tomofNV said:
Your basic description of regen is correct.
The operating mechanism differs in details between AC induction motors, variable reluctance motors, and permanent magnet motors but all of these motors have a stator with poles, each pole having its own windings. Voltage is applied across these pole windings sequentially, sinusoidally in time, to create a rotating magnetic flux. Magnetic flux then varies sinusoidally in time across the air gap between a given stator pole and the rotor.
There are two main parameters of importance: The magnetic flux density through the air gap between the rotor and stator, and the frequency of the rotating magnetic field of the stator. The flux density is proportional to the voltage applied to the stator poles and inversely proportional to the frequency.
Glossing over differences in details of operation, when the stator frequency is greater than the rotor frequency of rotation the stator field results in a torque on the rotor. When the rotor has the greater rotational frequency it induces an emf (electromotive force, which is a voltage) in the stator, driving current to the batteries.
To accelerate the vehicle when the accelerator pedal is pressed, the frequency and voltage are increased to increase the rotational frequency of the stator magnetic field while maintaining the required air gap magnetic flux.
When you back off on the pedal the stator frequency is reduced. If it is reduced such that it is lower than the rotor frequency the time varying magnetic flux of the rotor then induces an emf in the stator windings (described by Faraday's law) and current flow into the batteries - regenerative braking. The more you "relax" (back off) the pedal, the greater the emf and work done in slowing the vehicle.
The frequency and voltage of the motor controller are varied differently in D and L mode when the accelerator pedal is relaxed, resulting in different amounts of regenerative braking. Basically, the "slip", the difference in rotor and stator frequencies, is not permitted to become as great in D mode when regeneratively braking so a smaller emf is induced in the stator and less work is done to decelerate the vehicle.
Although I agree with the technical aspects, that doesn't actually explain the different levels of regen since the motor is not connected directly to the battery, and also doesn't explain how it would vary that amount. That description is much more accurate for how the car controls the speed of the car, but not the amount of regen.
Somewhat less technically, you have this:
Motor/Generator <--> SPIM <--> Battery
(there's a distribution block in there but that's irrelevant to this discussion)
The SPIM (single power inverter module) switches power between DC (battery side) and 3-phase AC (motor side).
A generator, like any power source, provides a voltage. That voltage is generated quickly, but is meaningless unless you have something to draw current. As soon as you start drawing current you create resistance which requires force to maintain, otherwise the speed will reduce, which reduces the voltage.
You may have seen exhibits at your local science center, but if not imagine this: you have a hand crank flywheel, and a few old incandescent lights that you can turn on and off with some switches. Crank the wheel with no lights on and it takes a little bit to get up to speed, at which point it's easy to keep turning (voltage is created but no current draw = no additional resistance other than normal friction). Turn a light on and suddenly you can feel that it's harder to keep the wheel spinning. Turn 2 lights on and it's harder still. 3 lights on and it's really hard to keep the wheel turning, and it starts to slow down.
Or, look at it another way - get the wheel spinning as quickly as you can and stop. It may take 60 seconds to fully stop with no lights on, 20 seconds to stop with 1 light on, 10 seconds to stop with 2 lights on and 5 seconds to stop with all 3 lights on.
Increasing resistance due to added load = more force required = more slowdown. So if you can control the amount of load (current drawn) then you can control the amount of resistance, or force required, or slowdown. I can control the load based on the number of lights that I turn on, but what if I only have 1 light?
What if I put an automatic switch that turns the light on/off with a given frequency. If the switch has a 50% duty cycle (turns on and off say once per second) then the resistance will be halved, and will require half of the force, so the wheel will slow down half as quickly. If the duty cycle is 25%, then I require 1/4 the force. Unfortunately we're going to see the light flicker as it goes from off to full brightness once per second.
Instead of once per second, however, what if I do that 100 times per second, or 100,000 times per second? Instead of seeing it flicker, I'll just see it at about half brightness with a 50% duty cycle. At 25% duty cycle, the brightness will still appear constant, but at 1/4 the brightness as full.
This is what our SPIM does during regen through the use of MOSFETs (and a bunch of other complicated stuff, but that's the main relevant part).
The force is the kinetic energy of the vehicle. The car monitors the power output of regen, and tells the SPIM to control the duty cycle so that it maintains the exact amount of power (deceleration) that it wants.
