At each start, the control system closes the direct-on-line starter, the 11 kV supply energizes the stator, the air-gap field induces rotor current, and the cage converts that current into shaft torque. Follow that path to find where the 4.35 MVA motor accumulated damage across approximately 4,000 starts in 20 years—an average of 200 starts per year. The resolving branch is the one that reduces rotor thermal cycling without extending acceleration or losing the torque required by the load.
Where does starting stress enter the rotor?
At standstill, slip is at its maximum and the rotor cage carries high induced current. The bars heat faster than the surrounding rotor structure, while electromagnetic forces and transmitted torque load the cage mechanically. Repeated differential expansion and contraction produces fatigue at bars, joints, and end-ring connections. Abrupt load-torque changes add cyclic mechanical stress.
A soft starter can reduce current and magnetic force by lowering terminal voltage. At fixed frequency, however, induction-motor torque varies approximately with the square of voltage. Reducing voltage can therefore lengthen acceleration. A longer high-slip interval may increase rotor heating even when the measured line-current peak is lower. Judge every starting method by acceleration time, rotor thermal duty, and torque margin—not peak current alone.
Is the incoming electrical path adding stress?
Layer one first. Capture all three line voltages and currents at the motor supply during a complete start. Use instrumentation rated for the 11 kV system and follow the site switching and protection procedures. Compare the phases and correlate every discontinuity with breaker, starter, or process events.
| Reading | Outcome | Next check |
|---|---|---|
| Phase voltage balance during acceleration | Imbalance drives unequal stator and rotor heating | Trace the imbalance through the supply, switching equipment, and connections |
| Three phase-current traces | Unequal or distorted traces identify supply, winding, or mechanical asymmetry | Compare with voltage traces, then perform offline winding and rotor tests |
| Voltage notches or transitions | Abrupt flux and torque changes can load the cage mechanically | Identify the switching event and review the starter sequence |
| Acceleration time | An extended high-slip interval raises rotor energy dissipation | Measure speed and load torque through the same interval |
If the supply is balanced and switching is clean, continue along the path to the start profile. If it is not, correct the upstream condition before selecting a different starter; otherwise the new equipment can mask rather than remove the initiating stress.
Does the start profile overheat the cage?
Record current, terminal voltage, speed, and time from energization to settled running speed. Obtain the motor manufacturer's permitted hot and cold start duty, minimum time between starts, acceleration limits, and rotor thermal model. The historical average of 200 starts per year does not describe clustering; several closely spaced starts can be more damaging than the same count distributed across months.
| Observed profile | Meaning | Decision |
|---|---|---|
| Short, repeatable acceleration with adequate torque margin | Starting duration is controlled | Examine start frequency and torque shocks |
| Long acceleration near constant current | The rotor remains at high slip for too long | Check load torque, inertia, voltage depression, and starter settings |
| Restarts before thermal recovery | Bars begin the next cycle hot | Add process or control interlocks based on the approved start duty |
| Lower current but longer soft start | Peak reduction has traded against heating time | Recalculate acceleration and rotor thermal duty before accepting the setting |
Use the measured waveform to compare start severity. A simple stator I²t calculation can help compare repeatable events, but it is not a substitute for a rotor thermal model because rotor resistance, current distribution, slip, and cooling change during acceleration.
Does the driven train impose torque shocks?
Measure shaft speed, vibration, and process load through starting, loading, unloading, and stopping. A sudden load pickup, coupling engagement, jam, or process transition can produce torsional stress even when electrical traces look normal. Correlate the exact time of each speed or vibration disturbance with current and process signals.
If the disturbance originates at the load, modify the operating sequence or coupling arrangement. An electromagnetic shaft coupling can soften torque transfer. A flywheel can smooth a short torque disturbance, but its added inertia also increases acceleration energy and may extend starting time. Model the full speed-torque curve before adding inertia.
Can cracked bars be detected without opening the motor?
Use more than one indicator because load oscillation, supply imbalance, air-gap eccentricity, and winding asymmetry can resemble rotor-cage faults.
| Method | Operating state | What to evaluate |
|---|---|---|
| Motor current signature analysis | Online at stable load | Rotor-related sidebands, phase consistency, and change from a baseline |
| Rotor influence check (RIC) | De-energized; rotor moved through positions | Changes in winding response associated with rotor bars, air gap, or windings |
| Vibration and speed analysis | Online | Components that track slip, torque pulsation, or mechanical looseness |
| Start-current trending | During acceleration | Repeatable modulation or phase asymmetry as the rotor passes through the speed range |
For current signature analysis, broken-bar asymmetry can create components around the supply frequency at , where f1 is supply frequency and s is slip. Record speed or slip and load with the current spectrum; a spectrum without operating condition is weak evidence. Trend amplitude under comparable load rather than accepting one snapshot.
The RIC cited for this application class is available from PdMA and can be performed without dismantling the motor. It requires an outage and the ability to reposition the rotor. Motor service companies may provide the test as a contracted service. Use offline RIC findings with online current and vibration trends before committing to another rebuild.
Which mitigation branch should be implemented?
| Option | Primary benefit | Recurring pitfall |
|---|---|---|
| Reduce start count | Directly removes thermal cycles | Process changes may create long idle running or new operating constraints |
| Medium-voltage soft starter | Limits voltage and current ramp | Reduced torque can prolong high-slip heating |
| Variable-frequency drive | Controls frequency, voltage, torque, and acceleration | Requires evaluation of motor insulation, cooling, bearings, harmonics, and system protection |
| Reduced-voltage reactor or autotransformer start | Provides staged current reduction | Fixed voltage reduction lowers available torque and transition events may introduce shocks |
| Coupling or process-sequence change | Reduces abrupt shaft torque | Does not correct rotor heating caused by excessive starts or long acceleration |
| Double-cage replacement motor | Redistributes starting current and losses through the cage design | Suitability depends on the actual load curve and required start duty |
With a rebuild costing almost 70% of a new motor, compare the complete lifecycle alternatives: repaired motor plus starting equipment, a drive retrofit, and a new motor designed for the measured duty.
- Capture synchronized voltage, current, speed, vibration, and process-load data for representative starts.
- Document starts per shift, clustered restarts, initial motor temperature, acceleration time, and abnormal load transitions.
- Perform online current signature analysis and an offline RIC or equivalent rotor assessment to establish the present cage condition.
- Give the measured supply and load speed-torque data to the motor and starter suppliers. Request acceleration, rotor thermal-duty, and permitted-start calculations for each candidate.
- Select the branch that keeps accelerating torque above load torque while reducing start count, rotor heating, or torque shock.
- Commission the selected system and repeat the original measurements under comparable load. Accept the change only when acceleration remains within the approved duty, phase traces are balanced, no new torque discontinuity appears, and the rotor-condition baseline is recorded.
FAQ
Can I prevent rotor bar cracks by installing a soft starter?
A soft starter can reduce current and magnetic force, but reduced voltage also lowers available torque approximately with voltage squared. Accept it only after measured acceleration and rotor thermal-duty calculations show that the longer ramp will not increase cage heating.
Can I detect cracked rotor bars without opening the motor?
Yes. Trend online motor-current signatures and vibration at known load and slip, then use a de-energized rotor influence check during an outage. Agreement between methods separates cage asymmetry from supply, winding, air-gap, and load effects.
Does lower starting current prove the rotor is better protected?
No. Repeat the synchronized current, voltage, speed, vibration, and load capture after commissioning; the final verification is acceptable acceleration with balanced phase traces, no new torque discontinuity, and a recorded rotor-condition baseline.