Test Procedures and Evaluation Tools

for Passenger Vehicle Dynamics

Master’s thesis in Automotive Engineering

ANDERS KARLSSON
Department of Applied Mechanics
Division of Vehicle Engineering and Autonomous Systems
Vehicle Dynamics Group
CHALMERS UNIVERSITY OF TECHNOLOGY
Göteborg, Sweden 2014
Master’s thesis 2014:23
MASTER’S THESIS IN AUTOMOTIVE ENGINEERING

Test Procedures and Evaluation Tools

for Passenger Vehicle Dynamics

ANDERS KARLSSON

Department of Applied Mechanics

Division of Vehicle Engineering and Autonomous Systems
Vehicle Dynamics Group
CHALMERS UNIVERSITY OF TECHNOLOGY
Göteborg, Sweden 2014
Test Procedures and Evaluation Tools
for Passenger Vehicle Dynamics
ANDERS KARLSSON

c ANDERS KARLSSON, 2014

Master’s thesis 2014:23

ISSN 1652-8557
Department of Applied Mechanics
Division of Vehicle Engineering and Autonomous Systems
Vehicle Dynamics Group
Chalmers University of Technology
SE-412 96 Göteborg
Sweden
Telephone: +46 (0)31-772 1000

Cover:
Example of summary graphs produced using the analysis tool written within this master thesis work.

Chalmers Reproservice
Göteborg, Sweden 2014
Test Procedures and Evaluation Tools
for Passenger Vehicle Dynamics
Master’s thesis in Automotive Engineering
ANDERS KARLSSON
Department of Applied Mechanics
Division of Vehicle Engineering and Autonomous Systems
Vehicle Dynamics Group
Chalmers University of Technology

Abstract
Both within education and research it is necessary to be able to ensure that the real or virtual vehicle used,
behaves as supposed. In order to verify this a massive set of test procedures is needed. However, to create a
ruff overview showing the basic behaviour, some test procedures are implemented into an evaluation tool. The
user of this tool is presented the results in a form of a report with numbers and diagrams

Within this thesis a set of test procedures have been selected and implemented in an evaluation tool divided in
to virtual test and analysis in order to both be able to use it for simulations as well as for post processing of
real test data.
This report focuses on the choice and implementation of a limited number of test procedures as well as the
design of the evaluation tool developed.
The tests that have been implemented are: Steady state cornering, Sine with dwell, Continuous sinusoidal
input and Straight ahead acceleration.
Other tests that were discussed and were suitable to be implemented at a later stage are: Step input, Random
input, Sinusoidal input, one period, Pulse input, Braking with split coefficient of friction, Brake in turn and
Straight braking.
In order to give the user freedom to adjust the tests based on current requirements, all test procedure constants,
e.g. times, velocities and distances, have been parametrized using an excel reference document to ensure good
adaptability
The tool developed within this test is easy to set up and supports batch processes which enables it to run
several simulations using different models. This have been done with one model from the research project
Balance Active and Passive Safety (BAPS).
The evaluation tool is based on an combination of Matlab and Simulink.
The results from the analysis part of the tool are presented as an pdf document.

The purpose with creating this tool is to provide a method to easily and quickly get results on the be-
haviour of a real vehicle or a vehicle model during a few standard tests. This purpose is fulfilled with the tool
created within this thesis.

Keywords: Test Procedures, Vehicle verification, Model verification, Vehicle dynamics

i
ii
 Preface
The work with the master thesis has been conducted at the Division of Vehicle Engineering and Autonomous
Systems under supervision by Bengt Jacobson and Gunnar Olsson.
This thesis have been written at the Division of Vehicle Engineering and Autonomous Systems facilities.

Acknowledgements
First and foremost would like to express my gratitude to my examiner Professor Bengt Jacobson for all his
interest, important input and good ideas regarding my thesis. I would like to thank Gunnar Olsson for his
valuable input regarding the design, layout and choice of test manoeuvres, Ulrich Sander for all the help with
the BAPS model and its vehicle parametrisation and Vehicle dynamics at Volvo Cars, for the productive
discussions regarding the choice of test procedures.

I would as well like to thank all my friends within Chalmers Student Union, FestU, KL, and Balliancen that I
have had the privileged to work and spend time with during my studies. An extra acknowledgement goes to
Björn, for all cooperation during our courses, Karin for all good discussions and help with proofreading, and of
Andreas for all good times during my time in Gothenburg.
Finally would as well like to thank Louis Tellier for all interesting and beneficial discussions, as well as all the
fun times at the department during this thesis.

iii
iv
 Nomenclature
Abbreviations
BAPS Balance Active and Passive Safety [Bapa]
BMAP Brake Mean Effective Pressure
rpm Revolutions Per Minute
FMU Functional Mock-up Unit [Fmi]
FMI Functional Mock-up Interface [Fmi]
ESC Electronic stability control

Matlab variables
StopSignal Stop signal for simulation
T ime Time vector
SW A Steering wheel angle
T hrottle Acceleration pedal position
Brake Deceleration pedal position
LongAcc Longitudinal acceleration
LatAcc Lateral acceleration
Y awAcc Yaw acceleration
LongV el Longitudinal velocity
LatV el Lateral velocity
Y awV el Yaw velocity
X position Vehicle position along environment X-axis
Y position Vehicle position along environment Y-axis
P si angle Vehicle angle compared to staring position
P itchAcc Pitch acceleration
P itchV el Pitch velocity
P itchAng Pitch angle
RollAcc Roll acceleration
RollV el Roll velocity
RollAng Roll angle
alphaF L,F R,RL,RR Tyre slip angle
F xW heelF L,F R,RL,RR Wheel force along tire X-axis
F zW heelF L,F R,RL,RR Wheel force along tire Y-axis

v
vi
 Contents

Abstract i

Preface iii

Acknowledgements iii

Nomenclature v

Contents vii

1 Introduction 1
1.1 Objective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Limitations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

2 Background and Theory 2

2.1 Literature review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
2.2 Development tools . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
2.3 Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

3 Selection of test procedures 4

3.1 Method & Development Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
3.2 Selection of tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
3.2.1 Implemented test procedures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

4 Evaluation Tool 8
4.1 Virtual test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
4.1.1 Method & Development Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
4.1.2 Results regarding virtual test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
4.2 Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
4.2.1 Method & Development Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
4.2.2 Results regarding analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

5 Simulations with BAPS model 15

5.1 Set up options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
5.1.1 Parametrized vehicles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
5.1.2 Adjustable safety systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
5.2 Results from the simulations with the BAPS model . . . . . . . . . . . . . . . . . . . . . . . . . . . 16

6 Conclusions 16
6.1 Future work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16

References 18

A Example report BAPS model

B Example report SAAB Sim whith ESC and EVAL

C Parametrization of BAPS model

D Batch run results using BAPS model

E Folder strucutre of Evaluation tool

vii
viii
1 Introduction
This master thesis is written within the Automotive engineering master program at Chalmers University of
Technology. It focuses on the choice and implementation of a limited number of test procedures as well as the
design of the evaluation tool developed within this thesis work.
The meaning with creating this tool is to provide a method to easily and quickly get results on the behaviour
of a real vehicle or a vehicle model during a few standard tests.
The tool developed within this thesis work should be possible to use both on more advanced research models
as well as simpler educational models.
In order to formulate and verify requirements on complete vehicle dynamics behaviour a massive set of test
procedures are needed. Same test procedures are preferable used for verification with virtual model in real
vehicles. Again same procedure are preferably used when validating a virtual model vs the real vehicle.

1.1 Objective
The two aims of this Master Thesis are:

• to define a set of test procedures for evaluation of a passenger vehicle, real or virtual, from a vehicle
dynamics perspective, to be used in education or research.

• set up a verification tool based on these test procedures.

Deliverables

• Determine a set of test procedures that can be used in order to evaluate the overall dynamics properties
of a vehicle. The chosen test procedures should reflect commonly used types of vehicle requirements.

• Reproduction of test procedures in simulation environment using Matlab. The simulation framework
should provide easy set up and import of models.

• A script that generates a pdf report after each test with results of each requirement and saves needed
data.

1.2 Limitations
In order to ensure the feasibility within the given time frame, several limitations has been made within this
master thesis.

Test procedures limitations

The following limitations has been made in order to limit the types and amount of test procedures that will be
included in this thesis work.

• Limited amount of test procedures, five to ten

• No advanced tracks

• No advanced driver models

• To ensure the result quality no subjective tests will be included.

• Only lateral and longitudinal dynamics will be considered.

• No tests that requires unreasonable advanced models or model environment, e.g. side winds or ADAS
sensors.

• No tests that requires steering wheel force feedback.

1
Model limitations
In order ensure that the tool can be used with different types of models, some overall limitations have been
declared.
• Limited number of model inputs,
– Steering wheel angle
– Acceleration pedal position, value between 0-1
– Deceleration pedal position, value between 0-1
• When using the parametric models from the BAPS research project [Bapa] they will be regarded as black
box models. No changes will be made to the vehicle model.
• In and outputs as decided in this thesis. The decision is inspired by the model from the BAPS project.

Choice of platform
The platforms used in this masters thesis are Matlab [MAT13] and Simulink [Sim13], this due to that the final
tool should be accessible for student as well as staff at Chalmers University of Technology without the need
of investments in additional licences. A Matlab/Simulink solution is as well FMI compliant which opens the
possibility of using models from multiple platforms [Fmi]. Using an FMI solution requires however a licence
from Modelon.

