Doc #96 — Bearing Repair and Fabrication

---

RECOMMENDED ACTIONS (BY ACTUAL URGENCY)

First week:

1. Classify all bearing distributor warehouse stocks as Category A strategic reserve (Doc #1). Major NZ distributors include CBC (Complete Bearing and Component Solutions), BSC (Bearing Service Centre), NTN Bearing Corp NZ, SKF NZ, and Schaeffler NZ/Australia agents.³ These stocks are irreplaceable.
2. Issue guidance to all machine shops, maintenance facilities, and equipment operators: do not discard any bearing, regardless of condition. Failed bearings are a source of balls, races, and cages for reconditioning.

First month:

3. Begin detailed bearing inventory as part of the national asset census (Doc #8). Capture: distributor stock by type and size, institutional maintenance stocks, and (critically) bearing sizes installed in Priority 1 national equipment (hydro stations, grid infrastructure, water treatment, food processing).
4. Identify NZ personnel with bearing reconditioning experience — some automotive machine shops and industrial maintenance operations recondition bearings. These skills are rare and valuable.
5. Inventory all precision grinding equipment in NZ (cylindrical grinders, internal grinders, centreless grinders, surface grinders) — this equipment is essential for both bearing reconditioning and eventual manufacture (Sections 4 and 5).

First 3 months:

6. Establish a national bearing allocation system. Bearings from the strategic reserve are allocated by equipment priority, not first-come-first-served. Priority 1: hydro station and grid infrastructure. Priority 2: water treatment and food processing. Priority 3: essential transport. Priority 4: machine tools and workshops. Priority 5: all other applications.
7. Begin systematic bearing reconditioning trials at the best-equipped NZ workshops — establish cleaning, inspection, re-greasing, and component-swap procedures.
8. Begin plain bearing production at NZ foundries (Section 3) — cast bronze bushes for common sizes, starting with the most frequently needed agricultural and industrial bearing housings.
9. Assess NZ’s capacity to produce bearing bronze alloys — inventory copper and tin stocks, assess recycling sources (Doc #59).

First year:

10. Establish a bearing repair centre with dedicated precision measurement and grinding capability. Locate at or near the best-equipped NZ precision engineering firm.
11. Develop standard plain bearing designs for the most common applications — publish specifications and distribute to NZ foundries and machine shops.
12. Begin experimental program on bearing steel production — assess whether Glenbrook (Doc #89) can produce a high-carbon steel suitable for bearing applications, even if inferior to AISI 52100.
13. Commission an engineering study on which ball bearing applications can be converted to plain bearings without unacceptable performance loss, and which cannot.

Phase 3–4 (Years 3–7):

14. Scale up plain bearing production to meet routine demand for slow-speed applications.
15. Develop ball and race grinding capability — precision grinding to the tolerances required for functional (if not original-specification) bearings.
16. If trade with Australia develops, bearing steel and finished bearings should be priority import items. Australia has no bearing manufacturing either, but may have access to Asian production via trade routes.

Phase 4–5 (Years 7–15+):

17. Attempt first locally manufactured ball bearings using NZ-produced or NZ-hardened steel, NZ-ground balls, and NZ-ground races. Accept that early production will have wider tolerances and shorter service life than imported bearings.
18. Develop NZ bearing steel production — requires either chromium import or development of alternative hardening pathways (see Section 8).

---

ECONOMIC JUSTIFICATION

The cost of bearing failure

Bearings are small, inexpensive components that support enormously valuable systems. A 6205 bearing costs approximately NZ$10–30 in normal times. The hydro turbine generator it supports produces electricity for thousands of homes. A tractor wheel bearing costs $20–50; the tractor it supports feeds hundreds of people. The economic ratio — bearing cost to system value — is extreme, which means bearing failure has consequences out of all proportion to the component’s size and cost.

Cascading failure is the central economic argument. Bearings do not fail in isolation. When a water pump bearing fails and cannot be replaced, the pump stops. When the pump stops, the water treatment plant loses a process stream. When water treatment loses capacity, the distribution system pressure drops and contamination risk rises. When bearing failure stops a hydro generator, downstream workshops lose electricity and cannot run machine tools needed to fabricate replacement parts for other failing equipment. These cascades are not hypothetical — they are documented in industrial collapse scenarios and in the early phases of past economic disruptions where maintenance supply chains were cut.⁴ Each bearing failure that cannot be remedied with a local substitute multiplies into downstream failures across the entire equipment network. The economic case for bearing repair and fabrication capability is not the value of individual bearings but the value of preventing these cascades.

Person-years by role

The workforce required for bearing maintenance capability differs substantially across the three tiers:

Precision machinists — the binding constraint at all tiers:

NZ’s precision machinist population (operators capable of work to ±0.01 mm on cylindrical and boring operations) is estimated at 3,000–5,000 nationally, including automotive machine shop operators, tool room machinists, and precision engineering staff.⁵ These are not general engineers — they are people who can reliably hold tolerances tight enough for bearing work. Under recovery conditions, this population competes for time across every precision manufacturing priority in the Recovery Library (Doc #91, Doc #93, Doc #97). Precision machinists are among the most severely scarce skilled trades in NZ’s recovery economy.

Bearing work claim on this population, by tier:

Activity	Setup investment	Ongoing demand	Skill level required
Plain bearing production	2–4 person-years (design + tooling)	0.5–2 FTE per production site	Competent machinist (±0.02 mm)
Ball bearing reconditioning	1–2 person-years (procedures + training)	2–5 FTE (national centre)	Precision machinist (±0.005 mm) + measurement skill
New ball bearing manufacture	10–20 person-years (R&D)	10–20 FTE at production scale	Master-level precision machinist + process engineer

Metallurgists:

Bearing steel development (Section 6 and Section 8) requires metallurgical expertise that NZ has only in small numbers — primarily at Victoria University’s School of Chemical and Physical Sciences, Callaghan Innovation, and within Glenbrook’s process team (Doc #89). The specific task of developing a workable bearing steel from NZ-producible materials is genuinely research-level work. Estimated demand: 2–4 metallurgists for 5–10 years before the material question is resolved. NZ has perhaps 20–40 metallurgists with relevant experience nationally. Bearing steel development would consume a significant fraction of this pool.⁶

Quality inspectors:

Bearing quality control requires precision measurement to at least ±0.002 mm (2 micrometres) for reconditioning work, and ±0.0005 mm (0.5 micrometres) for eventual manufacture. This is instrument calibration and measurement skills, not general engineering inspection. NZ has metrology capability concentrated in calibration laboratories (Measurement Standards Laboratory, industrial calibration services) and in precision engineering firms. An estimated 100–200 people in NZ currently work at this level of measurement competence.⁷ Bearing reconditioning and manufacture would require perhaps 5–15 of these people at national scale. This is a manageable but non-trivial claim on a small pool.

Breakeven analysis

Every rotating machine in NZ depends on bearings. Electric motors (powering water pumps, ventilation, refrigeration, machine tools), hydro generators (the primary electricity source), vehicle drivetrains (essential transport), agricultural equipment (food production), and manufacturing machinery all require bearings to function.

The breakeven calculation: Bearing repair capability costs approximately 5–10 FTE at the reconditioning stage (including setup) and prevents cascading failures in equipment valued in aggregate at hundreds of millions to billions of dollars. The ratio of prevented loss to investment cost is large enough — likely several orders of magnitude — that the economic case is strong. The question is not “whether” to invest in bearing capability but “how much and how fast.”

Specific breakeven scenarios:

• One hydro generator (e.g., Manapouri, generating ~850 MW) failing due to a bearing failure that could not be resolved would cost NZ the equivalent of approximately 2–5 person-years of bearing maintenance staff per week of outage, in lost electricity generation and downstream industrial capacity.⁸
• One water treatment plant pump failing due to a bearing failure in a major urban centre creates a public health emergency within days, requiring emergency response costs and infrastructure that dwarf the cost of maintaining a bearing reconditioning capability.
• Loss of machine tool bearings (Section 7.1) stops the manufacture and repair of every other mechanical component — it propagates across the entire industrial recovery.

