Resonance and Beats: When Forcing Meets Frequency

Picking up the story from the frequency response

By the end of the previous guide you had the full picture of a driven oscillator. A periodically forced spring-mass-damper, m x'' + c x' + k x = F0 cos(omega t), splits cleanly into a transient that dies away and a steady state that keeps ringing at the DRIVING frequency omega, not the system's own. You also met the frequency response curve: a plot of steady-state amplitude against omega that rises to a hump somewhere near the natural frequency and falls off on either side. This guide is about what happens at that hump — and what happens just beside it.

Two characters drive everything here. One is the natural frequency omega0 = sqrt(k/m), the rhythm the system swings at when left alone — this is the natural frequency you defined back in the simple-harmonic-motion guide. The other is the driving frequency omega, the rhythm the outside world imposes through the forcing term. Resonance and beats are both stories about the GAP between these two numbers. When the gap is zero you get resonance; when it is small but nonzero you get beats. That single difference is the whole plot.

Beats: two close rhythms that breathe

Start with the cleaner case: zero damping, and a driving frequency omega close to omega0 but not equal to it. Solving m x'' + k x = F0 cos(omega t) gives a particular piece oscillating at omega plus a homogeneous piece oscillating at omega0, and if the system starts from rest you can combine them into one tidy product. Using a standard trig identity, the sum of two cosines of nearly equal frequency becomes a fast oscillation multiplied by a slow one — that is the algebra behind beats.

cos(omega t) - cos(omega0 t)
     =  2 * sin( (omega0 - omega)/2 * t ) * sin( (omega0 + omega)/2 * t )
            \_______ slow envelope _______/   \____ fast carrier ____/

   fast carrier  ~  the average rhythm   (omega0 + omega)/2
   slow envelope ~  the half-difference  (omega0 - omega)/2   <-- tiny when omega ~ omega0

Listen to what this product means. The fast factor wobbles at roughly the average of the two frequencies — you barely notice it changing. The slow factor, oscillating at the tiny half-difference (omega0 - omega)/2, acts as a slowly opening and closing AMPLITUDE: it swells the motion, then squeezes it almost to nothing, then swells it again. Two guitar strings tuned a hair apart produce exactly this — a single tone that throbs louder and softer. The closer omega creeps toward omega0, the slower the envelope, and the longer each swell lasts.

Resonance: when the gap closes to zero

Now let omega slide all the way onto omega0. The beats envelope, swelling ever more slowly, never gets a chance to come back down — the amplitude just keeps climbing. To see it cleanly, look at the undamped equation driven exactly on resonance, m x'' + k x = F0 cos(omega0 t). Here the usual guess x = A cos(omega0 t) FAILS, because cos(omega0 t) is already a solution of the homogeneous equation. This is precisely the resonant forcing situation: the drive is feeding energy in at the system's own rhythm, and the standard trial solution is exhausted.

The modification rule from undetermined coefficients tells you exactly what to do: multiply the trial solution by t. The particular solution becomes something like x_p = (F0 / (2 m omega0)) * t sin(omega0 t). Stare at that factor of t out front. It means the amplitude is not a fixed number but GROWS LINEARLY with time — the oscillation gets wider and wider without bound. This unbounded growth is pure resonance, and it is the mathematical signature of a frictionless system driven at its natural frequency.

There is a lovely continuity between the two phenomena. Beats and pure resonance are not separate effects — pure resonance is the LIMIT of beats as omega approaches omega0. Watch the beat envelope: as the half-difference shrinks toward zero, the slow swell stretches out longer and longer, climbing higher before it would turn around. In the exact limit it never turns around at all, and the slow sine straightens into a line: the t sin(omega0 t) growth IS the first endless half-swell of a beat whose period has gone to infinity.

What real damping does to the picture

The unbounded t-growth is an idealization, and it is important to be honest about it: no real system has zero damping. Restore even a little friction, c > 0, and the t sin(omega0 t) runaway is replaced by a steady-state oscillation of large but FINITE amplitude. Here is the common misconception worth correcting head-on: resonance does NOT require zero damping. A damped forced oscillation still resonates — its frequency response curve still peaks — the peak is simply tall and finite instead of infinite.

Two honest refinements come with damping. First, the peak no longer sits exactly at omega0: with damping the maximum response occurs at a slightly LOWER frequency, omega_r = sqrt(omega0^2 - (c^2)/(2 m^2)), and for heavy enough damping the hump flattens out entirely and there is no resonant peak at all. Second, the height and sharpness of the peak are governed by one dimensionless number, the quality factor Q. A high Q means a tall, razor-thin peak — a lightly damped bell that rings for a long time; a low Q means a short, broad hump that barely qualifies as a resonance.

The same drama in a circuit, and a method recap

Remember the analogy that opened this rung: every word here translates straight into electronics. The resonance of a mechanical spring is the resonance of an RLC circuit, where L plays the mass, 1/C plays the spring constant, and R plays the damping. Drive the circuit with an AC voltage at its resonant frequency omega0 = 1/sqrt(LC) and the current swells to a peak — this is exactly how a radio tunes: the dial sets omega0, and only the station broadcasting near that frequency drives a large response while all others stay off the peak. The quality factor here decides how selectively the radio separates one station from its neighbour.

1. Identify the natural frequency omega0 = sqrt(k/m) (mechanical) or 1/sqrt(LC) (circuit), and the driving frequency omega from the forcing term.
2. If omega is near but not equal to omega0 and damping is negligible, expect beats: write the response as a fast carrier times a slow envelope at the half-difference frequency.
3. If omega equals omega0 with zero damping, the cosine guess is a homogeneous solution — apply the modification rule, multiply the trial by t, and get unbounded t sin(omega0 t) growth: pure resonance.
4. With real damping, drop the runaway and find the finite steady-state amplitude; locate the response peak near omega0 and read its height and sharpness from the quality factor Q.

Step back and see what this rung built. You started with one equation — a mass on a spring, equally a charge in a circuit — and followed it from free simple harmonic motion, through the three damping regimes, into driven motion and the frequency response, and finally to resonance and beats. The payoff is that a single second-order linear ODE quietly explains a pendulum, a tuned radio, a shattering wine glass, and a bridge that must never be marched across in step. The next rung lifts this same linear machinery to order n; the physics intuition you have built here travels with you.