Chapter 1

Relativistic dynamics

1.1 Principle of relativity

We are familiar with the idea that physical laws must be valid in all inertial reference frames;
this is called the principle of relativity. Mathematically, this is guaranteed in different ways
for different types of equations. A law written in terms of relativistic scalars
scalar = scalar, (1.1.1)
satisfies this condition, since scalars are the same in all reference frames. If the law is written
in terms of 4-vectors, then we have a covariant relationship
4-vector = 4-vector , (1.1.2)
which means that the left and right sides vary in the same way with changes of coordinate
system. This also satisfies the principle of relativity.
An example of the first is the definition of mass m in terms of the energy-momentum 4-vector
p = (p, E):
m2 = −p · p = E 2 − p2 . (1.1.3)
Both sides of this equation are scalars; the mass m by definition and the dot product of the
4-momentum with itself via the properties of the dot product. The relationship between the
wave 4-vector of a matter wave and the 4-momentum is an example of the second type:
p = ~k. (1.1.4)
The components in different reference frames are different, but the components on the two
sides change in similar ways since both sides of the equation are 4-vectors.

1.2 Relativistic form of Newton’s second law

The relativistic form of Newton’s second law is
dp
F = . (1.2.1)
dt
1
CHAPTER 1. RELATIVISTIC DYNAMICS 2

This is correct, but it is not expressed in covariant form because (1) it is a relationship
between space vectors only and (2) the dt is the timelike component of a displacement 4-
vector and not a scalar. This makes the conventional definition of force highly inconvenient
in the relativistic context.
If the object in question is moving at speed v, such that γ = (1 − v 2 )−1/2 (again, we set the
speed of light c = 1), then we can rectify the second problem by eliminating the increment
in time dt in favor of the increment in proper time dτ along the trajectory of the object,
dt = γdτ , so that Newton’s second law becomes
dp
M = γF = (1.2.2)
dτ
where M = γF is called the Minkowski force. Notice that the Minkowski force and dτ differ
respectively from the ordinary force F and the ordinary time increment dt when the object
is moving at relativistic speeds.
This completes the first step in making Newton’s second law covariant. The second step is
inferring the timelike component of this equation. This is fairly obvious, as p is the spacelike
component of the 4-momentum vector, the timelike component of which is the energy E:
dE
Mt = (1.2.3)
dτ
where Mt is the timelike component of the Minkowski 4-force. The above equation indicates
that Mt is just the proper time rate of change of energy, or something like the power. The
complete, covariant form of Newton’s second law is thus
dp
M= (1.2.4)
dτ
where M = (M , Mt ).
Given some force F , it is clear how to compute the spacelike part of the Minkowski force; just
set M = γF . However, equation (1.2.3) is not a particularly convenient way to compute the
timelike part Mt . An easier way comes from dotting equation (1.2.4) with the 4-momentum
p:
dp 1 d(p · p) 1 dm2
p·M =p· = =− =0 (1.2.5)
dτ 2 dτ 2 dτ
where we have used the fact that the length squared of p is minus the mass m squared. Since
the mass of the object is a constant scalar (no virtual masses here!), it doesn’t change with
time, resulting in
p · M = p · M − EMt = 0. (1.2.6)
Since p/E = v, we find that
Mt = v · M = γv · F . (1.2.7)
In words, Mt is just γ times the power, or work done per unit time (not proper time) by the
force F on the object.
CHAPTER 1. RELATIVISTIC DYNAMICS 3

1.3 The 4-velocity and 4-acceleration

If the position of an object is x, then the velocity is
dx
v= . (1.3.1)
dt
This expression suffers the same problems as the traditional form of Newton’s second law,
and we solve these problems in the same way. Eliminating dt in favor of dτ , we get
dx
u ≡ γv = (1.3.2)
dτ
where u = γv is the spacelike component of the 4-velocity u = (u, ut ) = (γv, γ). The
timelike component is just
dt
ut = = γ, (1.3.3)
dτ
so
dx
u= (1.3.4)
dτ
where the spacetime position vector is x = (x, t) as usual. Note that u is truly a 4-vector,
as its length is a scalar:
u · u = γ 2 v 2 − γ 2 = −1. (1.3.5)
It is also easy to show that the 4-velocity is related to the 4-momentum by

p = mu. (1.3.6)

