HW 6, Question 1: Forces in a Rotating Frame

Consider a solid platform, like a “merry-go round”, that rotates with respect to an inertial
frame of reference. It is oriented in the xy plane such that the axis of rotation is in the
z-direction, i.e. ω = ωẑ, which we can take to point out of the page. Now, let us also
consider a person walking on the rotating platform, such that this person is walking radially
outward on the platform with respect to their (non-inertial) reference frame with velocity
v = vr̂.

In what follows for Question 1, directions in the non-inertial reference frame (i.e. the frame
of the walking person) are defined in the following manner: the x0 -direction points directly
outward from the center of the merry-go round; the y 0 points in the tangential direction of
rotation in the plane of the platform; and the z 0 -direction points out of the page, and in the
same direction as the z-direction for the inertial reference frame.

(a) Identify all forces.

Since the person’s reference frame is non-inertial, we need to account for the various fictitious
forces given the description above, as well as any other external/reaction forces. The angular
rate of rotation is constant, and so the Euler force Feul = m(ω̇ × r) = ~0. The significant
forces acting on the person are:

• the force due to gravity, Fg = mg = mg(−ẑ 0 ),

• the (fictitious) centrifugal force, Fcen = mω × (ω × r) = mω 2 rx̂0 ,

• the (fictitious) Coriolis force, Fcor = 2mv × ω = 2mωv(−ŷ 0 ),

• the normal force the person experiences from the rotating platform N = N ẑ 0 , and

• a friction force acting on the person, f , in the x0 y 0 plane.

In order for the forces to balance, the friction force f must counteract all other forces acting on
the system. In other words, f has components in the x0 - and y 0 -directions: f = fx (−x̂0 )+ fy ŷ.
If we add up all components and set all sum-of-force components equal to zero, we get:

fx = mω 2 r, (1)
fy = 2mvω, (2)
N = mg. (3)

(b) Find the point of slippage.

The person with velocity v will not slip so long as the friction force balances the combination
of centrifugal and Coriolis forces, and the friction force does not exceed a value at which

1
the fictitious forces dominate. However, slippage will occur when f reaches its maximum
value: fmax = µs N , where µs is the coefficient of static friction. We can use this limit and
Equations 1-3 to find r as a function of all other known quantities:

f 2 = fx2 + fy2
= (mω 2 rslip )2 + (2mvω)2
1p 2 2
= µ2s (mg)2 =⇒ rslip = µs g − 4v 2 ω 2 (4)
ω2

(b) Find value for rslip .

We’re asked to consider a scenario where ω = 1 rad s−1 , v = 1 m s−1 , and µs = 1/2. We can
take g = 9.806 m s−2 , the local value of acceleration due to gravity. Plugging all of these
values into Equation 4, we find that

rslip ≈ 4.48 m (5)

2
 HW 6, Question 2: Coriolis Force on a Moving Train
Let’s consider a train in France at a latitude θ = 45 degrees (i.e. in the northern hemisphere)
that travels in the North direction at a constant speed v. As mentioned in the homework
prompt, the center of mass is located at a height z0 and the two rails are separated by a
distance w0 .

(a) Find ratio of forces exerted by the rails onto the wheels.

We ultimately want to examine the forces experienced by the rails that the train travels on.
In order to do so, we need to identify all forces acting on the rigid body:
• the force due to gravity, Fg = mg, acting at the center of mass of the train,
• the (fictitious) centrifugal force, Fcen = mω × (ω × r), acting at the center of mass of
the train,
• the (fictitious) Coriolis force, Fcor = 2mv × ω, since the train is moving,
• the reaction force that the eastern rail exerts on the eastern wheel, f1 , and
• the reaction force that the western rail exerts on the western wheel, f2 .
In the case of Earth’s rotation, the rotation rate |ω| = ω = 7.3 × 10−5 rad/s is very small, so
that terms that are proportional to ω 2 are negligible when compared to the Coriolis force,
since the latter term is proportional to ω. We therefore only need to consider the significant
forces acting on the system, which add to zero since the train keeps moving at a constant
speed v:

F ≈ mg + 2mv × ω + f1 + f2 = 0 (6)
We can take components of Equation 6 in the vertical and horizontal directions in the local,
non-inertial frame of reference. The direction of the Coriolis force can be determined since
the definition of the coordinate system dictates that ω points in the direction of the axis of
rotation, and that v points due North; the right hand rule requires that the vector v × ω
points due East. It has a magnitude of |Fcor | = Fcor = 2mvω sin θ. Since we neglect the
small contribution of Fcen , the local value of the gravitational acceleration g = 9.806 m s−2
and it points straight down towards the center of the Earth.

