 1 for the trajectory. This is an ellipse with semi-axes A, B. The particle moves
counterclockwise around the ellipse. If the t axis goes vertically, the world-line is a right helix, whose base
is the ellipse and over the period of revolution the helix goes up by 2π.
Find the absolute value and direction of the particle’s acceleration.
Answer: Differentiating twice yields ax = −ω 2 x, ay = −ω 2 y. Thus a = −ω 2 r, so the acceleration is always
directed towards the centre.
π
Find the curvature at at times t1 = 2ω , t2 = 2009π
ω . (Use the formula for the acceleration and fact that there
is no tangential acceleration at these points – why?)
Answer: As a = −ω 2 r, at these points the tangential component of the acceleration is zero. Curvature at
these points follows from the acceleration formula
dv
a= τ + Kv 2 n,
dt
where τ and n are the vectors tangent and normal to the trajectory, v is the speed. But a is directed
collinear with r, so at the times t1 , t2 , when the particle is at (0, B) and (−A, 0), respectively, the first term
in the above formula is zero. Hence at these points (but not in general)
1
K= |a|
v2
Using v = Aω or Bω, respectively (by differentiating x(t) and y(t) and substituting the vales t = t1,2 ) we have
|a| = ω 2 B or ω 2 A, respectively. It follows that the curvatures at times t1 , t2 are K1 = B/A2 , K2 = A/B 2 .
4. A particle moves counterclockwise in the plane along the ellipse x2 /A2 + y 2 /B 2 = 1, starting at rest at the
point (A, 0), so that its speed increases at a constant rate α. Find the particle acceleration vector every
time it returns to (A, 0), given that the circumference of the ellipse equals L (the general formula for the
circumference of the ellipse involves so-called elliptic functions which look much worse than, say the formula
for the area, which is just πAB).
p
Answer: The particle returns at times 2L/α k, k = 0, 1, 2, . . . by the motion with constant acceleration
formula. Indeed, for the first return the particle has to travel the distance L,
p with constant acceleration α.
The curvature is A/B 2 (the previous problem), the speed upon return is αk 2L/α. So, by the acceleration
formula,
A
a = −2Lk 2 α 2 i + αj.
B
5. A projectile is shot from the origin with the initial velocity v at the angle α to the horizon and moves in the
xy-plane. It is to be destroyed by a laser beam also emanating from the origin at the angle β < α with the
horizon. Find the coordinates of the point where the projectile will be shot down by the laser beam, just in
case assume v ¿ c.
Answer: The equations of motion of the projectile, which is motion with constant acceleration −gj, are
x
x(t) = v cos αt, y(t) = v sin αt − gt2 /2. Eliminating t = v cos α from the second equation yields the path
2
y(x) = x tan α − 2v2gxcos2 α . The cannonball will be shot down at the point where this path (a parabola)
intersects the straight line y = x tan β – the laser beam. So

gx2 2v 2 cos2 α
x tan α − = x tan β → x = 0 or x = (tan α − tan β).
2v 2cos2 α g
Clearly, we choose the latter value of x for the point of the projectile’s demise, and to get the y-coordinate
of this point just multiply that x by tan β. Note: to get the time it took the projectile to reach its end, it
would suffice to divide the latter value of x by v cos α, the constant horizontal speed of the projectile.
6. A ball that was lying on the edge of the table of height h is set in free fall motion with a horizontal speed
v. The ball hits the floor and starts jumping. Rebounds are instantaneous, and every time the ball hits

2
the floor its horizontal speed does not change, while the ratio of the vertical speed after and before every
rebound is some constant 0 ≤ α < 1. Show that at the horizontal distance
s
1 + α 2h
x=v
1−α g

from the edge of the table jumping will cease.

q
Answer: This is motion with constant acceleration g. The time before the first collision is t0 = 2hg . The
vertical speed immediately after the first rebound is v1 = αgt0 The time t1 it takes to reach the maximum
height is v1 /g = αt0 . Then, by induction, if ti is the time between the ith collision and the following
maximum, we have ti = αi t0 . Let T is the time it takes to stop jumping.
µ ¶ s
2 α 1 + α 2h
T = t0 + 2t1 + 2t2 + . . . = t0 (1 + 2(α + α + . . .) = t0 1 + 2 = .
1−α 1−α g

Multiplying by the constant horizontal velocity v gives the horizontal distance, as desired.