Professional Documents
Culture Documents
ON
Submitted by
CHINMAY PAMECHA
2009UME453
in partial fulfillment for the award of the degree
of
BACHELOR OF TECHNOLOGY
IN
MECHANICAL ENGINEERING
AT
MALAVIYA NATIONAL INSTITUTE OF TECHNOLOGY
JAIPUR
(2009 2013)
Submitted to:
Prof. P. K. Saxena
Prof. G. Agrawal
Dr. T. C. Gupta
i
SYNOPSI S
Rollover crashes cause more than 10,000 fatalities and nearly 30,000 serious
injuries per year in the U.S. alone. This is due to the fact that the vast majority of
vehicles, including commercial, police, and military, lack the roof strength to preserve
occupant survival space and protect their occupants in a rollover. Recent statistical
and epidemiological studies have shown a significant relationship between roof
crush and injury. This rollover occupant protection problem is well known to
industries with large vehicle fleets; until now, this problem has eluded solution.
Within these various industries a wide variety of rollover occupant protection systems
(ROPS), both internal and external, have been designed, purchased, manufactured,
installed, and maintained locally with little expert consultation. A wide variety of
designs have emerged with an alarming variance in "assumed" crashworthiness.
Couple this alarming trend with the risk of rendering the existing occupant protection
features (e.g., airbags) ineffective, which has resulted in vehicles with inadequate
crashworthiness. This paper describes how rollover damage to a vehicle with a weak
roof and the resulting reduction of occupant headroom can be minimized to an
inconsequential amount using an innovative externally retrofitted rollover load
distribution device. This system was based on an understanding of road crash data,
empirical evidence, and innovative state of the art testing and analysis to provide
effective external ROPS structures for the commercial, police and military fleets.
In this review paper I am presenting a robust controller design methodology
for vehicle rollover prevention utilizing active steering, active suspension, differential
braking, center of gravity estimation and uses of multiple models and controllers.
Control design is based on keeping the magnitude of the vehicle load transfer ratio
(LTR) below a certain level in the presence of driver steering inputs; also an exact
expression for LTR is developed. The proposed controllers have a proportional-
integral structure whose gain matrices are obtained using the results of Pancake,
Corless and Brockman. These controllers reduce the transient magnitude of the LTR
while maintaining the steady state steering response of the vehicle. The controllers
can be designed to be robust with respect to vehicle parameters such as speed and
centre of gravity height. The paper also provides a modification to the controllers so
that they only activate when the potential for rollover is significant. Numerical
simulations demonstrate the efficacy of the approach and the resulting controllers.
A methodology based on multiple models and switching for real time
estimation of center of gravity (CG) position and rollover prevention in automotive
vehicles is also presented here. Based on a linear vehicle model in which the
unknown parameters appear nonlinearly, a novel sequential identification algorithm
to determine the vehicle parameters rapidly in real time is proposed. The CG height
estimate is further coupled with a switching controller to prevent untripped rollover
in automotive vehicles. The efficacy of the proposed switched multi model/controller
estimation and control scheme is demonstrated via numerical simulations.
An anti-rollover control algorithm based on the Time-To-Rollover (TTR) metric
is proposed in this paper. A simple model with steering and direct yaw moment
ii
control inputs was constructed to calculate the TTR in real-time. Numerical
simulations on a nonlinear vehicle model show that both control structures effectively
track the driver intended yaw rate during extreme manoeuvres while also limiting the
peak roll angle. During ordinary driving, the controlled vehicle behaves identically to
an ordinary vehicle. These preliminary results show that for double lane change
manoeuvres, it is possible to limit roll angle while still closely tracking the drivers
intent.
