Traffic and Granular Flow ’03

Conventional traffic flow theory dictates that flow on a freeway is usually constrained only by a small number of critical locations or bottlenecks. When active, these bottlenecks cause queues that can stretch for several miles and reduce flow on other parts of the network. Bottlenecks are often thought to arise over short distances and are usually modeled as if they occur at discrete points since the resulting queues are thought to be much longer then the bottleneck region. This paper presents evidence that the delay causing phenomena may actually occur over extended distances. Some of which may occur downstream of the apparent bottleneck where drivers are accelerating away from the queue, while related phenomena are observed in the queue, over a mile upstream of the apparent bottleneck. It is shown that lane change maneuvers are responsible for some of the losses, reducing travel speed and consuming capacity when vehicles enter a given lane. These losses in one lane are not fully balanced by gains in other lanes.

This contribution reports a recently recorded data-set that helps to understand the car-following process better. Since the empirical basis of most traffic flow models can be called weak, these findings may help to design better models. They demonstrate, that the process of car-following seems to have a very interestic stochastic dynamics. Especially, and different from most of the existing models the data show clearly that car following cannot be described by a noisy fixed point dynamics. This is because the acceleration of the cars is smooth and roughly constant for a certain time, with fast changes to a new value at so called action points.

Furthermore, the data can be used for calibration and validation purposes of the various models. This has been done with very interesting results indicating that all the models tested behave very similar and cannot describe the data better than with 15 % difference between models and data.

We present an analysis of traffic data of the highway network of North-Rhine-Westphalia in order to identify and characterize the sections of the network which limit the performance, i.e., the bottlenecks. It is clarified whether the bottlenecks are of topological nature or if they are constituted by on-ramps. This allows to judge possible optimization mechanisms and reveals in which areas of the network they have to be applied. Our results support previous empirical observations and theoretical studies indicating that the overall travel-time of vehicles in a traffic network can be optimized by means of ramp metering control systems.

The importance of physical viewpoint for understanding the phenomena of freeway traffic is emphasized using the simulations of a mathematical model and the experiment. The traffic flow is understood as a many-body system of moving particles with the asymmetric interaction and the emergence of jam is the pattern formation as the phase transition of non-equilibrium system by the effect of collective motions.

The temporal data of the expressway traffic data are complex mixtures of various time scales. The periodic appearances of the congestion, for example, reflect social activities. We need some filters to extract proper fluctuations from the raw data. We employ the method of detrended fluctuation analysis for studying the time sequence of the traffic flow. We find the long-range correlation from 1 hour to 24 hours.

The aim of this paper is to present recent progress in calibrating ten microscopic traffic flow models. The models have been tested using data collected via DGPS-equipped cars (

D

ifferential

G

lobal

P

ositioning

S

ystem) on a test track in Japan. To calibrate the models, the data of a leading car are fed into the model under consideration and the model is used to compute the headway time series of the following car. The deviations between the measured and the simulated headways are then used to calibrate and validate the models. The calibration results agree with earlier studies as there are errors of 12 % to 17 % for all models and no model can be denoted to be the best. The differences between individual drivers are larger than the differences between different models. The validation process leads to errors from 17 % to 22 %. But for special data sets with validation errors up to 60 % the calibration process has reached what is known as “overfitting”: because of the adaptation to a particular situation, the models are not capable of generalizing to other situations.

In this paper we present results relative of our cellular automata simulation on Italian highway traffic. Based on the previous two lane highway model that we have realised some time ago this is a three lane model without periodic boundary conditions. In particular this model has on and off-ramps; it provides drivers with an unbound braking ability in order to avoid car accidents and it can take account of road works and lane reductions. It implements the Nagel and Schreckemberg rules and it can reproduce well known experimental phenomena observed in real traffic, such as the typical three regions in the fundamental plot, lane inversion or jams that propagate backward like waves.

In this paper we present results obtained with an original traffic data collector named BirdEye, comprising infrared emitters-detectors, microcomputers using a linux operating system and GSM or cable communication systems. It is currently under test, below a portal, along the A1 Italian highway, between Bologna and Florence. Data analysis has revealed the presence of the various phases of traffic; the formation of traffic slowing downs and jams and the highway overcrowding by lorries during working days. In particular we have analysed speed and length spectra, flow and fundamental plot, which reproduces experimental phenomena known from literature. Finally, we present results on the time sequence of vehicles transit which confirm the hypothesis of the existence of a fractal dimension of traffic.

