Reluctance Motor (SRM & SynRM)

Reluctance — the magnetic analogue of resistance

Magnetic flux flowing through a circuit obeys the same form of Ohm's law that current does in an electrical one. Magnetomotive force (NI, in ampere-turns) drives flux (Φ, in webers) through a magnetic reluctance (ℜ, the magnetic analogue of resistance):

NI = ℜ · Φ          (magnetic Ohm's law)
ℜ = ℓ / (μ · A)     (reluctance scales with path length, inversely with permeability and area)

Air has a reluctance roughly 1000× higher than silicon steel. If you make a rotor that presents alternating iron and air paths around its circumference and then put it inside a stator that creates a rotating magnetic field, the flux paths will be much shorter (lower reluctance) when the iron parts of the rotor align with the field. The rotor twists itself into that alignment because the system minimises its stored magnetic energy.

The torque on the rotor is the rate at which stored magnetic energy changes with rotor angle. For a constant current i flowing through windings whose self-inductance L(θ) depends on rotor position, the co-energy is W' = ½ · L(θ) · i² and the torque is dW'/dθ = ½ · (dL/dθ) · i². No magnets needed — just a non-circular inductance profile.

Switched Reluctance Motor (SRM) — salient on both sides

The crudest and most direct reluctance machine has salient (sticking-out) poles on both the stator and the rotor. A 6/4 SRM has six stator teeth and four rotor teeth; an 8/6 SRM has eight stator and six rotor. The stator carries simple concentrated windings, one phase per pole pair. The rotor is purely shaped steel — no magnets, no copper, no slip rings.

6/4 SRM, 3-phase
   stator: phase A poles at 0° and 180°
           phase B poles at 60° and 240°
           phase C poles at 120° and 300°
   rotor:  4 teeth at 90° spacing

  energise A → rotor snaps to align with A axis
  energise B → rotor snaps to align with B axis (advance 30°)
  energise C → rotor advances another 30°
  → 90° rotor rotation per 3 phase commutations

The inverter switches phases sequentially in synchronism with rotor angle (read by a Hall sensor, an optical encoder, or sensorless back-EMF estimation). Each phase pulls the nearest rotor tooth into alignment, then turns off as the next phase takes over. The mechanism is essentially that of a stepper motor scaled up to industrial power, and the family resemblance is strong: low-speed cogging, high torque ripple, audible whine, simple drive electronics.

What SRM gains is brutal simplicity. No magnets, no rotor windings, fully fault-tolerant (a failed phase just reduces torque), and the rotor can spin to extraordinarily high speeds without flying apart — the only thing on the rotor is solid steel. The Dyson V11 cordless vacuum motor uses a 4/2 SRM at 125,000 rpm; a permanent-magnet machine could not survive those centrifugal stresses.

Synchronous Reluctance Motor (SynRM) — modern industrial workhorse

SynRM is the sophisticated cousin of SRM. The stator is the same as an induction motor's — smooth bore with distributed three-phase windings. The rotor is laminated steel sculpted with curved internal flux barriers (effectively air slots) that create a strong magnetic anisotropy: the rotor's d-axis (along the flux barriers' aligned direction) has a much higher inductance than its q-axis (perpendicular). A typical industrial SynRM rotor has Ld/Lq ratios of 6 to 12.

Three-phase sinusoidal currents in the stator create a rotating magnetic field. The rotor's anisotropy means it experiences a torque that pulls its d-axis into alignment with the rotating field. Once locked in, the rotor turns synchronously with the field at the same speed — no slip, unlike induction. The drive runs a field-oriented control loop in the d-q frame, just like for a PMSM, but with the torque equation simplified because there is no magnet flux term:

T_em = (3/2) · p · (L_d - L_q) · i_d · i_q       (SynRM reluctance torque only)
       (PMSM also has λ_PM · i_q term from the permanent magnet)

ABB launched its first SynRM industrial drive in 2011 (the M3BL family). By 2015 the technology had pushed past induction on efficiency at every common power rating from 7.5 kW to 350 kW. By 2024 ABB's IE5 SynRM range is rated 96 percent efficiency at full load versus 92 percent for IE3 induction at the same frame size — four percentage points that translate into roughly 30 percent less loss energy, or about 1 MWh per year for a continuously running 50 kW industrial pump.

