Experimental Identification of Noise Paths and Mitigation Methods for a Device in a Submerged Environment

Open Access
Clark, Adam Leonard
Graduate Program:
Master of Science
Document Type:
Master Thesis
Date of Defense:
April 24, 2014
Committee Members:
  • Martin Wesley Trethewey, Thesis Advisor
  • Acoustics
  • Vibrations
  • Noise Control
  • Common-Mode Choke
  • High Frequency
  • Underwater Noise Control
  • Source-Path-Receiver Model
A common concern in the design of electromechanical devices are the noise emissions. In this thesis an electromechanical device that demonstrates an objectionable radiated tonal noise caused by a forced vibration from the system’s electric actuators is studied. The effort concentrates on first determining the transmission paths between the source and receiver. Once the primary path is determined, mitigation techniques are investigated. A small scale water test environment is developed to easily evaluate the radiated noise. The limits and capabilities of using the small scale reverberant environment are determined. A controlled acoustic source is used to evaluate the variation in mean radiated spectrum level measurements at multiple locations within the environment. From this study an optimal location for a receiver hydrophone is determined having a precision interval of ± 1.1 dB when measuring the controlled acoustic source. The problematic tonal noise is determined to be related to the pulse-width modulation switching frequency of the electric actuators. The characterized reverberant environment is used to determine the dominant noise path within the electromechanical device in a submerged water state. A source-path-receiver model is developed with five possible noise paths. Path breaks are methodically applied to rank their order of relative importance. The dominant noise path is determined to be between the electric actuators through the actuator mounts to the unit housing. When the pathway is broken, a 39 dB reduction is observed at the first problematic harmonic frequency and a 26 dB reduction at the second harmonic. These results confirm the dominant noise path from the source to the receiver. Mitigation techniques are examined for the electromechanical device and evaluated to show noise reduction proof of concept. A literature search of viable noise and vibration control techniques is performed. The application of a common-mode choke to the electric actuators appeared the most promising due to it being applied at the source, cost effective, and simple to design. The common-mode choke proof of concept consists of two 4.7 mH inductors and one 1 μF capacitor. When evaluated in the submerged small scale test environment a 23.7 dB and 5.4 dB reduction is observed at the first and second harmonic of the pulse-width modulation switching frequency, respectively.