Figure ?? shows the various force vectors acting at their respective points in the rigid body
of the train. With these directions, we can write the general y-component and z-component
force equations,

ΣFz = f1,z + f2,z − mg = 0 (7)

3
One initial approximation that we can make is that that the eastern rail experiences a
much larger horizontal reaction force that opposes the Coriolis term than the western rail.
From Equation 8, we therefore require that the sum f1,x + f2,x ≈ f1,x ; by extension, this
approximately means that f2 points strictly in the vertical direction, since it has no significant
horizontal component and so f2 ≈ f2,y ẑ, while f1 has both horizontal and vertical components.
If we make the substitution |f2 | ≈ f2,y = f2 , we can re-write the component equations as

ΣFz ≈ f1,z + f2 − mg = 0 (9)

Equations 9 and 10 contain three unknowns, so we cannot uniquely solve for each compo-
nent using these two equations alone. However, we can proceed further with a few more
approximations using two slightly different methods:

Method 1: Equilibrium of Forces and Torques

Since the train remains on the track and moves at a constant speed, the sum of all forces
and of all torques must cancel to yield no net terms. We can add all moments of force and
require that their sum be zero; in order to do so, as with calculations of angular momentum,
we must select a reference point to measure moments from. In Figure ??, the reference point
is chosen (arbitrarily) to be at the western wheel (where f2 acts). The sum of all moments
relative to this point are:

ΣM = (0×f
2 ) + (rCOM × Fg ) + (rCOM × Fcor ) + (r2 × f1 )

     
x̂ ŷ ẑ  x̂ ŷ ẑ  x̂ ŷ ẑ 
     
=  0 w0 /2 z0  +  0 w0 /2 z0  +  0 w0 0 
0 0 −mg   0 Fcor 0   0 f1,y f1,z 
 
mgw0
= x̂ − − z0 Fcor + w0 f1,z (11)
2
=0

Using Equation 11, we can find the value of f1,z in terms of the forces we can compute and
in terms of the distances where the forces act:
mg z0
f1,z = + Fcor (12)
2 w0
Method 2: Force Approximations

Instead of invoking torques, we can make further approximations of the force components
through physical arguments. Let’s the consider the simplest, “everyday” case, where we ig-
nore the non-inertial forces and simply consider the force due to gravity and the two reaction

4
forces on the rails. In the simplest case, the reaction forces are equal, and so f1 = f2 = mg/2
and each point in the ẑ direction, so that the sum of the two forces perfectly balance the
weight of the train.

In the case at hand, where we consider the Coriolis force to first order in ω, Equations
9 and 10 suggest that f1 and f2 are slightly modified. From Equation 10, we see that
f1,y = f1 sin φ = Fcor , where tan φ = Fcor /Fg (see Figure 1). Given the assumptions of how
the reaction forces are generated, f1,z should be equal to mg/2 with the addition of a term
that is proportional to Fcor . From Equation 9, we see that the Coriolis term enters in the
z-component Equation as f1 cos φ = (f1 sin φ) cot φ = Fcor cot φ. If we further assume that
this component of the Coriolis term, as with the weight, is distributed evenly between the
two rails, then we can write the first-order equation for f2 as
mg 1
f2 = − cot φFcor
2 2
since one would expect the eastward force to slightly lessen the reaction exerted by the
western rail. For an order-of-magnitude calculation, we can require that tan φ ∼ z0 /(w0 /2) =
2z0 /w0 . Therefore, with the above equation for f2 , we can find the modified equation for
f1,z using Equation 9:
 
1
f1,z ≈ mg + cot φFcor
2
mg z0
= + Fcor (13)
2 w0
Conclusions

The combination of Equations 9, 10 and 12 allows us to uniquely solve the system for the
components we want. Since we’re asked to find the ratio of the magnitudes of rail forces,
we must first compute the magnitude of f1 ; we can do this by adding the components in
quadrature and substituting the results from Equations 10 and 12:

f12 = f1,y
2 2
+ f1,z
 2
2 mg z0
= Fcor + + Fcor
2 w0
z02 (mg)2 mgz0
 
2
= Fcor 1 + 2 + + Fcor
w0 4 w0
(mg)2 mgz0
≈ + Fcor (14)
4 w0
2
where, in the final form of Equation 14, we note that Fcor ∝ ω 2 and drop this term since ω
is very small. In order to find the ratio f1 /f2 , we need to use Equation 9 and 12 in order to
find f2 ,

5
 mg z0
f2 = mg − f1,z = − Fcor (15)
2 w0
and use the same approximation method we employed when deriving Equation 14 to find
f22 . The square of the ratio of force magnitudes becomes

(mg)2 mgz0
f12 4
+ w0
Fcor
= (mg)2
f22 − mgz0
Fcor
4 w0
  −1
4z0 4z0
= 1+ Fcor 1− Fcor
mgw0 mgw0
 2
4z0
≈ 1+ Fcor (16)
mgw0

where we noted that mg >> Fcor on the Earth, and once again that ω is very small (and so
2
neglecting terms proportional to Fcor ). We can finally find the ratio of rail forces by taking
the square root of Equation 16 and plugging in the expression for Fcor in terms of g and ω:

f1 8ωz0
=1+ v sin θ (17)
f2 gw0
(b) Compute value of the ratio.

We’re asked to consider a scenario where v = 100 m s−1 , θ = 45 degrees, z0 = 2 m, w) = 1.5

f1
≈ 1.006 (18)
f2
So, even when deriving Equation 17 by only considering terms that are up to first order in
ω, we find that the difference in magnitudes for the two reaction forces are pretty small.

(b) Find the angle of tilt.

The resultant force due to gravity and the Coriolis effect will only slightly differ from the
weight of the train. In other words, we should expect a inclination angle of the platform
needed for the resultant force to be perpendicular. We can compute this angle by noting
that

6
  
Fcor
ψ = arctan
Fg
 
2mωv sin θ
= arctan
mg
 
2vω sin θ
= arctan (19)
g

which, when plugging in the numbers from part (b), comes to

ψ ≈ 0.06 degrees. (20)

7
 z

Fcor
rCOM
z0

f1,y
f2 f1 f1,z

Fg
y
r2
−w0 /2 w0 /2