The MRT power system supplies electricity. The propulsion system turns that electricity into controlled wheel torque.
Between the conductor rail or overhead supply and the steel wheel sits a chain of power electronics, motors, mechanical transmission and control that must produce exactly enough tractive force for the requested train motion.
A train does not accelerate because electricity “pushes” it directly.
The chain is:
electrical supply → onboard conversion and control → traction motor electromagnetic torque → gearbox or mechanical transmission where used → wheel torque → wheel–rail tangential force → train acceleration
During regenerative braking, much of the chain reverses.
train kinetic energy → wheel torque → motor generating torque → electrical conversion → usable returned electrical energy where the network can accept it
Singapore’s public railway material confirms this broad architecture. LTA describes Downtown Line trains as using energy-efficient regenerative brakes, and its Bukit Panjang LRT renewal material explicitly identifies new propulsion equipment with regenerative-braking capability. Earlier LTA rolling-stock safety testing also publicly described train propulsion as a rated system that must be able to move a failed train under specified recovery testing.
This article owns the onboard conversion from electrical power to motor torque to wheel force. The infrastructure supply network belongs to How MRT Power Supply Works Using Mathematics; whole-line energy scheduling belongs to Energy Optimisation; contact mechanics belong to Wheel–Rail Contact.
The RFE — What Is the Propulsion System For?
The weak answer is:
make the train move
The actual job is more disciplined.
The RFE of propulsion is to convert available electrical energy into precisely controllable wheel torque across the train’s speed range, while respecting adhesion, thermal, comfort, energy and braking constraints.
Prompt 1 — How Does Motor Torque Become Train Force?
Let motor torque be τm.
With gearbox ratio G and mechanical efficiency ηg:
τwheel ≈ G ηg τm
For wheel rolling radius r:
Ftractive = τwheel/r
If several powered axles contribute:
Ftrain = Σ Faxle
Then longitudinal acceleration is determined by:
meff a = Ftrain − Fresistance − Fgrade − Fbrake
The motor is therefore one component inside a force balance.
Prompt 2 — Why Does Torque Change With Speed?
Traction drives have finite electrical voltage, current and thermal capacity.
A simplified traction-motor envelope has two familiar regions.
Low speed: approximately constant torque.
τ ≈ τmax
Since mechanical power is:
P = τω
power rises with rotational speed ω.
Higher speed: approximately constant power.
τ ≈ Pmax/ω
Torque falls as speed rises.
This explains why trains can produce strong acceleration from low speed but acceleration naturally falls as speed increases, even before resistance rises.
Prompt 3 — What Do Inverters and Power Electronics Do?
The electrical supply available to the train is not automatically the exact voltage, current and frequency waveform required by the traction motor.
Power electronics synthesize the motor input.
At a public conceptual level:
DC traction supply → switching converter / inverter → controlled multi-phase motor currents → electromagnetic field → rotor torque
For a simplified three-phase motor, electrical input power is approximately:
Pelec ≈ √3 V I cosφ
with efficiency ηm:
Pmech = ηm Pelec
Modern drives use closed-loop control of motor torque and speed. The detailed control implementation is train-specific and is not reproduced here.
Prompt 4 — Why Can the Motor Produce More Torque Than the Rail Can Use?
The wheel–rail interface limits useful tractive force.
|Ftractive| ≤ μeff N
If commanded wheel torque tries to exceed available adhesion, wheel slip or slide tendency rises.
The propulsion controller therefore needs feedback.
requested acceleration → requested torque → measure wheel/train response → detect excessive slip tendency → reduce or redistribute torque → measure again
Motor capability is therefore not the same as usable train force.
The rail, not the motor datasheet, ultimately decides how much torque can become acceleration.
Prompt 5 — How Does Regenerative Braking Reverse the Energy Flow?
A moving train stores kinetic energy:
Ek = ½mv²
During regenerative braking, motor torque opposes wheel rotation.
The motor acts as a generator.
Pregen = ηregen Fbrake v
where ηregen captures conversion losses.
Usable returned energy is:
Eusable = ∫ min(Pregen, Preceptive) dt
The train can generate power that the electrical network cannot absorb at that instant. Whole-line receptivity therefore belongs to the power and energy pillars.
Prompt 6 — Why Is Heat a Propulsion Constraint?
Motors and inverters are not perfectly efficient.
Losses become heat.
Ploss = Pelec − Pmech
A simple thermal state is:
Cth dT/dt = Ploss − h(T−Tamb)
where Cth is thermal capacity and h represents effective cooling.
Repeated hard acceleration can raise temperature faster than it can be rejected.
Thermal limits can therefore reduce available continuous torque even when instantaneous electrical hardware could produce more.