There is discussion if traffic displays spontaneous breakdown. This paper presents computational evidence that stochastic car following models can have a control parameter that moves the model between displaying and not displaying spontaneous phase separation for some densities. Those phases can be called “laminar” and “jammed”. Models with spontaneous phase separation show three states as a function of density: a first state at low density, where those models are homogeneously laminar; a second state at high density, where they are homogeneously jammed; and a third state at intermediate density, where they consist of a mix between the two phases (phase coexistence). This is the same picture as for a gas-liquid transition when volume of the gas is the control parameter.

Although the gas-liquid analogy to traffic models has been widely discussed, no traffic-related model so far displayed a completely understood

stochastic

version of that transition. Having a stochastic model is important to understand the potentially probabilistic nature of the transition. Most importantly, if indeed models with spontaneous phase separation describe certain aspects correctly, then this leads to an understanding of spontaneous breakdown. Alternatively, if models without spontaneous phase separation describe these aspects better, then there is no spontaneous breakdown (= no breakdown without a reason). Interestingly, even models without spontaneous phase separation can still allow for jam formation on small scales, which may give the impression of having a model with spontaneous phase separation.

Resolution of LWR model is considered. The method proposed in this paper differs from (continuous) characteristic based analytical resolutions or from (discretized) finite difference schemes. It is based on an approximation of the flow-density relationship which yields a solution with piecewise constant density. This solution is then exactly calculated by tracking waves and handling their collisions. Extensions for incorporating boundary conditions such as traffic signals or discontinuity of the flow-density diagram are considered and provide an illustration of the potentiality of the method.

This paper is an analysis of the various scales at which traffic flow can be represented, from vehicular to continuum flow models, with various time and space resolutions. The paper first investigates the question of scales and scale separation in traffic flow, both in modelling and measurement, using analogy with other disciplines. It then evaluates the representation of vehicles heterogeneities at various scales. It concludes on the interest of developing multiscale modeling.

Based on a microscopic theory of spatial-temporal congested traffic patterns at freeway bottlenecks, diagrams and features of the congested patterns at bottlenecks due to on-ramps, merge bottlenecks (a reduction in freeway lanes), and off-ramps are studied and compared.

Congested traffic patterns at a symmetric merger of two single-lane roads into one road are studied in the KKW-model, a cellular automaton that traces all observed congestion patterns back to two basic concepts: A typical distance below which a driver decides to adjust his speed to the one of the car in front, and the minimal distance between cars that is compatible with secure driving. The diagram of congested patterns in the flow-flow plane whose co-ordinates are the inflow rates onto the two merging roads is determined. Depending on the parameters, every congestion pattern can occur on either road, but not all combinations of patterns on both roads are realized. There is no indication of spontaneous symmetry breaking: Equal demand on both roads leads to the same type of pattern on both sides.

The transport of products between different suppliers or production units can be described similarly to driven many-particle and traffic systems. We introduce equations for the flow of goods in supply networks and the adaptation of production speeds. Moreover, we present two examples: The case of linear (sequential) supply chains and the case of re-entrant production. In particular, we discuss the stability conditions, dynamic solutions, and resonance phenomena causing the frequently observed “bullwhip effect”, which is an analogue of stop-and-go traffic. Finally, we show how to treat discrete units and cycle times, which can be applied to the description of vehicle queues and travel times in freeway networks.

We present a macroscopic traffic flow model that explicitly builds on continuous individual driving behaviour. Not only do we start from classical car-following rules (like the kind that are used in microscopic simulators), the model also explicitly accounts for the finite reaction times of drivers, anticipation behaviour, anisotropy in driver responses and the finite space requirement of drivers in the stream. Moreover we allow variations in driver psychology, which lets drivers adopt different ‘driving styles’ dependent on traffic conditions, like the presence of a merging zone.

We illustrate the potential of such model by simulating a busy highway with an on-ramp. Plausible assumptions about driver psychology allow us to reproduce the so-called ‘capacity funnel’, i.e. the onset of congestion typically occurs some distance downstream of the merge area.

In this paper we describe a relation between a microscopic stochastic traffic cellular automaton model (i.e., the STCA) and the macroscopic first-order continuum model (i.e., the LWR model). The innovative aspect is that we explicitly incorporate the STCA's stochasticity in the construction of the fundamental diagram used by the LWR model. We apply our methodology to a small case study, giving a comparison of both models, based on simulations, numerical, and analytical calculations of their tempo-spatial behavior.

