High-intensity noise testing

Many aircraft and spacecraft components are exposed to high levels of noise-induced vibration at take-off and in flight. This can cause problems such as:

Failure of micro-electronic component lead wires
Chafing of wires
Cracking of printed circuit boards
Malfunction or failure of waveguides or Klystron tubes
Vibration of optical elements
Failure of joints in structures made from composite materials.

High intensity noise sources in the large reverberant room

Our high-intensity noise source in the Large Reverberation Room

To make sure components will not fail under this acoustic stress they are subjected to a high level noise field.

With spacecraft components, the stress occurs for a short time at take-off, typically one or two minutes. During testing the noise exposure is as close to real conditions as possible.

Aircraft components are more likely to be stressed for perhaps thousands of hours, so that real-time testing would be impracticable. In these cases accelerated testing is carried out with far higher noise levels but for much shorter test periods, typically one or two hours.

For spacecraft components the excitation spectrum is usually specific to the launch vehicle, but standard levels have been accepted for aircraft parts. These are specified in various documents, for example MIL-STD-810, or BS 3G 100. In either case overall levels of up to 160 dB may be required in various frequency ranges between 25 Hz and 10 kHz.

Levels of up to 147 dB can be generated in the Large Reverberation Chamber (348 cubic metres). In the Small Reverberant Chamber (131 cubic metres) levels of up to 158 dB can be attained, depending on the sound spectrum required.

Even higher levels, up to 170 dB, are attainable, but in limited frequency bands and normally only on flat or nearly flat panels using a Progressive Wave Tube. Such tests are normally continued until the component under test shows signs of damage which can take many tens of hours.

The test sample is normally monitored by accelerometers or strain gauges. When using a large number of these transducers their outputs are usually recorded on magnetic tape for subsequent analysis, but a limited number of channels can be monitored in real time. The recorded data can be reduced to narrow-band spectra, one-third octave band or octave band spectra.

If you require any further information please contact:

Dave Rawlinson or John Fithyan
ISVR Consulting
University of Southampton
Highfield
Southampton SO17 1BJ
UK

E-mail: consultancy@isvr.co.uk