- Email: [email protected]
Reduced ride comfort caused by beating idle vibrations in passenger vehicles
International Journal of Industrial Ergonomics 57 (2017) 74e79
Contents lists available at ScienceDirect
International Journal of Industrial Ergonomics journal homepage: www.elsevier.com/locate/ergon
Reduced ride comfort caused by beating idle vibrations in passenger vehicles Jinhan Park a, Junwoo Lee a, Sejin Ahn b, *, Weuibong Jeong a a b
Department of Mechanical Engineering, Pusan National University, Jangjoen-dong, Kumjung-ku, Pusan 609-735, Republic of Korea Division of Energy & Electrical Engineering, Uiduk University, Gyeongju, 780-713, Republic of Korea
a r t i c l e i n f o
a b s t r a c t
Article history: Received 14 December 2015 Received in revised form 21 November 2016 Accepted 10 December 2016
Idle vibrations that occur in passenger vehicles at a stop cause passengers to feel discomfort, thus leading to reduced ride comfort. Beating vibration takes place when the frequency of engine vibrations is similar to the frequency of vibrations due to auxiliary parts in the engine room, and this produces a rapid deterioration in ride comfort. This study analyzed beating idle vibrations in an actual vehicle, and performed experiments to assess changes in ride comfort in relation to the amplitude ratio of two vibrations having similar frequencies and their beating frequency. At most frequencies, ride comfort was poorer when the beating vibrations had higher amplitude ratios. A higher level of discomfort was found when the beating frequency fell in a range of 1e4 Hz. Beating vibrations with an amplitude ratio of 0.8 and a beating frequency of 1.0 and 2.0 Hz caused greater discomfort by 3.8 dB compared to single-frequency sinusoidal vibrations. © 2016 Elsevier B.V. All rights reserved.
Keywords: Idle vibration Beating vibration Ride comfort Passenger vehicle Electro-magnetic exciter Cooling fan
1. Introduction The engine RPM of a vehicle at a stop must be set to optimize vehicle vibration and fuel efficiency while ensuring stable operation of the engine. Idle vibrations refer to engine vibrations occurring when the vehicle comes to a stop, such as when waiting for a signal change. The frequency of idle vibrations is determined by the engine type and engine RPM when idling. For general passenger vehicles, the frequency of idle vibrations falls in a range of 20e40 Hz. (Iyer et al., 2011). The cooling fan of the radiator is triggered when the heated engine exceeds a certain temperature, and vibrations result from an imbalance of blades as the fan rotates. In general, the number of rotations of the cooling fan is determined in consideration of blade size and shape, which are designed to meet the requirements for engine cooling performance. Blade Pulsation Frequency (BPF), the main source of vibrations of the cooling fan, is determined by the number of fan rotations and the number of blades. The vibration frequency of the cooling fan is often inevitably designed to have a similar range as that of idle vibrations due to the engine. The two vibration types, with a slight difference in frequency, cause a
* Corresponding author. E-mail address: [email protected] (S. Ahn). http://dx.doi.org/10.1016/j.ergon.2016.12.003 0169-8141/© 2016 Elsevier B.V. All rights reserved.
regular variation in amplitude known as beating. (Rao and Yap, 1995). The frequency range of vibrations associated with ride comfort in a seat with a seat back is between 0.8 Hz and 10 Hz. (Griffin, 1990; ISO, 1997). ISO 2631-1 assigns a high weighting(Wk) to a range of 4e10 Hz for top-bottom vibrations of the seat surface that comes into contact with the buttocks, and a high frequency weighting (Wd) to a range of 0.8e1.6 Hz for front-back and left-right vibrations. For the seat back that comes into contact with the back, a high weighting (Wd) is assigned to a range of 0.8e1.6 Hz for left-right and topbottom vibrations, while a high weighting (Wc) is assigned to a range of 1e5 Hz for front-back vibrations. These frequency weightings should be applied with flexibility depending on the seat shape, seat material, and properties of the input vibrations. Basri and Griffin (2014) examined the relationship between SEAT(Seat Effective Amplitude Transmissibility) value and discomfort when seated subjects were exposed to white noise vibrations in a range of 0e50 Hz. This study, performed on various seat types, revealed slight differences in SEAT value by frequency, but the highest SEAT value was observed for most sheets at 2.5e6 Hz. Thuong and Griffin (2011) studied the discomfort experienced by subjects when exposed to vibratory forces of 0.5e16 Hz in various directions, and showed that both seated and standing subjects felt discomfort in similar frequency ranges. For vibratory forces in the longitudinal direction, the
J. Park et al. / International Journal of Industrial Ergonomics 57 (2017) 74e79
75
