Ford's "Virtual Vehicle Sound Simulator" looks like every video game racer's wet dream; three wide-screen monitors, force-feedback controls and Sennheiser headphones shame most setups, but this is actually a heavy duty in-vehicle sound simulator — an acoustic engineer's wet dream.
The driving simulation is pretty lame by even late 90s gaming standards; it's a road, with some trees and cars if you want, but you get the idea when you learn it can be set of automatic steering mode. It's a visualization to provide psychological and visual feedback surrounding the real purpose of the system. The VVSS may look like a game, but it's really a high-powered tool used by Ford's NVH group to develop a vehicle's sound signature before the first prototype ever hits a wind tunnel. Sound tuning has long been the last bastion of the craftsman engineer. The NVH profession has relied on a tuned ear and intuition to determine sources of noise and the paths they take into the cabin, and then recommend adjustments to either eliminate them or make them more pleasant. With this system, Ford is beginning to move that work into the proactive, digital realm to join nearly every other aspect of vehicle design.
The system is a first a North American manufacturer and found its first uses in the 2010 Ford F-150, it's currently being used in key programs surrounding EcoBoost, but is targeted at all models by the 2012 to 2013 time frame. In essence, the sound simulator is a mix of predicted and recorded acoustics, overlaid, and subjected to simulated vehicle inputs from the driver and the environment, outputting a simulated in-cabin noise experience.
The simulation has three stages. In the first stage, well before prototypes are even built, early design data from the chassis components, engine, driveline, and associated dampers and mounts is all modeled to reflect the harmonic performance of each part, 76 different simulated sounds in total. The components are all put into the simulation as independent equations and the vehicle is built as a vibration model. The model is then subjected to driver inputs, variations in loading on the engine, and road imperfections and as the models individually output frequency and amplitude, they're all summed into an ambient cabin noise which the tester can hear and can visualize on screen. The sources can be sorted and parsed to find sources of harshness or unpleasantness and the offending parts and paths can be determined. The team then works with design and release engineers to modify the part or system to improve the acoustic performance.
In the second stage, an actual prototype vehicle has been built. Since it's been optimized using the system, there's considerably less work in fixing NVH problems since the low-hanging fruit has already been picked. Now engineers can sample the environment directly, recording ambient sound to replace the simulated version. The entire model can then be run as a blend of virtual and real data to improve the resolution of the results. Engineers can also observe the variation between the simulated model and the real and make tweaks to their modeling methods. The tuning on individual parts can also be optimized to improve the experience of the car, not just eliminate squeaks. A throatier exhaust, better intake rush, reduced harshness at 40% throttle under a medium load because of changes to the shape of a transmission mount. These are all things that can be done with this system.
In fact, the Taurus SHO has a sound amplifier hooked to the intake ahead of the turbos which allows for a more authoritative in-cabin sound, otherwise it apparently sounds like a vacuum cleaner. The 2010 Mustang has a pipe with a tuned mechanical amplifier leading from the intake tube after the filter into the cabin to improve in-cabin noise while leaving the exhaust note alone. With the tube removed the car is apparently shockingly quiet inside. Oftentimes these measures are taken so the customer has the proper aural experience while allowing the car to pass the Federal exterior measure sound limit to 80 dB.
In the final stage of the simulation, a completed car is measured with all aspects of the simulation isolated and recorded independently. It's all placed into something like an archival record of the car, which allows the next vehicle program to benchmark itself and improve the sound performance, while maintaining the character of the car. You wouldn't want the next Mustang to accidentally sound like a Taurus, would you? The end result, and the goal of the whole endeavor, is to move sound engineering from reactive artisan engineering to more of a scientifically based, proactive, effort. It's a lot of up-front work, but in the end, Ford is hoping it will pay dividends.