電動機械式選換檔執(zhí)行機構方案的設計【AMT】,AMT,電動機械式選換檔執(zhí)行機構方案的設計【AMT】,電動,機械式,換檔,執(zhí)行機構,方案,設計 400 Commonwealth Drive, Warrendale, PA 15096-0001 U.S.A. Tel: (724) 776-4841 Fax: (724) 776-5760 Web: www.sae.org SAE TECHNICAL PAPER SERIES 2004-01-1322 Review and Development of Electromechanical Actuators for Improved Transmission Control and Efficiency A. J. Turner and K. Ramsay Ricardo Driveline and Transmission Systems Reprinted From: Transmission hydraulic, electrohydraulic, and electromechanical systems. Transmission data taken from tests undertaken by Ricardo shows hydraulic systems may represent 50% of the total transmission loss, but have the advantage in terms of actuator size and weight, with the hydraulic pressure source placed elsewhere in the transmission where space is at less of a premium. Electrohydraulic actuation (i.e. with an electric power source) attempts to provide a compromise between electromechanical systems and hydraulic technology. For example, the replacement of the mechanical drive to a hydraulic pump with an electric motor offers an improvement in duty cycle control, allowing better tracking of open circuit hydraulic pressure demand. However, losses associated with leakage and flow through components is still present. An alternative to this is a closed circuit hydraulic system 1 where an electric motor is able to vary hydraulic pressure by changing its speed and direction. Electromechanical devices can be highly efficient, 95% in some cases, and also offer excellent duty cycle control. However, their performance at high temperatures and in terms of force/torque density is generally poor. Because of this, electromechanical devices may need to be located remotely from the point of use, and unless some mechanical system is used to trade actuator speed, which is generally high in electrical devices, for force/torque capability, actuator dimensions may be unacceptable for a given specification. ELECTROMECHANICAL TECHNOLOGY Electromechanical technology can be separated in to two main classes; rotary devices such as conventional electric motors, and linear devices such as solenoids and linear motors. ROTARY ELECTRICAL MACHINES Rotary electrical machines can be further subdivided into two groups, limited angle actuators and motors. Motors can be precisely controlled in terms of torque, acceleration, speed and position, and there are a number of well established electric motor technologies each having different characteristics. Limited angle actuators may include devices such as rotary solenoids. Torque production in rotary devices is accomplished by two means; the alignment of magnetic fields (excitation torque) and the alignment of ferromagnetic materials (reluctance torque), and all machines use one or both of these mechanisms. This paper deals with three different motor topologies, brushed DC, brushless permanent magnet and switched reluctance (SR) motors. The induction motor is not considered here as traditionally this motor is used in high power applications where weight is less of an issue or the cost of using permanent magnet machines would be too great because of motor size. In automotive applications where power requirements are relatively low, there is little cost benefit in using induction motors, and their lower efficiency and torque density results in larger motors for a given torque requirement. Brushed DC motors have dominated motor technology used in the automotive environment, and this is largely due to cost. However, as motor abundance and power requirement has increased, it is becoming necessary to move away from brushed DC technology in some areas because of efficiency and torque density issues. In general, the torque produced by an electrical machine is proportional to the rotor volume, the electric loading and the magnetic loading in the machine. The electric loading is limited by the thermal properties of the motor, with force cooled machines being able to stand higher electric loadings. Magnetic loading is governed by the magnetic circuit design and type of materials used in the machine construction. Brushed DC motors This machine is constructed using either a stator winding or magnets and a rotor winding that is supplied with current via a commutator and brush gear. For low power applications, the stator winding is usually replaced with ferrite magnets. The brushed DC motor is cheap, but compared to other designs, has many drawbacks. Although brushed motors account for the majority of electric motors in automotive applications, the commutator and brushes are the limiting feature of the motor as they limit both maximum armature winding current (and hence torque) and the maximum motor speed. Therefore, for high power, fast response applications, brushed motors may be too large and slow, and alternative technologies need to be employed. Brushless motors The brushless motor class can be split into a number of different categories but only three main designs are considered here. These are the brushless