Design of Electric LIght Vans for Environment-impact Reduction

Project Information

DELIVER

Grant agreement ID: 285296

Closed project

Start date 1 November 2011

End date 31 January 2015

Funded under

• Specific Programme "Cooperation": Transport (including Aeronautics)

Total cost € 4 234 444,61

EU contribution € 2 799 594,00

Coordinated by

RHEINISCH-WESTFAELISCHE TECHNISCHE HOCHSCHULE AACHEN
Germany

Final Report Summary - DELIVER (Design of Electric LIght Vans for Environment-impact Reduction)

Executive Summary:
The DELIVER project pursed two main objectives: 1) raising the efficiency of a light commercial vehicle by 40 % compared to conventional vehicles in this category and 2) considerably enhancing the transport performance. Both objectives were to be based on a modern and exceptional design. The efficiency improvement of 40 % necessarily requires an electric drivetrain, especially to reduce the vehicle’s losses of energy conversion. Further, the very quiet engine and the freedoms in packaging due to in-wheel motors add to the advantages of this aim. Nowadays typical cargo vans are deducted from passenger cars or pick-up trucks. Vehicles with a dedicated design are, at least in Europe, hardly ever used, which leads to ergonomic strains for the drivers. Thus, the DELIVER project aimed to increase the ergonomics for the drivers, the design of the payload area as well as the means of transportation, in short to enhance the efficiency of transport. The interior and exterior design of the DELIVER vehicle was an important part of the project, as a vehicle’s character is mainly shaped through its exterior design. Many companies use their fleets to present themselves, especially in urban traffic. Therefore, the project aimed to create a modern and outstanding vehicle. The described use case of urban delivery is one relevant requirement for this vehicle category. There are many other cases, like the usage as a minibus or by craftsmen. They were thus also examined in a derivate analysis. The project was completed with economic analyses regarding the cost-effective production in a small volume. Based on the requirements collected in the early phase of the project a vehicle concept with a completely new design utilising a fully electric drivetrain was developed, with the driver’s requirements always in mind. The innovative drivetrain features two in-wheel motors with a two-speed transmission and locates the high voltage battery pack securely in the middle of the floor. The resulting freedom of design for the front vehicle enables optimisation of the passive safety systems. Thanks to large scale energy absorbent elements, acting in case of a frontal crash, the load path leading across the front wheels and the side sill within the door area is no longer necessary. The “walk-in door” concept reduces potential obstacles caused by the door sills. Furthermore the construction of the rolling chassis with a steel spaceframe and the integration of the drivetrain made a transmission tunnel redundant. The additionally available space was used to construct a swivelling driver seat that enables the driver entering and exiting the vehicle also on the kerb side. The drivetrain’s two in-wheel motors are synchronous machines on the rear axle. Each engine provides a maximum power of 57 kW with a maximum torque of 42 Nm. These so-called Motorized Wheels include a two-speed transmission ensuring an adequate driving torque and maximum speed of 100 km/h at all times. The high voltage battery pack consists of 80 prismatic Li-NMC cells with an overall energy content of 21.6 kWh to ensure the minimum range of 120 km (half ladden). The vehicle’s 12 V net is supplied by the high-voltage battery via a 3.5 kW DC/DC converter. The successful completion of the prototype vehicle in April 2014 was followed by a testing phase, which was conducted by various project partners on different proving grounds. The performed tests focussed on driving dynamics, efficiency, acoustics/NVH, electromagnetic compatibility and ergonomics, and widely approved the results from the virtual development process.
Project Context and Objectives:
DELIVER aims to develop an optimised electric LCV architecture, integrating and optimising all subsystems relevant to energy efficiency and to other key end-user requirements including affordability. The project aims to do this using a highly realistic method of design research, i.e. exploring the best design approaches and conceptualising a highly optimised electric LCV that, at the end of the project, is prototyped into a driving vehicle. The project furthermore aims to connect closely with the final users of urban LCVs in order to ensure that the resulting optimised concept is not just the best compromise regarding the energy efficiency perspective, but also from the perspective of major stakeholders, i.e. owners and users that need to work with these vehicles on average at least 8 hours a day.
The specific objectives of the DELIVER project are:
- Explore and identify a broad range of conceptual design options in a structured and well-documented development of electric light commercial vehicle concepts/ architectures and their modular structure using a systematic and holistic approach.
- One detailed vehicle concept built into a driving vehicle, which will be enabled by and specifically designed to utilise the optimised composition and integration of subsystems. This will be quantitatively assessed on all key performance parameters using physical real life testing in combination with highly accurate simulation tools. The prototyped vehicle will also be used to stimulate the dialogue between relevant stakeholders of operator (cities, postal services companies etc.) and vehicle as well as component manufacturers about the needs and possibilities of a more sustainable future.