discomfort was proportionate to velocity in a frequency range of 0.5e3.15 Hz, and to acceleration in a range of 3.15e16 Hz. Langer et al. (2015) experimentally analyzed the occupational whole-body vibration of an agricultural tractor and found that the condition of driving downhill with 4WD influences the longitudinal dynamics and intensifies the whole body vibration exposure on the tractor. Sekuli c et al. (2016) found that the middle part of an intercity bus is more comfortable than its front or rear part in terms of vibration exposure. In the present study, the authors hypothesize that the level of discomfort caused by idle vibrations increases rapidly when the frequency of beating vibrations coincides with the frequency that the human body is sensitive to. Here, beating vibrations are vibrations that occur when idle vibration of the engine overlaps with vibrations of the cooling fan or other auxiliary parts. The purpose of this study is to conduct a quantitative evaluation of how changes in frequency and amplitude of beating vibrations affect ride comfort in a passenger vehicle. Beating idle vibrations were measured from an actual vehicle, and these measurements were used to generate various beating vibration signals. The beating vibration signals were entered into an exciter having the same seat as that of the vehicle used in measurements, and nine seated subjects were asked to provide a subjective evaluation. 2. Beating idle vibrations 2.1. Idle vibrations from an actual vehicle This study measured and analyzed beating idle vibrations occurring in an actual passenger vehicle to ensure the effectiveness of the evaluation. Fig. 1 shows measurements of idle vibrations from the driver's seat track of a mid-sized sedan (Mileage: 200 km, Engine size: 2,000 cc) available in Korea and overseas. Comparisons are made between the idle vibrations before and after operating the cooling fan. The gear was in drive, and the vehicle was stopped for a certain period of time to heat up the engine, which then triggered the cooling fan. From the time-frequency map of Fig. 1 (a), we can see that the cooling fan operates at about 13.5 s, and vibrations (23.1 Hz) caused by fan imbalance are added to idle vibrations (21.7 Hz) of the engine. Beating is also observed in Fig. 1(b) and (c), which show the spectrum of acceleration and time signals before and after operation of the cooling fan. In Fig. 1(b), when the cooling fan was not in operation, a vibration of 98.6 dB was measured at a single frequency. When the cooling fan was in operation, vibrations of 99.5 dB and 97.1 dB were measured at two frequencies. Fig. 1(c), which shows time signals of vibrations measured from the test vehicle, gives an RMS of 101.2 dB when the cooling fan is not in operation and 103.0 dB when the cooling fan is in operation. Vibrations measured before and after operation of the cooling fan differ by 1.8 dB, falling below the human-perceivable level of 3 dB. However, some experts and general drivers participating in a pilot test prior to this study apparently experienced beating vibrations after the cooling fan was triggered, and reported a significant reduction in ride comfort. These findings highlighted the need for research on reduced ride comfort in relation to beating idle vibrations arising from operation of the cooling fan. 2.2. Generation of beating vibration signals This study generated beating signals to evaluate the extent to which ride comfort is reduced in relation to properties of beating vibrations. A beating signal is created when two sinusoidal signals having similar frequencies occur simultaneously, as shown in Eq. (1)
Fig. 1. Comparison of idle vibration measured on seat track of a passenger car with and without beating by cooling fan.
76
J. Park et al. / International Journal of Industrial Ergonomics 57 (2017) 74e79
aðtÞ ¼ A cosð2pftÞ þ B cosð2pðf þ fb Þt þ fÞ
(1)
here, fb is the beating frequency, or the difference in frequency between the two sinusoidal signals. f is the basic frequency of the sinusoidal signal, and 4 is the phase difference between the two signals. A and B represent the amplitude of each sinusoidal signal. A generated beating signal, as delineated in Eq. (1), creates vibrations with a beating frequency denoted by fb. For subjective evaluation of reduced ride comfort caused by beating vibrations, this study used the beating frequency fb and amplitude ratio B/A as control variables. The phase difference 4 of the two sinusoidal signals was set as 0. Fig. 2 shows the time signals in relation to the amplitude ratio of the beating signals (B/A) and beating frequency (fb). Fig. 3 is a comparison of the spectrum of the beating vibrations measured from an actual vehicle and the spectrum of beating signals created from Eq. (1). The generated beating signals were designed to have an f of 21.7 Hz, fb of 2 Hz, and B/A of 0.4 to match the frequency and size of the primary and secondary peaks of beating measurements. In the measured idle vibrations, vibrations of minor peaks over 25 Hz were considered negligible since they cannot be detected by the driver. A pilot test was conducted to determine whether the subjects were able to differentiate between the two signal types compared
Fig. 3. Spectrum of beating signal produced for imitation of the idle vibration measured on a passenger car (Primary peak frequency: 21.7 Hz, Secondary peak frequency: 19.7 Hz).
in Fig. 3. When seated in the car seat mounted on the exciter alternating between idle vibrations in the vehicle and selfgenerated beating vibrations, all the nine subjects were unable to differentiate between the two. For a subjective evaluation, the beating vibration signals were designed to have an fb of 0.5, 1, 2, 4, and 8 Hz for a B/A of 0.4 and 0.8, amounting to a total of 10 cases. The overall value of RMS from all beating vibrations was 101.2 dB, coinciding with that of actual vibrations.
3. Experimental method 3.1. Equipment A schematic diagram of the exciter system, which generates various beating vibrations, is shown in Fig. 4. The i-220 electric exciter developed by IMV Corporation was used to produce beating vibrations for the subjects. Table 2 shows the mechanical specifications of the vibration shaker (IMV-i220) used in the experiment. The various beating vibration signals are entered into the Ni-PXI
Fig. 2. Comparison of beating signals with different magnitude ratio and beating frequency.
Fig. 4. Excite system for generating idle vibration.
