Powertrain Electrification for Jerk Reduction and Continuous Torque Delivery

Publication Type:
Thesis
Issue Date:
2019
Full metadata record
A mild-hybrid electric powertrain is proposed for the principal purpose of providing continuous drive torque using a single, dry-plate clutched transmission. The powertrain is optimised to deliver several benefits, in relation to cost, complexity, vibration (jerk), as well as dynamic and emissions performance. The powertrain proposed is a post-transmission type, with the motor being placed inline with the transmission output shaft, prior to the differential. This allows the powertrain to be controlled for providing continuous drive torque to the wheels during gear shifting and take-off, eliminating the “torque hole” due to disengagement of the clutch plate, and providing a degree of damping during clutch re-engagement. A pressure-based clutch model is used to modulate the electric drive torque to minimise torsional vibration during the gear shifting process, whilst engine speed is controlled proportionally to road speed to minimise discontinuity of rotational velocity during the re-engagement process. The system is designed as a driver assistance function, but can optionally be implemented with automatic clutch and gear actuation units. A rule-based energy management strategy (EMS) allows the powertrain to be additionally controlled for drive torque supplementation, battery recharging, brake energy recuperation, and electric vehicle (EV) crawl. System optimization is conducted on several levels. The system architecture is optimized to minimize modification cost from a typical conventional vehicle (CV) by careful consideration of powertrain topology. The selection of the post-transmission (P3) architecture was made to eliminate the cost and complexity of a dual-motor configuration whilst maximising the utility of a single electric motor (EM) using a sophisticated EMS. The electric power components, primarily the electric motor and battery are optimized for component size and cost based on benchmarking criteria and power needs analysis. A V-cycle development process using model-based design was followed. A hardware-in-the-loop (HIL) vehicle model was built in a virtual environment, allowing testing of performance and comparison with the CV by running the software model through standard drive cycles in Advanced Vehicle Simulator (ADVISOR). Certain model parameters were tested on a HIL bench and refined, and then the model was downloaded onto a real-time controller (dSPACE MicroAutoBox II) for implementation in the prototype validation stage. The powertrain is designed to meet the requirements of a typical light vehicle. The prototype powertrain was built into a 1990 Mazda MX-5 (Miata) body, which was modified to fit the additional powertrain components selected through the optimization process. These components include a 1.2 KWh, 96 V LiFePO₄ battery pack, a 10 KW cont./30 KW pk. permanent magnet motor, four quadrant 600 A motor controller, battery management system, electronic throttle system, and supervisory controller. The vehicle was instrumented for clutch pedal position, clutch line pressure, gear lever position, brake pedal position, brake line pressure, throttle pedal and butterfly position, engine manifold vacuum, transmission output torque, transmission output speed, and electric motor torque. The battery is also instrumented through the battery management system (BMS) and is capable of logging individual cell voltages and temperatures, as well as pack statistics including state of charge, depth of discharge, current and voltage. As implemented, the system is designed to suit low-end vehicles typically sold in developing nations, and serves as a way to reduce fossil-fuel dependency, introduce fleet electrification (particularly in areas where access to electricity is unreliable), and improve urban air pollution whilst also improving vehicle driveability through powertrain refinement. In developing the vehicle for such purpose, a tight manufacturing cost control of no more than 105% of the manufacturing cost of the base vehicle is imposed. With changes to the benchmarking criteria and control, the powertrain architecture could also be used for dynamic performance enhancement. Results of experimental testing of the prototype against the CV are presented and discussed. The experimental testing encompasses acceleration, jerk, torque continuity, and emissions. Results validate the modelled system to a high degree, showing that the powertrain meets its design objectives, effectively providing continuous drive torque, substantially reducing torsional drivetrain vibrations manifested as longitudinal jerk. Based on the test results of the prototype, a number of refinements, optimizations, and further works are suggested. Principally, the major system improvements include the implementation of an auto-clutch system (ACS), computerized gear selection, or the combination of both in the form of an automated manual transmission (AMT). These improvements eliminate the need for predictive algorithms required to fill the torque hole, as the target speed and torque are known at all stages during the gear selection process. Further refinements include optimization of the traction battery, new approaches to motor control, and further cost reductions in the transmission componentry through the use of electronic synchronization control.
Please use this identifier to cite or link to this item: