SAFETY

VEHICLE DYNAMICS , CONTROL AND STABILITY

In the last years the number of electronic controllers of vehicle dynamics applied to chassis components has
increased dramatically. They use lookup table of the primary order vehicle global parameters as yaw rate, lateral
acceleration, steering angle, car velocity, that define the ideal behavior of the vehicle. They are usually based on
PID controllers which compare the actual behavior of every measured real vehicle data to the desired behavior,
from look up table. The controller attempts to keep the measured quantities the same as the tabled quantities by
using ESP, TC (brakes and throttle), CDC (control shocks absorbers), EDIFF(active differential) and 4WS (rear
wheels active toe). The performances of these controls are good but not perfect. The improvement can be achieved
by replacement of the lookup tables with a fast vehicle model running in parallel to the real vehicle. This model will
allow to anticipate the vehicle reactions to driver maneuvers or the effect of road conditions, and will simplify the
regulations of the controller making hopefully them more smoother (feed forward control). Firstly a low dof model
of suspensions kinematics with a reduced number of bodies can be used. Then a more complicate and fast model
with an accurate representation of tires, steering, suspensions and powertrain components will be used and
running inside of the central vehicle dynamics ecu. The subjects of measuring and calculating the real car
secondary order parameters like forces and side slip on the contact patches to have more detail information to
correct the car model when it diverts from the real car behavior are described in this article.

Modeling and Validation of ABS and RSC Control Algorithms for a

64 Tractor and Trailer Models using SIL SIMULATION.
A Software-in-the-Loop (SIL) simulation is presented here wherein control algorithms for the Anti-lock Braking
System (ABS) and Roll Stability Control (RSC) system were developed in Simulink. Vehicle dynamics models of a
64 cab-over tractor and two trailer combinations were developed in TruckSim and were used for control system
design. Model validation was performed by doing various dynamic maneuvers like J-Turn, double lane change,
decreasing radius curve, high dynamic steer input and constant radius test with increasing speed and comparing
the vehicle responses obtained from TruckSim against field test data. A commercial ESC ECU contains two
modules: Roll Stability Control (RSC) and Yaw Stability Control (YSC). In this research, only the RSC has been
modeled. The ABS system was developed based on the results obtained from a HIL setup that was developed as a
part of this research. The RSC system was developed after a careful study of the field test data obtained from the
vehicle manufacturer in which the ESC was activated. Both the control models were then validated by performing
dynamic maneuvers and straight ahead braking tests and the simulation results were compared against the field
test data. The results show that the ABS and RSC control models were able to mimic the performance of
commercially available ECUs.

The Three Suspension Roll Centers and their Application to

Vehicle DYNAMICS.
The automotive industry commonly uses two definitions of the suspension roll center, the Kinematic Roll Center
(KRC) - of interest in studying suspension geometry, and the Force-based Roll Center (FRC) - of interest in
studying steady-state vehicle dynamics. This paper introduces a third definition, the Dynamic Roll Axis (DRA) - of

interest in studying transient vehicle dynamics. The location of each one of these roll centers has a unique
application to vehicle design and development.
Although the physical meaning of each roll center is significantly different, the generic term roll center is often
used without proper specification. This can lead to confusion about how roll centers influence vehicle behavior. This
paper hopes to clarify some of this confusion and is organized into three parts: (1) Describes calculation methods
for each of the three vehicle roll centers (for independent suspensions) as well as their relevance to vehicle
dynamics; (2) Explains the relationship between the kinematic and force-based roll centers; (3) Offers
recommendations on considerations for choosing roll center(s) location during vehicle design.

Robust Control of Anti-Lock Brake System for an Electric Vehicle

Equipped with an Axle Motor
As the main power source of the electric vehicle, the electric motor has outstanding characteristics including rapid
response, accurate control and four-quadrant operation. Being introduced into the dynamic chassis control of
electrified vehicles, the electric motor torque can be used not only for driving and regenerative braking during
normal operating conditions, but also offers a great potential to improve the dynamic control performance of the
anti-lock braking under emergency deceleration situations.
This paper presents a robust control algorithm for anti-lock braking of a front-wheel-drive electric vehicle equipped
with an axle motor. The hydraulic and regenerative braking system of the electric vehicle is modeled as a LPV
(linear parameter varying) system. The nonlinearities of the control system are considered as uncertain parameters
of a linear fractional transformation. A static-state feedback control algorithm which is robust against the
uncertainties is designed to achieve the maximum braking capability of the vehicle. To validate the control
performance of the proposed algorithm, simulations are carried out. Based on the simulation results, the proposed
anti-lock braking control algorithm can achieve the good robustness and control performance under different road
adhesion coefficients, guaranteeing the brake stability of the electric vehicle.

The Rolling Quarter Car Model a Method to Incorporate Dynamic

Tire Response in Grip Optimization
Selection of springs and dampers is one of the most important considerations when finalizing a race car suspension
design. It is also one of most complex due to the dynamic interaction of the vehicle with the ground. Current tuning
methods for spring and dampers' effect on vehicle ride can be based on simplified dynamic models of the vehicle,
such as the quarter-car model. While efficient computationally, the traditional quarter-car model does not account
for the non-linear variation in grip seen by a fluctuating contact-patch. Both amplitude and frequency of suspension
oscillation contribute to loss of tire grip. The method can be improved by incorporation of a dynamic tire model,
though resulting in non-linear effects. An improved rolling quarter-car model is created, which includes the effect
of dynamic tire forces in the analysis of improved grip. Using typical Formula SAE race car, characteristics as a test
case, a linearized dynamic model is made. The effect of suspension parameters on the dynamic tire forces
produced are surveyed. Improvements made by the new model are presented.

A Novel Method for Side Slip Angle Estimation of Omni-Directional

Vehicles
The omni-directional vehicle is an innovative vehicle, in which the in-wheel steering motor and in-wheel driving

motor are integrated into each wheel of the vehicle so that each wheel can be independently controlled to have
traction, braking, and turning motions to improve the vehicle's mobility, handling and stability. To realize good
performance, various control strategies have been proposed, like the active steering control and the direct yaw
moment control, where the accurate slip angle information is critical to these control strategies. However, in
practice, the side slip angle is hard or expensive to be measured for a passenger vehicle, therefore, different
estimation methods have been proposed in the literature. In this paper, a novel side slip angle estimation method is
proposed for the omni-directional vehicle that has four independent steering motors. This method includes the
estimation of total alignment torque and side slip angle, and only needs the measurements of steering angles and
steering motor currents, which are available by using angle sensors and current sensors embedded with the inwheel steering motors. Numerical simulations are used to validate the effectiveness of the proposed side slip angle
estimation approach.