permanent magnet AC and DC machine, and the Switched Reluctance (SR) motor. Permanent magnet brushless DC and AC machines are almost identical in construction but the main distinguishing feature between the designs is the back-EMF waveforms, with brushless AC (BLAC) machines having a sinusoidal back-EMF waveform and brushless DC (BLDC) having trapezoidal back-EMF waveform. Although the power electronic drives used to control these two machines are similar in layout, their control is somewhat different and each requires rotor position information of differing degrees. As the AC machine requires sinusoidal currents, accurate rotor position information is necessary and is usually measured using an encoder or resolver. The DC machine, which requires square wave current waveforms, needs to experience a change in coil currents every 60 (electrical) or rotation, therefore coarse rotor position information can be tolerated which may be achieved using Hall effect sensors. However, if the machines are to be used in position servo applications, then a position sensor such as that used in the BLAC case must also be used for rotor position feedback in the DC machine case, giving no significant advantage in using a DC machine over an AC machine. If the application were speed regulated, then the DC machine would fulfill the requirements without the accurate position feedback needed by AC machines. There are also a number of subtle differences between the operation of the machines leading to slight differences in torque density, and hence torque/rotor inertia ratio, speed range and control drive electronics VA rating 2. Brushless DC motors may also be supplied with unidirectional currents which has a number of advantages and disadvantages 3. As cost reduction is of great importance, and as the silicon cost of a drive system can be a significant portion of the total material and production cost, the use of unipolar current drives to reduce the number of silicon devices needed is of interest. There are however, a number of significant drawbacks to this control scheme. The windings are poorly utilised as they may only be excited for half the maximum available interval for torque production. Some unipolar drive configurations do not allow four-quadrant operation (motoring and generating in forward and reverse directions). Inductive energy is dissipated in the windings of the motor and not returned to the supply as with conventional bipolar drives and they can exhibit greater levels of torque ripple than bipolar driven machines. However, bipolar drive problems include the risk of shoot-through faults where both switches of a phase may conduct simultaneously, resulting in a short circuit across the DC link, a situation that does not effect unipolar drives. The performance of unipolar drives may be improved by increasing the number of phases but at the expense of increasing cost due to the increased silicon device count in the drives, the greater coil count in the motor and more complicated lamination. Permanent magnet motors can have surface mounted magnets or have magnets embedded in the rotor. A rotor with embedded magnets exhibits saliency effects that can add to the peak torque of the machine. The switched reluctance machine can be described as a doubly salient, singly excited motor. The salient rotor construction leads to the machine having a high torque/rotor inertia ratio as sections of the rotor are removed to create the rotor teeth. Torque production is brought about by the alignment of stator and rotor teeth and the phases are electronically commutated to produce a continuous torque with the sequence of the phase excitation determining the direction of rotation of the rotor. As the motor functions with unipolar currents, a simpler drive may be used to control the machine, pointing to a reduction in cost and making fault mitigation easier. The lack of permanent excitation field (magnets) gives the machine a high degree of fault tolerance with benign failure modes, and operation is still achievable when the machine is in a fault state. However, the pulsed nature of torque production leads to a number of undesirable properties such as a high degree of torque ripple that causes speed fluctuations which is exacerbated by the low rotor inertia. SR machines may have a torque ripple of the order of 80% 4 and the machine can also be acoustically noisy. Motor Type Torque density Speed range Efficiency Cost Brushed DC Low Low Low Low BLDC/ BLAC High Medium High High SR Med High Med Med Table 1 Relative comparison of motor types LINEAR DRIVES Linear drives create force in a similar way to that in which rotary machines create torque, and as such are subject to the same drawbacks, i.e. they have poor force capability relative to hydraulic devices. However, they do remove the need for mechanical linkages if the force requirement is not too severe. They can be classified into groups as follows: Solenoids Basic solenoids provide attractive forces only and the solenoid is reset to the off position by a spring. However, bi-directional motion is possible from these devices with correct design. The force profile of a solenoid is particularly non-linear, being highly dependent on the working airgap between the solenoid plunger and the coil housing which makes their control demanding. Due to the force/displacement characteristic, solenoids are mostly used in a bistable mode of operation. Some proportional controllability is possible, but the penalty in this is a reduction in maximum generated force. To a first order, solenoid forces can be described using equation (1). 