Gears are among the most demanding tribological components in mechanical engineering. The gear tooth surface must simultaneously resist contact fatigue (Hertz stress up to 1500–2000 MPa at the pitch line), bending fatigue at the root fillet, adhesive wear during sliding contact at the addendum and dedendum, and occasional shock overloads, all within the constraints of a carburised case of 0.7–2.0 mm depth over a tough ferritic core. When a gear fails, the failure mode — whether pitting, bending fracture, spalling, scuffing, or overload — contains encoded information about the root cause: insufficient case depth, grinding burn, lubricant failure, design overload, or manufacturing defect. This tutorial provides a systematic, step-by-step methodology for conducting a microstructure-based gear failure investigation from receipt of the failed component through root cause identification and corrective action specification.

Step 1: Receipt, Documentation, and Initial Macroscopic Examination

The investigation begins before any sample is sectioned or cleaned. The condition of the failed component as received encodes critical forensic information that is irreversible once lost.

Step 2: Failure Mode Classification

Accurate failure mode classification from macroscopic examination is the central diagnostic step — it determines the root cause category and dictates which metallurgical investigations are required. The six primary gear failure modes have characteristic macroscopic signatures.

Reading the Fracture Surface: Fatigue vs. Overload

The single most important diagnostic distinction is between fatigue fracture and overload fracture on a gear tooth. Fatigue fracture develops progressively over many load cycles with three characteristic zones visible on the fracture surface:

• Fatigue origin: A relatively smooth, thumbnail-shaped area at or near the surface at the highest stress concentration (typically the compressive/tensile transition at the root fillet). Multiple origins (ratchet marks) indicate high applied stress relative to material strength. A single, small origin indicates lower stress with a specific local stress concentrator (pit, nick, grinding mark).
• Fatigue propagation zone: Smooth, flat surface with concentric beach marks (arrest lines) showing the crack front positions at different stages of growth. The spacing of beach marks indicates relative crack growth rate per load block. Under magnification (SEM), striations at nanometre spacing are visible on the fracture surface, with one striation per load cycle.
• Final fracture zone: Rough, irregular surface where the remaining cross-section was insufficient to carry the applied load and failed by ductile tearing (fibrous, dull) or brittle cleavage (crystalline, reflective). The relative size of the fatigue zone to the final fracture zone inversely indicates the applied stress: large fatigue zone + small final fracture = low applied stress (material fatigue limit was marginally exceeded); small fatigue zone + large final fracture = high applied stress (single or few overload cycles).

Step 3: Material and Hardness Verification

Before any sectioning, verify that the gear material and heat treatment meet specification. This eliminates or confirms material-related root causes.

Standard material requirements for carburised transmission gears:

Typical steel grades for carburised gears:
  Automotive:   SAE 8620, SAE 4120, 20MnCrS5 (ISO 683-17)
  Industrial:   20MnCr5, 17CrNiMo6, SAE 9310 (aerospace)
  Heavy duty:   EN36C (En36C), SAE 4820

Carburised and case-hardened (per AGMA 2101 / ISO 6336-5 MQ grade):
  Surface hardness:       58–62 HRC (700–750 HV1) on tooth flank
  Core hardness:          30–42 HRC (300–415 HV) (or per drawing)
  Effective case depth:   Per gear module and application (see table below)

AGMA 2101 minimum ECD guidelines vs. normal diametral pitch:
  Module 1.5–3: ECD 0.3–0.6 mm
  Module 3–5:   ECD 0.5–1.0 mm
  Module 5–8:   ECD 0.8–1.5 mm
  Module 8–12:  ECD 1.2–2.0 mm

Verification tests on retained material:
  1. OES spectrometry — verify alloy chemistry vs. mill certificate
  2. Cross-sectional hardness traverse (HV0.5 at 0.1mm intervals)
  3. Case depth measurement to 550 HV1 (ECD) and core hardness (TCD)
  4. Metallographic examination — retained austenite estimate (etching)
  5. XRD retained austenite (ASTM E975) — quantitative measurement

Step 4: Metallographic Investigation

Metallographic sectioning provides the definitive structural evidence for failure cause. Two section planes are required: a longitudinal section through the tooth (perpendicular to flank surface through tooth centreline) and, for fatigue failures, a section through the fracture plane from the origin region.

