Optimization for squeak and rattle reduction using vibration analysis of passenger car door trim

Introduction

Customer satisfaction is a key factor for automobile manufacturers. This sector always takes steps to solve problems from minor to major ranges. While driving the vehicle one can hear a sound that arises from the engine, aerodynamic parts, tire, etc. Apart from this, there are some abnormal sounds that may be annoying, and irritating. One such group of unwanted sounds is classified as buzz, squeak and rattle (BSR). The range of normal noises are reduced significantly over the years and nowadays electric powertrain doesn’t even make noise as combustion powertrains. Normal noises are a part of operating system of vehicle and they are acceptable up to certain levels but abnormal noises are independent of speed, they cause by sliding or impacting of various assemblies or parts on each other while the vehicle is in operation. For premium segment vehicles abnormal noises are not at all acceptable as such vehicles are purely built on luxury and comfort.^1–3

Squeak is noise generated due to sliding of mating parts over each other. Deformation of surfaces in contact stores some static energy which gets released when static friction exceeds kinetic friction, with this energy release squeak sound is produced. ³ Rattle is sound produced when two parts hit each other, collision between two parts with poor geometric tolerances leads to rattle generation. ⁴

An automobile door is complex assembly of sheet metal panel, plastic trim interior, opening latch, audio system, electronic control panel, glass slider, air tight locking mechanism, etc. All of these parts are assembled together with some pre-defined tolerance limit in between these parts. These parts may impact or slide on each other during a course of dynamic loading. J. Weber, et al. have defined the definition of squeak and rattle planes. As per Weber the relative displacement in the plane is squeak and the relative displacement in normal plane is rattle. ⁵ Modern vehicles are equipped with infotainment systems, control panel, etc which are covered with plastic covering and installed on dashboard, during dynamic conditions these sub-assemblies start impacting with dashboard structure and generate rattle noise. BSR is not only about occurrence but also about how frequently occurrence is obtained. ¹

Weber et al., have defined about Squeak and Rattle evaluation line in two points. Firstly, the gap or contact line between the two parts and second as use of displacement output instead of Squeak and Rattle index. ⁵ Farokh Kavarana et al., mentioned to prevent Squeak and Rattle adding stiffness is better solution but due to cost and heaviness one should focus of minimizing excitation. ³

Design approach to reduce rattle in door trim

The proposed method is carried out on door trim and speaker assembly. The primary focus of this method is to reduce the relative displacement in these two assemblies. As shown in Figure 1 speaker and trim is possible area of rattling. The red marked line shows the location of speaker in trim which have a 0.5 mm gap in between them. Every time during a dynamic condition when speaker and trim exceed this gap there will be an impact between them which will lead to rattle.OPEN IN VIEWER

As per Kavarana, ² very common approach to reduce BSR is find and fix. During this approach insulation foam is used as this foam don’t allow to surfaces to impact on each other and rattle is no longer induced in particular area. Another approach to reduce BSR is adding the stiffening element between two surfaces to prevent them from impacting on each other. In some cases two parallel sheets with smaller gaps are installed which may impact on each other hence to avoid such impact some small spacers are installed at various locations which helps in avoiding impact between them. In BIW(Body in White) structures metallic sheets installed on each other starts to sliding on each other which produces squeak, to avoid that adhesion is provided with help of adhesives or other type of locking systems such as parallel bolting.^6,7 Very commonly used method to avoid rattle is to putting PU foam in between two surfaces. Although foaming method is very cheap compared one discussed in this particular paper, one can’t assure reliability for such foam since vehicle may use in different conditions. To avoid such problem automotive OEMs are now focusing to solve the problem in design phase and BSR is part of “right first time” approach of vehicle design.

Modal analysis and shape optimization

A modal analysis calculates the modal or natural frequencies of component, during the modal analysis process time-history response of system is not required. Once the component is manufactured and its constrained are finalized then its modes are confirmed. They don’t change unless any parameter gets changed. The natural frequency of component can be defined as frequency at which structure vibrates when excitation is provided.^8,9

In the initial stage of design, modal analysis is performed to find modal frequencies of trim. Once the modal frequencies are identified, shifting these frequencies in higher region is a target for optimization. ¹⁰ Generally operating frequency for road input ranges from 0.5 Hz to 70 Hz, aim during optimization is to shift modal frequencies of trim beyond operating frequency or towards the higher region. Road input is highly distributed in lower frequency range and its density reduces towards the higher region, hence target is set to shift lower range modal frequencies in higher range without disturbing other parameters such as structural integrity, space, ergonomics, etc. ⁴ Figure 2 shows the result of shape optimization performed for trim using Altair Optistruct. Objectives set for optimization are to maximize the modal frequency and to reduce mass. Figure 2 shows the results for optimization performed. Left hand side of image shows the thickness of trim at particular instance, initially trim have a throughout 6 mm constant thickness which is now reduced up to 3.1 mm during six iterations of optimization. Figure 3 shows the change in frequency and mass of trim during each iteration.OPEN IN VIEWER

OPEN IN VIEWER

Figure 3.