J. Park et al. / International Journal of Industrial Ergonomics 57 (2017) 74e79
77
Table 1 Physical data of subject employed in pilot test. Item
Mean
S.D.
Max/Min
Weight(Kg) Height(cm) Age(yr) Driving(yr)
69.0 174.1 26.4 1.6
8.1 4.7 1.9 0.8
85/60 180/167 31/25 3/0.5
Table 2 Specifications of vibration shaker. Model
IMV i-220
Type Rated force Frequency range Maximum displacement Maximum payload
Electro-dynamic 5.6 kN 3e3300 Hz 51 mm p-p 200 kg
system through Labview, and applied to the exciter via an amplifier following D/A conversion. Vibration signals experienced by the subjects were monitored in real-time by measuring the acceleration at the top of the exciter platform, and the amplitude of vibrations was adjusted using control signals in Labview and the amp gain value. While a phase difference between seat and floor vibrations led to slight differences in evaluations of ride comfort (Jang and Griffin, 1999), this study consistently excluded floor vibrations by separating the floor from the exciter platform.
Fig. 5. Seat setup and seating posture of subject.
3.2. Subjects and sitting posture For a subjective evaluation of ride comfort in relation to beating vibrations, this study relied on nine male subjects without spinal injuries. Table 1 shows the height, weight, age, and driving experience of the subjects. To simulate actual driving conditions, the exciter platform was mounted with the same fabric seat as the vehicle from which measurements were taken. The transmissibility of fabric seat generally has the main peak at around 5 Hz, whereas difference of the values over 20 Hz is quite little with a magnitude ranged from 0.2 to 0.5 (Van Niekerk, 2003; Griffin, 1990). The contact pressure at the lumbar region was different according to the geometry of the seat, and this geometry influenced the ride comfort. (Guo et al., 2016). The seat cushion and seat back angle were adjusted to 103 . As shown in Fig. 5, the seat height was adjusted to allow subjects to comfortably place their feet while maintaining a knee angle of 120 . The subjects were asked to look forward with their hands resting comfortably on their knees. They were given hearing protectors to minimize the effect of surrounding noise on their assessment of ride comfort. 3.3. Method of subjective evaluation This study applied the three-down one-up method to identify test signals of single-frequency sinusoidal vibrations having the same discomfort level of 101.2 dB as reference signals of beating vibrations. As shown in Fig. 6, the subjects were asked to evaluate the level of discomfort after being exposed to 20 s of beating reference signals and 20 s of sinusoidal test signals, with a stop time of 5 s between them. To eliminate bias, the order of the reference signals and test signals was randomized each time. One second of fading was given to the start and end of each signal to prevent the results from being influenced by shock vibrations. As mentioned earlier, the reference signals were beating vibration signals created
Fig. 6. Magnitude estimation method of test and reference signal randomized in order.
by combining two amplitude ratios and five beating frequencies. All reference signals were 101.2 dB (ref.:106 m/s2), corresponding with idle vibrations measured from the actual vehicle. The test signals were sinusoidal vibration signals having a single frequency, and 10 steps were prepared from 90 dB to 110 dB at 2 dB intervals in magnitudes. Fig. 7 shows the flow of obtaining equivalent discomfort based on the three-down one-up method, first proposed by Dixon and Mood (1948). This can be used to eliminate the effect of nerve adaptation under continuous stimuli. By repeating the steps presented in Fig. 7, the test continues until three downs and three ups occur, as shown in Fig. 8. The last four values are averaged, and this becomes the amplitude of equivalent discomfort of the single-frequency sinusoidal vibration to the beating. Mansfield and Griffin (2000) applied the three-down one-up method to difference thresholds, which were used to differentiate between the amplitudes of the two signals delivered via a car seat.
78
J. Park et al. / International Journal of Industrial Ergonomics 57 (2017) 74e79
Fig. 7. Flow chart of three-down one-up method.
Fig. 9. Magnitude of sinusoidal vibration with equivalent discomfort to beating vibration of 101.2 dB.
Fig. 8. Example of three-down one-up method.
