Diesel Hybrid Powertrain Concept Study for Long Truck Combinations Master’s Thesis in the Master’s programme Automotive Engineering ASHWINKUMAR UMASANKAR DEVANSH MEHTA Department of Applied Mechanics Division of Vehicle Engineering and Autonomous Systems Vehicle Dynamics group CHALMERS UNIVERSITY OF TECHNOLOGY Göteborg, Sweden 2014 Master’s thesis 2014:62 Space for picture Replace this square with a picture illustrating the content of the report. This picture should be “floating over the text”, in order not to change the position of the title below (clic on Format: Object: Layout, and chose “In front of text”) Instructions for use of this template Replace the yellow marked text with your own title, name etc on the first nine pages. Replace only the text and not the return-signs, comment marks [mp1] etc. Update all field in the document by choosing Edit: Select All (Ctrl A) and then clicking the F9-button. Write your report using the formats and according to the instructions in this template. When it is completed, update the table of contents. The report is intended to be printed double-sided. MASTER’S THESIS IN AUTOMOTIVE ENGINEERING Diesel Hybrid Powertrain Concept Study ASHWINKUMAR UMASANKAR DEVANSH MEHTA Department of Applied Mechanics Division of Vehicle Engineering and Autonomous Systems Vehicle Dynamics group CHALMERS UNIVERSITY OF TECHNOLOGY Göteborg, Sweden 2014 Diesel Hybrid Powertrain Concept Study for Long Truck Combinations ASHWINKUMAR UMASANKAR DEVANSH MEHTA © ASHWINKUMAR UMASANKAR, DEVANSH MEHTA, 2014 Master’s Thesis 2014:62 ISSN 1652-8557 Department of Applied Mechanics Division of Vehicle Engineering and Autonomous Systems Chalmers University of Technology SE-412 96 Göteborg Sweden Telephone: + 46 (0)31-772 1000 Cover: Duo2, ett samarbetsprojekt mellan företag och myndigheter. Duo2, A collaborative project between companies and authorities. Chalmers Reproservice / Department of Applied Mechanics Göteborg, Sweden 2014 CHALMERS, Applied Mechanics, Master’s Thesis 2014:62 V Diesel Hybrid Powertrain Concept Study for Long Truck Combinations Master’s Thesis in the Master’s programme Automotive Engineering ASHWINKUMAR UMASANKAR DEVANSH MEHTA Department of Applied Mechanics Division of Vehicle Engineering and Autonomous Systems Vehicle Dynamics group Chalmers University of Technology ABSTRACT The commercial vehicle industry faces strong challenges in the forthcoming years; such as reducing the dependency on fossil fuels, lowering energy consumption, fulfilling stricter legislation regarding emissions and also increased congestion on the roads. Thus, there is a high demand for more economically and environmentally viable transport solutions. Long haul applications are interesting for fuel consumption reduction studies, due to the fact that even 1-2% reduction in fuel consumption can result in significant savings, since the distances and the mileage in context are large. The current hybridization and downsizing of existing combustion engines for long-haul applications can lead to decreased fuel consumption, but not enough to motivate the higher cost of the hybrid powertrain. Also, it is not favoured by the market as it affects the vehicle’s longitudinal propulsion performance. Hence, longer vehicle combinations are being evaluated, as they lead to less energy consumption per unit of mass transported, because of increased cargo carrying capacity. The market will also have the advantage of saving in labor costs such as driver wages because fewer drivers are needed to transport the cargo capacity in question. The thesis aims to electrify an intermediate unit in a typical long combination, called a dolly, and simulate the resulting powertrain. Feasibility studies are conducted for packaging the resulting components from powertrain sizing. 5 different concepts are constructed for the Dolly powertrain from the selected components. Rule-based control strategies are formulated for the concepts, which have a layout comparable to that of a Parallel Hybrid. The powertrain models are combined with the vehicle dynamics model of the long combination to create a simulation platform, and typical vehicle maneuvers are studied. Results in terms of fuel consumption and performance related parameters are recorded and compared for the different concepts. Conclusions are derived from the results and the most interesting concepts for different parameters are evaluated and presented. Key words: Hybrid Powertrain, Control Strategy, Long combinations CHALMERS, Applied Mechanics, Master’s Thesis 2014:62 VI Konceptstudie av Dieselhybrid Drivlina för Långa Fordonskombinationer Examensarbete inom Master’s programme Automotive Engineering ASHWINKUMAR UMASANKAR DEVANSH MEHTA Institutionen för tillämpad mekanik Division of Vehicle Engineering and Autonomous Systems Vehicle Dynamics group Chalmers Tekniska Högskola SAMMANFATTNING Den kommersiella fordonsindustrin står inför stora utmaningar under de kommande åren: Att minska beroendet av fossila bränslen, sänka energiförbrukningen, uppfylla strängare lagstiftning för avgasemissioner och hantera ökad trängsel på vägarna. Således finns en stor efterfrågan på mer ekonomiska och miljömässigt hållbara transportlösningar. Långdistanstransporter är högintressanta för studier i bränsleförbrukningsreduktion. Även en minskning av bränsleförbrukningen med 1-2% leder till betydande besparingar eftersom avstånd och körsträcka i sammanhanget är stora. Dagens hybriddrivlina med elektrisk motor, energiåtervinning till ett batteri och nedskalning av den befintliga förbränningsmotorn leder till minskad bränsleförbrukning även på långdistansapplikationer men inte tillräckligt för att motivera den högre kostnaden för hybriddrivlinan. Konceptet efterfrågas inte heller av marknaden eftersom det i vissa situationer påverkar fordonens framdrivningsprestanda. Längre fordonskombinationer studeras av flera fordonstillverkare då de ger lägre energiförbrukning per massenhet transporterat gods på grund av ökad lastkapacitet. Ytterligare fördelar är minskade arbetskostnader, för t.ex. förarlöner eftersom färre förare behövs för att transportera en given godsmängd samt ger även mindre trängsel på vägarna. Längre fordonskombinationer innebär dock större krav på stabilitet och körbarhet. En del av fordonet man studerar är en s.k. ”dolly”, en liten släpvagn som används som styraxel till en semitrailer och då förvandlar semitrailern till en släpvagn. En reglerbar dolly skulle kunna bidra med ökad stabilitet, startbarhet och vändbarhet för den långa kombinationen. Detta examensarbete syftar till att elektrifiera en dolly och simulera den resulterande drivlinan. En packningsstudie har genomförts, med komponenter som idag finns tillgängliga inom Volvo. Fyra olika hårdvarukoncept har konstruerats med de valda komponenterna. Styrstrategier har formulerats för den kompletta fordonskombinationen vars layout blir jämförbar med en parallellhybrid. Drivlinemodeller har sedan kombineras med en fordonsdynamisk modell av fordonskombinationen för att skapa en simuleringsplattform där typiska fordonsmanövrer studeras map stabilitet och körbarhet. Bränsleförbrukning och prestandarelaterade parametrar har simulerats och utvärderats för de olika hårdvarukoncepten och styrstrategierna. Simuleringsresultaten och konceptens olika egenskaper utvärderas och presenteras och slutsatser dras. Nyckelord: Hybriddrivlina, kontrollstrategi, Långa kombinationer CHALMERS, Applied Mechanics, Master’s Thesis 2014:62 VII Contents ABSTRACT V SAMMANFATTNING VI CONTENTS VII ACKNOWLEDGEMENTS X PREFACE XI About Volvo Group Trucks Technology XI About Powertrain Engineering XI About Powertrain Control Systems XI About Drivelines & Hybrids XI About Chassis & Vehicle Dynamics Engineering XI About Chassis Strategy & Vehicle Dynamics XII NOTATIONS XIII 1 INTRODUCTION 1 1.1 Background 1 1.2 Vehicle configuration 3 1.2.1 Dolly 3 1.2.2 A-double combination 4 1.3 Problem Definition 5 1.4 Objective 6 1.5 Limitations 6 2 LITERATURE REVIEW 7 2.1 Hybrid Powertrain Classification 7 2.1.1 Hybrid System functions 7 2.1.2 Rate of hybridization 7 2.1.3 Power/Energy flow through a hybrid driveline: 9 2.2 Driveline Topology 11 2.2.1 AMT 11 2.2.2 ISAM Configuration 11 2.3 Energy Storage System 12 2.3.1 Battery Selection and Sizing 12 2.3.2 Influence of Depth of Discharge on Battery Degradation 13 2.3.3 Energy Throughput 13 2.4 Energy Management & Control Strategies 14 2.4.1 Heuristic Control Strategies 15 2.4.2 Optimal Control Theory 19 CHALMERS, Applied Mechanics, Master’s Thesis 2014:62 VIII 2.4.3 Battery Discharging Strategies 20 2.4.4 GPS-based Transmission Control 21 2.5 Performance standards for Long Heavy Vehicle Combinations 24 2.5.1 Rearward Amplification 24 2.5.2 Low Speed Swept Path 24 2.5.3 High Speed Transient Offtracking 25 2.5.4 High Speed Steady-state Offtracking 26 3 METHODOLOGY 27 4 VEHICLE SPECIFICATIONS 28 5 VEHICLE SIMULATION MODELS 29 5.1 Powertrain Model (GSP) 30 5.2 Vehicle Transportation Models (VTM) 30 6 DRIVING CYCLES & VEHICLE MANEUVERS 31 7 CONCEPT STUDY 34 7.1 Packaging Study 34 7.2 Powertrain Sizing 35 7.3 Concept Definition 36 8 CONTROL STRATEGIES 37 8.1 Hybrid powertrain control functionalities 37 8.1.1 Torque Distribution Control 37 8.1.2 Torque Abilities EM 37 8.1.3 Charge Balance Control 38 8.2 Sailing 41 8.3 Ecoroll 42 9 RESULTS 43 9.1 Longitudinal Simulations 43 9.1.1 Fuel Consumption 43 9.1.2 Energy Throughput 47 9.1.3 Regenerated Energy 48 9.1.4 Performance & Drivability 49 9.1.5 Comparison with 32ton conventional truck 53 9.2 Lateral Simulations 60 9.2.1 High Speed Turn 60 9.2.2 High Speed Lane Change 61 9.2.3 Low Speed Roundabout Maneuver 63 9.2.4 Custom Roundabout Maneuver 64 CHALMERS, Applied Mechanics, Master’s Thesis 2014:62 IX 10 DISCUSSION 66 10.1 Longitudinal Simulations 66 10.1.1 Fuel Consumption 66 10.1.2 Energy throughput 67 10.1.3 Regenerated energy 67 10.1.4 Performance & Drivability 67 10.1.5 Comparison with 32ton conventional truck 68 10.2 Lateral Simulations 69 10.2.1 High Speed Turn 69 10.2.2 High Speed Lane Change 69 10.2.3 Low Speed Roundabout Maneuver 69 10.2.4 Custom Roundabout Maneuver 70 11 CONCLUSIONS 71 12 FUTURE WORK 74 13 BIBLIOGRAPHY 75 14 APPENDICES 77 14.1 Concept Study 77 14.1.1 Legend 77 14.1.2 Concept 1 78 14.1.3 Concept 2 78 14.1.4 Concept 3 79 14.1.5 Concept 4 79 14.2 Packaging feasibility study 80 CHALMERS, Applied Mechanics, Master’s Thesis 2014:62 X Acknowledgements This report is the result of the master thesis work carried out in the Electrical Propulsion Controls department of Volvo Group Trucks Technology, towards the partial requirement for the Master’s degree in Automotive Engineering at the Department of Applied Mechanics at Chalmers University of Technology. The thesis was performed from February 2014 till August 2014, at the Volvo Group Trucks Technology headquarters in Lundby, Gothenburg. We would like to thank Andreas Roupé, Group Manager, Electric Propulsion Controls, for giving us this opportunity and all the necessary resources for carrying out this thesis work in his group. We would like to thank Kristoffer Rydquist and Mattias Åsbogård of the same group for volunteering to be our supervisors for this thesis work, for without their support and insight, this level of progress with the thesis work would not have been possible. We would also like to thank our examiner Bengt Jacobson and supervisor Sixten Berglund for supporting us as and when required. Last but not the least, we would like to express our gratitude to the various personnel at Volvo who offered advice and invaluable feedback on the thesis work, such as Andreas Jansson, Mattias Hansson, Rasmus Andersson, Sachin Janardhanan, and Leo Laine. This thesis work has been a great learning experience for us, and has provided us a very good foundation for our professional