OPTICAL MEASUREMENT OF TREAD DEFORMATION FOR ROLLING RESISTANCE STUDIES

Transcription

1 Abstract OPTICAL MEASUREMENT OF TREAD DEFORMATION FOR ROLLING RESISTANCE STUDIES Yi Xiong, Ari Tuononen Vehicle Engineering, Department of Engineering Design and Production Aalto University P.O. Box Espoo, Finland address of lead author: Understanding the tread deformation behaviour of pneumatic tyres within the contact patch is of great significance in revealing the mechanism of rolling resistance. This paper demonstrates an optical tyre sensor platform developed for the tread deformation measurement. A short literature survey of rolling resistance study methods is presented. The methodology and corresponding procedure for an optical measurement platform are introduced. Validation experiments were conducted on a chassis dynamometer and an instrumented passenger car. An asymmetric tyre tread deformation along the contact patch was observed. The influences of the wheel load, velocity, and inflation pressure on the deformation of the tyre tread were examined. 1 INTRODUCTION Nowadays, the economic and ecological efficiencies of the transportation sector are receiving considerable attention as a result of high oil prices and increasing public awareness about global warming. Tyres, which act as one of the most significant components in ground vehicles, account for 17-21% of the total energy consumption [1]. Therefore, there is growing interest and demand for customers, tyre manufacturers, and governments to reduce the energy losses through tyres. According to the experimental results in [2], the hysteretic losses of tyre compounds dissipate 90-95% of tyre losses. Additionally, 2-10% of the losses are due to the tyre-road friction and % of the losses are due to the aerodynamic resistance. In another study carried out in the US [1], it is claimed that a 10% reduction in the average rolling resistance promises a 1 to 2% increase in the fuel economy. Furthermore, in both the European Commission White Paper for Transportation and the European Green Car Initiative roadmap, the improving of the rolling resistance of tyres is listed as a key issue. In order to further increase public awareness, the European Commission has introduced a new tyre labelling system, which, from November 2012, made it a legal obligation to show the rolling resistance performance, together with other performances, including the pass-by noise and wet grip. 1.1 Rolling Resistance Study Methods Within recent decades, rolling resistance has been extensively and intensively studied in both the tyre design and pavement design fields. Although several approaches have been proposed, these could generally be summarised as experimental methods and numerical methods. For experimental methods, comparison studies are mainly carried out on the basis of various measurement methodologies and results. In the ISO standard: 2009, the rolling resistance is defined as the loss of energy per unit of distance travelled. Since the rolling resistance is a quantity of energy, it is difficult or even impossible to measure it directly with state-of-the-art instruments and methodologies. In practice, an equivalent force F rr, which is required to move the rolling tyre in the desired direction with the same velocity, is measured instead. Several standards and procedures have been proposed and established for measuring the rolling resistance. According to the principles, those could be divided into four basic categories: the Laboratory Drum (LD) method, Trailer (TR) method, Coast-down (CD) method, and Energy Consumption (EC) method [3]. On the other hand, numerical methods such as finite element (FE) analysis provide powerful tools for modelling and predicting the rolling resistance. In [4] [5], a non-linear steady-state rolling analysis of a FEM tyre model is presented that is based on the Modal Arbitrary Lagrangian-Eulerian (M-ALE) approach. Moreover, two principal components, the smooth road rolling resistance and the road texture-induced rolling resistance, are studied. In another study [6], a waveguide finite element model is adopted to predict the tyre energy loss resulting from the rolling resistance. Nonetheless, those methods have their specific focus and limited scopes. For the experimental methods, even though most of them are well established, several limitations, such as the feasibility for tyre or pavement studies, 1