- Limited exploration and real-life experimental validation of usability, performance and environmental, health and safety aspects. The results will show that the concept can realistically compete in tomorrow’s markets, with regards primarily to usability, driveability, ergonomics and sustainability (including 40 % increased energy efficiency) and secondarily in terms of production costs and modularisation, low maintenance and aesthetics without compromising safety and reliability.
The DELIVER project specifically focuses on innovative electric LCV concepts for mass application by 2020, based on the enabling technologies and market demands that can be expected by then.
The DELIVER project covers the entire design cycle, starting with the establishment of sets of requirements and the likely trade-offs that need to be made, taking into account also detailed knowledge of customer preferences sourced from other projects. It further covers initial, rough design phases (translating requirements into sketches, layouts) and their translation into vehicle concepts. It proceeds with a best selected concept and develops this further into a physical prototype vehicle. The virtual and physical model will allow quantitative performance assessment on a range of performance indicators such as energy efficiency, ergonomics, structural performance, EMC, economics and sustainability both in production and in use.
Project Results:
Architecture
In course of the DELIVER project possible sketches of solutions for the driver seating position, the battery position, the side door concept and different axle and motor types are prepared in order to evaluate as many ideas as. The cab-over concept is initially favoured due to a better room efficiency (load length vs. overall length) as well as the command driving position and the novelty factor. Due to the above disadvantages, the central seating position is not pursued. The cab-over concept is examined and compared to the cab-rearward concept in order to evaluate expected advantages and disadvantages. With the help of concept CAD models and accompanying studies it is decided to realise the demonstrator vehicle as a cab-rearward concept.
After having defined the vehicle architecture, the seating position is evaluated. The goal is to reach a good load length while maintaining good ergonomics and enough space for the driver. As a result, the seating position chosen is more forward and inward than in today’s production vehicles.
Ergonomics
The ergonomic analysis of virtual DELIVER concept verifies the compliance with the set specifications. The following aspects are there for analysed:
- Roominess
- External visibility
- Cluster’s visibility
- Accessibility
- Ergonomic advantages
The postural assessments are carried out using Abita2000 software, based on ergonomic criteria for the driving posture in vehicles. The code uses a population of mathematical segmental dummies morphologically constructed according to the Human Scale scheme, which provides a normotype distribution of the articular segments by placing the same percentile value to each segment; i.e. the 50th percentile in height will be the 50th percentile also for the other segments. The statures of these dummies are updated in latest years according to the growth trend of the reference population (secular trend). The population is considered referencing to the years of the decade 2010-2020.
These dummies are able to interact with the elements that characterise the interior, using its control parameters (seat and steering wheel) and to take the proper body posture according to the defined joints ranges that constitute the implemented criteria of comfort.
Further analysis tool is the Ramsis software that implements a three-dimensional manikin, biomechanically complex, for the evaluation of the visibility of the dashboard and for the generation of the reachability areas of the controls
The assessments made can be summarised as follows, with recommendations taken into account in the final concept as much as possible:
- Roominess evaluation: The roominess parameters are set correctly for allowing accommodation in comfort for the whole scale of percentiles. However, it is recommended to reset the accelerator plate angle to 46° in order to improve the support of the shoe.
- Front visibility evaluation: The angle of lateralisation of the A-pillar is within acceptable values, while the visual obstruction determined by the pillar itself in addition to the structure of the door is excessive (15.6°). Therefore, it is suggested to reduce the dead angle in the range from 9° to 11° max. The viewing angles upward and downward are set correctly.
- Lateral visibility evaluation: The viewing angle through the side window is slightly affected by the presence of the B-pillar. It is suggested to move back the pillar of measure between 7 and 37 mm to achieve the minimum requirements.
- Cluster visibility evaluation: The position of the instrument panel provides visibility through the steering wheel rim by the whole scale of percentiles.
- Accessibility evaluation: For the size of the door aperture, it is suggested to increase the space for the passage of the feet. The same criticality occurs for the encumbrance of the steering column, which reduces too much the clearance for the passage of the knee. The longitudinal dimension of the seat cushion is correctly sized.