Section Preparation Protocol

1. Preserve fracture surface: Before sectioning adjacent to a fracture, cast the fracture surface in epoxy or vacuum-impregnate to preserve crack morphology and any debris embedded in the fracture. Never grind or polish a fracture surface that needs SEM examination.
2. Mark the section location: Mark the cutting plane on the tooth surface with an indelible marker before sawing. Record the distance from a reference feature (tooth centre, pitch line) to ensure the section is taken through the intended location.
3. Cut with appropriate tool: Abrasive cut-off wheel with continuous water coolant at low feed rate to minimise cutting heat. For precision sections adjacent to the failure surface, use a low-speed diamond saw. Overheating during cutting can produce a false grinding burn signature.
4. Mount in conductive phenolic or epoxy resin: Edge retention is critical for surface hardness measurement. Use epoxy + alumina filler for maximum edge retention, or ensure phenolic resin completely supports the section edge.
5. Polish through 400, 600, 1200 grit SiC, then 6, 3, 1 µm diamond: Final polish with 0.05 µm colloidal silica or alumina. The section must be flat and scratch-free for accurate microhardness indentation.
6. Etch with 2% nital for 5–15 seconds: Reveals martensite, retained austenite (white, unetched), carbides (dark), grain boundaries, IGO, and any grinding burn zones (dark bands near surface).

Microstructural Features and Their Significance

Step 5: Contact Stress Analysis and the Pitting Mechanism

When pitting is the primary failure mode, a quantitative contact stress calculation is essential to determine whether pitting occurred within the design envelope (progressive pitting from acceptable service) or resulted from overload, material deficiency, or lubricant failure. The Hertz contact stress for a spur gear tooth-on-tooth contact approximates cylindrical line contact:

Hertz contact stress — cylindrical contact (spur gear approximation):

Contact half-width b:
  b = √(4 × W × R_e / (π × L × E'))

Peak contact pressure p₀:
  p₀ = 2 × W / (π × b × L)     or equivalently
  p₀ = √(W × E' / (π × R_e × L))

Where:
  W   = normal load (N)
  L   = face width (contact length) (m)
  R_e = equivalent (reduced) radius (m)
       R_e = (R₁ × R₂) / (R₁ + R₂)
       R₁, R₂ = radii of curvature at pitch point of pinion and gear
       = r₁ × sinφ, r₂ × sinφ   (r = pitch circle radius, φ = pressure angle)
  E'  = reduced elastic modulus = 2 / [(1−ν₁²)/E₁ + (1−ν₂²)/E₂]

Subsurface shear stress maximum (initiates pitting):
  τ_max ≈ 0.304 × p₀   at depth z ≈ 0.48 × b below surface
  (This is the von Mises max shear stress for line contact)

Contact fatigue limit for carburised steel (AGMA 2101-D04):
  σ_c,allow = σ_c,lim × Z_N × Z_W / (S_H × K_T × K_R)
  σ_c,lim for carburised gear (Grade 3 material): ~1700 MPa (250 ksi)
  → If calculated p₀ > σ_c,allow: pitting is expected (overload or insufficient case)
  → If p₀ < σ_c,allow: look for material defects, lubricant failure, or subsurface inclusions

Subsurface Crack Initiation in Contact Fatigue

The classical contact fatigue mechanism initiates at the depth of maximum orthogonal shear stress, typically 0.1–0.3 mm below the pitch-line surface. The physical mechanism involves:

