Why a bearing handbook is required reading

Around 10 billion bearings are made every year, and they carry essentially everything that rotates: pumps, motors, fans, gearboxes, conveyors, mills. When a rotating machine fails, the post-mortem very often ends at a bearing – which makes bearing failure analysis one of the highest-leverage skills a maintenance team can have.

SKF's Bearing damage and failure analysis is the reference most plants actually use for it. It is free, generously illustrated, and organized around a powerful idea: every failure mode leaves its own fingerprint on the bearing. Learn the fingerprints and a failed bearing stops being scrap – it becomes a written confession about your lubrication, your fitting practices, your shaft currents or your operating loads.

Core idea

Don't throw the evidence in the scrap bin. The raceway path patterns, the discoloration, the dent shapes and the fracture surfaces tell you the root cause in the majority of cases. A 15-minute inspection of the failed bearing is the cheapest RCA you will ever run.

Bearings rarely die of old age

The handbook opens with a number that should reframe how your site thinks about bearings.

What happens to bearings in service ~90% Outlive the machine they're installed in retired with the equipment, still serviceable ~9.5% replaced preventively swapped before failure for security reasons ~0.5% are damaged or fail in service ≈ 50 million bearings per year — almost all of them prematurely, with an identifiable cause figures from SKF, Bearing damage and failure analysis (pub. 14219, diagram 1)
Fig. 1 — Bearing outcomes in service, redrawn by Rob Reliability from SKF publication 14219. Only a tiny fraction of bearings fail – and when they do, it is rarely because they reached the end of their fatigue life.

The punchline: a bearing that fails early was almost always killed – by lubrication problems, contamination, poor fitting, misalignment, excessive or unexpected loads, or stray electrical current. Industry estimates popularized by the major bearing makers attribute roughly a third of premature failures to lubrication, with contamination and mounting errors taking most of the rest – genuine material fatigue is a minority. Each of those killers is preventable, which is why this document pairs so naturally with precision maintenance and the I-P interval logic.

The six failure modes (ISO 15243)

The handbook classifies all rolling bearing damage into six families – the ISO 15243 standard – each with sub-modes, each recognizable by eye or with a magnifier.

Bearing failure modes ISO 15243 classification 1 · FATIGUE Rolling contact fatigue · Subsurface initiated — spalling · Surface initiated — from poor lubrication or contamination Look for: flaking, spalled craters on raceways 2 · WEAR Progressive material loss · Abrasive — hard particles grinding · Adhesive — smearing, metal transfer under sliding Look for: dull grey surfaces, mirror smearing 3 · CORROSION Chemical attack · Moisture corrosion — rust stains · Fretting corrosion — fit interfaces · False brinelling — standstill vibration Look for: red/black oxide, dents at ball pitch 4 · ELECTRICAL EROSION Current through the bearing · Excessive voltage — craters · Current leakage — VFD shaft currents, fluting washboard Look for: fluting lines, micro-craters 5 · PLASTIC DEFORMATION Permanent indentation · Overload — true brinelling · Indentation from debris · Indentation by handling Look for: dents, nicks from mounting blows 6 · FRACTURE & CRACKING Rings or cages break · Forced fracture — overstress, fits · Fatigue fracture — cyclic bending · Thermal cracking — friction heat Look for: through-cracks, broken ribs/cages
Fig. 2 — The six ISO 15243 failure mode families and their main sub-modes, drawn by Rob Reliability after SKF publication 14219 (ch. 4) and ISO 15243:2017. The original handbook pairs each sub-mode with photographs – worth keeping on the workshop bench.

Why the classification matters: "the bearing failed" is not a failure mode. Fatigue from normal duty, fluting from a VFD shaft current and false brinelling during transport all end in the same seized machine – but the corrective actions are completely different (resize/re-lubricate vs. install shaft grounding vs. fix transport locking). Naming the mode correctly is what makes the fix stick. This is exactly the failure-mode discipline RCM and FMECA demand (see our RCM II summary).

Reading the evidence: path patterns

Before looking for damage, the handbook teaches you to read the load path pattern – the polished track the rolling elements leave on the raceways. It records how the bearing was actually loaded.

RACEWAY UNROLLED — the polished band is the load path ✓ Normal radial load centred, uniform band — the healthy baseline Misalignment band runs diagonally shoulder to shoulder Oval compression two opposite widened zones — out-of-round housing or shaft Excessive axial load heavy band displaced to one side read the load picture first — then look at the damage; an off-centre path turns a "fatigue" diagnosis into an alignment problem
Fig. 3 — Load path patterns on an unrolled raceway, drawn schematically by Rob Reliability after SKF publication 14219 (path pattern chapter). The polished track records how the bearing was actually loaded – compare it with the table below before naming a failure mode.
Path pattern you seeWhat it means
Centred, uniform band on the stationary ring, all the way around the rotating ringNormal radial load – the healthy baseline to compare against.
Band off-centre or running diagonally from one shoulder to the otherMisalignment between shaft and housing – check alignment and housing faces.
Path widened in two diametrically opposite zones on the stationary ringOval compression – out-of-round housing or shaft, or pinched bearing.
Heavily loaded path on one side, axially displacedExcessive axial load – check preload, fits, thermal expansion allowance.
Pattern all the way around both rings on a bearing with one stationary ringThe "stationary" ring is creeping/turning – fit is too loose.