Vehicle Aerodynamics Simulation for the Next Generation on the K

Computer: Part 1 Development of the Framework for Fully
Unstructured Grids Using up to 10 Billion Numerical Elements
A simulation framework for vehicle aerodynamics using up to 10 billion fully unstructured cells has been developed
on a world-fastest class supercomputer, called the K computer, in Kobe, Japan. The simulation software
FrontFlow/red-Aero was fully optimized on the K computer to utilize up to 10,000 processors with tens of
thousands of cores. A hybrid parallelization method using MPI among processors and OpenMP among cores inside
each processor was adopted. The code was specially tuned for unsteady aerodynamic simulation including largeeddy simulation, and low Mach number approximation was adopted to avoid excessive iterations usually required
for the fully incompressible algorithm.
The automated mesh refining system was developed to generate unstructured meshes of up to 10 billion cells. In
the system, users only generate unstructured meshes in the order of tens of millions of cells directly using
commercial preprocessing software. Then the original elements are refined in the preconditioning process using the
K computer, and billions of cells are automatically generated. In the same way, the CFD results of the billions of
elements are mapped on the original coarse grid points during post processing, and the users can transport the
data from the K computer without any difficulty visualizing the results using conventional post-processing software.
The simulation framework was validated on vehicles with different rear slant angles, and dependence of the
aerodynamic forces on the unstructured-mesh resolution as well as the grid topology near the body surface was
discussed. It was demonstrated that the simulation of the multi-billion cell mesh on the K computer agreed with
wind-tunnel measurements within a small percentage of error, and validity of the fully unstructured finite volume
method using multi-billion cells was confirmed.

A Cost-Effective Alternative to Moving Floor Wind Tunnels in

Order to Calculate Rolling Resistance and Aerodynamic Drag
Coefficients
This study investigates the practicality of vehicle coast down testing as a suitable replacement to moving floor wind
tunnel experimentation. The recent implementation of full-scale moving floor wind tunnels is forcing a reestimation of previous coefficient of drag determinations. Moreover, these wind tunnels are relatively expensive to
build and operate and may not capture concepts such as linear and quadratic velocity dependency along with the
influence of tire pressure on rolling resistance. As a result, the method elucidated here improves the accuracy of
the fundamental vehicle modeling equations while remaining relatively affordable. The trends produced by
incorporating on road test data into the model fit the values indicated by laboratory tests. This research chose

equipment based on a balance between affordability and accuracy while illustrating that higher resolution frequency
equipment would further enhance the model accuracy.

Computational Process for Wind Noise Evaluation of Rear-View

Mirror Design in Cars
A computational approach to evaluate rear-view mirror performance on wind noise in cars is presented in this
paper. As a comfort metric at high speeds, wind noise needs to be addressed, for it dominates interior noise at
mid-high frequencies. The impetus on rear-view mirror design arises from its crucial role in the flow field and the
resulting pressure fluctuations on the greenhouse panels. The motivation to adopt a computational approach arises
from the need to evaluate mirror designs early in vehicle design process and thus in conjunction with different
vehicle shapes. The current study uses a Lattice Boltzmann method (LBM) based computational fluid
dynamics(CFD) solver to predict the transient flow field and a statistical energy analysis(SEA) solver to predict
interior noise contribution from the greenhouse panels. The accuracy of this computational procedure has been
validated and published in the past. Realistic car geometry is chosen and the transient flow field around the vehicle
resulting from mounting two different rear-view mirror designs is analyzed. The mirrors are ranked based on the
interior noise contribution from the greenhouse panels. In each case, pressure fluctuations on the panels are
decomposed using a wavenumber filter to visualize the turbulent and acoustic fields. Analysis leading to flow
features and geometry responsible for the contributing noisy regions is performed.

The Effect of Upstream Turbulence on an Exposed Wheel Wake

As open-wheeled racing cars frequently race in close proximity, a limiting factor on the ability to overtake is the
aerodynamic performance of the vehicle while operating in a leading car's wake. Whilst various studies have
examined the effectiveness of wings operating in turbulent flow, there has been limited research undertaken on the
aerodynamic effect of such conditions on wheels.
This study describes the influence of upstream turbulence on the wake flow features of an isolated wheel, since the
flow field of a wheel will generally be turbulent (due to the wakes of upstream cars and/or bodywork). Pressure
distributions and velocity vector plots are examined, which were obtained using a four-hole pressure-sensitive
Cobra probe on a traverse 2.5 diameters downstream of the wheel axle line, in smooth and turbulent flow.
This analysis also compares the effect of upstream turbulence on the wake for the rotating and stationary wheel;
as well as investigating the sensitivity of the wake to the wheel-to-road gap in smooth and turbulent flow.
The study found that, at 2.5 diameters downstream, the overall wake width of the stationary wheel decreased with
increased levels of turbulence, it also showed that the effect on the wake of increasing the wheel-to-road gap was
independent of the level of turbulence in the flow. Both rotating and stationary wheel wake maps also displayed a
higher minimum pressure with nominally turbulent flow, than those displayed by the corresponding wake maps
with nominally smooth flow, and a much more gradual transition to free-stream pressure.

Comparison between Casting and Alloy Rim Pattern on Brake Disk

Cooling Using Numerical Simulation
Efficient function of brake system is considered a crucial stage in the vehicle development process. Heat exchange
reduction can decrease the operational condition of braking system. Although the rims patterns have a significant

role on vehicle aerodynamics, they can also have effect on air flow around the brake disk. So, selecting a rim is
vital from both a safety and an aerodynamic point of view. In this paper the effect of air flow around a brake disk
for two different types of rims was studied; a steel casting and an aluminum alloy rim. Numerical simulation was
used for this analysis. First, the flow field around brake disk with iron casting rim was investigated at different
velocities. Second, the flow field around the same brake disk system with an aluminum alloy rim was modeled at
the same velocities. Finally, the effect of rim design on flow pattern over brake disk was compared. It was found
that changing rim design has significant influence on velocity distribution around brake disc and pads.

