Your Gear Service Life Is Consumed by Only 5% of Load Cycles

Reader Question:
“The time‑domain contact force analysis from our previous article works well. However, service life is my main concern. Once the force values are calculated, how many operating hours can this reducer last? Moreover, should I reduce torque or cut down impact loads if modifications are needed?”

This hits a core point. Predicting total service life and identifying improvement measures are two separate matters, and the latter guides your engineering actions. We have developed a second tool to address both questions.

A Counter‑intuitive Observation
Real‑world gearboxes almost never operate continuously at rated torque. Take a reducer with the following load‑spectrum distribution: 50 % operating time at rated torque, 30 % at 1.4× rated torque, 15 % at 1.8× rated torque, and merely 5 % operating time under heavy‑duty 2.6× rated torque.

Intuition suggests most damage would originate from the rated‑torque condition that occupies half of total runtime. Yet actual calculation yields the opposite result. In the animation, the left side shows contact stress under sequential load‑spectrum operation, while the right side presents cumulative damage. The red segment corresponds to that 5 % heavy‑load condition.

▲ Figure 1: With each cursor step forward, rain‑flow counting and damage accumulation are recalculated for the completed time‑history segment.

Calculation outcome: The heavy‑load condition, accounting for only 5 % of total cycles, generates 32 % of cumulative damage. By contrast, the rated‑torque condition covering half the runtime contributes less than 14 % of total damage. The overall reducer service life reaches approximately 20 000 hours.

This breakdown delivers far higher practical value than a single “20 000‑hour service‑life” figure. It clearly indicates: to extend service life, mitigate those occasional heavy‑load events instead of optimizing the frequently‑running rated‑torque operating point. This is why damage‑contribution decomposition per load‑amplitude bin is set as the primary output of this tool. Most free‑of‑charge fatigue calculators only output a single‑value lifespan, offering no guidance for subsequent engineering adjustments.

Calculation Workflow
Four textbook‑standard steps with zero black‑box logic:

1. Contact‑force time‑history:
Based on the cycloidal‑pin gear model introduced earlier, contact force on each pin tooth is computed sequentially according to the load spectrum.
2. Hertzian contact stress:
Convert contact force into stress. Key input parameters including contact width, radius of curvature and equivalent elastic modulus are fully editable. These parameters directly govern stress magnitude; concealing them would hide critical calculation assumptions.
3. Rain‑flow counting:
Decompose irregular time‑history signals into complete load cycles following ASTM E1049 four‑point algorithm.
4. Miner’s linear damage accumulation:
Obtain allowable cycle counts from S‑N curve for each individual cycle, then accumulate damage incrementally.

✅ Built‑in self‑check runs for every calculation.
Rain‑flow counting has a strict validity criterion: the sum of all counted‑cycle amplitudes must precisely match the total variation of the original time‑history signal (amplitude conservation). Algorithm defects will almost certainly break this conservation rule. This validation result is directly displayed on the user interface. Validation also triggers upon tool startup; the tool refuses to run if checks fail. You do not need to trust our implementation — simply check the validation indicator.

⚠️ Honest Remarks on Material Data
Rain‑flow counting represents the straightforward portion of fatigue assessment. The credibility of final results hinges heavily on material S‑N curves and stress conversion methodologies.

Therefore, this tool contains no arbitrary “plausible” default material datasets. Only traceable reference material entries are included, with source documentation explicitly displayed within the interface. Uncertain parameters are clearly marked for user verification. For instance, the contact‑fatigue strength σHlim ≈ 1500 MPa for carburized‑quenched steel is cross‑referenced from ISO/TS 6336‑5 and GB/T 3480 series standards, and both references are documented. S‑N curve exponent values are noted as common industry conventions rather than newly‑validated test data.

Why place such emphasis on data traceability? Fatigue life shows extreme sensitivity to these parameters. Under an exponent of 6.6, a 10 % deviation in σlim yields approximately 1.9‑fold difference in service life. Selecting reference cycle counts of 3×10⁷ versus 5×10⁷ brings a 1.67‑fold lifespan variance. A tool that outputs instant‑life figures without clarifying data sources delivers precise‑yet‑wrong answers, which is more hazardous than providing no result at all.