To content

Benchmarking Sound Radiation on a Mounted Car Engine

The engine bay often bias the acoustic radiation estimated with traditional sound pressure based solutions. Consequently, most current methods require dismounting the engine from the car and installing it in an anechoic room through a laborious process in order to measure the sound power emitted. Two particle velocity based methods were proposed to characterize the sound radiated from an engine while it is still installed in the car. Particle velocity sensors are much less affected by reflections than sound pressure microphones when the measurements are performed near a radiating surface due to the particle velocity's vector nature, intrinsic dependency upon surface displacement and directivity of the sensor. Therefore, the engine does not have to be disassembled, which saves time and money.

Extracted from: Comesana, D.F., Tijs, E. and Kim, D., 2014. Direct sound radiation testing on a mounted car engine. SAE International Journal of Passenger Cars-Mechanical Systems, 7 (2014-01-2088), pp.1229-1235.


  • Measure the entire engine with a large sensor array
  • Withstand demanding temperature conditions
  • Perform pressure contribution ranking
  • Validate the effectiveness of the proposed methodology


Benchmark car engines based on their acoustic output without disassembling any components

Sound Power & TPA methods

Two methods involving particle velocity measurements near the radiating object are proposed to calculate the sound pressure emitted by the engine:

  • Airborne TPA : free-field transfer functions measured in an anechoic chamber combined with information captured with particle velocity sensors in the engine bay.
  • A sound power based methodology using only particle velocity sensors.

Roller test bench setup

Thirty eight particle velocity sensors were distributed at all sides of the engine to ensure that all sound radiated is captured. Twenty-one transducers were mounted in probe dampers and then attached to the surface of the engine and the gearbox. The remaining seventeen sensors were mounted near the engine with goosenecks.

Acoustic excitation

The quality of the particle velocity measurements is assessed due to its crucial importance for both methodologies. The data is arranged into seven groups depending on the sensor position during the tests. The figure on the shows the particle velocity signals acquired with a dynamic range of 100 dB. The sensor signals are only affected by airflow induced noise below 125Hz, which can be seen as an increase of the self‑noise level mainly at the front part of the engine.

Free-field sound pressure prediction

A comparison of the sound pressure estimated with a TPA method spatially averaged at four positions at one meter from the engine and the sound power method. The results of the two methods are similar between 300 Hz and 4 kHz, and both are in good agreement with the measured sound pressure performed in the anechoic chamber with the dismounted engine.Although satisfactory results are obtained with both approaches, the sound power method seems to be more convenient to the simplified measurement process.


Two methods based on particle velocity have been introduced to calculate the sound radiated from an engine. The first method combines particle velocity measurements with acoustic transfer paths previously acquired in an anechoic room. The second method estimates sound power using the measured particle velocity together with an acoustic impedance model. Both methods have been compared to a direct sound pressure measurement in an anechoic room. The results achieved are fairly similar, therefore validating both in-situ measurement procedures to estimate the noise emission of a complex noise source using multiple particle velocity sensors.