Lee et al. (2012) used the same method in his study of objective human body reactions to vibrations and subjective discomfort in a passenger vehicle at a stop. 4. Experimental results and discussion Fig. 9 shows the amplitudes of single-frequency sinusoidal vibrations corresponding to the ten beating vibrations of 101.2 dB obtained by applying the three-down one-up method for the nine subjects, along with their median values. For all beating frequencies, beating vibrations induced greater discomfort than single-frequency sinusoidal vibrations. (p < 0.0001, Wilcoxon signed rank test). Differences in level of discomfort were found when the two vibrations had different beating frequencies. (p ¼ 0.047, Friedman test). A statistical analysis showed that changes in discomfort became more prominent at smaller
amplitude ratios. (B/A ¼ 0.4: p ¼ 0.0013, B/A ¼ 0.8: p ¼ 0.043, Friedman test). When the amplitude ratio (B/A) was 0.4, the greatest difference in equivalent discomfort was 3.2 dB at a beating frequency of 4.0 Hz. When the amplitude ratio was 0.8, the greatest difference was 3.8 dB at a beating frequency of 1.0 and 2.0 Hz. For the two amplitude ratios, the difference in equivalent discomfort was relatively small at a beating frequency of 0.5 Hz and 8.0 Hz. For the same beating frequency, more discomfort was caused by a larger amplitude ratio. (p < 0.001, Wilcoxon signed rank test). While ISO 2631-1 associates frequency weighting of top-bottom vibration in the seated position with greater discomfort at 4e10 Hz, beating vibration in this study tends to cause greater discomfort at relatively lower frequencies. Morioka and Griffin (2006) examined how the human body perceives vibrations in the tri-axial translational direction in a frequency range of 2e315 Hz, and found a greater sensitivity to vibrations less than 2e10 Hz. Ahn and Griffin (2008), Ahn (2010) studied equivalent discomfort and body response to shock signals in a range of 0.5e16 Hz, and reported
J. Park et al. / International Journal of Industrial Ergonomics 57 (2017) 74e79
greater discomfort at low frequencies below 1 Hz and resonant frequencies of 4e10 Hz. Discomfort level and body response had non-linear characteristics in relation to the amplitude of shock vibrations. The results of this study, which evaluated the discomfort caused by beating vibrations under actual seat conditions, were different from past research involving rigid body seats and single-frequency sinusoidal vibrations. This study found that the level of discomfort varied with the amplitude ratio of beating vibrations and beating frequency. As future work, rigid seats can be used to evaluate the discomfort caused by beating vibrations while eliminating the influence of seat characteristics. 5. Conclusion This study analyzed the deterioration in ride comfort caused by beating vibrations, which are generated when the frequency of idle vibrations is similar to the frequency of vibrations due to the cooling fan and other auxiliary parts in engine room. Various beating vibration signals were created by reflecting the amplitude of idle vibrations measured from an actual vehicle, and they were entered into an exciter. Using the subjective evaluations of 9 subjects and the three-down one-up method, the amplitudes of single-frequency sinusoidal vibration which is equivalent to a beating vibration were derived. Beating vibrations were associated with greater discomfort even when they were of the same amplitude as single-frequency sinusoidal idle vibrations. For idle vibrations with the same beating frequency, greater discomfort was induced at a higher amplitude ratio (B/A). Acknowledgement We would like to express our thanks to Renault-Samsung Motor Company which assisted to carry out the experimental study.
79
References Ahn, S.J., 2010. Discomfort of vertical whole-body shock-type vibration in the frequency range of 0.5 to 16 Hz. Int. J. Automot. Technol. 11 (6), 909e916. Ahn, S.J., Griffin, M.J., 2008. Effects of frequency, magnitude, damping, and direction on the discomfort of vertical whole-body mechanical shocks. J. Sound Vib. 311 (1), 485e497. Basri, B., Griffin, M.J., 2014. The application of SEAT values for predicting how compliant seats with backrests influence vibration discomfort. Appl. Ergon. 45 (6), 1461e1474. Dixon, W.J., Mood, A.M., 1948. A method for obtaining and analyzing sensitivity data. J. Am. Stat. Assoc. 43 (241), 109e126. Griffin, M.J., 1990. Handbook of Human. Academic Press, London. Guo, L.X., Dong, R.C., Zhang, M., 2016. Effect of lumbar support on seating comfort predicted by a whole human body-seat model. Int. J. Ind. Erg. 53, 319e327. ISO, I., 1997. Mechanical Vibration and Shockeevaluation of Human Exposure to Whole-body Vibrations (2631-1). International Standards Organization. Iyer, G., Prasanth, B., Wagh, S., Hudson, D., 2011. Idle Vibrations Refinement of a Passenger Car (No. 2011-26-0069). SAE (Technical Paper). Jang, H.K., Griffin, J.M., 1999. The effect of phase of differential vertical vibration at the seat and feet on discomfort. J. Sound Vib. 223 (5), 785e794. Langer, T.H., Ebbesen, M.K., Kordestani, A., 2015. Experimental analysis of occupational whole-body vibration exposure of agricultural tractor with large square baler. Int. J. Ind. Erg. 47, 79e83. Lee, J.Y., Jeon, G.J., Ahn, S.J., Jeong, W.B., 2012. The study of correlation between objective human response and subjective discomfort evaluation of idle vibration on passenger vehicle. Trans. Korean Soc. Noise Vib. Eng. 22 (5), 422e428. Mansfield, N.J., Griffin, M.J., 2000. Difference thresholds for automobile seat vibration. Appl. Ergon. 31 (3), 255e261. Morioka, M., Griffin, M.J., 2006. Magnitude-dependence of equivalent comfort contours for fore-and-aft, lateral and vertical whole-body vibration. J. Sound Vib. 298 (3), 755e772. Rao, S.S., Yap, F.F., 1995. In: Mechanical Vibrations, vol. 4. Addison-Wesley, New York. Sekuli c, D., Dedovi c, V., Rusov, S., Obradovi c, A., Salini c, S., 2016. Definition and determination of the bus oscillatory comfort zones. Int. J. Ind. Ergon. 53, 328e339. Thuong, O., Griffin, M.J., 2011. The vibration discomfort of standing persons: 0.5e16Hz fore-and-aft, lateral, and vertical vibration. J. Sound Vib. 330 (4), 816e826. Van Niekerk, J.L., 2003. The use of seat effective amplitude transmissibility (SEAT) values to predict dynamic seat comfort. J. sound Vib. 260 (5), 867e888.