2 Background and Theory

When designing a vehicle an extensive set of vehicle requirements and verification methods are used. Within
research projects and education there is as well a need for requirements and verification procedures, even if not
as extensive. Since research and education often focuses on a specific function in the vehicle it is important to
have a more generic requirement set in order to ensure the overall behaviour of the vehicle. In order to verify a
vehicle, real or virtual, it is necessary to have test procedures related to the requirements. If validating, or
verifying, a virtual vehicle is conducted through comparison with test results from a real vehicle. Same set of
test procedures should be used for both real and virtual vehicles.

2.1 Literature review

In order to use relevant test procedures a literature study has been preformed. Within this study the following
sources have been evaluated in order to determine relevant test manoeuvres.
• Swedish Standards Institute containing ISO standards
• SAE International Digital Library
• FMVSS, Federal Motor Vehicle Safety Standards (Sine with dwell FMVSS216)
• EuroNCAP (information regarding ESC standard)
• NHTSA (information regarding ESC standard)
• ISO/TC 22 Road vehicles (Relevant parts regarding passenger cars.)
• Motor sport Magazines (e.g. Acceleration 0 - 100 km/h)
The general result from the review was that it were possible to find ISO standards containing tests as well as
legislations such as FMVSS216. However it were significantly harder to obtain information comparing different
tests as well as evaluating the tests and discussing advantages and disadvantages regarding them.
During the literature review information regarding different ways to export the simulation or test results
from Matlab to a pdf document were investigated, as well as methods to coordinate simulations between Matlab
and Simulink.

2
2.2 Development tools
Within this thesis work several programs have been used. The programs that are essential for running the
evaluation tool are: Matlab, Simulink and a Latex installation, e.g. MikTex.

MathWorks
For all simulations and post processing Matlab [MAT13] and Simulink [Sim13] have been used. Matlab have
been used to prepare the simulations, change settings and execute a batch process using a simulink model. The
post processing have been preformed using Matlab and the results were exported to a .tex document in order
to be compiled to a pdf-document from Matlab using MiKTex [MiK].

Dymola
The platform used for the BAPS model, presented in chapter 5, is the modelling and simulation environment
Dymola [Dym13]. This has as well resulted in the use of FMI Toolbox for MATLAB/Simulink [FMI13]

MiKTex
To compile .tex files the Latex installation MiKTex [MiK] has been used together with the following latex
packages: babel, verbatim, inputenc, microtype, mathtools, booktabs, float and geometry.

2.3 Models
In order to ensure that the tool can be used with different types of vehicle models, four different vehicle models
have been used for this thesis work. All vehicle models except the BAPS models has been built within this
thesis work.

Single track model

A simple bike model based on values from a parametrisation of Saab 9.3, with a tire model based on Pacejkas
”Magic Formula” tire model [Pac02] have been used. This model does not account for roll or pitch behaviour.

Two Track Model

The two track model used in this thesis is based on models created during the Vehicle Dynamic Advances
course at Chalmers University of Technology [Doc]. It is using reference values from a SAAB 9.3 and with a
longitudinal force output described in equation 2.1, where FOutput is the longitudinal output force on the tires,
CM axLongF orce is the maximum longitudinal output force and PAccelP edal is the acceleration pedal position
(value between 0 and 1 where 1 is full throttle). It can be used as a tricial model of the drive train. The tire
model used in this two track model is based on Pacejkas ”Magic Formula” tire model [Pac02].

FOutput = CM axLongF orce · PAccelP edal (2.1)

Two Track Model with Drive-train

Based on the previous explained two track model, an modified version with a more realistic power-train have
been designed.
It is equipped with a very simple engine based on a example BMAP curve from a sedan vehicle as well as a
gearbox designed without delays and which changes gears at pre determined engine rotational speeds. This
model has been designed to test acceleration cases in order to be able to evaluate different typed of drive-trains
and gearboxes.

Two Track Model with ESC

This is a second modification of the previous presented two track model but with ESC. The ESC is designed to
compensate for the yaw velocity error by adding braking forces to specific wheels if the yaw error exceeds a
boundary condition.

3
BAPS model
Balancing Active and Passive Safety (BAPS) [Bapa] is a research project between Autoliv, Volvo Car Corporation,
VTI, Chalmers and Semcon. The main objective with this project is ”to develop a methodology for the estimation
of how much present and future active and passive safety measures, and combination of these, will reduce the
risk for people involved of sustaining injuries of different severities” [Bapb].
An Functional Mock-up Unit (FMU) from the BAPS project have been used as one of the vehicle models
within the work of this masters thesis. This both to evaluate the model as well as verify the compatibility of
the evaluation tool.
During this project the BAPS model will be treated as an black-box model. No modifications will be made
to the model.
This model have been used with several sets of parameters, the different parametrizations are presented in
Appendix C
The parameters altered during the simulations with the BAPS model are shown in Table 2.1. This set of
parameters have been provided from the BAPS project.

Table 2.1: Parameters adjusted for each parametrization during the simulations of the BAPS model.

Parameter: Unit:
Weight [kg]
Distance CoG FrAxle [m]
Distance CoG ReAxle [m]
Height CoG [m]
Track Width Front [m]
Track Width Rear [m]
Inertia about x-axis [kg·m2 ]
Inertia about y-axis [kg·m2 ]
Inertia about z-axis [kg·m2 ]

3 Selection of test procedures

In this section the methods and selection of a suitable set of test procedures will be discussed. The target with
the set of test procedures are to supply a ruff overview of a, real or virtual, vehicle and conclude which test
procedures to implement in the evaluation tool.

3.1 Method & Development Process

In order to select the most beneficial test manoeuvres for implementation in the tool, two purposes with the
tests were considered, tests that can be used to validate a real vehicles behaviour and that can be used to verify
a vehicle model and show the models limitations.
Within this masters thesis an attempt to get a ruff overview of the vehicle or models behaviour will be
conducted by using only a few tests. By preforming five to ten tests it is not possible to get a complete overview
of a vehicles behaviour.
The used test manoeuvres should preferably be possible to preform both in simulations and during testing
of a real vehicle, in order to be able to compare simulation and test data.

Considered test manoeuvres

The following tests were considered to be used as test manoeuvres on the information found regarding them
during the literature review. Some of the tests that will be discussed are not ISO standards and this will be
taken in to account when choosing which test manoeuvres to implement. At this stage all test manoeuvres
using the surroundings, e.g. test of lane departure warning, have been removed.

4
Steady state Steady state manoeuvres reveals important information about the vehicles be-
cornering haviour. The results from this manoeuvre are uses as reference values in several
ISO 4138 other test manoeuvres such as a sine with dwell [Nat07] and Continuous sinusoidal
input.
This manoeuvre revels the vehicle under/over-steer behaviour, roll angle at steady
state, steering as function of lateral acceleration and side slip as function of lateral
acceleration.
The steady state cornering manoeuvre can be preformed in three different ways, us-
ing constant radius, constant velocity or constant steering angle. More information
regarding the manoeuvre is available in:[ISOe]

Sine with dwell Test of over-steer intervention and responsiveness. Used to test ESC. ISO standard
FMVSS126 S7.9 is under development. Efficient manoeuvre to excite an over-steer response from a
vehicle [FEO05] More information regarding the manoeuvre is available in:[Nat07]

Fishhook This manoeuvre original designed to quantify on-road, untripped roll overs prop-
NHTSA erties. It might not provide the inputs needed for evaluating vehicle with ESC.
More information regarding the manoeuvre is available in:[FEO05]

Sine steer Base on Sinusoidal input, one period, but with a 1.3 times larger amplitude the
increased second half cycle. Similar results as Sine with dwell, however inconstant results of
amplitude different vehicles [FEO05].

Double lane To determine behaviour at a double lane change and the road holding ability of
change passenger cars. More information regarding the manoeuvre is available in:[ISOc]
ISO 3888-1

Sinusoidal Lateral acceleration related to steering wheel angle and yaw velocity related to
input, one steering wheel angle, in the time domain. Vehicle transient response to one period
period of sinusoidal steering input. Not fully representative to real driving but similar to
ISO 7401 (ISO lane change manoeuvres. More information regarding the manoeuvre is available
8725) in:[ISOf]

Obstacle Defines behaviour of vehicle at a severe lane change in order to avoid an obstacle.
avoidance More information regarding the manoeuvre is available in:[ISOd]
ISO 3888-2

Step input Lateral acceleration related to steering wheel angle and yaw velocity related
ISO 7401 to steering wheel angle, in the time domain.Gives transient response to step
input, including response times and overshoots. More information regarding the
manoeuvre is available in:[ISOf]

Pulse input Lateral acceleration related to steering wheel angle and yaw velocity related to
ISO 7401 steering wheel angle, in the frequency domain. Provides frequency response (gain
and phase angle functions) More information regarding the manoeuvre is available
in:[ISOf]

Random input Lateral acceleration related to steering wheel angle and yaw velocity related to
ISO 7401 steering wheel angle, in the frequency domain. Applies where the vehicle behaviour
(ISO 8726) is assumed to be linear. Provides high amount of information over a limited range
of lateral acceleration correlating to normal public road driving. More information
regarding the manoeuvre is available in:[ISOf]

5
 Continuous Lateral acceleration related to steering wheel angle and yaw velocity related to
sinusoidal input steering wheel angle, in the frequency domain. More information regarding the
ISO 7401 manoeuvre is available in:[ISOf]

Stopping Straight line braking from 100 kph down to 0 kph. Gives information about
distance at braking distance. It does as well show the stability of the vehicle at straight line
straight-line brak- braking. If the vehicle does not have ABS brakes this test can still be used to get
ing with this information. More information regarding the manoeuvre is available in:[ISOb]
ABS
ISO 21994:2007

Braking with The purpose of this test is to determine course holding and directional behaviour.
split coefficient The results are compared to the vehicle steady state cornering behaviour [ISOe].
of friction More information regarding the manoeuvre is available in:[ISOa]
ISO 14512

Brake in a turn Results in information regarding yaw stability change in path and change in lateral
ISO 7975 acceleration compared to steady state. More information regarding the manoeuvre
is available in:[ISOg]

Power off Gives information about how the vehicle behaves when releasing the acceleration
reaction of a pedal during cornering More information regarding the manoeuvre is available
vehicle in a turn in:[ISOh]
ISO 9816

Acceleration Provides information regarding drive-train, traction control, shifting properties,

0-top speed time 0-100 kph, time 402 m and acceleration margin.