The 4-acceleration is a 4-vector defined as follows:

du
α= . (1.3.7)
dτ
We know that it is a 4-vector, because the right side of the equation defining it is also a
4-vector. The 4-acceleration is normal to the 4-velocity. The Minkowski form of Newton’s
second law can be written
M = mα. (1.3.8)
In the rest frame, it may be shown that the 4-acceleration of some moving object is
du  2
= γ a + γ 4 (v · a)v, γ 4 (v · a)

α= (1.3.9)
dτ
where v and a = dv/dt are the actual velocity and acceleration of the object in the rest
frame. The form of the 4-acceleration depends very strongly as to whether the acceleration
is parallel or normal to the pre-existing velocity. If the acceleration is parallel to the velocity,
this reduces to
α = γ 4 a, γ 4 va


(1.3.10)
CHAPTER 1. RELATIVISTIC DYNAMICS 4

while for an acceleration normal to the velocity we have

α = γ 2 a, 0 .


(1.3.11)

An inertially co-moving reference frame relative to a moving and possibly accelerating object
is a frame moving at the velocity of the object at a specified instant in time, but not
accelerating. In an inertially co-moving frame, the 4-velocity of the object is zero and the
4-acceleration is
α = (a0 , 0) (1.3.12)
where a0 is the ordinary acceleration of the object in this frame. Note that a0 is distinct
from the above acceleration a in the rest frame. However, the Lorentz transformation can
be used to rewrite the components of α in the rest frame (or any other frame).

1.4 Assessment of different approaches

Newton’s second law is used to compute the motion of objects assuming that the force is
known. Both the traditional form of Newton’s second law, as given by equation (1.2.1), and
the Minkowski form, as given by equation (1.2.4), are useful under different circumstances.
The traditional form is valid for tracking the relativistic motion of an object from a fixed
rest frame in which the forces are known. For instance, the Lorentz force on a particle with
charge q
F = q(E + v × B) (1.4.1)
can be used to find the relativistic motion of such a particle relative to the fixed frame. How-
ever, transforming results to alternative frames is difficult, because the rules for transforming
the ordinary force to a new frame are complicated.
The Minkowski force approach is best if one is interested in computing the motion of an
object from the object’s own perspective, since integrating equation (1.2.4) yields a result
in terms of the proper time of the object τ rather then the time t of the fixed coordinate
system. The covariant form of equation (1.2.4) also allows easy changes of reference frame,
which is sometimes useful.

1.5 Electromagnetic force in covariant form

From our perspective it is easiest to write the electromagnetic force in covariant form using
the scalar and vector potentials combined as a 4-potential than to do it in terms of the
electric and magnetic fields. This is because the components of these fields actually form the
components of a 4-tensor, which carries us to a level of mathematics that is more complex
than we desire at this point.
We recall that the electric E and magnetic B fields are written in terms of the scalar φ and
vector A potentials as follows:
∂A
E = −∇φ − B = ∇ × A. (1.5.1)
∂t
CHAPTER 1. RELATIVISTIC DYNAMICS 5

1.5.1 Lorenz condition and the 4-gradient

Recall also that the scalar and vector potentials are not completely independent of each
other, but are related by the Lorenz condition:
∂φ
∇·A+ =0 (1.5.2)
∂t
where we again remind ourselves that we are setting c = 1. The scalar and vector potentials
are really the timelike and spacelike components of a 4-potential
A = (A, φ) (1.5.3)
and the Lorenz condition can be written in invariant form as
·A=0 (1.5.4)
where  
∂
= ∇, − (1.5.5)
∂t
is called the 4-gradient. (The minus sign on the time derivative is needed to make  a
4-vector. Test this by taking the 4-gradient of a plane wave ψ = exp[i(k · x − ωt)]; the result
should be a 4-vector.)

