Measurement and Analysis of Vehicle Comfort and Road Quality
Measurement and Analysis of Vehicle Comfort and Road Quality
Measurement and Analysis of Vehicle Comfort and Road Quality
Abstract
The suspension system of a vehicle is an essential component because it is related to the dynamic performance of the vehicle. It
ensures the tire-grounding contact, absorbing the oscillations excited by the road, and maintains the stability of the vehicle by
distributing the wheel forces, providing comfort and safety to vehicle occupants. Structurally it links the vehicle sprung mass to
the unsprung mass, and it connects to other subsystems, such as the steering and the brake components. In order to analysis and
design a suspension system, several models that simulate the suspension dynamics behavior are available, however each has a
different level of complexity and the accuracy of the results varies. Thus, the aim of this study was to compare a quarter car model
with the real dynamics of a vehicle using a Matlab simulation. A test vehicle was instrumented with three-axis accelerometers,
through which data on the acceleration of the sprung mass and unsprung mass was obtained when the vehicle was excited by a
speed bump. From this simulation, it was possible to perform a model update to obtain the approximate vehicle parameters. The
road surface quality plays a fundamental role in the vehicle oscillation and therefore in the comfort of the vehicle occupants.
Using the instrumented vehicle and the adjusted parameters of the model, some routes along different pavement types in the city
of Joinville were chosen to collect the vibration data in a vehicular test. The results were used to determine the international
roughness index (IRI), an index to evaluate road quality. It was verified that the city roads with high IRI values, for instance, on
streets paved with cobble stones, are associated with greater discomfort to passengers inside the vehicle, reflecting a low quality
of the pavement. In contrast, lower IRI values are obtained for roads with a higher quality surface, resulting in greater comfort
for the passengers.
MacPherson system. As described by Gillespie (1992)[10], vehicle (sprung mass). This condition is analyzed by Liang
for a two lumped-parameter dof model for independent et al. (2013)[16] for an excitation caused by speed-control
suspensions, an excitation from the road reaching the wheel humps.
on one side would not interfere with the dynamics of the
wheel on the opposite side. Due to the associated costs and user requirements, the
quality of roads has become an important concern. Road
The MacPherson suspension is widely employed in the managers, for instance, aim to maintain the quality of paved
automobile industry because it is a very simple assembly road surfaces through periodic maintenance. Thus, methods
and has a compact size. It is mounted with a telescopic have been developed to identify and monitor road features to
shock absorber and a concentric spring along the strut ensure that the conditions are appropriate for use. The main
structure. A lower control arm supports the strut and it links tool employed is the international roughness index (IRI),
the spindle to the vehicle chassis. Also, the independent which is widely applied to assess the quality of roads in the
suspension has a stabilizer bar connecting the two sides of United States [12].
the lower control arms. Reportedly, this is one of the biggest
advantages of MacPherson suspension[10] and it does not In this context, to obtain data on vehicle suspension systems
take up much of the engine space. and also information regarding road conditions, sensors such
as inertial accelerometers are used by automotive engineers
As mentioned above, in the automotive area comfort has and researchers. These data can then be used to adjust a
been the focus of many studies to improve the quality of the numerical model of the vehicle, and with this model the IRI
vehicle ride and reduce interior noise, which forms part of value for the road can be determined and the road conditions
the vibro-acoustical system[11]. The ride comfort is defined can be classified. The objective of this paper was to obtain
by Genta and Morello (2009)[12] as the ability of the tires vibration acceleration data from a low-cost data acquisition
and suspension system to filter external vibration excitations system, using an Arduino Uno microcontroller board, and
caused by the road irregularities. Among the main factors determine the condition of a road pavement in the city of
that result in complaints are noise and vibration, grouped Joinville (Brazil).
together in noise, vibration and harshness (NVH) studies.
These oscillations of low frequency, if not damped, can lead 2. EXPERIMENTAL SETUP
to passenger discomfort and are often responsible for
customer dissatisfaction[13]. In this study, an Arduino Uno microcontroller board and
two 3-axis accelerometers (ADXL345) were used to
The vehicle dynamic response to pavement roughness can measure the vertical vibration of sprung and unsprung
be measured by placing accelerometers on the sprung and masses of a test vehicle. In the measurements performed, the
unsprung mass. The correlation of the measured data with accelerometers were configured to have a range of ±8g to
user satisfaction is proposed in the standard ISO 2631 avoid overloading in terms of the signals and a sampling
(1997)[14], which regulates the parameters of the vehicular rate of 55Hz. The electrical circuit of the Arduino board and
comfort and safety. This standard provides information on ADXL345 accelerometers is represented in Figure 1.
the parameters correlated with the vibration frequency and
the vertical acceleration, in addition to the sensitivity of the
human body to vibration from different directions. The
standard presents, for example, a scale of discomfort due to
vertical acceleration (Table 1). Table 1 shows the relation
between the root mean square (RMS) acceleration of the
sprung mass and the subjective sensation of the driver and
passengers with regard to the comfort of the vehicle ride.