Accelerating with Similar procedure to ”Braking with µ-split, ISO 14512”. This test could provide
split information regarding directional stability and yaw stability.
coefficient of
friction

Accelerating in Similar procedure to ”Brake in a turn, ISO 797”. Results in information regarding
a turn yaw stability change in path and change in lateral acceleration compared to steady
state.

Within this thesis there have been a focus on lateral dynamics since that is the main focus area of the
vehicle dynamics group at the division of Vehicle Engineering and Autonomous Systems. The has contributed
to the decision that it is suitable to have one test trigging the ESC system. In order to ensure the longitudinal
dynamics it were considered suitable to have one acceleration test.
Based on the limitations discussed in section 1.2 Limitations several tests were discarded, e.g. even thou
double lane change is an efficient way to test the transient behaviour of a vehicle or model it is not selected in
order to avoid dependency of advanced driver models. The selection of tests are further discussed in section 3.2
Selection of tests.

3.2 Selection of tests

Due to the limitation of five to ten tests it is not possible to get a complete overview of the behaviour. However
when limited to this amount of tests, a ruff overview can be provided. This give the user an idea of the behaviour
and for example the information needed to determine if a model behaves in reasonable way and is working.
When choosing test procedures a target have been to focus on all areas of the vehicle, from acceleration, braking,
steady state cornering, evaluate the vehicle stability and ESC system as well as evaluate the transient behaviour.

6
It were as well considered the necessary of that the selected tests could be preformed both on virtual and real
vehicles.
Except with regard to the Sine with dwell manoeuvre there have been a considerable shortage of information
regarding how efficient different manoeuvres are to evaluate behaviours as well as comparison between different
manoeuvres. This have resulted in that the decisions have been largely based on discussions with Bengt
Jacobsson at Chalmers University, Gunnar Olsson at LeanNova and Chalmers as well as discussions with
representatives for Handling and Braking at Volvo Cars.
The chosen tests within this thesis are shown in table 3.1. Due to the time limitation, not all test procedures
have been implemented which will be further discussed in section 3.2.1

Table 3.1: Chosen test procedures and which test procedures that have been implemented in the evaluation
tool.

Test procedure: Implemented:

Steady state cornering Yes
Acceleration 0-max speed Yes
Sine with dwell Yes
Continues Sinusoidal input. Yes
Step input No
Random input Only post process
Sinusoidal input, one period. No
Pulse input No
Braking with split coefficient of friction No
Brake in turn No
Straight braking No

3.2.1 Implemented test procedures

The implementation of tests have been done in accordance with existing standards. More detailed information
regarding how the test procedures are implemented in the Evaluation tool are availiable in section 4, Evaluation
tool.
Several of the tests have specified standard conditions, e.g. Steady state cornering with constant speed is in
the standard case driven at 100 kph however can be changed in 20 km/h increments according to the standard
[ISOe]. Another adjustable parameter when running Steady state cornering with constant speed is the increase
in steering wheel angle. Similar adjustable parameters can be found in all test procedures and therefore all test
procedures implemented in this thesis needs to have several parametrised settings in order to supply the user
with relevant information.
Another type of adjustable parameters that occurs for all implemented tests are adjustable PID regulators
for acceleration pedal, deceleration pedal and for steering wheel. In the acceleration pedal case the PID
regulator is needed since many test procedures are initialized or preformed at constant predefined speed. While
using a model with a regular engine behaviour a standard parametrization can be used, but if necessary the
parametrization can be adapted based on the model.
The adjustable parameters which are specific for each test will be described in the remaining subsections in
sections 3.2.1.

Steady state cornering

In order to have a stable and simple control, steady state cornering with constant speed have been chosen.
Resulting in the need of only one regulator to control the acceleration pedal position in comparison with
constant radius turn where two regulators are needed, one for steering and one for acceleration pedal. By using
two regulator there is an increased risk for designing an unstable system since the desired steering angle is
dependent of the velocity, and the longitudinal velocity is affected by the steering angle. The manoeuvre with
constant steering angle will only be correct if the steering is completely linear, which makes it unsuitable.

7
 During steady state cornering in the virtual testing part of the evaluation tool it is possible for the user
to select several test speeds, e.g. to run the test at 80, 100 and 120 kph. Apart from the test velocity the
user can as well as adjust the steering input magnitude and duration in order to control the change in lateral
acceleration.

Sine with dwell

In order to be able to evaluate the stability of a vehicle as well as the ESC system a Sine with dwell manoeuvre
have been used. The chosen manoeuvre is based on FMVSS126 [Nat07] since this manoeuvre have produces
consistent results and shown to be an efficient way to trigger an over-steer behaviour according to NHTSA
[FEO05].
The Sine with dwell manoeuvre implemented in the virtual testing part of the evaluation tool allows the
user to change test velocity, the size of the steering wheel angle increments, aswell as change the turn frequency,
dwell length and maximum steering wheel angle.
In order to determine the steering wheel angle used in a Sine with dwell manoeuvre the results from steady
state cornering is used.

Acceleration
This test provides information about the behaviour of the vehicles or models power-train, both regarding
acceleration capabilities and gearbox behaviour. It were implemented as the second test manoeuvre due to
its simplicity to simulate. This have resulted in the possibility at an early stage test to run multiple test
manoeuvres in a simulation and thereby ensure that the chosen way of implementation of test procedures works.
The implementation of the acceleration test manoeuvre allows the user to either search for top speed or to
find a specific speed, e.g. if only 0-120 is interesting in a study there is no point in spending simulation time to
find the top speed.

Continuous sinusoidal input

In order to evaluate the frequency response of a model there is three different test methods: puls input, random
input and continuous sinusoidal. The reason for using continuous sinusoidal input instead of random or pulse
input is that it is possible to program in an efficient and representative way, as a difference to random input. It
consists as well of a sweep containing several different frequencies which is beneficial compared to puls inputs
single frequency.
Due to very similar analysis methods for random and continuous sinusoidal input, there is a possibility to
use the same analysis part of the evaluation tool on the logged data from a random input test.
The implementation of continuous sinusoidal input allows the user to adjust test velocity, low and high
frequency, step size, as well as the number of periods that should be preformed at each frequency.

4 Evaluation Tool
As stated in section 1.1 Objectives, an framework for simulation, referred to as Virtual test, and a data analysis
and report script, referred to as Analysis, should be produced within this thesis work. Figure 4.1 shows the
interaction between these two deliverables. Both the simulation part and the post analysis part should work
independently of each other. The results from the virtual testing preformed by the framework for simulation
should be possible to replace with results from real tests.
In order to build a structured tool a clear and well organized file structure have been used. This structure is
presented in figure E.1 in Appendix E.

4.1 Virtual test

This section will discuss the development process, design targets and the results of the virtual testing part of
the evaluation tool written within this thesis.
For clarity of this chapter the following definitions are needed.

8
 Genrated report

Real test Analysis Including plots on

standard format for
Post processing
and presentation
easy comparisons
script
Conducts data For comparison
Virtual test analysis and - Model to Requirements
generates pdf - Vehicle to Requirements
Framework for reports - Model to Vehicle
simulations - Model to Model
Used to test run - Vehicle to Vehicle
vehicle models and
save the results

Saved data

Figure 4.1: Verification of real vehicles, models with respect to requirements.

Vehicle model The model of a vehicle, without driver or logging tools.

With inputs such as Steering wheel angle, throttle position and brake.
Outpus such as Lateral acceleration, longitudinal acceleration and position.
Simulink inter- Within this thesis it is defined as the simulink blocks containing logging code and
face the virtual driver.
Simulink model This refers to the complete simulink file, including both the simulation interface
and the vehicle model.

4.1.1 Method & Development Process

As stated in section 1.2, Limitations, the platforms used in this thesis-work are Simulink and Matlab. To have
an efficient tool that allows further development the following key features have been regarded:

Usability An very important property is that the tool should be attractive to use in order to
verify a model or to quickly check the behaviour of a model. It is necessary that it
is easy to set up and operate, otherwise the potential users will not be motivated
to use this tool.

Adjust ability The tools needs to be compatible with different models . Also test parameters
needs to be adjustable.

Debugging of All information needed to run a test case should be possible to included in the
model simulink file. This provides the possibility to run the simulink model separately in
order to debug the model as well as verifying new test cases.

Allow To ensure that many tests can be preformed in a efficient way with for example
accelerated different parameter settings, it is important that the tool can be executed in
simulations simulinks accelerated mode. To ensure this all input settings needs to be pre
defined. However the possibility of using accelerated mode is depending on the
vehicle model as well.

Allow future In order to allow future development and addition of new test scripts it is necessary
development to have a standardised way of implementing the test scripts as well as clear program
structure in order to enable other persons then the original author of the evaluation
tool to understand and implement additional test scripts in to the tool.