career. The time we spent interacting with the various employees and infrastructure has helped us cultivate a sound analytical and technical frame of mind. Göteborg September 2014 Ashwinkumar Umasankar Devansh Mehta CHALMERS, Applied Mechanics, Master’s Thesis 2014:62 XI Preface About Volvo Group Trucks Technology Volvo Group Trucks Technology (GTT) is a worldwide entity supporting the Group Trucks and Business Area's within the Volvo Group. It provides state-of-the-art research, engineering, product planning and purchasing excellence to final delivery of complete products and also supports the products in the aftermarket [12]. The GTT organization is spread across the world. Its 10 000 employees are multi- brand, multi-cultural and work in global teams on international projects. About Powertrain Engineering Powertrain Engineering is a global organization within Volvo GTT with 2000 colleagues in six countries; Brazil, France, Japan, USA, India and Sweden. The scope of work includes the engineering and design of engines, gearboxes and axles for Volvo Group customers. About Powertrain Control Systems Powertrain Control Systems, within Powertrain Engineering is globally responsible for the design of all powertrain embedded electronic systems. It includes HW and SW developments. Engine, after treatment and transmission management systems are the most well- known electronic platforms developed and delivered to all group products and installations. Control Systems contributes to the innovation assets of the Volvo Group by providing competences and technical solutions in the area of controls theory, advanced driveline control systems and support for new technology such as E- mobility and alternative fuels. About Drivelines & Hybrids The Driveline and Hybrid technology area is responsible for systems and components within transmission and electro-mobility development. This includes; transmissions, clutches, propeller shafts driven axles, Energy Storage System and Motor Drive System. This Technology Area is also responsible for the platform development of these products. Transmissions are in-house developed. There are two platform sites directly linked to the technology area that perform the core component development and three applications sites linked to the sites that install the systems into the vehicles. About Chassis & Vehicle Dynamics Engineering Chassis & Vehicle Dynamics Engineering is a global organization with highly skilled teams located in six countries. They propose and develop profitable and competitive technical solutions for each truck company in the Volvo Group. CHALMERS, Applied Mechanics, Master’s Thesis 2014:62 XII About Chassis Strategy & Vehicle Dynamics This department of Chassis strategy & Vehicle Dynamics is responsible for performing the following functions: • Support development of mid and long term content in Technology Strategies, Technology Roadmaps and required activities in the CVDE AE Portfolio. • Develop, secure and verify complete vehicle features (i.e. active safety, durability, transport efficiency) by use of analysis. • Establish and support a structured approach to differentiated targets on modules/systems for different brand requirements and operating conditions. This Master thesis project involved interaction with both Drivelines & Hybrids and Powertrain Control Systems, due to the nature and scope of the thesis work. The Drivelines & Hybrids department provided support for the powertrain packaging feasibility study and also provided data for the various components. The Electric Propulsion Control team within Powertrain Control systems provided support on simulation models and control strategies. Also, the Complete Vehicle Control team under the Chassis Strategy & Vehicle Analysis department offered support for the vehicle dynamics simulation models and provided information about the important vehicle performance criteria used in the thesis work. CHALMERS, Applied Mechanics, Master’s Thesis 2014:62 XIII Notations Abbreviations AE Advanced Engineering AER All Electric Range AMT Automated Manual Transmission AMT PS Automated Manual Transmission Power Shift BLB Boras Landvetter Boras CAD Computer Aided Design CDCS Charge Depleting Charge Sustaining DoD Depth of Discharge EB Energy Buffer ECU Engine Control Unit EM Electric Motor EMS Energy Management System ESS Energy Storage System EV Electric Vehicle GPS Global Positioning System GSP Global Simulation Platform GTA Global Transport Application HCU Hybrid Control Unit HEV Hybrid Electric Vehicle HW Hardware IC Internal Combustion ICE Internal Combustion Engine IMF International Monetary Fund ISAM Integrated Starter Alternator Motor ISG Integrated Starter Generator KE Kinetic Energy MDS Motor Drive System MHD Medium Heavy Duty MHEV Mild Hybrid Electric Vehicle PBS Performance Based Standards PE Potential Energy PHEV Plug-in Hybrid Electric Vehicle SoC State of Charge SW Software TCU Transmission Control Unit VTM Vehicle Transportation Models CHALMERS, Applied Mechanics, Master’s Thesis 2014:62 XIV CHALMERS, Applied Mechanics, Master’s Thesis 2014:62 1 1 Introduction 1.1 Background The fuel prices in the world today are about 4 times of that in 2000 [16]. For example, the cost of gasoline in Sweden has doubled in the past 20 years [17]. Regarding future price of fuel, a study by IMF [1] predicts an increase in fuel price of 80% over the next 10 years, whereas a similar study by Volvo [Volvo Internal] estimates a more conservative figure of 30.Thus, there is a need to develop vehicles that consume less fuel per kilometer of distance travelled, in order to have cost-efficient transport [1]. Moreover, heavy-duty trucks are the largest consumers of fuel per year, as seen in the statistics below. This is because they provide the highest fuel consumption per kilometer of distance travelled (although they have the highest fuel economy per ton Figure 1: Volvo estimated fuel price increase [Volvo Internal] Figure 2: IMF estimated fuel price increase in USD per barrel [1] CHALMERS, Applied Mechanics, Master’s Thesis 2014:62 2 per km and highest thermal efficiency) and also have the highest mileage per year. Hence, a reduction in fuel consumption of even as little as 5% in long-haul heavy duty truck applications can have a great effect on the overall fuel consumption, running costs, and emissions at a macroscopic level. Figure 3: Mileage and Fuel economy of Vehicles in the US Market Figure 4: Fuel Consumption of Vehicles in the US Market CHALMERS, Applied Mechanics, Master’s Thesis 2014:62 3 1.2 Vehicle configuration This thesis deals with long truck combinations, which can transport a much larger cargo than conventional single trailer trucks. In order to have long truck combinations, a device called as a Dolly must be used to connect two or several trailers. There are various types of vehicle combinations present in the market or in the concept phase, but this thesis will be limited to the A-double configuration. 1.2.1 Dolly A dolly is a small trailer that can be coupled to a truck or trailer so as to support a semi-trailer. A semi-trailer is a trailer without a front axle. A large proportion of its weight is supported by a road tractor, a dolly, or the tail of another trailer. [11] The dolly consists of a bogie equipped with a kingpin and a fifth wheel, to which the semi-trailer is coupled [12]. Depending on design style, dollies may have a single- or double-tow-drawbar arrangement for coupling to the towing trailer. In either case, the tow bars terminate in a simple, rugged towing eye. The towing trailer is equipped with one or two pintle hitches consisting of a hook and locking mechanism, which engages and secures the eye(s), thereby supporting and towing the dolly. There are two types of converter dollies, which are distinguished by the number of tow bars, are illustrated in the above figure. Figure 5: Picture of a typical truck dolly Figure 6: Diagram of an A-type and C-type Dolly http://en.wikipedia.org/wiki/Tow_hitch http://en.wikipedia.org/wiki/Truck http://en.wikipedia.org/wiki/Semi-trailer http://en.wikipedia.org/wiki/Bogie http://en.wikipedia.org/wiki/Fifth_wheel_coupling CHALMERS, Applied Mechanics, Master’s Thesis 2014:62 4 1. A-dolly: The defining quality of the A-dolly is its single-point tow bar. The A- dolly is the most common type of converter dolly; over 99 percent of the dollies in use in the U.S. are of this type. The single hitching point allows the dolly to articulate in yaw (steering), pitch (fore/aft rotation), and roll (side-to- side rotation) with respect to the towing trailer. The advantage of an A-dolly is its excellent low-speed maneuverability and turning capability. However, its ability to provide yaw and roll articulation leads to considerable rear-ward amplification and dynamic roll instability. Hence, this type of dolly is more suited to low speed applications, such as city driving [2]. 2. C-dolly: The defining quality of the C-dolly is its double-tow-bar configuration. The C-dolly originated in Canada. Its attractive quality is its ability to improve the stability of multiple-trailer combination vehicles. This is accomplished because the double-tow-bar hitching arrangement eliminates yaw and roll articulation with respect to the lead trailer. Eliminating yaw, in particular, can degrade low-speed maneuverability and produce excessive hitch forces and tire scrubbing during tight turns at low speeds. To mitigate these low-speed problems, the wheels of the C dolly are allowed to steer by a caster mechanism. However, a centering mechanism provides mechanical resistance to this self-steering action as required for dynamic stability at highway speeds. Hence, this dolly type is more suited to high-speed applications, especially those involving multiple trailer combinations.[18] 1.2.2 A-double combination Table 1: Mass & Payload Distribution in an A-Double Combination Vehicle Unit 1 2 3 4 Mass (kg) 900 31000 3000 17000 Payload (kg) - 24000 - 10000 The A-double combination will be used for vehicle simulations in this thesis. This combination consists of a tractor, two identical standard semitrailers and a converter dolly. It has a total length of 31.5 m. The distance from the pintle hitch coupling to the end of the semitrailer is 1.5 m. The first semitrailer is fully laden and the rest of the payload is on the second semitrailer. Hence, the total payload for this 60ton vehicle is 34tons. It is not possible to achieve 25 % of the total weight on driving axles with even load distribution [3]. Figure 7: Diagram of an A-Double vehicle combination CHALMERS, Applied Mechanics, Master’s Thesis 2014:62 5 1.3 Problem Definition The motivation behind performing this thesis work is the potential benefits that can be achieved by hybridization of truck powertrains for long haul applications. Due to the indispensable nature and the sheer distances covered during long haul applications, a reduction of fuel consumption by even 1% can result in considerable economic and environmental savings. Increasing the total capacity of such long haul applications by increasing the length of vehicle combinations remain an interesting option for study due to increased productivity, decreased road footprint of the vehicle fleet for the same amount of cargo, and a reduction in amount of fuel required per mass of cargo transported for a mission, as well as the number of drivers required to move the same amount of cargo. An additional way of reducing the fuel consumption per mass of cargo transported per unit distance is to use an electrified dolly. An electrified dolly is basically a dolly with one or two axles driven by an electric motor(s). There is a battery pack attached to the bottom of the dolly that supplies energy to the electric motor(s). More about the electric dolly concept will be described later in Chapter 7 of the report. Previous studies have been done on electrification of the trailer unit, and although the fuel consumption reduction results were promising, they haven’t proved feasible. This is because the trailer is often not owned by the company which owns the truck and is interchanged between transporters. Electrifying the dolly, however, seems to be a commercially viable option if the dolly can be owned by the customer and can be attached to various semitrailers. An electric dolly is a modular solution, as it can be sold as an add-on product along with the truck which can be disconnected at a loading bay for in-city applications or any sort of transport that involves lower cargo capacities. Hence it is also a more flexible and adaptable solution than hybridization of either the tractor powertrain or a complete trailer. Also, it could improve startability of the combination as it would have more driven wheels. Figure 8: A-Double combination with an Electric Dolly CHALMERS, Applied Mechanics, Master’s Thesis 2014:62 6 1.4 Objective The objectives of the thesis work are encapsulated in the following points: • Construction of a complete powertrain simulation model of the vehicle combination, in order to study the vehicle’s fuel consumption and performance. • Sizing of the components of the Electrified Dolly (refer Chapter 7.2) with respect to performance requirements and driving cycle, taking basic geometrical limitations into account by means of a feasibility study. • Rule-based Control Strategies for the Hybrid Powertrain, taking into consideration the performance requirements and limitations imposed by vehicle dynamics. • Simulation results with regard to performance, fuel consumption and vehicle dynamics. 