Electrically Propelled Semi-trailer

Dimensioning and Packaging of an Electrical Powertrain in a

Semi-trailer
Master’s thesis in Mobility Engineering

Anton Gustafsson
Eric Olsson

DEPARTMENT OF MECHANICS AND MARITIME SCIENCES

C HALMERS U NIVERSITY OF T ECHNOLOGY
Gothenburg, Sweden 2023
www.chalmers.se
 Master’s thesis 2023

Dimensioning and Packaging of an Electrical

Powertrain in a Semi-trailer

Anton Gustafsson
Eric Olsson

Department of Mechanics and Maritime Sciences

Division of Vehicle Engineering and Autonomous Systems
Chalmers University of Technology
Gothenburg, Sweden 2023
Dimensioning and Packaging of an Electrical Powertrain in a Semi-trailer
Anton Gustafsson
Eric Olsson

© Anton Gustafsson, 2023.

© Eric Olsson, 2023.

Supervisor: Clive Misquith, Afry

Supervisor: Per-Axel Ohlsson, Afry
Supervisor: Emil Pettersson, Volvo Group Trucks Technology
Supervisor: Lena Larsson, Volvo Group Trucks Technology
Examiner: Fredrik Bruzelius, Department of Mechanics and Maritime Sciences

Master’s Thesis 2023

Department of Mechanics and Maritime Sciences
Division of Vehicle Engineering and Autonomous Systems
Chalmers University of Technology
SE-412 96 Gothenburg

Cover: First generation E-Trailer

Typeset in LATEX, template by Kyriaki Antoniadou-Plytaria

Printed by Chalmers Reproservice
Gothenburg, Sweden 2023

iv
Dimensioning and Packaging of an Electrical Powertrain in a Semi-trailer
Anton Gustafsson
Eric Olsson
Department of Mechanics and Maritime Sciences
Chalmers University of Technology

Abstract
This thesis set out to dimension an electrical powertrain and package it in an ex-
isting chassis. The thesis started with benchmarking an existing E-Trailer together
with the responsible people. The pre-study started with a market analysis of other
companies to better understand what they were developing. But to get a deeper
understanding of what the companies were developing the market analysis trans-
formed into a patent search. The pre-study resulted in a list of requirements for the
E-Trailer.
To find out what dimensions the motor needed to execute functions such as increas-
ing the startability and reversing only using the trailer’s powertrain several hand
calculations were made. To optimise the powertrain for fuel efficiency on different
routes several drive cycles simulations were made. Once the dimensions of the motor
and the batteries were decided it was time to select components that were available
or soon to be available within Volvo Group. When the components were selected a
rough CAD assembly was created to find out if the powertrain would work in the
special chassis. The E-Trailer is expected to save up to 42% fuel compared to a
conventional setup on a predetermined route.

Keywords: Electromobility, E-Mob, E-mobility, E-Trailer, Electric Trailer, Electric

Powertrain

v
Acknowledgements
First of we would like to thank our supervisors Clive Misquith, Emil Petterson,
Lena Larsson and Per Axel Ohlsson for their support through out this thesis.
The insights and knowledge they have shared have been of great help in this thesis.

We would also like to thank the following persons for their support and advice during
this thesis:
• Per Björe, CPAC
• Olof Cronquist, CPAC
• Lennart Cider, Expert Product Dev./Engineering, Volvo Technology
• Niklas Fröjd, Expert Technology Specialist-Handling, Volvo Technology
• Geert Iven, Goodyear
• Klara Pålsson, Product Architect, Volvo Technology
Lastly, we would like to thank the HCT group at Volvo Technology for including
us in their group and giving us the support needed to make the best out of this thesis.

Anton Gustafsson
Eric Olsson
Gothenburg, June 2023

vii
 List of Acronyms

Below is the list of acronyms that have been used throughout this thesis:

E-Trailer Semitrailer with electric powertrain

SoC State of Charge
ICE Internal Combustion Engine
HCT High Capacity Transportation
GCW Gross Combination Weight
GTT Group Trucks Technology
E-Axle Driven axle with integrated Motors
GBG Gothenburg
EDB Engineering Database
AVP Automated Vehicle Packaging
EM Electric Machine
CAD Computer Aided Design

ix
 Nomenclature

Below is the nomenclature of parameters and variables that have been used through-
out this thesis.

Parameters

cf Rolling resistance coefficient

mi Mass of the component i
Af rontal Frontal area of the truck
cd Drag coefficient
ρair Density of the air
g Gravity
ashif t Acceleration during shifting
µ Friction coefficient
rw Radius of the wheel
Costk Cost of the component k
tl Time of instance l
vmax Maximal Velocity
Bwindow SoC window
θ Angle of friction force
Vstart Initial velocity
tshif ting Time to shift
ln Length of component n
hm Height of component m
αslope Angle of the slope

Variables

xi
 Pij Power on/from component ij
Nji Normal force on/from component ji
Fj Force on/from component j
dmax Maximal distance traveled in a day

xii
 Contents

List of Acronyms ix

Nomenclature xi

List of Figures xvii

List of Tables xix

1 Introduction 1
1.1 Aim . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.2 Limitations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2

2 Background 3
2.1 Use Cases for an E-Trailer . . . . . . . . . . . . . . . . . . . . . . . . 3
2.1.1 Reversing and arranging at terminals . . . . . . . . . . . . . . 3
2.1.2 Startability . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2.1.3 Stability while descending . . . . . . . . . . . . . . . . . . . . 4
2.2 Regulations and Standards . . . . . . . . . . . . . . . . . . . . . . . . 5
2.2.1 Standard Weights . . . . . . . . . . . . . . . . . . . . . . . . . 6
2.2.2 Standard Lengths & widths . . . . . . . . . . . . . . . . . . . 7
2.2.3 High Capacity Transportation . . . . . . . . . . . . . . . . . . 8
2.3 Naming conventions and basic concepts . . . . . . . . . . . . . . . . . 9
2.3.1 Axle variants and configurations . . . . . . . . . . . . . . . . . 9
2.3.2 5th Wheel and Kingpin . . . . . . . . . . . . . . . . . . . . . . 11
2.4 Maneuverability and Stability . . . . . . . . . . . . . . . . . . . . . . 11

3 Method 13
3.1 Method Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
3.2 The Prestudy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
3.2.1 Internal Reports, Experiences, and Testing . . . . . . . . . . . 14
3.2.2 Market Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . 15
3.2.3 Patent search . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
3.3 List of Requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
3.3.1 Vehicle Dynamics . . . . . . . . . . . . . . . . . . . . . . . . . 21
3.3.1.1 Baseline truck and trailer . . . . . . . . . . . . . . . 21
3.3.1.2 Base cargo weight . . . . . . . . . . . . . . . . . . . 22
3.3.1.3 Reverse straight . . . . . . . . . . . . . . . . . . . . 22

xiii
Contents

3.3.1.4 Power shifting . . . . . . . . . . . . . . . . . . . . . . 26

3.3.1.5 Reverse with angled truck . . . . . . . . . . . . . . . 28
3.3.1.6 Sufficient power at 90 km/h . . . . . . . . . . . . . . 30
3.3.2 Other requirements . . . . . . . . . . . . . . . . . . . . . . . . 30
3.4 Simulations & Dimensioning of Components . . . . . . . . . . . . . . 30
3.4.1 Simulation Logic Description . . . . . . . . . . . . . . . . . . . 31
3.4.2 Simulation Tool Validation . . . . . . . . . . . . . . . . . . . . 32
3.4.2.1 Diesel Truck Comparison . . . . . . . . . . . . . . . 32
3.4.2.2 E-Trailer Comparison . . . . . . . . . . . . . . . . . 32
3.4.3 Sweep simulations . . . . . . . . . . . . . . . . . . . . . . . . . 33
3.4.3.1 Parameter Determination . . . . . . . . . . . . . . . 34
3.4.3.2 Route selection . . . . . . . . . . . . . . . . . . . . . 36
3.4.4 Powertrain Comparisons . . . . . . . . . . . . . . . . . . . . . 40
3.4.5 Final Powertrain . . . . . . . . . . . . . . . . . . . . . . . . . 42
3.4.5.1 Selection of components . . . . . . . . . . . . . . . . 42
3.4.5.2 Final Powertrain Simulations . . . . . . . . . . . . . 42
3.5 Concept ideas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
3.5.1 Axle configuration . . . . . . . . . . . . . . . . . . . . . . . . 43
3.5.1.1 Choice of driven axle . . . . . . . . . . . . . . . . . . 43
3.5.2 Powertrain layout . . . . . . . . . . . . . . . . . . . . . . . . . 48
3.6 Packaging of the powertrain . . . . . . . . . . . . . . . . . . . . . . . 54
3.6.1 Obtaining CAD models of components . . . . . . . . . . . . . 54
3.6.1.1 Chassis . . . . . . . . . . . . . . . . . . . . . . . . . 55
3.6.1.2 Electrical powertrain . . . . . . . . . . . . . . . . . . 55
3.6.1.3 Mechanical powertrain . . . . . . . . . . . . . . . . . 55

4 Results & Discussion 57

4.1 List of Requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
4.1.1 Vehicle Dynamics . . . . . . . . . . . . . . . . . . . . . . . . . 57
4.1.1.1 Straight Reverse . . . . . . . . . . . . . . . . . . . . 57
4.1.1.2 Angled Reverse . . . . . . . . . . . . . . . . . . . . . 58
4.1.1.3 Powershift . . . . . . . . . . . . . . . . . . . . . . . . 59
4.2 Simulation & Dimensioning . . . . . . . . . . . . . . . . . . . . . . . 60
4.2.1 Simulation Tool Validation . . . . . . . . . . . . . . . . . . . . 60
4.2.1.1 Diesel Truck Validation . . . . . . . . . . . . . . . . 60
4.2.1.2 E-Trailer Validation . . . . . . . . . . . . . . . . . . 61
4.2.2 Sweep Simulations . . . . . . . . . . . . . . . . . . . . . . . . 62
4.2.2.1 3 × Gothenburg-Borås-Gothenburg . . . . . . . . . . 62
4.2.2.2 2 × Tampere-Helsinki-Tampere . . . . . . . . . . . . 63
4.2.2.3 Kurtalan-Bahçesaray-Cizre . . . . . . . . . . . . . . 64
4.2.2.4 Sensitivity analysis of charging price . . . . . . . . . 65
4.3 Final Powertrain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
4.3.1 Powertrain selection . . . . . . . . . . . . . . . . . . . . . . . 66
4.3.2 Final Powertrain Simulations . . . . . . . . . . . . . . . . . . 68
4.4 Packaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
4.4.1 Chassis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73

xiv
 Contents

4.4.2 Mechanical powertrain . . . . . . . . . . . . . . . . . . . . . . 75

4.4.3 Electrical powertrain . . . . . . . . . . . . . . . . . . . . . . . 76
4.4.4 Requirement fulfillment . . . . . . . . . . . . . . . . . . . . . 77

5 Conclusion 79
5.1 General conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
5.2 Simulations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
5.3 Packaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81

Bibliography 83

A Appendix A I

B Appendix B III

C Appendix C V

D Appendix D VII

E Appendix E IX
E.1 Second and third axle lifted . . . . . . . . . . . . . . . . . . . . . . . IX
E.2 First and second axle lifted . . . . . . . . . . . . . . . . . . . . . . . . X

F Appendix F XI

xv
Contents

xvi
 List of Figures

2.1 Reverse assist for longer combinations. Idea by Lena Larsson and
Emil Petterson. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2.2 Components that could be used in a vehicle combination [7] . . . . . 5
2.3 Bearing capacity classes of roads within Gothenburg [10] . . . . . . . 6
2.4 Visualization of the longest vehicle combinations from the 25.25m
regulation [14] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2.5 Turning radius requirement . . . . . . . . . . . . . . . . . . . . . . . 8
2.6 Carbon footprint of different weight limited transportation. Green
boxes indicate measured fuel savings[7] . . . . . . . . . . . . . . . . . 8
2.7 Decrease of vehicles needed for the same volume transported [7] . . . 9

3.1 Powertrain layout of Trailer Dynamics E-Trailer. . . . . . . . . . . . . 16

3.2 Powertrain layout and components description of ZF’s E-Trailer. . . . 17
3.3 Powertrain layout and specifications of VAK’s E-Trailer. . . . . . . . 18
3.4 Schematic view of the base truck and trailer combination with mea-
surements. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
3.5 Free body diagram of the truck and trailer configuration, not includ-
ing kingpin forces. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
3.6 Free body diagram of the truck. . . . . . . . . . . . . . . . . . . . . . 25
3.7 Free body diagram of the trailer. . . . . . . . . . . . . . . . . . . . . 26
3.8 Forces that act on the combination while shifting in a slope. . . . . . 27
3.9 Free body diagram of the vehicle combination doing a angled reverse 29
3.10 Free Body diagram of a tire in the angled reverse scenario . . . . . . 29
3.11 Fuel consumption comparison of E-trailer weight and a conventional
trailer when loaded with 15 ton payload . . . . . . . . . . . . . . . . 35
3.12 Altitude profile of the route Tampere to Helsinki to Tampere . . . . . 37
3.13 Altitude profile of the route 2x Tampere to Helsinki to Tampere . . . 37
3.14 Altitude profile of the route from Kurtalan to Bahçesaray to Cizre . . 38
3.15 Altitude profile of the route from Kurtalan to Bahçesaray to Cizre to
Kurtalan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
3.16 Altitude profile of the route Göteborg to Viared to Göteborg . . . . . 39
3.17 Altitude profile of the route 3x Göteborg to Viared to Göteborg . . . 40
3.18 Truck and trailer combination with second and third axle lifted. . . . 44
3.19 Truck and trailer combination with first and second axle lifted. . . . . 45
3.20 Truck and trailer combination with first and third axle lifted. . . . . . 45
3.21 Free body diagram of the trailer with no axles lifted. . . . . . . . . . 46

xvii
List of Figures

3.22 Free body diagram of the trailer with first and second axle lifted. . . 47
3.23 Propeller shaft angle, α, is dependent of the vertical position of the
axles. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
3.24 Side and top view of the distances needed to calculate the propeller
shaft length. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
3.25 Propeller shaft angle, β, is not dependent on the vertical position of
the axles. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
3.26 Trailer Dynamics packaging of their powertrain. . . . . . . . . . . . . 53
3.27 Battery position on the first generation of the E-Trailer at Volvo Trucks. 53
3.28 Radiator placement on the first E-Trailer . . . . . . . . . . . . . . . . 54
3.29 ISO view of the Parator Chassis. . . . . . . . . . . . . . . . . . . . . 55

4.1 Angled reverse simulation results . . . . . . . . . . . . . . . . . . . . 59

4.2 Energy needed from the trucks diesel tank with E-Trailer compared
to a conventional trailer on the route Gothenburg to Borås . . . . . . 69
4.3 Energy needed from the trucks diesel tank with E-Trailer compared
to a conventional trailer on the route from Tampere to Helsinki . . . 70
4.4 Energy needed from the trucks diesel tank with E-Trailer compared
to a conventional trailer on the route in eastern Turkey . . . . . . . . 71
4.5 Isometric view of the E-Trailer . . . . . . . . . . . . . . . . . . . . . . 72
4.6 Side view of the E-Trailer . . . . . . . . . . . . . . . . . . . . . . . . 72
4.7 Isometric view of the chassis . . . . . . . . . . . . . . . . . . . . . . . 74
4.8 Side view of the chassis . . . . . . . . . . . . . . . . . . . . . . . . . . 74
4.9 Side view of drive axle in full droop and pusher in full bump. . . . . . 75
4.10 Top view of drive axle in full droop and pusher in full bump. . . . . . 75
4.11 Isometric view of the mechanical powertrain . . . . . . . . . . . . . . 76
4.12 Isometric view of the electric powertrain. . . . . . . . . . . . . . . . . 77
4.13 Top view of the assembly with components marked out. . . . . . . . . 77

D.1 Side view. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VII

D.2 Top view. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VII
D.3 Side view. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VIII
D.4 Top view. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VIII

E.1 Free body diagram of the trailer with second and third axle lifted. . . IX
E.2 Free body diagram of the trailer with first and second axle lifted. . . X

xviii
 List of Tables

3.1 Base truck specifications. . . . . . . . . . . . . . . . . . . . . . . . . . 21

3.2 Base truck specifications. . . . . . . . . . . . . . . . . . . . . . . . . . 22
3.3 Lengths of truck and trailer. . . . . . . . . . . . . . . . . . . . . . . . 23
3.4 Constants and their values used for calculating the force needed to
start reversing straight. . . . . . . . . . . . . . . . . . . . . . . . . . . 24
3.5 Values used for calculating the energy lost during an up-shift. . . . . 28
3.6 Values of αslope and mtrailer used for testing. . . . . . . . . . . . . . . 28
3.7 Angles used for iterative testing. . . . . . . . . . . . . . . . . . . . . . 28
3.8 Constants and their values used for calculating the force needed to
reverse with the truck in a angle. . . . . . . . . . . . . . . . . . . . . 29
3.9 Values used in the simulation tool to validate the diesel consumption
of a conventional truck. . . . . . . . . . . . . . . . . . . . . . . . . . . 32
3.10 Values used in the simulation tool to validate PT part of code. . . . . 33
3.11 Selection of motors that were used in the sweep simulations. . . . . . 34
3.12 Ratios for calculating the mass of motors and batteries depending on
peak power and capacity. . . . . . . . . . . . . . . . . . . . . . . . . . 35
3.13 Values of constants used to calculate trailer weight. . . . . . . . . . . 36
3.14 Prices used to calculate energy costs. . . . . . . . . . . . . . . . . . . 42
3.15 Values of base trailer constants used to calculate the kingpin forces,
Fkp . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
3.16 Positive distance indicates the compression of the suspension and the
axle moves up. A negative indicates the opposite and the axle moves
down. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50