Chassis and Suspension
Essential for the economic production of a commercial vehicle is the use of competitive components. The team thus focuses on using production suspension parts. Modified components of the Volkswagen Caddy are applied in the concept.
The axle tube of the rear suspension has to be prepared for the connection of the Michelin Motorized Wheels, but apart from that remains unchanged. The steering column is adapted to the new seating position. Those modifications are displayed in the demonstrator.
The vehicle dynamics are evaluated in theory, allowing implementing concept changes if necessary. After completion, the demonstrator undergoes dynamic tests to prove the safe driving behaviour of the concept.

Drivetrain
The drivetrain comprises the Michelin Motorized Wheels and the high voltage battery system provided by CADEM and Mobit, while Volkswagen provides the 12 V board net.
Michelin develops the main electrical architecture.
The architecture is divided according to following repartition:
- Vehicle Supervisor: Michelin
- Battery and BMS: CADEM & Mobit
- High Voltage Distribution: Michelin
- Low Voltage Distribution: Volkswagen/Michelin
- Electrical harness: Volkswagen
- Ignition and vehicle high voltage: Volkswagen
- Heater: LEC2
- Liquid cooling system: Michelin
In a dedicated inverter box, Michelin integrated:
- Traction inverter
- Vehicle supervisor
- HV connections and fuses
- LV fuses and breakers
The dedicated Motorized Wheels have the following features (per wheel):
- Michelin M3 Motor (max. power: 57 kW, max. torque: 42 Nm, max. speed 13,600 rpm)
- Mechanical transmission ensured by a two-gear transmission
- Brake functionality, including parking brake
The special feature of this wheel is the possibility of shifting two gears. A servomotor realises the choice of first or second gear positions using a fork. The position of the fork is known by an absolute position sensor. Before engaging the gear, the speeds of both parts are synchronised by the traction motor.
The chosen gear ratio enables a maximal torque of 1,030 Nm per wheel in the first gear and 575 Nm in the second gear. A special attention is paid on the gears design and the teeth geometry to reduce whining noise. The total weight of the wheel is 63.6 kg including a 195/65 R16 steel rim of 9.1 kg.
The information provided speed sensor is used by the ABS and for the synchronisation of the gears. The speed of the motor is provided by the resolver inside of the M3 motor. The electric motor chosen is the Michelin M3-002. The main characteristics of this dedicated machine are:
Three-phase permanent magnet synchronous motor
Mechanical data:
Maximum speed: 13,600 rpm
Motor weight (filled with coolant): 6.9 - (7.05) kg
Rotor inertia: 9.7 x 10-4 kg m2
Standard IP rating: IP68
Electric data:
Back EMF at 1,000 rpm (phase to phase): 15.6 Vrms
Phase resistance at 22 °C: 0.032
Mean phase inductance: 0.24 mH
Number of poles: 6
Windings coupling: star connection
Cooling conditions: circulating liquid (water / glycol-30%)
Cooling liquid flow rate: 5 l/min
Pressure drop (hydraulic connectors included): 0.14 bar
Cooling liquid temperature: 65°C
Maximum windings operating temperature: 200°C
Coil temperature sensor for thermal protection: integrated
The traction inverter used for this application is a specific Michelin double inverter (MD300I450U).
During the first test of the Motorized Wheel (August 2013), before the installation on the DELIVER rolling chassis, it is tested on a roller type test bench. On this bench however, it is not possible to apply the maximum torque at the maximal speed of the Motorized Wheel (29.9 kW) since the maximal power of the bench is 18 kW. It is tested instead at maximal power only until 4,000 rpm of the motor (indeed 13,600 rpm) or only at 13 Nm at maximal speed (indeed 42 Nm).
In order to verify and optimise the speed change and to validate the reliability and performances of components, two wheels are installed on another test bench. These two wheels are mounted one against the other. The first wheel is speed driven (simulation of the road) and the second is torque driven (simulation of the assembly to test). This bench allows testing the wheel at all ranges of torque and speed.
In April 2014, several tests are executed:
- 50 times acceleration until max. speed & deceleration under a torque of 20 Nm with gear change
- 50 times acceleration until max. speed at first gear (62 km/h) & deceleration under the maximal torque (1,133 Nm at the wheel)
- 24 hours non-stop testing of cycle "FTP 75 city driving" with four gear changes during each cycle (one cycle lasts 31 minutes).
- 120 hours of non-stop testing of cycle "FTP 75 city driving" with four gear changes during each cycle (one cycle lasts 31 minutes).
After all these tests, a complete verification of each element is completed, allowing equipping the vehicle with the Motorized Wheels.
Based on previous experiences around the automotive Motorized Wheel, Michelin develops a dedicated wheel for the DELIVER project. This 16 inches wheel is designed to comply with the leaf suspension already existing.
The HV battery is one of the most important components in an electric vehicle regarding both safety and performance characteristics. The development of the battery package is based on these criteria mainly.
In order to provide high-level specifications with the battery pack all partners put in their experience regarding the requirements of the battery pack and the battery management system (BMW) during the development of the concept. Safety line with emergency shut off function, crash sense, HV plugs connection sensing, isolation measurement and related safety functions are all included in the concept.
The studies on package space for the battery pack focus on placing the pack between the front and the rear axle to provide maximum crash protection. Available and suitable cell types are investigated in order to fit in this volume while providing a good range for a delivery vehicle.
Lithium polymer cells are selected due to their energy capacity advantage against Li-FePh cells. The selected cells are prismatic shape and packaged into steel sheet. The battery pack consists of eight battery modules where each module has ten serially connected cells, which sum up to 80 serially connected cells in the complete battery pack. Each module is monitored by a separate module unit and all module units are connected to the main board via CAN bus. The main board receives voltage and heat values from each module board and runs various evaluations before informing the supervisor of the vehicle and inverter of the traction motors. The battery package is covered by a steel box with air-cooling function.
The HV battery package provides the energy for the traction motors and the HV heat pump for cooling system. It is also connected to the DC/DC converter, which charges the 12 V auxiliary battery and the on-board charger, which charges the HV battery and 12 V auxiliary battery during charging.
The BMS, which is positioned in the HV battery box, has three main responsibilities:
1. Monitoring the condition of the cells in the battery pack regarding heat and voltage of cells
2. Communicating with the vehicle and other components through CAN bus for various conditions like starting, running, charging phases of the vehicle to provide necessary usage limits, warnings and information to the supervisor of the vehicle
3. Taking necessary precautions like shutting off the HV for emergency or limited usage cases, managing cooling functions etc.

