Edited by Sara Dorfner
Eric Sattler
BASF Corp. Plastic Materials
Mount Olive, N.J.
It’s no secret that plastics offer many design benefits compared to metals, including weight reduction, parts consolidation, low manufacturing costs, and the ability to achieve better design flexibility. Just pop the hood of your car and you’re sure to find many examples of plastic components that were once made of metal. Intake manifolds are one of the more recent applications targeted by designers. Though the sounds or noise produced by plastic manifolds are different than for metals, plastics still offer the flexibility of tailoring designs to achieve specific noise qualities.
Noise abatement is becoming a major design thrust in many industries. Often, the goal is not to eliminate noise, but rather to tune it to produce a specific noise quality. Engineers strive to achieve a noise that the consumer will like to hear, such as automobile engines that “purr,” computer keyboards that “click,” and power tools that “hum.”
BACK TO THE DRAWING BOARD
There are two basic approaches to controlling noise quality. The passive method uses insulating materials, such as foam, fiber mats and other low-density noisedamping materials. An active approach, in contrast, involves modifying component designs so they produce a different sound. The two design methods can be combined or used independently.
In recent years, engineers have favored the active approach by developing new methods that let them more easily control noise quality by modifying part designs. It is now possible to analyze the sound a customer wants and design a part to closely match this sound when in use.
The analysis starts by evaluating noise sources. Noise comes from vibrations. Therefore, modifying the surface velocity of each part surface alters the noise made by the part.
Some design factors that affect noise are:
• Stiffness — higher stiffness yields less noise.
• Weight — more weight results in less noise.
• Damping — higher damping cuts noise.
• Transform function — a mathematical equation that uses amplitude and frequency data to create a vibration “signature” for a part design.
Because engineers strive to reduce component weight and the damping qualities of a particular resin can’t be changed, only two parameters can be manipulated — stiffness and the transform function.
The first step to optimizing stiffness and the transform function is to analyze the noise produced by the plastic part. In order to accurately assess the noise characteristics, the component must be tested in its finished assembly under the same operating conditions anticipated for final use.
Take, for example, an engine intake manifold. The cylinder heads emanate vibrations through to the manifold, which results in audible noise. Internal-pressure variations in the manifold also cause vibrations, leading to a secondary noise source. A third factor is vibrations transferring to the manifold by other underhood components and systems within the engine compartment.
The reaction and interaction of vibrations and sounds is a difficult relationship to envision. For this reason, experimental acoustical testing and analysis of the manifold on an engine is essential. In addition, it’s important that analysis used to reduce and modify noise qualities be conducted under conditions closely replicating a component’s operating environment.
RECORDING AND ANALYZING THE NOISE DATA
The analysis of sounds starts with recording the noise made by a part in service. Sound coming from a system can be recorded in one of two ways. The first is by using a directional microphone held close to the plastic component. The second involves a test-dummy head with microphone “ears.” Sounds are received by an audio station, which combines special computer-based software with a fast Fourier-transform analyzer. By recording sound, engineers can review and analyze plastic components because we distinguish differently between pleasant and unpleasant noises not only on the basis of objective physical parameters, but also on the basis of subjective aspects, so-called psychoacoustic parameters. Typical examples are differently perceived loudness levels and the intensity or tone-pitch patterns.
To avoid the subjective nature of noise analysis, audio station signals are transferred to a computer where they are converted to values. Using digital-band pass filters at the audio station, bands of the original sound can be blocked to help uncover undesired noises.
The digitized sound data show noise levels, measured at close range, in a contoured envelope curve. Also, sound intensity can be shown for discrete frequencies. Analyses for octaves, one-third octaves, or any other frequency band are also possible.
Once noise frequency and sources are evaluated, vibration behavior of components can be determined using modal analysis. Design variations can be studied much more easily and closely with the modal analysis data than would be possible with psychoacoustic analysis.
GETTING GOOD VIBES
The next step in noise analysis is to determine the source of undesired noise. Data from the computerized model are used to produce a simplified schematic drawing that pinpoints noise intensity across part surfaces.
Experimental modal analysis provides more specific data on vibration sources and the primary vibration modes. This test analyzes the part using an accelerometer. The accelerometer is fixed in one location, while a manually held instrumented hammer subjects parts to a known vibration at several discrete locations. The instrumented hammer represents an input vibration that covers all noise frequencies. One drawback, however, is the modal analysis is valid only for the points where the data were taken and cannot give any information in between these points.
To obtain vibration data across an entire part, engineers use holographic interferometry. This method outperforms the hammer method for higher primary vibration modes. The holographic analysis provides design tips for areas that produce specific sound levels and qualities.
Once all of the data have been collected and evaluated, it’s time to modify part designs. To optimize the sound quality of plastic components, engineers overlay data, collected from modal analysis or holographic interferometry, on the schematic drawing of the part. This lets them identify which areas of the part are emitting particular frequencies whether it be the whole part or specific areas.
After noise sources are identified, designers can modify designs by adding ribs and changing wall thickness at specific locations. Using these techniques, noise levels can be reduced and “tuned” to produce more ear-pleasing qualities.