The inventory argument: NZ’s existing bearing stock is large but finite and non-renewable until domestic manufacture is achieved. Every bearing that cannot be reconditioned and is discarded represents a permanent reduction in NZ’s rotating machinery operational life. The bearing reconditioning programme extends the effective service life of the national bearing inventory by an estimated factor of 2–5x for common sizes (the range reflects that some bearings can be cleaned and re-greased multiple times while others yield only one reconditioning cycle before damage is too severe), at a workforce cost that is small relative to the industrial value preserved.

Opportunity cost: precision machinists are extremely scarce

The honest economic constraint is not money — it is people. Precision machinists at the level needed for bearing reconditioning and eventual manufacture are among the scarcest skills in NZ’s recovery economy. They are simultaneously needed for machine tool maintenance (Doc #91), hydroelectric plant maintenance (Doc #66), steam plant work, and dozens of other precision engineering priorities.

Allocating 5–10 precision machinists to a national bearing reconditioning centre is a genuine opportunity cost. The tradeoff is explicit: those machinists are not available for other precision work during the time they spend on bearing reconditioning. The economic justification must acknowledge this rather than treating bearing work as a free good.

The case for prioritising bearing capability despite the opportunity cost:

1. Bearing maintenance is a multiplier, not a competitor. A machinist working on bearing reconditioning enables every other machine that uses those bearings to continue operating — including the machine tools operated by other machinists. The leverage ratio is very high.
2. Plain bearing production is the lowest-cost intervention. Converting slow-speed applications to plain bearings requires only standard lathe skills and releases ball bearing inventory for applications that cannot be converted. The marginal cost of this work is low because it can be absorbed into existing foundry and machine shop operations.
3. The alternative is equipment failure. If bearing capability is not developed and the imported stock depletes, the failure mode is not gradual degradation — it is equipment stopping. The economic cost of inaction scales with every machine that cannot be returned to service.

Honest assessment: The opportunity cost argument points toward establishing the bearing reconditioning centre as a national facility with dedicated staff, rather than spreading the work across multiple workshops as part-time activity. Concentrated capability is more efficient, protects the precision skills base from fragmentation, and produces better quality outcomes.

---

1. NZ’S BEARING INVENTORY

1.1 Bearings in service

Every electric motor contains at least two bearings. Every pump, every fan, every vehicle wheel, every gearbox, every conveyor, every alternator. The total number of bearings in service across NZ is unknown but is estimated in the tens of millions — possibly exceeding 100 million.⁹

Rough order-of-magnitude estimates:

Category	Approximate units	Bearings per unit	Total bearings (est.)
Motor vehicles (approximately 4.4 million registered)10	4,400,000	20–40 (wheel, transmission, engine, alternator, A/C, pumps)	90–175 million
Electric motors (industrial, commercial, domestic)	5–10 million (est.)	2	10–20 million
Agricultural equipment (tractors, implements, milking, feed-out)	200,000–400,000	10–30	2–12 million
Hydro, wind, and geothermal generators	~200 major units	2–4 (large, critical bearings)	~500–800
Machine tools	10,000–30,000 (est.)	5–15	50,000–450,000
Pumps (water, sewage, industrial, domestic)	500,000–1,000,000 (est.)	2–4	1–4 million
Household appliances (washing machines, dryers, fans)	Several million	2–6	10–30 million

These figures are rough estimates. The actual numbers are unknown and should be established through the national asset census (Doc #8), with priority given to bearings in Category 1 national infrastructure.

The key observation is that the vast majority of bearings — perhaps 80–90% — are in motor vehicles. Under fuel rationing (Doc #1), most vehicles are mothballed. Their bearings become a strategic reserve: accessible through cannibalisation (Doc #1) when needed for higher-priority applications, provided the bearing size matches.

1.2 Distributor and warehouse stocks

NZ bearing distributors hold inventory covering the most common bearing sizes and types for immediate supply to industry. Major distributors include:¹¹

• CBC (Complete Bearing and Component Solutions): Multiple branches nationally. Broad range of bearings, seals, power transmission components.
• BSC (Bearing Service Centre): Branches in main centres.
• NTN Bearing Corporation NZ: Direct presence of the Japanese manufacturer.
• SKF NZ: Agent for the Swedish manufacturer — also provides condition monitoring and maintenance services.
• Schaeffler (FAG/INA) agents: German manufacturer representation.
• General industrial distributors (Motion, Konnect Fastening Systems, others) carrying bearing lines.

Total NZ distributor bearing stock is commercially sensitive data not publicly available. A rough estimate based on import volumes and distribution chain depth suggests tens of thousands to low hundreds of thousands of individual bearings across all sizes and types.¹² Common sizes (6205, 6206, 6207, 6208, 6305, 6306 — the workhorses of industrial and agricultural equipment) will be better stocked than unusual sizes.

1.3 The standardisation advantage

Ball bearings are one of the most standardised products in industrial manufacturing. The ISO dimension series (ISO 15, formerly ISO 492/199) specifies bearing dimensions internationally — a 6205 bearing is the same dimensions regardless of manufacturer: 25 mm bore, 52 mm outside diameter, 15 mm width.¹³ This standardisation means:

• Bearings can be interchanged between manufacturers
• Common sizes serve many different machines
• Cannibalised bearings from one machine can serve another if the size matches
• Plain bearing replacement bushes need only match the bore and housing dimensions

This standardisation is a significant advantage for the reconditioning and substitution strategies described in this document.

1.4 Depletion timeline

Under normal conditions, NZ consumes bearings at a rate determined by the running hours and operating conditions of the national machine population. Under recovery conditions, with most vehicles mothballed and industrial activity reduced, bearing consumption drops — but does not stop. Essential equipment runs continuously: hydro turbines, water pumps, grain mills, essential transport, machine tools.

The critical question is not “when will NZ run out of bearings?” (the total stock is very large) but “when will NZ run out of bearings in the specific sizes needed for critical equipment?” A hydro generator bearing is not interchangeable with a car wheel bearing. The depletion problem is size-specific and application-specific, not aggregate.

Honest assessment: The aggregate NZ bearing stock is sufficient for many years — probably decades — under rationed use with reconditioning. But specific critical sizes may become scarce much sooner, particularly large bearings for hydro generators, main spindle bearings for important machine tools, and specialised bearings for essential industrial equipment. The inventory and allocation system (Recommended Action #3 and #6) exists to identify and manage these specific vulnerabilities.

---

2. BEARING FAILURE MODES AND LIFE EXTENSION

Understanding how bearings fail is essential for extending their life. Most bearing failures are not sudden catastrophic events — they are progressive degradation that can be detected and managed.

2.1 Ball bearing failure modes

Fatigue spalling: The most common failure mode under proper lubrication. Repeated stress cycles cause subsurface cracks in the race material, which propagate to the surface and flake away (spall). The spalled area generates vibration and noise that increases as the damage spreads. A spalling bearing can often continue to operate for some time — the question is how much additional damage accumulates.¹⁴

Lubrication failure: The most common preventable failure mode. Insufficient lubricant, wrong lubricant type, contaminated lubricant, or expired lubricant allows metal-to-metal contact between balls and races. This causes rapid wear, heat generation, and eventual seizure. Under recovery conditions, where lubricant stocks are limited and bio-lubricant substitutes may not perform as well (Doc #34), lubrication failure risk increases.

Contamination: Dirt, water, abrasive particles, and corrosive substances entering the bearing accelerate wear. Damaged seals and shields — or operation in dirty environments without adequate protection — are the primary contamination pathways.