1.5 Limitations The limitations of the thesis are defined in such a way so as to make the scope of the thesis work clearer, and work towards reaching the goals while keeping in mind various factors such as geometrical constraints, feasibility, practicality and cost. They are listed out as follows: • The Packaging study to be done to determine the feasibility of packaging the propulsion components in the dolly, hence it is approximate. • Off-the-shelf propulsion and energy storage components have been used. • The energy storage system to be used is of battery type. • The long combination layout is considered as the so called A-Double configuration (refer chapter 1.2.2). • The powertrain layout is simplified to be a Parallel Hybrid layout for formulating the control strategies, with one conventional diesel engine in the truck and electric motor(s) on the dolly. • Engine-based charging mode is not possible, due to the lack of physical connection between the dolly powertrain components and the IC engine. • The power consumption required for heating/cooling the electric powertrain is considered to be negligible. • Charging from grid during cycle has not been considered. • Startability and Gradeability have not been investigated. CHALMERS, Applied Mechanics, Master’s Thesis 2014:62 7 2 Literature Review 2.1 Hybrid Powertrain Classification A hybrid powertrain is one which utilizes two different sources of energy for propulsion of the vehicle. The most common type of hybrid uses an internal combustion engine which derives energy from a fuel, and an electric motor which derives energy from a battery. Such a fuel-electric hybrid can be classified in different ways, based on: 2.1.1 Hybrid System functions Hybridization of a powertrain allows new functionalities that determine the class of HEV, such as: Start/stop: A start-stop system automatically shuts down and restarts the internal combustion engine to reduce the amount of time the engine spends idling, thereby improving fuel economy. Regenerative braking: This feature allows converting the kinetic energy of the vehicle into electrical energy by using the electric motor as a generator. This ‘free energy’ is often stored into a battery and reused later to propel the vehicle. In an HEV, this feature is fundamental and generates a major part of the fuel savings. Torque assist: The electric energy contained in the battery can be used to propel the vehicle through the electric motor. In case of torque assist, the power delivered by the ICE is reduced by an amount equal to the power supplied by the electric machine. Thereby the vehicle achieves the same performance as a conventional would. Boost: This feature differs from the torque assist in a sense that the sum of the ICE torque and EM torque exceeds the maximum torque capacity of a conventional ICE powered vehicle. Thus the vehicle can achieve better accelerations than a non- hybridized drivetrain. Electric only: In electric only mode, the hybrid vehicle is driven by the electric machine only. Meanwhile the engine can be turned off. This feature is particularly useful at low speeds when ICE has low efficiency. Also, the power needed at these low speeds is low enough to be within the EM’s maximum power limit. In consequence, no fuel is consumed and the vehicle achieves a zero-pollution level. Note also that shutting down the engine considerably reduces the noise generated by the vehicle (which makes it suitable for city driving, amongst other things). Plug to grid: Such a feature offers the possibility to plug a hybrid vehicle directly onto the electrical power grid. The batteries can thereby be charged when they are nearly depleted. This allows the vehicle to run longer in Electric mode, Boost mode and Torque assist mode as compared to a non-plugin hybrid. [15] 2.1.2 Rate of hybridization The contribution of the electric machine to the propulsion of the vehicle can be quantified through the rate of hybridization. Rate of hybridization (Hr) is a measure used to describe how strongly a parallel powertrain is hybridized. It is defined in the equation below as the ratio of maximal electric power to the maximal power deliverable by the powertrain and is often expressed as a percentage [Volvo Internal]. 𝐻𝐻 = 𝑃𝐸𝐸𝑚𝑚𝑚 𝑃𝐼𝐼𝐼𝑚𝑚𝑚 + 𝑃𝐸𝐸𝑚𝑚𝑚 ∗ 100 CHALMERS, Applied Mechanics, Master’s Thesis 2014:62 8 Table 2: Common Rates of Hybridization Type Rate of hybridization (%) EV Hr = 100 PHEV Hr > 50 HEV 25 < Hr < 50 MHEV 10 < Hr < 25 µHEV Hr < 10 Conventional Hr = 0 Table 3: Examples of hybrid vehicles in the market with their rate of hybridization [Volvo Internal] Vehicle PEM max(kW) PICE max(kW) PICE max(hp) Hr (%) Type Toyota Prius II 50 58 78 46 HEV Toyota Prius III 60 73 99 45 HEV V ol vo H yb rid s City Bus SD-DD 120 160 215 42 HEV Refuse Truck EU 120 246 330 32 HEV Distribution Truck EU 120 223 300 34 HEV Mack Truck US 120 242 325 33 HEV Long Haul 120 120 343 460 25 MHEV Long Haul 60 60 343 460 14 MHEV Figure 9: Examples of typical Rates of Hybridization for different types of vehicles CHALMERS, Applied Mechanics, Master’s Thesis 2014:62 9 2.1.3 Power/Energy flow through a hybrid driveline: There are three main types depending on the layout [4]: 1. Series Hybrid A series hybrid is one in which only one energy converter can provide propulsion power. The IC Engine acts as a prime mover in this configuration to drive an electric generator that delivers the power to the battery or another form of energy storage, and the propulsion motor. Figure 10: Layout of a Series Hybrid Powertrain 2. Parallel Hybrid A parallel hybrid is one in which more than one energy conversion device can deliver propulsion to the wheels. The IC Engine and the electric motor are configured in parallel with a mechanical coupling that blends the torque coming from the two sources. Hence, the IC Engine and electric motor can be used either simultaneously or separately to meet the power demand. Figure 11: Layout of a Parallel Hybrid Powertrain CHALMERS, Applied Mechanics, Master’s Thesis 2014:62 10 3. Power-split Hybrid In a Power-split hybrid, the power flow from the engine to the wheels can either be mechanical or electrical. Hence, the IC Engine can be also used to charge battery through the electric motor, and then send power to the wheels through the electric motor. Thus, it has the functionalities of both a series and parallel hybrid. Figure 12: Layout of a Power-split Hybrid Powertrain CHALMERS, Applied Mechanics, Master’s Thesis 2014:62 11 2.2 Driveline Topology 2.2.1 AMT AMT stands for Automated Manual Transmission. It is a standard gearbox with a dry clutch, but the clutch pedal and the gear lever is removed. The control of the clutch and the selection of the proper gears are done by software and actuators. The gearbox consists of an unsynchronized mainbox with three forward and one reverse gear, a synchronized range gearbox which splits every gear into two fairly similar gears, and a synchronized range gearbox which gives every gear a high speed and a low speed range. So, combining these three gearboxes result in a gearbox with a total of 12 forward gears and 4 reverse gears. 2.2.2 ISAM Configuration In the ISAM (Integrated Starter Alternator Motor) layout, the electric motor is placed between the clutch of the ICE and the transmission input shaft. A special electric motor is used, which has a large diameter but short in length. This provides a shorter but wider powertrain layout. The electric motor has to be placed between the gearbox and the extra-clutch, and is always connected to the driveline. It is possible to connect the ISAM directly on the ICE (without a clutch) but this solution (sometimes referred to as ISG) has limited functionalities since it is impossible to physically disconnect the electric motor from the engine. Figure 13: Layout of an ISAM configuration CHALMERS, Applied Mechanics, Master’s Thesis 2014:62 12 2.3 Energy Storage System An energy storage system is one which is used to store, charge and discharge energy as and when required. They are usually either batteries or supercapacitors in automotive applications. The energy storage type has been limited to batteries in this thesis. Batteries are devices that transform chemical energy to electrical energy and vice versa [5]. Desirable attributes of traction batteries for EV and HEV applications are high specific power, high specific energy, long calendar and cycle life, low initial and replacement costs, high reliability, and high robustness. Among other current technical challenges, a key point is developing accurate techniques to determine the capacity or the state of charge (SoC) of batteries during their operation [5]. The capacity of a battery, usually expressed in Ah, is the integral of the current that can be delivered under certain conditions. A dimensionless parameter is the state of charge, which describes the amount of charge remaining in the battery, expressed as a percentage of its nominal capacity. Another key design parameter is the specific energy, i.e., the energy that can be stored in the battery per unit mass, typically expressed in Wh/kg [5]. Lithium-ion batteries have been used as the ESS in this thesis. Their high specific energy and specific power make them suitable for HEVs. They have a carbon based anode, usually made up of graphite, in which lithium ions are intercalated in the interstitial spaces of the crystal. Hence, the cathode is a lithium oxide, and the electrolyte is a lithium salt solution. The cell voltage is usually 3.6 V [5]. 