𝑀. 𝑍 + 𝐾𝑠 . 𝑍 − 𝑍𝑢 − 𝐶𝑠 . 𝑍 − 𝑍𝑢 = 0 (1)
𝑚. 𝑍𝑢 + 𝐾𝑡 . 𝑍𝑢 − 𝑍𝑟 − 𝐾𝑠 . 𝑍 − 𝑍𝑢 (2)
+ −𝐶𝑠 . 𝑍 − 𝑍𝑢 = 0
Fig 3: Test vehicle.
_______________________________________________________________________________________
Volume: 07 Issue: 09 | Sep-2018, Available @ www.ijret.org 74
IJRET: International Journal of Research in Engineering and Technology eISSN: 2319-1163 | pISSN: 2321-7308
3.2 Model Adjustment two cycles of Figure 7 have similar decrements in the
frequency and amplitude; however, after this period the
The numerical adjustment of the model was performed effect of the rear suspension vibration adds a different
based on a previous experiment on the test vehicle. Here, the frequency component to the signal. Since the quarter car
acceleration of the sprung mass can be related to the model model represents two dof dynamics, the simulated response
vibration amplitude. A known road profile (a bump) is would not identify the presence of the new frequency.
applied to excite the front suspension on the right side at a
vehicle speed of 9km/h. For the simulation, the same profile The parameters of the adjusted vehicle model are given in
used in the experiment is measured and described using a Table 3. This resulting numerical adjustment uses a slightly
parabolic function of the bump height. An ultrasonic tape different stiffness value (less than 24%) in the function for
measure supported on a horizontal structure was used for the modelling the response.
discretization of the profile, as shown in Figure 5.
Equation (3) presents an approximated formulation obtained Fig 7: Comparison between simulationand measurement
from the measured points to describe the profile of this data.
bump. The curve is shown in Figure 6, where 𝑦 is the bump
height and 𝑥 is the traverse distance in meters. Table 3: Parameters of the vehicle according to the model
adjustment.
𝑦 = −3.7143. 𝑥 2 + 0.9229. 𝑥 − 0.0014 (3) Variables Symbol Value
Suspension stiffness 𝐾𝑆 55000.0N/m
Damping coefficient 𝐶𝑆 800.0N.s/m
Sprung mass 𝑀 294.0kg
Unsprung mass 𝑚 56.0kg
Tire stiffness 𝐾𝑡 550000.0N/m
_______________________________________________________________________________________
Volume: 07 Issue: 09 | Sep-2018, Available @ www.ijret.org 75
IJRET: International Journal of Research in Engineering and Technology eISSN: 2319-1163 | pISSN: 2321-7308
surface will be. In the figure, an IRI above 10.0 means 4.2 Analysis of road quality
unpaved road characteristics and an IRI below 4.0 indicates
good pavement surface (i.e., interstate highway or airport According to Du et al. (2014)[19], there is a strong
pavement quality). relationship between the IRI and the vertical acceleration of
the vehicle, since it is based on the quarter vehicle model.
According to [12], using a new quarter vehicle model with
two degrees of freedom, it is possible to calculate the IRI
value. This is called the golden car model, and the input
parameters are defined in Equations (5), (6), (7) and (8). In
this study, the golden car model was defined for an adjusted
sprung mass of 𝑀 = 294 𝑘𝑔.
𝐾𝑠 (5)
= 63.3 𝑠 −2
𝑀
𝐾𝑡 (6)
= 653.0 𝑠 −2
𝑀
𝑚 (7)
= 0.150
𝑀
𝐶𝑠 (8)
= 6.147 𝑠 −1
𝑀
Fig 8: Relationship between IRI and road characteristics From the motion equations of the golden car model, the
vibration response for road excitation is determined through
4.1 Estimation of surfaceroughness a numerical solution. The integration of ordinary differential
Based onthe vehicle data collected, a surface roughness equations is solved using the Runge-Kutta method. The IRI
value can be obtained. Isolating the excitation term, 𝑍𝑟 , of value is calculated through Equation (9), defined by [19], for
Equation (2) results in the road profile, as given in Equation a reference speed of 80km/h.
(4).
𝐿 (9)
1
𝑚. 𝑍𝑢 + 𝐾𝑡 . 𝑍𝑢 − 𝐾𝑠 . 𝑍 − 𝑍𝑢 − 𝐶𝑠 . 𝑍 − 𝑍𝑢 (4) 𝐼𝑅𝐼 = 𝑍 − 𝑍𝑢 𝑑𝑡
𝑍𝑟 = 𝐿
𝐾𝑡 0
Vehicular
Speed test
bump •Comparison
•Estimation of • Golden car
• Matlab • Sprung and model
• Experiment vehicle unsprung mass
• 2 dof model parameters • IRI values
• Sprung mass vibration
vibration • Road profile
Numerical Calculation of
simualtion Model
adjustment IRI
5. EXPERIMENTALRESULTS
After the model updating procedure conducted with the test
vehicle, on-road testing to verify the IRI calculation was
performed in the city of Joinville in October 2017. After the
data collection, signal processing was carried out using a
digital filter. This procedure was configured for a second-
order high pass Butterworth filter with a cutoff frequency of
1.0Hz.