Body Structure
The body structure of the DELIVER vehicle provides an extra load paths between the upper and the lower one. In case of a side impact the driver’s door and the B-pillar route the crash loads into the floor and roof structure.
After having chosen the cab-rearward concept from the prior examinations, the detailing of the CAD model is continued. The first step is a wire type concept of the frame considering the battery space.
Specific to the DELIVER concept is a sandwich type chassis with two levels of longitudinal and transversal beams. Therefore, the torsional and bending stiffness of the underbody is already very high. Due to the inward seating position, it is necessary to move the doorsills inward. The structure for offset front impact and side impact is executed by reinforcements in the door and a honeycomb structure behind the side panels.
With the detailed frame model and a substitution design of the upper body, it is possible to calculate the stability and rigidity of the frame so critical areas can be identified and reinforced.
The upper body is designed as a profile intensive steel construction. Such constructions can be economically manufactured even with smaller production numbers and they provide advantages with regard to derivatisation. Deep drawn sheets are only applied in areas that would be common for all derivates such as the rear wheelhouses and the front end of the vehicle.
The global stiffness behaviour of the vehicle body structure is tested under two load conditions. For global torsional stiffness measurements the rear strut mount is fixed on both the driver and passenger side while a torsional moment of 2,000 Nm in X direction is applied to the front strut mounts. For global bending stiffness measurements, all strut mounts are fixed while load forces are applied in negative Z direction in the wheelbase middle (3,000 N in total).
As the results show, the bending stiffness is on target while the torsional is not; so additional structural optimisation is necessary to meet the target for torsional stiffness. The eigenfrequency analysis is executed for the same body assembly as used for the static stiffness determination. The most important eigenfrequency is the 1st torsional mode, which is critical to avoid resonance frequencies for the vehicle body.
In order to increase the structure torsional stiffness and reach the required target, a thickness optimisation of some vehicle components is performed, in particular the upper body frame and the chassis. A variation of body panel thickness is not considered in this study.
This analysis assigned a new thickness distribution to the concerned structural elements in order to obtain the required torsional stiffness (> 700,000 Nm/rad) and minimise the eventual/expected added mass of the vehicle at the same time. An appropriate number of simulation loops in which the structural analyst managed them manually, i.e. without using an optimisation solver conducts this optimisation process. With this new configuration, the structure is 1.6 kg heavier, but the torsional stiffness is 735,000 Nm/rad. With a mass increase of 0.3 % the requested target is achieved: an increase of 22.5 % of torsional stiffness and an increase of 13.3 % of bending stiffness.
The dynamic stiffness of the trimmed body represents an important value of the structural performance as well. The model developed for the analysis of the bending and torsional stiffness is reduced to the parts that can have an impact on the performances of the panels and, in particular, of the attachment points. The reduced model suitable for static global performances needs an update for a correct dynamic evaluation. In particular, all significant masses of the model must be kept into account for this type of evaluation. Since it is not meaningful to have a detailed modeling of all masses attached to the vehicle, the following components are added with a reduced modeling strategy, keeping into account their dynamic contribution by including their position of the center of gravity, their mass and their inertial properties.
It must be noted that not all the suspended parts (the ones with elastic bushing between the body and the part) are considered in the so-called trimmed body. On the contrary, the windshield has been added due to its contribution to the stiffness of the front shock tower attachment points.
No engine attachments points are present in this vehicle since its traction is based on electric motors directly placed on rear axle. Inertance values of the front suspension system attachment points are in target. In almost all points, inertance values are assessed around 8,000 N/mm between 80 and 200 Hz with the exception of the vertical inertance of the front lower arm attachment point, in which the value is slightly lower. The overall stiffness is quite good with the exception of some peaks due to global eigenmodes (the lower, the better).
Regarding the rear suspension attachments points, two critical areas are identified immediately:
- Lateral dynamic stiffness of the front spring leaf attachment point
- Lateral dynamic stiffness of the shock absorber attachment point
Extending the welding lines not only at front part but also laterally towards the lateral rail, the performances increases from 3,000 N/mm to an interesting value of around 24,000 N/mm.
For the front leaf spring attachment point, a similar procedure is applied and, once more, extending the welding lines between the solid cast node and the closed sections of the chassis in the near body elements allows curing the problem and increasing the stiffness to an interesting value of 12,000 N/mm.
The analysis of the point inertances shows a quite good level of dynamic stiffness, in fact, most of the attachment points are between 8,000 and 16,000 N/mm, therefore ensuring a potentially acceptable filtering level of the vibrations coming from the road.
However, it must be advised adding some further welding lines to the attachment points of the rear suspension (front point of the leaf spring and rear damper fixing point) in order to avoid potential problems of rattle and noise/vibration. This should be considered for a production design.
To assess the passive safety properties of the DELIVER vehicle loadcases are chosen that are considered to be critical. The front crash scenario according to Euro NCAP is used to evaluate the structural behaviour in case of a frontal impact. The pole impact setup also according to Euro NCAP serves for the assessment of lateral collisions.
The behaviour of the full vehicle during a frontal impact according to the Euro NCAP test protocol is shown as an important fact that the cabin remains intact and provides sufficient survival space for the occupants. The largest intrusion that takes place in the upper left side of the firewall is 61 mm. The maximum value of acceleration is 28.1 g.
The behaviour of the full vehicle during a pole impact according to Euro NCAP is shows a maximum intrusion of 271 mm. The maximum value is 24.5 g. The results represent acceptable levels considering the prototype character of the vehicle concept.