2.3.1 Battery Selection and Sizing Battery selection can be done based on: power discharge capacity, energy capacity, allowed volume, allowed mass, load profile, price, nominal voltage, resistance, current charge and discharge. A cost function is minimized in order to select the most adequate battery. For a plug-in hybrid, the battery can be sized such that it has enough capacity to be able to complete a certain number of transport tasks per day. 𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡 ∗ 𝑤𝑤𝑤𝑤𝑤𝑤𝑤_𝑑𝑑𝑑𝑑𝑝𝑝𝑝𝑝𝑝𝑝𝑝 ∗ 𝑡𝑡𝑡𝑡_𝑙𝑙𝑙𝑙𝑙ℎ = 𝑎𝑎𝑎𝑎𝑎𝑎_𝑚𝑚𝑚𝑚𝑚𝑚𝑚 Figure 14: SoC versus Time for a battery in a typical hybrid vehicle application [Volvo Internal] CHALMERS, Applied Mechanics, Master’s Thesis 2014:62 13 2.3.2 Influence of Depth of Discharge on Battery Degradation Battery degradation increases with the amount of energy flowing through the battery, i.e. the total and maximum energy throughput during the battery lifetime. The energy throughput strongly depends on the depth of discharge. The DoD is the difference between the minimum and maximum value of SoC in a battery during the driving cycle of an electrified vehicle. The lifetime of a battery Ncmax (DoD) is defined as the maximum number of charge cycles after which the battery capacity reduces to less than 80% of the original capacity, when run at a certain depth of discharge. The battery life greatly depends on the DoD, as show below: 2.3.3 Energy Throughput The total energy throughput of a battery is the total amount of energy in watt-hours that is put into and taken out of the battery over all the cycles in its lifetime. It can also be measured for each operation of the vehicle, i.e. for each driving cycle. The average energy throughput (kWh/h) of a battery is the amount of energy that in average is stored and released during a driving cycle [Volvo Internal]. It can be calculated as below: 𝐴𝐴𝐴𝐴𝐴𝐴𝐴 𝐸𝐸𝐸𝐸𝐸𝐸 𝑇ℎ𝑜𝑜𝑜𝑜𝑜ℎ𝑝𝑝𝑝 (𝐴𝐴𝐴) = ∫ |𝑉∗𝐼|𝑡 0 1000∗𝑡 kWh/h Where, V= Battery Voltage (Volts) I= Battery Current (Amperes) t= Time taken to complete driving route (hours) The maximum energy throughput of the battery changes as the battery life decreases due to wear as shown below: The total energy throughput of a battery can be calculated by the formula: Figure 15: Battery Life versus Depth of Discharge [Volvo Internal] Figure 16: Maximum Energy Throughput versus Battery Life [Volvo Internal] CHALMERS, Applied Mechanics, Master’s Thesis 2014:62 14 𝐸𝑙𝑙𝑙𝑙 = 𝑁𝑐𝑐𝑐𝑐𝑐𝑐 𝑚𝑚𝑚(𝐷𝐷𝐷) ∗ 𝐸𝑏𝑏𝑏𝑏 ∗ 𝐷𝐷𝐷 The maximum energy throughput of a battery can be calculated by the formula: 𝐸𝑡𝑡𝑡𝑡𝑡(𝐷𝐷𝐷) = 2 ∗ 𝑉 ∗ 𝑄 ∗ 𝑁𝑐𝑐𝑐𝑐𝑐𝑐 𝑚𝑚𝑚(𝐷𝐷𝐷) ∗ 𝐷𝐷𝐷 2.4 Energy Management & Control Strategies An energy management strategy (EMS) is required in a hybrid powertrain as it has more degrees of freedom than a conventional one, and should utilize the IC Engine, electric motor, and battery in the most energy efficient manner. The basic idea behind most control strategies is to be able to run the IC Engine at its most efficient point with the help of the electric motor, to efficiently regenerate energy, and to be able to supply sufficient power to the wheels in order to maintain the required performance. Also, the ICE can be shut off completely if enough energy and power can be provided by the EM. The various types of energy management strategies can be classified as follows: Table 4: Control Strategies and their description [Volvo Internal] Aspect Description Rule or model based Examples of rule-based methods are neural networks and fuzzy logic. Model based methods uses a mathematically (state space formulization) described plant model to derive control signals. Predictive or non- predictive Predictive strategies need future information, for example the predicted vehicle power need. The future information can be guessed or derived from an ‘intelligence gathering system’, for example a camera. Non-predictive strategies just use the present information. Length of prediction horizon Full or partial. Seconds, minutes or hours. Figure 17: Plots of Maximum number of cycles and Maximum Energy Throughput versus Depth of Discharge [Volvo Internal] CHALMERS, Applied Mechanics, Master’s Thesis 2014:62 15 Minimization function Fuel, wasted energy, battery wear, or combinations of these are examples of possible objective functions. Hybrid or plug-in hybrid If grid power is fed into a hybrid vehicle, it is a plugin hybrid. In such a case a charge depleting control strategy is needed. It means that the battery energy level in the end of transport task shall be lower than in the beginning. Adaptive or non-adaptive An adaptive control automatically adapts its behaviour during vehicle operation. One can for example think of a strategy that uses less battery power if the battery health is bad. This aspect is a bit vague and much coupled to if a control is model based or not. Updating a parameter, for example battery capacity, in the plant model of a model-based strategy will make it adaptive. The EMS can split the demand for traction power between the ICE and Energy Buffer (EB), based on three signals: SoC, desired vehicle velocity (v), and demand for traction power (Pdem). The difference between the demanded tractive power and power supplied by the ICE can be used to charge/discharge the battery [6]. 2.4.1 Heuristic Control Strategies The strategies described below are relatively simple and thus easy to implement as they are rule-based, and can lead to a significant reduction in fuel consumption. 2.4.1.1 Heuristic Energy Management In the heuristic methods, the energy management controller is often put together directly using engineering intuition as a set of rules. Even if these kinds of controllers lack a guaranteed level of optimality; the intuitive approach remains as a common way to design energy management controllers. The heuristic control strategy described below has been implemented in the Mild Hybrid Long Haul project in 2009. In this strategy, the power split between the ICE and the EM is calculated by comparing the current SoC to a reference SoC value. The reference SoC calculation is based on the vehicle speed as shown in the figure below. The basic principle of this curve is to: a) Keep the SoC low when the vehicle speed is high since a braking phase is likely to happen later (and therefore be prepared to charge the battery). The goal here is to recover as much energy as possible when the vehicle decelerates. b) Maintain the SoC at high value when the vehicle speed is low to provide EM torque assist to the ICE with the electric machine and therefore deplete the battery during acceleration to a higher speed. CHALMERS, Applied Mechanics, Master’s Thesis 2014:62 16 However this control strategy can be overridden in some occasions. For example during braking phases, the energy recuperation is prioritized. Therefore the controller will aim at filling up the battery with as much free regenerated energy as possible. The recuperated energy is then used to propel the vehicle and feed the electrical auxiliaries. Note that this strategy does not aim at improving the performance of the vehicle: the hybrid vehicle is supposed to achieve the same performance as the equivalent conventional vehicle [Volvo Internal]. The reference SoC is used in the rule based controller to split the energy contained in the fuel tank and in the battery. This description is based on an example presented on figure 25 below and divided into 4 cases: Case 1: braking demand Case 2: propelling demand and SoC > SoCref Case 3: propelling demand and SoCref > SoC > SoCwin_min Case 4: propelling demand and SoC = SoCwin_min Figure 19: Plots of Vehicle Speed, Torque and SoC versus Time [Volvo Internal] Figure 18: Look-up plot of Reference SoC versus Vehicle Speed [Volvo Internal] CHALMERS, Applied Mechanics, Master’s Thesis 2014:62 17 Case 1 When braking power is required, regeneration is prioritized. Therefore the reference SoC is totally disregarded in this case. The goal is to maximize the energy regeneration and to charge the battery. Note that, if the SoC comes to the maximum SoC window (60% for example), the braking power is not supplied by the electric machine anymore but by the service brake. This phenomenon is referred as ESS saturation. Case 2 In this case, the vehicle power demand requires propelling the vehicle. Since the energy level in the battery is higher than the reference value, the EM will be partly used to propel the vehicle. The EM power demand is given by the following formula: An example of this control strategy is given on the figure 27 below with SoCref = 35%. The higher the SoC, the more power will be used by the EM. This way the battery will be depleted faster at high SoC and slower at low SoC. Figure 20: [Volvo Internal] EM power demand as a function of SoC for SoCref = 35% when SoC > SoCref The ICE torque is the result of the difference between the torque supplied by the electric machine and the total torque demanded. This ensures that the mild hybrid vehicle achieves the same acceleration performance of a conventional vehicle. As a result, the battery is depleted and a significant amount of fuel is saved. This kind of phase is referred to as Torque assist. The energy contained in the battery is also used to feed the electrical auxiliaries. Case 3 The goal here is to keep depleting the battery but slower than in case 2. The electric machine is not used anymore to deplete the battery but the electrical auxiliaries are still fed by the battery. This will make the energy level decrease in the battery. Since the electric machine is not used at all to propel the vehicle, the powertrain demand is only fulfilled by the ICE. CHALMERS, Applied Mechanics, Master’s Thesis 2014:62 18 Case 4 When the SoC has reached the lower limit of the SoC window (typically 30%), it has to be maintained at this level for several reasons: a) The first reason to avoid falling below the lower limit of the SoC window (below SoCwin_min) is that exceeding the SoC window will hasten the ageing process of the battery and reduce therefore its life length. b) The other reason is to be able to recover as much energy as possible if a braking phase would occur. To make this SoC control possible, the electric machine is driven by the ICE to supply the electric power consumed by the electric auxiliaries and the idle losses of the electric machine. In this way, the SoC in the battery remains constant. These phases should be avoided since the electrical energy is provided by the ICE through the electric machine. The global efficiency of the powertrain is known to be low since the EM works at low-efficiency operating points and thus energy conversion losses will occur [Volvo Internal]. The control strategy described above had been partly adopted in the controller developed later in the thesis, in the sense that it has been also designed around trying to follow a reference SoC and using the ICE to provide the torque difference between the EM and the total demand. The feature which allows the ICE to charge the batteries via the EM will not be used in the thesis as the ICE and EM are not connected to each other in the selected vehicle configuration. 2.4.1.2 Pure Electric Propulsion at Low Torque The EM is used to propel the vehicle when torque demand is below a certain threshold. The purpose of this is to avoid using the ICE at low loads, where the efficiency is low. However, the system is usually underused as the torque demand in heavy vehicles is seldom below ~400Nm and thus there are no considerable fuel savings. Hence, this strategy might be of more use in passenger vehicles. 2.4.1.3 Sailing Electric energy is used to propel a vehicle after a downhill regeneration phase. The ICE engine is also kept disconnected by disengaging the clutch or selecting the neutral gear in order to avoid engine friction losses (a function called Ecoroll in the conventional application), until the vehicle goes below a certain velocity threshold. The EM keeps propelling the vehicle even above the target speed in order to extend the Ecoroll phase. This strategy leads to a considerable decrease in fuel consumption of about 5% as compared to a conventional vehicle without Ecoroll. This is expected since engine internal friction leads to almost 20kW of power loss at cruising speed, and 9kW at idle speed [1]. 