Experimental Comparison of the Aerodynamic Behavior of

Fastback and Notchback Driver Models
The experimental investigation was conducted with a 25%-scaled realistic car model called DrivAer mounted in a
wind tunnel. This model includes geometric elements of a BMW 3 series and an Audi A4, accommodating modular,
rear-end geometries so that it represents a generalized modern production car. The measurements were done with
two different DrivAer rear end configurations (fastback and notchback) at varying side-wind conditions and a
Reynolds number of up to Re=3.2106. An array of more than 300 pressure ports distributed over the entire rear
section measured the temporal pressure distribution. Additionally, extensive flow visualizations were conducted.
The combination of flow visualization, and spatially and temporally resolved surface pressure measurements
enables a deep insight into the flow field characteristics and underlying mechanisms. Moreover, static pressure
fluctuations indicate regions with a high turbulence level due to flow separation and interaction between different
vortical structures. A directional decomposition of the pressure coefficient is demonstrated and acts generally
normal to the surface of the vehicle, in the drag and lift direction.

An Influence Matrix Method for Determining Boundary

Interference on Automotive Models in Solid-Wall Wind Tunnels
The solid-wall wind tunnel boundary correction method outlined in this paper is an efficient pressure-signature
method that requires few wall-mounted pressures. These pressures are used to determine the strengths of modeland wake-representing singularities that are used with the method of images to calculate the longitudinal and
lateral velocity increments induced by the wind tunnel walls. Two force correction models are presented that
convert these velocity increments to force and moment corrections. The performances of the correction procedures
are demonstrated by their application to data from two sets of four, geometrically identical, differently sized,
simplified automotive models.

Differences between Air-Dam Spoiler Performances in Wind

Tunnel and On-Road Tests
An air-dam spoiler is commonly used to reduce aerodynamic drag in production vehicles. However, it inexplicably
tends to show different performances between wind tunnel and coast-down tests. Neither the reason nor the
mechanism has been clarified.
We previously reported that an air-dam spoiler contributed to a change in the wake structure behind a vehicle. In
this study, to clarify the mechanism, we investigated the coefficient of aerodynamic drag C D reduction effect, wake
structure, and underflow under different boundary layer conditions by conducting wind tunnel tests with a rolling
road system and constant speed on-road tests.
We found that the air-dam spoiler changed the wake structure by deceleration of the underflow under stationary
floor conditions. Accordingly, the base pressure was recovered by approximately 30% and, the C D value reduction

effect was approximately 10%. The ratio of the base pressure recovery to the C D value reduction effect was
approximately 90%, suggesting that the main mechanism is the base pressure recovery produced by changing the
wake structure. In contrast, the air-dam spoiler did not change the wake structure in on- road tests and under
moving floor conditions. Consequently, the base pressure recovery and C D value reduction effect decreased to
approximately 20% and 4%, respectively, owing to the deceleration of the underflow being insufficient to change
the wake structure. In other words, the boundary layer of the floor helped to decelerate the underflow and to
change the wake structure under stationary floor conditions, suggesting that we overestimated the air-dam spoiler
performance under stationary floor conditions.

Numerical Simulations of Flow around a Realistic Generic Car

Model
This paper focuses on the numerical simulations of flow around a realistic generic car model called the DrivAer
body. This new open-source model is based on the geometries of two medium sized cars, the Audi A4 and the BMW
3 series, and possesses more representative car features as the well-known generic Ahmed body. In this paper,
only the fastback geometry is investigated. The flow solver used is ISIS-CFD developed by CNRS and Ecole
Centrale de Nantes. This solver is based on a finite-volume method, and two turbulence modelizations are used:
the Explicit Algebraic Reynolds Stress Model (EARSM) and a Detached Eddy Simulation (DES). Two meshes are
used. For one, the walls are described with a wall function and the mesh contains 19 million cells. This mesh is
called Mesh 1. For the second mesh, a low-Reynolds number turbulence model for the walls is used. In this case,
the mesh contains 39 million cells, and is called Mesh 2. For the simulation with the EARSM model, the drag
coefficient is well predicted while the lift coefficient is over-predicted with however a value lower with the Mesh 2.
With both meshes, an automatic grid refinement has been used. In this case, from the Mesh 1, the new mesh
contains 25 million cells while from the the Mesh 2, the new mesh contains 45 million cells. With these new
meshes, the drag coefficient is similar and the lift coefficient est lower than the values obtained with the initial
meshes but the predicted values over-predict the experimental value. DES simulations have been performed with
all meshes. The aerodynamic coefficients are similar and in good agreement with the experimental values, even for
the lift coefficient.

The Aerodynamics of a Small Car Overtaking a Truck

The influence of a large truck on the aerodynamics of a small passenger car in an overtaking manoeuvre on the
motorway was considered, many years ago, during the 1970's, to be a potential problem for the vehicle
aerodynamicist. The concern never became significant as vehicle architecture evolved and car weights increased.
The current drive for improved fuel economy is advocating that a considerable reduction in vehicle mass is
desirable and therefore it may be time to readdress the significance of the truck passing manoeuvre. A quasisteady experiment has been undertaken at small model scale to examine the aerodynamic characteristics of a
small car in proximity to a large truck. Measurements at yaw were included to crudely simulate the effects of a
crosswind. The wind tunnel data is presented and the limitations of the experimental procedure are discussed. An
estimate of the increased aerodynamic input on a car in a real world overtaking manoeuvre with relative velocity
between the two vehicles is introduced. The data is compared with the steady state yaw characteristics of the car
alone to relate truck passing with the general concern of crosswind sensitivity. Under the same crosswind
conditions the truck passing manoeuvre is shown to be a more extreme, but rarer event.

Aerodynamic Shape Optimization Based on the MIRA Reference

Car Model
Automobile industry is facing the great challenge of energy conservation and emission reduction. It's necessary to
do some researches on some surface components of a car body to find out which of them may affect aerodynamic

drag remarkably. This will help an aerodynamic engineer modify an initial car model more clearly. We also hope to
reduce the cost during the process, including time and resources. In this paper, with the purpose of developing an
aerodynamic shape optimization process and realizing its automation, a MIRA reference car model was studied and
three commercial softwares were integrated-Altair HyperStudy, HyperMesh and CD-adapco STAR-CCM+. The
optimization strategy in this paper was: firstly, a DOE (design of experiment) matrix, which contained four design
factors and thirty levels was created. The baseline model was morphed according to the DOE matrix. Then the
morphed model's aerodynamic drag coefficient (Cd) and lift coefficient (Cl) were calculated via CFD software. Both
of them were used as the response variables. Secondly, after finishing all the CFD simulations, an approximate
model was established on the basis of the DOE matrix. Finally, an optimization algorithm was performed. As a
result, an optimized combination of the four design variables was obtained and it met the demands: Cd was
minimum and Cl was less than zero. The combination was tested via a true CFD simulation to ensure the result's
accuracy. In the end, the drag coefficient decreased by 19.2%.