Corrosion: Bearings stored in humid environments without protective lubricant coating develop surface corrosion (rust on steel components). Corroded bearing surfaces act as stress concentrators for fatigue spalling and increase friction and noise. NZ’s maritime climate makes corrosion a significant storage degradation mechanism.

Overloading: Loads exceeding the bearing’s rated capacity cause premature fatigue. This includes both static overload (denting the races — a condition called brinelling) and dynamic overload (accelerated spalling).

Misalignment: If the shaft and housing are not properly aligned, the bearing carries uneven load. One side of the race is stressed more than the other, accelerating localised fatigue.

Electrical damage (arcing): In electric motors, stray electrical currents can arc across the bearing, creating pits in the race surface (electrical discharge machining of the bearing). This is a common failure mode in VFD-driven (variable frequency drive) motors and can occur in generators.¹⁵

2.2 Life extension strategies

Correct lubrication (highest impact): Proper lubricant type, correct quantity, and appropriate relubrication intervals extend bearing life more than any other single factor. SKF estimates that lubrication-related failures account for approximately 36% of all premature bearing failures.¹⁶ Under recovery conditions, this means:

• Identify the correct lubricant for each critical bearing application (Doc #34)
• Maintain relubrication schedules — do not skip or extend intervals
• Where petroleum grease is unavailable, use the best available substitute (calcium grease for slow-speed bearings, castor oil-based grease for higher speeds — see Doc #34) and shorten the relubrication interval
• Keep lubricant clean. Contaminated lubricant is worse than marginal lubricant.

Contamination control: Maintain bearing seals and shields. Replace damaged seals. Keep bearing housings clean. In dusty or wet environments, add external protection (labyrinth seals, flingers, protective covers). Clean workshop practice during bearing installation.

Condition monitoring: Regular monitoring of bearing condition — by vibration, temperature, and sound — allows detection of developing problems before catastrophic failure. Vibration analysis is the most sensitive method and requires portable vibration measurement equipment (imported, finite stock, but relatively robust instruments). Simple methods also work:

• Sound: A stethoscope (or even a screwdriver pressed to the bearing housing and held against the ear) can detect changes in bearing noise. Experienced fitters develop an ear for bearing condition.
• Temperature: A bearing running hotter than normal may be failing. Temperature comparison between similar bearings on the same machine is a useful diagnostic.
• Visual inspection: Disassemble accessible bearings periodically. Look for spalling, discoloration (heat damage), corrosion, cage damage, and lubricant condition.

Proper installation: Many bearing failures originate during installation. Hammering a bearing onto a shaft (transmitting force through the balls, damaging the races) is a common cause of premature failure. Bearings should be pressed onto shafts using force applied only to the ring being fitted (inner ring when mounting on a shaft, outer ring when pressing into a housing). Heating the bearing to 80–100°C in an oil bath or induction heater expands the bore, allowing it to slide onto the shaft without force.¹⁷

Speed reduction: Bearing life is inversely related to speed, because lower speed means fewer stress cycles per unit time. Operating equipment at reduced speed — where the application allows it — extends bearing life proportionally. A bearing running at half speed accumulates fatigue cycles at half the rate, roughly doubling its service life, all else being equal.¹⁸

Load reduction: Similarly, reducing load on a bearing extends its life dramatically. The relationship is approximately cubic — halving the load increases theoretical life by a factor of eight. Where equipment can be operated at reduced capacity (lighter cuts on a machine tool, reduced pump throughput, lighter vehicle loads), this extends bearing life.

---

3. PLAIN BEARING PRODUCTION

3.1 What plain bearings are

A plain bearing (also called a journal bearing, bush bearing, or sleeve bearing) is the simplest type of bearing: a cylindrical sleeve in which a shaft rotates. The shaft and sleeve are separated by a film of lubricant. There are no balls or rollers — load is carried by the lubricant film itself (hydrodynamic lubrication) or, at low speeds and high loads, by direct contact between the shaft and bearing surface with boundary lubrication.

Plain bearings predate ball bearings by thousands of years — wooden journal bearings appear in watermills from antiquity, and bronze sleeve bearings were standard in sailing ship, steam engine, and early industrial machinery applications.¹⁹ They are less efficient than ball bearings (higher friction, more heat generation, more lubricant required) but they are simpler, more tolerant of contamination, capable of carrying very high loads, and — critically — they can be made with NZ capabilities.

3.2 Materials

Bearing bronze (the primary NZ option): Bronze alloys containing copper, tin, and sometimes lead, zinc, or phosphorus are the traditional plain bearing material. The most common bearing bronzes:²⁰

Alloy	Composition	Properties	Applications
Phosphor bronze (CuSn10P)	90% Cu, 10% Sn, trace P	Hard, strong, excellent wear resistance	Medium-duty bearings, bushings, thrust washers
Leaded bronze (CuPb10Sn10)	80% Cu, 10% Pb, 10% Sn	Softer, good conformability, embeds contaminants	Heavy-load low-speed bearings
Gunmetal (CuSn5Zn5Pb5)	85% Cu, 5% Sn, 5% Zn, 5% Pb	Good all-round properties, castable	General-purpose bushings, valve parts

NZ materials availability: Copper is mined in small quantities in NZ, primarily as a byproduct of gold mining in Waihi and the West Coast. Total domestic copper production is modest — NZ imports most of its copper.²¹ Tin is not mined in NZ and must come from existing stocks or future trade. Under recovery conditions, the primary copper and tin supply for bearing bronze is recycled material: scrap bronze fittings, plumbing fittings, old bearings, electrical equipment (copper wire), and existing bronze and brass scrap stocks. NZ holds significant quantities of copper in its built environment (wiring, plumbing, motors) and in scrap yards. The national copper and tin inventory should be assessed as part of the census (Doc #8).

Babbitt metal (white metal): A family of tin-lead or lead-tin alloys developed by Isaac Babbitt in the 1830s, specifically designed for bearing surfaces. Babbitt has excellent conformability (it deforms around slight misalignment), embeddability (hard particles sink into the soft surface rather than scoring the shaft), and anti-seizure properties. It is the traditional lining for large journal bearings — including hydro turbine bearings.²²

Babbitt alloys typically contain:

Type	Composition	Applications
Tin-based (ASTM B23 Grade 2)	89% Sn, 7.5% Sb, 3.5% Cu	Premium bearing surfaces, high-speed
Lead-based (ASTM B23 Grade 15)	83% Pb, 15% Sb, 1% Sn, 1% As	Lower-cost, heavy-duty, lower-speed

NZ babbitt availability: NZ has existing babbitt stocks in current bearing linings and in maintenance supplies. Tin is the binding constraint for tin-based babbitt. Lead is available from recycled batteries (Doc #35) and other sources. Antimony is not mined in NZ but is present in existing babbitt linings and some lead alloys. Recycling existing babbitt linings from replaced bearings preserves the alloy composition.

Cast iron: For very low-speed, lightly loaded applications (hinges, non-critical pivots), grey cast iron can serve as a bearing material. It is self-lubricating to a degree (graphite flakes in the iron matrix provide dry lubrication). NZ can cast grey iron (Doc #35).

3.3 Plain bearing production process

Bronze bush casting and machining:

1. Pattern-making: A wooden or metal pattern of the bush (slightly oversized to allow for machining) is produced. For standard sizes, patterns can be made once and reused hundreds of times. Pattern-making is a skilled trade that should be part of foundry training (Doc #93).

2. Sand casting: The pattern is used to create a sand mould. Molten bronze is poured at approximately 1,000–1,100°C. For simple bushes, this is straightforward sand casting (Doc #93). For longer bushes with tight concentricity requirements, centrifugal casting produces superior results — the spinning mould forces denser metal outward, producing a more uniform and defect-free bore.²³

3. Rough machining: The casting is chucked on a lathe and rough-bored to near final dimension, with approximately 0.5–1 mm left for finishing. The outside diameter is turned to match the housing bore.