2.4.1.4 Shift strategy to increase regeneration power in downhill driving The basic strategy is to downshift during a slope in order to increase the EM speed, and thus increase the amount of power regenerated. The transmission can be upshifted again when the acceleration pedal is pressed or the brake is released. The downshift should be avoided if the ESS is saturated, if the brake demand is less than the maximum EM torque, or if the braking phase is too short (in order to avoid frequent gear changes) [Volvo Internal] CHALMERS, Applied Mechanics, Master’s Thesis 2014:62 19 2.4.2 Optimal Control Theory Optimal control theory deals with the problem of finding a control law for a given system such that a certain optimality criterion is achieved. A control problem includes a cost function J that is a function of state variables x(t) and control variables u(t). The optimal control policy can be derived using Pontryagin’s minimum principle (necessary condition), or by solving the Hamilton-Jacobi-Bellman equation (sufficient condition) [Volvo Internal]. 2.4.2.1 Optimal controller This controller assumes that the driving cycle is exactly known. It solves an optimal control problem where the optimal control signal is used to minimize the fuel consumption over a given drive cycle. Numerical methods such as dynamic programming can be used to solve this problem. However, several significant drawbacks make this controller inapplicable to a real vehicle: a) Computational time: Dynamic programming is well-known for needing huge computational resources especially when the time-horizon is large and when the number of state variable increases (also known as the curse of dimensionality). It turns out that the implementation of such controllers into current ECU’s is almost impossible. b) In real driving, most of the time there is no information available of the route ahead, or at least not an exact prediction of the full drive cycle which makes this controller inapplicable to a real vehicle. For instance, it is not possible to predict a drop in speed due to an accident on the road or varying traffic situations. Despite these drawbacks, optimal control is very valuable. Since the drive cycle is fully known in a simulation environment and the time-horizon finite, this method can be used to assess the optimal fuel consumption. Even if this controller is not practically implementable, it has the strength to provide a reference lowest possible fuel consumption that can be used to benchmark other controllers [7]. 2.4.2.2 Acausal optimal controller An acausal system is defined in as “a system that is not a causal system, i.e. one which depends on some future input values and possibly on some input values from the past or present. This is in contrast to a causal system which depends only on current and/or past input values.” This kind of controller uses prediction of the future drive cycle to control the energy in the battery. This type of controller can, for instance, anticipate the desired depletion of the buffer before a downhill (since the forthcoming downhill will allow recuperation of energy). Methods such as stochastic dynamic programming can be used to design such controllers [7]. 2.4.2.3 Causal optimal controller A causal system (also known as non-anticipative system) is a system where the output depends on past/current inputs but not future inputs. A solution to design such controllers consists in simplifying the optimal control problem into two parts: 1. Solve the static optimization defined in (25) for every possible combination of ωreq and Treq (i.e. find the optimal torque split in all cases). CHALMERS, Applied Mechanics, Master’s Thesis 2014:62 20 2. Define the equivalence factor W with a heuristic formula since no information is available ahead. Equivalence factor calculation Since no information is available ahead, the equivalence factor W has to be defined with a heuristic formula as a function of the state variables (SoC in this case). The tangent function suits to this kind of application, and thus W can be expressed as follows [7]: W (SoC) = p1*tan (p2*SoC+p3) + p4 2.4.3 Battery Discharging Strategies This section deals with various strategies that can be used to control the state of charge of the battery in such a way so as to utilize the SoC window of the battery in an efficient manner in parallel hybrid applications [7]. 2.4.3.1 CDCS Strategy Most effective PHEV’s have a battery sized to give an AER of only 20-30km. The average number of trips taken by an average commuter usually exceeds this AER, which is why it is interesting to optimally discharge the battery along the trip such that it is almost fully depleted by the end of the trip. This can be implemented in a simple way by the CDCS (Charge Depleting charge sustaining) strategy. This strategy consists of a “charge-depleting” phase during which the battery is nearly depleted, followed by a “charge-sustaining” phase during which the engine is run in order to keep the SoC around the SoCref level [7]. 2.4.3.2 Blended Cost-Optimal Strategy A cost-optimal discharge strategy can also be implemented, i.e. blended strategy can be used by applying optimal control theory. This blended strategy can achieve lower fuel costs than a CDCS strategy. However, a blended strategy is highly dependent on the trip length, as it is not cost-optimal to end the trip with a high SoC. The predicted length of the trip can be modeled as a stochastic variable and dynamic programming can be used to calculate the blended strategy which minimizes the expected total fuel cost. The main disadvantage with this strategy is that prior information about a future trip is required. Thus, the EMS can be operated in two modes based on whether or not a trip has been recognized. When no trip has been recognized, CDCS can be used, Figure 21: Plot of SoC versus Distance for both the CDCS and Blended Strategy CHALMERS, Applied Mechanics, Master’s Thesis 2014:62 21 where SoCref= SoClow. When a trip is recognized, the blended mode can be used, and the reference SoC is calculated as [7]: 𝑆𝑆𝑆𝑟𝑟𝑟(𝑡) = 𝑆𝑆𝑆(𝑡∗) −𝑚𝑚𝑚 �1, 𝑧𝑐(𝑡) − 𝑧𝑐(𝑡∗) �̂� − 𝑧𝑐(𝑡∗) � ∗ �𝑆𝑆𝑆𝑚𝑚𝑚 − 𝑆𝑆𝑆(𝑡∗)� Where, 𝑡∗: 𝑡𝑡𝑡𝑡 𝑤ℎ𝑒𝑒 𝑡𝑡𝑡𝑡 𝑖𝑖 𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟 𝑧𝑐(𝑡) = 𝑧(𝑡) + ∆𝑧∗(𝑡): corrected vehicle distance position along route To prevent overestimation of trip length, the trip length reference is set to �̂� = 𝑧̅ − 2𝜎𝑧 The overall advantages of using such a blended strategy are: • It reduces average battery current (thus leading to higher battery life). • It avoids ICE-based charging (thus leading to lower fuel consumption). • Its benefit will increase with battery aging, as the resistance losses increase with age. 2.4.4 GPS-based Transmission Control ZF Friedrichshafen AG has developed the Prevision GPS transmission shift program for the Traxon transmission system. This program aims to enable gear selection on any route in the same anticipatory way as an experienced truck driver with excellent route knowledge. The system also contributes to reducing driver fatigue and improving driving safety. The program requires a constant feed of GPS data, which includes not only street maps, but also information such as speed limits, traffic signs, traffic signals, curves, roundabouts, and topography [14]. The ECU prepares topographic information for the oncoming road section, and transmits messages to the TCU. Various evaluation criteria are then used by the Prevision GPS program in order to derive the optimal shift strategy. The implementation of this system can lead to two major advantages: Figure 22: Schematic of a GPS-based Transmission system [14] CHALMERS, Applied Mechanics, Master’s Thesis 2014:62 22 1. Avoiding of unnecessary and frequent gear shifts. The Prevision program prevents an upshift when an uphill gradient is directly ahead, or if the vehicle is passing through a short section of flat road during long uphill tours. Thus, this system prevents loss of tractive force and speed which normally occurs due to frequent gear changes and hence leads to additional fuel savings. Fewer gear changes also leads to reducing wear of the transmission and clutch system. Additionally, the program accepts a short-term lower vehicle performance (such as by selecting a higher gear before the end of an uphill gradient, if a flat/downhill section is to follow) in order to keep the engine running at more efficient operating points [14]. 2. Truck rolling function This function allows the kinetic and potential energy of the vehicle to be exploited under certain conditions in order to reduce fuel consumptions and CO2 emissions. For example, this function enables the truck to drive from a downhill to flat gradient with an open driveline, which prevents loss of the vehicle’s kinetic energy by engine friction. However, this would not be beneficial if an uphill gradient was to follow. Figure 23: Plot of Gear position versus Actual speed with and without Prevision GPS [14] Figure 24: Plot of Neutral gear position versus Actual speed with and without Prevision GPS [14] CHALMERS, Applied Mechanics, Master’s Thesis 2014:62 23 Thus, this system uses the available topography information in order to calculate when and for how long this rolling function should be activated. Moreover, the combination of Prevision GPS with intelligent cruise control can be used to provide for speed reductions and rolling before driving through roundabouts or road-signs. The above described functions can cause an average fuel saving of 2-3% and consequently a reduction in CO2 emissions [14]. CHALMERS, Applied Mechanics, Master’s Thesis 2014:62 24 2.5 Performance standards for Long Heavy Vehicle Combinations The increasing interest towards long heavy vehicle combinations due to possibilities of fuel consumption reduction and lesser road footprint, lead to a necessity to improve road safety and protect road infrastructure. Thus there exists a set of vehicle performance-based regulations which put restrictions on the vehicle design. Performance Based Standards (PBS) is an initiative introduced by the National Road Transport Commission in Australia to achieve this goal [8]. The following are definitions of important performance based characteristics that must be met by such long heavy vehicle combinations in order to enable them to participate in road transportation [9]. It is important to note that the following characteristics are the ones which are decided to be studied for the vehicle dynamics behavior within the scope of this thesis work. 2.5.1 Rearward Amplification Rearward amplification is defined as the ratio of the maximum value of the motion variable of interest (e.g. yaw rate or lateral acceleration of the center of gravity) of the worst excited following vehicle unit to that of the first vehicle unit during a specified maneuver at a certain friction level and constant speed [9]. When a sudden lateral movement is made, as in a turn, each unit in the combination experiences different lateral acceleration, and this is amplified towards the end of the vehicle. Lower values of rearward amplification imply better performing combination. 2.5.2 Low Speed Swept Path Low speed swept path is defined as the maximum width of the swept path between the outermost and innermost points of the body of the vehicle combination in a low speed turn with a certain outer radius at a certain friction level and a certain angle between entry and exit [9]. Figure 25: Rearward Amplification of a vehicle combination [9] CHALMERS, Applied Mechanics, Master’s Thesis 2014:62 25 It is important to minimize the road space occupied by the long combination vehicles at intersections so that the safety risk is managed in such conditions. A high value implies that the vehicle needs more space that is available space. The vehicle is likely to collide with objects or other vehicles in the road or run off the road during turning maneuvers. 