Effect of Rear-End Extensions on the Aerodynamic Forces of an

SUV
Under a global impulse for less man-made emissions, the automotive manufacturers search for innovative methods
to reduce the fuel consumption and hence the CO2-emissions. Aerodynamics has great potential to aid the
emission reduction since aerodynamic drag is an important parameter in the overall driving resistance force.
As vehicles are considered bluff bodies, the main drag source is pressure drag, caused by the difference between
front and rear pressure. Therefore increasing the base pressure is a key parameter to reduce the aerodynamic
drag. From previous research on small-scale and full-scale vehicles, rear-end extensions are known to have a
positive effect on the base pressure, enhancing pressure recovery and reducing the wake area.
This paper investigates the effect of several parameters of these extensions on the forces, on the surface pressures
of an SUV in the Volvo Cars Aerodynamic Wind Tunnel and compares them with numerical results. To decrease the
dependency of other effects within the engine bay and underbody, the SUV has been investigated in a closedcooling configuration with upper and lower grille closed and with a smoothened underbody. These results might
change if the study would be conducted with a less smooth underbody and in an open-cooling configuration.
Extensions with different shapes and dimensions have been placed around the perimeter of the base exterior. The
chosen design philosophy of the extensions allowed for different combinations with variable inclination angles
depending on their position along the base perimeter, multiple extension lengths and shapes to be investigated.
The results show that the extension shape is an important factor in reducing the aerodynamic drag. Significant
drag reductions could be obtained while maintaining the vehicle's rear lift within acceptable levels for stability with
a kicker attached to the extension. The investigation shows the reduction with a kicker holds for up to 7.5 yaw
angles. With a beneficial shape, the extension length can be significantly reduced. The reduced drag is visible in the
wake by a more concentric wake.

Aerodynamic Structure and Development of Formula 1 Racing Car

Wakes
For the modern Formula 1 racing car, the degradation in aerodynamic performance when following another car is
well documented. The problem can be broken into two parts; firstly the wake flow generated by these vehicles and
the subsequent interaction a following car has with this field. Previous research [ 1,

2& 3] has focused upon

investigating the later without completely characterizing the former. This paper seeks to address this deficiency
with initial data from a newly commissioned 30% scale Formula One wind tunnel model built to the 2011 technical

regulations. Experimentation was carried out in the Industrial Wind-Tunnel (IWT) at RMIT University. In the
absence of a rolling road an elevated ground plane was implemented; the results obtained show good agreement
with the limited published material available. Using a high frequency response, four-hole pressure probe the aft
body flow was investigated at multiple downstream locations. Time-averaged velocity, turbulence and secondary
flow vectors were plotted. These data were supplemented through flow visualization techniques to better
understand the salient aerodynamic features. The influence of geometric changes to the aerodynamic devices upon
the rear of the car (i.e. drag reduction system currently used within Formula 1) was also investigated.

Aerodynamic Characterization of the Design Changes for the

Facelift of the VW Bora
The recent facelift of the Chinese version of the VW Bora incorporated several changes of the styling of the upper
body. In particular, front facia, A-Pillar and rear end were subject to design changes. As major effects on the
aerodynamics performance were not expected, extensive wind tunnel testing for the upper body design changes
was not included in the development plan except for final performance evaluation. Nevertheless, an aerodynamic
study of the effects of the design changes was undertaken using a CFD based process. At the same time, the
facelift offered the opportunity for reducing the aerodynamic drag by improving the underbody flow. The design of
the engine undercover and the wheel spoilers were considered in this effort. For this purpose the CFD based
aerodynamic study was extended to include respective design features. The whole study was carried out using a
response surface method as a mathematical model to characterize and understand the effects of the design
changes and their interactions. This approach allowed achieving best possible insight at a minimal effort. In
addition, possible optima for the underbody optimization were identified. This helped to determine a possible tradeoffs between cost and performance and also allowed a reduced effort in the wind tunnel for fine tuning the
underbody design changes.

Methodology for the Design of an Aerodynamic Package for a

Formula SAE Vehicle
Recent changes to the rules regarding aerodynamics within Formula SAE, combined with faster circuits at the
European FSAE events, have made the implementation of aerodynamic devices, to add down-force, a more
relevant topic. As with any race series it is essential that a detailed analysis is completed to establish the costs and
benefits of including an aerodynamic package on the vehicle. The aim of the work reported here was to create a
methodology that would fully evaluate all aspects of the package and conclude with an estimate of the likely gain in
points at a typical FSAE event. The paper limits the analysis to a front and rear wing combination, but the approach
taken can be applied to more complex aerodynamic packages.
An initial wind tunnel investigation of the potential flow interactions between the driver's helmet and rear wing
using a multi-hole pressure probe is reported and the data used in a two-dimensional CFD calculation to provide an
accurate prediction of the likely down-force from the wing package. The chosen configurations are tested in a
comprehensive wind tunnel program and a map of potential setups generated. The potential aerodynamic
configurations are assessed in both quasi-static and dynamic handling analysis to demonstrate the effects of
aerodynamic lift, lift distribution, aerodynamic drag, and the effect of additional weight, weight distribution and
height of the center of gravity. The paper includes a description of the wind tunnel model and scaling considerations
and a description of the handling model and the lap simulation methods.
The results shows that a simple front and rear wing combination, providing relative low down-force (1000N at
110kph on a vehicle of 280kg, including driver), gives an improvement in performance equating to a gain of
approximately 40 points during the event. The largest improvements are seen during the autocross and endurance
events. This gain is considered sufficient to justify the inclusion of the package on the vehicle.