4.1 Results of hand-calculations for reversing straight . . . . . . . . . . . 58

4.2 Values of mtrailer and αslope and the resulting power need from the
EM at 40km/h initial Velocity . . . . . . . . . . . . . . . . . . . . . . 60
4.3 Diesel consumption from Autofreight vehicle compared to simulation
tool . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
4.4 Results from real test data and simulation data . . . . . . . . . . . . 61
4.5 Results of sweep simulations on the route 3xGBG-Via-GBG . . . . . 62
4.6 Results of sweep simulations on the route 2xtamp-helsi-tamp . . . . . 63
4.7 Results of sweep simulations on the route Kurtalan-Bahçesaray-Cizr . 64
4.8 Sensitivity analysis of charging price for GBG-Viared. . . . . . . . . . 65
4.9 Sensitivity analysis of charging price for Helsinki. . . . . . . . . . . . 65
4.10 Sensitivity analysis of charging price for Turkey. . . . . . . . . . . . . 66

xix
List of Tables

4.11 Selected in-house components. . . . . . . . . . . . . . . . . . . . . . . 67

4.12 Specification for the selected components used in the final simulation. 67
4.13 Specifications of the batteries used in the final simulations . . . . . . 67
4.14 Results of final powertrain simulations on the route 3xGbg-Borås-Gbg 69
4.15 Results of final powertrain simulations on the route 2xTampere-Helsinki-
Tampere . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
4.16 Results of final powertrain simulations on the route Kurtalan-Bahçesaray-
Cizre-Kurtalan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
4.17 Final axle distances for the E-Trailer. . . . . . . . . . . . . . . . . . . 73
4.18 Resulting kingpin force with different combinations of lifted axles
using base trailer distances. . . . . . . . . . . . . . . . . . . . . . . . 73
4.19 Resulting king pin force with different combination of lifted axles
using final distances. . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
4.20 Resulting values of lprop and α from the three scenarios tested. . . . . 76

xx
 1
Introduction

New long-term goals of global emissions such as the ’Paris Agreement’, which partly
states that there should be a decrease of emissions such that the general temperature
only rises with 2°C [5] and the ’European Green Deal’ which states that until 2030
Europe should have a decrease of 55% compared to the carbon footprint of 1990 as
well as being CO2 -neutral by 2050[3] creates the need of change within the mobility
industry. There is an interest in realizing projects that can help Europe become
the first climate-neutral continent. One of the many steps towards this is to make
transportation more sustainable and less dependent on resources that contribute to
the negative effects on the climate.
The move towards more sustainable transport solutions is making the transport
industry move towards different energy sources than traditional fossil fuels. Volvo
Trucks plans to have a fossil-free product range by 2040 with the help of battery-
electric and fuel-cell electric vehicles [16]. This change in energy sources also brings
changes in how truck manufacturers look at their product range. These new energy
sources are less energy dense compared to liquid fossil fuels, like diesel, which means
that the available transportation range is more limited than before. This requires
the transportation solution to be more tailored to the specific type of transport and
route. Together the fact that the customer does not want to pay for, potentially,
over-dimensioned components that could be a result of limited customization, shows
that the need for more customization will be even more preferable in the future. This
can be hard to meet at the large scale of a truck manufacturer, like Volvo Trucks,
due to each truck being built requiring more manual work than before to meet the
customer’s expectations and requirements. A commonly used solution to this is to
make the trucks modular. With the change in fuel and powertrains, this appears to
be even more important, to make a competitive package for the new trucks compared
to the old variants.
A promising idea that should decrease the carbon footprint is to integrate an elec-
trical propulsion system in a semitrailer, further E-Trailer. This makes it possible
for the semitrailer to help the truck during several operations. One example of an
operation is highway driving. The highways often have elevation differences where
the E-trailer could make it possible to help propel the truck during uphills while
recuperating energy during downhills. The downhill recovery will also help with
improving the stability of the truck-trailer combination compared to a regular semi-
trailer. Other benefits could be to use the propulsion system in the semitrailer to
help with operations such as reversing and pushing extra power during heavy starts,
which would decrease the energy usage from the diesel engine.

1
1. Introduction

1.1 Aim
The aim of this thesis is to present a deeper understanding of the capabilities,
benefits, and general pros and cons of having an electrically propelled trailer and
how, or if, they change depending on the route. This knowledge will help and
guide the future development of the E-Trailer at Volvo Trucks. The thesis also aims
to present a design suggestion of how a modular packaging solution for such an
E-Trailer could look like.

1.2 Limitations
This thesis work will be done during the spring of 2023 and covers 30 credits per
person, in total 60 credits for two students.
• Existing simulation tools for powertrain simulation will be used.
• Only one type of diesel truck will be considered when simulating.
• Only currently or soon to be on the market component from Volvo will be
considered in the design and simulations.
• Supplier proposals and quotations from suppliers not included.

2
 2
Background

In the following chapter background knowledge used in this thesis will be described
as well as some necessary central concepts.

2.1 Use Cases for an E-Trailer

In the introduction, it was stated that an E-Trailer can be used to reduce fuel
consumption by hybridization of existing trucks, but it is not the only way it could
be beneficial. In this section, some of these secondary features will be brought up to
show some other benefits that an E-Trailer could come with, beyond fuel savings.

2.1.1 Reversing and arranging at terminals

Traditionally, when reversing, the truck is pushing the trailer creating a twisting
torque over the axle groups. This makes the trailer steer in the opposite direction
of the steering wheel input. For an experienced driver, this phenomenon is not
hard to counter when reversing a single trailer. However high capacity transport,
HCT, projects at Volvo Trucks are testing with longer combinations, coupling two
regular trailers together in the Autofreight project [6] for example. This increases
the difficulty of reversing due to more joints being introduced and the input of the
traction force and steering now comes from the last unit in the combination. In
this scenario, it would be beneficial to have the last unit supplying the majority of
the traction force required to reverse the combination. With an E-Trailer as the
last unit, this would be possible, making the reversing of such a long combination
much easier in theory. The truck would still add enough traction force to support
itself making the electric powertrain of the E-Trailer only have to pull the trailers.
Expanding on this, the last axle of the last trailer in the combination could also be
steered to act as the main steering axle when reversing, and having axles on the
second trailer also steered would increase the maneuverability of the combination
as well. Having lifted axles on the trailers would also decrease the twisting torque
required to steer the trailers making the combination even more maneuverable. In
figure 2.1 a two-trailer combination reversing using the E-Trailer powertrain can be
seen as well as where steerable axles could be placed to increase maneuverability.

3
2. Background

Figure 2.1: Reverse assist for longer combinations. Idea by Lena Larsson and
Emil Petterson.

2.1.2 Startability
Naturally, the E-Trailer adds more traction force to the combination. This is be-
cause the normal force from the trailer’s mass now can be used as traction force.
This allows the combination to start from steeper slopes or to be able to start with
higher gross combination weights. By having an axle on the trailer the traction force
is more spread out over the entire vehicle combination. Because of this, the vehicle
combination will not be as sensitive to surface friction. The driven axle/axles on
the truck could have too little traction to deliver the required traction force to move
the combination. With the help of the E-Trailer, the extra traction force could be
applied to a surface with better friction and propel the combination.

The E-Trailer also would allow driving uphills at a more constant speed since the
electric powertrain could support with extra traction force. This extra traction force
would also allow the truck to run in higher gears while going up the hill, making
the ICE run at lower RPM and thereby saving fuel.

2.1.3 Stability while descending

The HCT group at Volvo Trucks also works with running heavier combinations to
decrease the number of transports required and thereby decreasing the emissions.
Making heavier combinations able to stop and brake efficiently is of course a concern,
especially in long descents where brakes easily can get overheated, which reduces
their efficiency. A solution to this problem is that the E-Trailer uses its electric
motor to brake. By running the electric motor in reverse, the energy that would
have been transformed into heat is instead transformed into usable electricity. This

4
 2. Background

means that the conventional brakes can be used less, since less energy is needed to
be transformed into heat and that the conventional brakes can be used for longer
periods of time without getting overheated due to heavy usage. Applying the brake
force at the end of the combination will also increase the stability of the combination
making it less prone to jack-knife, which is a scenario where the trailer pushes the
truck, making it spin and fold towards the trailer.

2.2 Regulations and Standards

The modular concept is a way of building vehicle combinations with the help of
different types of trailers and trucks [14]. The types of trailers are usually full
trailers, semi-trailers, link trailers, converter dollys, and center-axle trailers while
the trucks usually are rigid trucks and tractors. All these road train components
are illustrated as a table in figure 2.2.

Figure 2.2: Components that could be used in a vehicle combination [7]

5
2. Background

2.2.1 Standard Weights

How much a truck and trailer combination is allowed to weigh depends on what
bearing capacity class the road itself has. The bearing capacity classes differ be-
tween countries and roads within the specific country. The different national trans-
portation agencies determine what bearing capacity class the roads can take with
regard to several factors from the vehicle combinations, mainly from the axle loads
which include the total number of axles, axle configurations, distances between the
axles/axle configurations, and total vehicle combination lengths but also from the
overall condition of the road as well as its maintenance schedule [13].
Sweden has 4 different bearing capacity classes [8], BK1-BK4.
• BK1, max 64 tons, can be lower depending on distances between axles and
their axle loads
• BK2, max 51.4 tons, can be lower depending on distances between axles and
their axle loads
• BK3, max 37.5 tons, can be lower depending on distances between axles and
their axle loads
• BK4, max 74 tons, can be lower depending on distances between axles and
the same axle loads as BK1
The Swedish public road network mainly consists of roads that are classified as
BK1. This is because there are a lot of country roads within the Swedish road
network has the BK1 rating. The road ratings within cities are usually rated as BK2
which can be seen in figure 2.3 that illustrate the road network classifications within
Gothenburg, Sweden [10]. Roads that are particularly soft or include obstacles that
have a specific load capability such as bridges usually gets a BK3 rating. Since the
BK4 classification allows for the highest max loading the routes that are BK4 rated
are limited and almost exclusively highways.

Figure 2.3: Bearing capacity classes of roads within Gothenburg [10]

6
 2. Background

2.2.2 Standard Lengths & widths

The longest length of a vehicle combination that is allowed on public Swedish roads is
25.25m long, considering they are following the common rules within the European
Union. The European standard contains two different lengths and two different
widths of trucks and trailers [14]:
• 7.82m long
• 13.6m long
• 2.55m wide
• 2.60m wide, only for temperature-controlled structures with minimum 45mm
insulated walls
This makes it possible to create the most common combinations which are visible in
figure 2.4, where the trailers named "A" has the 7.82m lengths and the trailers named
"B" has the 13.6m lengths. Figure 2.4 also shows how the different combinations of
link trailers, semi-trailers, dollys, and center-axle trailers should be used.

Figure 2.4: Visualization of the longest vehicle combinations from the 25.25m
regulation [14]

To ensure that the truck and vehicle combination can safely maneuver the roads in
Sweden there are turning requirements. These turning requirements are for both
the truck itself, but there are also requirements on the entire vehicle combination.
The requirements differ a bit between the two cases considering the different lengths
that are in place. In figure 2.5 the outer radius and inner radius requirements for
an entire vehicle combination are shown. The vehicle combination has to be able to
turn with an outer radius of 12.5m and an inner radius of 2.0m. The only difference
between the entire vehicle combination and the requirements for a single truck is
that the inner radius requirement increases from 2.0m to 5.3m.

7
2. Background

Figure 2.5: Turning radius requirement

There are also several rules and exceptions that are not mentioned regarding the
distances from the kingpin to the axle configurations, the total length of the cargo
spaces, the distance from the middle axle to the rearmost part of the cargo space
(overhang), braking requirements, rules of the fifth wheel on all trailers/trucks/-
dollys, etc. More about the regulations can be read on the national transportation
agency’s web page [9] and their explanatory legal loading document [13].

2.2.3 High Capacity Transportation

High capacity transportation (HCT) is a study on how even heavier and longer com-
binations affect the climate compared to the already legal combinations. According
to the report about HCT combinations[7] where figure 2.6 is taken from, the fuel
savings of a 90ton weight limited transportation compared to a 60ton should be
around 22%.

Figure 2.6: Carbon footprint of different weight limited transportation. Green

boxes indicate measured fuel savings[7]

The same report also mentions that for volume-limited transportation the total
length of the vehicle combination is more important. The decrease in vehicles needed
for the same amount of transported cargo can be seen in figure 2.7. The longer

8
 2. Background

combination also has a lower carbon footprint and fewer drivers which could lead to
cheaper and more environmentally friendly transportation.

Figure 2.7: Decrease of vehicles needed for the same volume transported [7]

The fuel savings shown makes the argument to open up more of the road network to
a higher rating than BK4 (which currently is the highest-rated road). Sweden has
since 2020 started tests with vehicle combinations up to 100 tons and 32.5m. Finland
has already opened up a lot of its public roads to 76 tons and 34.5m combinations,
while they are also testing up to 104 tons combinations [7].
Heavier transportation could see the benefit of having functions such as better sta-
bility, reversing, and power shifting. These functions could all be done with the help
of an E-Trailer.

2.3 Naming conventions and basic concepts

In the trucking and transport industry, there are some commonly used names to
describe different types of axle configurations for how many axles a truck or trailer
has, which are driven, which are liftable, which axles are steerable, etc. These
naming conventions will be used throughout this report and are therefore important
to fully understand.

2.3.1 Axle variants and configurations

On a truck, the number of axles is counted from the front of the truck and backward,
the same applies for trailers. A three-axled truck with two of its axles driven is
denoted as a 6x4 truck. The first number, 6, describes the total amount of wheel
hubs the truck has. The latter number, 4, describes the number of driven wheel
hubs the truck has in total.

9
2. Background

There are several axle configurations possible on a truck since the number of axles
can be as much as 5. Some common configurations for trucks are 6x4, 6x2, and
4x2. When there are two rear axles on a truck, such as the 6x2 or 6x4 variants,
the rear axle configuration is known as a bogie. For semi-trailers, the conventional
configuration is to have 3 axles. These configurations can be built up with different
numbers of driven axles and also different positions of the driven axles. A truck
or trailer might also have liftable and/or steerable axles to suit their transport
requirements better. Because of this, there are some terms and concepts that need
to be known.

Tandem

A tandem bogie is defined as a bogie where both axles are driven. This will generally
allow the traction force to be distributed, which could improve the traction if the
ground surface is slippery. A tandem might have the possibility to disengage one of
the driven axles if needed. This decoupling will allow for the possibility of lifting one
of the axles. According to the "Tandem Axle Lift" in appendix F, this is especially
beneficial for energy consumption and also allows for better maneuverability or
stability.

Tag axle

A tag axle is defined as a non-driven axle behind the driven axle, it "tags along" the
driven axle. It might be liftable, steerable, or both.

Pusher axle

A pusher axle is defined as a non-driven axle in front of the driven axle. The driven
axle is pushing the axle in front of it. Like the tag axle, it can also be lifted and
steered, however, it can only be steered hydraulically. This is because of the lateral
dynamics of a truck, where if the pusher axle were to be self steered it would always
contradict the wanted lateral movement.

Liftable axles

As the name suggests, a liftable axle is an axle that can be lifted if necessary. If it
is the first axle in an axle combination it will improve the turning radius by moving
the rotation point close to the combination. By moving the rotation point closer to
the vehicle the low-speed maneuverability of the entire combination becomes better,
meaning that the inner turning radius will increase. But because the rotation point
is closer to the vehicle combination the high-speed stability will be worse than if
the axle were not lifted. Lifting axles is also a very common practice to reduce fuel
consumption. Another benefit is that lifting axles shifts the loads to other axles.
This could be beneficial if the trailer has a driven axle and needs an increase in
traction force for a start.

10
 2. Background

Steerable axles
Steerable axles come in two different layouts, actively steered or self-steered. The
benefit of steerable axles comes when the truck and trailer are loaded to a point
where lifting the axles no longer is an option. The hydraulically steered tag axle
uses a hydraulic system to actively steer the axle. The axle’s steering angle can
reach a maximum of 11.5◦ . According to the datasheet "Hydraulically steered axle"
in Appendix F the axle allows for the maximal steering angle up to a vehicle speed
of 0-15 kph. When the truck reaches 15 kph the steering angle starts to decrease
until it reaches 38 kph when it becomes fixed. The self-steered axle is pneumatically
controlled and allows for steering up to 25 kph. According to the datasheet "Self
Steered Axle" in appendix F the control cylinder locks itself in the rigid position
when it reaches 25 kph. When the truck reaches 25 kph the control cylinder locks
itself and becomes fixed.

E-Axle
An E-Axle is a driven axle that could house part of an electrical powertrain. The E-
Axles come in several different configurations. The minimal requirement is that one
or several electrical motors should be mounted to the axle. This could be in order
to save space, reduce the need for propeller shafts, etc. It can include a gearbox or
even inverters and other power electronics.

2.3.2 5th Wheel and Kingpin

Using a fifth wheel combined with a kingpin is the conventional method of mounting
semi-trailers to a tractor. The kingpin is a solid cylinder that is mounted to the
bottom side of the frame of the trailer. The fifth wheel sits on top of the frame of the
tractor and provides a flat surface that supports the trailer frame. The fifth wheel
also provides a locking mechanism that will hold the kingpin to the fifth wheel. This
setup allows the trailer to rotate around the kingpin’s axis when the truck is turning.

2.4 Maneuverability and Stability

Where and how axles are placed on a truck, trailer or a combination of both depends
on many things. First are the legal requirements for how much axle load a certain
axle is allowed to have. After that, it is a trade-off between high-speed stability and
maneuverability. A longer wheelbase, the longitudinal distance between the axles,
will result in a more stable vehicle or combination. However, the combination will
have a larger turning radius making it harder to maneuver in tight spaces. In the case
of a short wheelbase the effects are reversed, meaning that it will be more unstable
at higher speeds, but easier to maneuver. For a truck and trailer combination, this
trade-off between long and short wheelbase becomes more important since such a
combination is inherently large, but the necessity for maneuverability still exists. To
solve this problem it is common to use liftable and steerable axles to get the wanted
traits from both a long and a short wheelbase, as mentioned earlier.