9
 Data storage All data from the simulations needs to be automatically stored using an clear
structure and then saved. The data structure needs to provide a possibility for
post processing of the data at a later stage.

Allow batch In order to simulate several models or parametrizations the simulation tool needs
processes to allow batch processes. This results in that the entire evaluation tool should be
designed in order to allow this.

Allow FMU The virtual test tool should allow the user to import FMU models based on
models international standardized format FMI. These models can be used in several
different programs such as Dymola, Matlab/Simulink and CarMaker [Fmi].

Only require The tool for virtual testing should not require special licenses. Matlab/Simulink as
standard uni. well as MS Office are assumed to be available. LaTeX with standard user-packages
licenses are also assumed to be available.

In order to ensure a clear and efficient structure of the scripts activity based flowcharts have been used. The
flowcharts have been based on the Activity Diagrams within the OMG Unified Modeling Language Specification
v 1.5 [Uml].

4.1.2 Results regarding virtual test

In this section the final virtual test scripts designed and results are presented.

Tool layout
The over all layout of the simulation framework have been based on the key features mentioned in section 4.1.1.
In order to reduce the complexity for the user a specific set-up and initialisation script is being used as shown
in figure 4.2.
This provides the user with an efficient method of changing vehicle model, use batch scripts to change model
settings and select test manoeuvres to preform. It has however the disadvantage of a slightly more complicated
script structure.
The run part shown in figure 4.2 controls the selection of model, test and settings to be used and loads the
model and changes it settings in order to run the selected test. All tests using a specific model with a specific
parametrization are executed before the current model is closed and the next model is loaded. Using this
solution the loading time of models were reduced and as well as its contribution to the over all simulation time.
In order to ensure that the Simulink interface can run without recompiling the Simulink models for each
test the input settings are specified as a predefined vector that is assigned to a constant block in the Simulink
interface. This as well results in that the Simulink can run as a stand alone model during development of new
test manoeuvres as well as during debugging of vehicle models.
By specify the test manoeuvres as functions outside Simulink it is possible to add or change a manoeuvre
without having to do any adjustments to the Simulink files. This ensures that the same test manoeuvre code is
used for all models when simulating several models at once and reduces the risk of issues with different code
versions. The alternative to using external Matlab functions would be to build the functions with Simulink
blocks, however it would most likely result in that not only the current test case, but all test cases, would
calculate there driver input information and only the information from the current test would be used. The
disadvantageous with the current design with external Matlab functions is that the Simulink model can not
run unless the folders containing the Matlab functions are added to Matlabs file paths.
An important design change during the development were to export the test settings each run instead of
only once per test. The reason for it were to enable the possibility to check previous data and if necessary abort
a test, e.g. Sine with dwell, when it has failed to pass a test. Resulting in the possibility to check each run if
the model passes a criteria before it start the next run. The possibility to access previous runs is necessary for
several tests manoeuvres since many of them uses a steering angle based on a specific lateral acceleration at a
certain velocity.
After all tests are preformed for a model, all relevant data is saved. The reason for saving after each model
is to ensure that a batch processes can run several models without occupying to much of the RAM. The results

10
 Simulation framework
Simulations settings,
List of vehicles,
Choice of test procedures and
Initiate post process script

Import vehicle model

Imports the vehicle model and

initiates parametrization if
needed.
Simulink
Simulink model
Export test settings
Test Procedure and Run loop

containing vehicle
All test specific settings is model.
This model contains
Vehicle model loop

exported to the simulink

model. matlab scripts
Main script

controlling the driver

outputs as well as
controls when to
Run simulink simulation
terminate the
simulation.
The results from
Recives and store
Simulink are exported
Recives and stores simulation to matlab as timeseries.
results from simulink. The
results is saved in a structure.

Saving results

The results from the all test

runs with the model saved as a
.mat file in the results folder.

Figure 4.2: Figure showing the overall layout of the simulation framework tool as well as its simulation loops
and interaction with the simulink model.

for each model is saved as an .mat file with a array of structures containing all saved data from a test run, as
well as information about the test manoeuvre.

Adjust ability
In order to have a good adjust-ability an excel document containing the adjustable parameters for all test have
been used. The benefit of using an excel file is that it is providing an good overview of the settings and its
contents can easily be changed. When running the tool, the valus from the adjustable parameters are imported
from the excel file to matlab.
Even if excel requires a specific licence it is considered to be standard at most companies as well as
universities. If no license is accessible it is as well possible to access the document through open source
programs.

Model interface
The model interface have been designed based on three blocks as shown in figure 4.3. This is done in order
to ensure that it is easy to change model and make necessary set-ups. The input block creates and sends
the driver inputs to the vehicle model block. The vehicle model block sends the behaviour information on to
the logging block, which after the simulation exports it to Matlab. The logging block sends as-well necessary
feedback signals back to the input block.
In order to change vehicle model, it is only needed to replace the vehicle model block with new block

11
 Figure 4.3: Figure showing the simulink model contaning both simulink interface and vehicle model.

containing the same in and out-puts. This makes it easier for the user to change and adapt a model without
affecting the way the test is preformed. The disadvantages with this type of solution is that it is slightly more
complicated to add more output signals and that there is no inputs from the surrounding environment, as
stated in Limitations (section 1.2).
If a variables is not specified in the model, e.g. Gear, it should be set to N aN in order to ensure that both
simpler and more advanced models can be used.

Accelerated simulation
In order to avoid algebraic loops so that accelerated and rapid acceleration mode can be used, a memory box is
mounted on the main feedback loop. This results in a possibility to avoid errors due to algebraic loops when
using accelerated simulations, but with the disadvantage of using information from last simulation step when
calculating the current input signals. This disadvantage is however being considers to be very small due to
small simulation steps.

Allowing future development

In order to ensure the possible of continuing to add test procedures to this evaluation tool, the scripts have
been clearly divided for each test. The procedure needed is to create the script controlling the behaviour of
the vehicle, add it to the main input function, add this to the script file, containing a list of all tests, and
add the specific settings to the excel file. All input and output signals to the test scripts have the same size
independently of what test is run in order to reduce the need of recompiling the simulink model. It will still
require that the programmer have a large understanding of matlab and simulink as well as take the time needed
to understand the program, however thanks to this simplification, the need of understanding all evaluation tool
scripts should be significantly reduced.

12
4.2 Analysis
In this section of the report the design of the auto-generated reports as well as the design of the post processing
script will be regarded. There is a large value in establish a standardized format for review of the test results,
which is why the auto-generated report is a very important feature of the tool.

4.2.1 Method & Development Process

In order to ensure that the results are presented as standardized as possible the existing ISO standards that
were discussed in Chapter3 were used when writing the post processor.
The key features that have been regarded when writing the post processing and presentation part of the
evaluation tool is the following:

Option of using In order to allow the user to view and zoom in on a graph if needed, this option is
plot windows important. However if this option is not activated by the user no plots should be
displayed in order to improve the calculation speed

Export to pdf The final results needs to be exported to a pdf document so it easily can be printed
or distributed.

Export graphs All graphs needs to be exported in format suitable for use in a presentation.

Export In order to easily debug a model or to further investigate its behaviour, evaluation
evaluation plots plots containing all logged variables should be presented in a recognisable way for
easy evaluation.

Auto-generating When running virtual tests it should be possible to automatically post process the
of reports results immediately after running the simulations without further need of setup.

Adjustable The tool should allow the user to make adjustments to whether to plot evaluation
settings plots or to save all graphs.

4.2.2 Results regarding analysis

Design of auto generated reports
The auto generation of reports have been written so that it creates a separate report for each vehicle. The
disadvantage with this solution is that is is harder to compare two specific vehicle since the user needs to take
the two reports, find the graph to compare, and do it manually. However this results in the possibility to
compare reports that have been generated at different instances as well as comparing as many different reports
as the user would like.
The reports are generated in the same way independently if the information used for the report is from a
simulation of a model or from real vehicle testing.
In order to ensure that the test conditions for a specific test can be known when evaluating the result, the
path or reference number of the file where the test conditions have been recorded can be displayed on the front
page of the report. If now reference is defined, a line will be displayed allowing the user to manually add the
reference.
The tests manoeuvres that are based on ISO tests will be presented based on the specifications in the ISO
standard. The reason for this is as mentions in section 4.2.1 to ensure that the results are presented in a
standardized way, in order for the user to read and understand the results.
In those test manoeuvres that are not specified as a standard, the choice of graphs and information that is
presented are based on available information regarding the manoeuvre as well as the author experience from
courses as well as meetings within this thesis work.
In order to provide the user with as much important information as possible, two pages showing the logging
signals have been added. Giving the user more extensive information about the vehicles behaviour

13
 Analysis script

b Specifies post process settings as

well as specifing files to analyse

Import vehicle data

Imports the vehicle data and

evaluates contence.

Run test specific post

Test Procedure post process

processes

Vehicle data loop Test specfic post processor is

initilised and creates nessasary
Main script

graphs as well as calculates key

values.

Store results
Saves all graphs as well as
stores key values and graph
names in order to save them to
a .tex file at a later stage.

Export to .tex and compile

The results are exported to a

.tex file that is compiled using
pdflatex in to a pdf document.

Figure 4.4: Figure showing the overall layout of the Analysis part of the Evaluation tool.

Design of post processing tool.

