<aside> 🪳 Normal mode: of a system is one in which all parts of the system are oscillating at the same frequency. Normal modes also have the property that they are orthogonal in a vector space. There is no coupling or exchange of energy between them, thus if a system is oscillating in one of its normal modes it stays in that mode
</aside>
The most general motion of the system can be described as a superposition of motion in its normal modes
💼 Case: Consider two identical simple pendula consisting of point masses $m$ suspended by massless strings of length $l$ a distance $L$ apart. The masses are connected together by massless springs of natural length $L$ and spring constant $k$
The equilibrium position is $x_1=0$ and $x_2=0$ where the string is at its natural length.
We will consider small oscillations about this equilibrium position $x_1\sim x_2 \ll l$
If a system obeys a symmetry the normal modes are eigenfunctions of that symmetry,.
In this case with mirror symmetry the normal modes are either symmetric or antisymmetric
We can work out the frequencies of the normal mode
To obtain the Lagrangian we have to calculate the gravitational potential energy of each pendulum when it is displacement a small horizontal distance $x$ from vertical.
This corresponds to an angle of displacement $\theta \approx x/l$ at which point the height of the mass $m$ below the pivot is $l\cos\theta \approx l(1-\frac 12 \theta^2)=l-\frac 12 x^2/l$.
Neglecting the constant term we have a potential energy $\frac 12 mg/(l(x_1^2+x^2_2)$
When the two masses are displaced by $x_1$ and $x_2$ the spring is stretched to length $x_1-x_2+L$ and so its potential energy is $\frac 12 k(x_2-x_1)^2$
Thus the Lagrangian is
$$ L=\frac 12 m(\dot x^2_1+\dot x^2_2)-\frac{mg}{2l}(x^2_1+x^2_2)-\frac 12 k(x_2-x_1)^2 $$
The equations of motion corresponding to $x_1$ and $x_2$ are then
$$ \begin{aligned} \frac{\text d}{\text dt} \left ( \frac{\partial L}{\partial \dot x_1} \right )&=m\ddot x_1 =\frac{\partial L}{\partial x_1}=-\frac{mg}{l}x_1+k(x_2-x_1) \\ \frac{\text d}{\text dt} \left ( \frac{\partial L}{\partial \dot x_2} \right )&=m\ddot x_2 =\frac{\partial L}{\partial x_2}=-\frac{mg}{l}x_2+k(x_1-x_2)
\end{aligned} $$
We will solve these equations in 3 ways of increasing formality and generality
By adding both equations we get
$$ m(\ddot x_1+\ddot x_2)=-\frac{mg}{l}(x_1+x_2) $$
or
$$ x_1+x-2=-\frac gl (x_1+x_2) $$
💎 Conclusion: the combination $x_1+x_2$ oscillated with $\omega^2=g/l$
By subtracting we get
$$ m(\ddot x_1-\ddot x_2)=-\frac{mg}{l}(x_1-x_2)-2k(x_1-x_2) $$
or
$$ x_1-x_2=-\left ( \frac{g}{l}+\frac{2k}{m} \right )(x_1-x_2) $$
💎 Conclusion: the combination $x_1-x_2$ oscillates with $\omega^2=g/l+2k/m$
🗒️ Note: A normal mode implies that all components of the system oscillate at the same frequency, but with different amplitudes or phases.
We assume that in the $n$th normal mode the frequency is $\omega^2_n$ and write the displacement of each component as
$$ x_j=C_je^{i\omega_n t} \quad i\in\N^* $$
where $C_j$s are constants
🗒️ Note: in the present problem $C_j$ can be real but in general it should be complex to allow different phase between different components of a system