2 on-vehicle testing, and lifetime energy performance evaluation, still exist. Moreover, the state-of-the-art experimental methods cannot provide an insight into the rolling resistance mechanism. On the other hand, for numerical methods, the simplifications of the material properties in the FE simulation are not factual in real situations. The lack of proper instruments also causes difficulties in the validation of FE models. Therefore, it is necessary to propose an innovative method for rolling resistance studies of tyres that can be used for different purposes and under different conditions. Table 1 Comparisons of rolling resistance study methods Methods Conditions Tyre Design Investigations Pavement Design Investigations On-vehicle testing Remarks Laboratory Drum Difficult or Impossible tyre-flat road case Different with the In Laboratory Possible Impossible Method Trailer Method On Road Possible Possible Impossible Difficult for calibration procedure Coast-down Method Road the velocity effect In Laboratory/On Difficult to check Possible Possible Possible Energy Consumption Method On Road Difficult Difficult Possible Not accurate Numerical Method On Road Possible Possible Impossible Needs Validation Optical Measurement Method Road In Laboratory/On Possible Possible Possible Under development 1.2 Rolling Resistance Mechanism A general accepted theory [2] [7] [8] to explain the rolling resistance mechanism is described as follows: when a tyre is rolling at a constant speed, it is periodically deflected within the contact patch with the ground. Because of the hysteretic property of tyre compounds, the stress-strain relation means that the stresses are different at a constant strain within the loading phase and unloading phase. Therefore, the normal stress in the leading half of the contact patch is higher than that in the trailing half. This phenomenon results in a force shift from the central point of the contact patch to the leading half. The resultant moment caused by the shifted force is the so-called rolling resistance moment. Therefore, in order to estimate or predict the rolling resistance, the material deformation, the volume of material deformation, and the strain-stress relation need to be known. For elastic materials, the strain-stress relation is governed by Hooke s law. For rubber, a well-known viscoelastic material, the only way to determine the stressstrain response is using experimental methods. However, it should be noticed that a tyre, as a composite product, is built from various materials and complex structures. The mechanical properties of those materials are difficult to measure. One feasible way is to measure the tyre interface stress and tyre deformation separately. For tyre interface stress studies, research studies on both theoretical models and experimental results have been conducted. In [9], a dynamic model using modal parameters was proposed and experimentally validated. It should be noticed that such a model, which is only based on a geometrical relationship, is not accurate enough for in-depth normal pressure distribution analysis. In [10], a transient contact algorithm is developed, consisting of an analytical belt model, a non-linear sidewall structure, and a discretised viscoelastic tread foundation. Elegant experimental methods can be found in [11] [12] [13]. For rolling resistance analysis, a classic elastic ring model is introduced in [7]. In this model, a radial pneumatic tyre is simplified as an engineering structure with three principal components: 1. a tread band foundation (the inflated sidewall); 2. an elastic ring (the composite tread band), and 3. the tread components (attached to the tread band through spring-dashpot connections). For the inflated sidewall which has been modelled as a set of radial springs, the pressure distribution could be determined from the radial deformation of the tyre along the contact patch. As shown in Figure 1, this deformation results in a convex pressure distribution with a peak located in the centre of the contact patch. The bending of the elastic ring changes the curvature of the tread band from 1/r to 0. Such boundary conditions lead to a pressure distribution with two idealised concentrated forces applied to the ends of the contact patch. Moreover, taking into account the incompressibility of tyre rubber (Poisson ratio 0.5), it is necessary to consider the normal pressure changes caused by the shear deformation of the tread. Therefore, continuous pressure distributions with two peaks located at both ends of the contact path are introduced by the bending and shear deformation of the tyre. Several other factors, such as the velocity effect and the buckling effect under heavy loads, should also be considered. 2