Prompt 7 — How Does Propulsion Affect Timetable and Energy?
Train running time depends on the acceleration profile.
For constant acceleration a from rest to speed v:
t = v/a
Distance used is:
s = v²/(2a)
Higher acceleration can reduce run time but increases peak traction power, passenger jerk demands and thermal load.
The propulsion objective is therefore not maximum acceleration.
meet timetable + preserve comfort + stay inside adhesion + stay inside electrical and thermal limits + minimise unnecessary energy
Green CBTC work on the NSEWL illustrates this connection at system level: optimised automated running curves can reduce traction energy while preserving punctuality. Propulsion executes the torque profile that such higher-level control requests.
Prompt 8 — How Does the Railway Know the Propulsion System Is Healthy?
A generic propulsion-health vector may include:
x(t)=[motor current, voltage, torque estimate, inverter temperature, motor temperature, vibration, wheel response, fault events]
Expected efficiency under operating point q is η̂(q).
Measured efficiency is η(q).
eη = η−η̂
A persistent deterioration may justify inspection.
But one high-current event can simply reflect a heavy acceleration demand rather than a fault.
Condition must be interpreted relative to operating context.
A Fictional Propulsion Example
Suppose one motor produces 1,200 N·m torque.
Gear ratio G=6 and mechanical efficiency ηg=0.96.
τwheel = 1200×6×0.96
= 6912 N·m
Wheel radius r=0.42 m.
F = 6912/0.42 ≈ 16.46 kN
With eight equivalent powered motor paths:
Ftotal ≈ 131.7 kN
Suppose effective train mass is 160 tonnes and total resistance at that moment is 25 kN.
a = (131.7−25)/160000 ≈ 0.667 m/s²
But if adhesion allows only 100 kN total tractive force, the motor request must be reduced.
a = (100−25)/160000 ≈ 0.469 m/s²
The motor was capable of more torque. The wheel–rail interface was not.
Deletion Tests
- Remove inverter control: available electrical supply cannot become precise motor torque.
- Remove torque-speed envelope: the motor produces maximum torque at every speed without limit.
- Remove wheel radius: torque and tractive force become identical quantities.
- Remove adhesion: unlimited motor torque always becomes acceleration.
- Remove thermal state: repeated hard acceleration produces no heat consequence.
- Remove regeneration: braking energy can never return electrically.
- Remove timetable: propulsion can optimise efficiency without a transport-time job.
- Remove World Return: motor condition never learns from actual current, temperature or efficiency.
Propulsion Paradoxes
- The most powerful motor does not always create the greatest usable tractive force.
- A motor can become a generator during braking.
- Higher acceleration can shorten run time while reducing whole-line energy efficiency.
- Motor hardware can be healthy while the train accelerates poorly because adhesion is low.
- A thermal limit can reduce torque even when voltage and current supply are available.
The Propulsion Audit
- What train acceleration or braking force is requested?
- What motor torque follows?
- What gearbox or mechanical ratio converts it to wheel torque?
- What wheel radius converts torque to force?
- How many powered axles contribute?
- What adhesion limit is available?
- What motor torque-speed region is active?
- What voltage, current and power constraints apply?
- What heat is being generated?
- What regenerative power can the network receive?
- What passenger comfort and timetable requirements bound the torque trajectory?
- What condition data show the propulsion system still behaves as expected?
World Return — Torque Meets the Real Train
requested acceleration → torque command → motor current → wheel torque → actual train acceleration → compare → adjust control / inspect condition
If commanded torque is correct but acceleration is consistently lower than predicted, the model investigates resistance, mass, adhesion and motor/mechanical efficiency.
If electrical power rises for the same delivered wheel force, efficiency may be degrading.
MRT propulsion works when electrical energy becomes exactly the wheel torque the real train can use—no more, no less, and reversible when braking can return energy.
Key Equations
τwheel≈Gηgτmotor Wheel torque F=τwheel/r Tractive force meff a=Ftraction−Fresistance−Fgrade−Fbrake Train force balance P=τω Mechanical motor power τ≈Pmax/ω Constant-power torque region Pelec≈√3VIcosφ Three-phase input power scale Pmech=ηmPelec Motor conversion efficiency |Ftractive|≤μeffN Adhesion limit Ek=½mv² Kinetic energy Pregen=ηregenFbrakev Regenerative power scale Cth dT/dt=Ploss−h(T−Tamb) Thermal state eη=ηobserved−ηexpected World Return efficiency residual
Reader-safety note: This article does not reproduce train-specific inverter switching logic, propulsion software, protection thresholds, traction-control tuning, electrical schematics or restricted recovery procedures. All numerical examples are fictional unless attributed to public sources.