Abstract
Squeal noise is an on-going problem of the automotive industry and it has been researched via various methods. The approach of squeal noise through CAE-based Complex Eigenvalue Analysis (CEA) is commonly used in the current industry and is being utilized to derive prompt counter-measures. However, the results of a one-shot analysis based on a deterministic model have limitations that only show an average representation of various uncertainties. These restrictions can lead to distrust of CEA results such as deriving the counter-measure study.
Recently, various researches based on CEA have been continuously carried out, such as studies considering the scattering of uneven contact surface between the pad and the disc, random sampling that reflects variations in material properties and operating conditions as a probability density function(1)~(3). However, these studies are only used for the verification and the robustness of the squeal noise in the preceding stages. It takes a considerable amount of time in the analysis process to cope with the squeal noise issue that occurs in the development stage, and it also limited to utilize the results.
Therefore, in this research, when deriving counter-measures through CEA, a study on statistical approach through DOE (Design of Experiments)- based sampling analysis is conducted in order to overcome the limit of deterministic model. After deriving some counter-measure cases through a one-shot analysis, the validity of them is examined by analysing the improvement effect statistically with setting up a total of 972 full matrix combinations such as FRF distribution of calliper housing / bracket / disc / knuckle at each 3 levels, disc rotating speeds when braking at 3 levels and braking pressure at 4 levels.
A specific vehicle with field claim of squeal noise is selected. Due to the characteristics of the squeal noise being sensitive to environmental and braking condition, highly reproducible evaluation conditions are required. This is to compare quantitatively the degree of improvement and alteration of CEA-based countermeasures with the verification test results. As a result of comparing the improvement tendency of the statistical analysis and the test, it is confirmed that the statistical verification method is effective. In addition, the analytical statistic and the test results are examined by linear regression analysis and it is shown that there is a high correlation. Therefore, the statistical approach of CEA for squeal noise is validated clearly.
Based on the results of these studies, it is possible to suggest an analytically effective solution and it is expected that the trial and error can be greatly reduced. But it is necessary to review the efficiency of the work for the excessive analysis time.
KEYWORDS Squeal noise, Complex Eigenvalue analysis, DOE, Sampling analysis