3 Figure 1 Normal pressure distribution for a rolling tyre [7] However, to our knowledge, so far there is no efficient method to measure the tread deformation of the rolling tyre. In this paper, a novel optical tyre sensor platform for the measurement of the tread deformation is presented. The measurement results could be used further for rolling resistance studies. As listed in Table 1, the comparisons with other rolling resistance methods imply that the proposed measurement platform and procedures are feasible for investigations of rolling resistance from both the tyre and road aspects through either on-vehicle or laboratory testing. The rest of the paper is organised as follows. In Section 2, the methodology and setup of the optical tyre sensor platform are illustrated. Section 3 first presents the results of validation tests conducted on a chassis dynamometer drum under different operating conditions. Then, the results of a preliminary study on a normal public road are also demonstrated. Finally, Section 4 concludes all this and suggests proposals for the future. 2 MEASUREMENT PLATFORM AND PROCEDURE 2.1 Tread Deformation Measurement As discussed in the previous section, tread deformation is a complex phenomenon and is present in all three dimensions. The present work concentrates on the measurement of tread deformation in the normal direction (vertical). The methodology is illustrated in Figure 2. A tyre installed on a test rig is in contact with a drum. Under the assumption that the tread is in full contact with the surface of the drum, the contact profile is a curvature with a constant radius r. Thus, the tread thickness could be defined as the difference in height between the inner liner of the tyre and surface of the drum in the vertical direction. In order to represent the methodology mathematically, the two coordinates of the system are defined as: Rim Coordinate {R}: the origin is located at the centre of the axle, which does not rotate during rolling. The x- and y-axes point in the vertical and horizontal directions, respectively; Drum Coordinate {D}: the origin of this frame is affixed to the centre of the drum. The x- and y-axes are defined in the same manner as for the rim coordinate {R} The x-axes of both coordinates are aligned in an identical vertical line under the conditions of zero camber and slip angles (α=0 and γ=0 ). In other words, only a vertical translation {T} is needed when transforming the rim coordinate {R} to the drum coordinate {D}. In a matrix form, this can be described as: Dx 0 { R} { D} { T} { } { } (1) Dy r L 2 where L 2 is the distance from the axle to the surface of the drum in the vertical direction. This distance, which is also called the loaded radius, changes with the wheel load. For any element in the inner liner profile of the tyre, the distance ρ from the point to the axle could be measured. This is realised by measuring the distance L 1 from the steel rim to the inner liner and knowing the constant offset Z offset between the rim and the axle. Since the angular position θ of those points can be measured easily, it is convenient to describe the tyre inner liner profile as a polar coordinate, R inner-liner (ρ, θ). This could the be transferred into the Cartesian coordinate R inner-liner (x, y). Since the drum has a constant radius, the drum surface profile could also be described easily by the polar coordinate D drum (ρ, θ). In the same manner, this could also be described by the Cartesian coordinate D drum {x, y}. According to Equation 1, D drum {x, y} could be further transformed into R drum {x, y}. Thus, the tread thickness could be calculated as follows: 3

4 Z ( x) R ( x, y) R ( x, y) t inner linear With a known uncompressed tread thickness Z t0, the tread deformation during the contact is written as: drum Z Z t 0 Z t (3) Since the contact pressure is zero out of the contact, it is logical to assume that the vertical deformations only exist inside the contact patch. The contact patch length could be determined as the length of the non-zero deformation part. (2) 2.2 Optical Tyre Sensor Platform Figure 2 The principle of the measurement system Within the previous projects [14] [15] [16] [17], the optical tyre sensor platform for tyre-road studies in the Vehicle Engineering group at Aalto University has experienced two generations, namely the OTS sensor, which measures the movement of an LED glued onto the tyre inner liner, and the laser triangulation sensor, which acts as the basis for the study in this paper. On the basis of the aforementioned methodology in Section 2.1, an optical tyre sensor platform that included two laser triangular sensors, a slip ring package with an angular sensor, and a data acquisition and processing system was proposed (Figure 3 (a)). One laser sensor (Keyence IL-065) was capsulated and assembled on a steel passenger car rim to measure the distance L 1. The sampling frequency of the sensor was 3 khz. For more details regarding this sensor setup, please refer to [17]. In addition, another laser sensor (Keyence LK-H150) was selected for the loaded radius L 2 measurement. This sensor was fixed to a rigid mechanical support attached to the stator of the slip ring. The sensor has a small laser spot diameter (ϕ =120μm) and an ultra-high sampling frequency (392 khz), because of the requirement to capture the micro-roughness of the road surface at normal driving velocities. The slip ring package (Michigan Scientific SR10AW/T512) is able not only to transmit signals in multiple channels from a rotating tyre but also to measure the circumferential position of Laser 1. The data acquisition system was implemented through the NI DAQ system and the post-processing algorithm was developed in Matlab. The sampling frequency of the DAQ system was selected to be 10 khz. 4