Worked example — replacing a 75 kW induction motor with SynRM

Consider a continuously running 75 kW industrial pump driven by an IE3 induction motor at 92 percent efficiency. The annual loss energy is roughly:

P_loss = P_out × (1/η - 1)
       = 75 kW × (1/0.92 - 1)
       = 75 kW × 0.087
       = 6.5 kW

E_loss_annual = 6.5 kW × 8760 h/year = 57,000 kWh/year

Swap in an IE5 SynRM at 96 percent. Annual loss energy drops to:

P_loss = 75 × (1/0.96 - 1) = 75 × 0.0417 = 3.1 kW
E_loss_annual = 27,400 kWh/year
ΔE = 57,000 - 27,400 = 29,600 kWh/year saved

At industrial electricity prices of €0.15/kWh, that's about €4,440 per year per motor in energy savings. The price premium of a SynRM over an IE3 induction motor is typically €1500-3000 at this power level — payback in 6 to 12 months for continuously running loads. This is the math behind the EU's MEPS-2023 efficiency directive that effectively obsoletes IE2 and below for new industrial installations.

Reluctance vs induction vs PMSM — when each wins

Property	Induction (IM)	SynRM (IE5)	SRM	PMSM
Peak efficiency	92% (IE3)	96% (IE5)	88-92%	96-98%
Torque density (N·m/kg)	1.0×	0.85×	0.75×	1.4×
Power factor at rated load	0.85	0.70-0.78	poor	0.95+
Rotor losses	1-3% (slip)	~0 (no current)	~0	~0
Rare-earth content	None	None	None	0.2-0.5 kg Nd
Torque ripple	~3%	~5%	~15%	~1%
Acoustic noise	Low	Low	High	Low
Drive complexity	V/f or vector	FOC required	Phase-switched	FOC
Typical use	Pumps, fans	IE5 industrial drives	Vacuums, washers	EVs, robots, servos

The rare-earth-free motivation

High-performance permanent-magnet motors depend on neodymium-iron-boron (NdFeB) magnets. The standard grade requires a few percent dysprosium added to raise the Curie temperature and stop demagnetisation under hot operation. Dysprosium is a "heavy" rare earth and its supply is concentrated in southern China — at one point in 2011 the spot price spiked 40× in a year. Every Western automaker, every industrial drives company, and every appliance maker put real money into rare-earth-free alternatives after that shock.

SynRM and SRM are the two leading candidates because they need no magnets at all. The induction motor is the legacy alternative but is fundamentally less efficient. Tesla's response was a hybrid: the Model 3 rear motor is an internal-permanent-magnet synchronous reluctance machine (IPM-SynRM) that gets most of its torque from reluctance and the rest from cheap ferrite magnets, cutting dysprosium content to near zero. BMW's iX5 Hydrogen used a similar topology. ABB and Siemens dominate the rare-earth-free industrial drive market with SynRM.

SRM drives — the asymmetric half-bridge

An SRM drive looks unlike anything else in power electronics. Because each SRM phase only ever needs unidirectional current (you can pull the rotor pole, not push it), the inverter uses an asymmetric half-bridge with two switches and two diodes per phase, instead of the full H-bridge that SynRM and PMSM require. The drive is mechanically simpler and electrically more fault-tolerant — a failed switch isolates only one phase — but the trade is high torque ripple and acoustic noise.

SRM asymmetric half-bridge (per phase)
   +V_bus
     │
    [Q1]
     │
     ●─── phase A coil ───●
                          │
                         [Q2]
                          │
   ─V_bus

   Q1+Q2 on:  current rises into the coil
   Q1+Q2 off: current freewheels through diodes back to bus
   only Q1 on (or only Q2 on): current circulates in zero-volt loop

Phase commutation is synchronised with rotor angle: the controller turns on a phase a few degrees before its rotor pole reaches alignment ("turn-on angle") and turns off a few degrees before alignment to avoid generating negative torque ("turn-off angle"). Those two angles are the main control variables — tuning them for low ripple and high efficiency is the hard part of SRM commissioning.