As stated in section 1.2, Matlab have been used as a standard platform, however in order to export the results
in an efficient way to a pdf document other licence free software as Latex were early considered due to there
flexibility. In figure 4.4 the script structure is described.
The chosen design of the post processor are a matlab script that calculates all needed information, creates
plots which it saves, and then exports the information to a .tex file that is compiled using PdfLatex. This
results in an easily printed or distributed pdf document for each tested vehicle containing information from all
preformed tests. Example of reports can be found in Appendix A and B.
When post processing results from simulations the search path to the .mat. file is needed. This file contains
information about what type of runs that is stored in the file, if this information is missing a pop up window
will occur and ask the user for needed information. This is to ensure that the stored data are post processed in
the right way, giving the user the possibility to directly post process the results if using the simulation tool.
In order to be able to import results from tests with a real vehicle a script converting the test data to the
data structure used in this thesis is needed. Within this thesis a script for importing logged test data from the
Saab 9.3 possessed by the vehicle dynamics department have been written in order to ensure this possibility.

When post processing the results the tool evaluates one vehicle at a time and then steps thou the different
test manoeuvres specific post process scripts.
As described in section 4.2.1 one important feature were to have the ability to access graphs if requested.
This have been solved by creating a possibility of choosing whether to save the graphs after the pdf document
is created or if all .png files that contains the graphs should be removed.

14
5 Simulations with BAPS model
This chapter describe the setup of the BAPS model as well as presenting results from simulations using the
Evaluation tool.

5.1 Set up options

Before being able to run the BAPS model it is nessasary to conecting the FMU to the simulink interface
as shown in figure 5.1. It is as well several parameters that needs to be altered depending of what vehicle
specifications to use during the simulation.

Figure 5.1: Figure showing the BAPS model (FMU) connected to the simulink interface. This figure represents
the Vehicle block in figure 4.3

15
5.1.1 Parametrized vehicles
In order to test run the BAPS model set of vehicles have been provided by the BAPS project. The parameters
provided of these are not complete parametrisations of vehicles and are not completely representing the original
vehicles, however they are sufficient in order to verify the evaluation tool and the possibility to run a batch
script. The different parametrisations used within this thesis are presented in table C.1 in Appendix C

5.1.2 Adjustable safety systems

This model have some Active Safety functions. Of these, ESC, ABS and TCS can be tested in the tool in
present thesis. But another of them is AEB (Automatic Emergency Brake) is an ADAS (Advanced Driver
Assistance System) function and can consequently not be tested since the tool does not support information
about the vehicles environment. Table 5.1 shows what systems that can be used.

Table 5.1: Active Safety functions in BAPS model and whether they can be used within the evaluation tool.

Safety system: Supported: Comments:

ESC Yes
ABS Yes Not been tested within this thesis
TCS Yes Not been tested within this thesis
AEB No No information regarding surrounding environment available.

5.2 Results from the simulations with the BAPS model

The simulations using the BAPS model showed that it is possible to run batch simulations in an efficient way,
as well as the compatibility with FMU models.
However the simulations were preformed without ESC, which resulted in an misleading result regarding the
FMVSS 126 legislation. It should as well be noted that the model is not designed in order to evaluate frequency
responses, which results in unreliable results.
The results from these simulations are presented in Table D.1.

6 Conclusions
This thesis have developed an evaluation tool enables the users to easily verify and check a model as well as
easily post process logged data.
This possibility is beneficial both for research and educational purpose.
Within research it gives an efficient method to easily verify multiple vehicle models, as have been done with
the BAPS model. This can show if the vehicle model is a good representation of a real vehicle in order to
increase the reliability of the research results.
In education this thesis work can as well provide a useful tool for rapid simulations in order to compare
variables e.g. different trackwidths.
As an example, a beta version of the tool were successfully used in two other MSc theses, [San14] and
[Kar14].

6.1 Future work

It would be beneficial to develop several more test scripts in order to be able to get a better overview of a
virtual or real vehicle. A recommendation of test procedures for future implementation are presented in table
6.1. The reason why this tests are recommended are in order to complete the overview of the vehicle behaviour
and thereby increase the usefulness of the evaluation tool.

16
 Table 6.1: Proposed test procedures for future implementation.

Test procedure:
Step input
Sinusoidal input, one period.
Pulse input
Braking with split coefficient of friction
Brake in turn
Straight braking
Random frequency response

This type of tool could as well be beneficial to present as an open source solution and make it available
online. It could result in designers that uses the tool supplies scripts for different test procedures, as well as
scripts for different types of result analysis.
Similar tools could as well be beneficial for evaluation of trucks, truck combinations, electric bikes and
minicars, e.g. urban personal vehicles.

17
References
[Bapa] Balancing Active and Passive Safety. Autoliv, Volvo Car Corporation, Saab, VTI, Chalmers,
Semcon. url: http://www.chalmers.se/safer/EN/projects/pre-crash-safety/associated-
projects/balancing-active-passive.
[Bapb] Balancing Active and Passive Safety, Vinnova. Autoliv, Volvo Car Corporation, VTI, Chalmers,
Semcon. url: http://www.vinnova.se/sv/Resultat/Projekt/Effekta/Balancing- Active-
and-Passive-Safety/.
[FEO05] G. J. Forkenbrock, D. Elsasser, and B. O’Harra. NHTSA’s light vehicle handling and ESC effectiveness
research program. Paper Number 05-0221. National Highway Traffic Safety Administration, United
States, 2005.
[Fmi] Functional Mock-up Interface (FMI). Modelica Association Project. url: https : / / www . fmi -
standard.org/tools (visited on 05/12/2014).
[Kar14] A. Karsolia. Desktop driving simulator with modular vehicle model and scenario specification. Master
thesis 2014:06 Department of Applied Mechanics, Chalmers tekniska högskola. Department of
Applied Mechanics, Chalmers tekniska högskola, 2014. isbn: 1652-8557.
[MAT13] MATLAB. version 8.2.0.701 (R2013b). Natick, Massachusetts: The MathWorks Inc., 2013.
[MiK] MiKTex. version 2.9.4813 Basic. url: miktex.org (visited on 05/12/2014).
[Pac02] H. Pacejka. Tire and Vehicle Dynamics. R: Society of Automotive Engineers. 2002. isbn: 9780768011265.
[San14] M. Santoro. Development of a parameterized passenger vehicle model for longitudinal dynamics for
a desktop driving simulator. Ex - Institutionen för signaler och system, Chalmers tekniska högskola,
no: EX008/2014. Institutionen för signaler och system, Chalmers tekniska högskola, 2014.
[Sim13] Simulink. version 8.2 (R2013b). Natick, Massachusetts: The MathWorks Inc., 2013.
[Uml] Unified Modeling Language, v1.5. Object Management Group, Inc. 2003. url: http://doc.omg.
org/formal/2003-03-01.pdf (visited on 02/03/2014).
[Doc] Docent Mathias Lidberg. TME102 - Vehicle dynamics, advanced. url: https://student.portal.
chalmers.se/en/chalmersstudies/courseinformation/Pages/SearchCourse.aspx?course_
id=20630&parsergrp=3 (visited on 05/12/2014).
[Dym13] Dymola - Dynamic Modeling Laboratory. version 2014 FD01 (64-bit). Lund, Sweden: Dassault
Systèmes AB, 2013.
[FMI13] FMI Toolbox for Matlab. version 1.8. Lund, Sweden: Dassault Systèmes AB, 2013.
[ISOa] ISO 14512:1999. Passenger cars – Straight-ahead braking on surfaces with split coefficient of friction
– Open-loop test procedure. International Organization for Standardization, Geneva, Switzerland.
[ISOb] ISO 21994:2007. Passenger cars – Stopping distance at straight-line braking with ABS – Open-loop
test method. International Organization for Standardization, Geneva, Switzerland.
[ISOc] ISO 3888-1:1999. Passenger cars – Test track for a severe lane-change manoeuvre – Part 1: Double
lane-change. International Organization for Standardization, Geneva, Switzerland.
[ISOd] ISO 3888-2:2011. Passenger cars - Test track for a severe lane-change manoeuvre - Part 2: Obstacle
avoidance. International Organization for Standardization, Geneva, Switzerland.
[ISOe] ISO 4138:2012. Passenger cars - Steady-state circular driving behaviour - Open-loop test methods.
International Organization for Standardization, Geneva, Switzerland.
[ISOf] ISO 7401:2011. Road vehicles – Lateral transient response test methods – Open-loop test methods.
International Organization for Standardization, Geneva, Switzerland.
[ISOg] ISO 7975:2006. Passenger cars – Braking in a turn – Open-loop test method. International Organiza-
tion for Standardization, Geneva, Switzerland.
[ISOh] ISO 9816:2006. Passenger cars – Power-off reaction of a vehicle in a turn – Open-loop test method.
International Organization for Standardization, Geneva, Switzerland.
[Nat07] National Highway Traffic Safety Administration. Standard No. 126; Electronic stability control
systems. 2007. url: http://www.fmcsa.dot.gov/rules-regulations/administration/fmcsr/
fmcsrruletext.aspx?reg=571.126.

18
A Example report BAPS model
 Test object: Volvo XC90 2011
Test conditions:
Reference number: BAPS 30-Nov-2014
Testdate: 2014-11-30 12:57:07

Contents
1 Steady State Cornering, SS-ISO 4138:2012 2
1.1 Longitudinal Velocity during test . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.2 Steering angle as function on lateral acceleration . . . . . . . . . . . . . . . . . . 4
1.3 Sideslip angle as function on lateral acceleration . . . . . . . . . . . . . . . . . . 6
1.4 (αr − αf ) in relation to ay /g . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

2 Straight Ahead Acceleration 8

2.1 Longitudinal Velocity during test . . . . . . . . . . . . . . . . . . . . . . . . . . 8
2.2 Longitudinal Acceleration during test . . . . . . . . . . . . . . . . . . . . . . . . 9
2.3 Acceleration at different velocities . . . . . . . . . . . . . . . . . . . . . . . . . . 10

3 Sine With Dwell, FMVSS 126 11

4 Frequency response, SS-ISO 7401:2011 16

1 Steady State Cornering, SS-ISO 4138:2012
This test is preformed using steady state cornering according to SS-ISO 4138:2012. The used
method within this case is the Constant speed method. The centripetal acceleration is obtained
using the product of yaw velocity and horizontal velocity, as is described in section 9.2 b in
SS-ISO 4138:2012

Table 1: List of runs and active tests

Run Test type Reason for ending

1 Steady State Cornering Break due to high change in lateral acc.
2 Steady State Cornering Break due to high change in lateral acc.