Contents lists available at ScienceDirect
International Journal of Industrial Ergonomics journal homepage: www.elsevier.com/locate/ergon
Reduced ride comfort caused by beating idle vibrations in passenger vehicles Jinhan Park a, Junwoo Lee a, Sejin Ahn b, *, Weuibong Jeong a a b
Department of Mechanical Engineering, Pusan National University, Jangjoen-dong, Kumjung-ku, Pusan 609-735, Republic of Korea Division of Energy & Electrical Engineering, Uiduk University, Gyeongju, 780-713, Republic of Korea
a r t i c l e i n f o
a b s t r a c t
Article history: Received 14 December 2015 Received in revised form 21 November 2016 Accepted 10 December 2016
Idle vibrations that occur in passenger vehicles at a stop cause passengers to feel discomfort, thus leading to reduced ride comfort. Beating vibration takes place when the frequency of engine vibrations is similar to the frequency of vibrations due to auxiliary parts in the engine room, and this produces a rapid deterioration in ride comfort. This study analyzed beating idle vibrations in an actual vehicle, and performed experiments to assess changes in ride comfort in relation to the amplitude ratio of two vibrations having similar frequencies and their beating frequency. At most frequencies, ride comfort was poorer when the beating vibrations had higher amplitude ratios. A higher level of discomfort was found when the beating frequency fell in a range of 1e4 Hz. Beating vibrations with an amplitude ratio of 0.8 and a beating frequency of 1.0 and 2.0 Hz caused greater discomfort by 3.8 dB compared to single-frequency sinusoidal vibrations. © 2016 Elsevier B.V. All rights reserved.
Keywords: Idle vibration Beating vibration Ride comfort Passenger vehicle Electro-magnetic exciter Cooling fan
1. Introduction The engine RPM of a vehicle at a stop must be set to optimize vehicle vibration and fuel efficiency while ensuring stable operation of the engine. Idle vibrations refer to engine vibrations occurring when the vehicle comes to a stop, such as when waiting for a signal change. The frequency of idle vibrations is determined by the engine type and engine RPM when idling. For general passenger vehicles, the frequency of idle vibrations falls in a range of 20e40 Hz. (Iyer et al., 2011). The cooling fan of the radiator is triggered when the heated engine exceeds a certain temperature, and vibrations result from an imbalance of blades as the fan rotates. In general, the number of rotations of the cooling fan is determined in consideration of blade size and shape, which are designed to meet the requirements for engine cooling performance. Blade Pulsation Frequency (BPF), the main source of vibrations of the cooling fan, is determined by the number of fan rotations and the number of blades. The vibration frequency of the cooling fan is often inevitably designed to have a similar range as that of idle vibrations due to the engine. The two vibration types, with a slight difference in frequency, cause a
* Corresponding author. E-mail address: [email protected] (S. Ahn). http://dx.doi.org/10.1016/j.ergon.2016.12.003 0169-8141/© 2016 Elsevier B.V. All rights reserved.
regular variation in amplitude known as beating. (Rao and Yap, 1995). The frequency range of vibrations associated with ride comfort in a seat with a seat back is between 0.8 Hz and 10 Hz. (Griffin, 1990; ISO, 1997). ISO 2631-1 assigns a high weighting(Wk) to a range of 4e10 Hz for top-bottom vibrations of the seat surface that comes into contact with the buttocks, and a high frequency weighting (Wd) to a range of 0.8e1.6 Hz for front-back and left-right vibrations. For the seat back that comes into contact with the back, a high weighting (Wd) is assigned to a range of 0.8e1.6 Hz for left-right and topbottom vibrations, while a high weighting (Wc) is assigned to a range of 1e5 Hz for front-back vibrations. These frequency weightings should be applied with flexibility depending on the seat shape, seat material, and properties of the input vibrations. Basri and Griffin (2014) examined the relationship between SEAT(Seat Effective Amplitude Transmissibility) value and discomfort when seated subjects were exposed to white noise vibrations in a range of 0e50 Hz. This study, performed on various seat types, revealed slight differences in SEAT value by frequency, but the highest SEAT value was observed for most sheets at 2.5e6 Hz. Thuong and Griffin (2011) studied the discomfort experienced by subjects when exposed to vibratory forces of 0.5e16 Hz in various directions, and showed that both seated and standing subjects felt discomfort in similar frequency ranges. For vibratory forces in the longitudinal direction, the
J. Park et al. / International Journal of Industrial Ergonomics 57 (2017) 74e79
75
discomfort was proportionate to velocity in a frequency range of 0.5e3.15 Hz, and to acceleration in a range of 3.15e16 Hz. Langer et al. (2015) experimentally analyzed the occupational whole-body vibration of an agricultural tractor and found that the condition of driving downhill with 4WD influences the longitudinal dynamics and intensifies the whole body vibration exposure on the tractor. Sekuli c et al. (2016) found that the middle part of an intercity bus is more comfortable than its front or rear part in terms of vibration exposure. In the present study, the authors hypothesize that the level of discomfort caused by idle vibrations increases rapidly when the frequency of beating vibrations coincides with the frequency that the human body is sensitive to. Here, beating vibrations are vibrations that occur when idle vibration of the engine overlaps with vibrations of the cooling fan or other auxiliary parts. The purpose of this study is to conduct a quantitative evaluation of how changes in frequency and amplitude of beating vibrations affect ride comfort in a passenger vehicle. Beating idle vibrations were measured from an actual vehicle, and these measurements were used to generate various beating vibration signals. The beating vibration signals were entered into an exciter having the same seat as that of the vehicle used in measurements, and nine seated subjects were asked to provide a subjective evaluation. 