Also a tilting vehicle technology is described here. Theoretical analysis is
conducted to find Perfect Tilting Condition that is a geometric parameter set of link
mechanism that can provide both zero lateral acceleration and zero wheel load
imbalance simultaneously. This perfect tilting condition can be realized by changing
the variable length of link mechanism so that this tilting control system is a kind of
semi active control. To change the length of link mechanism on curved section,
optimal control theory is applied to the vehicle body tilting system with a newly
developed variable link mechanism. From computer simulation and experiment using
a scale model, it is clarified that the proposed tilting system is effective to suppress
both over-centrifugal acceleration and wheel load imbalance.
iii
LIST OF FIGURES
Sr. No. Figure No. Caption Page No.
1. 3.1 Positions of the distribution of loads[10]* 7
2. 4.1 Single track model with roll degree of
freedom.[5]
9
3. 4.2 Differential braking force as control input[3] 11
4. 4.3 SimMechanics vehicle model layout[3] 15
5. 4.4 Tire force components[3] 15
6. 4.5 Lateral force as a function of the slip angle for
various vertical loads[3]
15
7. 4.6 Steering wheel input history for simulated Elk-
Test[3]
17
8. 4.7 LTR comparison for bicycle, SM and linearized
SM models at v=20 m/s and s,max=60 deg[3]
18
9. 4.8 LTR comparison for bicycle, SM and linearized
SM models at v=20 m/s and s,max=80 deg[3]
19
10. 6.1 CG longitudinal position and tire stiffness
estimations[5]
24
11. 6.2 CG height and suspension parameter
estimations[5]
25
12. 6.3 Multiple model switched adaptive control
structure[5]
25
13. 6.4 Steering input and the resulting CG height
estimation[5]
28
14. 6.5 LTRd for the controlled and uncontrolled
vehicles.[5]
29
15. 6.6 Vehicle states for the controlled and
uncontrolled vehicles[5]
29
16. 6.7 Vehicle speed and the normalized control
force[5]
30
17. 7.1 Flow diagram of the PI active steering
controller[7]
35
18. 7.2 Comparison of the controlled (with fixed model)
and uncontrolled vehicles.[7]
37
19. 7.3 Comparison of the controlled (with fixed model)
and uncontrolled vehicle states and
trajectories.[7]
37
*The figure in [ ] denoted references given at the end of the report.
iv
20. 7.4 Comparison of the robustly controlled and the
uncontrolled vehicles (v = 70km/h peak =
150, and h = 0.45m).[7]
38
21. 7.5 Comparison of the trajectories and states of
robustly controlled and the uncontrolled
vehicles (v = 70km/h peak = 150, and h =
0.45m).[7]
38
22. 8.1 Top row: Vehicles CG drops during rollover
and then rises again;
Bottom row: Hoop maintains vehicles CG at
constant height, thus reducing roof load [1]
40
23. 8.2 Photograph of the major components of the
JRS[1]
41
24. 8.3 HALO Prototype 2; No noticeable
deformation to the roof structure[1]
41
25. 9.1 Government vehicle rollover victim David[1] 43
26. 9.2 Government vehicle rollover victim Luis[1] 44
27. 9.3 Military vehicle fleet - U.S. Marine transport
van case study[1]
45
28. 10.1 Rollover Air Bags[14] 47
29. 13.1 Comparison of the trajectories and states of
robustly controlled and the uncontrolled
vehicles (v0 = 140km/h, dpeak = 120, and h =
0.375m).[2]
50
30. 13.2 Comparison of continuous and switched robust
controllers at a non critical maneuver (v =
140km/h, dpeak = 50, and h = 0.375m) [7]
51
v
LIST OF TABLES
Sr. No. Table No. Title Page No.
1. 4.1 Model parameters and their
definitions[5]
9
2. 4.2 Model parameters [3] 16
3. 4.3 Conditions for wheel lift, |LTRd| = 1
and vehicle rollover [3]
19
4. 7.1 Fixed model parameters [7] 36
5. 8.1 Test results for the HALO
Prototype 2 [1]
42
vi
CONTENTS:
Synopsis.i
List of Figures..iii
List of Tablesv
1. Introduction.1
1.1 Around the world1
1.2 Types of rollovers...............................1
1.3 Control measures..2
2. Contributing Causes.4
2.1 The Physical factors4
2.2 The Driving factors..5
2.3 The human factors..5
3. Drivers Training6
3.1 Vehicle Center of Gravity6
3.2 Load Security.6
3.3 Radius of Curves and Slope of Roadways...6
3.4 Vehicle Speed .7
3.5 Trailer towing.7
3.6 Vehicle condition and preparation8
4. Vehicle Modelling and LTR
d.