Figure 1: Displays the path for the different runs. All mesurements are in meters.

2
1.1 Longitudinal Velocity during test
This plot is not defined in ISO 4138:2012 Steady state cornering, however it is presented in
order to display the variations in speed that occurs during this test.

Figure 2: Longitudinal velocities during Steady State Cornering test. The maximum velocities
were: Run 1: 80 [kph]. Run 2: 100 [kph]. And the final velocity were: Run 1: 80 [kph]. Run 2:
100 [kph].

3
1.2 Steering angle as function on lateral acceleration
Presented according to Steady State Cornering SS-ISO 4138:2012.

Figure 3: Steering angle as function of lateral acceleration in g. The graph shows as well the
linearisation of the different runs. The displayed velocities in the legend is the maximum speed
during the test.

4
1.2.1 Steering angle gradient as function on lateral acceleration
The gradient is calculated based on the change during one sample step divided with acceleration
change during this sample and the corresponding acceleration vector is calculated as mean value
during that the sample step.

Figure 4: Steering angle gradient as function of lateral acceleration.

When using the reference speed of 80 [kph] the linearisation is represented by 51.59 ∗ (ay /g).
When using the reference speed of 100 [kph] the linearisation is represented by 33.33 ∗ (ay /g).

5
1.3 Sideslip angle as function on lateral acceleration
Presented according to Steady State Cornering SS-ISO 4138:2012.

Figure 5: Side slip angle as function of lateral acceleration.

6
1.4 (αr − αf ) in relation to ay /g

Figure 6: Showing the relation between slip angles and lateral acceleration as well as the
linearisation.

When using the reference speed of 80 [kph] the linearisation is represented by 0.04 ∗ (ay /g).
When using the reference speed of 100 [kph] the linearisation is represented by 0.04 ∗ (ay /g).

7
2 Straight Ahead Acceleration

Table 2: List of runs and active tests

Run Test type Reason for ending

3 Straight Ahead Acceleration Break due to too low longitudinal acc.

2.1 Longitudinal Velocity during test

Figure 7: Longitudinal velocities acceleration test. The maximum velocity during the test were:
Run 3: 247 [kph].

Table 3: Displayes time between two velocities during ongoing full acceleration.

Velocity change Time, Run: 3

0-60 [kph] 6.79 [s]
0-100 [kph] 10.48 [s]
80-120 [kph] 4.56 [s]

8
 Table 4: Time for traveling the 402 meters.

Distance Time, Run: 3

402 [m] 18.09 [s]

2.2 Longitudinal Acceleration during test

Figure 8: Longitudinal acceleration during test.

9
2.3 Acceleration at different velocities

Figure 9: Longitudinal acceleration as a function of longitudinal velocity.

10
3 Sine With Dwell, FMVSS 126
This results are based on Laboratory Test Procedure for FMVSS 126, Electronic Stability
Control System (TP-126-03, 9th Sept. 2011). The methods to calculate key-values ate displayed
in section 13.10 in TP-126-03.

Table 5: List of runs and active tests

Run Test type Reason for ending

4 Sine with dwell Break due to manoeuvre done.
5 Sine with dwell Break due to manoeuvre done.
6 Sine with dwell Break due to manoeuvre done.
7 Sine with dwell Break due to manoeuvre done.

Table 6: Displays results from all sine withdwell test manoeuvre in terms of: Run number, Peak
steering wheel angle, Peak yaw velocity, Yaw velocity criteria 1.00 seconds after completion
of steering, Yaw velocity criteria 1.75 seconds after completion of steering, Lateral movement
criteria.

Run SWA YawVel COS+1.00[s] COS+1.75[s] Lat.movement

4 23 -0.20 Pass Pass Fail
5 31 -0.26 Pass Pass Fail
6 39 -0.34 Pass Pass Pass
7 47 -0.55 Fail Fail Pass

11
Figure 10: Steering wheel angle as well as Yaw velocity as a function of time after Beginning
of steer (BOS). COS+Time shows time after Completion of steer (COS) which is used as test
criteria.

12
Figure 11: Vehicle movement during manoeuvre in Run 6.

13
Figure 12: Steering wheel angle as well as Yaw velocity as a function of time after Beginning
of steer (BOS). COS+Time shows time after Completion of steer (COS) which is used as test
criteria.

14
Figure 13: Vehicle movement during manoeuvre in Run 7.

15
4 Frequency response, SS-ISO 7401:2011
This results are based on Frequency response from SS-ISO 7401:2011 It is valid according to
SS-ISO 7401:2011 for evaluating Continuous sinusoidal input or Random input.

Table 7: List of runs and active tests

Run Test type Reason for ending

8 Continuous sinusoidal input Break due to end of simulation time.

Figure 14: This graph presents steering wheel angle, phase angle and coherence as a function of
frequecy.

16
Figure 15: Steering wheel angle as well as Yaw velocity as a function of time after Beginning
of steer (BOS). COS+Time shows time after Completion of steer (COS) which is used as test
criteria.

17
B Example report SAAB Sim whith ESC and EVAL
 Test object: Model Saab 9.3
Test conditions:
Reference number: model 4wheel engine ESC 01-Dec-2014
Testdate: 2014-12-01 21:20:22

Contents
1 Steady State Cornering, SS-ISO 4138:2012 2
1.1 Longitudinal Velocity during test . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.2 Steering angle as function on lateral acceleration . . . . . . . . . . . . . . . . . . 4
1.3 Sideslip angle as function on lateral acceleration . . . . . . . . . . . . . . . . . . 6
1.4 (αr − αf ) in relation to ay /g . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
1.5 Test Evaluation plots: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8

2 Straight Ahead Acceleration 12

2.1 Longitudinal Velocity during test . . . . . . . . . . . . . . . . . . . . . . . . . . 12
2.2 Longitudinal Acceleration during test . . . . . . . . . . . . . . . . . . . . . . . . 13
2.3 Acceleration at different velocities . . . . . . . . . . . . . . . . . . . . . . . . . . 14
2.4 Test Evaluation plots: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15

3 Sine With Dwell, FMVSS 126 17

3.1 Test Evaluation plots: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29

4 Frequency response, SS-ISO 7401:2011 39

4.1 Test Evaluation plots: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
1 Steady State Cornering, SS-ISO 4138:2012
This test is preformed using steady state cornering according to SS-ISO 4138:2012. The used
method within this case is the Constant speed method. The centripetal acceleration is obtained
using the product of yaw velocity and horizontal velocity, as is described in section 9.2 b in
SS-ISO 4138:2012

Table 1: List of runs and active tests

Run Test type Reason for ending

1 Steady State Cornering Break due to end of simulation time.
2 Steady State Cornering Break due to end of simulation time.

Figure 1: Displays the path for the different runs. All mesurements are in meters.

2
1.1 Longitudinal Velocity during test
This plot is not defined in ISO 4138:2012 Steady state cornering, however it is presented in
order to display the variations in speed that occurs during this test.

Figure 2: Longitudinal velocities during Steady State Cornering test. The maximum velocities
were: Run 1: 80 [kph]. Run 2: 100 [kph]. And the final velocity were: Run 1: 80 [kph]. Run 2:
100 [kph].

3
1.2 Steering angle as function on lateral acceleration
Presented according to Steady State Cornering SS-ISO 4138:2012.

Figure 3: Steering angle as function of lateral acceleration in g. The graph shows as well the
linearisation of the different runs. The displayed velocities in the legend is the maximum speed
during the test.

4
1.2.1 Steering angle gradient as function on lateral acceleration
The gradient is calculated based on the change during one sample step divided with acceleration
change during this sample and the corresponding acceleration vector is calculated as mean value
during that the sample step.

Figure 4: Steering angle gradient as function of lateral acceleration.

When using the reference speed of 80 [kph] the linearisation is represented by 57.58 ∗ (ay /g).
When using the reference speed of 100 [kph] the linearisation is represented by 40.41 ∗ (ay /g).

5
1.3 Sideslip angle as function on lateral acceleration
Presented according to Steady State Cornering SS-ISO 4138:2012.

Figure 5: Side slip angle as function of lateral acceleration.

6
1.4 (αr − αf ) in relation to ay /g

Figure 6: Showing the relation between slip angles and lateral acceleration as well as the
linearisation.

When using the reference speed of 80 [kph] the linearisation is represented by 0.55 ∗ (ay /g).
When using the reference speed of 100 [kph] the linearisation is represented by 0.57 ∗ (ay /g).

7
1.5 Test Evaluation plots:
Show summary of key values in order to verify the results.

1.5.1 Evaluation plot from run:1

8
1.5.2 Evaluation plot from run:1

9
1.5.3 Evaluation plot from run:2

10
1.5.4 Evaluation plot from run:2

11
2 Straight Ahead Acceleration

Table 2: List of runs and active tests

Run Test type Reason for ending

3 Straight Ahead Acceleration Break due to too low longitudinal acc.

2.1 Longitudinal Velocity during test

Figure 7: Longitudinal velocities acceleration test. The maximum velocity during the test were:
Run 3: 280 [kph].

