Tire Wear and Pollutants: An Overview of Research

: Tire Road and Wear Particles are a major source of microplastic emissions. Tire and Road Wear Particles are important to study and understand as there are alarming amounts found in various environments. Currently, Tire and Road Wear Particles compared to other microplastics are not studied as rigorously in literature but are becoming a larger field of study due to their impact on emissions control. Tire Road and Wear Particles are commonly found as Styrene Butadiene Rubber Butadiene Rubber, and Natural Rubber To understand and quantify tire wear, experimental and mathematical models are developed to estimate tire wear. Tire wear can be measured experimentally using semi-empirical models, predetermined data, and sensor technologies. Tire wear is also measured mathematically using different modeling approaches and different friction models. This review discusses different and popular methodologies for estimating tire wear through experimental and simulated environments. Furthermore, discusses a review of the literature regarding tire wear emissions and its impact on the environment. Finally, it is evident that an accurate simulated tire wear model can be developed in the future alongside a driver model to predict tire wear emissions.


Introduction
According to recent statistics, there are approximately 1.4 billion ground vehicles today including passenger vehicles, trucks, buses, and motorcycles [1]. The majority of these vehicles are equipped with pneumatic tires that are worn over time as they roll over different surfaces and varying operating conditions. In 2020, it is estimated that there are up to 5,917,518 t/a of tire wear emissions globally, and up to 2.6 kg/capita per year in Europe alone [2]. The alarming climate crisis occurring globally has incentivized research on methods to estimate and reduce harmful emissions, including those given off by vehicles and tires. So much so, there are many initiatives and challenges sponsored by industry leaders and government associations to reduce harmful tire emissions such as the `Plastics Challenge: Mitigating the Release of Microplastics from Tire Wear' from Innovation, Science and Economics Development Canada.
To understand the effects of tire wear emissions, commonly referred to as Tire Road and Wear Particles (TRWPs), the particulates must be estimated or measured. Currently, there are two methodologies to approximate tire wear emissions. The first method estimates tire wear emissions through emissions factors, such as observing the mass of tire wear and average mileage. The second method estimates tire wear by measuring tire mass loss from friction and abrasion. A deeper understanding of how tire wear is estimated and measured through modeling and experimental methods can benefit in creating more accurate models in their respective categories or developing more comprehensive hybrid models [3]. In other tribology studies, surface texturing of the tire can be used as a method to control topography and friction at the tire-terrain contact. Surface texturing to control friction through tire topology has been a prominent research topic in the past 60 years [4].

______________________________________________________________________________
By controlling the surface texturing of the tire, it may be possible to retain adhesion to the terrain but reduce the amount of material expelled as emissions.
In the present day, reducing harmful emissions has been a priority on a global scale. Understanding tire wear and methods to measure/estimate the quantity of harmful emissions are critical to formulating strategies for mitigating the release of TRWPs [5,6].
This study aims to present a technical survey on approaches to estimating tire wear through experimental and modeling approaches, the methods of measuring tire wear pollutants, and the effects of tire wear pollutants on the environment. Currently, there is a lack of literature in regard to tire wear research, especially detailed wear models linking to accurate emission values. The simulated and experimental tire wear models shown in this review are methodologies to estimate wear rate, however, they may not show the emissions of the tire wear and its impact on the environment. Wear emissions and factors are separately discussed further in the consequent section.