2.5.3 High Speed Transient Offtracking High speed Transient Offtracking is defined as an overshoot in the lateral distance between the paths of the center of the front axle and the center of the most severely offtracking axle of any unit in a specified maneuver at a certain friction level and a certain constant longitudinal speed [9]. When a long heavy vehicle is turning at a high speed, there is a tendency for the rear axles to sway outside the front axle’s path. This tendency is referred to as high speed transient offtracking. A high value of this might lead to collision with the road objects or other vehicles. Figure 26: Swept Path Width of a truck [9] Figure 27: High Speed Offtracking of a truck [9] CHALMERS, Applied Mechanics, Master’s Thesis 2014:62 26 2.5.4 High Speed Steady-state Offtracking High speed steady-state offtracking is define as the lateral offset between the paths of the center of the front axle and the center of the most several offtracking axle of an unit in a steady turn at a certain friction level and a certain constant longitudinal speed [9]. Just like high speed transient offtracking, high speed steady-state offtracking is the lateral displacement of the rear end of the last trailer of a long vehicle combination from the final path of the front axle of the hauling unit can lead to collision with the road objects or other vehicles especially when the road lane width is narrow and traffic flow is high on the road. Figure 28: High Speed Steady-state Offtracking of a truck [9] CHALMERS, Applied Mechanics, Master’s Thesis 2014:62 27 3 Methodology The first phase of the thesis work was marked by literature review of the engineering reports from previous projects involving long haul hybridization, and learning of the simulation tools used at Volvo Group Trucks Technology, namely Global Simulation Platform (GSP) and Volvo Transportation Models (VTM) (refer Chapter 5). The simulation model of a baseline truck was chosen and modifications were performed on the model to develop the combined model of the truck and the electrified dolly, in GSP. The next step from there was to integrate the VTM model of the A-Double configuration with the previous GSP model. After the previous step, Packaging study and Powertrain sizing was performed in parallel with the building of detailed Powertrain control strategies, and simulation of the new control strategies with the developed GSP model. The packaging study was performed by collecting geometric data of the components and a typical dolly, and construction of mock-ups using CAD software CATIA. The next phase of the thesis work was the evaluation of the simulation results with the combined GSP + VTM model platform, and performing simulations on it. The final phase was analysis and conclusions, which involved documenting results, analysis of those results and drawing conclusions from it. A flowchart illustrating the above mentioned methodology is as follows. Figure 29: Flowchart of the thesis project methodology CHALMERS, Applied Mechanics, Master’s Thesis 2014:62 28 4 Vehicle Specifications The chosen vehicle configuration for the simulations were a result of discussion with the Volvo personnel on what the typical use case scenario of such a long combination would need, based on previous projects and experiences. Also considered in this discussion were the typical consumers of this application and their demands on the capability and the performance on the product. Relevance to the current market scenario and the technologies, as well as a thought on financial relevance, was highlighted. The resulting vehicle component specifications for the hybrid and conventional reference vehicle have been tabulated as follows. Table 5: Specifications of the Hybrid and Conventional Vehicle Component Hybrid Vehicle Conventional Reference Vehicles Engine 13 litre 500hp Euro5 Volvo Diesel Engine 13 litre 500hp Euro5 Volvo Diesel Engine Transmission 12 speed Direct-drive Automated Manual Transmission 2 speed Gearbox 12 speed Direct-drive Automated Manual Transmission - ESS Battery A 19.2kWh - MDS Motor A 120kW 800Nm Motor B 179kW 430Nm Motor C 110kW 800Nm - - - Total Length of Combination 35.5m 35.5m 18m Total weight of the Combination 60,000kg 60,000kg 32,000kg CHALMERS, Applied Mechanics, Master’s Thesis 2014:62 29 5 Vehicle Simulation Models The simulation models were designed with the help of proprietary tools that Volvo has developed. They are namely the Global Simulation Platform (GSP) and Volvo Transportation Models (VTM), which are used by the Powertrain and Vehicle Engineering departments respectively. The tools are present as add-ons to the library of Matlab Simulink. There are two types of vehicle simulations used in industry depending on how the tractive power demand is determined: A. Driver throttle and brake pedal behavior (Forward simulation): The demanded traction power can be calculated with the accelerator pedal position level (0 to 1), and the maximum power available that the current vehicle speed. The negative demanded braking power can be calculated as a multiplication of maximum regeneration force, which is scaled by the brake pedal position (0 to 1) and the current vehicle speed. This type of forward simulation is similar to how a vehicle is operated in real life. B. Power Demand Equation (Backward simulation): A more accurate method would be to calculate it by the equation: Ptract = (m*a + m*g*sinΘ + m*g*Cr*cosΘ + 0.5*Cd*A*ρ*v2)*v; which accounts for the acceleration/retardation, climbing resistance, rolling resistance, and aerodynamic drag. Hence, no driver model is required in this type of simulation. One disadvantage of using this estimated value of power demand is that the pedal position no longer controls the power flow. Although the pedal position is sent to many subsystems as a control signal, the powertrain torque demand will be incorrect if there is no feedback possibility. (Optimizing energy management and component sizing of hybrid powertrain) Method A, i.e. forward simulation is used in both the GSP and VTM simulation tools as it has the same causality as the real world and hence it can be readily implemented in a prototype. These tools were developed based on the needs of the respective departments, and hence the level of detail used in modeling the different components vary based on the need. The GSP models contain greater level of detail towards the powertrain components as the need is to study the longitudinal dynamics characteristics of the vehicle such as acceleration, gradeability, fuel consumption and the suspension and tyres are not modeled in detail. Whereas, in the VTM models, the lateral and vertical dynamics characteristics are of more concern, and hence represented in detail. The reasons behind this prioritization could be mostly attributed to the availability of computing power, as it is the limiting factor to how detailed the models can be. The major part of the thesis is to analyze the fuel consumption and performance of the resulting hybrid powertrain. Additionally, this thesis also aims to study the lateral dynamics behaviour of the vehicle combination after the electrification of the dolly, and if the resulting control strategy has any questionable impact on it. Hence, it is important to prioritize and employ the right simulation model towards the right application due to lack of computational power. The complete powertrain model of the A-double combination is developed in GSP and it is used to compute the fuel consumption and the related longitudinal performance parameters. This model is then united with the VTM model of the A- double combination and the vehicle dynamics behaviour is analyzed. CHALMERS, Applied Mechanics, Master’s Thesis 2014:62 30 5.1 Powertrain Model (GSP) GSP is the Volvo Group’s common interface for evaluation of fuel consumption and vehicle performance. It is a database repository for vehicle and machine simulation models .This database is of high standard that guarantees traceability and quality assurance. It provides common guidelines and a unified model structure that facilitates sharing of data and reuse of model components. The open platform architecture of GSP ensures transparency and enables the system to integrate efficiently with other systems and processes [10]. The base GSP model was studied to understand, familiarize with and use as a starting point to develop the hybrid powertrain of the A-double long combination. It was a simulation model of a conventional truck with the same engine and transmission that is to be used in the target vehicle. . The longitudinal simulations have been done on GSP models with different control strategies and powertrain concepts. Energy-related parameters such as fuel consumption, energy throughput of the ESS, regenerated energy have been calculated for two different driving routes, which will be explained in detail in the results section of this report. 5.2 Vehicle Transportation Models (VTM) The lateral simulations have been carried out in order to study the effect of electrification of the dolly axle(s) on the stability and performance of the vehicle during certain common maneuvers, such as lane-changing, taking a U-turn, and driving around a roundabout. The tool used in order to simulate these phenomena is by the name Vehicle Transportation Models, which is used by various departments at the Chassis & Vehicle engineering department at Volvo GTT. It consists of an extensive Simulink library consisting of SimMechanics and other components used in order to create an accurate mathematical description of the physical components used. The VTM model for the A-Double long combination was combined with the GSP model of the powertrain and this unified platform was used to evaluate the maneuvers for the lateral simulations which were set up. CHALMERS, Applied Mechanics, Master’s Thesis 2014:62 31 6 Driving Cycles & Vehicle Maneuvers The thesis work deals with long haul truck application, and hence the driving cycles chosen towards the simulation were typical long haul driving cycles taken from the GSP repository. The driving cycles considered were Borås-Landvetter-Borås, and German Highway. Driving cycles are a collection of data that describe the speed of a vehicle versus time. The driving cycles are used as an input for models of vehicles to perform simulations to analyze the parameters of the model. They are aimed to be a representation of the road. There exist different types of driving cycles, designed specifically to test a certain aspect of the performance of the vehicle, such as emissions or fuel consumption. Driving cycles can be broadly defined based on the speed profile as city driving cycles and long distance cycles. They can also be classified based on the type of vehicle to be tested, as passenger car driving cycles and commercial vehicle driving cycles. It is to be mentioned that altitude is also included in the dataset of the driving cycles, as it is interesting to study the effect of altitude on the control strategies for hybrids. Some important parameters of the driving cycles chosen are tabulated below. The plots showing the vehicle speed and altitude versus time follow the table. Table 6: Data on the Borås-Landvetter-Borås and German Highway cycles Driving Cycle Name Total Distance (km) Average Speed (km/h) Number of stops Max gradient (%) Borås- Landvetter- Borås 87 84.6 0 Uphill 5.3 Downhill -5.3 German Highway 546 85.8 4 Uphill 6.5 Downhill -7.1 As is evident from the table above, the Borås-Landvetter-Borås cycle is a constant speed cycle with no stops during the route. The gradient is also considerably significant, with a maximum value of 5.3%. This cycle was selected because it is very well known within Volvo Powertrain and is used as an internal