This habit – load picture first, damage second – is what separates a real analysis from guessing. A perfect fatigue spall in an off-centre path is not a fatigue problem; it's a misalignment problem that expressed itself as fatigue.

Symptom → likely cause → corrective action

A condensed field card from the handbook's "damage and actions" chapter. Confirm with the path pattern and the machine's history before acting.

What you findLikely mode & causeCorrective action
Flaking / spalling patches on raceway, started below surface Subsurface fatigue – load above design, or simply end of life Check actual loads vs. rating; consider higher-capacity bearing or reduce load; if life was acceptable, it's a normal replacement.
Surface distress, micro-spalls, dull frosted raceways Surface-initiated fatigue – lubricant film too thin (wrong grade, too hot, contaminated) Fix viscosity for operating temperature; cleaner lube handling; check seals; review relubrication interval and quantity.
Mirror-like smearing on rollers/raceway entry zones Adhesive wear – rollers sliding instead of rolling (light load + high speed, lube too thin) Minimum load check, correct viscosity, consider different internal clearance or preload.
Regular washboard "fluting" lines across the raceway Electrical erosion – VFD common-mode or shaft current passing through the bearing Shaft grounding ring, insulated bearing or housing, check cable/earthing per drive vendor guidance.
Dents at rolling-element pitch on a machine that was stored or transported False brinelling – standstill vibration with no oil film Lock shafts during transport; rotate spares periodically in storage; isolate standby machines from vibration.
Rust bands, etched patches, blackened fit surfaces Corrosion – moisture/process ingress, or fretting at the fits Seal upgrade, dry storage, correct fits and clamping; consider coated or stainless variants.
Localized dents with raised edges; damage at exactly cage-pocket spacing Plastic deformation – mounting blows transmitted through rolling elements Fit with proper sleeves/induction heater; never hammer the wrong ring; train and audit fitting practice.
Cracked ring through the bore; discoloration straw-blue Fracture / thermal cracking – excessive interference fit, or severe sliding heat Check shaft tolerances and taper drive-up; verify clearance class; investigate lubrication starvation.

Field tip

Photograph everything before cleaning, and keep the bearing. Date, machine, position (DE/NDE), orientation marks, lubricant sample. Half of all bearing RCAs die because the evidence went in the bin or the solvent bath before anyone looked at it.

Using it on site

How to turn a free PDF into measurably fewer bearing failures.

  1. Make it the standard. Put the ISO 15243 mode names in your CMMS failure code list for rotating equipment. "Bearing failed" should be impossible to enter.
  2. Inspect every failed bearing of critical machines. 15 minutes, the path-pattern check, photos, mode named. File it against the work order.
  3. Trend the modes, not the failures. Twelve "electrical erosion" entries on VFD motors is a shaft-grounding project, not twelve repairs. This is bad-actor analysis at component level (see our Bad Actor Diagnostic).
  4. Feed it back to the left of the curve. Most confirmed modes point straight at a precision standard: fitting tools, lube spec, alignment tolerance, storage rules. That's where the failure actually gets eliminated.
  5. Pair with condition monitoring. Each mode has a known signature (fluting and fatigue are textbook vibration finds). Knowing your dominant modes tells your analysts what to listen for – and your P-F intervals follow (see the P-F summary).

References & further reading

This summary is original explanatory writing. All concepts belong to their authors – go to the sources.

  1. SKF Group. Bearing damage and failure analysis. Publication BU/I3 14219 EN. Free PDF (SKF media hub)
  2. ISO. ISO 15243:2017 – Rolling bearings — Damage and failures — Terms, characteristics and causes. ISO standard page
  3. SKF. Bearing knowledge and maintenance resources. skf.com/group/support
  4. Rob Reliability. P-F & D-I-P-F summary – where bearing defect detection and prevention fit in your strategy.

Disclaimer. This page is an independent educational summary written entirely in Rob Reliability's own words. It is not affiliated with, sponsored by or endorsed by AB SKF or ISO. SKF® is a registered trademark of AB SKF; it is used here solely to identify the publication being discussed. No text or photographs from the SKF publication are reproduced; the diagrams are our own original illustrations, and the failure-mode classification summarised here follows the publicly available structure of ISO 15243 without reproducing the standard's text. If you are a rights holder and have any concern about this page, contact us at hello@robreliability.com and we will address it promptly.

Done for you

Want this applied to your site?

Trend the modes on your own work orders: the Bad Actor Diagnostic ranks your top 10 value destroyers with root causes, in 3-4 weeks. From $7,000.

Bad Actor Diagnostic
← All resources Hunt your bearing killers with us