A small mass \(\varepsilon\) is added to a rigid body with moments of inertia
\(I_{1}>I_{2}>I_{3}\) at the point \(\mathbf{Q}=x_{1} \mathbf{e}_{1}+x_{2}
\mathbf{e}_{2}+x_{3} \mathbf{e}_{3}\). Find the change in \(I_{1}\) and
\(\mathbf{e}_{1}\) with error \(\mathrm{O}\left(\varepsilon^{2}\right)\).
Solution. The center of mass is displaced by a distance of order
\(\varepsilon\). Therefore, the moments of inertia of the old body with sespect
to the parallel axes passing through the old and new centers of mass differ in
magnitude of order \(\varepsilon^{2}\). At the same time, the addition of mass
changes the moment of inertia relative to any fixed axis by order
\(\varepsilon\). Therefore, we can disregard the displacement of the center of
mass for calculations with error \(\mathrm{O}\left(\varepsilon^{2}\right)\).
Thus, after addition of a small mass the kinetic energy takes the form
$$
T=T_{0}+\frac{1}{2} \varepsilon[\Omega, Q]^{2}+O\left(\varepsilon^{2}\right)
$$
where \(T_{0}=\frac{1}{2}\left(I_{1} \Omega_{1}^{3}+I_{2} \Omega_{2}^{2}+I_{3}
\Omega_{3}^{2}\right)\) is the kinetic energy of the original body. We look for
the eigenvalue \(I_{t}(\varepsilon)\) and eigenvector
\(\mathbf{e}_{1}(\varepsilon)\) of the inertia operator in the form of a Taylor
series in \(\varepsilon\). By equating coefficients of \(\varepsilon\) in the
relation \(A(\varepsilon) \mathbf{e}_{1}(\varepsilon)=I_{1}(\varepsilon)
e_{1}(\varepsilon)\), we find that, within error
\(\mathrm{O}\left(\varepsilon^{2}\right)\) :
$$
I_{1}(\varepsilon) \approx I_{1}+\alpha\left(x_{2}^{2}+x_{3}^{2}\right) \text
{ and } \mathbf{e}_{1}(\varepsilon) \approx
\mathbf{e}_{1}+\varepsilon\left(\frac{x_{1} x_{2}}{I_{2}-I_{1}}
\mathbf{e}_{2}+\frac{x_{1} x_{3}}{I_{3}-I_{1}} \mathbf{e}_{3}\right)
$$
From the formula for \(I_{1}(\varepsilon)\) it is clear that the change in the
principal moments of inertia (to the first approximation in \(\varepsilon\) ) is
as if neither the center of mass nor the principal axes changed. The formula
for \(e_{1}(\varepsilon)\) demonstrates how the directions of the principal axes
change: the largest principal axis of the inertia ellipsoid approaches the
added point, and the smallest recedes from it. Furthermore, the addition of a
small mass on one of the principal planes of the inertia ellipsoid rotates the
two axes lying in this plane and does not change the direction of the third
axis. The appearance of the differences of moments of inertia in the
denominator is connected with the fact that the major axes of an ellipsoid of
revolution are not defined. If the inertia ellipsoid is nearly an ellipsoid of
revolution (i.e., \(I_{1} \approx I_{2}\) ) then the addition of a small mass
could strongly turn the axes \(e_{1}\) and \(e_{2}\) in the plane spanned by them.