11
2. Background

12
 3
Method

In the following chapter, the methods used in this thesis work will be presented. It
will begin with an overview of the methods used in this project followed by more
in-depth explanations of how and why each method was used.

3.1 Method Overview

The thesis work started with a prestudy of how Volvo Trucks built their first gen-
eration of E-Trailer, what weaknesses and strengths it had as well as what the
competitors’ solutions look like. The prestudy also included an inventory of what
electric powertrains Volvo already had in-house. From the prestudy a list of re-
quirements was set that included both requirements that a new E-Trailer should
fulfill, but also desirable features that were set as goals for the new E-Trailer. These
requirements focus both on the performance of the powertrain as well as how it will
be integrated and mounted on the trailer chassis. The performance requirements
were calculated using a baseline truck and trailer combination and simple vehicle
dynamics calculations in different scenarios.

With the list of requirements set, drive cycle simulations of different battery and
motor sizes on three different routes were conducted. The three routes consisted of
one mostly flat, one very hilly, and one route with a mix of both. The aim was to
get a better understanding of how different powertrain parameters affected different
performance parameters such as fuel consumption and profitability, compared to
the base combination on the same routes. This was done without any relation to
what was available at Volvo, but instead to see trends in which powertrain variables
affected performance parameters the most and if there were any breaking points
where certain powertrain variables stopped or had less effect on the performance
parameters. These trends were then used, together with the list of requirements,
to decide which existing powertrain solutions that could be suited for each route.
These trends also made it possible to leave recommendations for what an optimal
powertrain for an E-Trailer could look like.

With the selection of powertrain components completed, a packaging study was

carried out in order to make sure the selected components would be possible to
mount in the trailer. The decision on the powertrain layout was influenced by the
selected components interfaces, the list of requirements, and common practices. A
package solution for the E-Trailer was then done using PTC Creo. The packaging

13
3. Method

was done as a base to start from and to be changed further down the line so that
lessons learned from unexpected changes could be implemented.

3.2 The Prestudy

The prestudy was initiated by gathering information about how Volvo Trucks built
their first E-Trailer. This was done by reading internal documentation and reports
from the creation of Volvo’s first E-Trailer, see Appendix F. Reports and documen-
tation from other electromobility and high capacity transportation, HCT, projects
at Volvo were also included to get a better understanding of what similar concepts
had already been investigated and what was learned from them. Simultaneously,
there were ongoing discussions with the engineers involved in the creation of the
first E-Trailer about how and why certain decisions were taken, filling in informa-
tion gaps that the reports did not cover or what they thought was important to
investigate further in a future generation of the E-Trailer.

An analysis of the current market and competitors was also carried out during
the prestudy, with the intention of widening the understanding of the market and
current solutions. Since the concept of an electrically propelled trailer still was in
its initial stage of development, even for competitors, information was difficult to
gather. Volvo Trucks had done a market analysis for the development of the first
E-Trailer where the most information was gathered by networking, conferences, and
press releases. The result of this market analysis turned out to cover most of the
competitors and was deemed to still be valid and useful for this project. As an
extension of the market analysis, a patent search was also conducted to get a better
picture of what competitors could be working on or if any other competitors could
be expected to enter the market in the future.

3.2.1 Internal Reports, Experiences, and Testing

The overall conclusion drawn from the knowledge gathered was that it is possible
to package an electrical powertrain in a trailer. It also showed some very promising
results in terms of decreased fuel consumption. However, the battery’s SoC window
and motor torque were two limiting factors that were discovered in the first pro-
totype. Especially when it came to reversing the truck and trailer only using the
electric motor on the trailer. The motor could just supply enough power to reverse
the truck and trailer straight and on very flat ground. The SoC window of the
batteries was set between 30-70 % which was very conservative but still managed to
show a positive effect on the fuel consumption on Landsvägsbanan at Hällered. A
bigger SoC window or more capacity would have been interesting to be able to test
for longer instances and take more advantages of regenerative braking.

What was learned during the building of the E-Trailer was mainly that the usage of
two cooling mediums, requiring two separate loops was not recommended in future
builds. Why this was done in the first E-Trailer was due to the approach taken
during designing it. The approach was to build it with what was available "in the

14
 3. Method

yard" and make the best out of these components.

A similar project that was of interest was the E-Dolly, an electrically powered dolly
that was built a couple of years before the first E-Trailer. The most interesting
aspect of this project was the powertrain and its layout. It had the last axle driven
meaning that a propeller shaft passed over the first axle to the two electric motors.
Such a layout would open up the possibilities of axle configurations on a future
E-Trailer. Having different configurations possible could allow a selection from the
customer to optimize for their needs.

From the discussions with the engineers involved in the first E-Trailer, it was de-
cided that this E-Trailer should be a three-axled trailer like the first generation. Due
to it being a common configuration of axles for a trailer. The previous E-Trailer
was based on a container trailer and a decision was made to stick with this type of
trailer for this E-Trailer as well. A container trailer is a cheap chassis to buy, easy
to work on, and flexible to test with due to the ease of changing the gross combina-
tion weight, GCW, by changing containers or their contents. It is also decided that
powertrain components should be sourced from what was available at Volvo Group.
This was to make the building easier and less costly.

3.2.2 Market Analysis

The following subsection will briefly describe the competitors’ E-Trailers in terms
of specifications and choice of powertrain. It will also bring up how far they have
come in their development.

Trailer Dynamics

Trailer Dynamics has three different versions of their E-Trailer, all with continuous
power of 360 kW and peak power of 580 kW. The difference is the battery capacity
with a 300 kWh, a 450 kWh, and a 600 kWh configuration. Trailer dynamics in-
cludes a 44 kW charger, the system runs on 800 V and the 600 kWh configuration
has a claimed 21.4-ton payload capacity.
Trailer Dynamics uses an E-axle for their powertrain and it’s placed as the second
axle on the trailer. Their E-axle allows for a compact design that does not seem to
interfere with their axle placement. Batteries are located in front of the first axle
in a modular system. The inverters and other electronics seem to be placed behind
the third axle. Figure 3.1 shows the trailer and powertrain layout.

15
3. Method

Figure 3.1: Powertrain layout of Trailer Dynamics E-Trailer.

In April of 2023 Trailer.se wrote an article about DB Schenkers field testing of Trailer
Dynamics E-Trailers [1]. The claimed results showed a decrease of fuel consumption
between 24-55 % according to DB Schenker. They also claimed that the fuel savings
were just 0.7 to 0.9 % off from the expected values.

ZF

ZF bought WABCO on the 29th of May in 2020 and thereby acquired their E-Trailer
product. WABCO’s design, from 2019, had the motor powering the third axle with
the motor placed behind it. The batteries were placed in front of the first axle, like
Trailer Dynamics. The inverter and the cooling system were both placed behind the
third axle on each side of the motor. The ZF layout, presented in 2021, was slightly
different from the WABCO design. It seemed to have the second axle driven with
an E-axle and the inverters packaged between the batteries and axles.
The specification of the motors and batteries were not disclosed by either WABCO
or ZF. Judging from the short clips and images that were available it seemed like
it is a smaller battery and motor compared to other competitors. This assumption
was also backed up by a ZF press release seen in Appendix B (also have this link to
the same thing, where they estimate up to 16 % fuel and CO2 savings on shorter
routes and up to 7 % on longer routes. Figure 3.2 shows the trailer and powertrain
layout.

16
 3. Method

Figure 3.2: Powertrain layout and components description of ZF’s E-Trailer.

Randon

The Brazilian trailer manufacturer Randon also provides an E-axle solution. It

consists of two electric motors mounted to the axle. They have the possibility to
have the driven axle as either the first, second, or third. In a commercial clip,
however, the axle is placed as the second axle with batteries and inverters in front
of the first axle. Randon also seems to go down the route of smaller batteries and
motors making their powertrain very compact and easier to package on the trailer.
This should also result in less cost for the final product which is also an important
aspect.

VAK

The Finish trailer and body builder VAK is currently working on a pilot project
for an E-Trailer that they aim to commission during 2023. It consists of an E-axle
propelling the second axle. Battery and cooling are mounted in front of the first
axle and by the presented specifications the capacity is 15 kWh. VAK estimates 2-10
% lower fuel consumption with this powertrain. VAK also claims that the whole
powertrain will only add 1000 kg to an existing trailer. Figure 3.3 shows the trailer
and powertrain layout.

17
3. Method

Figure 3.3: Powertrain layout and specifications of VAK’s E-Trailer.

3.2.3 Patent search

As a part of the market analysis, a patent search was also carried out to obtain more
information about the market and the competitors. Due to the E-Trailer concept
still being in the research and design phase a patent search could give more infor-
mation about which companies that could be venturing into the market and in what
way.

The database Espacenet was used and the keywords used are found in the list be-
low. Not all keywords resulted in relevant results and therefore all results will not
be covered in this report. Instead, a summary of the most relevant results will be
presented in the following subsections.

• Electric
• Semi-Trailer
• Semitrailer
• Trailer
• Propelled
• Driven
• ZF
• Trailer Dynamics
• Actuated
• Self propelled
• Self driven
• Electrified

From the patent search, a few patents stood out as being more interesting or relevant.
Some patents found even described the whole concept of an E-Trailer and gave a
glimpse into what the competitors might be trying to develop. It was also discovered

18
 3. Method

that many patents cover the concept of a driven trailer and that these patents were
quite similar. Therefore not all of them will be covered here and instead, only the
ones that are more applicable to the thesis. In the following sections, these patents
will be shortly described.

DE102021202321A1
Published: 15-09-2022
Keywords: "ZF" and "trailer" and "propelled"

This patent from Zahnradfabrik Friedrichshafen, ZF, describes a truck towing (with
a traction battery and a charger) and a trailer (which is also carrying traction
batteries and a charger). The patent covers how the distribution of energy could be
done during charging and during driving, i.e. which traction battery to use up first,
if the second one should charge the first one, etc. It also mentions that the trailer
could include a motor to provide extra power/energy when needed.

DE102020108391A1
Published: 30-09-2021
Keywords: "ZF" and "trailer" or "propelled" or "driven"

Similar to DE102021202321A1 this patent, also from ZF, describes a trailer that
includes an electric drive unit. The patent focuses more on the control methods of
the electric machine, how to know how much energy to deploy, and how to deploy
it.

DE102010042268A1
Published: 12-04-2012
Keywords: "ZF" or "trailer" or "propelled"

This patent from ZF describes a hydraulically driven trailer where each of the two
driven wheels are powered separately. This gives the possibility to control the torque
on each wheel resulting in less tire wear and added control of the trailer. The
hydraulic propulsion system can also be used as a generator which provides electrical
energy while braking.

EP4059755A1
Published: 11-03-2022
Keywords: "trailer dynamics"

Trailer Dynamics is as company developing E-Trailers. This patent describes how

their driven axle is packaged and how the power is transmitted from the electric
motor or motors to the wheels through gearboxes and driveshafts. It also describes
the layout of the powertrain. This results in a fairly compact powertrain which
becomes easier to integrate into a trailer.

19
3. Method

CN105691479A

Published: 22-06-2016
Keywords: "propelled" and "semi-trailer" and "semitrailer" and "trailer"

This patent describes the basics of an electrically propelled trailer towed by an

electric truck. It includes both battery mounting locations on the truck and trailer
as well as a trailer with either one or two electrically driven axles.

GB1385172A

Published: 26-02-1975
Keywords: "electric" and "semi-trailer" and "semitrailer" and "trailer"

A solution where each driven axle on both truck and trailer is powered by hub
motors. The generator is located on the truck and provides power to the trailer.
The patent mentions the benefit of the electric motors on the trailer when braking
due to them reducing the risk of "jack-knifing".

US11453292B2

Published: 27-09-2022
Keywords: "propelled trailer"

In this patent, a driven trailer is described. The method of propulsion covers hy-
draulic motors, pneumatic motors, and electric motors. It mentions that the driven
axle also acts as an electric brake when deceleration is needed. The patent aims to
solve driving scenarios where there is no space for both a truck and a trailer. Since
the trailer is smaller this patent wants the trailer to be equipped with a full pow-
ertrain that can therefore propel itself into smaller spaces. The patent also covers
how the transmission of power from the motor to the wheel will be carried out.

3.3 List of Requirements

The prestudy resulted in a list of requirements that the new E-Trailer should fulfill,
see appendix A. The list covered both general requirements such as ground clearance
and the number of axles on the trailer to specific scenarios where the E-Trailer was
required to meet certain performance targets. These scenarios and performance
targets were based on testing on the current E-Trailer together with discussion with
the thesis supervisors and other engineers at Volvo Trucks and CPAC. Some wishes
were also added to the list. These were features or functions that were not deemed
absolutely necessary for the E-Trailer to fulfill its main goal of lowering the energy
consumption from the ICE in the truck. All requirements and wishes were based on
information gathered in the prestudy.

20
 3. Method

3.3.1 Vehicle Dynamics

The following scenarios were used to set requirements on the output torque and
rotational speed of the powertrain. To get a measurable value on the scenario or
requirement some vehicle dynamics modeling and calculations were done. Below
are the scenarios and requirements listed and in the following subsections it is also
described how they were modeled and calculated.
Scenarios and requirements modeled and calculated:
• Being able to start/reverse straight with a base-loaded combination on level
ground up to 4 km/h.
• Not losing any speed when up-shifting gears by having the electric motor
support during the time it takes to shift.
• Contribute sufficient to power up to 90 km/h.
• Maximize the possible angle between the truck and trailer when reversing by
only using the E-Trailer.

3.3.1.1 Baseline truck and trailer

To ensure that the calculations and simulations resulted in realistic values, a baseline
truck and trailer were used. Since a truck and trailer can vary a lot in weight, axle
distances, power output, etc, it was decided to select a truck and trailer that was
representative of the routes and cargo hauled on the route and not change them be-
tween the calculations. The same truck and trailer were also used in the simulation
to get a baseline consumption on the route. This also allowed the simulation results
to only be isolated to the difference in the powertrain and not the combination.
Another benefit was also the fact that the results could be more representative of
how the different powertrains should perform in real life.

The base truck was based on a truck used in the Auto-Freight project at Volvo
Trucks. The truck was operating the route between the Gothenburg harbour and
Viared in Borås daily. This was a very well-documented route for fuel consumption
which made it easier to compare and validate the simulated values and by that also
get a better approximation of what an E-Trailer could perform on the same route.
The truck’s usual commission, to transport containers from Gothenburg harbour to
Viared, was also deemed as a realistic scenario where an E-Trailer could be used in
the future. In table 3.1 are some of the specifications for the truck listed. These
are mostly taken from the real truck using Transportstyrelsen [12], otherwise taken
from in-house reports.

Table 3.1: Base truck specifications.

Chassis 6*4
Axle distance, 1-2 axle 3600 [mm]
Axle distance, 2-3 axle 1370 [mm]
Weight 10000 [kg]
Engine d13, 375 [kW]
Gearbox Automatic, 12-speed

21
3. Method

The base trailer was decided to have the same axle distances as the first generation
of the E-Trailer. The weight was specified as the weight of the first-generation E-
Trailer without any powertrain components installed. In table 3.2 the specifications
for the base trailer are presented.

Table 3.2: Base truck specifications.

Chassis 3 axles
Axle distance, kingpin to 2 axle 7900 [mm]
Axle distance, 2-3 axle 1350 [mm]
Weight 6500 [kg]

3.3.1.2 Base cargo weight

Similar to the base truck and trailer a base cargo weight was also set to keep the
simulations and calculations comparable and was set to 15 tons. This was derived
partly from discussions with Emil Pettersson and Clive Misquith about their experi-
ence of what type of cargo weight a semi-truck and semi-trailer are transporting on
average. According to Lennart Cider, an engineer at Volvo Trucks, most transports
are volume-limited and not weight limited. It was also partly derived from data from
the Auto-Freight project of transports on the Gothenburg-Viared route. The data,
see "Autofreight Project Data" in Appendix F, covered the GCW of a duo-trailer
combination that partially also operated on the route with only one trailer. When
only one trailer was used it could be seen that the average GCW was 29 tons. From
this, the baseline GCW was set to be 31 tons to have some margin.
Using the baseline truck and trailer, mentioned earlier, a baseline cargo weight,
mcargo , could be calculated with equation 3.1.

mcargo = GCW − mtruck + mtrailer (3.1)

This resulted in a mcargo of 15 tons which were used in all simulations and calcula-
tions in the thesis.

3.3.1.3 Reverse straight

In this scenario, the E-Trailer was reversed in a straight line by only using the
trailer’s electric motor. The goal was to calculate how much traction force and
thereby how much torque the electric motor needed to provide to reverse the base
truck and trailer with the base cargo weight. Figure 3.4 shows a schematic picture
of the combination. Some simplifications and assumptions were also done to make
the calculations easier to carry out. The assumptions and simplifications for this
were:
• The three-axle trailer is seen as a single-axle trailer, but with the rolling re-
sistance of a three-axle trailer.
• The normal force of the trailer is acting on the middle axle of the trailer.
• The truck is a 6x4 configuration but the last axle pair is seen as one.

22
 3. Method

• The traction force of the truck is seen as one common force from the two driven
axles.
• The normal force on the driven axle pair is acting in the middle of the two
axles.
• The truck and trailer are viewed as one solid body when connected. No cou-
pling forces are taken into consideration.
• The truck and trailer are in steady state and no load transfer is taken into
consideration.
• The truck and trailer are standing on level ground.
• Both truck and trailer have a weight distribution rear/front of 50/50.
• Aerodynamic forces are not considered since we are considering the start-
ability of the combination.

Figure 3.4: Schematic view of the base truck and trailer combination with
measurements.

