Tire ABS-Braking Prediction with Lab Tests and Friction Simulations
The invention and application of antilock braking systems (ABS) has resulted in a significant improvement of traffic safety and a reduction of stopping distance, especially on wet roads [1]. The reason for this success is rather clear: ABS is designed to steer the braking process in the most efficient way by keeping an optimal level of tire slip. At the same time, it must be clear that ABS uses braking forces generated in the tire footprint, and really good braking is possible only with high-performance tires. The best way to probe tire performance is to build tires and test them. This is, however, a long and an expensive procedure, so prediction of ABS performance based on results of some simple experiments is a very attractive supplement to the development process. Tire-braking performance is related to the friction of rubber on a surface. Relevant friction mechanisms can include adhesion, rubber hysteresis, and various kinds of viscous friction. All of these phenomena depend on the local sliding velocity, load, and temperature of tread rubber. Tire blocks pass the footprint area of a braking tire very rapidly, but their dynamics are indeed influenced by ABS. All of these aspects make the problem of ABS-braking prediction very intricate. In this publication, we present an approach for prediction of the ABS-braking performance. The approach links friction measurements conducted in laboratory to tire tests results. The friction of six specially designed compounds was measured on dry and wet surfaces using a high-speed linear friction test rig. Obtained experimental results are analyzed with the aid of rubber friction theory [2,3] involving both surface and rubber as input parameters. It is demonstrated that lab friction test procedures can be used for prediction of ABS wet braking performance.ABSTRACT
(a) EU tire label. (b) Wet grip index and corresponding label class. Estimation of residual vehicle velocity and remaining stopping distances in comparison to the best label class A. Please note that the label classes D and G are not used in the EU label.
Example of LFT measurements: time-dependent (a) velocity, (b) acceleration, (c) sample displacement, (d) friction force, (e) normal force, (f) friction coefficient. Average steady-state friction coefficient μS is plotted in (f) with dashed line.
Sketch of a rubber block sliding on a rough surface. Friction mechanisms illustrated are (i) hysteresis energy loss due to rubber deformation, (ii) adhesion, (iii) viscous friction in media confined in between surface and rubber, (iv) interlocking of rubber block edge with surface asperities. Contact between rubber and surface appear only on asperities. Contact and friction depend on normal pressure p and sliding velocity v [Eq. (4)].
Real contact of a tire with smooth (left) and rough (right) asphalts [5,6].
(a) Storage and loss moduli of rubber compound and (b) the corresponding tan δ.
Correlation between relative ABS-braking performance on wet (a) and dry (b) roads and the corresponding lab friction measurements μ s by v = 0.5 (m/s) and p = 0.3 MPa. Points represent measured data with error bars. Numbers nearby denote compounds described in Table 1. Solid line is a linear correlation law y = ax + b found by means of the regression model. Dashed lined are the corresponding confidence bounds.
Comparison between lab friction measurements and theoretical simulations at v = 0.5 (m/s) and p = 3 (bar). (a) Theory described by the hysteresis friction only is plotted versus LFT measurements on wet surface. (b) Theory accounting both for hysteresis and adhesion friction mechanisms is compared with LFT measurements performed on wet (circles) and dry (diamonds) surfaces. Dashed line is an eye guideline for the perfect match μTHEORY = μLFT.
Contributor Notes