Low Drag Automotive Mirrors Using Passive Jet Flow Control

This paper introduces and proves a novel automotive mirror base drag reduction method using passive jet flow
control. The new concept is to open an inlet at the front part of the mirror, introduces the airflow via a converging
duct, and ejects the jet surrounding the mirror surface at an angle toward the center of the mirror. The jet
harnesses the energy from the free stream by jet mixing with the main flow via large coherent structures, entrains
the main flow to energize the base flow, reduces the wake size and turbulence fluctuation, and ultimately
significantly decreases the drag. Above phenomena are proved by wind tunnel testing with PIV and drag force
measurement and CFD large eddy simulation (LES) calculation. Two jet mirrors with different inlet areas are
studied. The jet mirror tunnel 1 has a smaller inlet area, and the jet mirror tunnel 2 has a 4.7 times larger inlet
area. The wind tunnel testing is only done for the baseline and jet mirror tunnel 1. LES is used to study all the
three mirror configurations. Both the wind tunnel testing and LES indicate that the jet mirror tunnel 1 reduces the
drag by about 18% with smaller wake width. The LES indicates that the jet mirror tunnel 2 with larger inlet area
further reduces the wake and achieves a drag reduction of 39%. This paper is only for proof of the concept and no
design optimization is done. It is believed that there is a large room to further reduce the drag with a systematic
design optimization.

Benefit Estimation Method for Lane Departure Warning Systems in

the American Traffic Environment
We develop a simulation tool which reproduces lane departure crashes for the purpose of estimating potential
benefits of Lane Departure Warning (LDW) systems in the American traffic environment.
Tools that allow a fast evaluation of active safety systems are useful to make better systems, more effective in the
real world; however accuracy of the results is always an issue. Our proposed approach is based on developing a
simulation tool that reproduces lane departure crashes, then adding the effect of the LDW to compare the cases
with and without the safety system, and finally comparing the results of different settings of the safety system.
Here, the accurate reproduction of the relevant crashes determines the reliability of the results.
In this paper, we present the reproduction of the lane departure crashes occurred in American roads in one year, by
using data distributions obtained from retrospective crash databases. We analyze data from NASS/GES 1 and
NASS/CDS2 to obtain the characteristics of lane departure accidents in the USA. Then, we model lane departure
crash scenarios by dividing them in 5 phases: (A) before departure, (B) starting the departure, (C) running offlane, (D) impact and (E) after the impact. This division of the departure sequence allowed an easy recognition of
the required parameters to reproduce each phase. The reproduced crashes were validated by comparing them to
the real crashes.
The reproduction of lane departure crashes will allow the estimation of potential benefits of LDW systems in the
USA, which is an important step towards the development of more effective countermeasures.

Adaptation of the Mean Shift Tracking Algorithm to Monochrome

Vision Systems for Pedestrian Tracking Based on HoG-Features

The mean shift tracking algorithm has become a standard in the field of visual object tracking, caused by its real
time capability and robustness to object changes in pose, size, or illumination. The standard mean shift tracking
approach is an iterative procedure that is based on kernel weighted color histograms for object modelling and the
Bhattacharyyan coefficient as a similarity measure between target and candidate histogram model. The benefits of
the approach could not been transferred to monochrome vision systems yet, because the loss of information from
color to grey-scale histogram object models is too high and the system performance drops seriously. We propose a
new framework that solves this problem by using histograms of HoG-features as object model and the SOAMST
approach by Ning et al. for track estimation. Mean shift tracking requires a histogram for object modelling. In the
proposed framework a set of high dimensional HoG-features is clustered via K-means and features inside the object
area are matched to the cluster-centers via a nearest neighbor search. This procedure is comparable to a Bag of
Words algorithm. The proposed system is evaluated for advanced driver assistance systems and it is shown that
the framework can be used as a reliable visual tracking system for a pedestrian recognition module.

Development of Lens Condition Diagnosis for Lane Departure

Warning by Using Outside Camera
Driver safety continues to be improved by advances in active safety technologies. One important example is Lane
Departure Warning (LDW). European regulators soon will require LDW in big cars to reduce traffic accidents and
New Car Assessment Programs in various countries will include LDW in a few years. Our focus is on rear cameras
as sensing devices to recognize lane markers. Rear cameras are the most prevalent cameras for outside
monitoring, and new Kids and Cars legislation will make them obligatory in the United States from 2014.
As an affordable sensing system, we envision a rear camera which will function both as a rear-view monitoring
device for drivers and as an LDW sensing device. However, there is a great difficulty involved in using the rear
camera: water-droplets and dirt are directly attached to the lens surface, creating bad lens condition.
The purpose of this study is to improve the durability of lane recognition systems when water-droplets and dirt are
deposited on the lens surface. First, we developed various diagnostic logics under various lens conditions. We then
analyzed the results of various diagnosis and expressed the lens conditions by using two evaluation axes. After
that, we improve the durability of the lane recognition system including a judgment function that determines
whether to stop the LDW system under heavy dirt and water-droplets.
We conducted driving tests and captured evaluation movies in the United States, Europe, and Japan. We evaluated
the lane recognition rate for a total of 8 hours of evaluation movies under various weather conditions. We achieved
a lane recognition rate of 95% and improved the durability of the lane recognition system.

Multi-Modal Image Segmentation for Obstacle Detection and

Masking
A novel multi-modal scene segmentation algorithm for obstacle identification and masking is presented in this
work. A co-registered data set is generated from monocular camera and light detection and ranging (LIDAR)
sensors. This calibrated data enables 3D scene information to be mapped to time-synchronized 2D camera images,
where discontinuities in the ranging data indicate the increased likelihood of obstacle edges. Applications include
Advanced Driver Assistance Systems (ADAS) which address lane-departure, pedestrian protection and collision
avoidance and require both high-quality image segmentation and computational efficiency. Simulated and
experimental results that demonstrate system performance are presented.

360 Surround View System with Parking Guidance

In this paper, we present a real-time 360 degree surround system with parking aid feature, which is a very
convenient parking and blind spot aid system. In the proposed system, there are four fisheye cameras mounted
around a vehicle to cover the whole surrounding area. After correcting the distortion of four fisheye images and
registering all images on a planar surface, a flexible stitching method was developed to smooth the seam of
adjacent images away to generate a high-quality result. In the post-process step, a unique brightness balance
algorithm was proposed to compensate the exposure difference as the images are not captured with the same
exposure condition. In addition, a unique parking guidance feature is applied on the surround view scene by
utilizing steering wheel angle information as well as vehicle speed information.