Tire Parameters
Preliminarily, before understanding different methodologies to measure and estimate tire wear, fundamental tire parameters should be established. This section provides an overview of important tire forces and moments.  Figure 1 depicts the forces and moments that a tire may experience during operation according to SAE J670 JAN2008 [7]. Moreover, these forces are also referred to as tire-terrain interaction characteristics as they are only present when the tire is in contact with a surface. Here, the tire is torqued in the longitudinal direction, and hence Fx is a tractive force which is also commonly referred to as a more general term, the longitudinal force [3,5,6]. Alternatively, for a free-rolling tire, the rolling resistance force can be seen in the opposite direction of the wheel heading in Figure 2.
The longitudinal force is quantified through the tire rolling or sliding over a terrain. Fy is the lateral experienced by the tire during a cornering maneuver. The lateral force occurs when there is a difference in the direction of the wheel heading and travel. The angle about the tire center of this difference is the slip angle. In practical terms, a lateral force arises when a vehicle performs a cornering maneuver or experiences external lateral forces like crosswind. Lateral forces are considered a dynamic characteristic as it is directly proportional to the vehicle's lateral acceleration. , is the normal or vertical force that is also equivalent to the load experienced on the tire during regular operations. Mx is the overturning moment which is produced due to surface irregularities and tire cambering. The overturning moment is present due to the uneven distribution of vertical pressure throughout the tire width on the contact patch and creates a torque about the longitudinal direction My, is the rolling resistance moment in the case of a free-rolling tire and is the torque in the case of a driven tire. The rolling resistance moment occurs as there are more vertical forces near the leading edge of the tire, which torques against the tire's rolling direction. Mz is the aligning moment which occurs during cornering maneuvers and is generally equivalent to the slip angle of the tire. The self-aligning moment is a function of the pneumatic trail and is highly dependent on the contact patch and tread contact with the surface. The pneumatic trail is present due to cornering forces. The cornering force is behind the applied side force, and the difference between the two applied forces is the pneumatic trail. This difference generates a torque that tends to realign the tire in the direction of motion [7][8][9][10]. According to longitudinal vehicle dynamics, the normalized traction force for a tire, ρ, can be defined as [10]: Where the peak quantity of this normalized traction force is called the friction coefficient, μ. While all the discussed tire parameters play a role in tire wear models, the wear models discussed in this review focus on frictional forces including F x and F y . As many derivations for energy ______________________________________________________________________________ dissipation are relative these parameters and their dependent variables such as α.

Experimental Tire Wear Modeling
Experimental-based methodologies to estimate tire wear involve quantitative measuring tools such as sensor data. The data is used to then relate to tire-road frictional data to estimate a wear model. Khaleghian et al. [3] describe the experimental approach in three steps: obtaining wear-related parameters, correlating the data through wear-related formulations, and estimating the tire wear. In 2002, Grosch [11] developed an abrasion testing machine for empirical tire wear simulations. Grosch's experimental procedure uses a sample piece of a tire attached to a mount that rolls over an abrasive disc. As seen in Figure  3, this machine is titled the Grosch machine, or the Laboratory Abrasion and Skid Tester (LAT) 100. In this test, the lateral force at a given slip angle is observed. In Grosch's analysis, tire wear was quantified as a function of abrasion, lateral force, and speed. Abrasion is defined as energy dissipation in the contact area of the wheel as seen in below: Where abrasion, A, is a function of A uo , the abrasion energy loss/dissipation of the compound at reference energy U uo at experimentally set speed and n, the power index that is dependent on the tread and disc interaction. The energy dissipation, U is determined by the tire's lateral force and predetermined slip angle as in the equation below: = . sin Where F is the lateral force at a given slip angle, α. Abrasion was also formulated as a function of tire speed to analyze the wear at various operating conditions. The equation below shows the abrasion, A, relative to A uo , the abrasion energy dissipation of the compound at reference speed v uo , and the experimentally set energy dissipation level relative to slip angle and load, m [11].

= ( )
Grosch's studies showed that the LAT100 can determine wear rate factors comparable to real road tire wear data under limited conditions. Although the average datasets were comparable to actual tire data, the conditions in which the machine tested wear were within a finite range. This analysis used a small sample piece of tire that was not inflated and therefore the damping characteristics of a pneumatic tire were not included. Furthermore, the operating conditions were limited to small slip angles and low energy dissipation. However, this machine is advantageous as it can output many wear-related parameters in a short amount of time and was validated against limited real tire datasets.
In 2015, Ivanov et al. [12] conducted a tire wear study by looking at the wear of rivets embedded into a truck tire. In this study, button head rivets were embedded along the width of a truck tire as seen in Figure 4. The embedded rivets were weighed before the experimental procedure began and then once again postexperiment. The difference in weight was how this study analyzed the wearing along the width of the tire. In the experiment, the tires were sliding across an asphalt surface pulled by a tractor in a straight motion. This study showed that the wearing was most prominent in the middle path of the tire where the peak wear rate was 30.1 mg/km for increasing inflation pressures. However, the rivets on the side paths had higher wear rates with lower inflation pressures. This study showed that the patterns of wear rate along the width were comparable to the wearing of a tire, ______________________________________________________________________________ however, does not observe how much the tire material wears. Additionally, it was tested in braking conditions as the tire slides over the asphalt rather than a driven or rolling tire.
Another popular method of experimental tire wear modeling is analyzing the tire tread at the block level. In 2021, Wu et al. [13] analyzed an aircraft tire on a block level by varying the draft angle and root radius of the tread block as seen in Figure 5. The machine in this study used a normal loading mechanism for the tire sample and a horizontal feeding mechanism for the concrete and asphalt terrain. The wear width was found to peak at 5.5mm at lower draft angles.