Table 3: Displayes time between two velocities during ongoing full acceleration.

Velocity change Time, Run: 3

0-60 [kph] 5.39 [s]
0-100 [kph] 11.04 [s]
80-120 [kph] 7.05 [s]

12
 Table 4: Time for traveling the 402 meters.

Distance Time, Run: 3

402 [m] 17.88 [s]

2.2 Longitudinal Acceleration during test

Figure 8: Longitudinal acceleration during test.

13
2.3 Acceleration at different velocities

Figure 9: Longitudinal acceleration as a function of longitudinal velocity.

14
2.4 Test Evaluation plots:
Show summary of key values in order to verify the results.

2.4.1 Evaluation plot from run:3

15
2.4.2 Evaluation plot from run:3

16
3 Sine With Dwell, FMVSS 126
This results are based on Laboratory Test Procedure for FMVSS 126, Electronic Stability
Control System (TP-126-03, 9th Sept. 2011). The methods to calculate key-values ate displayed
in section 13.10 in TP-126-03.

Table 5: List of runs and active tests

Run Test type Reason for ending

4 Sine with dwell Break due to manoeuvre done.
5 Sine with dwell Break due to manoeuvre done.
6 Sine with dwell Break due to manoeuvre done.
7 Sine with dwell Break due to manoeuvre done.
8 Sine with dwell Break due to manoeuvre done.
9 Sine with dwell Break due to manoeuvre done.
10 Sine with dwell Break due to manoeuvre done.
11 Sine with dwell Break due to manoeuvre done.
12 Sine with dwell Break due to manoeuvre done.
13 Sine with dwell Break due to manoeuvre done.
14 Sine with dwell Break due to manoeuvre done.
15 Sine with dwell Break due to manoeuvre done.
16 Sine with dwell Break due to manoeuvre done.
17 Sine with dwell Break due to manoeuvre done.
18 Sine with dwell Break due to manoeuvre done.
19 Sine with dwell Break due to manoeuvre done.
20 Sine with dwell Break due to manoeuvre done.
21 Sine with dwell Break due to manoeuvre done.
22 Sine with dwell Break due to manoeuvre done.
23 Sine with dwell Break due to manoeuvre done.
24 Sine with dwell Break due to manoeuvre done.
25 Sine with dwell Break due to manoeuvre done.
26 Sine with dwell Break due to manoeuvre done.
27 Sine with dwell Break due to manoeuvre done.
28 Sine with dwell Break due to manoeuvre done.
29 Sine with dwell Break due to manoeuvre done.
30 Sine with dwell Break due to manoeuvre done.
31 Sine with dwell Break due to manoeuvre done.
32 Sine with dwell Break due to manoeuvre done.
33 Sine with dwell Break due to manoeuvre done.

17
Table 6: Displays results from all sine withdwell test manoeuvre in terms of: Run number, Peak
steering wheel angle, Peak yaw velocity, Yaw velocity criteria 1.00 seconds after completion
of steering, Yaw velocity criteria 1.75 seconds after completion of steering, Lateral movement
criteria.

Run SWA YawVel COS+1.00[s] COS+1.75[s] Lat.movement

4 26 -0.18 Pass Pass Fail
5 34 -0.24 Pass Pass Fail
6 43 -0.29 Pass Pass Pass
7 52 -0.34 Pass Pass Pass
8 60 -0.38 Pass Pass Pass
9 69 -0.42 Pass Pass Pass
10 77 -0.46 Pass Pass Pass
11 86 -0.50 Pass Pass Pass
12 95 -0.54 Pass Pass Pass
13 103 -0.58 Pass Pass Pass
14 112 -0.60 Pass Pass Pass
15 121 -0.62 Pass Pass Pass
16 129 -0.64 Pass Pass Pass
17 138 -0.64 Pass Pass Pass
18 146 -0.64 Pass Pass Pass
19 155 -0.63 Pass Pass Pass
20 164 -0.61 Pass Pass Pass
21 172 -0.60 Pass Pass Pass
22 181 -0.58 Pass Pass Pass
23 189 -0.56 Pass Pass Pass
24 198 -0.54 Pass Pass Pass
25 207 -0.52 Pass Pass Pass
26 215 -0.49 Pass Pass Pass
27 224 -0.47 Pass Pass Pass
28 232 -0.46 Pass Pass Pass
29 241 -0.45 Pass Pass Pass
30 250 -0.44 Pass Pass Pass
31 258 -0.43 Pass Pass Pass
32 267 -0.42 Pass Pass Pass
33 275 -0.42 Pass Pass Pass

18
Figure 10: Steering wheel angle as well as Yaw velocity as a function of time after Beginning
of steer (BOS). COS+Time shows time after Completion of steer (COS) which is used as test
criteria.

19
Figure 11: Vehicle movement during manoeuvre in Run 6.

20
Figure 12: Steering wheel angle as well as Yaw velocity as a function of time after Beginning
of steer (BOS). COS+Time shows time after Completion of steer (COS) which is used as test
criteria.

21
Figure 13: Vehicle movement during manoeuvre in Run 13.

22
Figure 14: Steering wheel angle as well as Yaw velocity as a function of time after Beginning
of steer (BOS). COS+Time shows time after Completion of steer (COS) which is used as test
criteria.

23
Figure 15: Vehicle movement during manoeuvre in Run 20.

24
Figure 16: Steering wheel angle as well as Yaw velocity as a function of time after Beginning
of steer (BOS). COS+Time shows time after Completion of steer (COS) which is used as test
criteria.

25
Figure 17: Vehicle movement during manoeuvre in Run 26.

26
Figure 18: Steering wheel angle as well as Yaw velocity as a function of time after Beginning
of steer (BOS). COS+Time shows time after Completion of steer (COS) which is used as test
criteria.

27
Figure 19: Vehicle movement during manoeuvre in Run 33.

28
3.1 Test Evaluation plots:
Show summary of key values in order to verify the results.

3.1.1 Evaluation plot from run:6

29
3.1.2 Evaluation plot from run:6

30
3.1.3 Evaluation plot from run:13

31
3.1.4 Evaluation plot from run:13

32
3.1.5 Evaluation plot from run:20

33
3.1.6 Evaluation plot from run:20

34
3.1.7 Evaluation plot from run:26

35
3.1.8 Evaluation plot from run:26

36
3.1.9 Evaluation plot from run:33

37
3.1.10 Evaluation plot from run:33

38
4 Frequency response, SS-ISO 7401:2011
This results are based on Frequency response from SS-ISO 7401:2011 It is valid according to
SS-ISO 7401:2011 for evaluating Continuous sinusoidal input or Random input.

Table 7: List of runs and active tests

Run Test type Reason for ending

34 Continuous sinusoidal input Break due to end of manouvre.

Figure 20: This graph presents steering wheel angle, phase angle and coherence as a function of
frequecy.

39
Figure 21: Steering wheel angle as well as Yaw velocity as a function of time after Beginning
of steer (BOS). COS+Time shows time after Completion of steer (COS) which is used as test
criteria.

40
4.1 Test Evaluation plots:
Show summary of key values in order to verify the results.

4.1.1 Evaluation plot from run:34

41
4.1.2 Evaluation plot from run:34

42
 ]

]
·m 2

·m 2

·m 2
]
]

[m
[m

[kg

[kg

[kg
]
le

]
le

t[ m

[m
Ax
Ax

xis

is
i
ar

-ax
on

-ax
Re

x-a
Fr

Re
m]

Fr

ty

tz
oG

oG

t
h

ou

ou

ou
oG
Table C.1: Parametrizations used together with BAPS model.

idt

idt
]

eC

eC
[kg

ab

ab

ab
tC

kW

kW
nc

nc
ht

tia

tia

tia
igh
sta

sta
eig

ac

ac

er

er

er
He
Di

Di
W

Tr

Tr

In

In

In
Make, Model and Model Year
BMW 535i 2011 1646 1.508 1.338 0.603 1.600 1.600 982 3273 3273
Volkswagen Passat 2 0TDI 2011 1687 1.443 1.386 0.626 1.552 1.552 990 3301 3301
Parametrization of BAPS model

