I would like to find out how quickly and frequently a reversible electrical machine (EM) can be made to switch functions from a motor to generator and vice-versa - and what are the parameters of the EM that determine this.
-
\$\begingroup\$ How quickly can you do the mechanical switching, between driving the shaft or loading the shaft? \$\endgroup\$– Elliot AldersonSep 6, 2018 at 19:38
-
3\$\begingroup\$ Not sure what you mean, but a rotating motor generates EMF anyway, to drive it electrically the driver overcomes it with higher voltage. If you want to instantly take energy from the motor- no problem, just stop diving it. Takes few microseconds at most. This is well implemented in hybrid cars. \$\endgroup\$– user76844Sep 6, 2018 at 19:39
-
1\$\begingroup\$ Welcome to EE.SE! Keep in mind that "Is it possible ...?" is a yes/no question. In this case, the answer is "Yes". If you're asking us to design it for you, that would be too broad. You would need to specify something about the level of performance you expect as well as what constraints you have on the implementation. What is your specific question? \$\endgroup\$– Dave TweedSep 6, 2018 at 20:21
-
\$\begingroup\$ Apply PWM and you're doing exactly that. The angular velocity can be measured via the back-emf generated during PWM off-periods. \$\endgroup\$– ChuSep 6, 2018 at 21:30
4 Answers
This sounds like an x-y problem.
There are two main timeconstants associated with an electrical machine
- Mechanical \$\tau = \frac{J}{b}\$
Where J is the rotor+system inertia and b is the damping coeficient - Electrical \$tau = \frac{L}{R}\$
Where L is the inductance of the stator and R is the resistance of the stator.
Likewise it is highly dependant on the two sources of power. The electrical source to permit motoring and the mechanical source to permit generation. Likewise the ability of both domains to absorb the power flow from the other.
how quickly and frequently a reversible electrical machine (EM) can be made to switch functions from a motor to generator and vice-versa
Well what do you class as rotating? Different disciplines may have a different requirement of the minimum definition of rotation... conceptually any arc movement is rotating and thus "how quickly" then is dominated by the electrical time constant and the ability to change the direction of the stator current
What about a +-10degree twitch? 359degree? With an ever larger arc being considered, the mechanical time constant become more and more important. If you define "rotation" as a 180degree movement then the load torque as well as the ability of the motor to generator torque will drive the time needed to switch between the two quadrants.
So besides the limitations being within the electrical domain to change between sink & source and the mechanical domain to change between sink and source, one limitation on electrical machines is the spline and at which point it begins to exhibit signs of plasticity.
-
\$\begingroup\$ Good formulae to compare (+1), but comes down to frequency response and attenuation of freq depending on n Poles per Rev within 360 deg at some Hz x60 =RPM \$\endgroup\$ Sep 6, 2018 at 20:48
It depends on the motor characteristics.
Mainly, the coil inductance, ( core type and permeability), coil resistance, flat wire wound, low L; shunt or series, synchronous or not, Reluctance, Induction or BLDC, etc.
Time Constant of Current Step response
Basically, the response time to current is T=L/R and the rotational time depends on the number of phases per revolution, RPM and moment of rotational inertia vs torque that affects acceleration.
So, in theory, it could respond within a rotation or may not swing much compared to full scale. It is not something you choose to do carelessly.
Shoot-thru
Shoot-thru is a fault condition when drivers are active in both directions at the same time. This is avoided with dead-time on the order of 1us. But for changing directions twice the max current can be created due to doubling the PP voltage swing and thus power limits may be exceeded.
So dual half bridges must have a dead-time for commutation in one direction and another dead-time for changing current flow directions dependent on VI product losses in the so-called Safe Operating Area SOA) limits.
- e.g. you have a choice of reversal;
- shunting to Vbat to regenerate quickly if needed
- shunting to ground to brake harder
- shunting to opposite Vbat voltage to brake twice as hard and 2x more heat.
- oscillating back and forth at some frequency within a range of RPM for some reason within 1 cycle?
- then one can ask, why? commutate the load, generate frequncy faster than the pole frequency.
- There may be nearly zero effect of change of angular momentum, in reponse unless very low mass high pole count.
Analogy
It may be possible in theory, but in practice is like expect a large voltage change from a 1Henry inductor in 1 microsecond. with a negative pulse.
Loads can have 0 step rise time, but current will be L/R=T. The question ought to define the "dead time" and/or hysteresis between full bridge commutation forward and reverse in terms of phase at max RPM.
Mechanical issues
Motors have bearings that operate at some harmonic of RPM due to diameter ratios. Stimulating higher frequency step functions on current and thus torque generate harmonics of the step or a square wave of change in torque. Thus bearing resonance can be stimulated by harmonics or RPM and high frequency now of load within 1 cycle thus making bearings vulnerable to excess wear from the vibration.
Historical examples
Massive generating stations are power factor corrected to compensation for grid loads with for power factors instead of racks of plastic caps or oil filled tank caps for passive PFC. However modern Grid-tied Inverters produce pure sine waves now with active PFC inverters so the current generated is always sinusoidal in-phase with the voltage.
-
\$\begingroup\$ If you dont like (-1) the answer, explain why \$\endgroup\$ Sep 6, 2018 at 20:43
I would like to find out how quickly and frequently a reversible electrical machine (EM) can be made to switch functions from a motor to generator and vice-versa - and what are the parameters of the EM that determine this.
simulate this circuit – Schematic created using CircuitLab
Figure 1. A simple peramanent magnet DC machine can be switched from one mode to the other as quickly as you can toggle the switch.
The inertia of the motor and load will determine the amount of energy available. The resistance of the load will determine how quickly this is used.
-
\$\begingroup\$ What is the dI/dΦ of a light bulb? (-1) \$\endgroup\$ Sep 6, 2018 at 20:44
Not sure what you mean, but a rotating motor generates EMF anyway, to drive it electrically the driver overcomes it with higher voltage. If you want to instantly take energy from the motor- no problem, just stop diving it. Takes few microseconds at most. This is well implemented in hybrid cars.