4. Finish machining: The bore is finish-bored or reamed to final dimension. Typical bearing clearance (the gap between shaft and bore) is 0.001 mm per mm of shaft diameter — a 50 mm shaft requires a bore of approximately 50.05 mm. This requires machining to ±0.01–0.02 mm, which is within the capability of a well-maintained manual lathe with an experienced operator (Doc #91).²⁴

5. Oil groove cutting: Lubrication grooves are cut into the bore surface, typically using a small boring tool or graver on the lathe. Groove geometry affects lubricant distribution and bearing performance — standard groove patterns are documented in bearing engineering references.²⁵

6. Fitting: The finished bush is pressed into the housing (interference fit on the outside diameter) and may be finish-bored in situ for best concentricity with the housing.

Dependency chain:

• Copper and tin (recycled — Doc #90 addresses alloy sourcing)
• Foundry capability (Doc #93)
• Lathe with boring capability (Doc #91)
• Precision measurement — micrometers, bore gauges (Doc #34, Section 10)
• Engineering knowledge — shaft fits, clearances, groove geometry
• Lubricant for the finished bearing (Doc #34)

Babbitt re-lining of journal bearings:

Many large plain bearings (hydro turbine bearings, large pump bearings, marine stern tube bearings) use a steel or cast iron shell with a babbitt lining. When the babbitt surface wears, the bearing is re-lined — the old babbitt is melted out, the shell is cleaned and tinned (coated with a thin layer of tin for bonding), and fresh babbitt is poured or centrifugally cast into the shell. This is a well-established maintenance procedure that was routine in NZ engineering workshops until the 1980s–1990s, when the availability of cheap imported replacement bearings made re-lining less common.²⁶

The re-lining skill exists in NZ but is rare and aging. Any NZ fitters or engineers who have poured babbitt should be identified and their knowledge captured (Doc #160). The process is documented in standard bearing engineering texts but involves enough craft skill — particularly in achieving a good bond between the babbitt and the shell, and in scraping the finished surface to fit the shaft — that textbook knowledge alone is insufficient.

Babbitt re-lining process:

1. Remove old bearing. Note shaft dimensions, bearing clearances, and any alignment data.
2. Melt out old babbitt (if re-using the alloy, collect and clean it).
3. Clean the shell thoroughly. Remove all old babbitt, flux residues, and contaminants.
4. Tin the shell — apply a thin layer of tin (using a tinning flux and molten tin) to the bearing surface of the shell. This ensures the new babbitt bonds metallurgically to the shell rather than sitting loose.²⁷
5. Preheat the shell to approximately 200°C to prevent thermal shock when babbitt is poured.
6. Pour molten babbitt (heated to approximately 400–450°C for tin-based, 350–400°C for lead-based) into the shell. For half-bearings (split bearing shells), pour in a jig that positions a mandrel to form the bore. For full cylindrical bearings, centrifugal casting produces the best results.
7. Allow to cool slowly — rapid cooling causes shrinkage cracks in the babbitt.
8. Machine the bore to final dimension on a lathe.
9. Hand-scrape the bore to achieve contact with the shaft — using Prussian blue marking compound and the actual shaft (or a gauge shaft) as a reference. The scraping process removes high spots until the babbitt surface shows uniform contact across approximately 70–80% of the bearing area.²⁸

Babbitt re-lining dependency chain:

• Babbitt alloy (recycled from existing linings, or tin + antimony + copper for tin-based; lead + antimony for lead-based — see Doc #90 for metals sourcing)
• Tinning flux and tin for shell preparation
• Heat source capable of sustained 400–450°C (gas furnace, coke furnace, or electric furnace — Doc #93)
• Lathe with boring capability for final machining of the poured lining (Doc #91)
• Prussian blue marking compound (imported, finite stock — substitute: lamp black in light oil)
• Hand scraping tools and a practitioner trained in scraping technique (Doc #156)
• Precision measurement — micrometers, bore gauges, shaft gauges (Doc #91)
• Centrifugal casting jig for cylindrical bearings (fabricated locally from steel)

3.4 Performance comparison: plain vs. ball bearings

Plain bearings substitute for ball bearings with real performance trade-offs that must be understood:

Property	Ball Bearing	Plain Bearing (Bronze)
Friction coefficient (running)	0.001–0.005	0.01–0.10 (depending on speed, load, lubricant)
Starting friction	Low	High (no hydrodynamic film until shaft is turning)
Speed capability	High (10,000+ RPM for small bearings)	Moderate (typically <3,000 RPM; <500 RPM for large shafts)
Load capacity (radial)	Moderate to high	Very high (for properly designed journal bearings)
Lubrication requirement	Grease, infrequent relubrication	Continuous oil supply or frequent greasing
Contamination tolerance	Poor (particles damage races)	Good (soft metals embed particles)
Misalignment tolerance	Low (causes edge loading)	Moderate (babbitt conforms)
Noise	Low when healthy	Moderate
NZ producibility	Not feasible near-term	Feasible now

The friction penalty is real but manageable. A machine converted from ball bearings to plain bearings will consume more power and generate more heat at the bearing. For a small electric motor, the efficiency loss is estimated at 2–5%, based on the friction coefficient difference in the table above (plain bearing friction is roughly 5–20x higher than ball bearing friction, but bearing friction is only a fraction of total motor losses).²⁹ For a hydro turbine bearing, the parasitic loss is more significant. The trade-off is acceptable when the alternative is no bearing at all.

Speed is the hard constraint. Plain bearings work well at low speeds but have a practical upper speed limit set by the lubricant’s ability to maintain a hydrodynamic film and dissipate heat. A bronze bush without pressurised lubrication is typically limited to PV values (pressure times velocity) below approximately 1.8 MPa·m/s.³⁰ High-speed applications — electric motor bearings at 3,000 RPM, machine tool spindles at 1,000–5,000 RPM — are at or beyond the practical limit for plain bearings unless pressurised lubrication is provided. These applications are where ball bearing reconditioning and eventual new manufacture matter most.

---

4. BALL BEARING RECONDITIONING

4.1 Why reconditioning works

A “failed” ball bearing is not necessarily a complete loss. Typically, one component has failed while others remain usable:

• Spalling on one race, but balls and the other race are intact
• Cage (retainer) broken, but balls and races are serviceable
• Seal or shield damaged, but the bearing elements are fine
• Lubricant degraded, but the bearing itself is undamaged

Reconditioning recovers usable components and reassembles them into functional bearings. This is not the same as manufacturing — it is maintenance and parts recovery.

4.2 Reconditioning procedures

Level 1 — Clean and re-grease (simplest, widest application):

1. Remove the bearing from the machine. Note any damage or unusual wear patterns.
2. Remove seals or shields (carefully — if reusable, preserve them).
3. Clean all components thoroughly in solvent (petroleum solvent while available; later, ethanol from Doc #51, or heated canola oil). Remove all old grease and contaminants.
4. Inspect all components visually and by feel. Roll each ball between clean fingertips — any flat spot, rough spot, or irregularity is detectable. Examine races under magnification for spalling, corrosion, or discoloration.
5. If all components pass inspection, re-grease with appropriate lubricant (Doc #34) and reassemble.
6. Reinstall seals or shields. If original seals are damaged, fabricate replacements from available elastomer or felt material, or operate the bearing with external protection (labyrinth seal, flinger).

Level 2 — Component swap (requires matching bearings):

When one component of a bearing is damaged but others are serviceable, components from multiple bearings of the same size can be combined. This requires:

• Multiple bearings of the same dimension series (e.g., several 6205 bearings)
• Precision measurement to verify component compatibility — ball diameter, race groove dimensions, clearance
• Clean assembly conditions

For example: Bearing A has a spalled inner race but good outer race and balls. Bearing B has a spalled outer race but good inner race and balls. The good inner race from B and good outer race from A can be combined with the best available balls and cage to produce one serviceable bearing from two failed ones.