From

probabilistic point of view

we investigate a quite classical dynamical system given by stochastic differential equations, i. e. ordinary differential equations driven by multiplicative noise. Based on this Langevin approach the probability density distributions of vehicular velocities as well as headway distances are calculated and discussed.

Our work is a continuation of a stochastic theory of freeway traffic based on a Master equation approach presented first at

Traffic and Granular Flow

'97 as the one—cluster model. The extension to our multi—cluster model can be found at

Traffic and Granular Flow

’99.

A new type of noised-induced phase transitions that should occur in systems of elements with motivated behavior is considered. By way of an example, a simple oscillatory system

$$ \left\{ {x,v = \dot x} \right\} $$

with additive white noise is analyzed numerically. A chain of such oscillators is also studied in brief.

Since the year 2000, the Centre of Applied Informatics and the Institute for Transport Research at the German Aerospace Centre develops a microscopic road traffic simulation package named ’SUMO’ — an acronym for “Simulation of Urban MObility”. Meanwhile, the simulation is capable to deal with realistic scenarios such as large cities and is used for these purposes within the institute's projects. The idea was to support the traffic research community with a common platform to test new ideas and models without the need to reimplement a framework that handles road data, vehicle routes, traffic light steering etc. To achieve this goal, the simulation code is available as open source. Within this publication, we would like to demonstrate how most attributes of traffic flow can be simulated. This should be mainly interesting for educational purposes.

We show that some simple urban traffic flow equations can be approximated by equations which are equivalent to a Schrödinger equation. For a simulation of the Schrödinger equation as well as for analytical computations it is useful that waves of traffic which travel along a road are not reflected at the boundaries of the simulated region. We present the non-reflecting boundary condition for a corresponding one-dimensional Schrödinger equation, and show simulation results for a wave package of traffic moving towards such a boundary.

Continuum models for traffic flows may serve as a tool for the optimal management of traffic. From the mathematical point of view, the development of a control theory for conservation laws is still at its beginning. From the engineering point of view, both the optimality criteria and the selection of the controllable parameters seem far from being standardized.

This note presents first some recently introduced continuum models, then some optimal management problem are discussed.

This note discusses the role of source terms in the modeling of vehicular traffic through conservation laws. As is well known, a source term in the equation for the vehicular density may represent entries or exits. When a second conservation laws is present, suitable source terms may describe various inhomogeneities of the road, such as ascents, descents or fog banks. In each of these cases, we provide an analytical framework. Finally, numerical experiments show specific phenomena that may not be reproduced by purely conservative models.

The aim of this article is to study how the obstruction caused by slow vehicles (like trucks for example) on other ones can be modelled in a two-flow macroscopic model based on the Lighthill-Whitham-Richards (LWR) theory. This will be done by using the results of moving bottleneck models which describe the effect of a single slow vehicle on the rest of the flow.

This contribution re-visits the classical optimal-velocity model of Bando

et al

[

1

], describing vehicles on a loop, and explains the different wave types and meta-stable behaviour from the mathematical viewpoint of bifurcation theory. We apply the numerical continuation package AUTO for a rapid derivation of phase diagrams to investigate the relation between coexisting states, flow transitions and hysteresis in traffic flow. Some extensions which involve (i) multiple look-ahead effects and (ii) delay are also discussed.

A traffic model based on a cellular automaton (CA) approach with the focus on limited braking capabilities is presented. It is shown that the known empirical features of traffic flow can be reproduced including even the so called

pinch effect

being the object of actual investigations.

The object of the paper is to analyze intersection modeling in the context of macroscopic traffic flow models. The paper begins with a brief review of classical boundary conditions of the Dubois-LeFloch and the Bardos-Nédélec-LeRoux type, and their relation to the concepts of local traffic supply and demand. It will be shown that the local traffic supply and demand concept extends and simplifies these classical approaches. The resulting constraints on phenomenological intersection models will be discussed. Several examples of intersection models are deduced. Some of these recapture earlier models; others are specifically designed for congested traffic conditions and take into account the bounds on car acceleration. The last part of the paper is devoted to network modeling and to applications to network traffic management.

The present situation in transport policy is that many authorities have a responsibility for a part of the transport system and that each authority has its own goals and priorities. Central steering is only on main policy goals. Public transport and municipal, regional, provincial and national road authorities have their own policies, targets and instruments. The interaction between different policies themselves exists indirectly, but policies are ‘granular’.