Figure 3.4 shows the majority of the measurements used during the calculations.
The measurements have the following values:

Table 3.3: Lengths of truck and trailer.

Constant Value Unit

h1cog 1 m
h2cog 1.5 m
hKP 1 m
Lf 3.6 m
La 1.37 m
L1cog 3.0425 m
L2cog 3.275 m
L3 7.9 m
lt 1.35 m

In the figure 3.5 the forces that are acting on the combination can be seen. The
truck has the mass of a standard 6x4 Volvo truck which weighs roughly 10000 kg.

23
3. Method

The mass of the trailer is divided into several masses since it is easier to compare the
E-Trailer to a conventional combination. The weight of the trailer itself is roughly
6500 kg, the weight of the powertrain is 2000 kg and the payload is 15000 kg. The
gradient of the slope αslope is 0°for the base calculations. αslope could, however, be
used iteratively to find the maximum gradient this maneuver could be performed in
with a specific powertrain.

Table 3.4: Constants and their values used for calculating the force needed to
start reversing straight.

Constant Value Unit

mtruck 10000 kg
mbasetrailer 6500 kg
mpayload 30000 kg
mpowertrain 2000 kg
αslope 0.0 °
cf 55/(g*1000) [-]

Figure 3.5: Free body diagram of the truck and trailer configuration, not
including kingpin forces.

From the free body diagram in figure 3.5 both a vertical and a horizontal equilibrium
equation can be extracted:

mtrailer = mbasetrailer + mpayload + mpowertrain [kg] (3.2)

55
cf = [N/N ] (3.3)
1000 ∗ g
Fy1 = mtruck ∗ g ∗ cos(αslope )[N ] (3.4)
Fy2 = mtrailer ∗ g ∗ cos(αslope )[N ] (3.5)
Fx1 = mtruck ∗ g ∗ sin(αslope )[N ] (3.6)
Fx2 = mtrailer ∗ g ∗ sin(αslope )[N ] (3.7)

24
 3. Method

Frrx = FN x ∗ cf [N ] (3.8)

FN 0 + FN 1 + FN 3 = Fy1 + Fy2 (3.9)

FT 1 + FT 3 = Frr0 + Frr1 + Frr3 + Fx1 + Fx2 (3.10)

With the help of the free body diagrams, in figures 3.6 and 3.7, the following equa-
tions were extracted:

FN 0 ∗ Lf = Fy1 ∗ L1cog (3.11)

FN 3 ∗ L3 = Fy2 ∗ (L3 − L2cog ) (3.12)

By doing a moment equilibrium equation of the truck around point A in figure 3.6,
equation 3.11 was derived. It was done at this point since the kingpin and traction
force would be excluded, making the calculations easier. The equation 3.12 was
acquired by doing a moment equilibrium equation around point B in figure 3.7. It
was done at this point to exclude both the traction force on the trailer and the
kingpin force.

Figure 3.6: Free body diagram of the truck.

25
3. Method

Figure 3.7: Free body diagram of the trailer.

To solve the system of equations consisting of 3.2 to 3.12 the split of traction force
from the truck and the trailer needed to be determined. To determine the minimum
traction force that is required from the trailer’s electrical powertrain the traction
force from the truck is set to zero.

3.3.1.4 Power shifting

The power shifting criteria is wanted because it would make it possible to perform
an up-shift without ever losing any velocity. This would make it possible to save
fuel since there is no need of accelerating back up to the velocity from where the
shift started. A secondary benefit is that this would enable a smoother shifting ex-
perience for the driver since the driver would not feel any difference in acceleration.
Power shifting could possibly increase the life of components within the powertrain.
The power shifting criteria set a demand on what peak power the electrical motor
can output. It sets a demand on the peak rather than the continuous power because
a shift to a higher gear never exceeds 10 seconds (a very small duration).

To calculate the energy that was lost by changing gear the existing energy of the
vehicle combination had to be calculated. Figure 3.8 shows the forces that act on
the vehicle combination during a gear shift in a slope. Equation 3.13 is the total
energy the vehicle combination has before the gear change. There is no potential
energy due to the height reference point being set when the gear shift starts. One of
three possible losses is the aerodynamic drag, the force produced by the aerodynamic
drag is described in equation 3.14. The second loss is the rolling resistance, the force
created by the rolling resistance is estimated by the rolling resistance coefficient cf
and the total load that acts on the tires. The force created by the rolling resistance
can be seen in equation 3.15. To ensure that equation 3.19 was as accurate as
possible the addition of the truck’s inertia was added. Equation 3.16 shows a simple
expression of how the inertia of the truck can be calculated. In order to validate
equation 3.19, the variable ashif t was determined from iterative testing. This was

26
 3. Method

done by comparing real-life data of how long it takes to shift gears in a truck with a
trailer attached and what velocity it had after the up-shift. The velocity the vehicle
combination should have after the up-shift is described by equation 3.18. Equation
3.19 overestimates the energy lost compared to the real data due to the magnitude
of the variable ashif t , which acts as a safety factor. Equation 3.20 then calculates
the peak power needed from the electric motor in order to "powershift". The power
needed with regards to different angles of the slope and different masses, the values
of αslope and mtrailer used are shown in table 3.6.

Figure 3.8: Forces that act on the combination while shifting in a slope.

fig

2
EKE1 = (mtrailer + mtruck ) ∗ Vstart [J] (3.13)
2
cd ∗ Af rontal ∗ ρair ∗ Vstart
Faero = [N ] (3.14)
2
Frr = cf ∗ (FN 3 + FN 1 + FN 0 )[N ] (3.15)

Finertia = (mtrailer + mtruck ) ∗ ashif t [N ] (3.16)

EP Elost = (mtruck + mtrailer ) ∗ g ∗ Vstart ∗ tshif ting ∗ tan(αslope )[W s] (3.17)

s
EKE1 − ELost
Vaf ter = [m/s] (3.18)
mtruck + mtrailer

ELost = EP Elost + Vstart ∗ tshif ting ∗ (Frr + Faero − Finertia )[W s] (3.19)

Elost
PneededEM = [W ] (3.20)
tshif ting

27
3. Method

Table 3.5: Values used for calculating the energy lost during an up-shift.

Constant Value Unit

mtruck 10000 kg
mtrailer 38000 kg
cd 0.8 [-]
cf 55/(g*1000) [-]
Af rontal 10 m2
ρair 1.293 kg/m−3
Vstart 19.5 m/s
ashif t 0.3 m/s2

Table 3.6: Values of αslope and mtrailer used for testing.

αslope 0 1 2 3 4 5
mtrailer 10 15 20 30 40 50

3.3.1.5 Reverse with angled truck

The end scenario of reversing with the E-Trailer is to replace the usage of the diesel
engine when arriving at terminals. To make sure that the electric machine can be
used for even longer periods of time (compared to straight reverse) the decision was
made to calculate how much more torque would be needed to enable corrections
while reversing (steering maneuvers).

Calculation
In order to turn the combination there has to be an angle between the trailer and the
truck, this is the articulation angle α, and a steering wheel angle β. Because of the
angles β and α in figure 3.9 there will be a force component acting as friction on the
wheels of the truck. This will increase the total force needed compared to a straight
reverse. This friction force’s magnitude is particularly hard to estimate because it
requires deep knowledge about how tires work and also depends on the angles α and
β, loads, and velocities. The resulting force due to this friction force and its angle θ
is visualized in figure 3.10. Because no data on the tires were available, the different
θ angles were changed iteratively in the equation 3.21 with the values seen in the
tables 3.7 and 3.8.

Thub = rw ∗(cf ∗(FN 0 +FN 1 +FN 3 )+FN 1 ∗µ∗sin(θrear )+FN 0 ∗µ∗sin(θf ront )) (3.21)

Table 3.7: Angles used for iterative testing.

θf ront [deg] 0 1 0 1 2 1 2 3 2 3 4 3 4 5 4 5
θrear [deg] 0 0 1 1 1 2 2 2 3 3 3 4 4 4 5 5

28
 3. Method

Table 3.8: Constants and their values used for calculating the force needed to
reverse with the truck in a angle.

Constant Value Unit

mtruck 10000 kg
mbasetrailer 6000 kg
mpayload 30000 kg
mpowertrain 2000 kg
cf 55/(g*1000) [-]
µ 0.8 [-]
rw 0.5 [m]

Figure 3.9: Free body diagram of the vehicle combination doing a angled reverse

Figure 3.10: Free Body diagram of a tire in the angled reverse scenario

29
3. Method

Simulation
The angled reverse was also simulated in a Volvo tool called VTM with the help of
Niklas Fröjd. A meeting with Niklas was set up to discuss the case of reversing with
an angle and having the trailer pull the truck instead of having the truck pushing
the trailer. The simulations were done using the base trailer. The simulation starts
with the wheels being straight as well as the entire combination. The trailer then
starts to pull the truck, while the truck starts doing the normal reversing maneuvers.
The simulations, therefore, tested different steering angles and articulation angles
and resulted in different torques on the driven trailer axle.

Testing
The case of reversing with articulation and a wheel angle was also tested with the
first-generation E-Trailer. This was done by driving the truck and the first gen-
eration E-Trailer into different articulation angles while having the steering wheel
straight (same as the articulation angle). From this position, the E-Trailer at-
tempted to reverse the entire combination.

3.3.1.6 Sufficient power at 90 km/h

To make it possible for the electrical powertrain to support the diesel engine at all
times it is vital to make sure that the powertrain is able to deliver power at any
velocity of the vehicle combination. To make sure this was the case equation 3.22
was used. All of the gear ratios (from the motor output shaft to the hub) need to
be known as well as the motor’s maximal rotational speed. The minimal velocity
the powertrain has to be able to output is 80 kph since it is the maximal velocity a
heavy truck is allowed to drive in Sweden.

velocity = ωmotor /(itotal ∗ 2 ∗ π ∗ rw ) (3.22)

3.3.2 Other requirements

The remaining requirements and wishes on the list were based on information gath-
ered from discussions with the people who were involved in the building of the
first-generation E-Trailer.

3.4 Simulations & Dimensioning of Components

A simulation program developed in-house by Per Björe at CPAC was used as the
base tool to simulate different powertrain configurations and routes. Some modifica-
tions were made to the program to make it easier to change variables and analyze the
result, but the base logic and calculations were kept the same. The program returns
estimated results of saved carbon dioxide and fuel compared to a regular truck and
trailer combination. For this thesis, the amount of saved fuel, or efficiency gained,
was used as the value to compare the different motor and battery configurations.
The choice of comparing saved fuel was due to it being the main goal for Volvo

30
 3. Method

Trucks with the E-Trailer concept. Therefore the simulation program was created
to have it as one of its outputs, making the comparison easy to do. Estimated
profitability on the route was also deduced from the results form the simulations.
This was done to evaluate if there was any potential money to be saved with the
use of an E-Trailer. In the case that many powertrain combinations had similar fuel
savings, the profitability was used as a second value to decide which combination
was the most beneficial on a specific route.

The simulations were done in two main stages; the sweep simulations and the final
simulations. Both types will be described more in-depth in section 3.4.3 and section
3.4.6 respectively. In short, the sweep simulations were done with a fixed set of
motor sizes and a fixed set of battery sizes. All combinations were simulated on
all routes and the motors and batteries had no connection to existing components.
The goal here was only to see if there were any trends in what type of combinations
were more beneficial on a route or if there were any other trends to be seen. The
trends and results found from these sweep simulations were then used as a guide
for selecting an existing motor and battery that would be the most optimal for a
selected route. These selected components were then to be simulated to estimate
how much fuel that could be saved on this route with this final powertrain.

3.4.1 Simulation Logic Description

In the following section, the basic logic of the simulation will be described as well
as some comments and limitations that the set logic causes.

Truck and Trailer Propulsion Separate

The simulation tool was built in a way that the trailer only propels its own weight
and never helps propelling the truck. This is because of safety reasons but also a
big limitation of the program. An example when this matters is when the trailer
is unloaded, this means that the motor only propels a small portion of what it is
capable of. This results in that a powerful motor is not going to save any more
energy than a less powerful motor, considering they both have enough energy from
the battery. This could be optimised further to decrease the fuel consumption.

Deployment Strategy

The deployment strategy used is very basic and can be improved immensely. It
works is by setting a target velocity the vehicle combination will deploy energy to
maintain it. If the battery has energy stored and the vehicle is in a short uphill we
can deploy peak power from the motor, otherwise it outputs continuous power. If
there is no energy in the battery the motor will only act as a generator until there
is enough energy to start acting as a motor again. The rest of the required energy
has to come from the truck’s diesel engine.

31
3. Method

Mass Never Unloaded

Transportation of goods are usually transported from one place and then offloaded.
The truck can pick up other cargo and bring it back to where it started In the
simulation tool, there is no way of changing the mass during the drive cycle making
the scenario simulated less accurate to reality.

Battery/Motor power limit

The program also takes into account what component will be the limiting factor. For
example, if the battery has a lower continuous power output than the motor it will
limit the total output to the battery’s continuous output. This impacted the way
the sweeping simulations were done. Since the max/continuous power output from
the motor was a parameter that affected the results heavily the max/continuous
power from the battery also had to be changed to match the motors.

3.4.2 Simulation Tool Validation

To see how accurate the simulation tool was, a validation study was conducted.
The validation was done by comparing both a simulation of a truck and trailer com-
bination without an electrical powertrain and by simulation of the first-generation
E-Trailer on a route that has measurements from the real vehicles available.

3.4.2.1 Diesel Truck Comparison

The first comparison was to compare the results from the code with the data col-
lected from the Autofreight vehicle. The Autofreight vehicle does not have an E-
Trailer which means that this comparison is only for the truck’s diesel engine and
weight.
The data collected from the Autofreight vehicle is shown in "Autofreight Project
Data" in Appendix F. The data includes which route, all the axle loads, container
weights, distance traveled, weather, and fuel consumption. The simulations were
then conducted with a standard trailer.

Table 3.9: Values used in the simulation tool to validate the diesel consumption
of a conventional truck.

Attempt Truck mass Trailer mass

1 10 [ton] 17 [ton]
2 10 [ton] 42 [ton]
3 10 [ton] 25 [ton]

3.4.2.2 E-Trailer Comparison

The other comparison was to simulate the powertrain part and compare it to real
data collected from testing of the first-generation trailer. To be able to do this the
data had to come from where the E-Trailer already had been tested, which means

32
 3. Method

that the elevation profile has to be taken from Hällered proving ground. This route
at Hällered had already been integrated into a file by Per Björe at CPAC. The
weight and target speed were transferred to the simulation tool and the simulations
were run with the powertrain settings of the first-generation E-Trailer. The masses,
SoC, and target velocity that was used can be seen in table 3.10. The simulations
were also run with the correct masses but without the electric powertrain turned
off in order to get a baseline. The fuel consumption with the electric powertrain
turned on was then compared to the baseline result. The validation data for the
fuel savings were taken from a test of the first-generation E-Trailer. This test was
done by driving 9 laps around the test track at Hällered. 3 laps were done with a
higher SoC, 3 without the powertrain activated, and 3 with the SoC being at the
lowest limit. The three laps with the powertrain deactivated resulted in average fuel
consumption of 3.87L diesel. All of the real tests of the powertrain on the E-Trailer
are compared to this average consumption.

Table 3.10: Values used in the simulation tool to validate PT part of code.

Attempt Truck mass Trailer mass Initial SoC Target velocity

1 10 [ton] 37 [ton] 37% 70 [km/h]
2 10 [ton] 37 [ton] 32% 70 [km/h]
3 10 [ton] 37 [ton] 28% 70 [km/h]

3.4.3 Sweep simulations

This section of simulations were done by "sweeping" over a selected range of battery
capacities and motor sizes. It was done by selecting one motor size and then changing
the battery capacity, "sweeping" over the batteries, for each simulation. When all
battery capacities were simulated with that motor, the motor was changed and the
process was redone. This ensured that all combinations of motor torque and battery
capacities were methodically simulated. Only these two parameters were changed
between the simulation to keep the combinations to a minimum and also to make
the result easier to analyze.
The selected motor sizes and battery capacities were, as mentioned earlier, not
related to any existing components. The range of motor sizes ranged from 250 Nm
of peak torque to 2500 Nm of peak torque. The batteries ranged from 50 kWh to
2000 kWh. In table 3.11 the full list of motor sizes is presented. To avoid limiting
the motors, by having lower power ratings from the battery, the battery’s peak, and
continuous power were changed to the same as the motor’s.

33
3. Method

Table 3.11: Selection of motors that were used in the sweep simulations.

Peak torque [Nm] Continuous torque [Nm]

250 250
500 300
750 400
1000 600
1500 800
2500 1400

3.4.3.1 Parameter Determination

In order to make the simulation results more realistic parameters needed to run
the simulations had to be derived. The following section will describe how this was
done.

Base Trailer Weights

A conventional trailer is neither equipped with an electrical powertrain, heavy-duty

drive axles nor a second chassis. This means that there is a lot of added mass,
which is not only coming from the electrical powertrain. To make the simulations
as realistic as possible the E-Trailer was compared to a conventional semi-trailer
and not the E-Trailer without its powertrain. A conventional container semi-trailer
weighs approximately 5 tons while the first generation E-Trailer without its power-
train weighs roughly 6.5 tons according to internal reports. To find the difference,
a simulation was done to compare the different masses. The simulation showed
that an additional 50 kWh was needed to compensate for the E-Trailer’s additional
weight on the route that goes from Gothenburg to Borås three times, which can be
seen in figure 3.11.

34
 3. Method

Figure 3.11: Fuel consumption comparison of E-trailer weight and a conventional

trailer when loaded with 15 ton payload

Powertrain Weights
With higher battery capacities and motor torques also comes a higher weight. To
take this into consideration a ratio between capacity and weight as well as between
peak power and mass was calculated. To calculate the battery ratio, values of ca-
pacity and weight from an existing battery currently used in the Volvo group were
used. The same was done for the motor ratio. Here the motor is one of the ones cur-
rently used in electric Volvo trucks. In table 3.12 the values and ratios are presented.

