1
Stability Analysis and Rollover Scenario Prediction
For Tractor Semi-Trailer
Mohamed Bouteldja12, Nacer K. M’Sirdi1,
S´ebastien Glaser 2, VictorDolcemascolo2
1 Laboratoire de Robotique de Versailles, CNRS/FRE2659
10-12 avenue de l’Europe, 78140 VELIZY, FRANCE
2Laboratoire Central des Ponts et Chauss´ees
58 Bd Lef`evre, 75014 PARIS, FRANCE
Abstract—In this paper, we present a stability analysis
of heavy vehicle through variation of the parameters
influencing the rollover. A 5-DOF nonlinear dynamic
simulation model for tractor semi-trailers is proposed.
The proposed model is used to validate the prediction of
rollover. Threshold accelerations deduced using quasistatic
analysis. The predicted threshold accelerations
are compared with simulation threshold accelerations.
Index Terms—Modelling, Heavy Vehicle, Rollover,
Prediction, Stability.
I. Introduction
The tractor semitrailer represents a population of risky
vehicles, both for themselves as well as other road users.
Moreover, for a limited capacity of the infrastructure,
the number and density of heavy vehicles grow faster
than those of cars. Accidents involving heavy lorries have
serious consequences for road users, and incidents induce
major congestions or damage to the environment or the
infrastructure at disproportionate economic costs. For instance,
the risk of having a dead people is 2.4 higher when
lorries are involved in accident [1].
To reduce the number of accidents and to improve
safety, several solutions have been studied in some programs
dealing with the idea and activity of Intelligent
Transportation Systems (US NAHSC program, California
PATH Program, Japan’s AHSRA, European programs:
ADASE, REPONSE and CHAUFEUR-driven, French
PREDIT and ARCOS programs, etc.) [2]. Some orientations
of these programs may consist of driver assistance,
active safety systems (lateral control) and passive safety
systems (detection and warning message under hazardous
conditions).
Some commercial systems installed in the infrastructure
before a dangerous cornering, are able to measure the
truck speed. In case of overspeed leading to a rollover in
the cornering, a warning is sent to the driver in order
to incite him to decrease its speed. Other systems are
installed aboard the truck. They can, using informative
(passive) nature as measuring lateral acceleration, issue a
warning signal to the driver when it goes beyond some
risky thresholds. In case of active systems, the concept
is to minimize the lateral acceleration by braking action,
steering action, suspension action or anti-roll action or a
combination of all.
Naleck and al [3] have defined, for the detection of
rollover, a measurement based on the energy reserve until
to have a rollover situation (Rollover Prevention Energy
Reserve). At the same time, Dunwoody simulated the
steady state cornering performance of a tractor semitrailer
fitted with an active roll control system. The control
system required the measurement of the trailer lateral
acceleration and the relative roll angle between the tractor
and the trailer. The study stated that such a system could
raise the static rollover threshold by 20-30% [4].
Dahlberg gave the definition of the remaining energy
margin that lead to rollover [5]. This margin is the dierence
between the critical potential energy and the sum of
the potential and kinetic rollover energy.
Chen et al. defined a short predictive criterion called
TTR (Time to Rollover) which takes into account the
dynamic state evolution of the vehicle [6].
Ackermann et al. [7], [8] have studied the influence of
the inertia force and the centre of gravity height on the roll
dynamics of the vehicle heavy. He defined the load transfer
ratio (LTR) for an axle as the ratio of the dierence
between right and left impact force by the sum of right and
left impact force. If LTR is greater than 0.9, rollover risk is
high. He showed that LTR is approximately a function of
the lateral acceleration (easy to measure) and the center
of gravity height (for which no sensor is available)
Lin et al. then investigated roll control system design
using an optimal state feedback technique and a steering
input power spectrum based on road alignement data and
pseudo-random lane changes [9]. The system performance
was marginally superior to that of the lateral acceleration
feedback controller.
This work is developed as part of the ”safety of heavy
vehicle” subject of ARCOS 2004 program. The objectives
are to develop a strategy to detect critical situations and
trigger a warning signal under dangerous conditions such
as unstable yaw or rollover, and allow real-time detection
.....