5 (a) (b) Figure 3. The optical measurement platform for the tyre tread deformation study: (a) a CAD model; (b) the real system on a chassis dynamometer drum As shown in Figure 3 (b), the experiments were conducted in a test rig on a chassis dynamometer drum [18] which was covered with Safety-Walk (roughness ~ P 60). The measured radius r of the drum was mm. The wheel load was adjustable through hydraulic cylinders controlled by p/q controllers. Moreover, the test tyre could be driven in either forwards or backwards directions, which is important in this rolling resistance study. In the present work, the slip angle and the camber angle were both adjusted to 0 degrees. 2.3 Sensor Location Calibration As mentioned in the previous section, the circumferential position θ of Laser 1 is measured by an angular sensor and defined in the polar coordinate {R}, which has its zero degree located at the 6 o clock position. However, the zero degree position for the angular sensor coordinate {S} is not the same as the one within the rim coordinate {R}. Therefore, it is necessary to calibrate the angular offset θ offset between these two polar coordinates. Here, a software calibration procedure is presented. The assumptions are based on the observations in [16] that: (i) the largest tyre carcass deflections are located in the centre region of the contact patch; (ii) the largest tyre carcass deflections are symmetric about the centre of the contact patch in both directions of rotation. For the calibration experiments, the tyre was rotating at 5 km/h, under a 300 N load in both the clockwise and counterclockwise directions. Figure 4 (a) shows the carcass deflections of five repeated test runs in the clockwise and counterclockwise directions, respectively. The angular coordinates of the peaks for each run were determined with a peak detection algorithm and are shown in Figure 4 (b). Figure 4 (a) Inner linear displacement as a function of circumferential position. (b) Calculated offset angle of 10 calibration runs (the 1 st -5 th runs in a counterclockwise direction, the 6 th -10 th in a clockwise direction) Therefore, the angular offset θ offset is calculated as the average of those angular coordinates of peaks for equal numbers of calibration runs in both directions: n n offset ( clock counter clock ) 2n (4) i 1 i 1 5

6 where θ clock and θ counter-clock are the angular coordinates of the peaks during one single calibration run in the clockwise and counterclockwise directions, respectively. Moreover, another quantity that needs to be calibrated is the offset Z offset. The calibration is conducted as follows: (i) two laser sensors are installed on the rim without a tyre; (ii) the sensors are oriented in the same vertical direction for a horizontal surface measurement. For this case, the geometry information is simplified as: 3 RESULTS AND DISCUSSION Z tread L2 Zoffset L1 * cos 0 0 Zoffset L2 L1 (5) The experiments were conducted on the aforementioned optical tyre sensor platform with a commercially available passenger car tyre (205/55R16). The original thickness from the inner liner of the tyre to the outer surface of the tread was measured as 15.9 mm. Four variables, namely the direction of rotation, wheel load, velocity, and inflation pressure, were selected to examine their potential effects on tyre tread deformations. In order to reduce coupling effects, only one variable was selected in each experiment by keeping all the other variables constant. The observations are presented and discussed in detail in the following section. 3.1 Asymmetric Tyre Tread Deformation Figure 5 demonstrates the measured tyre thickness and deformation within the contact patch for both the clockwise (CW) and counterclockwise (CCW) directions. Both contours show the same observation that there is a larger deformation at the leading edge and a smaller deformation at the trailing edge. Such an observation displays a good match with our analysis and prediction in Section 2. Moreover, the asymmetry caused by the rolling resistance changes with the direction of rotation. In an ideal case, the tyre deformation should be identical for both directions of rotation under the same experimental conditions. However, the results of the experiment show that the tyre deformation in a clockwise direction is slightly smaller than that in a counterclockwise direction. The possible explanations include: (i) the anisotropic behaviours of a real tyre rolling in two different directions; (ii) the difference between the nominal and actual applied loads caused by the tyre test rig control system. Figure 5 The asymmetric tread deformations for a radial tyre (wheel load 2 kn, velocity 10 km/h, inflation pressure: 2.5 bars) 3.2 Influence of Wheel Load on Tyre Tread Deformation In order to examine the effects of the wheel load on tyre tread deformation, experiments were conducted under three different loads (2000N, 4000N, and 6000N). As shown in Figure 6, the experimental results imply that a higher load leads to a longer contact patch length and vice versa. Meanwhile, the tread deformation appears to be smaller under a higher load. This might be due to the shift of the load from the tyre crown to the tyre shoulder, which has been documented in many other studies [10] [11] [13]. On the other hand, the largest deformation point shifts forwards with the leading edge. For gaining an understanding of the rolling resistance mechanism of 6