Tradeoffs in the Evaluation of Light Vehicle Pre-Collision Systems

Pre-collision systems (PCS) use forward-looking sensors to detect the location and motion of vehicles ahead and
provide a sequence of actions to help the driver either avoid striking the rear-end of another vehicle or mitigate the
severity of the crash. The actions include driver alerts, amplification of driver braking as distance decreases
(dynamic brake support, DBS), and automatic braking if the driver has not acted or has not acted sufficiently
(crash imminent braking, CIB).
Recent efforts by various organizations have sought to define PCS objective test procedures and test equipment in
support of consumer information programs and potential certification. This paper presents results and insights from
conducting DBS and CIB tests on two production vehicles sold in the US. Eleven scenarios are used to assess the
systems' performance. The two systems' performance shows that commercial systems can be quite different. Also
demonstrated is the experience with test equipment, including a towable target that has been designed for
resiliency and radar signature, a braking robot, and bumper guard.
While improvements continue in evaluation procedures, the community may still be a distance away from an ability
to test these systems with realistic surrogate targets in speed conditions similar to those that lead to serious injury
and fatalities. Recommendations are offered to help improve PCS evaluation.

Clustering and Scaling of Naturalistic Forward Collision Warning

Events Based on Expert Judgments
The objectives of this study were a) to determine how expert judges categorized valid Integrated Vehicle-Based
Safety Systems (IVBSS) Forward Collision Warning (FCW) events from review of naturalistic driving data; and b) to
determine how consistent these categorizations were across the judges working in pairs. FCW event data were
gathered from 108 drivers who drove instrumented vehicles for 6 weeks each. The data included video of the driver
and road scene ahead, beside, and behind the vehicle; audio of the FCW alert onset; and engineering data such as
speed and braking applications. Six automotive safety experts examined 197 valid (i.e., conditions met design
intent) FCW events and categorized each according to a taxonomy of primary contributing factors. Results indicated
that of these valid FCW events, between 55% and 73% could be considered nuisance alerts by the driver. These
were the FCW alerts presented in benign conditions (e.g., lead-vehicle turning) or as a result of deliberate driver
action (aggressive driving). Only 16% of the FCW alerts were attributed to driver distraction and all of these cases
involved a driver looking away from the road scene at an inopportune time. The consistency or agreement in
categorization performance of the 6 experienced safety professionals was also examined. Working in pairs, there
was either majority or unanimous agreement on 88% of the 197 FCW cases reviewed by the three pairs of judges
working independently from one another.

Lateral Control for Automated Vehicle Following System in Urban

Environments

In contrast to highway, there are some sections not well maintained in urban roads. In these sections, there may
be faint lane marks or static obstacles due to construction or some other reasons. Therefore, an automated vehicle
following system such as traffic jam assistant should consider these sections to guarantee the safety of the system.
In order to achieve this purpose, a model predictive control (MPC) scheme has been developed. The objectives of
MPC are to compute the sequence of optimal steering input for vehicle following with obstacle avoidance. For this,
the MPC uses the lead vehicle's state and obstacle's position obtained by lidars. For this purpose, a simplified
nonlinear model of the vehicle was used to predict the future evolution of the system. Based on this prediction,
performance index is optimized under operating constraints at each time step. A test vehicle equipped with two
lidars on left and right corner of the front bumper has been developed. And the performance of the proposed MPCbased steering control algorithm has been investigated via vehicle test. Test results show the robust performance
of vehicle following in urban environments.

Independent Control of Steering Force and Wheel Angles to

Improve Straight Line Stability
This paper describes a control method to improve straight-line stability without sacrificing natural steering feel,
utilizing a newly developed steering system controlling the steering force and the wheel angle independently. It
cancels drifting by a road cant and suppresses the yaw angle induced by road surface irregularities or a side wind.
Therefore drivers can keep the car straight with such a little steering input adjustment, thus reducing the driver's
workload greatly.
In this control method, a camera mounted behind the windshield recognizes the forward lane and calculate the
discrepancy between the vehicle direction and the driving lane. This method has been applied to the test car, and
the reduction of the driver's workload was confirmed.
This paper presents an outline of the method and describes its advantages.

Independent Left and Right Rear Toe Control System

Honda has developed an Independent Left and Right Rear Toe Control System that can achieve stable cornering
performance and agile handling. We believe the issue that should be resolved in the next generation of ESC is the
expansion of stability and agility into the general operation area. We examined how to accomplish this aim, and
control of the independent rear toe angle was decided to be an appropriate method. In addition, a method for
mounting the system without using a dedicated suspension was proposed. If left and right toe angles can be
controlled independently, toe angle control and normal 4WS control become possible at the same time. In this
paper, we will discuss the fundamental principle of independent toe angle control and the system configuration.
Also, INOMAMA Handling (at driver's will) achieved by this system, as well as the fun and safe driving that are
achieved as a result will be shown.

Consideration about Meshing of Worm Gear Based on MUB

(Meshing Under Base-Circle) Theory for EPS
This paper will discuss the stress reduction of the worm wheel for an electric power steering (EPS) system. The
research discussed in this paper focused on the worm wheel, the EPS component that determines the maximum

diameter of the system. If the stress of the worm wheel could be reduced without increasing in size, it would be
possible to reduce the size of the worm wheel and EPS system. In order to reduce the stress of the worm wheel,
the conventional design method has extended the line-of-action toward outside of the worm wheel to increase the
contact ratio of the gears and these method lead to an increase in the outer diameter.
In order to address this issue, past research proposes the basic concept to extend line-of-action toward the inside
of the worm wheel. And this new meshing theory was named MUB (Meshing Under Base-circle) theory.
In this paper, characteristics of meshing of the gear formed by MUB theory are determined in more detail. After
that, an example of design guideline of the worm gear profile based on MUB theory is explained. And this worm
gear was manufactured, installed into an EPS system, and tested. This paper will report on the findings made
during the course of this research

Preview Ride Comfort Control for Electric Active Suspension

(eActive3)
This paper reports the results of a study into a preview control that uses the displacement of the road surface in
front of the vehicle to improve for front and rear actuator responsiveness delays, as well as delays due to
calculation, communication, and the like. This study also examined the effect of a preview control using the
eActive3 electric active suspension system, which is capable of controlling the roll, pitch, and warp modes of
vehicle motio

Development of A New Model for Roll-Plane Active Hydraulically

Interconnected Suspension
In this paper, a more sophisticated mathematical linear model for a roll-plane active hydraulically interconnected
suspension (HIS) system was developed. Model parameters tuning were then carried out, which resulted in a
model that is capable of producing rather accurate estimation of the system, with significant improvements over
models built previously. For the verification of the new model, two simulations and corresponding experiments are
conducted. Data comparisons between the simulations and experiments show high consistent responses of the
model and the real system, which validated the robustness and accuracy of the new mathematical model. In this
process, the characteristics of the pressure response and the rise time inside the actuators have been revealed due
to the presence of the flow.