1.5.2 Electromagnetic Minkowski force

Substituting equations (1.5.1) into equation (1.4.1) and using equation (1.2.2) gives us the
spacelike part of the Minkowski force for electromagnetism:
 
∂A
M = γF = γq −∇φ − + v × (∇ × A) . (1.5.6)
∂t
From vector calculus we have
v × (∇ × A) = ∇(v · A) − v · ∇A (1.5.7)
where it is important to realize that the derivatives in the gradient operator apply only to
A; the velocity v is a constant under spatial differentiation.
In general A = A(x, t). We define a special kind of total time derivative of A which is the
derivative following the position of the charged particle on which the electromagnetic force is
acting. For the purposes of this derivative, we take x = xparticle (t), so that dx/dt = v where
v is the particle velocity. Thus, the total time derivative of the vector potential, obtained
using the chain rule, is
dA ∂A
= + v · ∇A. (1.5.8)
dt ∂t
Combining equations (1.5.6), (1.5.7), and (1.5.8) results in
dA
M = −γq + q∇(γv · A − γφ). (1.5.9)
dt
CHAPTER 1. RELATIVISTIC DYNAMICS 6

We recognize γv · A − γφ as a dot product between the 4-velocity u = (γv, γ) and the

4-potential A = (A, φ). Using also the fact that dt = γdτ , equation (1.5.9) simplifies to
dA
M = −q + q∇(u · A). (1.5.10)
dτ

Extending this to the Minkowski 4-force is a simple matter of writing down a covariant
expression for M that has a spacelike component given by equation (1.5.10).This is pretty
clearly
dA
M = −q + q(u · A). (1.5.11)
dτ
The final step is to note that the proper time derivative of 4-momentum in equation (1.2.4)
can be combined with the first term on the right in equation (1.5.11), resulting in the very
compact relativistic form of Newton’s second law for electromagnetism
dΠ d

≡ p + qA = q(u · A). (1.5.12)
dτ dτ
The quantity
Π = p + qA (1.5.13)
is the total 4-momentum discussed in Physics 222. Equation (1.5.12) shows that the total
4-momentum of a particle is conserved if u is normal (in a spacetime sense) to A.

1.6 Problems
1. Prove that the 4-velocity and the 4-acceleration of an object are normal or perpendic-
ular in the relativistic sense.

2. Four-acceleration:

(a) Prove equation (1.3.9).

(b) Prove equation (1.3.10).
(c) Prove equation (1.3.11).

3. An object of mass m is moving with velocity v = v î relative to a stationary frame at

some instant and its ordinary acceleration in a co-moving inertial frame at that instant
is a0 = a0 î where a0 is constant.

(a) Find its 4-velocity at that instant in the stationary frame.

(b) Find its 4-acceleration at that instant in the stationary frame. Hint: Make a
Lorentz transformation of the 4-acceleration from the co-moving inertial frame to
the stationary frame.
(c) Show that the acceleration in the stationary frame is a = a0 /γ 3 . Hint: Compare
the above results for α resolved in the rest frame with equation (1.3.10).
CHAPTER 1. RELATIVISTIC DYNAMICS 7

(d) Integrate dv/dt = a to get v(t), assuming that v(0) = 0. (You will probably have
to look up the integral.)

4. A spaceship of mass m is subject to an ordinary force F = F î where F is constant.

(a) If the spaceship starts from rest at time t = 0, compute its velocity as a function
of time, assuming that the force acts long enough for it to reach relativistic speeds.
(b) Compare the time dependence of the velocity in this case with that obtained in
the previous problem. For comparison purposes, define a = F/m.
(c) Recall that dτ /dt = 1/γ. Using the fact that γ can be written in terms of at,
integrate this equation to find τ = τ (t), thereby obtaining elapsed proper time on
the spaceship relative to rest frame time. (You will probably have to look up the
integral.) Obtain an approximate form of this equation for at  1. Comment on
relative aging rates of individuals on the spaceship relative to those at rest.

5. Imagine a particle with mass m and charge q in relativistic circular motion in the x-y
plane under the influence of a magnetic field B in the z direction.

(a) Relate the radial component of the Minkowski force to the radial component of the
4-acceleration using the covariant form of Newton’s second law. Then eliminate
the Minkowski force for the regular radial magnetic force, which has magnitude
qvB as usual, where v is the tangential component of the particle velocity.
(b) Eliminate the 4-acceleration in favor of the regular radial circulation, noting that
this is a case in which the acceleration is normal to the velocity.
(c) The radial acceleration still equals v 2 /r in relativity, where r is the radius of the
circle, since this is just geometry. Thus obtain the angular frequency of revolution
ω = v/r in the relativistic case, and compare it to the non-relativistic case.