.9
4.1 Vehicle model and control design..9
4.2 The load transfer ratio (LTR
d
)...11
4.3 Actuators, sensors and parameters13
4.4 A high fidelity non-linear simulation model..14
4.5 Simulation of SM model17
5. Vehicle Parameter Identification20
5.1 Identification of lateral dynamic parameter20
5.2 Identification of roll dynamic parameter..21
5.3 Numerical Analysis22
6. Switching Rollover Controller (Center of Gravity Estimation)24
6.1 Rollover controller based on single model26
6.2 Switched rollover prevention26
6.3 Numerical analysis.27
7. Active Steering..31
7.1 State feedback controller for robust disturbance attenuation.31
7.2 Rollover control design..34
7.2.1 Actual steering PI controller with known parameters34
8. ROPS (Rollover Occupants Protection System).39
8.1 The Concept39
8.2 The Apparatus39
8.2.1 The JLS39
8.2.2 HALO
TM
Prototype 2. Development and testing41
9. Case Studies.43
9.1 U.S. Border Patrol vehicles..43
9.2 U.S. Marine transport van.44
10. Global Scenario..46
vii
10.1 ESC (Electronic Stability Control)..46
10.2 Rollover air bags.........47
11. Indian Scenario.48
12. Future Aspects48
13. Conclusion..50
References..a
Bibliographyb
Annexure.c
P a g e | 1
1. INTRODUCTION
1.1) Around the world:
It is well known that vehicles with a high center of gravity such as light trucks
(vans, pickups, and SUVs) are more prone to rollover accidents than other types of
passenger vehicles. According to recent statistical data, this class of vehicles were
involved in nearly 70% of all the rollover accidents in the USA during 2004, with
SUVs alone were responsible for about half of this total. The fact that the
composition of the current automotive fleet in the U.S. consists of nearly 36% light
trucks, along with the recent increase in their popularity worldwide, makes rollover an
important safety problem.
Each year in the United States (U.S.), approximately one in every three traffic-
related deaths is the result of a rollover crash. Australia has similar annual statistics;
about one in every four occupant fatalities occurred in vehicles that rolled. In Europe,
around 15% of vehicle occupant fatalities involve some element of a rollover[1]. The
high incidence of rollover crashes have become a concern to global mining and
petroleum companies, U.S. border patrol police and U.S. military personnel because
of occupational health and safety requirements. The vehicle is considered a
workplace environment. Hence, duty of care to provide a safe work environment
extends to employees, police and military personnel travelling in either a company or
government vehicle, including protection in a rollover crash. This is particularly true if
the driver was travelling at the posted speed limit, all occupants were seat belted,
and the crash was not due to driver or occupant error or negligence.
1.2) Types of rollovers:
There are two distinct types of vehicle rollover: tripped and un-tripped[2]. A
tripped rollover commonly occurs when a vehicle slides sideways and digs its tires
into soft soil or strikes an object such as a curb or guardrail. Driver induced un-
tripped rollover however, can occur during typical driving situations and it poses a
real threat for top-heavy vehicles. Examples are excessive speed during cornering,
obstacle avoidance and severe lane change manoeuvres, where rollover may occur
as a direct result of the lateral wheel forces induced during these manoeuvres.
Rollover has been the subject of intensive research in recent years, especially by the
major automobile manufacturers, and the majority this work is geared towards the
development of rollover prediction schemes and robust occupant protection systems.
While robust active rollover control systems achieve the goal of preventing this type
of accidents, such a control approach may be too conservative, and it can potentially
compromise the performance of the vehicle under noncritical driving situations. It is
however, possible to prevent rollover accidents in an effective and efficient manner
by continuously monitoring the car dynamics and applying the proper and sufficient
control action to recover handling of the vehicle in emergency situations.
