Automotive interior noise has become a critical consideration for manufacturers. Many studies have shown that interior noise in a vehicle has a major impact on the customer's perception of operation, performance, and quality. However, to reach desired interior sound levels, many manufacturers have needed to add sound absorption, damping, and barrier materials to the vehicle. These additions lead to higher weight and costs.

Automated Analysis Corp. offers a new capability for optimizing vehicle interior noise treatments. The new approach combines test data, finite element analysis (FEA), and boundary element methods (BEMs) to produce the best available process for vehicle interior acoustic optimization.

Figure 1: Vehicle Surface Velocities

Ideally, one would minimize noise generation at the source, whether it is the engine exhaust, tire-road contact (road noise), wind noise, or any of a hundred other sources. Often, treatments along the path of sound energy transmission are used to reduce noise. These may take the form of vibration isolation, sound barriers, sound absorption materials, or damping materials.

The new approach - Panel Acoustic Contribution Analysis or PACA - employs Automated Analysis Corp.'s COMET software combined with FEA and/or measured results. COMET uses the boundary element analysis process to build very precise relationships between the surface velocity of panels and noise at any interior point. These sensitivities allow the precise calculation of the contribution from a given panel.

Since COMET models the precise relationship between sound waves arriving at the point of interest, it is possible to determine which waves contribute positively to the total sound level, which have no contribution, and which contribute negatively or reduce the sound. By tracing these contributions back to the panel from which they originated, COMET allows the optimization of the noise treatment.

Where there is a negative contribution, one may reduce the noise treatment by removing barriers or damping treatments. The noise from those panels is actually reducing the noise level at the point of interest, and if possible, this effect should be increased. A significant side benefit of defining such panels is that vehicle weight reduction is possible by removing the sound treatments that are counterproductive.

Perhaps the best way to illustrate the usefulness of PACA is to run through an example. This instance begins with interior surface velocities that are obtained from a combination of measurements and FEA analysis. As is often the case with automotive vehicles, FEA analysis can be used to provide reliable predictions of results to only 200 to 250 Hz. Above these frequencies, the computational requirements get to be larger than what is practical. In this case, experimental measurements were made to determine surface velocities to 600 Hz.

COMET then was used to perform the PACA analysis with the driver's ear as the point of interest in the interior. Using a vertical line of icons, COMET takes the user from the entry of data through to the final analysis. Along the way, data and model checks are performed when appropriate, and the user is prompted for all information required. If the user desires, the computer interior noise with improved treatments can be played back using wave files to actually hear the results of recommended changes.

In this particular case, there are large surface velocities in the front passenger corner of the interior cabin (Figure 1). This is shown by the red-yellow color concentration in the lower right corner. Using traditional methods, this is an area that would receive additional acoustic treatments in an effort to lower the interior noise.

Figure 2:Panel Contributions to Interior Noise

The PACA analysis for the same vehicle at the same frequency shows that the contributions from panels look quite different (Figure 2). This contribution calculation of sensitivity is expressed in a range from positive values to negative values representing positive to negative contributions to the noise at the driver's ear. In this case, the change from yellow to green represents the transition from positive to negative. The high velocities in the front passenger corner actually contribute negatively to the interior noise at the point of interest. The best approach would be to not alter or reduce the noise treatment in this area to reduce the noise level at the point of concern.

Midway along the length of the vehicle on the passenger side, as shown in Figure 1, there is another area of high surface velocity. In this case, the sensitivity shown in Figure 2 is positive, meaning that the high surface velocities are a significant contributor to interior noise. Improved noise treatment is necessary in this area to improve interior noise at the driver's ear.

One important concern in such optimization programs is that the optimum obtained be robust enough to handle actual vehicle and occupant variations. Using COMET, it is possible to examine the distribution of sound pressure along the length or width of the vehicle. This sort of analysis can show how sharp variations occur and whether the optimum reach for treatment is robust enough for variations in passenger seating or other factors (Figure 3).

Figure 3: Sound Distribution in the Vehicle Interior

Clearly, the PACA approach to vehicle interior treatment optimization has some academic value, but one may ask how it works with real-world problems. This approach has been used on several vehicle development programs. In one case, an automobile manufacturer was introducing a new sport-utility vehicle and had implemented a large amount of noise control treatment to meet ambitious noise performance goals. The goal was to achieve interior noise levels as good as a "best in class" luxury vehicle. Unfortunately, along the way, too much weight was added to the vehicle.

The development program had reached a stage at which the vehicle weight targets were exceeded and the noise objective was not quite achieved. In an attempt to reduce vehicle mass and continue to work toward the noise performance goal, PACA was employed on a trial basis. The result was a reduction of 2 dB in interior noise, with a weight reduction of 10 pounds. The vehicle was able to meet the ambitious noise performance target at a lower weight than what was thought possible.

This work was performed by James K. Thompson, Ph.D, PE, for Automated Analysis Corp., Ann Arbor, MI. For more information, call 313-973-1000; Fax: 313-973-1190;