7 the tyre, this observation is significant as it reveals the changes in deformation under different load conditions. It can be seen that a larger load will result in a greater rolling resistance force, which shows a good match with both daily experience and experimental observations [2]. Furthermore, in the contours for all three different loads, a buckling-like deformation drop could be observed. This phenomenon, which depends on the load condition, needs to be investigated further. Figure 6 Comparison of measured tread deformation under different wheel loads (velocity 5 km/h, inflation pressure 2.6 bars, counterclockwise) 3.3 Influence of Velocity on Tyre Tread Deformation The tyre treads suffer a cyclic deformation in which the excitation frequency is determined by the speed of rotation of the wheel and roughness of the pavement. Moreover, according to e.g. [19], the damping property of rubber is highly dependent on the excitation frequency. Hence, the velocity of the wheel has significant impacts on the tyre tread deformation. Figure 7 illustrates the fact that a higher velocity enlarges the difference in thickness between the tread at the leading edge and at the trailing edge. Such differences are amplified by the greater compression at the leading edge and lesser compression at the trailing edge. Similar observations of the effects of velocity on the normal pressure and rolling resistance can be found in [20] [21]. In the same manner, for a higher load, the largest deformation point also shifts forwards within the leading edge. For a tyre-road case, the simulation results in [10] [22] show that the velocity has an insignificant effect on the contact length for a radial tyre. However, as shown in Figure 7, for a tyre-drum contact, the velocity has an important influence on the contact patch length. A similar experimental result can be found in [11]. The different contact profile shape could account for such differences. This might be an interesting topic in the future to help gain an understanding of and analyse the differences in the rolling resistance measurement on the drum and on the flat road. 7

8 Figure 7 Comparison of measured tread deformation at different velocities (wheel load 5 kn, inflation pressure 2.6 bars, counterclockwise) 3.4 Influence of Inflation Pressure on Tyre Tread Deformation It is well known that the inflation pressure has a strong effect on the rolling resistance and fuel consumption. Comparisons of the tread deformation for different inflation pressures indicate that a higher inflation pressure (3.0 bars) will shorten the contact patch length (Figure 8). Meanwhile, the maximum deformation level also decreases. Moreover, the largest deformation point for the lowest pressure shifts forward into the leading edge which will result in a greater rolling resistance moment. To conclude, also from this point of view, the proper inflation pressure is the easiest effective way to reduce the rolling resistance. Figure 8 Comparison of measured tread deformation at different inflation pressures (velocity 5 km/h, wheel load 5 kn, counterclockwise) 3.5 Preliminary Study for On-board Measurement By assuming that the radius r goes to infinity on a flat road, the aforementioned methodology for drum measurement is also applicable for on-board measurement. A preliminary study for on-board measurement was carried out on an instrumented passenger car (VW Golf V Variant) at a slow speed for both the clockwise and coun- 8

9 terclockwise directions. Moreover, the experiments were conducted on a city street with asphalt pavement in Espoo, Finland. As shown in Figure 9, the measurement result for on-board measurement clearly manifests the asymmetric tread deformations in two different directions. Furthermore, compared with drum measurement, it demonstrates a smoother change in the thickness from the leading edge to the trailing edge. Considering the curvature of the dynamometer drum, the resultant larger and more complex deflections may account for the differences. More results related to this topic, including various inflation pressures, road conditions, and wheel loads, will be presented in a future paper. It should also be noted that the force and torque transmission is different for the tyre on the chassis dynamometer and on the test vehicle. For the first case, the torque is transmitted from the drum to the tyre through the drum-tyre interface. On the other hand, for the on-board test, the torque from the driveline is transmitted from the rim to the tyre through the rim-tyre carcass interface. Thus, the windup phenomenon probably has different effects on these measurement results. Figure 9 Measured tread thickness of on-board measurement for different directions 4 CONCLUSIONS AND FUTURE WORK In summary, an optical tyre sensor platform for tread deformation measurement was presented in this paper. The asymmetric tread deformations along the contact patch as a result of the rolling resistance were clearly observed. The factors which affect the degree of asymmetry, including the directions of rotation, wheel loads, velocities, and inflation pressure, were examined. This study provides a powerful tool for revealing the mechanism of rolling resistance. Furthermore, the durability of this platform has also been examined by means of high-speed tests and on-board tests. The main contribution of this work lies in the development of an optical tyre sensor platform for rolling tyre tread deformation studies. In the future, several related tasks will be carried out. Firstly, the small spot size and ultra-high sampling frequency of the laser sensor make it possible for it to be used in future studies to examine the possible effect of the roughness of the road on tyre tread deformations. Moreover, as mentioned in several other publications, the pressure distribution along the cross-section of the tyre is also asymmetric and the shape depends on both tyre operation and design variables. Therefore, it is worthwhile to further examine the three-dimensional pressure distribution with a 2D laser sensor. ACKNOWLEDGEMENTS The authors wish to thank Prof. Matti Juhala for his helpful discussions and Mr. Panu Sainio for his support. We are particularly grateful to Mr. Pekka Martelius for his help in building the test environment. The research leading to these results was performed within the LORRY project ( and received funding from the European Community's Seventh Framework Programme (FP7/ ) under grant agreement No This work was also partially funded by the Academy of Finland (grant number ). The support is gratefully acknowledged. 9