A New Adaptive Controller for Performance Improvement of

Automotive Suspension Systems with MR Dampers
A control algorithm is developed for active/semi-active suspensions which can provide more comfort and better
handling simultaneously. A weighting parameter is tuned online which is derived from two components - slow and
fast adaptation to assign weights to comfort and handling. After establishing through simulations that the proposed
adaptive control algorithm can demonstrate a performance better than some controllers in prior-art, it is
implemented on an actual vehicle (Cadillac STS) which is equipped with MR dampers and several sensors. The
vehicle is tested on smooth and rough roads and over speed bumps.

Durability Design Method of New Stopper Bush Using New Theory

(Friction and Spring) for Electric Power Steering
In the automobile industries, weight reduction has been investigated to improve fuel efficiency together with
reduction of CO2 emission. In such circumstance, it becomes necessity to make an electric power steering (EPS)
more compact and lightweight. In this study, we aimed to have a smaller and lighter EPS gear size by focusing on

an impact load caused at steering end. In order to increase the shock absorption energy without increase of
stopper bush size, we propose new theory of impact energy absorption by not only spring function but also friction,
and a new stopper bush was designed on the basis of the theory. The profile of the new stopper bush is cylinder
form with wedge-shaped grooves, and when the new stopper bush is compressed by the end of rack and the gear
housing at steering end, it enables to expand the external diameter and produce friction. In this study, we
considered the durability in the proposed profile. In particular, the theoretical study was made on the effect due to
sliding from the standpoint of the contact pressure and the sliding distance, and durability testing was carried out
to check for the durability of the performance after predetermined operation time. After the durability testing, we
have examined the force-displacement characteristics, the change in characteristics by thickness reduction agree
with the theory. We will report the design method, since the design procedure of the smaller and lighter EPS gear
size could be confirmed useful.

Influence of Shim Bending Mode on Damping Force Variation of a

Hydraulic Twin Tube Shock Absorber
Automotive shock absorber shims are subjected to deformation while generating the pressure differential across
the rebound and compression chambers. Considering the contact, large deflection, and material this shim stack
deformation will be nonlinear throughout the working velocity of shock absorbers. The deformation of shim stack
mainly depends on number and geometry of deflection disk, number and geometry of ports, and clamping disk
geometry on which shims are rested.
During the rebound and compression stroke of the shock absorber, the oil flows through the piston and base valve
ports. High pressure oil developed during mid and high velocity of shock absorber results in deflection of shim stack
in piston and base valve assembly. This deflection leads to oil leakage through the shim stack which results in
change in damping force by the shock absorber. The fluid pressure from the flow passage (well) acting over surface
area of shim differs while causing an elastic bending of the shims. The oil leakage depends on shim bending mode
which in turn depends piston port position and resting/clamping size of back up disk on which shims are rested.
The bending mode of shim changes, if the piston/base valve port positions are varied which in turn changes the
pressure differential across shims. This will results in variation of damping force trend. This paper presents
different porting positions of piston/ base valve and its effect on bending mode of shim stack & damping force
trend. Structural simulation is carried out to substantiate the bending mode shape and experimental trails are
carried out to demonstrate the damping force changes.

Simulated Driving Assessment: Case Study for the Development of

Drivelab, Extendable Matlab Toolbox for Data Reduction of
Clinical Driving Simulator Data
Driving simulators provide a safe, highly reproducible environment in which to assess driver behavior.
Nevertheless, data reduction to standardized metrics can be time-consuming and cumbersome. Further, the validity
of the results is challenged by inconsistent definitions of metrics, precluding comparison across studies and
integration of data. No established tool has yet been made available and kept current for the systematic reduction
of literature-derived safety metrics.
The long term goal of this work is to develop DriveLab, a set of widely applicable routines for reducing simulator
data to expert-approved metrics. Since Matlab is so widely used in the research community, it was chosen as a
suitable environment. This paper aims to serve as a case study of data reduction techniques and programming
choices that were made for simulator analysis of a specific research project, the Simulated Driving Assessment.

The initial set of Matlab routines was successfully tested by analyzing recognized metrics, such as Distance
Headway, Time Headway, Time-To-Collision, and Reaction Time. The programming choices are described in this
paper. This foundation of processes and Matlab routines will be used as the base for a DriveLab. The newly
developed Matlab toolbox will be validated using the Realtime Technologies Incorporated (RTI) simulator.It will be
generalized to other driving simulators so as to be a robust research tool for use by the scientific community.

A Comparison of Three Different Approaches to Image Depth in

Driver Information Clusters: 2D Computer Graphics, 3D Computer
Graphics and 3D Imaging
Displays that support complex graphics in driver information (DI) systems allow for the presentation of detailed
visual data by employing a range of static (fixed image) and/or dynamic (moving image) design approaches. Such
displays are gaining market share across a wide range of mainstream vehicles as the availability and cost of such
technologies improves.
Although a range of 2D, rendered 3D, and 3D imaging (orstereoscopic) information displays have been
demonstrated throughout the automotive industry in recent years, there is limited empirical research examining
consumer preference of the respective approaches or their influence on driving related tasks.
The vehicle environment is known to be a demanding context for efficiently displaying information to the driver.
Research in 3D [1, 2] reveals some of the factors that influence its acceptance and effective use, but there is
limited research on the effects of 3D-related design elements when used in a driver-vehicle interface. The purpose
of this research is to improve our understanding of these display approaches on consumer preference and drivingrelated task performance.
Participants in these studies completed a set of driver-vehicle tasks that involved a text-based item search in a
custom-designed interface that employed 2D, rendered 3D and stereoscopic 3D imaging in both static and
simulated driving conditions.
Analysis of the results of the two studies reveals developmental and procedural benefits from the use of realtime
rendered 3D graphics in a driver information display.
The specific type of 3D imaging examined in this study wasautostereoscopic (or glasses-free) 3D. Sentiment
toward these effects varied considerably, with participants divided between strongly liking, or strongly disliking, this
particular display approach.
The results of this study suggest that real-time rendering enabled by modern graphical processing units (and
associated software tools) represents an important step forward in providing more effective graphics in driver
information displays.

