As NVH gains importance in the quality of off-highway machines’ performance and operator comfort, it is essential to understand every aspect of machine noise and then reduce the noise to a level that does not affect the operator’s comfort and performance. Booming noise, a low-frequency NVH phenomenon below 200 Hz, defined as continuous bass drum roll, distant thunder sound, or a deep resonant sound like an explosion, is one of the concerns of off-highway machines.
Motor graders are commonly used in the construction and maintenance of dirt and gravel roads. They are used for the preparation of flat base course (on which asphalt is laid) in the construction of gravel roads and in the removal of snow/ice from roads for maintenance. The V-pick located in front of the blade is often used to break any hard objects in front of the blade. The contacts between the blade/V-pick and the road generate significant impact forces, transmitting considerable low-frequency energy into the cab and thereby generating booming noise.
Due to the low-frequency characteristics of the booming noise, it is difficult to absorb the energy once it is generated in the machine cab. The best way to deal with booming noise is to prevent the noise from being radiated.
To improve the NVH performance of a motor grader, a series of nonstandard NVH tests were carried out in field conditions to understand the noise characteristics in the cab. The machine running conditions included traveling on a dirt road (roading) and working on a gravel road with the blade and V-pick down (blading). Subjective evaluations from machine operators confirmed the presence of detectable and annoying booming/rumbling noises.
As any NVH problem, this booming noise can be represented by a source-path-receiver model as illustrated: Y(f)=H1(f) x F1(f) (1), where Y(f) stands for the cab sound response, H1(f) is the transfer function between the sound pressure level (SPL) acquisition locations and the machine blade/V-pick, and F1(f) is the force generated from the blading process.
A modification to any of the source, path, or receiver elements of the model can provide a solution to the NVH problem. However, physical constraints and the product environment limit the possibilities to reduce the sound levels and achieve the NVH target. The excessive blading forces between the blade/V-pick and the ground can be reduced by decreasing the machine speed but would considerably decrease the productivity of machines. Due to high productivity demands on the machine, it seemed more feasible to modify the energy flow path between the machine blade/V-pick and the operator ear.
Modifications to the machine frame and cab mounting system to mitigate cab booming noise would require extensive further structural validation of the proposed changes, thereby increasing project time and cost. In consideration of this, the response Y(f) can be represented in another source-path-receiver model: Y(f)=H2(f) x F2(f) (2), where H2(f) is the vibra-acoustic characteristics of the cab structures and F2(f) is the force input to the cab at the cab mount interface.
Now that the force input to the cab is assumed to be fixed, the only possible solution to decrease the booming noise would be some structural changes that can lower the cab sound radiation efficiency.
To verify if the cab structure is the booming noise root cause and can be modified to reduce or eliminate it, an experimental cab modal analysis was carried out. Countermeasures were then recommended from the test results. A cab structural/acoustic simulation was also performed to complement the test results and predict the effectiveness of the modifications.
Experimental modal analysis (EMA) is the process of determining the natural frequencies and associated damping and mode shapes of a structure from a set of frequency response functions (FRF) measured from the structure. The purpose of an EMA on the motor grader cab was to determine if the cab panels had natural frequencies close to the booming noise frequencies. The panels are the prime candidates for modifications if they have natural frequencies close to the booming noise frequencies and are the potentials to be treated to reduce/eliminate the booming noise.
A few modes were found to be at the very vicinity of the booming noise frequencies, which indicates that these modes are the potential root causes of the booming noise. The cab mode at 39 Hz is the first flexible mode of the roof. The cab mode at 40.5 Hz indicates a breathing mode of the side windows as well as between the front and rear windows with some residues of the roof mode at 39 Hz. Modes around 100 Hz indicate some disengagement between the two roof layers and the flexibility of the cab windows.
Besides the EMA, a preliminary cab acoustical modal test was conducted, as well as an FEA prediction of the cab acoustic modes. During the preliminary acoustical modal test, the SPL at various locations in the cab was measured with a broadband sound source placed inside and then outside the cab. The SPLs at different locations were then compared to identify the peak frequencies. The preliminary acoustical modal test showed cab acoustic mode between 100 and 110 Hz, which align with the cab panel structural modes and engine firing frequencies.
Based on the results of EMA, acoustical test, and FEA, it was determined that the booming noise is generated from a combination of the cab acoustic mode, cab panel breathing modes, disengagement of the cab roof layers, excitations from the engine, and the blade/V-pick blading force when the machine is at work. With the knowledge that the cab cavity acoustic modes do not change unless the cab cavity is modified, it is logical to consider a solution that results in stiffer panels with higher damping due to the limitations of the changes to the blading force and mounting system.
The modal analysis results suggest thicker and/or laminated window glasses were potential solutions to reduce or eliminate the booming noise. Before new thicker glass was installed, a simple test with window mass dampers was done to confirm the solution would be effective. In this test, a few mass dampers (steel plates) were attached to each window. It was expected that the mass dampers would effectively shift the 40 Hz booming noise to a lower frequency if those window panels were the root cause of the booming noise.
FEA based structural modal analysis was also performed on the cab. The 39 Hz roof panel mode matches well with the modes at 39 to 40 Hz from modal test. One easy way to stiffen the roof and engage the two layers of the roof panels is to add more welding locations. FEA results show that additional welds stiffened the roof and increased the first roof mode frequency from 38 to 44 Hz.
Based on the experimental and FE modal analysis and preliminary countermeasure test and FE analysis, a recommendation to change the side and rear window glasses to thicker laminated glasses and increase welding points on the roof to fully bond the two roof layers was proposed. The simple and low-cost changes on the cab structure eliminated the booming noise and lowered the sound pressure levels by 8 to 10 dB at the booming frequencies. The improvement was also confirmed by subjective operator evaluation.
This article is based on SAE Technical Paper 2011-01-1729 by Jiantie Zhen, Chunhui Pan, Ashish Jangale, and Brad Salisbury of Caterpillar Inc.