The measurement challenge: Component swapping requires verifying that the balls from one bearing match the races from another — specifically that the radial internal clearance (the play between balls and races when assembled) is within acceptable limits. This requires precision measurement to at least ±0.005 mm. Micrometers and bore gauges (Doc #91) are essential.

Level 3 — Race repair (most difficult):

Minor spalling on a race surface can sometimes be stoned out (polished with a fine abrasive stone) if the damage is shallow. This does not restore the race to original specification, but it may extend serviceable life. The repaired area will have slightly altered geometry and will re-spall eventually, but the bearing may provide months or years of additional service in a lower-duty application than its original installation.

Deep spalling, brinelling (dent marks), or significant corrosion damage generally cannot be repaired to produce a reliable bearing. These components become donors for balls or cages.

4.3 Reconditioning infrastructure

A functional bearing reconditioning operation requires:

• Clean workspace: Bearing reconditioning must be done in a clean environment. Airborne particles that enter a bearing during assembly will accelerate wear. A dedicated clean area within an existing workshop — enclosed, with filtered air or positive pressure — is adequate. Full clean-room conditions are unnecessary.
• Cleaning equipment: Solvent wash stations, ultrasonic cleaner (if available — imported, finite), compressed air for drying.
• Magnification: A magnifying glass (5–10x) or stereo microscope for visual inspection of races and balls.
• Precision measurement: Micrometers (to 0.001 mm), bore gauges, ball gauges, internal clearance measurement fixtures.
• Lubricant: Appropriate bearing grease (Doc #34).
• Inventory and traceability system: Record which components came from which bearings, what inspection was performed, and what grease was used. A reconditioned bearing should carry a tag noting its history and any limitations.

---

5. THE BALL BEARING MANUFACTURING CHALLENGE

5.1 Why it is genuinely difficult

New ball bearing manufacture is not an incremental extension of existing NZ capability — it is a new industry that requires precision exceeding anything NZ currently produces. The difficulty lies in three converging requirements:

Material hardness: Bearing races and balls must be hardened to 58–64 HRC (Rockwell C scale) to resist fatigue and wear. This hardness is achieved by heat-treating high-carbon chromium steel (AISI 52100: approximately 1.0% C, 1.5% Cr, 0.35% Mn, 0.25% Si).³¹ The chromium provides through-hardenability — the ability to harden uniformly through the cross-section, not just on the surface. NZ Steel at Glenbrook (Doc #89) produces structural steel from ironsand, but does not currently produce high-carbon steel or alloy steel with controlled chromium content. Whether Glenbrook can be adapted to produce a suitable bearing steel is one of the critical uncertainties (Section 8).

Geometric precision: A bearing ball must be round to within 0.0001 mm (0.1 micrometre). A bearing race must have a groove profile accurate to similar tolerances with a surface finish below 0.1 micrometre Ra (roughly 100 times finer than a typical machined surface). This level of precision is achieved through multiple stages of grinding and superfinishing, using precision grinding machines with extremely rigid spindles, high-quality grinding wheels, and active in-process measurement.³²

Consistency: Every ball in a bearing must be the same diameter to within 0.0001 mm. Inconsistent balls cause uneven load distribution, vibration, and premature failure. Mass production of consistent precision components requires process control at a level NZ has never attempted.

5.2 The manufacturing process (simplified)

For reference and planning, the basic sequence for ball bearing manufacture:³³

Ball production: 1. Cut wire or bar stock into slugs 2. Cold-head (forge) slugs into rough spheres in a heading die 3. Flash removal — grind or machine the forging flash from the equator 4. Heat treatment — harden and temper to 60–64 HRC 5. Hard grinding — grind between two grooved cast iron plates or in a ball grinding machine through progressively finer stages 6. Lapping — final finishing between hardened lapping plates with fine abrasive compound to achieve roundness and surface finish 7. Washing, inspection, and sorting by size to ±0.0001 mm grades

Race production: 1. Cut tube or bar stock to length 2. Forge or machine rough ring shape 3. Machine groove profile (turning) 4. Heat treatment — harden and temper to 58–62 HRC 5. Grind bore and outside diameter (cylindrical grinding) 6. Grind groove profile (internal/external grinding with formed wheels) 7. Superfinish groove (oscillating stone or tape finishing to <0.1 micrometre Ra) 8. Wash and inspect

Assembly: 1. Match inner race, outer race, and ball set for correct internal clearance 2. Insert balls through a loading slot or by tilting the inner race 3. Install cage (retainer) to space the balls evenly 4. Pack with grease 5. Install seals or shields 6. Final inspection — noise, vibration, running torque

Ball bearing manufacture dependency chain:

• Bearing steel: AISI 52100 or equivalent requiring chromium (Section 6), or alternative hardening pathway (case-hardened plain carbon steel, manganese steel)
• Steel bar or tube stock, hot-rolled and annealed — requires steelworks capability (Doc #89) adapted for alloy steel production
• Heat treatment furnace with controlled atmosphere (to prevent decarburisation during hardening) and oil quench tank — requires furnace construction and temperature control instrumentation
• Cold heading die for ball blanks — requires tool steel die blocks and die-sinking capability (Doc #91)
• Ball grinding machine or grooved cast iron lapping plates — no NZ source; must be fabricated or adapted from existing surface grinders
• Cylindrical grinding machine for race OD/ID (imported, finite stock — Doc #91 inventory)
• Internal grinding machine with formed wheel for race groove profile (imported, finite stock)
• Superfinishing equipment (oscillating stone or tape finisher) — may need to be fabricated locally
• Precision grinding wheels (aluminium oxide or CBN — imported, finite stock, consumed in use)
• Precision measurement instruments to sub-micrometre resolution (imported, finite stock)
• Clean assembly area with filtered air
• Cage (retainer) material — pressed steel or machined brass
• Grease and seals for finished bearings (Doc #34)

5.3 What NZ can develop, and when

Phase 4 (Years 7–15) — First-generation NZ bearings:

Realistic targets for early NZ production:

• Ball roundness: 0.001–0.005 mm (10–50x coarser than imported bearings)
• Race surface finish: 0.4–1.0 micrometre Ra (4–10x coarser)
• Dimensional tolerance: ±0.01 mm on critical dimensions
• Expected service life: 10–30% of equivalent imported bearing life

These bearings would be crude by global manufacturing standards but functional for many applications — particularly slow-to-moderate speed, moderate load applications where the alternative is a plain bearing or no bearing at all.

Phase 5+ (Years 15–30) — Improving precision:

With accumulated experience, better tooling, and possibly imported grinding wheels and measuring instruments via trade:

• Ball roundness: 0.0005–0.001 mm
• Race surface finish: 0.2–0.4 micrometre Ra
• Service life: 30–60% of imported equivalent

Full reproduction of imported bearing quality would require precision grinding machines that are themselves precision-manufactured — the machine tool bootstrap problem (Doc #91) applied to grinding machines. This is a multi-decade development.

---

6. BEARING STEEL: THE MATERIAL CONSTRAINT

6.1 Why standard bearing steel requires chromium

AISI 52100 steel — the universal bearing steel — contains approximately 1.5% chromium. The chromium serves a specific metallurgical function: it increases the hardenability of the steel, meaning the steel can be hardened uniformly through its cross-section by oil quenching from approximately 830°C. Without chromium, a plain high-carbon steel (say, 1095 with 0.95% C and no significant alloying) can be hardened — but only on the surface. The core remains softer because the carbon steel’s hardenability is low. For a bearing ball or race, which must be hard throughout to resist subsurface fatigue cracking, through-hardness is essential.³⁴

6.2 NZ’s chromium problem

NZ has no known chromium ore deposits and does not produce chromium or ferrochromium. All chromium in NZ is in the form of stainless steel objects, chrome plating, and chromium compounds in existing industrial stocks. This is a finite resource.