2. Beating idle vibrations 2.1. Idle vibrations from an actual vehicle This study measured and analyzed beating idle vibrations occurring in an actual passenger vehicle to ensure the effectiveness of the evaluation. Fig. 1 shows measurements of idle vibrations from the driver's seat track of a mid-sized sedan (Mileage: 200 km, Engine size: 2,000 cc) available in Korea and overseas. Comparisons are made between the idle vibrations before and after operating the cooling fan. The gear was in drive, and the vehicle was stopped for a certain period of time to heat up the engine, which then triggered the cooling fan. From the time-frequency map of Fig. 1 (a), we can see that the cooling fan operates at about 13.5 s, and vibrations (23.1 Hz) caused by fan imbalance are added to idle vibrations (21.7 Hz) of the engine. Beating is also observed in Fig. 1(b) and (c), which show the spectrum of acceleration and time signals before and after operation of the cooling fan. In Fig. 1(b), when the cooling fan was not in operation, a vibration of 98.6 dB was measured at a single frequency. When the cooling fan was in operation, vibrations of 99.5 dB and 97.1 dB were measured at two frequencies. Fig. 1(c), which shows time signals of vibrations measured from the test vehicle, gives an RMS of 101.2 dB when the cooling fan is not in operation and 103.0 dB when the cooling fan is in operation. Vibrations measured before and after operation of the cooling fan differ by 1.8 dB, falling below the human-perceivable level of 3 dB. However, some experts and general drivers participating in a pilot test prior to this study apparently experienced beating vibrations after the cooling fan was triggered, and reported a significant reduction in ride comfort. These findings highlighted the need for research on reduced ride comfort in relation to beating idle vibrations arising from operation of the cooling fan. 2.2. Generation of beating vibration signals This study generated beating signals to evaluate the extent to which ride comfort is reduced in relation to properties of beating vibrations. A beating signal is created when two sinusoidal signals having similar frequencies occur simultaneously, as shown in Eq. (1)
Fig. 1. Comparison of idle vibration measured on seat track of a passenger car with and without beating by cooling fan.
76
J. Park et al. / International Journal of Industrial Ergonomics 57 (2017) 74e79
aðtÞ ¼ A cosð2pftÞ þ B cosð2pðf þ fb Þt þ fÞ
(1)
here, fb is the beating frequency, or the difference in frequency between the two sinusoidal signals. f is the basic frequency of the sinusoidal signal, and 4 is the phase difference between the two signals. A and B represent the amplitude of each sinusoidal signal. A generated beating signal, as delineated in Eq. (1), creates vibrations with a beating frequency denoted by fb. For subjective evaluation of reduced ride comfort caused by beating vibrations, this study used the beating frequency fb and amplitude ratio B/A as control variables. The phase difference 4 of the two sinusoidal signals was set as 0. Fig. 2 shows the time signals in relation to the amplitude ratio of the beating signals (B/A) and beating frequency (fb). Fig. 3 is a comparison of the spectrum of the beating vibrations measured from an actual vehicle and the spectrum of beating signals created from Eq. (1). The generated beating signals were designed to have an f of 21.7 Hz, fb of 2 Hz, and B/A of 0.4 to match the frequency and size of the primary and secondary peaks of beating measurements. In the measured idle vibrations, vibrations of minor peaks over 25 Hz were considered negligible since they cannot be detected by the driver. A pilot test was conducted to determine whether the subjects were able to differentiate between the two signal types compared
Fig. 3. Spectrum of beating signal produced for imitation of the idle vibration measured on a passenger car (Primary peak frequency: 21.7 Hz, Secondary peak frequency: 19.7 Hz).
in Fig. 3. When seated in the car seat mounted on the exciter alternating between idle vibrations in the vehicle and selfgenerated beating vibrations, all the nine subjects were unable to differentiate between the two. For a subjective evaluation, the beating vibration signals were designed to have an fb of 0.5, 1, 2, 4, and 8 Hz for a B/A of 0.4 and 0.8, amounting to a total of 10 cases. The overall value of RMS from all beating vibrations was 101.2 dB, coinciding with that of actual vibrations.
3. Experimental method 3.1. Equipment A schematic diagram of the exciter system, which generates various beating vibrations, is shown in Fig. 4. The i-220 electric exciter developed by IMV Corporation was used to produce beating vibrations for the subjects. Table 2 shows the mechanical specifications of the vibration shaker (IMV-i220) used in the experiment. The various beating vibration signals are entered into the Ni-PXI
Fig. 2. Comparison of beating signals with different magnitude ratio and beating frequency.
Fig. 4. Excite system for generating idle vibration.
J. Park et al. / International Journal of Industrial Ergonomics 57 (2017) 74e79
77
Table 1 Physical data of subject employed in pilot test. Item
Mean
S.D.