Iterative returns of optimization.

Frequency response analysis

Frequency response analysis is performed to get displacement values at area of interest. Frequency response analysis is used to calculate structural response of system in range of frequencies. Critical frequencies are carried out by comparing given input and output at specified area of component against the transmissibility. Generalized equation of motion is defined as

M x ¨ + C x ˙ + K x = f e i w t

To perform frequency response analysis sine sweep signal is provided as input to component. During the sine sweep, sine wave of desired frequency range is used as signal, here amplitude of all the input frequencies is kept constant and output displacement is monitored. Another way of analysing sine sweep is to compare input displacement against the observed displacement at desired location.

For a sine sweep testing, 1g constant amplitude is provided against a sine wave from 0.5 Hz to 70 Hz. In Figure 4(a) Frequency response analysis is performed on base design. Maximum displacement of around 0.98 mm is observed. During this process frequency and displacement along rattle line are closely monitored. Rattle line here is considered conidered from left upper corner to right lower corner of speaker assembly as response for displacement is high in this region. The frequencies at which displacement is more than desired displacement are marked as critical frequencies. During the next phase of optimization these frequencies are isolated and new shape is generated. During this process lower frequencies are also shifted to higher frequencies since with increasing frequency amplitude decreases hence priority is given to lower excitation frequency. After optimization final shape is come in picture which gives displacement around 0.25 mm.OPEN IN VIEWER

Figure 5 shows the relative displacement observed along rattle line. After shifting lower frequencies in higher range relative displacement is lowered as these low frequencies don’t get excited during the operation. Before finalizing geometry after optimization frequency response analysis is reperformed on multiple versions of optimized shapes. Design which is more suitable for tooling and with lowest possible tooling cost is finalize.OPEN IN VIEWER

Figure 6 shows the final geometry of door trim. For the ease of tooling, thickness of trim is divided in four parts as 2 mm, 3 mm, 4 mm and 6 mm. It is also important that to cover maximum allowable stresses, fatigue life of component should not hamper, during optimization constraints for these factors were already provided. After finalizing the geometry structural analysis and fatigue analysis is performed on final geometry and safety of component is assured. After optimization one more frequency response analysis is performed and all other possible rattle lines were also cross checked to confirm that there is less possibility of Rattle event after changing the geometry of component.OPEN IN VIEWER

Transient response analysis

To validate the results transient response analysis or random response analysis is performed. During transient response analysis acceleration versus time or displacement versus time data is provided as input to the system. Response of the same is observed at point of interest or along the rattle line. Figure 7 shows the response to the road excitation on the earlier geometry. In earlier response it is observed that at multiple times allowable gap is crossed by component. The assumption here is that if component crosses the allowable gap it is impacting on another assembly making a rattle noise. Figure 7 shows the response to road input data. Graph is the plotted for the same node which shows max displacement during the sine sweep test. As the allowable gap during the assemblies was predetermined as 0.5 mm, 0.45 mm line is marked on graph to observe how many times during the input gap is crossed. First graph shows the displacement response to the input provided for original trim, here allowable gap is crossed at multiple instances. Second graph shows the response to optimized trim for same road input and at same node. Second graph shows reduced instances for crossing allowable gap. At some instances of time, component is crossing the allowable gap but to classify it as rattle there should be at least one hit in one cycle. As instances of gap crossing are reduced, one can classify that optimized design reduces instances of rattle.OPEN IN VIEWER

Conclusion

The proposed optimization methodology helps to reduce the rattle issue during the product design and development phase. This methodology uses an approach in shifting frequency with the help of the CAE optimization tool. In this paper earlier design with constant 6 mm thick trim is converted into variable thickness trim. Care is also taken during the process, weight is reduced and there is no effect on safety and fatigue life.

The new methodology will be helpful to develop parts which are least exposed to rattle issues in vehicles. Similar methodology can be used to reduce squeak as well only additional calculations of energy storage and the frictional coefficient is required instead of the GD&T gap method is used in this method. In this method shape optimization of components is used which allowed reducing the thickness even up to 66% of the original thickness. For BIW components or any sheet metal component, it is suggested to use topography optimization if a change in thickness is not possible.