Table 3.12: Ratios for calculating the mass of motors and batteries depending on
peak power and capacity.

Mass [kg] kWh or kW Ratio, i

Battery 563 66 8.5 [kg/kWh]
Motor 87 160 0.54 [kg/kW]

These ratios were then used to calculate the weight of the E-Trailer in each simu-
lation to determine how efficient each combination was. The weight was calculated
by using equation 3.23. The weight of the trailer, mtrailer , and the mass of auxiliary
components, maux , were estimated from the current E-Trailer. These corrections
mainly were done to the weight of the auxiliary components such as; cooling, in-
verters, and mounting and were done consent with the thesis supervisor from Afry,

35
3. Method

Clive Misquith, who is involved in the E-trailer project as well. The values used are
shown in the tables 3.12 and 3.13.

mE−T railer = mtrailer + maux + (imotor ∗ Ppeak ) + (ibattery ∗ Ebattery ) (3.23)

Table 3.13: Values of constants used to calculate trailer weight.

Constant Value Unit

mtrailer 6500 kg
maux 250 kg

Power Scaling from wanted Torque

The gear reduction built into the motors was not changed during the sweep simu-
lations. This variable is useful if the rpm range of the electrical motor is too large
and needs to be reduced to fit existing final gear ratios. However, for the sweep
simulations, the max rpm was set constant to replicate a diesel engine.
The different motors were all scaled from the wanted torque output and the rpm
range was derived from the final drive ratio and wanted velocity. The max/con-
tinuous power output was calculated from the wanted torque and estimated rpm
range.

Initial SoC & SoC window

After some initial simulations, it became apparent that the SoC window and the
initial SoC affected the simulation results heavily. For all simulations, the SoC
window was set to 30%-70% and the initial SoC was set to 50% of the battery size.
This will affect the sweeping simulations since the larger capacity batteries will have
more initial energy and therefore an advantage. A fixed value would mean that all
benefits of a bigger battery would not be considered unless the max energy of the
battery would be reached during regenerative braking.

3.4.3.2 Route selection

To gather data on how different powertrains perform depending on the route, three
different routes were chosen. One route had an aggressive elevation profile, one had
a relatively flat profile, and one had a profile somewhere in between the two previous
routes.
To simulate a full day of driving equation 3.24 was used. The maximum allowed
time to drive each day is 9 hours [11] and with a maximum velocity of 80 km/h the
maximum distance traveled in a day will never exceed 720 km.

dmax = tmax ∗ vmax (3.24)

Tampere to Helsinki
The driving cycle chosen to simulate a quite flat elevation profile is the route between
Tampere and Helsinki. This route has an elevation difference of roughly 140m. This

36
 3. Method

is not very flat, but the big elevation difference is located close to Helsinki and is
very local, see figure 3.12. The driving cycle is both starting and ending in Tampere
(roughly 360 km), this is because in most cases the haulage contractors have a
location where they park their trucks. Starting and ending the route at the same
place also makes sure there is no natural gain of potential energy. The highest
elevation the route reaches is almost 140m and the lowest is sea level.

Figure 3.12: Altitude profile of the route Tampere to Helsinki to Tampere

2 × Tampere to Helsinki

To simulate a full day of driving the normal Tampere-Helsinki route was done twice,
resulting in 718km. This distance is barely possible on a normal day but can be made
possible through extra time according to Trafikverket [11]. The elevation profile is
shown in figure 3.13.

Figure 3.13: Altitude profile of the route 2x Tampere to Helsinki to Tampere

37
3. Method

Kurtalan to Bahçesaray to Cizre

These cities are located in the Taurus mountains in the southeast of Turkey and
were chosen in order to simulate how different powertrains would perform in hilly
conditions. The cumulative height gain of the route was almost 11000 m and can
be seen in figure 3.14 while the cumulative height loss was a bit more than 11000 m.
During this route there is some potential energy to gain, it is however considered
neglectable due to the characteristics of the route. The total distance of this route
is roughly 450 km and the elevation differs between 400-2500 m .

Figure 3.14: Altitude profile of the route from Kurtalan to Bahçesaray to Cizre

Kurtalan to Bahçesaray to Cizre to Kurtalan

In order to simulate a full day of driving the initial route in Turkey was extended
by going back to Kurtalan. This resulted in the trip being a total of 610km with
the elevation profile shown in figure 3.15. It also results in the route not gaining
any potential energy.

38
 3. Method

Figure 3.15: Altitude profile of the route from Kurtalan to Bahçesaray to Cizre
to Kurtalan

Gothenburg to Viared to Gothenburg

The third route was between the Harbour in Gothenburg and the industrial area
Viared which is located outside of Borås. This route is somewhere in between the
routes in Turkey and Finland. The elevation stretches from sea level to roughly
200m above and is almost 150 km long. The elevation profile can be seen in figure
3.16 and has a cumulative height gain of 1000m.

Figure 3.16: Altitude profile of the route Göteborg to Viared to Göteborg

3 × Gothenburg to Viared to Gothenburg

The trip from Gothenburg harbor to Viared can be made three times each day
according to the Autofreight project at Volvo Trucks Technology [6]. This can be
seen in the elevation profile shown in figure 3.17. Doing this trip three times a day
extends the distance traveled to 430km.

39
3. Method

Figure 3.17: Altitude profile of the route 3x Göteborg to Viared to Göteborg

3.4.4 Powertrain Comparisons

As mentioned earlier, the main goal of the E-Trailer is to save fuel for an ICE-
powered truck and therefore the fuel saved or increased fuel efficiency was mainly
compared. The increased efficiency was calculated with equation 3.25, where the
Econvetional is the energy required from the ICE to drive the route with a standard
ICE truck and trailer and EE−T railer is the required energy from the ICE to drive
the same route, but with the assist of an E-Trailer. Both values were calculated by
the simulation program and measured in kWh.

Econvetional
(1 − ) · 100 (3.25)
EE−T railer
A higher percentage indicated more diesel saved and potentially a better combi-
nation. However, since these batteries and motors were not based on existing or
near-future components some of these combinations were not realistic. To filter
these out and end up with more realistic combinations some other aspects were
taken into account.

Physical limitations
To narrow down the results even further, a maximum battery capacity of 750 kWh
was set. This was based on what Volvo currently have in their trucks. A similar
approach was taken to determine the maximum motor torque. A solution with
three motors is currently in use for Volvo Trucks’ electrical truck, the FM electric,
with a torque output of 1200 Nm. With this information, it was determined that
a realistic, maximum, motor torque would be 1500 Nm. The simulation results for
higher capacity batteries and higher torque motors were still used to identify trends
and to observe if any of the trends dropped off if the motor or battery capacity
got bigger. The results could also be used as pointers to what could be possible to
achieve in the future if the battery and electric motor technology becomes better
and more affordable.

40
 3. Method

Profitability
By taking the diesel and electricity costs into account it was possible to calculate if a
combination was also saving money and therefore could be seen as beneficial in that
aspect as well. This was done by calculating the cost of the diesel required to drive
the route with a conventional truck and trailer and comparing it to the diesel and
electricity cost required to drive with an E-Trailer on the same route. Equation 3.26
was used to calculate the cost of the consumed diesel for the conventional truck and
trailer. Equation 3.27 was used to calculate the cost of charging the E-Trailer and
using equation 3.28 the combined cost of driving an ICE truck with an E-Trailer.
The difference in cost was then calculated using equation 3.29.

Since these prices always fluctuate, some limitations had to be set to get a fixed
value that represents reality as well as possible. The price for the diesel was derived
from taking the diesel price of the day from the closest gas station in Sweden and
subtracting 25% for taxes.

According to InCharge [15] the cost for electricity, grid, and taxes results in a cost
between 2-4 SEK/kWh depending on which company supplies the electricity, in what
area of the country you are located, what time of day the charging takes place and
what electricity contract you have. Since the price of charging varied significantly
and had a big impact on the profitability of the E-Trailer a sensitivity analysis was
carried out. The analysis was done after the sweep simulations were done to select
a powertrain configuration that would be realistic, one for each route. The sweep
simulations were all done with a charging price of 2 SEK/kWh. All three routes
were included in the analysis and it was done by changing the charging price from
1-5 SEK/kWh and seeing how it affected the profitability on the selected powertrain
and the routes. It was also assumed that the charging also only took place at night
since the prices were lower and it also represented a realistic case, where the night
was usually the only time a trailer was standing still for a longer amount of time
and therefore could be charged a considerable amount. In table 3.14 the derived
prices for diesel and electricity can be found.
Costdiesel
Costconvetional = Econvetnional (3.26)
10

Echarged = Ebattery−capacity (Bwindow−high − Bwindow−low )SoCinitial (3.27)

Costdiesel
CostE−T railer = EE−T railer + Costelectricity Echarged (3.28)
10

Costdif f erence = Costconvetional − CostE−T railer (3.29)

With the help of these limitations, it was possible to narrow down which motor
and battery would be the most beneficial powertrain to reduce diesel consumption
without costing too much money and be realistic in the near future. The process of
deriving these powertrains was initiated by finding which powertrain that was within
the mentioned limitations. This was done using an if-statement in Excel. It checked

41
3. Method

Table 3.14: Prices used to calculate energy costs.

Source Price Unit

Diesel 16.5 SEK/liter excl. tax
Electricity 2 SEK/kWh excl. tax

if a powertrain’s Costdif f erence ≥ 0 and if its battery capacity and motor torque
were within the specified limits. From the remaining powertrains, that fulfilled
these requirements, the one which had the greatest fuel savings was selected as the
best possible powertrain for that specific route. This was done for each route to
determine the most optimal powertrain for each.

3.4.5 Final Powertrain

In this subsection, the setup for the simulation of the final powertrain will be de-
scribed. It will cover the route selection, powertrain parameters, and other param-
eters used in the simulations.

3.4.5.1 Selection of components

The component selection was based on the simulations for the selected route, 3 ×
Gothenburg-Viared-Gothenbrug, and narrowed down using the methods described
earlier. This route was selected since Volvo Trucks already has field vehicles operat-
ing it daily, making it well known and also well covered when it comes to reference
data. The optimal configuration specifications were then as closely matched as pos-
sible with in-house components, motors, and batteries, available at Volvo Group.

3.4.5.2 Final Powertrain Simulations

With the components selected, final simulations were carried out. The setup for
these simulations differed from the earlier, sweep simulations, in the fact that pa-
rameters that were previously unknown, and therefore left unchanged, now could
be changed due to real components being used. Which parameters were changed
depended on if they could be obtained or were described in the technical docu-
mentation related to the components. For the parameters that were not found the
previous values were used.
In these simulations, the initial SoC, payload, and routes were also altered to see
how different scenarios affected the performance of the powertrain.

Powertrain Parameters
To find the data needed to start the final simulations the datasheets for both the
battery and motors were used.

In the datasheet for the batteries, it was concluded that the peak power output is
173 kW and that the continuous power output is 91 kW.

42
 3. Method

In the datasheet for the motors, there is a constant for both the peak power and
the continuous power. Where the peak power is 150 kW for 150 seconds and for 10
seconds it is 160 kW. A mean value of these two where selected (155 kW). The motors
can deliver 105 kW continuously. The datasheet does not give the torque/power/rpm
curves of the motor. The motors can deliver 266 Nm continuously for 0-38% of the
speed range. This results in the motor’s starts to field weakening in order to increase
the rpm at this point (3800 rpm), which then gives us the graphs needed.

3.5 Concept ideas

The main focus of the design and packaging aspect of the thesis included the axle
configuration of the E-Trailer and the powertrain packaging. The specified require-
ments were the foundation of the concept and had a big influence on how the final
design was done.

3.5.1 Axle configuration

From the list of requirements, it was stated that the E-Trailer required a three-axle
combination. In the market analysis, it was also shown that a three-axle combina-
tion is the most common axle configuration for these types of trailers.

There were several different variants within the three-axle configuration, not includ-
ing the driven axle, as well. Each with its own benefits and drawbacks. The main
variants of axles are;
• Liftable
• Steerable
– Self steered
– Hydraulically steered
• Liftable and steerable
A conventional trailer usually does not have these variants, due to the increased
price of adding such features. However, it is becoming more common to have a
first liftable axle on the trailer to decrease rolling resistance and fuel consumption
according to L. Cider [2]. This together with the wish of having both steerable and
liftable axles on the E-Trailer, led to the decision to implement both liftable and
steerable axles in the concept.

3.5.1.1 Choice of driven axle

The choice of which axle would be the most beneficial to have driven was derived
from which axles would be preferred to have liftable. Inspiration was taken from
four-axle trucks and how they usually are configured as well as the reasoning behind
why such a configuration was the most beneficial in terms of weight distribution and
drivability. The following reasoning was based on the criteria that two of the three
axles were liftable on the trailer to fulfill the wishes from the list of requirements.

43
3. Method

Figure 3.18: Truck and trailer combination with second and third axle lifted.

If the first axle was unliftable, the overhang of the trailer, with the two last axles
lifted, would have been too large. This overhang would lower the kingpin pressure
by "lifting" the trailer at the kingpin. By lowering the kingpin pressure, the axle
pressure on the trailer will increase, potentially becoming too much, resulting in a
lower cargo weight. The lower king pin pressure also decreases the axle pressure
on the driven axle/axles on the truck decreasing the traction force and with that
a decreased cargo weight. For the drivability of such a combination, it would be
easy to maneuver at lower speeds due to its shorter wheelbase. However, it could
be unstable at higher speeds. Since these trailers mostly are driven on highways at
high speeds an unstable combination was not acceptable. In figure 3.18 a schematic
figure of this scenario is shown.

In the case of the third axle being the unliftable the problem with weight distribu-
tion becomes the opposite with an increased kingpin pressure due to there being too
little overhang. This could also result in a lower maximal cargo weight. When it
came to the drivability of such a combination, the combination would have a longer
wheelbase making it stable at higher speeds, which was preferable, but more diffi-
cult to maneuver at lower speed. In figure 3.19 a schematic figure of this scenario is
shown.

The last case of having the second axle unliftable became a compromise between the
two earlier cases. It created a smaller overhang, decreasing the kingpin pressure, as
well as moving more weight in front of the balancing point, by lifting the first axle,
increasing the kingpin pressure. In figure 3.20 a schematic figure of this scenario is
shown.

44
 3. Method

Figure 3.19: Truck and trailer combination with first and second axle lifted.

Figure 3.20: Truck and trailer combination with first and third axle lifted.

45
3. Method

Figure 3.21: Free body diagram of the trailer with no axles lifted.

This unchanged kingpin pressure was also tested by calculating the difference in
the kingpin pressure between having all three wheels on the ground compared to
having the first and third axles lifted. This was done by making a free-body diagram
of the trailer with all three wheels on the ground, figure 3.21. From the free body
diagram two equilibrium equations were derived, equation 3.30 and 3.31. From these
equations, the kingpin force, Fkp , was solved. The same was done for the case of
having the first and third axles lifted. Figure 3.22 shows the free body diagram of
that case and equation E.2 and E.2 was the derived equations from that free body
diagram. The same was also done for the first and second cases described above.
The free body diagram and equations for those cases can be found in appendix E.

F1 l1 F2 l2 l3 l4
− + + F3 (l2 + ) + F4 (l2 + l3 + ) − Fkp (l2 + l3 + l4 ) − FN l2 − FN (l2 + l3 ) = 0
2 2 2 2
(3.30)

3FN + Fkp − (F1 + F2 + F3 + F4 ) = 0 (3.31)

F1 l1 F2 l2 l3 l4
− + + F3 (l2 + ) + F4 (l2 + l3 + ) − Fkp (l2 + l3 + l4 ) − FN l2 = 0 (3.32)
2 2 2 2

FN + Fkp − (F1 + F2 + F3 + F4 ) = 0 (3.33)

In both cases, it was assumed that the load on the trailer was equally distributed
over the trailer and that the three axles also had equal load distribution. Equation
3.34 was used to calculate the distributed load coefficient, q. Axle distances and pay-
load were taken from the base trailer and calculated the payload used earlier. Table
3.15 shows the values of the constants used in the calculations. The moment was

46
 3. Method

Figure 3.22: Free body diagram of the trailer with first and second axle lifted.

Table 3.15: Values of base trailer constants used to calculate the kingpin forces,
Fkp .

Constant Explanation Value Unit

l1 distance from end of trailer to third axle 1.675 m
l2 distance from third to second axle 1.35 m
l3 distance from second to first axle 1.35 m
l4 distance from first axle to kingpin 7.05 m
mpayload mass of payload on the trailer 15000 kg

47
3. Method

taken around the contact patch of the third axles tire, marked with "A" in the figures.

mpayload g
q= [N/m] (3.34)
l1 + 2l2 + l3
Using the distributed load coefficient, q, the forces F1−4 was calculated using equa-
tions 3.35 - 3.38.

ql1
F1 = (3.35)
2
l2
F2 = q(l1 + ) (3.36)
2
3l2
F3 = q(l1 + ) (3.37)
2
l3
F4 = q(l1 + 2l2 + ) (3.38)
2
With all forces and distances known, the kingpin pressure, Fkp , for both cases were
solved for and compared using MATLAB.

Having the second axle unliftable, and thereby driven, also brings another benefit
when it comes to maneuverability with all axles down. The resulting rotating center
for the three axles ends up in the second axle. Having the second axle being the
driven axle helps the rotation of the trailer since the driving force is acting on the
same axles as the trailer rotates about. If the first axle would have been driven,
the axle would have pulled the trailer straight instead of helping the rotation. The
same logic applies if the third axle would have been driven, but instead of pulling
it straight, it would have pushed it straight. For 8x2 trucks, which also have a
three-axle combination in the rear, that operates on roads in good condition it is
common to have this axle configuration.