Aerodynamics
The scope of the analysis of the aerodynamics of the defined concept is to quantify the impact on the energy efficiency of the vehicle. All evaluations are executed by means of CFD simulations. The following chapters describe the applied simulation tool and methodology.
The starting point for the build-up of the simulation model is the CAD model. In this early stage of the product development process only the external flow around the vehicle is considered since this is sufficient to determine the drag value cx with appropriate accuracy. Hence, a watertight model of the exterior surfaces is build up in CAD. Special effort is spend in this phase as possible errors can be fixed easily in this stage while they would cause large efforts at a later stage.
Subsequently, the completed CAD model is imported into the CFD software STAR CCM+. It is then used to generate the three-dimensional simulation mesh and to define the boundary conditions of the simulation as well as the physic models that should be applied. The result of this procedure is the CFD model with about 61.1 million 3D elements. A moving road surfaces as well as turning wheels are considered. The inflow velocity is 90 km/h.
Three different yaw angels (β = 0°, β = 5° and β = 10°) are considered in the investigation. This is done for the representation of the real world effect of crosswinds. Especially in those cases where the vehicle is not hit by straight flow, large areas of flow separations can occur, which again leads to energy dissipation and higher energy consumptions in the end.
All simulations are executed as steady state simulations using the Reynolds-averaged Navier-Stokes equations (or RANS equations). bThe computation of one CFD model takes about 8 hours on IKA’s high performance cluster (3,000 iteration steps).
The representation of the ISO surfaces for the total pressure coefficient cp, tot reveals that the unsteady exterior geometry in the front of the vehicle, which is required to implement the ergonomic concept, causes large areas of flow separation and hence a negative impact on the overall aerodynamics. In addition, the exposed front wheelhouses and headlights are disadvantageous for the drag value of the vehicle. However, the most significant area for the overall aerodynamics is the rear of the vehicle. The side view shows that there is only a relatively small ware in this region, which is caused by the tapered roofline and the distinct tearing edges at the rear doors.
In summary, the drag value cx, 0° at 0° yaw angle is at 0.35 and thus in the same range as values of comparable LCVs. This shows that the disadvantages created by the unsteady front geometry needed for the ergonomic concept are compensated by the design measure in the rear of the vehicle.
The results for yaw angles of 5° and 10° are documented in deliverable D4.2. In principle, the same observations as in the 0° yaw angle case can be made here. As expected, the wake region behind the vehicle and the flow separation area on the leeward side are larger than in the 0° case. This leads to higher drag values. The drag value for 5° yaw angle is 0.375. For 10°, the value is 0.412.