Where each reluctance machine ends up

• ABB IE5 SynRM industrial drives. Pumps, fans, compressors, conveyors at 7.5 kW to 350 kW. Direct replacement for IE3 induction in continuously running OEM applications. 96 percent efficiency saves 30 percent of waste-heat energy compared with induction.
• Dyson V11 cordless vacuum. 4/2 SRM at 125,000 rpm, ~600 W mechanical from a 26 V Li-ion pack. The high speed lets the motor be physically tiny (about the size of a roll of quarters); a permanent-magnet rotor would fly apart at that speed.
• Tesla Model 3/Y rear-drive motor. IPM-SynRM hybrid producing 211 kW peak at 96 percent efficiency. Cuts dysprosium to near zero versus the Model S 90D's PMSM. The hybrid topology gets reluctance torque from rotor anisotropy and additional torque from ferrite IPM magnets.
• BMW iX5 Hydrogen drivetrain. Externally excited synchronous (EESM) and SynRM technologies share BMW's commitment to rare-earth-free traction. The iX5 prototype's rear-drive eAxle is an EESM that takes the rotor-field-free principle one step further (a wound rotor with no magnets, driven by brush-fed DC).
• White goods — washing machines and HVAC compressors. Bosch, Whirlpool, Mitsubishi-Electric and others have shipped SRM and SynRM washer drum motors. Trade-off: efficiency-cost vs acoustic noise; SRM-with-noise-shaping is a battle Korean and Chinese OEMs still fight.
• Mining truck traction. Komatsu and Caterpillar use SRM in some autonomous mining haul trucks at 1 MW+. The magnet-free rotor survives heat and vibration better than any PM alternative, and acoustic noise does not matter in a mining pit.

Common pitfalls and failure modes

• SRM acoustic noise. The radial magnetic pull on each rotor pole oscillates at the commutation frequency, exciting the stator iron ring like a tuning fork. Mitigation: lamination shape optimisation, current-shaping algorithms that smooth the radial force pulse, and external stator damping rubber. Cost: gives back some of the savings from eliminating magnets.
• SynRM low power factor. A SynRM at rated load typically runs 0.70-0.78 power factor versus 0.95+ for a PMSM. The drive inverter must therefore be sized for higher apparent power (VA) than would be needed for an equivalent-real-power PMSM. Mitigation: ferrite-assisted SynRM (small magnets in the flux barriers raise PF to 0.85-0.9) — but at the cost of some rare-earth-free purity.
• Sensorless commutation hard to start. Both SRM and SynRM rely on rotor angle for commutation. Sensorless schemes infer angle from back-EMF — which is zero at rest. Cold-start algorithms inject a brief test pulse to detect rotor position by inductance asymmetry, then ramp open-loop until back-EMF is detectable. Failure to start cleanly under high static friction is a common warranty issue.
• Saturation modelling for high-torque operation. At rated current the rotor iron starts to saturate magnetically, flattening the L_d versus i_d curve. FOC controllers tuned for the linear region will produce reduced torque at high load; modern controllers use saturation maps (lookup tables of L_d, L_q vs current) to maintain accurate torque control.
• Vibration coupling to bearings. Radial force pulses from SRM commutation couple into the bearings as side-load, shortening bearing life. Especially noticeable at part-load operation where the pull is most asymmetric. Mitigation: bearings with higher radial load rating, balanced double-row designs, and active force-shaping in the controller.

A short history — from Davidson's locomotive to ABB IE5

The reluctance principle is the oldest in electrical machines. Robert Davidson built a reluctance-motor-driven locomotive in Scotland in 1838 — possibly the first practical electric vehicle. The technology then lay nearly dormant for 130 years because induction motors and brushed DC motors were simpler and good enough for most applications, and high-quality semiconductor switching to drive a reluctance motor smoothly did not exist.

SRM revival began in 1972 when Peter Lawrenson and his group at Leeds University demonstrated a controlled SRM drive that could match induction motor performance on certain industrial loads. Industrial adoption was slow through the 1980s and 1990s — Switched Reluctance Drives Ltd. (later acquired by Emerson) shipped SRM-driven HVAC compressors and Dyson's first cordless vacuum SRM appeared in 2009. SynRM exploded in the 2010s when ABB perfected high-anisotropy laminated rotors and shipped industrial drives that beat induction efficiency at the same price point. By 2024, SynRM is the dominant new IE5 industrial drive technology in Europe, with Siemens, ABB, WEG, and Yaskawa all shipping ranges.