3.5.2 Powertrain layout

From the market analysis, it was clear that the powertrain of an E-Trailer could
be designed in many different ways and that every solution had its pros and cons.
Because of the list of requirements, some concepts could be ruled out. This was
mainly because the thesis aims to only use existing Volvo Group components.

The use of a propeller shaft was looked into. It was due to that current electric
Volvo trucks were using it. The components available are designed to be used with
a propeller shaft. A propeller shaft was also beneficial since it moved the motors
and eventual gearboxes away from the axles, where there is space for them.

Since the axles can move up and down, due to the suspension and road surface, as
well as be lifted it was important to ensure that the propeller shaft and the pusher
axle would not clash in any scenario. To test this, a worst-case scenario was set

48
 3. Method

up. If the propeller shaft cleared the pusher axle it was deemed to work in any
scenario. The test setup was inspired by a similar test that was conducted in Emil
Olsson’s report regarding the packaging and design of the first E-Trailer. In the
scenario where the first axle, the pusher, is lifted up and the second, driven axle,
was dropped. These values used were the maximum bump and droop distances that
the axles were designed for. By lowering the driven axle the propeller shaft angle,
α, decreases. This also decreased the clearance to the pusher axle and by having
the pusher axle lifted, the clearance became the smallest possible, making it the
worst-case scenario. In figure 3.23 this decrease of α is visualized. A normal ride
height scenario, with both axles at ride height, and a scenario with a full bump on
the driven axle and ride height on the pusher axle were also carried out to ensure
no unexpected clashes would appear. The change in travel from normal ride height
for each axle and scenario can be found in table 4.20

(a) The angle α during normal ride height.

(b) The angle α during full droop on the driven axle.

Figure 3.23: Propeller shaft angle, α, is dependent of the vertical position of the
axles.

49
3. Method

Table 3.16: Positive distance indicates the compression of the suspension and the
axle moves up. A negative indicates the opposite and the axle moves down.

Scenario First axle [mm] Second axle [mm] Explanation

1 90 -178.5 Worst case scenario
2 0 0 Normal ride height
3 0 90 Full bump on the second axle

Each scenario was tested by positioning the axles in CAD with the gearbox and the
propeller shaft in their respective positions. A drawing was then created to measure
clearance distances between the propeller shaft and the axle as well as to ensure
that the angles and distances of the propeller shaft were within the specified ranges.
The gearbox and motor placement were then changed in order to obtain the right
angles and distances as well as to prevent clashing.

Since the propeller shaft moves with the driven axle its length varies. By testing
these scenarios it also showed how much this movement affected the length of the
propeller shaft. Due to an offset between the gearbox flange and the driven axle,
the y-distance, and the height difference between the two, the z-distance and the
length of the propeller shaft, lprop , were needed to be calculated in 3D. To calculate
the length Pythagoras theorem in 3D was used, see equation 3.39. In figure 3.24 the
required distances to do the calculations are shown. To get both the full range of
lengths the propeller shaft could encounter the distances were measured in all three
cases described earlier in this section. Then the smallest and largest measurement
was taken to be the extremes for the propeller shaft length.

q
lprop = x2 + y 2 + z 2 (3.39)

Emil’s report stated some important limits of the propeller shaft. First, the maxi-
mum length of the propeller shaft should not exceed 2100 mm. Secondly, the working
range of the universal joints on the propeller shaft was between 1-6◦ with the axles
in normal ride height. This meant that the propeller shaft angle could exceed this
range in shorter periods of time if necessary. However, if the angle were to exceed
this range continuously the expected lifetime would decrease. As explained earlier
the propeller shaft operates in the 3D-space and there are two angles to take into
account, the α-angle described earlier, and the angle β shown in figure 3.25. The
latter only needed to be checked once since it does not change regardless of the
scenario. This is due to that the y-distance never changes since the suspension does
not move in the lateral direction while in use. Because of these requirements, the
motor/gearbox was placed and packaged first, since they largely affect the propeller
shaft’s angles and length.

50
 3. Method

(a) The x and y distances was measured from the top view.

(b) The z distance was measured from the side view.

Figure 3.24: Side and top view of the distances needed to calculate the propeller
shaft length.

51
3. Method

Figure 3.25: Propeller shaft angle, β, is not dependent on the vertical position of
the axles.

The scenarios were tested using CAD. The axles were put in position together with
the gearbox and the propeller shaft in their respective positions. A drawing was
then created to measure clearance distances between the propeller shaft and the
axle as well as to make sure that the angles and distances of the propeller shaft were
within the specified ranges. Gearbox and motor placement were then changed until
there was no clashing of the axles or propeller shaft.

The placement of the batteries was decided based on where space in the chassis was
available as well as how they would affect the center of mass of the trailer. Since
the batteries were the heaviest component their placement could affect both the
drivability and loading capacity of the trailer. Looking at competitors and where
they placed their batteries as well as where they were located on the first generation
E-Trailer, a possible solution was to place them between the first axle and the
support legs of the trailer. This space was not occupied by anything on a regular
trailer and was often left empty or used to package cooling units for the trailer. In
figure 3.26 and 3.27 the placement of the batteries can be seen.

52
 3. Method

Figure 3.26: Trailer Dynamics packaging of their powertrain.

Figure 3.27: Battery position on the first generation of the E-Trailer at Volvo
Trucks.

This location was also beneficial in terms of positioning the center of mass. Since it
was centered on the trailer making the potential shift of its position minimal. This
minimal change in center of mass also resulted in the axle and kingpin pressures
would remain similar to a regular trailer, just with a greater magnitude.

Other electrical components, like the inverter and charger, had greater degrees of
freedom in terms of positioning since they were smaller and lighter. However, looking
at an electrical schematic of an electric powertrain, the inverter will be located be-
tween the batteries and the motor and the charger will be connected to the battery.
Which made it beneficial to also package them close to their connected components.
For the inverters, it meant being close to the batteries and the motors. The charger
could, as seen in figure 3.26, be placed behind the third axle together with other
auxiliary components. It was preferred to have the charger closer to the batteries in
order to keep the entire powertrain as centered as possible.

53
3. Method

The positioning of the cooling components mainly focuses on the placement of the
radiators since they are the most important. The main requirement was to have
sufficient airflow through the radiators. The first E-Trailer positioned the radiators
in front of the batteries, one on each side, see figure 3.28. This feeds the radiators
with clear air without anything to obstruct it. Other positions for the radiators
could have been between the batteries and the first axle or behind the third. The
problem with these positions was that the front of the radiators would have been
blocked by either the batteries or the rear fender, decreasing the amount of air
naturally passing through.

Figure 3.28: Radiator placement on the first E-Trailer

3.6 Packaging of the powertrain

The packaging of the selected components was done in the CAD software CREO
PTC, a CAD software used at Volvo Trucks. Most of the components, like motors,
batteries and various mounts for these already exist and were used as much as
possible. This made the packaging and design process more simple since new parts
were kept to a minimum. Emil Olsson’s report on the first E-Trailer, was used as a
guide and reference material in this packaging study of this E-Trailer, see appendix
F.

3.6.1 Obtaining CAD models of components

The CAD model was structured into three main sub-assemblies. In the list below
are these main sub-assemblies listed.
• Chassis
• Electrical powertrain
• Mechanical powertrain
Since most parts included in this model were already existing components at Volvo
Group, they were located from already built trucks with the right configuration or
parts. The components were obtained using Volvo software.

54
 3. Method

3.6.1.1 Chassis
The chassis sub-assembly contained two assemblies, the Parator chassis and the
Volvo Chassis. The Parator chassis is the upper chassis that carries the containers.
It also included the kingpin to connect the trailer to the truck as well as the support
legs and lights. Figure 3.29 shows a CAD representation of the Parator chassis. It
was taken from the previous E-Trailer since it was not going to change and be built
by Parator. It was mounted to the Volvo chassis using the mounting method derived
from a previous thesis report [4]. The Volvo chassis was the lower chassis built by
Volvo Trucks. It consists of Volvo frames rails, cross members, towing members
etc. These parts were carried over parts from the previous E-Trailer design. Axles
and suspension components were sourced from existing trucks using the described
method above.

Figure 3.29: ISO view of the Parator Chassis.

3.6.1.2 Electrical powertrain

The electrical powertrain assembly contained the following components:
• Inverter
• Batteries
• Charger
• Junction boxes
• Radiators

3.6.1.3 Mechanical powertrain

The mechanical powertrain assembly contained the following components:
• Motors
• Gearbox
• Propeller shaft

55
3. Method

56
 4
Results & Discussion

This chapter contains the results found gathering information from different sources,
what the list of requirements included, what trends were accumulated from the
simulations, the dimensions of the final powertrain, the fuel savings on the different
routes and how the final powertrain is packaged inside the trailer. Once the different
results are presented there will also be a discussion with thoughts and reasoning
around the results.

4.1 List of Requirements

The result of the prestudy was the list of requirements found in appendix A. The
requirements and wishes could be split into two categories: general requirements
and target values that were calculated, see the Vehicle Dynamics section below.
The general requirements were derived from the experience of building and testing
the first generation of E-Trailer. Such requirements could be to have the charger
port on the right side of the trailer or have the ability to heat the batteries. Other
general requirements were more basic and tied to the problem that the E-Trailer
was created to solve, for example, reducing energy consumption from the ICE or
using an electric powertrain.

4.1.1 Vehicle Dynamics

The results from the hand calculations, simulations and the testing of the presented
vehicle dynamics scenarios will be shown in this chapter.

4.1.1.1 Straight Reverse

The result of the hand calculations is presented in table 4.1. The equations are used
to focus on calculating the torque needed on the hub, which then is divided with
different gear ratios to calculate what peak torque is needed from the motor. This
can be seen in table 4.1 where the torque needed from the motor, for a 33-ton GCW
is 220Nm with a 4.11 ratio on the rear axle.

57
4. Results & Discussion

Table 4.1: Results of hand-calculations for reversing straight

GCW [ton] Torque hub [Nm] Gear ratio 4.11 [Nm] Gear ratio 3.52 [Nm]
33 908 220 258
48 1320 321 375
63 1735 422 492

4.1.1.2 Angled Reverse

The function of reversing with an articulation and steering angle was calculated,
simulated and tested.

Calculations

The calculation part was done iteratively, and the result can be seen in table ??.
The torque is calculated at the hubs of the drive axle and is measured in the unit
kNm.

θf ront 0 1 0 1 2 1 2 3 2 3 4 3 4 5 4 5
θrear 0 0 1 1 1 2 2 2 3 3 3 4 4 4 5 5
Thub 1.3 1.8 2.6 3.1 3.6 4.4 4.9 5.3 6.1 6.6 7.1 7.9 8.4 8.9 9.7 10

Simulation

The x-axis in the figures 4.1 represents the time in seconds. The x-axis times are
the same for both 4.1a and 4.1b. The y-axis in figure 4.1a shows the articulation
angle and the steering wheel angle in °. The blue line represents the steering wheel
angle, which ranges from -5 °to 15 °. The black line in the same figure represents the
vehicle’s articulation angle during the reverse. The y-axis in figure 4.1b shows the
torque needed on each driven hub of the trailer. The torque is negative because of
the direction the combination has, since it is reversing the torque is negative. The
maximal torque needed according to the simulation is roughly 1.6 kNm on each hub.
Thus resulting in a total torque need of 3.2 kNm.

58
 4. Results & Discussion

(a) Steering wheel angle and articulation an- (b) Torque needed on each hub from the sim-
gle simulated ulated articulation and steering angles

Figure 4.1: Angled reverse simulation results

Testing

The result from testing the first generation E-Trailer was that the motor can not
pull the vehicle combination from any significant angle. It was also very sensitive
to the slope of the ground, where even small slopes made significant differences.

Discussion

The angled reverse is a very difficult area to fully comprehend. The calculations
depend on tire data that is determined from testing, according to the simulation
tool the first-generation E-Trailer is enough, and the testing of the first-generation
E-Trailer suggests that it definitely is not possible. If the function of reversing with
angles is a function that is wanted the E-Trailer has to be built and tested. An
E-Trailer with higher gear ratios from the motor to the ground will most definitely
deliver enough traction force to test this more thoroughly and thereby deliver the
minimum traction force needed.

4.1.1.3 Powershift

The calculations made for the powershift function resulted in the values that can be
seen in table 4.2. In the table the mass of the trailer (mtrailer ) is given in tons, the
angle αs is given in °and the power needed from the electric motor (PEM ) is given
in kW. Since an up-shift is not going to be more than 10s the power needed from
the electric motor can be assumed to be peak power.

59
4. Results & Discussion

Table 4.2: Values of mtrailer and αslope and the resulting power need from the EM
at 40km/h initial Velocity

mtrailer αs PEM αs PEM αs PEM αs PEM αs PEM αs PEM

10 0 45 1 84 2 122 3 160 4 198 5 236
15 0 50 1 98 2 146 3 194 4 241 5 289
20 0 55 1 113 2 170 3 227 4 284 5 342
30 0 65 1 142 2 218 3 294 4 371 5 447
40 0 75 1 171 2 266 3 362 4 457 5 553
50 0 85 1 200 2 315 3 429 4 543 5 658

Discussion
A quite powerful motor is needed to have the powershift function. For completely
flat roads the needed power is relatively low, however, it is important to know that
no roads are completely flat and the needed motor power increases significantly
between 0°and 1°.
A 300kW motor seems to be a size that can handle a spread of masses/angles very
well. It is important to know that the steps in table 4.2 will affect these values. If for
instance, the steps of the variable mtrailer would have been 5 tons the needed motor
size would be different. Therefore it is important to know the expected payload on
the trailer that the powershift function will be dimensioned for.
The initial velocity of 40km/h where chosen as the average speed of the powershift
calculations. This means that with larger velocities the peak power needed of the
motor will increase. The table 4.2 is therefore only valid for velocities of 40km/h or
below.

4.2 Simulation & Dimensioning

During this chapter, the results from the simulation tool validation and the sweeping
simulations will be presented. The results will also be discussed succeeding the
presented results.

4.2.1 Simulation Tool Validation

The simulation tool was validated in two different ways, validating towards a diesel
truck and the E-Trailer itself. The results and discussions of the validations will be
presented below.

4.2.1.1 Diesel Truck Validation

When running the simulation tool with the data from the Autofreight vehicle on the
same route (Gothenburg to Borås). Three simulations were made from real data
and their results are shown in figure 4.3.

60
 4. Results & Discussion

Table 4.3: Diesel consumption from Autofreight vehicle compared to simulation

tool

Attempt 1 2 3
Real Consumption [kWh] 22.4 40.4 27.4
Simulation Consumption [kWh] 29.4 47. 35.2

Discussion
The amount of fuel that is consumed seems to always be higher in the simulation
tool when compared to real data. We believe this might be because the simulation
savings of diesel is scaled from how many kWh of fuel is needed. This means that
the amount in liters is calculated by using the efficiency values of different parts of
the powertrain. These efficiency values are estimations and might be different in
reality. An example of this is the efficiency of the diesel engine, where the value for
the diesel engine’s efficiency was 30%. In reality, it might be higher which would
result in fewer liters of diesel consumed. As long as the simulation is consistent
and calculates the fuel consumption equally for the E-Trailer the percentage of fuel
savings should be similar. The exact amount of fuel consumed should however not
be trusted fully.

4.2.1.2 E-Trailer Validation

The fuel saved in percentage, can be seen in table 4.4, are quite close when the
battery is not reaching its "safe state" where it starts limiting the power output
which can be seen by the SoC value in attempt 1. This is, however, the only attempt
where the end SoC from the simulations differs from the real-life data. During the
other attempts the end SoC from the simulations follows the actual end SoC while
the fuel-saving percentage differs. The simulation tool saves up to 5% more fuel
than the real data in attempts 2 and 3.

Table 4.4: Results from real test data and simulation data

Real Fuel Savings [%] 23.8 18.6 15.5

Simulation Fuel Savings [%] 24.3 23.8 19.5
Real End SoC [%] 32 30 27
Simulation End SoC [%] 34 30 27

Discussion
In attempt 1 it can be seen that the simulation has not used as much energy com-
pared to the real data. This means that the simulation tool saves the correct amount
of fuel but is not draining the battery correctly. In attempts 2 and 3 it can be seen
that the SoC at the end of the route is following the real value while the fuel-savins
are more than in the real-life data. We think this means that the battery runs out
later than it should. This will lead to the battery lasting longer and the total fuel
savings at the end of a route are going to be more than it should be.

61
4. Results & Discussion

Having more real testing data would also be useful in order to simulate more and
higher levels of SoC. This could lead to more data and would make it easier to draw
a conclusion about the simulation tool.
The Simulation tool is most likely overestimating the electrical powertrain. We
decided due to time limitations that we would use this simulation tool with increased
mass as safety-factor and would not spend time on making changes in order to
improve it.
The conclusions from the simulation validation are that the tool itself is not perfect.
But since the intended simulation tool was not an option, we had to go with the
next best thing, which was this tool.

4.2.2 Sweep Simulations

In the following section, the results from the sweep simulations will be presented.
Due to the number of sweep simulations, only some powertrains and their results
will be presented in order to make the results easier to grasp. The full results from
the simulations can be found in "Simulation Results" in Appendix F.

4.2.2.1 3 × Gothenburg-Borås-Gothenburg
The best match of motor and battery size of the sweep simulations on the route
three times Gothenburg to Borås (Viared) back to Gothenburg can be seen in table
4.5. A maximal fuel saving of 38.9% is achieved by the motor with 1500 Nm of peak
torque when it is combined with a battery size of 650 kWh. Another powertrain
that was simulated consists of a motor that delivers 750 Nm peak torque and has a
battery with a capacity of 550 kWh. This powertrain will save roughly 30.2% fuel
compared to a conventional combination.

