The application of multiple-input models to analyze structurally generated noise from a forge hammer is discussed. Part I of this article presented the rationale for developing and interpreting multiple-input models for structural noise source identification. An investigation of the transducer requirements for characterizing the sound radiation from a monolithic element showed that a single, well-placed accelerometer may be sufficient for each element. Part II of this article analyzes the application of the multiple-input modeling technique to the structural noise source identification of a Chambersburg #8 die forger. A comparison of three-, five-, and seven-input models applied to the forge hammer under production conditions indicates that as few as five transducers would suffice to characterize the sound contributions of the five structural elements. Analysis of these models indicates that the ram is the dominant source of sound energy, the columns are secondary sources, and the yoke and anvil are minor sources when detected through a microphone at the operator's position. The analysis also shows that the coupling between the hammer structural elements is sufficient to render conventional wrapping identification methods unreliable for analyzing hammer noise.
All Science Journal Classification (ASJC) codes
- Acoustics and Ultrasonics
- Arts and Humanities (miscellaneous)