P a g e | 2
1.3) Control measures:
The height of CG along with the lateral acceleration are the most important
parameters affecting the rollover propensity of an automotive vehicle; while the
former is available as part of standard sensor packs, the CG height can not be
measured directly. With this background in mind, we first propose our CG estimation
method based on multiple models, and then use the technique in designing a
switching rollover controller. As part of the feedback implementation we utilize
multiple simplified linear models, which are parameterized to cover uncertainty in the
vehicle parameters. Switching between these models yields a rapid estimation of
unknown and time-varying vehicle parameters through the selected models, which is
then used to switch among a set of suitable controllers in order to improve the
performance of active rollover mitigation systems[4].
Our motivation for considering a switching controller implementation is
twofold. Firstly, switching controllers are the alternative option to the robust ones and
they can potentially provide higher performance. Robust controllers have fixed gains
that are chosen considering the worst-case that the plant undergoes; for the rollover
problem, the worst operating condition translates to operating the vehicle with the
highest possible CG position[3]. While choosing the controller gains for the worst-
case guarantees the performance (i.e., safety) under the designed extreme
operating condition, the feedback performance of the robustly controlled systems
under less severe or even normal operating conditions are suboptimal. Our second
motivation is related to the time constant of rollover accidents, which is on the order
of seconds. While conventional adaptive controllers are known to have slow
convergence rates and large transient control errors when the initial parameter errors
are large (a factor that renders these control approaches unsuited for use in rollover
mitigation applications), utilization of MMST type algorithms may overcome these
problems and provide high performance adaptive controllers. Therefore, when
improving the controller performance and speed for the rollover problem is
considered, MMST framework come an ideal choice as it can provide rapid
identification of the unknown parameters as part of the closed loop implementation.
This way we can rapidly switch to a controller that is more suitable for the vehicle
operating conditions, thus improving the overall safety of the vehicle without
sacrificing its performance.
In this paper a robust rollover prevention controller design methodology based
on differential braking is also presented. The proposed control design is an
application of recent results on the design of control systems which guarantee that
the peak values of the performance outputs of a plant do not exceed certain
thresholds when subject to bounded disturbance inputs. The main selected
performance output for the rollover problem is the Load Transfer Ratio LTRd. This
measure of performance is related to tire lift-off and it can be considered as an early
indicator of impending vehicle rollover[5]. The paper also includes the braking force
as a performance output to take into account limitations on the maximum braking
force. The aim of our control strategy is to maximize the magnitude of the allowable
disturbance inputs which do not drive the performance outputs outside their
P a g e | 3
prespecified limits; in this case the disturbance input is the driver steering input. We
also want to guarantee robustness with respect to the parameter uncertainty that
arises from changing vehicle speed. We indicate how our design can be extended to
account for other sources of uncertainty such as unknown vehicle center of gravity
and tire stiffness parameters.
In terms of serious injuries, the National Highway Traffic Safety Administration
(NHTSA) estimates around 24,000 serious injuries occur in approximately 273,000
rollover crashes. Recent work by Mandell et al investigated the Crash Injury
Research Engineering Network (CIREN) database and the National Automotive
Sampling System (NASS) CDS database for belted, adult (older than 15 years),
outboard occupants in rollover crashes from 1999 to 2009. They used a logistic
regression to establish a relationship between roof crush magnitude and severe-to-
fatal (AIS3) injury. They found that the risk of mortality, traumatic brain injury, spine,
and spinal cord injury increased with increasing roof crush. For spine injury,
increased risk was observed when roof crush exceeded 3 inches (8 cm).
The relationship between roof deformation and serious injury in rollovers was
also investigated by Rechnitzer and Lane in 1994. Their study of Australian crashes
concluded that roof crush was a significant factor in serious-to-fatal injury occurrence
in rollovers. It is worth noting that rollover crashes are the leading cause (17%) of
spinal cord injury in Australia.