Table 4.5: Results of sweep simulations on the route 3xGBG-Via-GBG

Motor Size [Nm] Battery Size [kWh] Maximal Fuel Savings [%]
250 250 10.7
500 450 22.4
750 550 30.2
1000 600 34.6
1500 650 38.9

Discussion
That the maximal fuel savings are archived by the biggest powertrain is not a sur-
prise since it will also be the most expensive one with its 650 kWh of battery capacity.
In table 4.5 it can be seen that the increase in battery capacity is not linear. This
is because of the different motors’ possibility to regenerate potential energy back
to electric energy. Where the bigger motors regenerate more energy, therefore the

62
 4. Results & Discussion

powertrains with bigger motors do not have the need for a linearly increased bat-
tery capacity. But a bigger motor will need an increase in battery capacity because
a bigger motor consumes more energy. In the same table, it is also evident that
the total fuel savings increases with the bigger powertrains, however, the amount it
saves decreases with each step of powertrain size. For example, between the 250 Nm
motor and the 500 Nm motor, the total fuel savings increased by 11.7 percentage
units. Between the 500 Nm motor and the 750 Nm motor, the fuel-saving increase
is 7.8 percentage units. From the 750 Nm motor to the 1000 Nm, the increase is 4.4
percentage units. While a bigger step of motor size (500 Nm instead of the previous
250 Nm steps) from the 1000 Nm to the 1500 Nm motor is only 4.3 percentage units.

It is also important to remember the price of the components. Bigger components

will cost more money. Increasing the battery capacity is the most costly part of the
powertrain. This means that the battery capacity should be minimized. Minimizing
the battery capacity will also decrease the cost of charging it fully. Since electricity
is expensive compared to diesel the bigger powertrains will in general cost more each
day. Increasing the battery size would also increase the total mass of the trailer. Due
to axle loads and distances the trailer has a maximal cargo weight. By increasing
the mass of the trailer the maximal cargo weight will decrease.

Determining which powertrain would be the best is difficult. If the cost of the trailer
is not in consideration the best powertrain would be the biggest powertrain since it
would save the most fuel. The most efficient when considering fuel savings per kWh
battery would be the 500 Nm motor. And a middle ground would be the powertrain
with the 750 Nm motor.

4.2.2.2 2 × Tampere-Helsinki-Tampere
The results of the simulations on the route that starts from Tampere, goes to Helsinki
and then back to Tampere 2 times in a day can be seen in the table 4.6. The maximal
fuel saving is made with the powertrain that consists of the motor with 1500 Nm
peak torque and a battery capacity of 1000 kWh. Another powertrain consists of
a motor that delivers 500 Nm peak torque combined with a battery pack with a
capacity of 750 kWh.

Table 4.6: Results of sweep simulations on the route 2xtamp-helsi-tamp

Motor Size [Nm] Battery Size [kWh] Maximal Fuel Savings [%]
250 500 13.3
500 750 22.8
750 1000 29.6
1000 1000 32.4
1500 1000 33.0

63
4. Results & Discussion

Discussion
The maximal fuel savings is still achieved by the largest powertrain possible. The
largest powertrain simulated can be seen in table 4.6 where the 1500 Nm and its
1000 kWh battery capacity saves roughly 33% fuel compared to the conventional
combination. But if the cost of the batteries and motors is considered this changes
quite drastically. This can also be seen in table 4.6 where the best battery capacities
for each motor size are shown. The increased fuel savings between the 1500 Nm
motor and the 1000 Nm motor is only 0.6 percentage units, and the fuel savings
between the 750 Nm and the 1000 Nm motors are only 2.8 percentage units. This is
most likely due to a bad match between the bigger motors and their batteries and
the bad match is a result of not simulating bigger batteries than 1000 kWh. This is
because it would be incredibly expensive to put more than 1000 kWh of batteries on
the trailer. The step from the 500 Nm motor to the 750 Nm motor is a significant 6.8
percentage units, but the 750 Nm motor requires a battery capacity of 1000kWh.
This battery size is right now unrealistic because of its specific energy and cost.
Batteries with a capacity over 750 kWh are therefore not fulfilling the requirements.
This means that the 750 kWh needed with the 500 Nm motor is considered the most
optimal for this route.

4.2.2.3 Kurtalan-Bahçesaray-Cizre
The results of the best powertrain combinations that were simulated can be seen in
table 4.7. The maximal fuel savings comes from the powertrain combination that
has a motor with 1500Nm peak torque and a battery capacity of 200 kWh. On
this route, the most optimal powertrain is also the powertrain with the highest fuel
savings.

Table 4.7: Results of sweep simulations on the route Kurtalan-Bahçesaray-Cizr

Motor Size [Nm] Battery Size [kWh] Maximal Fuel Savings [%]
250 50 -1.3
500 50 4.0
750 100 8.9
1000 150 13.7
1500 200 22.4

Discussion
This is the only route where the powertrain that saves the most also is quite realistic.
Table 4.7 shows that the maximal fuel savings are around 22.4% with a 1500 Nm
motor and a 200 kWh battery. However the table 4.7 does not show how the fuel
savings differs between different battery capacities and the same motor. In this case,
the best fuel savings for the 1500 Nm motor is the 200 kWh one, but it does not show
that the 100 kWh battery capacity saves 22.0% when it is compared to a conventional
combination. The fuel savings between a 200 kWh battery and a 100 kWh battery

64
 4. Results & Discussion

is therefore only 0.4 percentage units. In this case, the extra fuel savings from a 200
kWh battery does not make up for having an extra 100 kWh of battery capacity. It
is important to know that this route is an absolute extreme and has the extra cost,
both weight-wise and the price of the 100 kWh extra battery capacity. It might
be beneficial on routes that are similar but not exactly as extreme. Deciding an
absolute optimal is therefore more difficult on this route than the others, but the
trend of having a big motor combined with a relatively small battery is evident.

4.2.2.4 Sensitivity analysis of charging price

The result of the sensitivity analysis on each route is presented below as well as
what powertrain that was compared and its reduction in fuel consumption.

3 x GBG-Viared

Powertrain specs: 1000 Nm, 600 kWh battery and saving 34,6% fuel

Table 4.8: Sensitivity analysis of charging price for GBG-Viared.

Charging price [SEK/kWh] Profitability [SEK/day]

1 53,7
2 -186.3
3 -426.3
4 -666.3
5 -906.3

2 x Tempere-Helsinki

Powertrain specs: 750 Nm, 750 kWh battery and saving 23,7% fuel

Table 4.9: Sensitivity analysis of charging price for Helsinki.

Charging price [SEK/kWh] Profitability [SEK/day]

1 -7,95
2 -307.95
3 -607.95
4 -907.95
5 -1207.95

Turkey

Powertrain specs: 1500 Nm, 100 kWh battery and saving 22,0% fuel

65
4. Results & Discussion

Table 4.10: Sensitivity analysis of charging price for Turkey.

Charging price [SEK/kWh] Profitability [SEK/day]

1 474.8
2 434.85
3 394.8
4 354.8
5 314.8

Discussion

The charging price has a big effect on if it is possible to save money using an E-
Trailer due to lower fuel consumption. It can clearly be seen in the results above and
shows that the price of liquid fuel, diesel in this case, is very cheap for the amount of
energy it carries compared to electricity. The outlier here is that the route in Turkey
actually being profitable regardless of the change in charging price. This follows a
trend seen on the other routes as well being that the larger battery capacity the
trailer is equipped with, the less profitable it becomes. This can be traced to the
fact that the scenario tested is with a fully charged battery. The more the customer
has to charge the battery, the more energy needs to be regenerated to make up for
that initial cost. On routes where the "free" regenerative energy is less available,
such as more flat routes like Helsinki-Temper, fewer opportunities are there to gain
back that initial investment. On the other hand, on a route like Turkey where the
route is very hilly, the opportunity of using regenerative braking to get back electric
energy is easier. Once again, this shows that regenerative braking is very important
to make both more effective, but also potentially profitable as well. It is important
to remember that the E-Trailer is not aimed to make transport more profitable,
but rather reduce emissions by reducing fuel consumption, to meet potential future
regulations. Such an adaptation will most likely cost the customer and the question
becomes more of: "How much is a customer willing to pay to reduce emissions by
X%?". Which is a question only a potential buyer of an E-Trailer can answer.

4.3 Final Powertrain

This section will cover the selected powertrain components and their specifications
compared to the ones derived from the sweep simulations. It will also show the sim-
ulation results of the selected powertrain both on its designated route, Gothenburg-
Viared, with the standard 15 tons of payload, but also how it performs on the other
routes and with varying conditions such as SoC and payload.

4.3.1 Powertrain selection

In table 4.11 the selected powertrain components are listed. Less critical systems
such as cooling and mounting solutions are not included.

66
 4. Results & Discussion

Table 4.11: Selected in-house components.

Part No of units
Electric motor 2
Battery 4
Inverter 2
Gearbox 1

In table 4.12 and 4.13 the selected component’s specifications are compared to the
targeted specifications.

Table 4.12: Specification for the selected components used in the final simulation.

Specification Selected motor Target values

Peak torque [Nm] 800 750
Continuous torque [Nm] 510 600
Peak RPM 10000 3250
Peak power [kW] 310 125
Continuous power [kW] 210 75
Gearing First: 1:9.26, Second: 1:3.52 1:1
Peak torque at hub [kNm] First: 30.4, Second: 11.6 4.11
Continuous torque at hub [kNm] First: 19.4, Second: 7.4 2.5

Table 4.13: Specifications of the batteries used in the final simulations

Specification Selected battery Target values

Capacity [kWh] 360 550
Peak power [kW] 173 125
Continuous power [kW] 90 75
SoC window [%] 10-90 30-70
Usable SoC at 100% SoC 288 220

Discussion
The selected powertrain components do not match the specification of the targeted
powertrain derived from the sweep simulations perfectly. This has a lot to do with
the SoC window used in the sweep simulations. Because of the small SoC window,
a big increase in battery capacity was needed which also increases the mass of the
trailer.

Comparing the battery capacity it might look like the batteries are even further
away from what the initial simulations suggested. But in the end, they ended up
carrying more capacity than the wanted 550 kWh. This is due to their larger SoC
window compared to the simulated batteries. As mentioned earlier the SoC window
during the sweep simulations was between 30-70 % and the final batteries had a
simulated window of 10-90%, which also included a ∼ 5 % safety margin to what

67
4. Results & Discussion

their specified window was. This difference was due to the large safety margins
taken when testing the first E-Trailer where the SOC window was set to be 30-70%.
A range that became the only reference before the specification sheet for the se-
lected, "Cube", batteries were found. In the end, this was a positive discovery and
the effective capacity aimed for had to be calculated from the stated capacity in the
simulations to get the required usable capacity. Since the selected batteries had a
larger SoC window this meant a decrease in total battery capacity needed, fewer
battery modules needed, fewer components to package and a lighter E-Trailer.

The biggest difference we saw during the powertrain selection was the addition of
a gearbox. This addition is because there are no electrical powertrains available at
Volvo that do not include one. The most common powertrain includes an automatic
12-speed gearbox which would be unnecessary for the E-Trailer. The smaller trucks
have another gearbox that is only 2-speed. Because the gearbox with 12 gears was
deemed unnecessary the configurations with the 2-speed gearbox were selected. This
left the choice between EPT402 and EPT802. Both configurations come with the
same electric motor, where the EPT402 configuration only has 1 motor and the
EPT802 configuration has 2 of them. The motor can deliver a peak torque of 400
Nm, but together with the highest gear of the gearbox, the torque to the drive axle
is around 1400 Nm. According to the sweep simulations, the EPT402 configuration
would therefore be more than enough. But during some initial simulations of this
powertrain, it became evident that the EPT402 did not drain the batteries (with
the new battery capacity and mass). Instead of further decreasing the battery
capacity the EPT802 was simulated. The EPT802 drained the battery more and
was therefore a better match with the battery than the EPT402 configuration.
From these specifications, it can also be seen that the powertrain is power limited by
the batteries. Meaning that the motors can push and regenerate more power than
the batteries can handle. This is not ideal of course, especially when considering
the effect of regenerative braking. When the electrical powertrain is limited by the
batteries and not the motors, the effect of having bigger motors will be neglected
and thereby some energy will not be transferred back to the batteries.

4.3.2 Final Powertrain Simulations

The final simulations altered the different routes, the initial SoC and the payload to
see what fuel savings we could expect from the final powertrain. The first couple of
simulations were made on the route between Gothenburg and Viared. The results
are presented as a percentage fuel saving compared to a conventional combination
and can be seen in table 4.14. The maximal fuel saving reached 42.1% and is reached
when the E-Trailer is fully charged when leaving Gothenburg and pulling 15 tons of
payload. The fuel needed from the truck’s diesel tank is visualized in figure 4.2, the
figure also shows how much the conventional combination would consume.

68
 4. Results & Discussion

Table 4.14: Results of final powertrain simulations on the route

3xGbg-Borås-Gbg

Payload [ton] Initial SoC [%] Maximal Fuel Savings [%]

15 100 42.1
15 50 27.0
15 10 1.2
30 100 41.2
30 50 23.2
30 10 3.8
45 100 35.6
45 50 20.4
45 10 5.9

Figure 4.2: Energy needed from the trucks diesel tank with E-Trailer compared
to a conventional trailer on the route Gothenburg to Borås

The final powertrain was also simulated on the route that starts in Tampere and
goes to Helsinki then back two times. The results of the powertrain pushing different
payloads while having different initial SoC are presented in table 4.15. The maximal
fuel savings is 30.6% with the E-Trailer compared to the conventional trailer. This
was also done with the least amount of payload (15 tons) and a fully charged battery
and can be seen in figure 4.3.

69
4. Results & Discussion

Table 4.15: Results of final powertrain simulations on the route

2xTampere-Helsinki-Tampere

Payload [ton] Initial SoC [%] Maximal Fuel Savings [%]

15 100 30.6
15 50 13.4
15 10 -3.8
30 100 26.3
30 50 12.8
30 10 -0.6
45 100 18.3
45 50 12.0
45 10 1.0

Figure 4.3: Energy needed from the trucks diesel tank with E-Trailer compared
to a conventional trailer on the route from Tampere to Helsinki

The final simulations were conducted on the route that starts in Kurtalan and
then goes to Bahçesaray and Cizre while ending up back in Kurtalan. The fuel
savings that are simulated with the final powertrain with different payloads and
initial SoC is presented in table 4.16. The maximal fuel savings is 28.7% compared to
a conventional and the difference between the two different combinations is visualized
in figure 4.4.

70
 4. Results & Discussion

Table 4.16: Results of final powertrain simulations on the route

Kurtalan-Bahçesaray-Cizre-Kurtalan

Payload [ton] Initial SoC [%] Maximal Fuel Savings [%]

15 100 28.7
15 50 19.6
15 10 10.9
30 100 10.8
30 50 6.3
30 10 2.5
45 100 -2.0
45 50 -2.5
45 10 -3.6

Figure 4.4: Energy needed from the trucks diesel tank with E-Trailer compared
to a conventional trailer on the route in eastern Turkey

Discussion
The final powertrain did a lot better than we expected. We think that this mainly
has to do with the over-estimated SoC window when doing the first simulations.
By increasing the SoC window from 40% to 80% the needed battery capacity was
halved. Because of the decrease in needed battery capacity, the total mass of the
batteries will also decrease. Another reason is probably that the motor is bigger
than the sweep simulations. The final powertrain simulations, therefore, have even
better values than the initial sweeping simulations.

71
4. Results & Discussion

The decrease in mass is especially evident in the simulations on the Turkish route.
Where the final powertrain improves on all of the initial simulations where the best
fuel savings generally was made with a big motor and a small battery.
Even if the simulation tool is not the most accurate, having up to roughly 42% of
fuel savings should end up saving fuel on a real truck. To find out exactly how much
this E-Trailer can save it has to be built and tested.
It would be interesting to simulate different amounts of cube batteries. For example
reducing the amount to 3, 2 or even 1 and see how much a cheaper E-Trailer could
reduce the fuel consumption. Maybe even trying the EPT402 configurations together
with other amounts of batteries.

4.4 Packaging
The final packaging of the E-Trailer can be seen in figure 4.5 and 4.6. In the following
subsections, the final packaging of each subassembly will be presented.

Figure 4.5: Isometric view of the E-Trailer

Figure 4.6: Side view of the E-Trailer

72
 4. Results & Discussion

4.4.1 Chassis
The chassis contains both the Parator container chassis and the Volvo chassis. The
second axle was decided to be the driven one with the first and third liftable. To
increase the maneuverability of the truck and trailer combination when it was loaded
and can not lift any axles, the third axle was also decided to be self-steered. The
first axle was decided to only be liftable.

The axle distances can be found in table 4.17. The final axle configuration was a
3-axle trailer with a pusher axle configuration and the last axle was both liftable
and self-steered. The three-axle combination was packaged more similar to what a
four-axle truck would have been, with the second and third axle sharing mounting
points and the first axle still being a single, compared to the first E-Trailer which
had three single axles mounted together. This change should make the ordering and
assembly of the trailer slightly simpler since it would follow a more standardized
assembly. In figure 4.7 - 4.8 CAD visualization of the chassis is shown.

Table 4.17: Final axle distances for the E-Trailer.

Distance [mm]
Kingpin to Axle 1 7050
Axle 1 to Axle 2 1350
Axle 2 to Axle 3 1375
Axle 3 to Rear 1712

In table 4.18 and 4.19 the kingpin forces calculated for the four different cases with
varying lifted axles are presented. The results in table 4.18 use the base trailer’s
axle distances and table 4.19 shows the results using the final axle distances shown
in table 4.17. The payload used was the previously derived 15 tons for both trailers.

Table 4.18: Resulting kingpin force with different combinations of lifted axles
using base trailer distances.

Case Fkp [kN] Comments

1 5.88 All axles on the ground
2 3.28 Second and third axle lifted
3 7.75 First and second axle lifted
4 5.88 First and third axle lifted

Table 4.19: Resulting king pin force with different combination of lifted axles
using final distances.

