Road accidents increase during the first rain after a long dry period. It is widely accepted that the tire traction loss is due to fine particles accumulated on the road surface. Yet, the involved mechanisms at the tire/road interface are not clearly understood. This paper deals with the particle-induced lubrication on dry road. Tests are performed in laboratory. Sediments are extracted from a catchment area which collects runoff water. The process of extracting particles from sediments in laboratory by drying and sieving is described. Particles are separated into fractions characterized by their chemical composition and size distribution. Protocol to simulate the process of particles' build-up on the road surface is described. The effect of the particle's size and concentration is studied. Surfaces representative of real road surface textures are tested. Friction tests are conducted using the Skid Resistance Tester, which is widely used to assess friction characteristics of road surfaces. Samples are weighed before and after each friction run. On a surface initially covered by a compacted particle layer, successive friction runs are performed, without resupplying particles, to follow the evolution of the friction coefficient with the particle depletion. Friction drops significantly when a surface is covered by particles. Successive runs induce an increase of the friction coefficient until reaching a stable value which is still lower than that of a clean surface. Particles are considered as third bodies introduced artificially in the interface between the friction slider and the test sample. Analysis of particles' flows (particles trapped and ejected), through particles' mass extracted from weighing, helps to explain the friction variation. Assuming that friction is governed by the surface fraction (X) covered by particles, a model was developed. The calculated friction coefficient is a weighted sum of friction coefficients when the surface is respectively clean (X = 0) and fully covered by particles (X = 1). It was found that the surface fraction can be expressed as a function of the ratio of available particles' mass to initial particles' mass. Comparisons between model and experiment are satisfactory. Discussions are made in terms of interactions between surface textures and particles' size and concentration.

est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d'enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.

Road skid resistance decreases under traffic and climate actions. For a surface made of bituminous concrete, the skid resistance evolution involves a removal of the bitumen layer and a polishing of the aggregates. Previous researches have been focused on polishing mechanisms and less is known about bitumen removal phase. This paper presents a laboratory study to better understand the behavior of the stone mastic asphalt under simulated actions of traffic. Due to its formulation, this bituminous concrete contains more bitumen than other materials used for road surfacing. Polishing tests are performed on circular cores using tWehner/Schulze machine, which simulates the traffic-induced polishing by means of rolling rubber cones and measures the tire friction by means of sliding rubber pads. Tests are stopped at predefined numbers of passes for friction measurements, 3D cartographies of surface texture and scanning electron microscopic observations of the cores' surface. Statistical tests are performed to identify relevant texture parameters to explain friction evolution. During bitumen removal phase (i.e. before 20,000 passes of polishing), Ssc (average curvature of peaks) and Sdq (average quadratic slope of asperities) are correlated with friction values. After 50,000 passes, volume parameter Vvc is more adapted due to the abrasion of asphalt mix surface. Then, scanning electron microscopic observations show that a layer composed by bitumen and particles (small aggregates and sand) is a third-body surrounding aggregates, which evolves in thickness and size of aggregates during polishing. The system composed of the aggregates, the bitumen–particles layer and the polishing cones can be assimilated to a tribosystem with a tribological circuit at the interface. The bitumen–particles layer acts as an internal flow, before leaving the contact area. The movement of the layer under the shear stress induced by the polishing cones explains the ejection of big aggregates and the heterogeneity of the layer around aggregates.

Previous studies highlighted that friction value measured on real road surface textures covered by particles depends on particle sizes. This paper focuses on identification of particles behaviours at the tyre/road interface in the presence of particles. Identification is made by visual observations through high-speed camera, focus-variation microscopy and Scanning Electron Microscopy on the contaminated surface as scratch analysis on the surface. Two particle types were collected into samples picked next to roads. They are composed of clay and quartz which have different behaviours at the interface and affect the friction. Clay has a lasting effect due to its adhesive behaviour. Quartz has a scratching and rolling behaviour and is quickly ejected to the contact area due to a pinching effect.

This paper presents a laboratory study to investigate the lubrication induced by particles deposited on road surfaces which is responsible for the increase of accidents at the first rain after a long dry period. Particles are extracted from sediments collected at a catchment area and characterized by their chemical composition and size distribution. Protocol to simulate the particle deposit on the road surface and their compaction by the traffic is described. The test program includes variables such as particle concentrations, particles' size fractions and surface textures. Dry friction tests are conducted using the Skid Resistance Tester which simulates the sliding friction between a rubber slider and the test surface. Friction is lowered when the surface is covered by particles, compared with a clean surface. Repeated passages of the slider induce an increase of friction coefficient until reaching a stable value. Particles' flows during a friction run are separated into those ejected from the sample, those stored by the surface macrotexture and those that stay on the test surface; this study demonstrated that the friction coefficient is closely related to the quantity of particles available on the test surface. SEM analysis shows on the other hand that these particles are mainly trapped by the surface microtexture. Similar behavior between fine particles and powder in terms of lubrication mechanisms is observed.

This paper deals with the modeling of the coverage of surfaces contaminated by fine particles, the objective being the prediction of the skid resistance of road surfaces when it rains after a long dry period. The research methodology is based on the identification of particles' flows in the tribological circuit composed of the particles (3rd body) and the tire and the road (1st bodies). Experiments are conducted in laboratory where sliding friction is measured between a rubber pad (simulating a tire tread block) and a sandblasted aluminum surface (simulating a microtextured road surface) covered by particles. The test program includes particle concentrations representative of deposits of particles on the road surface at different dry periods and different particle size fractions. The test protocol consists in repeating passages of a rubber pad on the test surface and visualizing the particles' movements by means of high-speed cameras. Two particles' flows are identified: particles ejected from the contact and those raised by the rubber pad then fall back to the surface. Similarities are established with the removal and deposition of gas species in vapor-phase lubrication. An analytical model was derived to express the surface coverage as a function of the number of passages and two parameters (values between 0 and 1) called respectively the ejection and recirculation ratios. The proposed model is included in a linear rule-of-mixtures equation for the calculation of friction. Calculated friction coefficients compare favorably to experimental data and the model's parameters are determined. A master curve for different particles' sizes is obtained when relating the surface coverage to the mass of particles. Relationships between the ejection and recirculation ratios, the particles' characteristics and the surface texture are presented. Discussions are made in terms of transposition of the model to real road surfaces.

No abstract