Figure 5 Schematic of a Tire Block Diagram [13]
Similarly in 2021, Hartung et al. [14] also investigated the wearing of a tread block using a different mechanism. In this study, the sample tread was also pressed onto a test track and slid over the concrete terrain for a given velocity as seen in Figure 6. The tread samples post-procedure was processed using image processing tools to compare rubber profiles to evaluate tire wear along the contact surface of the tread. These profiles were then compared with Finite Element Analysis (FEA) models of the same tread block where the results aligned with the experimental quantities. At block level, the studies neglect damping effects that pneumatic tires inherently are equipped with as they are inflated with air. Additionally, these experimental tests are capable of estimating tire wear at particular conditions, however, it does not fully replicate a rolling or torqued tire as the tire tread slides across the terrain.
With the advances in sensor technology, many experimental tire wear studies have been conducted with the sensors attached to the tire carcass. Strain sensors are a popular tool to obtain tire forces within the carcass [15][16][17]. In 2016, Singh et al. [15] has done numerous studies using strain sensors on the sidewalls of the tire to correlate normal load and sideslip angles with algorithms to measure tire wear. In 2001, Miyazaki et al. [17] used multiple strain sensors in the tire to measure tire forces and compared these forces to a reference dataset to estimate tire wear. Pressure sensors are also used in tire wear studies such as those conducted by Hillenmayer et al. [18] in 2006. The pressure sensors were used to obtain a frequency response as the tire is driven over a surface. The distinct frequency response is then used to estimate the normal loading of the vehicle and determines the type of terrain the vehicle is rolling over which then predicts tire wear over the given surface.

Simulated Tire Wear Modeling
Simulated tire wear approaches are analysis of tire wear using theoretical, mathematical, or computer-aided simulated models. DaSilva et al. [19] used a similar approach of formulating wear with abrasion like in Grosch's work, but theoretically in the form of the bicycle model. This tire wear model was an estimation using only considering steady-state cornering maneuvers and a 2-degree-of-freedom bicycle model. In this model, any lateral load transfer was negligible, and no tire suspension characteristics were considered. Additionally, the model is only valid for low lateral accelerations. In this study, Silva describes tire wear as a function of steady-state longitudinal and lateral dynamic responses. This can be seen in the equation below, where W is the rate of wear, Ab is the abrasion of the tire-terrain interaction, Fy is the lateral force, α is the slip angle, S is the distance traveled by the vehicle during a maneuver, F N is the normal force, and F N0 is the reference normal force. This model depended on the bicycle model to estimate the lateral force and side slip angle [17]. Similarly, in 2018 Chen et al. [20] estimated tire wear using a 3-degree-of-freedom non-linear vehicle dynamic model. Chen used a power function that required the fitting of vehicle dynamic experimental data. To obtain the data for fitting, Chen used a hydraulic wear test system and cement terrain. The bicycle model can be seen to be a popular method for theoretical estimates of tire wear with the use of Kalman filters [21][22][23][24][25][26][27][28]. Where in many studies the Kalman filters are used to estimate side slip angles, cornering stiffnesses, yaw rates, speeds, and other tire forces derived from a bicycle model to analyze tire wear.
In 2021, Li et al. [29] developed a tire wear model based on a large number of simulated datasets. In this study, vehicle dynamic responses to vertical load, longitudinal speeds, and inflation pressure were evaluated. These vehicle ______________________________________________________________________________ responses are converted to modal frequencies which are fed into a three-layer neural network. It should be noted that this tire wear model is assessed for limited conditions and under steady-state rolling simulations.
In 2011, Chang et al. [30] developed an FEA tire wear model with a programmed subroutine using the Archard wear model. This model was a simple sliding abrasion relationship between two surfaces. This study analyzed the wear using Archard's wear model under straight free rolling and driven conditions and estimated tire wear as a function of mass and depth. In 2014, Zang et al. [31] combine a FEA tire wear testing and neural network tools to estimate tire wear. In this study, an FEA tire model using the Neo-Hookean tread material equipped with constant shear and unconditional stability is used in several varying simulations. This analysis looks at a particular material type and the neural network is trained for input variables of inflation pressure, speed, load, and different vibrational frequencies of the tire.
In 2022, Alfonse et al. [32] indirectly looked at the effects of tire wear using FEA. In this study, a truck tire with different groove numbers and tread depth to replicate the tire wearing to distinct levels was rolled over a given terrain at different operating conditions. This studied the rolling resistance coefficient of the tire at different tread depths over different moisture-content soils. The FEA truck tire model for different tread depths of a 3-grooved tire can be seen in Figure 7. In this study, the truck tire is operating over varying sandy loam soils, and it was concluded that the quarter-depth groove has the highest rolling resistance coefficient with respect to vertical loading. The effect of groove depth was significantly visible at higher soil moisture content. For example, at 62% moist sandy loam, the quarter groove tire exhibited 12% more rolling resistance coefficient than the full groove tire. The quarter groove tire model also exhibited a 3% more rolling resistance coefficient in comparison to the full groove tire model at 10 km/h tire speed and 40 kN vertical load. Generally, the full-groove and half-groove tire models have similar rolling resistance coefficients. It was also concluded that the effect of tread depth is greater than that of the number of grooves on the rolling resistance coefficient.