_______________________________________________________________________________________
Volume: 07 Issue: 09 | Sep-2018, Available @ www.ijret.org 77
IJRET: International Journal of Research in Engineering and Technology eISSN: 2319-1163 | pISSN: 2321-7308
_______________________________________________________________________________________
Volume: 07 Issue: 09 | Sep-2018, Available @ www.ijret.org 78
IJRET: International Journal of Research in Engineering and Technology eISSN: 2319-1163 | pISSN: 2321-7308
many irregularities, such as a cobbled street, provided IRI requirements. 2 ed. International Organization for
values around 15.0m/km (route 1), which is classified as a Standardization. Geneva, Swiss, 1997.
damaged pavement in Figure 8. In contrast, for a smoother [15] Loprencipe, G.; Zoccali, P. Use of generated artificial
road, such as an asphalt pavement, the IRI value was road profiles in road roughness evaluation. J. Mod.
2.9m/km (route 5), indicating anew pavement. Therefore, Transport. 25, 1, p.24–33, 2017.
this methodology focuses on the application of analytical- [16] Liang, S.; Sun, Y.; Zhu, Q.; Yang, Z.; Zeng, C. Ride
based models to determine an index to classify road surface comfort analysis of a nonlinear vehicle excited by the
quality, using low-cost instrumentation. consecutive speed-control humps. Journal of
Vibroengineering, 15, 4, p.1656-1664.
REFERENCES [17] Arhin, S.A.; Noel, E.C.; Ribbiso, A. Acceptable
International Roughness Index Thresholds based on
[1] Stone, R.; Ball, J.K. Automotive Engineering Present Serviceability Rating. Journal of Civil
Fundamentals. Warrendale, Pa: Society of Automotive Engineering Research, Washington, EUA. p.90-96,
Engineers, 2004. 2015.
[2] Knowles, D. Classroom manual for automotive [18] Abulizi, N. et al. Measuring and Evaluating of Road
suspension & steering systems. 5. ed. Clifton Park, Roughness Conditions with a Compact Road Profiler
NY: Delmar, 2011. and ArcGIS. Journal of Traffic and Transportation
[3] Wada, T.; Konno, H.; Fujisawa, S.; Doi, S. Can Engineering. p.398-411, 2016.
passengers’ active head tilt decrease the severity of [19] Du, Y. et al. Measurement of International Roughness
carsickness? Effect of head tilt on severity of motion Index by Using 𝑍-Axis Accelerometers and GPS.
sickness in a lateral acceleration environment. Human Shanghai, China. Jun. 2014.
Factors, 54, 2, p.226-234, 2012. [20] ASTM E1926-08; Standard Practice for Computing
[4] Azizan, A.; Fard, M.; Azari, M.F.; Benediktsdóttir, B.; International Roughness Index of Roads from
Arnardóttir, E.S.; Jazar, R.; Maeda, S. The influence of Longitudinal Profile Measurements. American Society
vibration on seated human drowsiness. Industrial for Testing and Materials International. 2015.
Health, 54, p.296-307, 2016.
[5] Walker, G. H.; Stanton, N. A.; Salmon, P. M. Humam
Factors in Automotive Engineering and Technology.
Farnham, Surrey: Ashgate, 2015.
[6] Wong, J. Y. Theory of Ground Vehicles. 3 ed. Canada.
2001.
[7] Gao J.; Niu Z.; Yang X.; Zhang K. Correlative analysis
of hydraulic bushing of suspension control arm and full
vehicle ride comfort. Journal of Highway and
Transportation Research and Development, 9, 4, p.102-
110, 2015.
[8] Kuznetsov, A.; Mammadov, M.; Sultan, I.; Hajilarov,
E. Optimization of improved suspension system with
inerter device of the quarter-car model in vibration
analysis. Arch. Appl. Mech., 81, p.1427-1437, 2011.
[9] Sezgin, A.; Arslan, Y.Z. Analysis of the vertical
vibration effects on ride comfort of vehicle driver.
Journal of Vibroengineering, 14, 2, p.559-571, 2012.
[10] Gillespie, T. D. Fundamentals of Vehicles Dynamics.
Warrendale, PA: Society of Automotive Engineers,
1992, 470 p.
[11] Genuit, K. The sound quality of vehicle interior noise:
a challenge for the NVH-engineers. Int. J. Vehicle
Noise and Vibration, 1, 1/2, 2004.
[12] Genta, G.; Morello, L. The Automotive Chassis:
Volume 2: System Design. Dordrecht: Springer, 2009.
[13] Tang, C.; Zhang, Y.; Zhao G.; Ma Y. Annoyance rate
evaluation method on ride comfort of vehicle
suspension system. Chinese Journal of Mechanical
Engineering, 27, 2, p.296-303, 2014.
[14] ISO 2631 - 1//1997; ISO 2631 - 1 – Mechanical
Vibration and Shock – Evaluation of Human Exposure
to Whole - Body Vibration – Part 1: General
_______________________________________________________________________________________
Volume: 07 Issue: 09 | Sep-2018, Available @ www.ijret.org 79