Calculations
Placing a diagram of a cross-section of
the rotating jar and water in a Cartesian plane as shown,
x0
our objective is to derive a function u(x) that models the shape
of the water by giving the height of the surface of the water above a given
point on the x-axis. We can derive such
a function by calculating the total energy from the system and minimizing it
using the Euler-Lagrange equation.
We consider only the rotational kinetic
and gravitational potential energy of the system, assuming all other sources of
energy to be negligible because they are relatively weak in this system (we
will later examine the effects of factoring surface tension into our
calculations). From mechanics of
physical systems, we have the equations
,
where
K is the rotational kinetic energy of the system, U is the gravitational
potential energy, I is the moment of inertial of the water, ω is the
rotational velocity the fluid (constant), ρ is the density of the water (constant),
g is the force of gravity on the water (constant), and h is the height of each
infinitesimally small cube of water dV above the bottom of the jar. We find the total kinetic and potential
energies of the system by integrating over all cubes dV of water.
We consider our kinetic energy
equation first. Since ω is constant
throughout the water, we can place it outside our integral, leaving us to
calculate I. By the formula for moment
of inertia, the value of I for each element dV of water is I = ρx2,
where x is the distance of dV from the axis of rotation (the u-axis in the
diagram above), so our total moment of inertia is
,
and
our total kinetic energy is
.
However,
the integral in this form is too difficult to solve, in part because we do not
have clear bounds and in part because we have not found a way to include u(x)
into the equation. We solve this problem
by changing the variables from Cartesian to cylindrical coordinates, giving us
,
where
R is the radius of the cylinder, x is the horizontal distance of each element
dV from the axis of rotation (the u-axis on the diagram at top), z is the vertical
distance of each element from the bottom of the jar, and θ is the angle of
dV from a fixed Cartesian plane that intersects the axis of rotation.
So we now have an integral that, if
evaluated, may be solved for u (x) but since kinetic energy is only part of the
overall equation (total energy) that we wish to derive, we will hold off and
solve it later as part of this larger equation.
We now consider our potential energy
equation. We have
.
We convert the variables to
cylindrical coordinates following the same logic as our kinetic energy
equation:
.
So our total system energy is
.
Kinetic
energy is negative because, by the law of the conservation of energy, kinetic
and potential energy have an inverse relationship.
Note, however, that we have a
constraint in the motion of the water inside the jar. Assuming that compression forces on the water
are negligible and observing that no water is lost during the system’s
rotation, our volume remains fixed.
Thus, our calculations are affected by the constraint
,
where
V is the total volume of the water. By
the shell method,
.
We
then factor this limitation into our total energy equation using a Lagrange
multiplier λ, so we have
.
Following the procedure for
implementing the Euler-Lagrange equation on an integrated function, we let F(z,
p) equal the integrand of the energy equation—that is,
,
where z = u(x) and p = u’(x) (note that this
is not the same z used in the triple
integrals above)—which we substitute into the Euler-Lagrange formula
to
get
.
So
we can solve for u(x) to get
.
However, we still need to solve for
λ since it’s the only parameter in the equation, so we examine the
function associated with λ, which in this case is our constraint function
,
which
states that the volume of water within the jar is constant for different
rotational velocities. If we allow the
volume of water when the system is at rest to be πR2h0,
where R is the radius of the cylinder and h0 is height of the
standing water from the bottom of the jar, then
πR2h0
=
.
Evaluating
the integral and solving for λ gives us
,
so our final
value of u(x) is
.