In this paper the granular transport policies are described as a game with many players. Integrated traffic control and traffic assignment problem are studied for a situation with two road authorities. The road authorities try to optimize their own objectives and the same is done by the travelers. This leads to a two-level three-player, multi-stage optimization problem with complete information. Game theory gives a suitable framework to analyze the problem and to find solutions for different situations, e.g. no cooperation, cooperation between the two authorities and a system optimum where all actors cooperate to minimize the total costs for all travellers.

In this paper two approached are used: an analytical one and an approach based on a simulation and assignment framework. Both approaches are described and used to study a simple example, for which the results are given and discussed.

The results show that separate or integrated anticipatory control gives better results than iterative reacting to the current situation. If one road authority takes the lead and anticipates the reactions of both the road users and the other road authority a sub-optimum is reached. The model calculations give evidence that cooperation of road authorities improves the utilization of the infrastructure and that a global optimization does not necessarily give a worse situation for one road authority.

The paper presents a real-time motorway network traffic surveillance tool called RENAISSANCE that enables traffic state estimation and short-term prediction, travel time estimation and prediction as well as queue length estimation and prediction based on a limited amount of real-time measurements. Both simulation testing and real data testing were conducted to evaluate the RENAISSANCE performances.

The paper reports laboratory experiments on a

minority game

with two routes. Subjects had to choose between a road

A

and a road

B

. Nine subjects participated in each session. Subjects played 100 rounds and had to choose between one of both roads. The road which the minority of players chose got positive payoffs. Two treatments with 6 sessions each were run at the Laboratory of Experimental Economics at Bonn University. Feedback was given in treatment I only about own travel time and in treatment II on travel time for road

A

and road

B

.

The paper reports laboratory experiments on a

day-to-day route choice game

with two routes and alternating construction areas on both routes. Subjects had to choose between a main road

M

and a side road

S

. The capacity was in every period greater for the main road. 18 subjects participated in each session. In periods without construction areas the equilibrium is 12 players on

M

and 6 players on

S

. In periods with construction areas on M the equilibrium is 10 players on

M

and 8 players on

S

. In periods with construction areas on S the equilibrium is 15 players on

M

and 3 players on

S

. Two treatments with 6 sessions each were run at the Laboratory of Experimental Economics at Bonn University. Feedback was given in treatment I only about own travel time and in treatment II on travel time for

M

and

S

. Money payoffs increase with decreasing time. Subjects are told that in each of 200 periods they have to make a choice between the two routes.

A recently introduced cellular automaton model for Internet traffic is investigated in the context of boundary induced phase transitions and the appropriate phase diagrams. Since the model allows multi allocation of sites there are some deviations in the internal dynamics as well as in global properties compared to known driven lattice gas models. As a consequence the yielding phase diagram derived by numerical simulations reveals some interesting new features like a capacity shift in dependence to the allocation number.

The impact of adaptive traffic light control is studied in the Chowdhury Schadschneider (ChSch) cellular automaton (CA) model for city traffic. Therefore, three adaptive strategies are presented being able to react flexible to the traffic conditions. It is shown that the adaptive control is capable to direct the system to its

system optimum

. Furthermore, the impact inhomogeneous densities is investigated with the aim to demonstrate the difference between a global traffic control and an adaptive one if varying traffic conditions are considered.

A computer simulation of a simple dynamical model of vehicles moving across a two-dimensional traffic network is performed. Traffic management is enforced through the roundabouts: providing the conditions are appropriate, vehicles can move through the roundabouts. When the density of the travelling traffic is low, the vehicles can move freely.

Increasing the vehicle density leads to structured patterns which initially enable a high density of vehicles to travel throughout the system but eventually leads to gridlock. A phase transition separates a congestion free phase from a congested one as the vehicle density increases.

Many insects like ants communicate chemically via chemotaxis. This allows them to build large trail systems which in many respects are similar to human-build highway networks. Using a stochastic cellular automaton model we discuss the basic properties of the traffic flow on existing trails. Surprisingly it is found that in certain regimes the average speed of the ants can vary non-monotonically with their density. This is in sharp contrast to highway traffic. The observations can be understood by the formation of loose clusters, i.e. space regions of enhanced, but not maximal, density. We also discuss the effect of counterflow on the trails.

