While there are detail differences in the implementations from each manufacturer, the basic operating principles are largely the same. The key to the efficiency advantage of hybrids is their ability to recover kinetic energy that is normally dissipated as heat when the vehicle is slowing down and then store and release that energy to provide propulsion later. This is known as regenerative braking, and we're going to discuss aspects of regenerative braking that apply equally to pure electric and extended range electric vehicles like the Nissan Leaf and Chevrolet Volt.
Non-hybrid vehicles use hydraulic pressure to apply a friction force to brake rotors or drums. That pressure is generated by the driver applying the brake pedal. The pressure generated by the driver is typically amplified by a vacuum booster. The amount of deceleration is then in proportion to how hard the driver applies the pedal. The goal of the electro-hydraulic system on hybrid vehicles is to maintain this relationship between vehicle deceleration and pedal apply.
Maintaining this relationship requires blending of the hydraulic friction braking and the regenerative braking. Therein lies the rub. Let's take a walk through these brake blending systems to see how they work.
These systems utilize a number of additional sensors to determine what the driver is asking for from the brakes and then manage regen and friction braking to achieve it. Typically, a brake pedal position and sometimes a force sensor are used to derive the driver's intent and calculate a desired level of deceleration. For example, if the driver presses the brake pedal down about half-way, that may equate to about 0.5 g of deceleration. That deceleration intent is converted into a desired amount of braking force or, more specifically, torque (rotational force).
At the same time, the system is monitoring the wheel speeds using the sensors that are also used in the anti-lock brakes. This information is used to determine the type of surface the vehicle is on, including ice, snow, gravel or dry pavement. The instantaneous behavior of the wheel speeds can provide a lot of clues about the nature of the road surface, but it isn't perfect. This "not quite perfect" reality will be an ongoing theme throughout this discussion.
If the driver's intent exceeds what the road surface can support, the total braking force is reduced to what is available. For example, if the car is on a snow-covered road, the maximum deceleration may only be 0.3 g so the total braking force applied needs to be reduced accordingly. Once the driver's intent and the road surface have been determined, the system has to figure out how much hydraulic pressure to apply to the friction brakes and how much regenerative braking energy to feed to the battery.
Because we are talking about hybrid vehicles where efficiency is a priority, the systems will always try to maximize the amount of regen, if possible. With the nickel metal hydride batteries used on current hybrids, the amount of regenerative braking is typically limited to about 0.3 g because of the rate at which the batteries can absorb energy. The amount of regeneration that is possible at any point in time is also limited by the battery state of charge. If the battery is fully charged, no regen is possible.
The amount of braking torque provided by regeneration is pretty straightforward to calculate and is directly proportional to the voltage feedback from the generator. The control system subtracts the regeneration torque from the total desired brake torque. The leftover amount of torque is then realized by the hydraulic system. This is where things get really hard.
In these electro-hydraulic systems, the amount of pressure on the friction brakes is semi-independent of the driver's demand on the pedal. In order to control this, the system must have a mathematical model of the relationship between pressure and torque for the brakes.
Hydraulic friction brakes have a lot to recommend them. They are a very effective means of providing the large quantities of brake torque needed to stop a vehicle. They are also comparatively inexpensive and generally very reliable. However, because they rely on the act of pressing a friction material against a moving surface, the physical characteristics change over time, sometimes a very short time.
Remember that the law of conservation of energy tells us that energy can neither be created or destroyed, only transformed. The whole premise of friction braking is that kinetic energy of the vehicle is converted to heat energy by the brakes. As the vehicle slows, the brake friction material heats up and as the relationship between apply pressure and braking force changes. Other factors also cause that relationship to change, including humidity and pad and rotor wear.
The control software must incorporate adaptive control algorithms to help compensate for these changes but, just like the road surface estimations, these are not perfect. Brake systems engineers spend thousands of hours testing and developing these algorithms to make them more robust and this is one of the major factors that leads to extending the time it takes to bring hybrid vehicles to market.
Once the desired friction brake force is determined, the hydraulic control unit tries to fill the gap between the total and regen force. One of the downsides of regeneration is that no braking force is available unless the wheels (and thus the motor/generator) are turning. Even at low speeds, regen is limited. As the vehicle slows to a stop (typically between five and seven miles per hour) the regenerative braking force is ramped down and the friction force is ramped up.
This is where Toyota appears to be having its problem. If the hydraulic model isn't completely accurate, it can cause changes in vehicle deceleration even if the driver is holding the brake pedal steady. If the hydraulic model is overestimating the brake torque, then it will apply too little pressure, causing a perceived loss of braking.
The control can also be influenced by rough pavement. The wheel speed sensors rely on a toothed ring moving past a magnetic or hall-effect sensor. At low speeds, the frequency of the sensor pulses decreases and with it so does the accuracy of the signal. If the car hits a bump or pothole at low speed, it can often appear as though the wheel is going to lock. If this happens, the brake pressure is reduced to compensate. This, too, can cause what appears to be a loss of braking to the driver.
Because the vast majority of braking in the real world is in the range of 0.2-0.3 g, any loss of braking tends to be quite noticeable. The actual difference in stopping distance may only be a foot or two, but in traffic that distance can be critical to avoiding a fender-bender or something worse.