Proposed automobile steering wheel test method for vibration

Abstract

This thesis proposes a test method for evaluating the perceived vibration which occurs at the driver's hand in automotive steering wheel interface. The objective of the research was to develop frequency weightings for quantifying the human perception of steering wheel hand-arm vibration. Family of frequency weightings were developed from equal sensation curves obtained from the psychophysical laboratory experimental tests. The previous literature suggests that the only internationally standardised frequency weighting Wh is not accurate to predict human perception of steering wheel hand-arm vibration (Amman et. al, 2005) because Wh was developed originally for health effects, not for the human perception. In addition, most of the data in hand-arm vibration are based upon responses from male subjects (Neely and Burström, 2006) and previous studies based only on sinusoidal stimuli. Further, it has been continuously suggested by researchers (Gnanasekarna et al., 2006; Morioka and Griffin, 2006; Ajovalasit and Giacomin, 2009) that only one weighting is not optimal to estimate the human perception at all vibrational magnitudes. In order to address these problems, the investigation of the effect of gender, body mass and the signal type on the equal sensation curves has been performed by means of psychophysical laboratory experimental tests. The test participants were seated on a steering wheel simulator which consists of a rigid frame, a rigid steering wheel, an automobile seat, an electrodynamic shaker unit, a power amplifier and a signal generator. The category-ratio Borg CR10 scale procedure was used to quantify the perceived vibration intensity. A same test protocol was used for each test and for each test subject. The first experiment was conducted to investigate the effect of gender using sinusoidal vibration with 40 test participants (20 males and 20 females). The results suggested that the male participants provided generally lower subjective ratings than the female participants. The second experiment was conducted using band-limited random vibration to investigate the effect of signal type between sinusoidal and band-limited random vibration with 30 test participants (15 males and 15 females). The results suggested that the equal sensation curves obtained using random vibration were generally steeper and deeper in the shape of the curves than those obtained using sinusoidal vibration. These differences may be due to the characteristics of random vibration which produce generally higher crest factors than sinusoidal vibration. The third experiment was conducted to investigate the effect of physical body mass with 40 test participants (20 light and 20 heavy participants) using sinusoidal vibration. The results suggested that the light participants produced generally higher subjective ratings than the heavy participants. From the results it can be suggested that the equal sensation curves for steering wheel rotational vibration differ mainly due to differences of body size rather than differences of gender. The final experiments was conducted using real road signals to quantify the human subjective response to representative driving condition and to use the results to define the selection method for choosing the adequate frequency weightings for the road signals by means of correlation analysis. The final experiment was performed with 40 test participants (20 light and 20 heavy participants) using 21 real road signals obtained from the road tests. From the results the hypothesis was established that different amplitude groups may require different frequency weightings. Three amplitude groups were defined and the frequency weightings were selected for each amplitude group. The following findings can be drawn from the research: • the equal sensation curves suggest a nonlinear dependency on both the frequency and the amplitude. • the subjective responses obtained from band-limited random stimuli were steeper and the deeper in the shape of the equal sensation curves than those obtained using sinusoidal vibration stimuli. • females provided higher perceived intensity values than the males for the same physical stimulus at most frequencies. • light test participants provided higher perceived intensity than the heavy test participants for the same physical stimulus at most frequencies. • the equal sensation curves for steering wheel rotational vibration differ mainly due to differences in body size, rather than differences of gender. • at least three frequency weightings may be necessary to estimate the subjective intensity for road surface stimuli.