This paper discusses basic findings on crowd movement and their application to simulation models. This includes empirical and experimental results concerning group behavior, pedestrian motion, and emergency egress. Next to a literature review, we will present own empirical investigations on walking speed distribution and the dependency of walking speed on group size.

The second part of the paper relates these findings to modelling and simulation of crowd movement. This comprises the representation of behavior, calibration and verification and the connection to many particle systems.

Finally, we will present an extension of the current microscopic theory (basically underlying all the “individual” models used for real world applications). This includes route-choice behavior and links microscopic and macroscopic behavior.

Microscopic simulation models predict different forms of self-organization in pedestrian flows, such as the dynamic formation of lanes in bi-directional pedestrian flows. The experimental research presented in this paper provides more insight into these dynamic phenomena as well as exposing other forms of self-organization, i.e. in case of over-saturated bottlenecks or crossing pedestrian flows. The resulting structures resemble states occurring in granular matter and solids, including their imperfections (so-called vacancies). Groups of pedestrians that are homogeneous in terms of desired walking speeds and direction appear to form structures consisting of overlapping layers. This basic pattern forms the basis of other more complex patterns emerging in multi-directional pedestrian flow: in a bi-directional pedestrian flow, dynamic lanes are formed which can be described by the layer structure. Diagonal patterns can be identified in crossing pedestrian flows. This paper both describes these structures and the conditions under which they emerge, as well as the implications for theory and modeling of pedestrian flows.

We provide a comparison of the relative merits of video and infrared based methods for collecting pedestrian movements from the real world and also from experimental environments. We describe the underlying technological basis of both methods and the tools we have developed to help in collection and analysis of the data. The desire to collect such data is driven by the need of modellers and simulation packages to use base data that is founded in valid empirical evidence, rather than some form of inspired supposition, as is the case with many of the current systems. In addition to the collection of speeds we are also interested in understanding and quantifying the ranges of distances people deviate from a straight-on path when confronted by some obstruction in front of them.

We consider a numerical model of pedestrian movement in a counter flow footway using a cellar automaton model. The enter rate of a pedestrian into the footway is the control parameter in our study. The probability of the panic occurrence is calculated statistically. If the width of footway is large, we find critical behavior of the panic probability which is similar to the phase transition obtained from statistical mechanics. We also consider the asymmetric case where the number of pedestrian moving one direction is different form that of the couter direction.

We propose a dynamical model for collective biological motion, which is based on 2-dimensional optimal velocity model. The property of the model is investigated by numerical simulation. Various patterns of group emerge by changing parameters of the model and density of organisms.

The floor field CA model for studying evacuation dynamics is extended in this paper. A method for calculating the static floor field, which describes the shortest distance to an exit door, in an arbitrary geometry of rooms is presented. The wall potential and contraction effect at a wide exit are also proposed in order to obtain realistic behavior near corners and bottlenecks.

Based on flow rate curves collected by two synchronized and separated cameras, the propagation and dispersion of pedestrian pulses were investigated. Microscopic simulations were performed to reproduce the real data. By means of a simple macroscopic analysis, it is possible to find microscopic parameters of the social force model. We found, for the case study that the ranges of these parameters are:

A

= 900−1500 N and

B

= 0.8−1.2 m.

Based on data collected by a turnstile line of an important subway station, the problem of decisions making was studied. Microscopic simulations were performed to reproduce the real behavior. A simple selection model was proposed to simulate how persons choose between different possibilities. The functionality of the criteria consider how far and how populated are each turnstile. The model was able to reproduce the utilization curve of a given layout and to predict the curve for a different geometry.

Computer simulations have become an important tool for analysing egress processes and assessing evacuation concepts. Especially so called microscopic models can by now be considered state of the art. In this paper we will describe the software PedGo which is based on a 2D cellular automaton and its application to the simulation of evacuations form large and complex structures. The focus is on the practical application to full-scale scenarios. As an example, we show results for the egress from a football stadium.

We review recent work characterizing force fluctuations and transmission in dense granular materials. These forces are carried preferentially on filimentary structures known as force chains. When a system is deformed, these chains tend to resist further deformation; with continued deformation, chains break and rearrange, leading to large spatio-temporal fluctuations. We first consider experiments on force fluctuations, diffusion and mobility under steady-state shear. We then turn to force transmission in static systems as determined by the response to a small point force. These experiments show that the packing structure and friction play important roles in determining the force transmission. Disordered highly frictional packings have responses that are similar to that of an elastic solid. Ordered packings show responses that may be described either by anisotropic elasticity or by a wave-like description.