Max/Min
Weight(Kg) Height(cm) Age(yr) Driving(yr)
69.0 174.1 26.4 1.6
8.1 4.7 1.9 0.8
85/60 180/167 31/25 3/0.5
Table 2 Specifications of vibration shaker. Model
IMV i-220
Type Rated force Frequency range Maximum displacement Maximum payload
Electro-dynamic 5.6 kN 3e3300 Hz 51 mm p-p 200 kg
system through Labview, and applied to the exciter via an amplifier following D/A conversion. Vibration signals experienced by the subjects were monitored in real-time by measuring the acceleration at the top of the exciter platform, and the amplitude of vibrations was adjusted using control signals in Labview and the amp gain value. While a phase difference between seat and floor vibrations led to slight differences in evaluations of ride comfort (Jang and Griffin, 1999), this study consistently excluded floor vibrations by separating the floor from the exciter platform.
Fig. 5. Seat setup and seating posture of subject.
3.2. Subjects and sitting posture For a subjective evaluation of ride comfort in relation to beating vibrations, this study relied on nine male subjects without spinal injuries. Table 1 shows the height, weight, age, and driving experience of the subjects. To simulate actual driving conditions, the exciter platform was mounted with the same fabric seat as the vehicle from which measurements were taken. The transmissibility of fabric seat generally has the main peak at around 5 Hz, whereas difference of the values over 20 Hz is quite little with a magnitude ranged from 0.2 to 0.5 (Van Niekerk, 2003; Griffin, 1990). The contact pressure at the lumbar region was different according to the geometry of the seat, and this geometry influenced the ride comfort. (Guo et al., 2016). The seat cushion and seat back angle were adjusted to 103 . As shown in Fig. 5, the seat height was adjusted to allow subjects to comfortably place their feet while maintaining a knee angle of 120 . The subjects were asked to look forward with their hands resting comfortably on their knees. They were given hearing protectors to minimize the effect of surrounding noise on their assessment of ride comfort. 3.3. Method of subjective evaluation This study applied the three-down one-up method to identify test signals of single-frequency sinusoidal vibrations having the same discomfort level of 101.2 dB as reference signals of beating vibrations. As shown in Fig. 6, the subjects were asked to evaluate the level of discomfort after being exposed to 20 s of beating reference signals and 20 s of sinusoidal test signals, with a stop time of 5 s between them. To eliminate bias, the order of the reference signals and test signals was randomized each time. One second of fading was given to the start and end of each signal to prevent the results from being influenced by shock vibrations. As mentioned earlier, the reference signals were beating vibration signals created
Fig. 6. Magnitude estimation method of test and reference signal randomized in order.
by combining two amplitude ratios and five beating frequencies. All reference signals were 101.2 dB (ref.:106 m/s2), corresponding with idle vibrations measured from the actual vehicle. The test signals were sinusoidal vibration signals having a single frequency, and 10 steps were prepared from 90 dB to 110 dB at 2 dB intervals in magnitudes. Fig. 7 shows the flow of obtaining equivalent discomfort based on the three-down one-up method, first proposed by Dixon and Mood (1948). This can be used to eliminate the effect of nerve adaptation under continuous stimuli. By repeating the steps presented in Fig. 7, the test continues until three downs and three ups occur, as shown in Fig. 8. The last four values are averaged, and this becomes the amplitude of equivalent discomfort of the single-frequency sinusoidal vibration to the beating. Mansfield and Griffin (2000) applied the three-down one-up method to difference thresholds, which were used to differentiate between the amplitudes of the two signals delivered via a car seat.
78
J. Park et al. / International Journal of Industrial Ergonomics 57 (2017) 74e79
Fig. 7. Flow chart of three-down one-up method.
Fig. 9. Magnitude of sinusoidal vibration with equivalent discomfort to beating vibration of 101.2 dB.
Fig. 8. Example of three-down one-up method.