2 0 2 2 )( g l ANI F = (1) Where N is the number of turns of wire in the solenoid coil, I is the coil current, A is the working cross sectional area of the solenoid, 0 is the permeability of free space and l g is the airgap length. From equation (1) it can be seen that as the working airgap l g reduces, the output force becomes very large, making control difficult. Moving coil actuators Moving coil actuators include devices such as speaker coils, which feature very fast responses. However, output forces can be limited due to the coil being in the working airgap, which limits the ability to cool it, robustness is also an issue. Moving magnet actuators In general, moving magnet linear actuators are very robust as they have no flying leads to the magnet shuttle, and the stationary winding arrangement on the stator allows efficient removal of any losses. The use of magnets could be problematic in hot environments because of their poor temperature capability. Linear motors Linear motors may be considered as multi phase moving magnet, coil or solenoid actuators, and are available in the same variants as rotary machines. As in rotary machines, phases are commutated to allow a greater degree of travel. POWER SUPPLIES AND 14/42 VOLT SYSTEMS FOR ELECTRIC ACTUATORS Ultimately, current levels drawn by electric devices are governed by the capability of the battery and the motor operating voltage. The nominal current demand and voltage also defines the type of power electronic devices required. In this respect, higher voltages and lower currents are more appropriate as they are easier to deal with from a power electronics point of view. There is also a greater degree of choice in terms of components compared to lower voltage / higher current rated devices. In motor design terms, once a satisfactory magnetic circuit has been designed and the device operating speed has been specified, the operating voltage dictates the form of the windings with regard to number of turns and wire diameter. For a given machine power the amount of current drawn by a 42 Volt machine will be 3 times less than for a 14 Volt machine. This dictates that the number of winding turns must be altered to satisfy the motor Amp-turns requirement, which is a constant set during the design of the magnetic circuit. For correctly designed 14 and 42 Volt machines, the winding losses should be approximately equal. The voltage levels cause differences in the drive losses and useful available voltage. The higher current demanded by 14 Volt machines leads to greater I 2 R losses in the drive electronics. The voltage dropped across each power semiconductor device reduces the applied winding voltage, and in the case of brushed machines, the brush voltage drop is also a factor. As these voltage drops are essentially fixed for both 14 and 42 Volt cases, the 14 Volt system suffers a larger percentage loss of useful applied voltage, giving the 42 Volt system the upper hand in terms of both efficiency and performance. However, if the duty cycle is sufficiently short, these losses should be relatively negligible providing the device has stable states where power consumption is low or zero. Figure 1 shows a comparison of the loss components between 14 and 42 Volt motor drives. 14/42 Volt comparison 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 14 Volts 42 Volts normaliesed losses Motor Drive Figure 1 Comparison of 14 and 42 Volt motor drive system losses AMPLIFICATION OF FORCE AND TORQUE FOR ELECTROMECHANICAL DEVICES As stated earlier, electromechanical devices produce relatively low levels of specific torque and force relative to conventional hydraulic actuation, and to compete with hydraulics, a system is needed to amplify forces and torques. The selection of a mechanism is important to ensure correct system dynamics, system efficiency (losses in the mechanism and whether it provides a self locking feature) and system reliability (wear rate). All three criteria are linked and good design should lead to all the factors being well optimised. Some examples of mechanical systems for transmission control can be found in 5. SIMPLE LEVERS An electromechanical device may be connected to the clutch release lever as is usually done with the cable from a manual transmission clutch pedal. However, this type of device offers no locking feature requiring a constant force from a linear device