The behaviour of granular material is strongly dependent on the stress level, so that small scale tests simulating large granular systems are not reliable in many cases. The stress dependent behaviour, however, can be simulated by placing the small scale test in a centrifuge. Due to the artificial gravity the stresses gradient in the small model can be made similar as in the prototype, resulting in a realistic behaviour of the granular material. Almost all types of granular systems can be investigated by this testing technique.

Hydrodynamic equations of motion for a monodisperse collection of weakly frictional spheres have been derived from the corresponding Boltzmann equation, using a Chapman-Enskog expansion around the elastic smooth spheres limit. The hydrodynamic fields required in this case are: the velocity field,

V

(

r

,

t

), the translational granular temperature,

T

(

r

,

t

), and the (infinite) set of number densities,

n

(

s, r

,

t

), corresponding the continuum of values of the spin,

s

. An immediate consequence of these equations is that the asymptotic spin distribution in the homogeneous cooling state for nearly smooth, nearly elastic spheres, is highly non- Maxwellian.

In this article we review the two basic mechanisms responsible for overcharging a single spherical colloid in the presence of aqueous salts. The first mechanism rests on energetic arguments, and deals only with the ground state configuration of the counterions. This mechanism can be qualitatively explained in very simple terms using freshmen electrostatics, and a very good quantitative description can be obtained using Wigner crystal theory. The second mechanism is driven by a combination of excluded volume correlations and the preceeding energetic arguments, and in the following called entropic overcharging. We demonstrate the validity of the proposed mechanisms with results of molecular dynamics simulations within the primitive model of electrolytes.

In dense arrangements of magnetic grains, cutting off the interaction potential gives a possibility to accelerate simulation algorithms. We argue that

R

≈ 5 particle diameters is a reasonable choice for dipole-dipole interaction cutoff in two-dimensional dipolar hard sphere systems, if one is interested in local ordering. As an application, we performed computer simulations based on a two-dimensional Distinct Element Method to study granular systems of magnetized spherical particles. The effect of the magnetization on the angle of repose, on the surface roughness of piles, and on particle avalanches were studied. We found a smooth transition in the avalanche formation from a

granular regime

to a

correlated regime

controlled by the magnetic interparticle force. This observation underlines the analogies between systems with magnetic and adhesive forces.

This paper concerns with a solution concept for three-dimensional simulations of the transport of discrete particles and their deposition to irregular surface structures. The gas phase, governed by the Navier-Stokes or Stokes equations, and potential fields of electrical forces are solved by lattice-Boltzmann methods. The motion of particles is described by a particle Monte-Carlo method including convection, stochastic diffusion (Brownian diffusion), van der Waals forces or electrical forces. A number of simulation results show the influence of different effects on the particle motion and deposition.

Electro HydroDynamic Atomisation (EHDA) disperses a liquid into small, highly charged droplets. We show that this method can be used to produce particles that release a drug at a desired rate. This is done by spraying a solution of bio-degradable polymers and an enzyme, which represents the effective drug. The release rate can be varied by modification of the polymer matrix. It is further demonstrated that the enzyme fully retains its functionality in the EHDA process. Practical use of this technique for medicine production requires a scaled-up design, which must be based on an adequate model of the particle flow in the charged droplet spray plume. As a step in this direction, the most important result is a scale-up relation that allows simulations of an experimental spray with millions of particles, using only a few thousand model particles. The experimental spray is examined with a Phase Doppler Particle Analyser (PDPA) set-up, and the resulting density and velocity profiles are compared to the numerical results. There is a qualitative agreement between experiment and model.

The electrical conductivity of a granular matter has been studied. Electromagnetic perturbations have been experimentally produced at the vicinity of the packing. The Branly experiment has been performed and quantified. It appears that the soldering of grains is induced by the electromagnetic waves. This explains the drop of electrical resistance, i.e., the Branly effect. The contacts between the grains is enhanced because the electromagnetic waves induce soldering between grains.