A retrofitted structural system that distributes roof loads and can minimize roof
crush during a rollover crash will best protect occupants. In order to distribute the
roof loads the geometry of the roof structure must be changed. This paper extends
the work of Grzebieta et al (2009). The hypothesis was that loads applied to the roof
structure and hence roof crush and injury risk can be reduced by reinforcing the roof
structure of a weak roof vehicle. The authors showed this using the proof of
concept device (HALO
TM
prototype 1)[1], a hoop mounted to an external frame to
hold the vehicle up as it rolls from one side of the roof to the other, thus changing the
geometry[1]. Results from the production vehicle and HALO
TM
Prototype 1 are
presented in Grzebieta et al (2009). This paper provides further proof of this concept.
The evolution of the HALO
TM
prototype 1 into the current HALO
TM
rollover load
distribution device is described, including the design, construction, and JRS testing
of 3 prototypes. A dolly rollover test of a HALO
TM
prototype is also presented here.
The innovative HALO
TM
rollover load distribution device was designed for and is well
suited for the Commercial, Police, and Military Vehicle fleets.
P a g e | 4
2. CONTRIBUTING CAUSES
Vehicles can roll over in several ways. These include excessive cornering
speed, tripping, collision with another vehicle or object, or traversing a critical slope.
Rollovers caused by excessive cornering speed occur when cornering forces
destabilize the vehicle. As a vehicle rounds a corner, three forces act on it: tire
forces, inertial effects, and gravity. The cornering forces from the tire push the
vehicle towards the center of the curve. This force acts at ground level. The force of
inertia acts horizontally through the vehicle's center of mass in the direction opposite
to the one it is turning. These two forces make the vehicle roll towards the outside of
the curve. The force of the vehicle's weight acts downward through the center of
mass in the opposite direction. When the tire and inertial forces are enough to
overcome the force of gravity, the vehicle starts to turn over. Most passenger
vehicles will slide or spin before this happens, but this is a common type of rollover
for taller vehicles, including light trucks (SUVs, vans and pickup trucks), buses, and
heavy trucks[9].
A vehicle may perform a complete rollover, rolling onto its side, roof, far side
and finally ending up on its wheels again, or even going on to roll further.
Tripping rollovers occur when a vehicle is sliding sideways, and the tires strike
a curb, dig into soft ground, suddenly regain traction, or a similar event occurs that
results in a sudden lateral force. The physics are similar to cornering rollovers.
A collision with another vehicle or object can cause a rollover. These occur
when the collision causes the vehicle to become unstable, such as when a narrow
object causes one side of the vehicle to accelerate upwards, but not the other,
causing the vehicle to rotate along its long axis. A side impact can accelerate a
vehicle sideways. The tires resist the change, and the coupled forces rotate the
vehicle. In 1983, crash tests showed that light trucks were prone to rolling over after
colliding with certain early designs of guide rail.
A rollover can also occur as a vehicle crosses a ditch or slope. Slopes steeper
than 33 percent (one vertical unit rise or fall per three horizontal units) are called
'critical slopes' and often contribute to rollovers.
A vehicle may roll over for other reasons, such as when hitting a large
obstacle with one of its wheels or when manoeuvring over uneven terrain.
These factors can be listed as follows[8]:
2.1) The Physical factors:
- The arc of the curve and the bank of the roadway.
- Steepness of the crown of the road.
- Liquid sloshing: worst case liquid sloshing occurs when tankers are filled
to 40-70% capacity, and the tanker is not equipped with baffles.
- Speed at which you enter the curve or ramp.
P a g e | 5
- You need friction to turn: pavement friction generated by truckes is less
than that generated by a passenger car, but the effective friction demand
for trucks is higher.
- High crosswinds.
2.2) The Driving factors:
- Tight radius turns.
- Short deceleration lanes preceding tight radius exits.
- Curbs on the outside of curves.
- Downgrades leading into ramps.