EMC and EMF
Electromagnetic Compatibility (EMC) and health effects from electromagnetic fields (EMF) are different aspects of electromagnetic fields. EMC has for many years been on the agenda in the vehicle development process. EMF has not been so much in focus since the electromagnetic fields in an ICE vehicle are generally low compared to recommended limits for avoiding harmful health effects. In a vehicle with an electric powertrain, this is not necessarily true, and EMF was therefore specifically included in this project.
Low emission levels of electromagnetic fields for avoiding harmful health effects (EMF) is not a legal requirement for E-marking of vehicles and the situation has been for a long time un-clear with a combination of international recommendations and national requirements. During the time of the project, the situation has become clearer with a European Directive that has been decided by the European Commission and shall be incorporated in national laws by 2016. Fulfilment of this directive is therefore declared to be the goal of the measurements in the project. A vital part of the effort concerning EMF is also to see if the field levels by computer simulations can be predicted before the vehicle is built.
Recommended techniques for designing and building the vehicle to achieve low levels of electromagnetic fields are presented in D2.8. Some of these rules are difficult to follow in a prototype vehicle were e.g. access to all systems must be prioritised over good shielding. The rules apply in principle for reducing both EMC and EMF effects. There are, however, some differences in how effective the methods are due to differences in frequency content etc. for EMC and EMF effects.
The main method to reduce EMF effects is to remove the potential sources as far as possible from the driver and passenger. In the case of the DELIVER vehicle, this results in positioning the inverter in the far back, and the battery output close to the inverter. Since the vehicle is back-wheel driven, this means that no high voltage lines carrying high currents during driving will be routed close to the driver or the passenger positions.
It would have been a further advantage for low EMF levels if the cargo bay floor had been a single metal sheet without any service opening. This would, however, be difficult on a prototype vehicle were access to all systems is vital.
The highest field levels in an electric vehicle will emanate from the drive train where the highest electric currents can be found. The frequency spectra of this current will show that most of the power is associated with low frequencies (<100 kHz).
CAD models of the complete vehicle are used as input to 3D field simulations of low frequency electromagnetic fields emanating from the electric powertrain. The models are imported to CST, a commercially available field simulation software package. Low frequency simulations are generally more reliable than high frequency simulations since the wavelength is long compared to most structures in the vehicle. Many small structures will therefore have a negligible influence on the results and can even be omitted from the simulation model. Minor differences between the real vehicle and the CAD model will only have a minor effect.
EMF levels are measured in a semi-anechoic shielded chamber with the vehicle on a dynamometer to simulate driving load conditions. The measurements show that the field levels in the driver position are well below the limits for health effects. The highest magnetic field measured is 0.74 µT at 20 kHz where the action level according to the EU directive is 100 µT.
The measurements also show good matching to the simulated values in field strength (0.18 and 0.16 µT/A respectively), and a reasonable match to the simulations in field pattern in the vehicle. This shows that simulating low frequency magnetic fields is a useful tool to predict EMF performance of vehicles.
The EMF work in DELIVER focuses on the actual vehicle that is built. Other FP7 projects have recently addressed EMF issues in electric vehicles with a more general approach. One example is EM Safety that finished in 2014. That project investigated techniques for measurements of magnetic fields in vehicles, design guidelines for reducing the magnetic field and EMF effects on mammalian cells. The last subject has not been addressed in DELIVER, but the first two issues have been included to various extent. The design guidelines developed in EM Safety are consistent with the guidelines developed in DELIVER. The measurements in DELIVER have been performed using standard set-ups and instrumentations. We have been comparing our data with the new EU directive that will be mandatory in 2016. The present text in the directive does not give any details on e.g. how to interpret measurements with a broad frequency spectrum (guidelines will be available in 2015, according to the directive). In the DELIVER project, we have used the algorithms suggested by ICNIRP in their guidelines, while EM Safety has worked with evaluations of these algorithms.
EM Safety has also been working with electromagnetic simulations, but not using CAD data of the actual vehicle, but instead simplified mock-ups. This part of the project has given us valuable experience on the whole process where a considerable problem is the process of converting mechanical CAD data to a usable model in the EM simulation system.
One important conclusion from both projects is that the levels of magnetic fields in electric vehicles can be kept within the limits that are considered to be safe for humans. There are many possible areas for future work. Practical guidelines to complement the EU directive in e.g. adding effects from multiple frequencies, and special averaging of external fields must be developed. There is also much to be done to develop methods to include EM simulations in the workflow of designing new vehicles.
Fulfilling requirements for Electromagnetic Compatibility (EMC) is a legal requirement in Europe and to be able to E-mark a vehicle the UN ECE R10 requirements should be fulfilled. This is also the goal for this project. As mentioned above, recommended techniques for designing and building the vehicle to achieve low levels of electromagnetic fields are presented in D2.8.
All electric systems are possible sources for EMC disturbances, and all electric systems are also possible victims. This makes it difficult to reduce the EMC coupling by increasing the distances as with EMF. Shielding and shielding strategy are therefore important tools. Achieving good shielding and maintaining the shielding performance during the build process in a prototype vehicle is extremely difficult. Good EMC practice is to mount all control units and to route all cabling close to metal structures. This is difficult to combine with the engineers’ need of good and frequent access to control units and cables. This is normal in a prototype vehicle, but the consequences of this become clear in the measurements.
EMC simulations are more challenging than EMF simulations at low frequencies, and were therefore not planned in the project. EMC issues are mainly noticeable at frequencies above 100 MHz and metal structures that have a negligible influence on low frequency simulations will be noticeable here. Detailed models for the electromagnetic emission from all control units and cables would also be needed. All deviations from the mechanical and electrical models would also affect the final result. The conclusion is that predicting EMC performance using full vehicle simulations is today not a reliable tool.
EMC measurements are performed in the same facility as the EMF measurements. The initial plan was to perform all the tests included for E-marking. However, the tests show that the DELIVER vehicle in the present status is not compliant. The immunity test that is performed shows compliancy but the emission levels are too high. This is to some extent expected since building a prototype vehicle is a balance between several conflicting requirements such as easy access, price and performance. The full test plan is not completed for several reasons; the early tests show that the vehicle is not compliant, and due to the high emission levels the immunity is questioned and we do not want to risk damage to any systems.
Potential Impact:
The dissemination activities of DELIVER aimed to spread the project objectives and findings to a wide range of target groups. To address these effectively, different categories of key stakeholders were identified:
- Automotive industry professionals
- Logistic operators
- Public authorities including decision makers
- Research institutions
- European Institutions
- General public
A project contact database was created and permanently updated. This was managed by IKA and together with contacts provided by Polis, it served as a basis for the newsletters. Moreover, DELIVER actively facilitated synergies with complementary projects and pilot programmes at European and national level. The purpose was to cover and enhance areas that complete the holistic view of urban delivery systems using electric LCVs.
IKA developed a first DELIVER leaflet in January 2012. This was distributed at the TRA conference in April 2012. A second version also served as the project’s identity card towards a wide range of stakeholders, and provided an update on the DELIVER concept.
Once the prototype was ready a video explaining the DELIVER concept and results was produced. In addition, a public gallery of images was made available.