reference. It is also quite easy to practically test a vehicle in this driving cycle. The German Highway cycle is a long distance route in Germany, with a constant cruising speed of ~86km. The major difference between the nature of this cycle and the BLB one is that this route involves 4 complete stops to a standstill. Also, this route has more frequent gradient variations as well as a wider range in gradients (maximum 6.5% and minimum -7.1%). This route was selected as it represents a typical European long-haul application, and has both stops as well as considerable gradient variations. CHALMERS, Applied Mechanics, Master’s Thesis 2014:62 32 Figure 30: Speed and Altitude versus Time for the Borås-Landvetter-Borås cycle CHALMERS, Applied Mechanics, Master’s Thesis 2014:62 33 Figure 31: Speed versus Distance for the German Highway cycle Figure 32: Altitude versus Distance for the German Highway cycle CHALMERS, Applied Mechanics, Master’s Thesis 2014:62 34 7 Concept Study 7.1 Packaging Study The packaging study is aimed at studying the feasibility of packaging the different powertrain components in the dolly. The 3D mock-up of the dolly and the powertrain components were built in CAD software by name CATIA, with the help of 2D component drawings and field measurements. The dolly was built from the 2D drawing of a steerable dolly from a certain supplier with the suggestion from Vehicle Engineering department at Volvo GTT. The 2D drawings and the field measurements of the powertrain components were obtained from Drivelines & Hybrids department. The components were assembled in the dolly and the observations for different components are as follows. Table 7: Observations regarding various components during the packaging study Category Component Comment ESS Battery A Extension of Wheelbase from 1.3m to 1.7m required to house the ESS. Battery B Ground clearance less than 150mm. MDS Motor A Ground clearance less than 150mm Motor B Packaging is possible. Motor C Ground clearance less than 150mm. Gearbox 2 speed Gearbox Packaging is possible. It is important to mention that the dolly modeled towards the packaging study is a typical dolly used in an A-double configuration, and that the results of this study are more of a suggestive nature towards the motive of packaging the powertrain components in the dolly. The aim was towards solutions which involve the least modification to the established dolly designs which are prevailing in the market today. The main parameter used to compare whether a component is desirable to study or not, was the ground clearance i.e., the distance of the ground from the lowest part of the component. This is a very important parameter as it is not wise to have an expensive and a vital component lying low, which makes it susceptible towards damage. Hence, on this parameter, Battery B, Motor A and Motor C were considered undesirable. CHALMERS, Applied Mechanics, Master’s Thesis 2014:62 35 7.2 Powertrain Sizing This thesis work was bound by the limitations that the usages of components were limited to the inventory, which made the calculation of the optimal powertrain sizing superfluous. Hence Powertrain sizing as per the scope of the thesis work is more related to the choice and the reasoning behind the selection of the components which were included in the concept study for the propulsion components of the Electric dolly. With reference to the MDS components considered, motor B with a speed range of 0- 11000 RPM was considered the most desirable of the selection, due to packaging ease. Motor C with a speed sweep of 0-6000 RPM was considered more suitable to the application than Motor A which had a speed sweep of 0-3000 RPM. The main reason behind this is that the Motor A was designed to work with the diesel engine in the Parallel hybrid layout by name ISAM (as explained in the Literature Review section), hence it has its speed range comparable to that of the diesel engine. Since the electric dolly is completely independent of any physical linkage with the diesel engine, it is possible to have a different range of speed for the electric motor. An additional reason why an electric motor with a small speed range was undesirable is the potential over-speeding of the electric motor in the regenerative mode, where the road is driving the electric motor. Due to the fixed gear steps involved, when the power flows from the road to the electric motor, often high speeds are achieved. This is more unlikely in the case of the other motors considered, which have higher speed ranges. A two speed gearbox was included to analyze the effect of having two gear ratios to operate with on the fuel consumption and performance. The two speed gearbox provides higher torque at low speeds and prevents overspeeding of the EM at higher vehicle speeds by shifting to a lower gear ratio. With respect to the sizing of the final gear ratios, the inventory of GSP was used to list out all the available final gear ratio units. The values of the final gear ratios were again converged to 6 units, namely 0.89, 2.5, 5, 10, 15, 20 and then simulations were run with these units to further converge the number of units. CHALMERS, Applied Mechanics, Master’s Thesis 2014:62 36 7.3 Concept Definition Based on the inputs from the packaging feasibility study and the powertrain sizing, the combinations of the electric powertrain components for the dolly are arranged into different concepts for the final phase of the concept study. The concepts are defined as follows: Table 8: Definitions of the hybrid concepts Concept Number Definition Driven Axles Thumbnail Concept 1 • 1x Battery A • 1x Motor B • 1x Final gear unit 1 Concept 2 • 1x Battery A • 2x Motor B • 2x Final gear units 2 Concept 3 • 1x Battery A • 1x Motor B • 1x 2 speed gearbox • 1x Final gear unit 1 Concept 4 • 1x Battery A • 2x Motor B • 2x 2 speed gearbox • 2x Final gear units 2 The illustrations explaining the configuration of the different concepts are provided in the Appendix (Chapter 14.2). The thumbnails of the same are provided in the table above. The concepts were defined having in mind various factors such as simplicity, cost, and the characteristics of the components. Each concept is significantly different to each other, and they are defined so to understand the effects and study how the different concepts behave towards the control strategies. Emphasis was to avoid redundancy and on how effectively the components can be used. Concept 1 and 3 share the fact that only one of the two axles of the dolly is driven, whereas in Concept 2 and 4, both the axles of the dolly are driven. This was defined so as to make use of the increased grip and traction offered when more axles are driven. As mentioned in the Powertrain sizing section, the introduction of a 2 speed gearbox into the concepts was to study the effect of the choice of two gear steps in the fuel consumption and performance during the different operation modes of the electric motor. The only difference between Concept 1 and Concept 3 is the inclusion of the 2 speed gearbox between the electric motor and the final gear unit. It is the same case with Concept 2 and Concept 4 as well. CHALMERS, Applied Mechanics, Master’s Thesis 2014:62 37 8 Control Strategies The control strategy for the hybrid powertrain is formulated with emphasis on rule based control methodology. The base functionalities were adapted from an existing HCU from a Volvo project. The functionalities which were then developed for this application were added in increments and they are as follows: 8.1 Hybrid powertrain control functionalities There are three control blocks within the powertrain simulation model which have been dealt with in detail and extensively modified to implement the desired control strategies in the powertrain. The Charge Balance Control block takes in data about the vehicle state and battery state, and thereby gives limits on the maximum and minimum SoC-based battery power. The Torque Abilities EM block requires inputs on the EM speed, EM current gear ratio, EM Power and torque limits, ESS power limits and auxiliary power consumption in order to set limits on the maximum and minimum EM torque that can be requested. Finally, the Torque Distribution Control block decides on how to split the torque demand between the ICE and EM, based on inputs of total powertrain torque demand, braking torque demand, and the EM and ICE torque limits. All three control blocks have been adopted from the existing HCU. The Charge Balance Control block and Torque Distribution block have been extensively modified, whereas the Torque Abilities EM block has been left untouched. 8.1.1 Torque Distribution Control This block controls the amount of torque demand distributed to the EM and ICE. The total powertrain torque demand is checked with the EM torque limit signals coming from the ‘Torque Abilities EM’ block, after which it is checked with the EM torque limit set by the ‘SoC Control’ blockThis control logic in this block is entirely new and differs greatly from the previous logic. 8.1.2 Torque Abilities EM This control block outputs the final limits on maximum and minimum EM torque. It finds the power limit of the system by comparing those of the EM and ESS. The power limit is converted to a torque limit by dividing it with the real-time EM speed. This block has been adopted from the existing HCU and has been left unmodified. CHALMERS, Applied Mechanics, Master’s Thesis 2014:62 38 8.1.3 Charge Balance Control 8.1.3.1 SoC Control strategy The aim of the SoC control block is set the SoC window, as well as to govern the SoC level of the ESS by setting a SoC target and trying to match the current SoC with it. Since this hybrid powertrain doesn’t have a charge-through-engine mode, the only way of charging the ESS is through regeneration. Also, the EM cannot be made to supply more torque than required to reach the required SoC, as the EM is directly connected to the driveline. Hence only an upper limit on the EM torque can be set. The two ways in which SoC target estimation has been done in this thesis have been described in the next section. The blocks within the Charge Balance Control block were adopted from the existing model. However, the SoC Target block has been changed completely, due to application of a new method of calculating SoC target. Also, the SoC based Discharge/Charge limits block has been replaced with new control logic. The SoC Target is converted into a power request by calculating a parameter called the SoC Ratio. Then, a look-up table is then used to provide the power request for a certain SoC ratio. When SoC > SoCtarget : 𝑆𝑆𝑆𝑟𝑟𝑟𝑟𝑟 = 𝑆𝑆𝑆 − 𝑆𝑆𝑆𝑡𝑡𝑡𝑡𝑡𝑡 𝑆𝑆𝑆ℎ𝑖𝑖ℎ − 𝑆𝑆𝑆𝑡𝑡𝑡𝑡𝑡𝑡 When SoC < SoCtarget : 𝑆𝑆𝑆𝑟𝑟𝑟𝑟𝑟 = 𝑆𝑆𝑆 − 𝑆𝑆𝑆𝑡𝑡𝑡𝑡𝑡𝑡 𝑆𝑆𝑆𝑡𝑡𝑡𝑡𝑡𝑡 − 𝑆𝑆𝑆𝑙𝑙𝑙 The SoC ratio calculated is fed as an input to the following lookup table, and the output is a percentage of the battery’s maximum power capacity that is sent as a power request. CHALMERS, Applied Mechanics, Master’s Thesis 2014:62 39 8.1.3.2 Look-Ahead SoC Target Calculation This SoC control strategy aims to use knowledge of the future driving route in order to ensure that the vehicle fully utilizes opportunities to recover energy and store it in the battery without completely saturating it, such as when braking to a standstill or driving down a hill. It also aims at enabling the EM to assist the ICE with extra propulsion during demanding situations, such as driving up a hill. The look-ahead strategy of estimating the SoC target uses information of the forthcoming driving route in order to calculate the kinetic and potential energy of the vehicle at a certain ‘X’ meters ahead from its current position. This is done by looking up the altitude and velocity of the vehicle at ‘X’ meters ahead, and then estimating the SoC Target based on the amount of