Potential chromium sources under recovery conditions:

• Stainless steel scrap (containing 10–18% Cr). If stainless steel scrap can be melted and alloyed with high-carbon steel, the resulting alloy might approach bearing steel composition. This requires precise metallurgical control at the foundry or steelworks level — possible at Glenbrook (Doc #89) if the metallurgy can be developed, uncertain elsewhere.
• Chrome plating solutions and residues — small quantities, probably insufficient for steelmaking.
• Import via trade — ferrochromium from South Africa (the world’s dominant producer) or Australia, if trade routes develop.

6.3 Alternative hardening pathways

If chromium is unavailable, NZ must explore alternative approaches to bearing-quality hardness:

Case hardening: Low-carbon steel can be carburised (soaked in a carbon-rich environment at high temperature) to produce a hard, high-carbon surface layer (case) over a tough, low-carbon core. The case can reach full hardness (60+ HRC) and can be ground to precision. Limitation: the case depth is typically 0.5–2 mm. If the bearing surface wears through the case, the soft core is exposed. For applications with moderate wear, case-hardened bearing surfaces may be adequate.³⁵

High-carbon plain steel, oil-quenched in small sections: Plain high-carbon steel (AISI 1095 or similar, producible from NZ Steel with process modification) can be through-hardened in small cross-sections — balls under approximately 8–12 mm diameter (depending on quench severity and exact carbon content), thin race cross-sections. For small bearings, this may be sufficient without chromium. Larger bearings require greater cross-sections and the hardenability problem resurfaces.

Manganese steel: Manganese increases hardenability, though less effectively than chromium. If NZ can source manganese (NZ has small manganese ore deposits, though they are not commercially developed), a high-carbon manganese steel might serve as a bearing steel substitute. The metallurgy would need to be developed through experimentation.³⁶

Honest assessment: None of these alternatives fully replaces AISI 52100 for bearing applications. Each involves a performance compromise. The bearing steel problem is real and may constrain NZ’s bearing manufacturing to smaller sizes and lower-performance applications until chromium becomes available through trade.

---

7. APPLICATIONS: WHICH BEARINGS TO PRIORITISE

Not all bearing applications are equally critical. The reconditioning and allocation strategy should reflect actual priority:

7.1 Highest priority — keep on ball bearings

These applications require the performance that only ball or roller bearings provide, and should receive reconditioning priority and bearing allocation:

• Hydro generator bearings: Large, critical, and with no acceptable substitute. Failure means loss of generating capacity. These bearings are often large-format thrust and radial bearings, some of which use babbitt-lined plain bearings already. The ball or roller bearing components within the generator assembly must be maintained.
• Machine tool spindle bearings: Precision and speed requirements exceed what plain bearings can deliver. A lathe or mill spindle running on plain bearings would produce parts of unacceptable accuracy for precision work.
• High-speed electric motor bearings (>3,000 RPM): Plain bearings generate excessive heat and friction at these speeds without pressurised lubrication systems.
• Vehicle wheel bearings on essential fleet: Tapered roller bearings in vehicle hubs carry high loads at road speed. Plain bearing substitution would require complete hub redesign and would produce inferior performance.

7.2 Convertible to plain bearings

These applications can operate on plain bearings with acceptable performance:

• Low-speed shaft bearings (<500 RPM): Agricultural implement bearings, conveyor bearings, gate mechanisms, slow-speed mixers.
• Pump bearings (moderate speed): Some pump designs can accommodate plain bearings if bearing housing dimensions are adapted.
• Non-precision electric motor bearings (moderate speed, <1,500 RPM): Larger motors running at lower speeds can tolerate plain bearings, particularly if lubrication is maintained.
• General industrial bearings: Hinge pivots, idler rollers, tensioner pulleys, non-precision rotating equipment.

7.3 Conversion procedure

Converting from ball to plain bearings requires:

1. Measurement: Measure shaft diameter, housing bore diameter, and operating clearance of the existing bearing. A 6205 ball bearing has a 25 mm bore and 52 mm outside diameter — the replacement bush must have a 25 mm bore (plus running clearance) and a 52 mm outside diameter (interference fit with housing).
2. Bush design: Determine appropriate running clearance (typically 0.025–0.075 mm for a 25 mm shaft), oil groove pattern, and material (bearing bronze for most applications).
3. Casting and machining: Cast and machine the bush as described in Section 3.3.
4. Housing modification: The existing bearing housing may need modification — the width of a plain bush may differ from the ball bearing it replaces. Thrust loads previously carried by the ball bearing’s shoulders may need separate thrust washers.
5. Lubrication provision: Ball bearings often run on sealed grease with long relubrication intervals. Plain bearings require more frequent lubrication — grease nipples, oil holes, or drip feeders may need to be added to the housing.

---

8. CRITICAL UNCERTAINTIES

Uncertainty	Why It Matters	How to Resolve
Total NZ bearing stock by size and type	Determines which sizes become scarce first	National asset census (Doc #8) with bearing-specific module
NZ Steel capacity to produce high-carbon steel	Determines whether bearing steel is locally producible	Engineering assessment at Glenbrook (Doc #8)
Chromium availability from stainless steel recycling	Determines whether 52100-equivalent steel can be made	Metallurgical research and trial melts
Surviving babbitt re-lining skills in NZ	Determines whether large bearing re-lining can continue	Skills survey as part of heritage skills preservation (Doc #160)
Precision grinding machine inventory and condition in NZ	Constrains bearing reconditioning and manufacture	Machine tool census (Doc #160)
Bearing bronze alloy raw material stocks (Cu, Sn, Pb)	Constrains plain bearing production volume	Metals inventory (Doc #8, Doc #90)
Feasibility of case-hardened plain carbon steel for bearings	Alternative to chromium-alloyed bearing steel	Metallurgical testing program
Actual service life of NZ-produced plain bearings in specific applications	Determines relubrication intervals and replacement schedules	Field trials in operational equipment
Ball bearing reconditioning success rate	Determines how many bearings can be extended	Track outcomes from reconditioning trials

---

CROSS-REFERENCES

• Doc #1 — National Emergency Stockpile Strategy (bearing stock requisition)
• Doc #8 — National Asset and Skills Census (bearing inventory, skills survey)
• Doc #34 — Lubricant Production (bearing lubrication — all types)
• Doc #66 — Hydroelectric Maintenance (turbine bearing requirements)
• Doc #66 — Geothermal Maintenance (turbine and pump bearing requirements)
• Doc #88 — Spare Parts Triage (bearing cannibalisation from donor equipment)
• Doc #89 — NZ Steel Glenbrook (bearing steel production potential)
• Doc #91 — Machine Shop Operations (machining plain bearings, grinding capability, measurement)
• Doc #92 — Blacksmithing (bearing housing fabrication)
• Doc #93 — Foundry and Casting (bronze bush and babbitt casting)
• Doc #105 — Wire, Fencing, and Nails (wire stock for ball blanks)
• Doc #138 — Sailing Vessel Design (marine bearing requirements)
• Doc #157 — Trade Training Priorities (bearing fitting and maintenance skills)
• Doc #160 — Heritage Skills Preservation (babbitt pouring and scraping techniques)

---

FOOTNOTES

---

1. Global ball bearing manufacturing is concentrated among a small number of large manufacturers: SKF (Sweden), NSK, NTN, and JTEKT/Koyo (Japan), Schaeffler Group/FAG/INA (Germany), Timken (USA), and several large Chinese manufacturers (ZWZ, LYC, C&U). NZ has no domestic bearing manufacturing. All bearings are imported as finished products.↩︎