Website
IKA was responsible for developing and maintaining the DELIVER website. This has been launched in month 3 (January 2012); the registered domain is www.deliver-project.org. Public deliverables and other information on the project were made available on this website as soon as all project partners and the European Commission granted an approval for publication. During the course of the project, 4,230 individuals visited the website 6,015 times. It encountered 8,699 page views, which is about 240 page views/months in average.

Newsletter
Polis coordinated the edition of four DELIVER electronic newsletters (see Deliverable 5.4) which served to keep interested stakeholders informed on the project’s activities and progress. Based on the input received from all partners, the newsletter was published according to the project progress and results. The newsletter allowed to further extend the project’s contact database, through the subscription option on the project website.
The newsletters were released on July 2012, September 2013, September 2014 and January 2015. These were sent by Polis to more than 2,000 contacts and were opened, on average, by 500 receivers.

Publications and participation in external events
The project partners presented the DELIVER results in different conferences and media, including, in relevant events, the exhibition of the prototype (see also the project website):
- Polis Conference, 27-28 November 2014, Madrid (Spain): the DELIVER prototype was part of the conference exhibition
- EUCAR Conference, 04 November 2014 Brussels (Belgium): prototype viewing, members of the DELIVER advisory board attended
- DELIVER - An Innovative Vehicle Concept for Increased Energy and Transport Efficiency
Aachen Colloquium Automobile and Engine Technology, 06-08 October 2014, Aachen (Germany)
- Design and Build-Up of an Electric Delivery Vehicle
Published first in ATZ 116 (2014), No. 9, pp. 34-40
- DELIVER - An Innovative Vehicle Concept for Increased Energy and Transport Efficiency
Aachen Body Engineering Days, 23-24 September 2014, Aachen (Germany)
- DELIVER - EMF & EMC
EMC Europe 2014, 01-04 September 2014, Gothenburg (Sweden)
- Smart Urban Freight Conference 2014
12 June 2014, Newcastle (UK), in cooperation with the EC funded projects on city logistics SMARTFUSION, BESTFACT, FREVUE, STRAIGHTSOL (http://www.straightsol.eu/) and SMARTSET (http://smartset-project.eu/)
- DELIVER - An Innovative Vehicle Concept for Increased Energy and Transport Efficiency
Outstanding Paper Award Winner
FISITA World Automotive Congress, 02-06 June 2014, Maastricht (The Netherlands)
- DELIVER: The Next Generation of Electric Delivery Vehicles presented at FISITA 2014
Press Release, DELIVER Consortium, 27 May 2014, Aachen (Germany)
- DELIVER - Fully Electric LCV with Intelligent Body Design Features
Strategies in Car Body Engineering 2014, 11-12 March 2014, Bad Nauheim (Germany)
- DELIVER - Design of Electric Light Vans for Environment-impact Reduction
Strategies in Car Body Engineering 2013, 19-20 March 2013, Bad Nauheim (Germany)
- Development of the DELIVER Vehicle Design Concept
Technical Paper, Coventry/Aachen, June 2012
- Development of a Fully Electric Light Duty Vehicle - The DELIVER Project
Transport Research Arena (TRA) 2012, 23-26 April 2012, Athens (Greece)

The planed exploitation activities of the results of the DELIVER project are listed in the following for each partner:

IKA
The main interest of IKA is to exploit the results by means of sharing the ideas and results among the scientific community, consulting services for interest parties and improving the education of the future generation of engineers. As a university partner IKA does not follow any profit-oriented approach.