- Unrealistic speed limits for ramps off of high speed highways.
- Curves with a negative bank (usually for drainage purposes).
- Tripping.
- Traversing a critical slope.
- Collision with another vehicle.
2.3) The Human factors:
- Driver fatigue.
- Driver inexperience in transporting loads with a high center of gravity.
- Drifting off the road, then quickly counter-steering.
- Failure to anticipate sharp ramp or curve.
- Entering a turn or ramp too fast.
- Driver distraction.
P a g e | 6
3. DRIVERS TRAINING
Start by training your drivers in the six important factors that can affect vehicle
stability. These six factors are[10]:
3 Vehicle Center of Gravity.
4 Load Security.
5 Radius of Curves and Slope of Roadways.
6 Vehicle Speed
6.2 Adjusted to environmental/road conditions.
7 Trailer Towing.
8 Vehicle Condition and Preparation.
* Should wear Safety Belts.
Vehicle control factors:
3.1) Vehicle center of gravity:
The height of a vehicle's center of gravity and the length of the
wheelbase determine the vehicle's stability.
3.2) Load security:
Improperly secured loads can change a vehicle's center of gravity and
its stability. Bulk tank trucks are inherently less secure because fluids can
surge when trucks brake or go around curves, thereby altering the center of
gravity. Also, a vehicle loaded with containers will have a higher center of
gravity. Additionally, it is important that payloads are secured as closely as
possible to the lateral centerline of the truck or trailer bed. If the payload is
not centered properly, the vehicle stability will not be equivalent when turning
to both the right and left. (see fig 3.1).
3.3) Radius of curves and slope of roadways:
These are important because they generate a centrifugal force that
acts sideways on the vehicle, thereby decreasing vehicle stability.
Figure 3.1 (a)
P a g e | 7
3.4) Vehicle speed:
As the vehicle's speed increases, the centrifugal force, or sideways
force increases. Faster speeds also result in decreased driver response
times. Speed is the factor over which the driver can exercise the most
control. When maneuvering through curves or sudden traffic situations, a
vehicle with a high center of gravity can easily turn over.
- Speed is even more important when the movement of liquid cargo is "in
phase" with the vehicle's maneuver. If the liquid is on one side during the first
curve, then shifts to the other side during the next curve, the liquid is
positioned to shift back to the first side with four times the side force it had
during the initial curve.
- Sudden vehicle maneuvers are especially risky because the combination of
speed and load shift makes the vehicle unstable.
3.5) Trailer towing:
Vehicles towing trailers are much more prone to roll over, especially in
curves and during sudden steering maneuvers, as a result of the exaggerated
motion of the trailer.
Figure 3.1 (b)
Figure 3.1 (c)
Figure 3.1 Positions of the distribution of loads [10]
P a g e | 8
3.6) Vehicle condition and preparation:
It is critical the vehicle is in good operating condition before starting
your mission, with particular attention paid to the tires condition and air
pressure.
P a g e | 9
4. VEHICLE MODELLING AND LTR
d
In this section we introduce the model that we use for controller design. We
define the rollover detection criterion LTRd and present the assumptions on the
sensors and actuators used in the design. We also present the higher fidelity
nonlinear multi-body simulation model to which the controllers will be applied. For
related notation, refer to Table 4.1.
Table 4.1: Model parameters and their
definitions [5]
Parameter
m
Description
vehicle mass
Unit
[kg]
v vehicle speed [m/s]
J
xx
J
zz
roll moment of inertia at CG
yaw moment of inertia at CG
[kg m
2
]
[kg m
2
]
a longitudinal CG position w.r.t. front axle [m]
b longitudinal CG position w.r.t. rear axle [m]
T vehicle track width [m]
h distance of CG from roll axis [m]
c
k
C
, f
suspension damping coefficient
suspension spring stiffness
linear tire stiffness for front tire
[N m s/rad]
[N m/rad]
[N/rad]
C
,r linear tire stiffness for rear tire [N/rad]
steering angle [deg]