2. AISI 52100 (also designated 100Cr6 under European standards) is the standard bearing steel globally. Its composition (~1.0% C, 1.5% Cr, 0.35% Mn, 0.25% Si) and heat treatment to 58–64 HRC are specified in ISO 683-17 and ASTM A295. Ball grade precision is specified in ISO 3290 — Grade 10 balls (common industrial grade) have a roundness tolerance of 0.25 micrometres and a diameter variation of 0.25 micrometres within a lot. Premium grades (Grade 3, Grade 5) are tighter still.↩︎

3. NZ bearing distributors and their branch networks serve industrial, agricultural, and automotive markets nationally. CBC (https://www.cbcbearings.com.au/ — Australian parent, NZ branches), BSC (https://www.bsc.com.au/ — Australian parent, NZ branches), NTN Bearing Corporation NZ, SKF NZ. Exact stockholding data is commercial and not publicly available.↩︎

4. The role of bearing and component supply failure in industrial collapse is documented in historical case studies of wartime industrial disruption and post-Soviet economic dislocation. A characteristic pattern is that equipment continues to operate until bearings fail, at which point an entire production chain stops simultaneously rather than degrading gradually. See: Overy, R.J., “War and Economy in the Third Reich,” Oxford University Press, 1994 (on the role of component supply in limiting German industrial output); and Davies, R.W. and Wheatcroft, S.G., “The Industrialisation of Soviet Russia,” Vol. 5 (on maintenance supply chain disruption). The Recovery Library does not have access to post-Soviet maintenance failure studies directly, but this pattern is well-attested in industrial history.↩︎

5. NZ precision machinist population estimate: Stats NZ ANZSCO classification 323111 (Precision Metal Trades Worker) and related codes (323312 Toolmaker, 323211 Metal Machinist) together account for approximately 8,000–12,000 workers in NZ’s engineering trades (2023 Census data, https://www.stats.govt.nz/). The subset capable of precision work to ±0.01 mm or better is smaller — estimated at 30–40% of the classified population based on the proportion employed in tool rooms, precision engineering shops, and aerospace/medical device supply chains. The 3,000–5,000 figure is an estimate; the skills census (Doc #8) should establish the actual number and geographic distribution.↩︎

6. NZ metallurgist population: Metallurgy is not separately classified in NZ’s ANZSCO scheme — metallurgists appear within Materials Engineers (225499) and Chemical and Materials Engineers (225411). IPENZ membership data and university graduate surveys suggest NZ has approximately 200–400 engineers with significant materials science training; of these, perhaps 50–100 have specific expertise in ferrous metallurgy (steel). The subset with expertise in bearing steel or tool steel — high-alloy, high-carbon, precision heat treatment — is likely 20–40 nationally. This estimate should be verified through the engineering skills census (Doc #8).↩︎

7. NZ precision machinist population estimate: Stats NZ ANZSCO classification 323111 (Precision Metal Trades Worker) and related codes (323312 Toolmaker, 323211 Metal Machinist) together account for approximately 8,000–12,000 workers in NZ’s engineering trades (2023 Census data, https://www.stats.govt.nz/). The subset capable of precision work to ±0.01 mm or better is smaller — estimated at 30–40% of the classified population based on the proportion employed in tool rooms, precision engineering shops, and aerospace/medical device supply chains. The 3,000–5,000 figure is an estimate; the skills census (Doc #8) should establish the actual number and geographic distribution.↩︎

8. Manapouri power station output capacity: Meridian Energy, Manapouri Power Station fact sheet — the station has a maximum output of approximately 800 MW from its seven generating units (sometimes cited as up to 850 MW depending on operating conditions and unit upgrades). Economic cost of lost generation is calculated at approximately NZ$75–120/MWh (Electricity Authority average wholesale prices, 2022–2024 range, adjusted for recovery-era scarcity), implying approximately NZ$60–100 million per week of full-station outage before accounting for downstream industrial losses. A single unit outage (one bearing failure) would be proportionally less, but the point remains that the ratio of bearing maintenance cost to prevented loss is extreme.↩︎

9. The total NZ bearing population estimate is derived from the component table in Section 1.1. The figure is highly uncertain — the vehicle fleet alone accounts for 90–175 million bearings depending on how comprehensively ancillary components (power steering pumps, A/C compressors, alternators, starter motors) are counted. The aggregate figure is an order-of-magnitude estimate only. The national asset census (Doc #8) should establish a more rigorous figure for critical bearing applications.↩︎

10. NZ motor vehicle fleet: Ministry of Transport, “Annual Fleet Statistics 2023,” https://www.transport.govt.nz/statistics-and-insights/fle... The 2023 figure was approximately 4.36 million registered vehicles (light passenger, light commercial, heavy vehicles, motorcycles). The figure used here is rounded to 4.4 million.↩︎

11. NZ bearing distributors and their branch networks serve industrial, agricultural, and automotive markets nationally. CBC (https://www.cbcbearings.com.au/ — Australian parent, NZ branches), BSC (https://www.bsc.com.au/ — Australian parent, NZ branches), NTN Bearing Corporation NZ, SKF NZ. Exact stockholding data is commercial and not publicly available.↩︎

12. NZ bearing import data is available through Stats NZ trade statistics under HS codes 8482 (ball and roller bearings). Exact current import volumes and in-country stock levels should be verified against the most recent trade data. The estimate of tens to hundreds of thousands of individual bearings in distributor stock is based on the distribution chain serving a fleet of 4.4 million vehicles plus industrial equipment.↩︎

13. ISO 15:2017 (Rolling bearings — Radial bearings — Boundary dimensions, general plan) specifies the standardised dimension series for ball and roller bearings. This standard ensures global interchangeability — a 6205 bearing from any manufacturer fits the same shaft and housing. The dimension plan was originally established by the Anti-Friction Bearing Manufacturers Association (AFBMA) and adopted internationally.↩︎

14. Bearing fatigue theory is based on the Lundberg-Palmgren model (1947, 1952), which relates bearing life to load, as expressed in the ISO 281 bearing life calculation standard. Fatigue spalling is the primary life-limiting failure mode for properly lubricated bearings operating within their design envelope. See: Harris, T.A. and Kotzalas, M.N., “Rolling Bearing Analysis,” 5th ed., CRC Press, 2006.↩︎

15. Electrical bearing damage (EDM — electrical discharge machining of bearing surfaces) is a well-documented failure mode in VFD-driven motors. The shaft voltage induced by the VFD’s switching frequency discharges through the bearing, creating micro-pits in the race surface. See: SKF, “Bearing damage and failure analysis,” Publication WL 82 102. Also relevant to generator bearings where induced shaft voltages may occur.↩︎

16. SKF bearing failure analysis data: SKF estimates that approximately 36% of premature bearing failures are related to lubrication (insufficient, incorrect type, contaminated, or degraded). The remaining failures are attributed to contamination (~14%), fatigue (~34%), and mounting/installation errors (~16%). See: SKF, “Rolling bearings,” Publication PUB BU/P1 10000/2 EN, various editions. https://www.skf.com/↩︎

17. Bearing installation best practice: SKF, NSK, and other major bearing manufacturers publish detailed installation guides. The fundamental principle — apply mounting force only to the ring being fitted — is universal. Heating bearings for installation (typically to 80–110°C) is standard practice for interference-fit applications. Induction bearing heaters are preferred; oil bath heating is the traditional method.↩︎

18. The relationship between speed and bearing life is captured in the ISO 281 rating life calculation. The reference speed and limiting speed ratings published by bearing manufacturers define the speed capability of each bearing type. Reducing speed reduces the frequency of stress cycles on the races and reduces heat generation from lubricant shearing and ball-race contact, both of which extend life.↩︎