VOLKSWAGEN AG
- Usage of the market, logistics and technology report for future ICE and electric light commercial vehicles (D1.1).
- Usage of the business case and calculation spreadsheets for evaluation of economic efficiency for future vehicle projects (D1.2). Especially the great effect of the right side entrance of Deliver on the TCO is very important.
- Requirements from customers in the advisory board as a basis for future vehicle concepts (D2.1)
- Evaluation of different vehicle layouts (cab forward vs. central seating position vs. conventional) as a basis for vehicle development in the concept phase (D2.2/D2.3).
- Development of an ergonomically optimized interior concept, new findings on optimized vision and ingress/egress by the ergo buck and in the demonstrator (D2.4).
- Interesting detail solutions from the project team that were not chosen for the selected concept (Michelin Active Wheel, materials [e.g. Inrekor], etc.)
- Detailed information on the performance and design of the Michelin Motorized Wheel.
- Compact knowledge of today’s and future EMC regulations and its fulfilment (D2.7/D2.8/D4.5)
- Generally interesting aspects of different approaches and forms of collaboration within the project team; enhanced network of possible new business partnerships

LEC 2 LIMITED
Partner terminated because of insolvency, no further exploitation of results.

MICHELIN
Objectives of the DELIVER program for Michelin
- Built up a demonstrator to evaluate Motorized Wheel technological feasibility on a new type of vehicle (Light Duty Vehicle, or LDV) with an original mechanical transmission design never tested before on Motorized Wheel: a two gear ratio set allowing a very high level of torque at low speed, compatible with the requirements of LDV
- To increase Michelin’s knowledge and performances in terms of acoustic (especially in the transmission pinion design)
- Technical learnings of the DELIVER program are positive for Michelin and, without European funding, it would have been more difficult or much slower to have the opportunity to test this new design in a concrete demonstrator application
- Some initial low performances observed in previous demonstrator applications (lack of dynamic performances at low speed; high noise emitted by the system and transmitted to the vehicle) have been analyzed and their negative impacts have been much reduced on DELIVER system. Nevertheless some technical challenges are not completely solved for market acceptance, e.g. noise. The dynamic performances reached by the motorized wheel for this application are a real success for this demonstrator.
- Despite the technical solution developed, Michelin observed in parallel the impossibility to find a profitable business model for Motorized Wheel technology on the markets and applications targeted by the DELIVER program compared to a more standard e-powertrain solution (central motor combined with a gearbox and differential)
- A two gear ratio set Motorized Wheel does have better performances than a one gear ratio solution (already designed and tested on previous R&D projects by Michelin) but is definitely increasing complexity and costs
- Industrial and costing topics are still to be addressed but today Michelin doesn’t get a viable business outlook for Motorized Wheel technology in the domain of automotive applications. Consequently Michelin decides not to investigate more in this technology in the future.
- Michelin’s ambition is definitely to stay a leading actor of sustainable mobility by pursuing in this domain R&D activities, internally or with partners.

SP
SP has gained new knowledge in the methods of using CAD data in electromagnetic low frequency simulations, and what results that can be expected. We have already submitted applications to national research funding programs for new projects were we will use and build further on this knowledge. One example is a project to study possible risks with wireless charging, a functionality that was not included in the DELIVER vehicle. Another example of further work is to implement the methods we are developing on a research level into the development process of the automotive industry.

HPL PROTOTYPES LTD
HPLP enjoyed the working partnership with the consortium. We are used to producing this type of model and there are always things that are learnt with any new project. The main benefits for HPL are as follows
- The integration of partners technology into a running prototype
- The development of the design and introduction of original ideas
- Manufacture of large screens
- Unique door configuration
- The manufacture of the bespoke steel roll cage structure for the vehicle
Working on this type of project has opened the way for other Euro project opportunities. HPLP recognises that our experience to be able to bring all the technology together at the end enables all the partners to show case their efforts and experience gained during the development of the programme.

CADEM & MOBIT
This study is the first complete battery pack study for the company and within this project, battery pack and BMS requirements were learnt from the rest of the experienced team members VW, Michelin, Lec2 and IKA.
Our company also find the courage to start developing industrialised versions of the BMS and battery pack for future applications. This will dramatically help the company to make deeper studies with EV applications.
Our company has its own ICE powered light weight commercial vehicle concept. This is still under development. After the completion of DELIVER we also developed a new version of battery pack with some progress in thermal management and higher energy density cells were used in that project. EV version of this vehicle is now running with our own developed battery pack on it.
DELIVER project researchs helped us to have a better understanding of EV’s and guide our studies for our own vehicle concept or for other local projects which may need BMS and complete battery pack.
Our next coming studies about battery pack will be based upon development of more safe, more energy dense, compact and affordable versions of the battery pack and BSM systems.
List of Websites:
www.deliver-project.org

Main contact/Project coordinator
Dipl.-Ing. Micha Lesemann
Institute for Automotive Engineering (ika)
RWTH Aachen University
Steinnachstr. 7
52074 Aachen
Germany
Phone: [phone]
Fax: [phone]
E-mail: [e-mail]