LMS consulting engineers used a combination of signature and running modes analysis to diagnose and solve a noise problem in a windshield wiper assembly in far less time than would have been required using conventional methods. In the past, the automotive supplier would have used trial and error methods to try and find a design that reduced noise to acceptable levels. In this case, LMS used signature analysis and identified a single frequency that was critical at all motor speeds, indicating that a structural mode was the root cause. They instrumented the motor with accelerometers and identified the critical mode as an ovalization of the motor housing. With the root cause identified, engineers suggested a modification of the housing that raised the frequency of that mode high enough that it was no longer excited during normal operation.A small problem with large consequences
Often noise can easily be traced to small components such as actuators or pumps or larger systems such as powertrains. Suppliers used to be able to ignore most of these noise problems because they were drowned out by engine and road noise. In recent years, the interior noise levels of the vehicle have been substantially reduced, so this type of noise is no longer acceptable to most OEMs. Suppliers are being forced to solve these problems or run the risk of creating serious negative quality perceptions.
Trial and error approach takes too much time
Once the problem is localized, the traditional approach involves selecting a component on the assembly generating the noise. Then simple modifications of1the component are tested to evaluate its sensitivity on the noise. Shields are placed around the assembly while measurements are made to compare noise levels to the original design. Typically, the first fix that is attempted doesn’t solve the problem so engineers keep trying many approaches until they find one that reduces the noise. One problem with this approach is that prototyping and testing various alternatives are expensive and take considerable amounts of time. This approach also provides little diagnostic information so engineers usually have to rely upon guesswork and intuition in identifying possible design changes.
Identifying the root cause
LMS engineers recognized that the key to solving the problem was understanding the dynamics of the noise source. They began by performing limited measurements while varying the rotational speeds and the weight on the wiper. Once they established the general parameters of the problem, they performed more detailed measurements, calculating the frequency spectrum of noise at different rotational speeds. LMS engineers analyzed the amplitude of the noise versus engine rpm and noise frequency. The results showed peaks at different orders of the engine rotational speed. More significantly, the troublesome 3 kHz frequency was seen to be excited at any speed, providing a clue that it was caused by a structural resonance.
Engineers still had no way of knowing at this point whether the motor housing or gearbox housing or some other component was the primary noise generator. So they instrumented the engine and gearbox with accelerometers and took measurements while running it through its full operating range. They carried out a Running Modes Analysis to analyze the structural deformations at the frequency causing the noise problem. They were able to clearly see that the noise was related to an ovalization mode of the engine housing.
Solving the problem
Once they identified the root cause, LMS engineers were able to solve the problem relatively quickly. They made simple modifications that stiffened the base of the engine housing and determined that these changes eliminated the noise problem. They worked with the supplier’s design and manufacturing engineers to discuss the other requirements for the engine housing and slightly modified their design to meet the additional requirements. Additional tests were performed to verify that the new design provided acceptable noise levels and the final design was implemented.