19. Plain bearing history: journal bearings in wood and bronze are documented in Roman-era watermills (Vitruvius, “De Architectura,” Book X, c. 25 BC) and in medieval European mill construction. Bronze sleeve bearings became standard in marine and industrial applications during the 18th and 19th centuries, before the widespread adoption of ball bearings in the late 19th century. See: Dowson, D., “History of Tribology,” 2nd ed., Professional Engineering Publishing, 1998.↩︎

20. Bearing bronze alloy compositions and properties: ASM Handbook, Volume 2, “Properties and Selection: Nonferrous Alloys and Special-Purpose Materials,” ASM International. Copper Development Association publications (https://www.copper.org/) provide detailed property data and application guidance for bearing bronzes. The alloys listed are among the most widely used globally.↩︎

21. NZ copper production is modest — primarily as a byproduct of gold mining operations (OceanaGold at Waihi, various West Coast mines). NZ imports the vast majority of its copper requirements. Under recovery conditions, the domestic copper supply is dominated by recycled copper from the built environment (electrical wiring, plumbing, motors, transformers). See: NZ Petroleum and Minerals, mineral production statistics, https://www.nzpam.govt.nz/↩︎

22. Babbitt metal and its application in journal bearings: Babbitt, I., US Patent 1252 (1839) — the original specification for white metal bearing alloy. Modern babbitt practice is documented in: Booser, E.R. (ed.), “Tribology Data Handbook,” CRC Press, 1997, and in machine design references such as Shigley’s “Mechanical Engineering Design,” various editions.↩︎

23. Centrifugal casting of bearing bushes: Janco, S.J., “Centrifugal Casting,” American Foundrymen’s Society. Centrifugal casting produces denser, more uniform castings than static sand casting because centrifugal force concentrates the heavier metal outward while pushing lighter inclusions and gas porosity inward (where they are machined away during boring). This is the preferred method for bronze bearing bushes and babbitt-lined bearing shells.↩︎

24. Plain bearing design and clearance calculations: Shigley, J.E. and Mischke, C.R., “Mechanical Engineering Design,” McGraw-Hill, various editions. The rule of thumb for bearing clearance (0.001 mm per mm of shaft diameter) is a starting point; actual clearance depends on speed, load, lubricant viscosity, and operating temperature. Oil groove geometry is covered in Rippel, H.C., “Design of Hydrostatic Bearings,” Machine Design series.↩︎

25. Plain bearing design and clearance calculations: Shigley, J.E. and Mischke, C.R., “Mechanical Engineering Design,” McGraw-Hill, various editions. The rule of thumb for bearing clearance (0.001 mm per mm of shaft diameter) is a starting point; actual clearance depends on speed, load, lubricant viscosity, and operating temperature. Oil groove geometry is covered in Rippel, H.C., “Design of Hydrostatic Bearings,” Machine Design series.↩︎

26. Babbitt bearing re-lining was a standard engineering workshop skill in NZ through the mid-20th century. As imported replacement bearings became cheap and readily available, the skill fell out of routine practice. Some NZ engineering firms and hydro station maintenance teams retain this capability, but the number of practitioners is small and declining. The skills census (Doc #8) should specifically identify babbitt pouring capability.↩︎

27. Babbitt metal and its application in journal bearings: Babbitt, I., US Patent 1252 (1839) — the original specification for white metal bearing alloy. Modern babbitt practice is documented in: Booser, E.R. (ed.), “Tribology Data Handbook,” CRC Press, 1997, and in machine design references such as Shigley’s “Mechanical Engineering Design,” various editions.↩︎

28. Hand scraping of babbitt bearings to achieve shaft contact: the scraping process is the same in principle as hand scraping of machine tool surfaces (Doc #156, Section 9.3), using Prussian blue marking compound to identify high spots and a hand scraper to remove them. The target is 70–80% contact area across the bearing surface, verified by blue transfer pattern. See: Connelly, R., “The Whitworth Three Plates Method” (for the general principle), and bearing-specific guidance in King, R.J., “Machine Tool Reconditioning.”↩︎

29. Electric motor efficiency loss from bearing substitution: in a typical small induction motor (1–5 kW), bearing friction accounts for roughly 5–15% of total motor losses, with the remainder being stator copper losses, core losses, and stray load losses. Replacing ball bearings (friction coefficient ~0.002) with plain bearings (friction coefficient ~0.02–0.05) increases the bearing friction component by an order of magnitude, but since bearing friction is a minority of total losses, the overall motor efficiency decrease is estimated at 2–5 percentage points. The actual figure depends on motor size, speed, load, and lubricant properties. See: Shigley (note 15) for friction coefficient data; motor loss breakdown from: Alger, P.L., “Induction Machines,” Gordon and Breach, 1970.↩︎

30. PV limits for plain bearings: the PV (pressure × velocity) value is a widely used empirical limit for plain bearing operation without pressurised lubrication. For bearing bronze operating with grease lubrication, PV limits of approximately 1.0–1.8 MPa·m/s are typical. Higher PV values require pressurised oil lubrication or materials with higher temperature capability. See: copper alloy bearing data from the Copper Development Association.↩︎

31. AISI 52100 (also designated 100Cr6 under European standards) is the standard bearing steel globally. Its composition (~1.0% C, 1.5% Cr, 0.35% Mn, 0.25% Si) and heat treatment to 58–64 HRC are specified in ISO 683-17 and ASTM A295. Ball grade precision is specified in ISO 3290 — Grade 10 balls (common industrial grade) have a roundness tolerance of 0.25 micrometres and a diameter variation of 0.25 micrometres within a lot. Premium grades (Grade 3, Grade 5) are tighter still.↩︎

32. Ball bearing manufacturing process: Hamrock, B.J. and Anderson, W.J., “Rolling Element Bearings,” NASA Reference Publication 1105, 1983. Also: Harris and Kotzalas (note 6). The multi-stage grinding and lapping process for balls (typically 5–8 stages, from rough grinding through fine lapping) is the core of bearing manufacturing technology and represents accumulated process knowledge that took the global bearing industry decades to develop.↩︎

33. Ball bearing manufacturing process: Hamrock, B.J. and Anderson, W.J., “Rolling Element Bearings,” NASA Reference Publication 1105, 1983. Also: Harris and Kotzalas (note 6). The multi-stage grinding and lapping process for balls (typically 5–8 stages, from rough grinding through fine lapping) is the core of bearing manufacturing technology and represents accumulated process knowledge that took the global bearing industry decades to develop.↩︎

34. The metallurgical role of chromium in bearing steel: chromium increases hardenability by retarding the austenite-to-ferrite/pearlite transformation during cooling, allowing the steel to transform to martensite (the hard phase) at slower cooling rates and through thicker cross-sections. Without chromium, only the surface of a thick section transforms to martensite during oil quenching; the core remains softer pearlite. See: ASM Handbook, Volume 4, “Heat Treating,” ASM International.↩︎

35. Case hardening (carburising) as an alternative bearing surface treatment: carburised steel bearings are used in some automotive applications (tapered roller bearings in wheel hubs are sometimes carburised rather than through-hardened). The case depth is limited (typically 0.5–2 mm), and if the case is worn through or chipped, the soft core is exposed. For applications with controlled wear and moderate loads, carburised bearings are a proven technology. See: ASM Handbook, Volume 4 (note 20).↩︎

36. Manganese as a hardenability element: manganese is less effective than chromium for hardenability improvement but is significantly cheaper and more widely available. AISI 1340 steel (0.40% C, 1.75% Mn) has useful hardenability for moderate cross-sections. For bearing applications, a higher-carbon manganese steel (say, 0.90–1.0% C, 1.5–2.0% Mn) might provide adequate through-hardenability for small bearing components. NZ manganese deposits are documented in GNS Science mineral occurrence records but are not commercially developed. See: https://www.gns.cri.nz/↩︎