Noise and Vibration Control Engineering
This domain tests whether you can choose and size the right control measure at the right point in the source–path–receiver chain, and defend the numbers.
Core rules
- Control hierarchy: reduce at source, then interrupt the path, then protect the receiver. Source treatments are usually the most cost-effective and the least dependent on worker behaviour.
- Mass law for a single homogeneous panel: TL ≈ 20 log10(m·f) − 47, with m in kg/m² and f in Hz. Doubling mass or frequency gives about +6 dB, but a coincidence dip and flanking paths cap real performance.
- Double-leaf construction with an absorptive cavity outperforms a single leaf of the same total mass, provided the leaves are structurally isolated and the cavity is truly sealed.
- Silencers: reactive types are tuned and narrowband; dissipative types are broadband but can suffer from flow-generated noise and contamination.
- Vibration isolation: fn ≈ 4.98/√δ(cm); transmissibility T = 1/|(f/fn)² − 1| for the undamped case. Isolation only occurs above f/fn = √2, and near resonance the mount amplifies.
Worked example
A 20 Hz machine sits on mounts with 5 mm static deflection. fn = 4.98/√0.5 = 7.0 Hz; f/fn = 2.86; T = 1/(8.2 − 1) = 0.141. Since transmissibility is an amplitude ratio, the reduction is 20 log10(0.141) = −17 dB. Choosing softer mounts to gain more deflection risks instability and a lower-frequency resonance.
Traps that cost marks
- Using 10 log10 instead of 20 log10 on a transmissibility or amplitude ratio.
- Specifying isolation below √2 fn, where the mount makes vibration worse.
- Ignoring leaks and flanking: a 1 % open gap can cap a wall's field TL near 20 dB regardless of mass.
- Designing a barrier or enclosure that is not sealed at joints and penetrations.