Granular matter exhibits an intricate mix of solid and liquid-like phenomena, some familiar, others remarkable, but almost always poorly understood [1–5]. In particular, when submitted to external stress, granular matter does not flow homogeneously like an ordinary fluid would. Instead, it forms rigid regions separated by narrow shear bands where the material yields and flows [1–7]. This shear localization may also be relevant for dense colloids, emulsions and foams [8–10], but for granular media it is ubiquitous- think of geological faults [11,12], avalanches [13,14] and silo discharges [2,15–17]. Empirically, shear bands are observed to be narrow, particle-shape dependent and often localize near a wall [1,2,6,7,11–21]. Here in contrast to this behaviour, we present experiments in which much wider and universal shear zones can be created in the bulk of the material. These shear zones exhibit Gaussian strain rate profiles, with position and width tunable by the experimental geometry and particle properties.

We investigate the stress response function of a layer of grains, i.e. the stress profile in response to a localized overload. The shape of the profile is very sensitive to the packing arrangement, and is thus a good signature of the preparation procedure of the layer. This study has been done by the use of molecular dynamics numerical simulations. Here, for a given rain-like preparation, we present the scaling properties of the response function, and in particular the influence of the thickness of the layer, and the importance of the location of the overload and measurement points (at the boundaries, in the bulk).

We report an experimental study of a binary sand bed under : (i) an oscillating flow and (ii) under a circular flow. In both cases, the appearance of a granular segregation is shown to strongly depend on the sand bed preparation. A reproducible protocol is proposed. In the oscillating case, a segregation in volume is observed in the final steady state. The correlation between this phenomenon and the fluid flow is emphasised. In the unidirectional case, a segregation in surface in observed instead of a segregation in volume. The relative difference in size between both sand species is highlighted as the physical parameter leading to the phase segregation. The difference between the characteristic times of appearance of both types of ripples is shown to be the major reason for the different behaviours leading to either a bulk segregation or a surface segregation.

The common experiment of a granular flow in a vertical tube is modified. The grains are submitted to the joint action of both gravity and upward air flux. While searching their equilibrium, the grains form clogs. A space-time analysis of the phenomenon is conducted. One-dimensional simulations of the experiment are presented. The influence of several parameters is discussed.

We have experimentally studied the dense flow of identical bubbles below inclined planes. The flow is driven by the fast motion of dislocations. As a function of the density of dislocations (controlled by the parameter

θ

), a transition occurs between a simple collective motion and a granular flow regime.

The dilute-to-dense transition of granular flow of particle size

d

0

is studied experimentally in a two dimensional channel (width

D

) with confined exit (width

d

). It is found that there exists a maximum inflow rate

Q

c

, above which the outflow changes from dilute to dense and the outflow rate

Q

(

t

) drops abruptly from

Q

c

to a dense flow rate

Q

d

. The re-scaled critical rate

$$ D/d_0 $$

is found to be a function of a scaling variable

λ

only, i.e.

q

c

∼

F

(

λ

), and

$$ \lambda \equiv \frac{d} {{d_0 }}\frac{d} {{D - d}} $$

. This form of

λ

suggests that the dilute-to-dense transition is a global property of the flow; unlike the jamming transition, which depends only on

$$ \frac{d} {{d_0 }} $$

. The transition is found to occur when the area fraction of particles near the exit reaches a critical value 0.65±0.03.

A description of highway traffic flow is proposed, based on a flux model originally developed for granular gases in a compartmentalized setup. Results from a pilot study of the highway A58 in the Netherlands are presented, followed by a discussion of possible improvements and applications of the model.

We study the behavior of inelastically colliding particles moving in a bistable potential, and driven by a stochastic heat bath. The system has the tendency to cluster at low drivings, and to fill completely the available space when vigorously shaked. In the case of just two particles, we show that the hopping over the potential barrier occurs following the Arrhenius rate, if the temperature is replaced by the granular temperature. For systems containing many particles we observe a strong competition between the excluded volume effect, which favors a symmetric distribution between the two wells, and inelasticity which on the contrary induces clustering.

We compare different techniques to modelize granular flows on two-dimensional inclined planes. An example of each simulation is described. some characteristics and the limitations of the different models are mentionned.

In the present work, we have introduced the presence of granular arches in the Tetris model. In this modified Tetris model, the friction between grains is the key parameter and a slipping threshold is defined. Two different regimes have been observed as a function of the slipping threshold.

The probability density function of velocity fluctuations of

glanulence

observed by Radjai and Roux in their two-dimensional simulation of a slow granular flow under homogeneous quasistatic shearing is studied by the multifractal analysis for fluid turbulence proposed by the present authors. It is shown that the system of granulence and of turbulence have indeed common scaling characteristics.