10 References [1] Tires and passenger vehicle fuel economy, Washington, D.C., [2] J.Y. Wong, Theory of ground vehicles, New York: John Wiley & Sons, [3] U. Sandberg, Rolling resistance-basic information and state-of-the-art on measurment methods, [4] S. Boere, Prediction of road texture influence on rolling resistance, Eindhoven University of Technology, [5] I. Lopez, Influence of material damping on the prediction of road texture and tread pattern related rolling resistance, in Proceedings of the International Conference on Noise and Vibration Engineering, [6] M. Fraggstedt, Vibrations, damping and power dissipation in Car Tyres, Stockholm: KTH, [7] S. K. Clark, Mechanics of pneumatic tires, U.S. Department of Transportation, National Highway Traffic Safety Administration, [8] E. H. David and J. C. Moreland, Fundamentals of rolling resistance, Rubber Chemistry and Technology, vol. 74, no. 3, pp , [9] D. Guan and C. Fan, The quantitative analysis and experimental verification of the tire static enveloping model using experimental modal parameters, Vehicle System Dynamics, vol. 44, no. 9, pp , [10] A. Tsotras, On the interaction between modal behaviour and shear force behaviour of a pneumatic tyre, Loughborough: Loughborough University, [11] W. Hall, Finite element modelling and simulation for a 'smart' tyre, Coventry: University of Warwick, [12] H. Shiobara, T. Akasaka and S. Kagami, Two dimensional contact pressure distribution of a radial tire in motion, Tire Science and Technology, vol. 24, pp , [13] M. De Beer, C. Fisher and J. F. J., Determination of pneumatic tyre/pavement interface contact stresses under moving loads and some effects on pavements with thin asphalt surfacing layers, in Eighth International Conference on Asphalt Pavements, Seattle, Washington, [14] A. Tuononen, On-board estimation of dynamic tyre forces from optically measured tyre carcass deflections, International Journal of Heavy Vehicle Systems, vol. 16, no. 3, pp , [15] A. Tuononen and L. Hartikainen, Optical position detection sensor to measure tyre carcass deflections in aquaplaning, International Journal of Vehicle Systems Modelling and Testing, vol. 3, no. 3, pp , [16] A. Tuononen, Optical position detection to measure tyre carcass deflections, Vehicle System Dynamics, vol. 46, no. 6, pp , [17] A. Tuononen, Laser triangulation to measure the carcass deflections of a rolling tire, Measurement Science and Technolog, vol. 22, no. 12, [18] H. Pirjola, Dynamic tyre force measuring device, Helsinki University of Technology, Espoo, [19] B. N. Persson, Sliding friction: physical principles and applications, Springer, [20] J. S. Dieterich, Tire rolling resistance: papers given at the symposium on tire rolling resistance at the 122nd meeting of the rubber division, Akron: American Chemical Society. Rubber Division, [21] S. J. Kim and A. R. Savkoor, The contact problem of in-plane rolling of tires on a flat road, Vehicle System Dynamics, vol. 27, no. S1, pp , [22] S. Gong, Study of in-plane dynamics of tires, Delft: Delft University of Technology,