A trace rings around its final value. You add damping, the ringing shrinks, and the result looks better. Keep increasing it and the motion becomes quiet but takes longer to arrive. The missing question is which part of the response you are trying to improve.
Start with forces, not a damping label
Our horizontal mass has displacement x, velocity v and external force F. The spring force is −kx and the viscous damper force is −cv. The damper opposes motion in either direction. Newton's law gives m x″+c x′+kx=F. With m=1 kg, k=16 N/m and a constant 2 N input, a resting equilibrium satisfies 16x=2, or x=0.125 m.
At rest, v=0, so damping contributes no static force. Changing c alone cannot change this equilibrium. It changes how the mass gets there. That distinction prevents a common mistake: judging the final displacement from how dramatic the early motion looks. At c=0, the same equilibrium exists mathematically, but a step from rest oscillates around it instead of converging.
Two poles explain the slow tail
The zero-initial-state transfer function has denominator ms²+cs+k. In this example its natural frequency is 4 rad/s and damping ratio is c/8. At c=1.6, the poles are −0.8±j3.919: the real part describes decay and the imaginary part describes ringing. At c=8, both poles are −4, the critical-damping boundary. A repeated pole gives a time response containing t exp(−4t), not just a single exponential.
At c=16, the poles separate to about −1.072 and −14.928. The fast mode disappears quickly, leaving the slow one. Increasing c further can move that slow pole closer to zero; for very large c it is approximately −k/c. The trace has less ringing but a longer tail. Critical damping is the fastest monotonic step response within this fixed-m, fixed-k family, not the fastest design under every metric.
| Damping c (N·s/m) | Ratio ζ | What to look for |
|---|---|---|
| 1.6 | 0.2 | Ringing and overshoot |
| 8 | 1 | Monotonic approach |
| 16 | 2 | A longer slow tail |
Compare the same question three times
Open One force, different transients. It starts at x=v=0, with m=1, k=16, F=2 and c=1.6. Predict whether a higher c changes the final position. Pin the run, change only c to 8, pin again, then try 16 and pin the third run. Use one replay time for the device and all traces. Check the equilibrium line, first crossing, maximum displacement and late tail.
Rise time measures the first 10% to 90% passage; it does not say that the response stays near the final value. Overshoot describes a peak. The displayed 2% settling time requires at least one second in the band through the end of this recording. If the window ends too early, Not settled is an honest result. Increase duration while keeping the other settings fixed before concluding that a stable response never settles.
A damper dissipates energy, but motion sets the rate
Choose Release, then watch the energy. With F=0, total stored energy is E=½mv²+½kx² and its derivative is −cv². Set c=0: energy stays constant while kinetic and spring energy exchange. Restore positive damping: total energy decreases. The energy panel also integrates dissipated energy and external work, so you can check E(t)−E(0)=W−D rather than infer energy from position alone.
The rate is not c alone; it is c times velocity squared. A very large damper can make motion so slow that energy drains over a long interval. At nonzero applied force, input power Fv can replenish stored energy. That is why a periodically forced mass can keep moving even with positive damping. A damper is not an energy-storage element like a mass or spring.
Ringing frequency is not the forced-response maximum
For the c=1.6 case, the undamped natural frequency ωn is 4 rad/s, free damped oscillation ωd is about 3.92, and the displacement-gain maximum ωr is about 3.84. The linked resonance preset compares sine inputs at 1, 3.84 and 8 rad/s. Inspect late motion and the normalized plant frequency curve. Its peak describes a sinusoidal particular response after startup transients become negligible, not every peak in a short recording.
For ζ≥1/√2, this force-to-displacement curve has no nonzero resonance maximum even though ωn remains defined. At zero damping and exact ωn, a nonzero sine input makes amplitude grow rather than approach a bounded steady state. These statements depend on the input and output being force and displacement; another transfer path can have a different frequency curve. This lab has no feedback controller, so plant resonance is not a gain-margin or phase-margin measurement.
Choose what you mean by better: less overshoot, faster first arrival or earlier sustained settling. Match the initial conditions and input before using damping to explain a difference.
CHECK YOUR UNDERSTANDING: Why do c=1.6 and c=16 have the same static equilibrium but different settling behavior?
At rest, cv=0 and kx=F. During motion c changes the poles, velocities and dissipation, so the route to that equilibrium differs.