to operate it. Rotary devices must also include a motion conversion linkage, but the conversion stage may be designed to be self-locking. POWER SCREWS Power screws cover a number of different designs that convert rotary to linear motion. The main types of screw can be divided into sliding (square or Acme thread type) and rolling contact (ball and roller screw type) variations. A power screw alone may not be able to provide the necessary reduction gearing to multiply the motor torque to a usable level. A simple gear set or epicyclic may also be needed between the machine and power screw, determined by the space available for the machine and the lead of the power screw. WORM GEARS Although worm gears have a low efficiency, they provide a method of packaging a set of gears with a large reduction in a small volume. However, worm gears introduce backlash, and the poor efficiency of the gears will increase the size of the motor due to increased torque requirement. A further linkage is required to convert the rotary output from the worm gear set to a linear force. ELECTROHYDRAULIC CYLINDER A closed circuit hydraulic system may be used so to allow electromechanical devices to be easily placed away from high temperatures or to areas where space is less limited. CLUTCH ACTUATOR DESIGN AND EVALUATION (a) (b) Figure 2 Basic design concept of for motor driven power screws showing (a) external and (b) integrated power screw arrangement The design and evaluation of a simple clutch actuator was conducted in terms of the rating the motor and conducting simulations to evaluate performance. The aim was to design an actuator such that the mechanical linkage used was integral to the torque producing mechanism to give an actuator that minimised volume, maximised output force and gave the required dynamic response. The actuator was based on integrating a power screw with a brushless DC motor, and figure 2 shows two examples of such designs. The brushless DC motor was chosen because it has a high torque density, low torque ripple and high efficiency. The power screw was chosen because it allows the multiplication of force and allows motion conversion in one unit. The basic requirements of the system should be: High power/weight ratio giving the lowest machine mass High torque/inertia ratio giving the best acceleration possible Smooth production of torque particularly at low speeds to minimise speed variation and achieve good positional accuracy. Controlled torque at zero speed High maximum speed of operation. High efficiency and power factor to minimise drive VA requirement Compact integrated design with the application Good frequency response Low backlash Low cost A balance of these attributes is required to optimise a particular system. Many factors influence the design considerations and use of permanent magnet machines, however, for simplicity the motor considered was a conventional surface mounted permanent magnet machine. Magnetic materials used in surface mounted rotors such as Neodymium Iron Boron (NdFeB), are susceptible to damage from high ambient temperatures and corrosive environments because of their high iron content. Ferrite magnets, which may also be used in brushless machines, are much cheaper, more robust and have a better temperature dependent characteristic than NdFeB, but are poorer in magnetic terms, leading to an increase in the amount of magnetic material required. Samarium Cobalt (SmCo) magnets produce fields of the same order as NdFeB and have a much better temperature characteristic, but are very expensive and are very brittle which makes handling them difficult. NdFeB magnets were selected because they offer the best performance/cost compromise, although temperature performance is still a significant issue. Figure 3 shows an example of how motor efficiency varies as a function of ambient temperature for two different types of NdFeB magnet. The higher grade, more expensive magnet exhibits greater efficiency through the entire temperature range. Motor performance vs temperature 0 10 20 30 40 50 60 70 80 90 100 20 70 120 170 Temperature (degrees C) Motor efficiency Low grade magnet High grade magnet Figure 3 Motor performance for two different types of magnet versus ambient temperatures MECHANICAL DESIGN Initially, a figure for the required output force was needed to establish the torque requirement of the motor, which in turn leads to an approximate motor volume. The required torque is dependent on the load, system friction, the screw lead, the screw diameter and the system inertia. The torque requirement for a power screw having a lead L, a coefficient of friction and a diameter D that provides a force F is: + = LD DLFD T 2 (2) To satisfy the criteria of low or zero power consumption whilst the screw is stationary, the screw was designed to lock. Power screws may become self-locking when LD or tan where is the lead angle of the power screw. The screw lead was selected to give high positional accuracy, repeatability, rigidity and