Lee et al. (2012) used the same method in his study of objective human body reactions to vibrations and subjective discomfort in a passenger vehicle at a stop. 4. Experimental results and discussion Fig. 9 shows the amplitudes of single-frequency sinusoidal vibrations corresponding to the ten beating vibrations of 101.2 dB obtained by applying the three-down one-up method for the nine subjects, along with their median values. For all beating frequencies, beating vibrations induced greater discomfort than single-frequency sinusoidal vibrations. (p < 0.0001, Wilcoxon signed rank test). Differences in level of discomfort were found when the two vibrations had different beating frequencies. (p ¼ 0.047, Friedman test). A statistical analysis showed that changes in discomfort became more prominent at smaller
amplitude ratios. (B/A ¼ 0.4: p ¼ 0.0013, B/A ¼ 0.8: p ¼ 0.043, Friedman test). When the amplitude ratio (B/A) was 0.4, the greatest difference in equivalent discomfort was 3.2 dB at a beating frequency of 4.0 Hz. When the amplitude ratio was 0.8, the greatest difference was 3.8 dB at a beating frequency of 1.0 and 2.0 Hz. For the two amplitude ratios, the difference in equivalent discomfort was relatively small at a beating frequency of 0.5 Hz and 8.0 Hz. For the same beating frequency, more discomfort was caused by a larger amplitude ratio. (p < 0.001, Wilcoxon signed rank test). While ISO 2631-1 associates frequency weighting of top-bottom vibration in the seated position with greater discomfort at 4e10 Hz, beating vibration in this study tends to cause greater discomfort at relatively lower frequencies. Morioka and Griffin (2006) examined how the human body perceives vibrations in the tri-axial translational direction in a frequency range of 2e315 Hz, and found a greater sensitivity to vibrations less than 2e10 Hz. Ahn and Griffin (2008), Ahn (2010) studied equivalent discomfort and body response to shock signals in a range of 0.5e16 Hz, and reported
J. Park et al. / International Journal of Industrial Ergonomics 57 (2017) 74e79
greater discomfort at low frequencies below 1 Hz and resonant frequencies of 4e10 Hz. Discomfort level and body response had non-linear characteristics in relation to the amplitude of shock vibrations. The results of this study, which evaluated the discomfort caused by beating vibrations under actual seat conditions, were different from past research involving rigid body seats and single-frequency sinusoidal vibrations. This study found that the level of discomfort varied with the amplitude ratio of beating vibrations and beating frequency. As future work, rigid seats can be used to evaluate the discomfort caused by beating vibrations while eliminating the influence of seat characteristics. 5. Conclusion This study analyzed the deterioration in ride comfort caused by beating vibrations, which are generated when the frequency of idle vibrations is similar to the frequency of vibrations due to the cooling fan and other auxiliary parts in engine room. Various beating vibration signals were created by reflecting the amplitude of idle vibrations measured from an actual vehicle, and they were entered into an exciter. Using the subjective evaluations of 9 subjects and the three-down one-up method, the amplitudes of single-frequency sinusoidal vibration which is equivalent to a beating vibration were derived. Beating vibrations were associated with greater discomfort even when they were of the same amplitude as single-frequency sinusoidal idle vibrations. For idle vibrations with the same beating frequency, greater discomfort was induced at a higher amplitude ratio (B/A). Acknowledgement We would like to express our thanks to Renault-Samsung Motor Company which assisted to carry out the experimental study.
79
References Ahn, S.J., 2010. Discomfort of vertical whole-body shock-type vibration in the frequency range of 0.5 to 16 Hz. Int. J. Automot. Technol. 11 (6), 909e916. Ahn, S.J., Griffin, M.J., 2008. Effects of frequency, magnitude, damping, and direction on the discomfort of vertical whole-body mechanical shocks. J. Sound Vib. 311 (1), 485e497. Basri, B., Griffin, M.J., 2014. The application of SEAT values for predicting how compliant seats with backrests influence vibration discomfort. Appl. Ergon. 45 (6), 1461e1474. Dixon, W.J., Mood, A.M., 1948. A method for obtaining and analyzing sensitivity data. J. Am. Stat. Assoc. 43 (241), 109e126. Griffin, M.J., 1990. Handbook of Human. Academic Press, London. Guo, L.X., Dong, R.C., Zhang, M., 2016. Effect of lumbar support on seating comfort predicted by a whole human body-seat model. Int. J. Ind. Erg. 53, 319e327. ISO, I., 1997. Mechanical Vibration and Shockeevaluation of Human Exposure to Whole-body Vibrations (2631-1). International Standards Organization. Iyer, G., Prasanth, B., Wagh, S., Hudson, D., 2011. Idle Vibrations Refinement of a Passenger Car (No. 2011-26-0069). SAE (Technical Paper). Jang, H.K., Griffin, J.M., 1999. The effect of phase of differential vertical vibration at the seat and feet on discomfort. J. Sound Vib. 223 (5), 785e794. Langer, T.H., Ebbesen, M.K., Kordestani, A., 2015. Experimental analysis of occupational whole-body vibration exposure of agricultural tractor with large square baler. Int. J. Ind. Erg. 47, 79e83. Lee, J.Y., Jeon, G.J., Ahn, S.J., Jeong, W.B., 2012. The study of correlation between objective human response and subjective discomfort evaluation of idle vibration on passenger vehicle. Trans. Korean Soc. Noise Vib. Eng. 22 (5), 422e428. Mansfield, N.J., Griffin, M.J., 2000. Difference thresholds for automobile seat vibration. Appl. Ergon. 31 (3), 255e261. Morioka, M., Griffin, M.J., 2006. Magnitude-dependence of equivalent comfort contours for fore-and-aft, lateral and vertical whole-body vibration. J. Sound Vib. 298 (3), 755e772. Rao, S.S., Yap, F.F., 1995. In: Mechanical Vibrations, vol. 4. Addison-Wesley, New York. Sekuli c, D., Dedovi c, V., Rusov, S., Obradovi c, A., Salini c, S., 2016. Definition and determination of the bus oscillatory comfort zones. Int. J. Ind. Ergon. 53, 328e339. Thuong, O., Griffin, M.J., 2011. The vibration discomfort of standing persons: 0.5e16Hz fore-and-aft, lateral, and vertical vibration. J. Sound Vib. 330 (4), 816e826. Van Niekerk, J.L., 2003. The use of seat effective amplitude transmissibility (SEAT) values to predict dynamic seat comfort. J. sound Vib. 260 (5), 867e888.