Case Fkp [kN] Comments

1 5.89 All axles on the ground
2 3.25 Second and third axle lifted
3 7.80 First and second axle lifted
4 5.86 First and third axle lifted

73
4. Results & Discussion

Discussion
The decision to only have the third axle steered was because the pushed axle can
not be self-steered. A hydraulically steered axle would have required an extra hy-
draulic system which was unwanted due to increased complexity and part count. A
hydraulically steered axle could, however, most likely have been implemented if the
need for it was expressed by a customer.

As seen in table 4.18 the kingpin force was the same between case 1 and case 4, as
expected. The other two cases also produced results that were expected with case 2
reducing the kingpin force due to creating an overhang that counteracts the initial
moment pushing down on the kingpin. A similar comment can be said about the
third case, where the moment arm, creating the moment pushing the kingpin down,
is increased. Since the load is assumed to be equally spread out over the trailer,
the increased distance also increases the force that pushes the kingpin down. The
same trends can be seen in the final distances with the difference that case 1 and
case 4 no longer being exactly the same. This is due to the slightly longer distance
between axles two and three compared to the base trailer. The distance from the
third axle to the rear of the final trailer is slightly longer on the base trailer making
the results differ slightly as well. None of these differences should be noticeable in
real life.

Figure 4.7: Isometric view of the chassis

Figure 4.8: Side view of the chassis

74
 4. Results & Discussion

4.4.2 Mechanical powertrain

The electric motors and gearbox are connected to the driven axle through a propeller
shaft. No support bearings were needed to clear the pusher axle due to its length.
The motor was placed as far rearwards as possible, towards the pusher axle, in order
to clear the pusher axle and make the propeller shaft shorter. In figure 4.9 and 4.10
the resulting clearances, angles and lengths for the propeller shaft in the worst-case
scenario are shown. The propeller shaft’s length and angle, α in the three scenarios
is shown in table 4.20. In appendix D the measurements for the two other cases are
shown. Both the gearbox flange and the axle flange are designed to be mounted to
the frame at a ∼ 4◦ angle in reference to the vertical plane.

Figure 4.9: Side view of drive axle in full droop and pusher in full bump.

Figure 4.10: Top view of drive axle in full droop and pusher in full bump.

75
4. Results & Discussion

Table 4.20: Resulting values of lprop and α from the three scenarios tested.

Scenario lprop [mm] α [◦ ]

1 1597,0 7,76
2 1583,8 1,85
3 1583,5 1,41

The propeller shaft was not visualized in CAD due to issues with the alignment of
the universal joints. Below, in figure 4.11 an isometric view of the subassembly can
be found. The mounts are the same used for this motor and gearbox combination
in production trucks.

Figure 4.11: Isometric view of the mechanical powertrain

4.4.3 Electrical powertrain

The electrical powertrain was packaged between the supporting legs and the first
axle on the trailer. Batteries were mounted using their original mounting solution
on the side of the trailer. Inverters and the charger were placed in between the
batteries. All smaller components were not included. This was because there was
no time to decide upon them. They will be added later and will most likely be
placed on top of the batteries for easy access, similar to the first generation of E-
Trailer. The radiators were placed in front of the batteries using carryover mounting
solutions from the first generation of the E-Trailer. They originate from a hybrid
powertrain from Volvo Busses. Figure 4.12 shows an isometric view of the electric
powertrain.

76
 4. Results & Discussion

Figure 4.12: Isometric view of the electric powertrain.

Figure 4.13: Top view of the assembly with components marked out.

4.4.4 Requirement fulfillment

The final choice of powertrain and packaging solution fulfills all of the requirements
that were investigated. Some requirements were unfortunately not investigated as
the thesis had time limitations. Other requirements were fulfilled because solutions
already existed. These requirements were:
• Maintain the rated temperature of the electrical components during driving.

77
4. Results & Discussion

• Be able to heat batteries.

The cooling of the batteries has been brought up earlier, but not fully investigated
in this thesis. From testing the first generation E-Trailer as well as the original
intended use of the radiators, it seems unlikely that they will be restricting the
powertrain. Similar can be said for the heating of the batteries. This function was
controlled by an inline heater in the cooling loop both for the first generation of the
E-Trailer as well as how the production solution works for electric trucks at Volvo
Trucks. Therefore, no effort was put into these two requirements due to the knowl-
edge that both were already solved by standard components that could be easily
implemented on the E-Trailer as well.

Most of the wishes were also fulfilled. The ones that are not seen as fulfilled are:
• Maximizing the articulation and steering wheel angle when reversing only using
the E-Trailers powertrain.
• Minimize weight.
• Minimizing the length of cooling hoses and cables.
As mentioned earlier no reliable value for the angled reverse case could be derived
in this thesis and therefore not possible to judge if the selected powertrains specifi-
cation could do it or at what angle. Instead, this is recommended to be tested with
the selected powertrain.

Due to time and also the requirement of using Volvo components, the wish to min-
imize the weight of the trailer chassis was not looked into. This was mostly due to
no real components design was done in this thesis since all significant components
were already designed.

The wish of minimizing the length of cooling hoses and cables was not looked into
either since the packaging never was mature enough to a point where cable and hose
routing was relevant.

78
 5
Conclusion

In this final chapter conclusions drawn from the simulation results and packaging
will be presented together with some thoughts on the whole E-Trailer concept in
general. Finally, some recommendations for future studies will be brought up to
cover areas that were discovered, but not fully explored in this thesis.

5.1 General conclusions

The thesis fulfilled its aim of presenting a deeper understanding of the E-Trailer
concept, what is possible to achieve with it in terms of reduced fuel consumption
and other beneficial features that come with adding a second, electric, powertrain to
a trailer. The thesis also delivered on the packing aspect of the project by present-
ing a packing solution that is using mostly standardized Volvo Group components
and fulfills most of the requirements. The two unfulfilled requirements both have
carry-over production solutions that should fulfill the requirements, but since noth-
ing was tested or validated for this design the requirement was deemed as unfulfilled.

Second generation ended up having a very similar powertrain configuration to the

one derived in this thesis. This was due to some late changes, which lead to problems
with compatibility between component software. Even without these problems, they
would have been similar. This is probably due to the very similar starting points
for the two different projects with regard to requirements and limitations. The
difference is that the thesis developed one having a stronger, proven, method of
coming up with the components and design. Compared to the second generation,
which was built more on engineering intuition and experience. The impressive part
is that both paths basically lead to the same product. It also shows that, with
the current components available at Volvo Group, this is probably the "best" E-
Trailer that can be built in the foreseeable future and that future improvements will
come from software, as mentioned earlier, or from testing in real-life scenarios. It is
therefore recommended to focus on these two areas to develop this concept further.
The data available is now analyzed and the result is ready to be tested in order to
get more information and learn new things about the E-Trailer concept.

5.2 Simulations
The concept of adding an electric powertrain to a trailer to reduce fuel consumption
is viable and the simulations show it very clearly. It might not come as a surprise

79
5. Conclusion

since it is effectively adding another source of energy to the combination. Since this
second source is an electric powertrain there will be a substantial decrease in fuel
consumption and thereby carbon emissions. The fact that an electric powertrain
adds the possibility to regenerate energy to be reused also adds to the reduction of
fuel consumption.

We thought that a more powerful powertrain would lead to increased fuel savings.
However, this is not always the case and it was shown in this thesis. The selection of a
powertrain is very route dependent and in this thesis, the main goal was to maximize
the savings in fuel consumption. This made the selection of the optimal powertrain
for each route simple, but maybe not the most realistic or easily defendable when it
comes to the investment of the E-Trailer.
Similar results can be had by different powertrain configurations on the same route.
This makes the decision of the optimal configuration difficult to decide since the
decision comes down to customer specific preferences.
Making the decision of which powertrain size to go with was difficult since it would
come down to customer preferences or requirements to truly get the best one. Some-
thing that was concluded to be the case for most things when it comes to the trucking
and transporting industry.

To make a general, simulation would not be possible and the only realistic way of
doing the simulation is to make specific cases to compare against. This is due to the
vast amount of variables and customer specific requirements that go into building
a truck. Therefore it would be interesting to continue this study with a transport
company to get these variables from them and then derive an optimal powertrain for
their specific use case and route. Without these customer requirements, it is hard
or near impossible to say which powertrain is the best for them, since it depends
too much between routes and their requirements.

An interesting trend that was discovered in the simulations was the importance of
regenerative braking and how it affected both fuel consumption and how it can make
the E-Trailer profitable for a customer. This is not only relevant for the E-Trailer,
but for any electric powertrain since the energy that otherwise would be lost now
can be transferred back to electrical energy. Making the required battery capacity
smaller, making the battery pack smaller and lighter. This sets some requirements
on both the motor and batteries as well and in this thesis it was shown that the
batteries were the limiting factor in terms of the ability to take care of the regener-
ated energy in the final selected components.

Regenerative braking is, to some extent, affected by the deployment strategy of the
electric powertrain. The strategy used in this thesis was very basic. However, it
was not impossible to see how a more adaptive or advanced strategy could save
even more fuel and is strongly recommended as something that should be looked
into, in order to push the development of the E-Trailer further. Especially since the
other limiting factors are hardware related, batteries and motors. Developing better
motors and batteries to increase the powertrain’s performance in terms of power

80
 5. Conclusion

delivery and regenerative braking is undergoing, but it takes time. Implementing

a more advanced deployment strategy should be a better return on investment at
the moment since it exists more advanced strategies or technology to make them.
The Volvo Trucks function "I-see" is one technology that could potentially be used
to develop a more advanced strategy.

It is difficult to say how accurate the simulations of the selected components actually
are. Since more specifications were known about the simulated components, the sim-
ulations should be more accurate than the sweep simulation, which also correlated
fairly well with the real test. Fortunately, the second generation of the E-Trailer
ended up with a very similar powertrain to the one derived in this thesis. Therefore
it is strongly recommended to do some or all of the test that was carried out in this
thesis with that one to have some good data for validation. Especially the arraigning
maneuvers described in the vehicle dynamics section should be quick and easy to
test since they can be done in closed of areas like Hällered. An important test there
would be the reversed at an angle since no real answer was found for that in this
thesis.

It is important to know that the simulations are very rough and real data is always
preferred. To know for certain if the motors are a good match to the battery capacity
is impossible without testing. Therefore we would recommend that this powertrain
is tested to gather real information about the powertrain size and how much fuel it
actually saves.

It would be very interesting to test the EPT802 motor/gearbox configuration with

only one motor running to simulate having an EPT402 configuration instead. An-
other interesting possibility is to try and match the battery capacity to the EPT402
powertrain. This could be a way of making the E-Trailer concept modular so that
customers that do not need the power and capacity from the EPT802 could have a
smaller alternative.

5.3 Packaging
The packaging portion of this thesis went as planned and the implementation of a
pusher axle on the trailer turned out to be doable. Since it was almost only packag-
ing of Volvo Trucks components, designed to fit together, the process went smoothly.
It did not fully reach the level of detail that was initially thought possible. This was
due to the process of finding the right components as well as knowing what other
auxiliary components, such as coolant pumps and reservoirs, were required. To have
this level of detail would have required schematics of the electrical and cooling sys-
tems, something that was out of the scoop for this thesis.

With the packaging now done, it would have been interesting to try to fit an E-axle
into the E-Trailer but there was no time. It would make the powertrain a bit more
compact due to the removal of the propeller shaft and maybe it could be proven
that more gears have some benefit in the future.

81
5. Conclusion

82
 Bibliography

[1] Ralph Andersson. Trailer med eldriven hjälpmotor sänker bränsleförbruknin-

gen. Apr. 2023. url: https://www.trailer.se/artikel/trailer- med-
eldriven-hjalpmotor-sanker-bransleforbrukningen.
[2] Lennart Cider. Expert Product Dev./Engineering at Volvo Group Trucks Tech-
nology. 2023.
[3] Directorate-General for Communication. Delivering the European Green Deal.
Jan. 2023. url: https://commission.europa.eu/strategy-and-policy/
priorities- 2019- 2024/european- green- deal/delivering- european-
green-deal_en.
[4] Lovisa Fernvik and Shiva Sateei. Framtagning av en ny E-DUO-koncepttrailer:
Ett hållbart koncept för framtida E-DUO-trailer. June 2021.
[5] Directorate General. Paris Agreement. 2023. url: https : / / climate . ec .
europa.eu/eu-action/international-action-climate-change/climate-
negotiations/paris-agreement_en.
[6] Annika Hansson. Autofreight. Mar. 2023. url: https : / / www . boras . se /
foretagare/natverkstraffarforetagarforeningarochprojekt/projekt/
autofreight.4.1a6791a316f1a92fec2b236.html.
[7] E. Pettersson, L. Larsson, and N. Fröjd. Svenska HCT Typfordonskombina-
tioner utvärderade mot år 2020 gällande regelverk för BK4. Tech. rep. Nordisk
Vejforum, 2021. url: https://nvfnorden.org/wp-content/uploads/2021/
04/2021-04-15_Svenska_HCT_Typfordon.pdf.
[8] Trafikverket. Bärighetsklasser (BK) på vägar och broar. Aug. 2022. url: https:
//bransch.trafikverket.se/for-dig-i-branschen/vag/bk--barighetsklasser-
pa-vagar-och-broar/.
[9] Trafikverket. Mått- vikt- och lastbestämmelser för fordon och fordonståg. 2021.
url: https://www.transportstyrelsen.se/sv/vagtrafik/Yrkestrafik/
Gods-och-buss/Matt-och-vikt/.
[10] Trafikverket. NVDB på webb. 2022. url: https://nvdb2012.trafikverket.
se/SeTransportnatverket.
[11] Trafikverket. Regler om kör- och vilotider. Jan. 202. url: https : / / www .
transportstyrelsen.se/sv/vagtrafik/Yrkestrafik/Kor--och-vilotider/
regler-om-kor--och-vilotider/.
[12] Transportstyrelsen. Fordonsuppgifter. Apr. 2023. url: https://fordon-fu-
regnr.transportstyrelsen.se/.
[13] Transportstyrelsen. Legal Loading. Apr. 2022. url: https://www.transportstyrelsen.
se/globalassets/global/publikationer-och-rapporter/vag/yrkestrafik/
lasta-lagligt/tran045-lasta-lagligt-eng-low.pdf.

83
Bibliography

[14] Transportstyrelsen. Modulsystemet. Nov. 2021. url: https://www.transportstyrelsen.

se/sv/vagtrafik/Yrkestrafik/Gods-och-buss/Matt-och-vikt/langd-
och-breddbestammelser/Modulsystemet/.
[15] InCharge Vattenfall. Hur mycket kostar det att ladda en elbil? Feb. 2022. url:
https://incharge.vattenfall.se/kunskapshubb/artiklar/hur-mycket-
kostar-det-att-ladda-en-elbil/.
[16] AB Volvo. Electromobility. Jan. 2023. url: https://www.volvotrucks.com/
en-en/about-us/electromobility.html.

84
 A
Appendix A

List of requirements
Requirements:

• Lower energy consumption from ICE.

• Electric powertrain.
• Built with existing Volvo components.
• Minimum ground clearance:
– Rear axle: 230mm
– General: 203mm
• All required components are installed on the trailer.
• Use same base mounting solution for different powertrain configurations.
• Being able to start/reverse with a straight base loaded combination on flat
ground at 1m/s. Result: provide continuous torque of 908Nm at the wheel
hub.
• Not losing any speed when up shifting gears by having the electric motor
support with the lost energy during shifting time. Result: 60 kW at 40 km/h
on flat ground.
rpmmax
• Be able to add power up to 90 km/h. igearbox ·if inalgear
≥ 90km/h
• Regenerate energy going downhill.
• Charging port on the right side of the trailer.
• External charging possible, DC.
• Maintain the rated temperature of the electrical components during driving.
• Be able to heat batteries.
• Three axle container trailer.

Wishes:

• Minimize weight.
• Two liftable axles.
• Steerable axles.
• Maximize the possible angle between the truck and trailer when reversing by
only using the E-Trailer.
• Minimize number of cooling medias.
• Minimize length of cables and cooling hoses.
• Maximize regenerated energy in slopes.

I
A. Appendix A

II
 B
Appendix B

Press release from ZF regarding their E-Trailer

III
B. Appendix B

IV
 C
Appendix C

List of motors and batteries at Volvo Group

V
C. Appendix C

VI
 D
Appendix D

Figures of propeller shaft angles and distances under

different scenarios.
Drive axle during full bump, 90mm, and pusher at ride height.

Figure D.1: Side view.

Figure D.2: Top view.

Both drive and pusher axle at normal ride height.

VII
D. Appendix D

Figure D.3: Side view.

Figure D.4: Top view.

VIII
 E
Appendix E

Free body diagrams and corresponding equations for

different combinations of lifted axles.

E.1 Second and third axle lifted

Figure E.1: Free body diagram of the trailer with second and third axle lifted.

Derived equations from free body diagram E.1.

F1 l1 F2 l2 F3 3l2 l3
− + + + F4 (2l2 + ) − Fkp (l3 + 2l2 ) − FN l2 = 0
2 2 2 2

FN + Fkp − (F1 + F2 + F3 + F4 ) = 0

IX
E. Appendix E

E.2 First and second axle lifted

Figure E.2: Free body diagram of the trailer with first and second axle lifted.

Derived equations from free body diagram E.2.

F1 l1 F2 l2 F3 3l2 l3
− + + + F4 (2l2 + ) − Fkp (l3 + 2l2 ) = 0
2 2 2 2

FN + Fkp − (F1 + F2 + F3 + F4 ) = 0

X
 F
Appendix F

Volvo Reports

XI
DEPARTMENT OF SOME SUBJECT OR TECHNOLOGY
CHALMERS UNIVERSITY OF TECHNOLOGY
Gothenburg, Sweden
www.chalmers.se