1. Cyclic loading produces shear stress reversals in the subsurface material beneath the Hertz contact ellipse. For spur gears, the maximum orthogonal shear stress (reversing) is approximately 0.25 × p₀ at depth 0.5b; the maximum von Mises shear stress is approximately 0.30 × p₀ at depth 0.48b.
2. At hard inclusions (MnS, Al₂O₃ stringers) or in regions of retained austenite, stress concentration causes early crack nucleation at or just below the depth of maximum shear stress.
3. The crack propagates at an angle of approximately 20–40° to the surface under mixed-mode (Mode I + Mode II) loading, driven by the combined effect of contact stress and EHD lubricant pressure pumping into the crack on each load cycle.
4. When the crack intersects the surface or connects to an adjacent crack, a pit spalls out. The characteristic pit shape (deeper at the leading edge, shallower at the trailing edge relative to the sliding direction) reveals the crack growth direction and lubricant pumping mechanism.

Step 6: Grinding Burn Detection and Residual Stress

Grinding burn is one of the most common and damaging root causes of premature gear failure, and one of the most frequently missed during incoming inspection because it is invisible to conventional hardness testing unless the hardness drop is severe. It requires targeted metallographic investigation at every gear failure analysis.

Etching Protocol for Grinding Burn Detection

The standard field method uses modified Barkhausen noise measurement, but the laboratory confirmation uses a controlled nital etch sequence:

1. Polish the tooth flank section to 1 µm diamond finish.
2. Etch with 4% nital for 30 seconds, then wash and dry immediately. In a grinding burn-affected zone, dark (overtempering) or white (rehardening) regions appear against the background of normally etched tempered martensite.
3. Re-polish to remove the etch (remove ~2–5 µm), and measure Vickers microhardness HV0.1 at 0.02 mm intervals in the surface zone. A hardness drop of more than 50 HV compared to adjacent undamaged material is diagnostic of temper burn. A hardness increase above 800 HV combined with a white etching layer indicates rehardening burn.
4. Apply acid ammonium persulfate (10% w/v, aqueous) for deeper-etching contrast of the white-etching rehardened layer if present. Alternatively, alkaline potassium permanganate solution (3 g KMnO₄ + 25 g KOH / 100 mL water) preferentially darkens the rehardened layer.

Step 7: Root Cause Analysis and Fishbone Diagram

Root cause analysis (RCA) for gear failures uses the evidence assembled in Steps 1–6 to trace from the observed failure mode back to the fundamental cause through a structured logic tree. The five root cause categories for gear failures follow the 5M structure (Material, Manufacturing, Maintenance, Misuse, Design):

Step 8: Report Structure and Corrective Action Specification

The failure analysis report must present findings in a structured format that enables a non-specialist reader to understand the conclusion while providing sufficient technical depth for a metallurgist or gear engineer to verify the analysis. The ASTM E2677 standard format for failure analysis reports provides the recommended structure:

1. Executive Summary: One page. Failure mode, root cause, and top three corrective actions. Written for management and decision-makers.
2. Background and Objectives: Component description, service history, operating conditions, failure description, and scope of investigation.
3. Investigation Methods: Techniques used, standards applied, equipment specifications. Enables third-party verification.
4. Results: Factual presentation of all findings — macroscopic examination, hardness data, metallographic observations, chemical analysis, contact stress calculations. Tables and figures referenced in text. No conclusions in this section.
5. Discussion: Interpretation of findings. How each piece of evidence supports or eliminates each potential root cause. Quantitative comparison of calculated contact stress vs. allowable, measured ECD vs. specified ECD, retained austenite vs. limits, etc.
6. Conclusions: Definitive statement of failure mode and root cause. Primary and contributing causes clearly differentiated. Conservative language if evidence is incomplete.
7. Recommendations: Specific, actionable corrective actions assigned to responsible functions (manufacturing, design, maintenance) with priority and target completion date.

For the metallurgical underpinnings of this tutorial, see the Martensite Formation in Steel article for the microstructural basis of case hardness, Quenching and Tempering for tempering response and retained austenite control, and Hardness Testing Methods for the full methodology of microhardness traverses and conversion to HRC. The Charpy Impact Test and Mechanical Testing articles cover companion characterisation methods used in gear material qualification.