The Influence of Solid Anisotropy on Flame Spread over Melting

Polymers

Flame spread over a melting thermally thick composite polymer is investigated in a channel flow above a
condensed fuel. The condensed fuel consists of an isotropic (melted layer of) liquid near the heated surface and an
anisotropic (not-yet-melted) solid surrounding it. The influence of the solid anisotropy is evaluated by changing the
solid conductivity (ksx or

ksy) in one particular direction (x in horizontal flame spread direction or y in vertical

direction, see schematics in Figure 1) while keeping the other properties fixed. Note that the liquid
conductivity kl has no isotropic behavior. Numerically, it is found that the flame spread rate decreases with either
increasing ksx or ksy. The decrease with respect to ksy is less than for a comparable case described by the de Ris
formula for an isotropic pure solid. The flame spread rate is more accurately determined by an analytical formula
derived for spread across a melting solid fuel. Qualitatively, the liquid layer extent decreases with either

ksx or ksy due to the role played by the solid conduction as a heat loss mechanism in a thermally thick
fuel. In addition, ksy more strongly influences the extent of the liquid layer than ksx in that it changes both the
increasing

horizontal and vertical boundaries of the liquid layer. The surface parameters and energy balance show the

ksx and ksy. Generally, the size of the spreading flame decreases with either
ksx or ksy. The influence of the condensed phase anisotropy is also studied by the influence of other

dependence of the flame size on

increasing

overall quantities such as the heat flux region, the mass flux region, and the ignition delay.

Comparison of Fires in Lithium-Ion Battery Vehicles and Gasoline

Vehicles
Electric vehicles have become more popular and may be involved in fires due to accidents. However, characteristics
of fires in electric vehicles are not yet fully understood. The electrolytic solution of lithium-battery vehicles is
inflammable, so combustion characteristics and gases generated may differ from those of gasoline cars. Therefore,
we conducted fire tests on lithium-ion battery vehicles and gasoline vehicles and investigated the differences in
combustion characteristics and gases generated. The fire tests revealed some differences in combustion
characteristics. For example, in lithium-ion battery vehicles, the battery temperature remained high after
combustion of the body. However, there was almost no difference in the maximum CO concentration measured 0.5
to 1 m above the roof and 1 m from the side of the body. Furthermore, HF was not detected in either type of
vehicle when measured at the same positions as for CO.
The experiment results did not reveal any reason to treat vehicles with lithium-ion batteries differently from
gasoline vehicles with regard to safety measures against harmful gases that may be generated around the burning
vehicle.

Vehicle Hot Surface Ignition and Mitigation Measures of R-1234yf

Refrigerant for MAC Systems
The European Commission (EC) as well as the United States Environmental Protection Agency (EPA) published
legislations to regulate or encourage the use of low Global Warming Potential (GWP) refrigerants applied to Mobile
Air Conditioning (MAC) systems. Europe mandates a GWP less than 150 of MAC refrigerants for new vehicle types.
The thermodynamic refrigerant properties of R-1234yf are slightly different from the properties of R-134a,
currently used in MAC systems. Although the basic material data show that R-1234yf is flammable, ignition tests
performed for an automotive engine under-hood environment reveal design and packaging influences of its ignition
behavior. After extensive collaborative research in 2009, the Society of Automotive Engineers Cooperative Research
Team (SAE CRP1234) concluded that R-1234yf is suitable for use in automotive applications. Further ignition risk

assessment regarding R-1234yf usage in MAC systems was done by SAE CRP1234-4 in 2013. They concluded that
risks are still very small compared to the risks of a vehicle fire from all causes and well below risks that are
commonly viewed as acceptable by the general public.
This paper will compare under-hood hot surface ignition behavior of five different vehicles under specific test
conditions. Out of these five, two vehicles were further investigated and the engine compartment designs were
assessed. A broad range of flammability testing was completed to study release of the R-1234yf and Polyalkylene
Glycol (PAG) oil mixture. Both vehicles were equipped with a transverse gas engine, front take down exhaust
routing, turbo charger, and single evaporator MAC system. This engine architecture was chosen due to the short
distance from a potential condenser leak and the hot exhaust system components in case of a frontal crash.
Although the engine architecture was the same for both vehicles, the engine design and packaging were different
and were described in this paper. Parameters that were considered include heat shield design, engine packaging,
exhaust system surface temperatures during refrigerant release, and the impact on the time to ignition of the R1234yf/PAG mixture. Mitigation measures were also identified and their potential integration into a vehicle were
discussed. As engine coolant will be released in many real-world frontal crash scenarios, the impact of vaporized
glycol water mixture on the ignition behavior was assessed. This paper further describes the MAC systems, the test
facility used, and the testing procedure applied.

Crash Test with a Hot Operating Engine and R-1234yf MAC System
As part of the launch of the refrigerant R-1234yf there were a number of studies done regarding the ignition
behavior of this new refrigerant in passenger cars. These tests were conducted by a number of automobile
manufacturers, component suppliers, and the refrigerant supplier under laboratory conditions at the component
and vehicle level. In November 2009 the international automotive industry concluded that the R-1234yf can be
used safely in automotive air conditioning systems.
Further tests were conducted by different automobile manufacturers, suppliers, and the refrigerant supplier under
various laboratory and vehicle operation conditions means hot surfaces in the engine compartment. A number of
vehicle manufactures have conducted full vehicle crash tests.
In this paper, real world accidents are analyzed using the German In-Depth Accident Study (GIDAS) database as
well as the thermal parameters for ignition of R-1234yf, i.e. concentration and surface temperature to create a
worst-case scenario. These parameters were considered while conducting a frontal impact test with a hot 1.4 liter
turbo gasoline engine with a critical overheated exhaust surface temperature and an operating R-1234yf Mobile Air
Conditioning (MAC) system. The test was conducted with an ECE R95 deformable moving barrier at a test speed of
50kph.