Figure 7 FEA Tire Wear Simulation Setup for 100%, 50% and 25% Groove Depth Truck Tires [32]
Mathematical tire models are used to analyze the connections between a tire's forces and moments with slip angles. Many different tire models have been developed throughout the years to better understand vehicle dynamics and estimate vehicle dynamic properties. With this being said, popular tire models such as the brush model have been adapted with tire wear models to estimate tire wear. The brush model, as its namesake, analyses the tire surface that is in contact with the terrain as bristles. Different regions in the contact surface experience different forces such as adhesion and friction forces. Three conventions are derived from the brush tire model: pure sideslip, pure longitudinal, and combined slipcases. Heinrich et al. [33] combined a hysteresis model with the brush model to analyze tire wear in an ABS braking situation. Specifically, this study looks at the dependence of normal loading and kinetic friction coefficient in the contact area of slipping tires. With the brush model, in an ABS braking condition, the shear deformation was assumed to be visualized in a parabolic pressure distribution in the contact region as seen in Figure  8. Where ξ is the longitudinal displacement of a tire, l is the contact length of the tire tread, pm is the maximum value of contact pressure distribution and a is the breakaway point. Yamazaki et al. [34] estimated tire friction similarly, however only using the partial sliding and pure sliding conventions of the brush model. On a similar note, in 2022, Lu et al. [35] developed a semi-empirical model using experimental values, tire footprint pressure distributions, and the brush model to estimate a tire block's wear. However, in 2011, Matilainen et al. [36], combined the brush model with the bicycle model to estimate tire wear friction. Specifically, the study looked at the potential friction the tire may encounter during a lateral driving maneuver using the bicycle model. Then using the tire forces acquired from the bicycle model to estimate friction using the brush model, however, assuming negligible tire stiffness.

Tire Road and Wear Particle Emissions
TRWPs are considered microplastics where the particles are smaller than 5 mm [37]. Compared to other microplastics, the literature in which TRWPs are studied as emissions is not studied rigorously [3]. However, due to alarming amounts of TRWPs found in different environments, the increase of studies regarding this topic has ______________________________________________________________________________ increased. In 2018, Wagner et al. [38] claimed that TRWPs contribute to over 50% of microplastic emissions in Denmark and Norway. Some studies show that TRWPs amount to the majority of microplastics found in some marine environments [38]. TRWPs can be mostly found to be Styrene Butadiene Rubber (SBR), Butadiene Rubber (BR), and Natural Rubber (NR) [39]. Where SBRs and BRs are mainly found in passenger car tire compositions and NRs in Truck tires [40]. Out of the two types of TRWPs, NRs are known to be more wear and impact resistant [41]. However, studies have also shown that truck tire wear emissions are a fraction of passenger car tire emissions at approximately a 1:11 ratio [38,42].
Park et al. [43] classified TRWPs into five distinct categories depending on the particle's geometry and size. This can be seen in Figure 9, where Type 1 particles are TRWPs in elongated tube shapes that are formed most commonly and are found at the tire-terrain contact. Type 2 particles are spherical in shape, averaged at around 2.5µ m. Type 1 and type 2 particles make up the majority of TRWPs compared to the five types in this study. The type 1 particle was seen to form into type 2 particles through secondary wear processes. Type 3 TRWPs are irregular in shape and are on average larger than type 1 and type 2. The study suggests that type 3 particles originate from severe tireterrain contacts such as critical cornering or braking situations where much higher friction between tire and terrain is experienced. Type 4 and 5 particles are much smaller in size compared to the other types and are formed by combining or adhering with other larger particles. The location in which TRWPs are concentrated can be variable as well, depending on roadways, waterways, and different paths the TRWPs may maintain location or move. That being said, studies do show that TRWPs can be mainly found in urban driving areas, compared to highway driving areas [44]. The frequency of braking, accelerating, and encountering varying road characteristics are much higher on urban roads than on highways [43]. This being said, TRWP emissions can be categorized into three locations: ambient air, water, and soil. The emissions through air can be taken in by humans and animals through inhalation, or just into the nearby environment through deposition.
Emissions into waters and soils can pose ecotoxicological risks or transport the TRWPs into different areas through runoffs. Some studies have estimated the origin of TRWPs found in road dust, sediment, and marine salt [44]. One study shows that car tire tread contributes most of the TRWPs as seen in Figure 10, which is concurrent to that of some studies [40,41], where Car Tire Treads and Truck Tire Treads are CTTs and TTTs, respectively. Although different studies estimate the ratio differently between CTTs and TTTs, it is undoubtedly that CTTs output more TRWPs than emissions of the two. Though present, TRWPs are not majorly found in ambient air where only 0.1-10% of TRWPs end up airborne [44,45]. Many TRWPs can be found in road dust and soils up to 210g/kg dry weight [47]. It was found that there are more pollutants found within tunnels compared to outside tunnels due to the dilation of TRWPs outside of a concentrated area. Another study showed that there can be found up to 117 g/kg dry weight of TRWPs in soils close to automobile roads and can be seen to drop to below 80% of microplastic emissions at 30m away from the road [49]. TRWPs can also be found in sewage sludge that may be applied to agricultural soils. In 2003, Fjallborg et al. [46] estimated the number of TRWPs may end up in agricultural soils by extrapolating the zinc emissions from tires into agricultural sewage. In road runoffs and receiving waters, studies show that TRWPs can be found from 0.3 to 197 mg/l [47].

