This article focuses on various features of a finite element analysis (FEA) program designed by IDIADA, a Barcelona-based company providing design, engineering, testing, and homologation services to the automotive industry. The program has been designed as a solution to the problem of squeaks and rattles in an automobile. FEA software used by engineers from automotive testing company IDIADA detects potential automotive noise. The company uses Abaqus Unified Finite Element Analysis from Dassault Systèmes’ brand Simulia. The team delivered a paper at the Simulia Customer Conference in Barcelona in May 2011 to present the latest improvements in their methodology, applied to rattle in a car instrument panel and correlated with real-world testing. Research has shown that a standard noise-and-vibration analysis method alone can’t model, or predict, the contact that will result in a rattle. The engineers need to come up with a simulation that can accommodate both frequency for noise and vibration and contact for squeak and rattle.
So you’ve finally bought your shiny, new car. But then one day the noise starts: an unidentifiable, repetitive, distracting sound coming from somewhere inside your car. In no time at all, that annoying sound is driving you crazy.
The problem is known to those in the automotive design industry as squeak and rattle, often written shorthand as S&R, and it's been driving the automotive industry crazy, too.
Engineers from IDIADA have found a way to put their FEA program to work on the problem, said Inés Lama, project manager for design engineering. The company is based in Barcelona, Spain, and offers design, engineering, testing, and certification services for automobiles.
She and her team have found that by isolating the source of squeaks and rattles through software analysis, automotive designers can come closer to doing away with the problem.
Paradoxically, while great progress has been made in the areas of vehicle noise and vibration, the fact that modern automobiles run more quietly than ever has made lingering S&R issues even more apparent. With engine and road noise diminished, smaller sounds that used to be hidden become amplified to the driver's ear. While designers and manufacturers can target specific squeaks or rattles, ongoing automotive trends toward lighter cars and new materials continue to work against total elimination of the problem.
Squeaks and rattles are often located in the interior trim of a vehicle, such as the dashboard, but the exact source can be hard to pinpoint. Squeaks happen when components periodically slip and stick together. Rattles occur when parts hit each other intermittently. Both noises are usually due to inconsistent assembly tolerances or lack of stiffness. Some are more detectable at slower driving speeds, and others get worse as the vehicle accelerates.
When buyers notice a squeak or a rattle during a test drive, they might see it as a sign of poor quality and decide not to buy the vehicle. When squeak and rattle issues surface after purchase, they’re often difficult to diagnose and expensive to fix. And the fixes may even lead to new S&R problems. Subsequent warranty claims can significantly impact vehicle manufacturers’ reputations and profits
“Squeak and rattle have become of greater concern for us in recent years as more auto manufacturers come to us with these problems,” Lama said. “Our customers are asking us if it's possible to use simulation to identify the potential for squeak and rattle earlier in their design processes, rather than later when it is more costly and time consuming to solve.”
As a mechanical engineer with a university concentration in vibration and noise and more than a decade spent working on multiphysics issues at IDIADA, Lama is well-versed in using simulation software to visualize and predict the complex material responses that arise when motor vehicles meet the open road.
She and her team have been using finite element analysis software for this purpose for years. The company uses Abaqus Unified Finite Element Analysis from Dassault Systèmes’ brand Simulia.
“Since we’d already been using Abaqus in vehicle cockpit design and testing for thermal, impact, and normal modal analyses, it made a lot of sense to simply develop a new load case for squeak and rattle inside Abaqus,” Lama said.
In 2008, her team began developing and squeak-and-rattle-specific simulation protocol.
The team delivered a paper at the Simulia Customer Conference in Barcelona in May 2011 to present the latest improvements in their methodology, applied to rattle in a car instrument panel and correlated with real-world testing.
The instrument panels—both physical and the virtual ones—were donated to IDIADA by Spanish car manufacturer SEAT. Designers had called upon an FEA model in the normal course of developing the instrument panel. Now they wanted to specifically simulate a rattle in the instrument panel.
Rattle arises when parts collide. The relative movement between them can generate noise if the surfaces adjacent to where the impact occurs radiate sound.
Developing a new load case to simulate such an event required some creative thinking, Lama said.
A classic automotive noise and vibration analysis of the effects of a car engine running, or tires rolling uses modal theory to predict at what frequencies certain parts of the vehicle will begin to vibrate.
“Modal theory is based on the hypothesis of linearity, without contact,” Lama said. “But an S&R event, although it happens within a frequency-dependent, noise-and-vibration-type setting, is also very nonlinear. The parts producing the squeaks and rattles are interacting with each other in three dimensions.”
A standard noise-and-vibration analysis method alone can’t model, or predict, the contact that will result in a rattle, she said.
The engineers needed to come up with a simulation that would accommodate both frequency for noise and vibration and contact for squeak and rattle.
Pau Cruz, along with colleagues Jordi Viñas and Andreas Rousounelos, all on the IDIADA computer-aided engineering team, realized a feature within the FEA program could be used as a virtual sensor to detect contact.
While the virtual sensors could provide a more accurate idea of the amplitude of movement, or interference, between two parts, amplitude alone didn’t predict the possibility of rattle, Lama said.
They also needed to determine how much the parts were interfering with each other, which is known as the value of penetration. The amount of the interference was, in turn, affected by the frequency at which the components were vibrating at various points on the instrument panel, Lama said.
By adding more geometric details to their models, including simulation of the masses of the radiators in the HVAC system, and by fine-tuning the stiffness of different connection points, the engineers were able to pinpoint three areas in the instrument panel that rattled the most.
These were the two connections of the HVAC duct leading to the two side diffusers and the connection of the HVAC duct leading to the central diffuser, Lama said.
The first simulations of these three areas predicted rattles with good correlation to physical tests, although the simulations detected many more rattling issues than heard in the real-world tests.
“In the future we’ll be working on rattle detection criteria improvement to differentiate between rattles that can be heard and those that can’t,” Lama said. “We’ll also continue to refine our analyses to include those zones in the vehicle cockpit where there can be more problems with tolerances.”
Down the road, as a result of this research, there may be a lot less squeak and rattle, and many more happy automobile manufacturers—and drivers.