total energy that will be lost or gained after ‘X’ meters. The look-ahead parameter X can be iterated for each driving route in order to reach an optimal value. Figure 33: A Look-up Plot of % Max ESS Power versus SoC Ratio Figure 34: A flow chart depicting the Look-ahead SoC Target Calculation CHALMERS, Applied Mechanics, Master’s Thesis 2014:62 40 8.1.3.3 Predictive SoC Target Calculation This SoC control strategy is similar to the look-ahead type as it also aims at using knowledge of the future driving route to maximize energy recovery and assist the ICE whenever possible. However, instead of looking up the future vehicle speed and altitude at a certain distance X, it utilizes the energy profile of the entire driving route. Hence, complete information of the entire driving route must be known in advance at the start of the journey in order to implement this strategy. The energy profile is basically a plot of the kinetic and potential energy of the vehicle during the entire driving route. Then, the KE and PE of the vehicle at certain important points are identified on the energy profile, such as the start and end of hill, and the start and end of an acceleration/deceleration phase. In order to use this information, the energy of the vehicle at the next ‘important’ point is compared with the vehicle’s current energy. The energy difference is calculated as a percentage of the battery’s capacity, and thus converted into a SoC target using the two formulae below: Figure 35: A flow chart depicting the Predictive SoC Target Calculation Figure 36: Energy Profile for the Borås-Landvetter-Borås driving cycle CHALMERS, Applied Mechanics, Master’s Thesis 2014:62 41 𝐹𝐹𝐹𝐹𝐹𝐹 𝐸𝐸𝐸𝐸𝐸𝐸 𝐺𝐺𝐺𝐺 (𝑘𝑘ℎ) = 𝐹𝐹𝐹𝐹𝐹𝐹 𝐾𝐾 𝑔𝑔𝑔𝑔 + 𝐹𝐹𝐹𝐹𝐹𝐹 𝑃𝑃 𝑔𝑔𝑔𝑔 = 𝑚 ∗ �𝑣𝑓𝑓𝑓𝑓𝑓𝑓2 − 𝑣2� + 𝑚 ∗ 𝑔 ∗ �ℎ𝑓𝑓𝑓𝑓𝑓𝑓 2 − ℎ2� 𝑆𝑆𝐶 𝑇𝑇𝑇𝑇𝑇𝑇(%) = 𝑆𝑆𝑆𝑖𝑖𝑖𝑖𝑖𝑖𝑖(%) − � 𝐹𝐹𝐹𝐹𝐹𝐹 𝐸𝐸𝐸𝐸𝐸𝐸 𝐺𝐺𝐺𝐺 (𝑘𝑘ℎ) 𝑇𝑇𝑇𝑇𝑇 𝐵𝐵𝐵𝐵𝐵𝐵𝐵 𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 (𝑘𝑘ℎ) � 8.2 Sailing This control strategy is designed to make use of the potential energy gained by the vehicle at the top of a hill. It is similar to the sailing concept described in section 2.3.1.2 of this report as it also keeps the driveline disconnected during a downhill phase, in order to prevent loss in kinetic energy due to engine friction losses. However, this strategy is not necessarily aimed at extending the coasting phase by keeping the EM propelling the vehicle above the target speed. The EM operation is directed only by the torque distribution control and the SoC control. The controller requires input signals of Current Road Angle, ICE torque Demand, and Current ICE Torque, and outputs the Clutch Disengagement Position to the TECU. This control strategy is entirely new and is not adopted from the existing HCU. The control logic is explained in the flowchart below: As seen in the above flowchart, the controller checks for three conditions that must be met in order to allow disengagement of the clutch: 1) Current road angle must be negative: Since sailing must be activated only while driving down a slope. 2) ICE Torque Demand must be less than 100 Nm: This condition must be met in order to make sure that sailing is activated only when there is negligible torque demand. In order to avoid frequent clutch disengagements which can affect vehicle performance, a certain optimal value for this must be reached, which was found to be 100 Nm after several iterations. 3) Difference between ICE torque demand and actual ICE torque produced must be less than 10Nm: This condition is required in order to prevent engine overspeeding when the clutch is disengaged. Figure 37: A flow chart depicting the Sailing control strategy CHALMERS, Applied Mechanics, Master’s Thesis 2014:62 42 On implementing this function, it was observed that it resulted in fuel consumption reduction only for a short and transient driving route, i.e. the MHD cycle (-1.2% f.c.). However, it lead to a slight increase in fuel consumption for the longer heavy duty driving routes such as the German Highway and BLB cycle (+1.2% f.c.). It was concluded that this was because of the frequent engaging and disengaging of the ICE at high output shaft speeds during cruising. Also, this function only reduces the engine inertia, and not the inertia of the transmission input shaft. 8.3 Ecoroll The Ecoroll function was then considered as a replacement because of the problems caused by the sailing feature mentioned in the previous section. This function aims to reduce the inertia of the driveline by allowing the transmission input shaft to rotate freely (i.e. engaging the neutral gear). This is done while going down a hill with negligible torque demand from the ICE. This function was present in the existing HCU, but the conditions for activation have been changed. This function is activated when the following conditions are met: 1. ICE Torque Demand < 100 Nm. 2. Vehicle Speed >= 60 km/h. 3. Gradient < 0. Figure 38: Altitude versus Time and Current Gear versus Time plots CHALMERS, Applied Mechanics, Master’s Thesis 2014:62 43 39 40 41 42 43 44 45 46 47 48 49 50 Base HCU Lookahead+Sailing Predictive HCU with Ecoroll Fu el C on su m pt io n (l/ 10 0k m ) Control Unit Type BLB Cycle Conventional Concept 1 (FGR=5) Concept 2 (FGR=5) 9% 10% 8% 7% 11.8% 12.2% 9 Results This chapter will present results from both longitudinal and lateral simulations that have been performed on the GSP and GSP+VTM models respectively. The results of the hybrid concepts have been compared with a conventional truck of 60tons and 32tons total weight. The longitudinal simulations have been done on GSP models with different control strategies and powertrain concepts. Energy-related parameters such as fuel consumption, energy throughput of the ESS, regenerated energy have been calculated for two different driving routes. Also, performance-related parameters have been calculated, such as average speed, total trip time, vehicle speed error, and the number of gearshifts. The fuel consumption and performance results of the hybrid models have been compared with both a conventional model for the same vehicle configuration, as well as with results for existing hybrid models built at Volvo earlier. The lateral simulations have been carried out in order to study the effect of electrification of the dolly axle(s) on the stability and performance of the vehicle during certain common maneuvers, such as lane-changing, taking a U-turn, and driving around a roundabout. Several lateral performance-based characteristics have been calculated for these maneuvers, and the results have been compared with the acceptable limits and with the results for the conventional model. 9.1 Longitudinal Simulations In this chapter, various results related to fuel consumption, vehicle performance and drivability have been presented and compared for the Conventional vehicle and the Hybrid concepts. Please refer to Chapter 7 for more information on the hybrid concepts. It must be clarified that the Conventional vehicle in this case is a 60ton truck with the same ICE, length, and cargo capacity, unless stated otherwise. Please refer to Chapter 4 for more information on the conventional reference vehicle. 9.1.1 Fuel Consumption 9.1.1.1 Control Strategies The HCU’s that have been compared in this section are the Base HCU, the HCU with Lookahead + sailing, and predictive HCU with Ecoroll. Figure 39: Fuel Consumption plots for various Control Strategies for the BLB cycle. The percentages represent fuel saving relative to conventional vehicle. CHALMERS, Applied Mechanics, Master’s Thesis 2014:62 44 It is observed in the figure 39 above for the BLB cycle that highest savings in fuel consumption were achieved for the predictive HCU with Ecoroll functionality, i.e. 11.8% and 12.2% for concepts 1 & 2 respectively. The base HCU taken from the Volvo project achieved a fuel saving of 9.1% & 10.1%. However, the fuel savings decreased to 8.15% & 6.7% on implementing the look-ahead & sailing function onto the base HCU. A possible reason for this is that the look-ahead distance was fixed as 1000m, and this parameter needs to be optimized for each driving cycle. As seen above, a similar result was achieved for the German Highway cycle as well, where the predictive HCU provided the highest fuel savings of 12.6% % 13.9%, and the Lookahead+Sailing HCU provided the least fuel savings of 7.6% & 6.6% for concepts 1 & 2 respectively. 9.1.1.2 Concepts Concept 4 has provided the least fuel consumption, which is 14.2% less than the conventional model. Concepts 1 & 2, which have a direct final gear ratio from the EM 40 42 44 46 48 50 52 54 Base HCU Lookahead+Sailing Predictive HCU with Ecoroll Fu el C on su m pt io n (l/ 10 0k m ) Control Unit Type German highway Cycle Conventional Concept 1 (FGR=5) Concept 2 (FGR=5) 8.8% 10.7% 7.6% 6.6% 12.6% 13.9% 38 40 42 44 46 48 50 Conventional Concept 1 Concept 2 Concept 3 Concept 4 Fu el C on su m pt io n (l/ 10 0k m ) Concepts BLB Cycle Ecoroll Disabled Ecoroll Enabled 11.75% 12.15% 14% 14.2% Figure 40: Fuel Consumption plots for various Control Strategies for the German highway cycle Figure 41: Fuel Consumption plots for various concepts for the BLB cycle CHALMERS, Applied Mechanics, Master’s Thesis 2014:62 45 38 40 42 44 46 48 50 52 54 Conventional Concept 1 Concept 2 Concept 3 Concept 4 Fu el C on su m pt io n (l/ 10 0k m ) Concepts German highway Cycle Ecoroll Disabled Ecoroll Enabled 12.6% 13.9% 14.3% 16.4% to the wheels, provide about 2% less fuel savings as compared to concepts 3 & 4 which have a 2 speed gearbox. The implementation of Ecoroll has resulted in an average increase in fuel savings of 0.44%, with the maximum effect seen in in the conventional model (0.69% more savings) and the least effect seen in concept 4 (0.29%). Also, it is interesting to note that there is no significant difference in fuel consumption values between concept 1 and 2, and between concept 3 and 4. Hence, the addition of an additional EM has not contributed towards significant fuel savings (only 0.3% and 0.2% increase in fuel savings). The reason for this was found to be the power limit of the ESS. It was discovered that although the two EM’s have a total capacity of 358kW, the ESS has a discharge limit of 170kW. This power limit translates to a torque limit on the EM’s, as depicted in the figure below. The blue line represents the EM torque limit, and the green line is the torque limit due to the battery’s maximum discharge power limit. A similar trend is observed in the results for the German Highway cycle. However, the fuel savings values are greater for all concepts in this application. The effect of Ecoroll has resulted in an average increase in fuel savings of 0.41%, with the Figure 42: Plot of the Battery and EM Torque Limits versus Time Figure 43: Fuel Consumption plots for various concepts for the German highway cycle CHALMERS, Applied Mechanics, Master’s Thesis 2014:62 46 43.4 43.6 43.8 44 44.2 44.4 44.6 44.8 45 FGR= 0.89 FGR=5 FGR=25 Fu el C on su m pt io n (l/ 10 0k m ) Final Gear Ratio of EM BLB cycle Concept 1 Concept 2 -0.98% + 0.045% -1.83% -0.93% maximum effect seen on the conventional model (0.64%), and the least effect seen on concept 1 (0.29%). The difference in fuel savings between the two driving routes is especially more for concepts 2 and 4 which have 2 EM’s (1.4% & 2.2% higher fuel savings as compared to the BLB cycle). 9.1.1.3 EM Final Gear Ratios Several iterations were done in order to arrive at fuel consumption-optimized final gear ratio for the EM. It was found that in the driving cycles with more speed variations, the higher final gear ratios such as 10, 15 and 20 were found to give better fuel consumption, whereas in driving cycles with constant cruising speeds over larger distances, the lower final gear ratios such as 0.89, 2.5 and 5 gave better fuel con