______________________________________________________________________________
Another study studied the TRWP concentration in five separate locations in runways and receiving waters through three stages of a storm. It was seen that there was two times less TRWP concentration in the receiving waters post-storm event indicating that the tire wear particles were flushed away from the initial concentration area [50].
It was noted that estimating tire wear particles through chemical indicators and chemical sampling may be inaccurate to how much TRWPs are actually emitted through vehicles [41]. Capturing as many TRWPs in real time may provide a more accurate estimate of TRWPs emissions. As such, one study [43] looks at capturing the TRWPs within an enclosed space to look at the TRWP size, mass, and type as seen in Figure 11. Similarly, in 2021, Tonegawa et al. [52] developed an indoor tire wear tester and an on-road tire wear tester to compare results. The on-road tire wear tester can be seen in Figure 12, where a nozzle is placed behind the tire to suction the generated tire abrasion and dust. The particles were filtered to type 1 particles and other types to compare results with the control indoor test machine. It was found that for a constant speed, the tire wear particles were similar in size, however at large lateral accelerations the results were dramatically different.

Figure 12 Enclosed Tire Road and Particle Wear Capturing
System [52] 6. Conclusion In conclusion, tire wear particles are a significant source of microplastic emissions that have been found to contribute to over 50% of microplastic emissions in some countries. Tire wear particles are composed of various materials, including Styrene Butadiene rubber, Butadiene rubber, and natural rubber. The concentration of tire wear particles can vary depending on location, with urban driving areas having higher concentrations compared to highways. Tire wear particles can be found in ambient air, water, and soil, posing potential ecotoxicological risks. While estimating tire wear particle emissions through chemical indicators and sampling may be inaccurate, capturing tire wear particles in real-time provides more accurate estimates. Developing solutions to mitigate tire wear particle emissions is necessary to reduce the impact of microplastic pollution on the environment and human health. Currently, there is much research developed in the field of the tire wearing through experimental methods, modeling methods, and semi-empirical methods. In literature, much experimental and modeling research is done on estimating tire wearing for dynamic and control problems through sensors and tests. These methods are accountable to estimate tire wear, however, may have trouble extrapolating their results as tire wear particle emissions. To account for this, estimating tire wear particles can be done by looking at indicators and markers of the wear particles in surroundings such as soil, water, and roads. However, due to different environmental factors, it may be difficult to discern the exact origins of the particulate, whether they be found in tires or other sources that are further away from automobile roads. Thus, it is important to continue research on estimating tire wear and methods of capturing tire wear to further reduce harmful emissions through tire wear particles. It can be seen that wear emissions estimated through trackers can be inaccurate, and experimental wear modeling can become expensive and time-consuming. Current simulated tire wear models in the literature are still simple and developed under limited operating conditions. Introspectively, a more optimized and capable tire model to simulate the tire wear model may be developed alongside a driver model in order to obtain a more accurate tire wear emissions value.