Mercedes Vivaro 2011 1830 1.536 1.475 0.781 1.615 1.615 1218 4059 4059
Fod Kuga TDCI 2011 1466 1.426 1.214 0.692 1.574 1.574 753 2510 2510
Skoda Yeti Tdi 4x4 2011 1552 1.368 1.165 0.685 1.541 1.541 728 2425 2425
Volvo XC90 2011 1863 1.396 1.396 0.720 1.634 1.634 1027 3422 3422
Skoda Fabia Kombi TSI 2011 1060 1.297 1.247 0.613 1.417 1.417 503 1678 1678
Skoda Octavia 2011 1354 1.354 1.301 0.595 1.541 1.541 699 2331 2331
Suzuki Swift 2011 1023 1.096 1.141 0.616 1.490 1.490 373 1244 1244
Smart for 2 2011 765 0.830 0.736 0.637 1.283 1.283 137 458 458
BMW 1er 2011 1242 1.380 1.176 0.612 1.471 1.471 597 1991 1991
BMW X3 2011 1613 1.350 1.350 0.673 1.581 1.581 866 2887 2887
Citr en C3 2011 968 1.122 1.168 0.621 1.470 1.470 374 1246 1246
Opel Astra Sports Tourer 2011 1387 1.392 1.338 0.625 1.544 1.544 757 2523 2523
Volkswagen Sharan 2011 1547 1.438 1.382 0.696 1.569 1.569 905 3016 3016
BMW 530 D Gran Turismo 2011 1998 1.452 1.452 0.634 1.611 1.611 1237 4122 4122
Seat Ibiza 1 2 Ecomotive 2011 1249 1.147 1.194 0.584 1.446 1.446 501 1670 1670
Mercedes CLS 350 CDI 2011 1730 1.568 1.335 0.582 1.581 1.581 1071 3569 3569
Volkswagen Passat Variant 2011 1452 1.414 1.358 0.628 1.537 1.537 818 2726 2726
Opel Astra 2011 1471 1.392 1.338 0.625 1.533 1.533 806 2687 2687
Volkswagen Polo TDI 2011 1045 1.130 1.177 0.595 1.463 1.463 409 1363 1363
Volkswagen Touareg Hybrid 2011 2101 1.422 1.421 0.700 1.624 1.624 1251 4170 4170
Honda Jazz 2011 1016 1.110 1.156 0.622 1.495 1.495 383 1278 1278
Volkswagen Golf VI 2011 1095 1.244 1.196 0.597 1.512 1.512 481 1602 1602
Seat Alhambra 2011 1605 1.410 1.410 0.686 1.598 1.598 939 3130 3130
BMW 3er Kombi 2011 1492 1.448 1.184 0.582 1.530 1.530 747 2489 2489
C Volkswagen Caddy 2011 1492 1.448 1.184 0.582 1.530 1.530 747 2489 2489
 C
SS
0
.00

0.0

of
nt
[s]

[s]

me
80

10

%
.00

.75

95
ve
h]
nt

nt

h]

+1

+1
h]

kp

mo
ta

h]

low
]

[kp
t

12
ph

[kp

kp
ns

ns

0[

OS

OS
]
[m

be
al
0[k

co

co

S
[
0

00

-12

tC

er
VS
10

-60

02

est
in
t8

0-1
A

lat
at
80

sa
SW

SW

FM

Ga
t

e0

e4

gt
ca

ca

ss
e

me

ss
as
im

tim

tim

pa

rin
Batch run results using BAPS model

pa
Ac

th
of

of

op

he
lA

i
t

wi

du
to
n

to
yw
ion

ion

ion

ion
al
Table D.1: Results summary for test objects.

t
a

tio

tio
ter

te r

nt

in
A
at

at

at

at

nc
isa

isa

SW

SW

SW
lie

Ga
La

La

ler

ler

ler

ler

ue
ar

ar

mp
ce

ce

ce

ce
ax

ax

ax

ax

ax
eq
ne

ne

in
Co
Ac

Ac

Ac

Ac

Fr
M

M
Li

Li
Test object:
BMW 535i 2011 BAPS 10 Jun 2014 0.94 0.64 55.99 37.40 7.37 11.07 4.11 18.43 Pass 271 271 42 1.07 155.97
Volkswagen Passat 2 0TDI 2011 BAPS 10 Jun 2014 0.64 0.62 53.30 34.85 7.00 10.58 4.15 18.11 - 56 56 40 0.92 110.51
Mercedes Vivaro 2011 BAPS 10 Jun 2014 0.57 0.52 56.63 36.96 6.99 10.69 4.49 18.24 - 43 43 43 0.91 114.76
Fod Kuga TDCI 2011 BAPS 10 Jun 2014 0.73 0.53 53.32 36.07 7.82 11.83 3.97 18.82 - 57 57 41 1.24 139.52
Skoda Yeti Tdi 4x4 2011 BAPS 10 Jun 2014 0.55 0.52 51.28 34.83 7.77 11.71 4.05 18.82 - 47 47 39 1.33 134.51
Volvo XC90 2011 BAPS 10 Jun 2014 0.59 0.58 51.59 33.33 6.79 10.48 4.56 18.09 - 39 39 39 0.86 80.61
Skoda Fabia Kombi TSI 2011 BAPS 10 Jun 2014 0.58 0.56 48.13 31.58 7.50 11.67 3.93 18.43 - 44 44 36 0.93 109.17
Skoda Octavia 2011 BAPS 10 Jun 2014 0.66 0.64 50.17 32.88 7.25 10.95 3.66 18.12 Pass 273 273 38 0.93 112.26
Suzuki Swift 2011 BAPS 10 Jun 2014 0.62 0.61 40.46 25.87 7.26 11.30 3.80 18.13 - 30 30 30 0.75 127.43
Smart for 2 2011 BAPS 10 Jun 2014 0.47 0.45 32.94 22.93 8.46 13.81 5.40 20.13 - 20 20 25 1.49 171.88
BMW 1er 2011 BAPS 10 Jun 2014 0.62 0.56 51.71 35.13 7.92 12.19 4.02 18.92 Pass 276 276 39 1.30 123.14
BMW X3 2011 BAPS 10 Jun 2014 0.61 0.59 49.90 32.26 6.95 10.46 3.99 17.98 - 45 45 38 0.84 139.34
Citr en C3 2011 BAPS 10 Jun 2014 0.61 0.59 41.42 26.49 7.29 11.43 3.91 18.21 - 31 31 31 0.75 123.22
Opel Astra Sports Tourer 2011 BAPS 10 Jun 2014 0.63 0.61 51.51 33.71 7.24 10.90 3.69 18.12 - 54 54 39 0.92 182.45
Volkswagen Sharan 2011 BAPS 10 Jun 2014 0.58 0.56 53.14 34.74 7.16 10.75 3.89 18.14 - 48 48 40 0.92 119.85
BMW 530 D Gran Turismo 2011 BAPS 10 Jun 2014 0.66 0.65 53.55 34.57 6.61 10.45 4.90 18.09 - 56 56 40 0.82 101.33
Seat Ibiza 1 2 Ecomotive 2011 BAPS 10 Jun 2014 0.64 0.63 42.27 26.98 7.05 10.69 3.51 17.83 - 32 32 32 0.75 150.87
Mercedes CLS 350 CDI 2011 BAPS 10 Jun 2014 0.78 0.79 58.68 40.05 7.47 11.25 4.29 18.62 Pass 273 273 44 1.14 131.27
Volkswagen Passat Variant 2011 BAPS 10 Jun 2014 0.62 0.61 52.28 34.20 7.19 10.79 3.75 18.10 - 47 47 39 0.92 142.90
Opel Astra 2011 BAPS 10 Jun 2014 0.62 0.61 51.50 33.70 7.17 10.77 3.78 18.09 - 47 47 39 0.92 109.31
Volkswagen Polo TDI 2011 BAPS 10 Jun 2014 0.64 0.62 41.67 26.62 7.21 11.18 3.74 18.04 - 31 31 31 0.75 122.31
Volkswagen Touareg Hybrid 2011 BAPS 10 Jun 2014 0.60 0.59 52.51 33.92 6.60 10.61 5.16 18.23 - 40 40 40 0.82 101.61
Honda Jazz 2011 BAPS 10 Jun 2014 0.62 0.60 40.98 26.19 7.26 11.31 3.82 18.13 - 31 31 31 0.75 125.03
Volkswagen Golf VI 2011 BAPS 10 Jun 2014 0.63 0.62 46.30 30.43 7.48 11.60 3.87 18.39 - 49 49 35 0.95 141.57
Seat Alhambra 2011 BAPS 10 Jun 2014 0.60 0.59 52.07 33.64 6.94 10.45 3.97 17.96 - 47 47 39 0.83 155.80
BMW 3er Kombi 2011 BAPS 10 Jun 2014 0.80 0.62 55.15 37.35 7.90 11.93 4.01 18.92 Pass 273 273 41 1.35 98.20
D Volkswagen Caddy 2011 BAPS 10 Jun 2014 0.52 0.51 48.50 31.81 7.37 11.13 3.74 18.27 - 37 37 37 0.96 116.64
E Folder strucutre of Evaluation tool
Both within education and research it is beneficial to be able to ensure that the vehicle used, real or virtual,
behaves as it should. In order to verify this a set of test procedures is needed. By implementing these test
procedures in to an evaluation tool the user is presented with an efficient method to verify a model through
simulation way and analysis of the results.
Within this thesis a set of test procedures have been selected and implemented in an evaluation tool divided
into virtual test and analysis in order to both be able to use it for simulations as well as for post processing of
real test data. This report focuses on the choice and implementation of a limited number of test procedures as
well as the design of the evaluation tool developed.
Within this thesis Steady state cornering, Sine with dwell, Continuous sinusoidal input and Straight ahead
acceleration have been implemented in to the evaluation tool. This gives the user an partial overview. Preferably
several more test procedures should be implemented in the tool in order to get a more comprehensive view
of the vehicles behavior. A recommendation for future work is to implement: Step input, Random input,
Sinusoidal input, one period, Pulse input, Braking with split coefficient of friction, Brake in turn and Straight
braking.
In order to give the user freedom to adjust the test procedures based on his specific requirements, all test
procedures have been implemented using reference document with parametrizations for each test procedure.
This is done in order to to ensure good adaptability
The tool developed within this test is easy to set up and supports batch processes which enables it to
run several simulations using different model. This have been done with one model from the research project
Balance Active and Passive Safety (BAPS). The evaluation tool is based on an combination of Matlab and
Simulink. The results from the analysis part of the tool are presented as an pdf document.

The purpose with creating this tool is to provide a method to easily and quickly get results on the be-
havior of a real vehicle or a vehicle model during a few standard tests. This purpose is fulfilled with the tool
created within this thesis.
Figure E.1: File and folder structure for Evaluation tool