← ArticlesRolling-Element Bearings: Selection, Fits and Installation: When Additives Change the GameEngineering · Machine DesignLesson 12/17← PrevNext →
GuidePublished 14 Aug 202622 min readBy Kevin JoginMachine DesignBearingsRolling-Element Bearings: SelectionFits and Installation

Engineering · Machine Design · Bearings

Rolling-Element Bearings: Selection, Fits and Installation: When Additives Change the Game

Engineering handbook for rolling-element bearings: selection, fits and installation, covering when additives change the game, common mistakes: the hall of shame,...

Executive summary

This handbook section converts the supplied engineering material into a practical, source-controlled reference. It concentrates on the following learning outcomes.

When Additives Change the Game
Common Mistakes: The Hall of Shame
Mistake #1: Selecting on Static Fit Alone
Mistake #2: Ignoring Axial Loads
Mistake #3: Using Room-Temperature Lubricant Data
Mistake #4: Forgetting Minimum Load Requirements

When Additives Change the Game

The manufacturer diagrams for the a₂₃ factor typically show a shaded region for lubricants with additives. EP (Extreme Pressure) additives and anti-wear additives can provide additional life benefit, especially when κ is marginal (between 0.4 and 1).

However, additives are not a substitute for correct viscosity. They provide a safety margin, not a permanent solution.



Common Mistakes: The Hall of Shame

the practitioner has a list on his office wall. Every bearing failure he's investigated has taught him something. Here are the patterns:


Mistake #1: Selecting on Static Fit Alone

The error: "The bearing fits the shaft, so it's the right bearing."

The reality: A bearing that physically fits may be wildly wrong for the actual loads, speeds, and conditions.


Mistake #2: Ignoring Axial Loads

The error: "It's a radial bearing on a horizontal shaft — there's no axial load."

The reality: Gear separating forces, thermal expansion, misalignment, and vibration all create axial loads. They might be small, but they're not zero.


Mistake #3: Using Room-Temperature Lubricant Data

The error: Specifying lubricant viscosity at 40°C when the bearing operates at 80°C.

The reality: At 80°C, many lubricants have lost 60–80% of their viscosity compared to 40°C.


Mistake #4: Forgetting Minimum Load Requirements

The error: Selecting a massively oversized bearing "just to be safe."

The reality: If the bearing doesn't carry enough minimum load, the rolling elements skid, causing surface damage and premature failure. Bigger is not always better.


Mistake #5: Ignoring Contamination

The error: "The bearing is sealed, so contamination isn't a concern."

The reality: Seals degrade. Environments change. Installation introduces debris. Contamination must be actively managed throughout the bearing's life.


Mistake #6: Not Converting Units

The error: Mixing imperial and metric values in calculations.

The reality: This is a universal problem. Always verify: loads in kN or N? Dimensions in mm? Speed in rpm? Life in millions of revolutions or hours? One misplaced decimal can make a bearing selection useless.

Key unit conversions for global practice:

From (Imperial) To (Metric) Factor
1 inch mm × 25.4
1 foot mm × 304.8
1 pound (mass) kg × 0.454
1 psi kPa × 6.89
1 hp W × 747
1 BTU kJ × 1.055
°F to °C °C = (°F − 32) / 1.8


The Sensitivity Factor: How Small Changes Create Big Consequences

This is the table that finally made the practitioner a believer. the practitioner generated it to show leadership exactly how sensitive bearing life is to design variables:


Load Sensitivity (Ball Bearing, C = 52.7 kN)

Applied Load P (kN) C/P Ratio L₁₀ (million revs) Relative Life
15 3.51 43 0.15×
12 4.39 85 0.30×
10 5.27 146 0.51×
8 6.59 286 1.0× (baseline)
6 8.78 678 2.37×
5 10.54 1,171 4.09×
4 13.18 2,286 7.99×

Key takeaway: Reducing load from 8 kN to 6 kN (a 25% reduction) more than doubles the bearing life. Going from 8 kN to 4 kN (50% reduction) gives nearly 8× the life. Load reduction is the most powerful design lever available.


Contamination Sensitivity (Bearing 6309, P = 8 kN, κ = 2.9)

Condition η_c a_the bearing supplier Life (million revs) Relative to Clean
Very clean 1.0 ~10 2,860 1.25×
Clean 0.8 8 2,288 1.0× (reference)
Normal 0.5 ~4 1,144 0.50×
Contaminated 0.2 1.2 343 0.15×
Heavily contaminated 0 ~0.1 29 0.01×

Read the bottom row. Under heavy contamination, even with excellent lubrication, bearing life collapses to 1% of what it would be under clean conditions. This is why the practitioner calls cleanliness the "multiplier of everything."



The Decision Framework: Which Method Should You Use?

Situation Recommended Method Why
Quick sizing during concept design Method 1 (ISO L₁₀) Simple, fast, conservative
Standard industrial design Method 2 (ISO Adjusted) Accounts for lubrication; more realistic
Critical or high-value applications Method 3 (New Life Theory) Most accurate; accounts for contamination
Unknown operating conditions Method 1 Don't assume favorable conditions you can't verify
Premium lubrication system Method 2 or 3 These methods reward good lubrication design
Harsh/contaminated environments Method 3 Only Method 3 penalizes contamination quantitatively
Design audit or failure analysis All three Compare results to understand risk exposure


Engineering takeaway

the practitioner's story ends well. The redesigned conveyor system has been running for over four years without a bearing failure. The changes weren't exotic — they were fundamental:

  1. He calculated, rather than guessed, the actual loads — including thermal expansion forces he'd previously ignored
  2. He selected a lubricant based on the operating temperature, not the catalog temperature — increasing κ from below 0.4 to above 2
  3. He specified sealed bearings — increasing η_c from 0.2 to 0.8
  4. He documented everything — so the next engineer who touches this system has the full picture

None of these changes cost significant money. The lubricant upgrade was marginal. The sealed bearings were the same price as the open ones. The real investment was time and knowledge.



Your Action Plan

Here's what to do next, based on where you are:

If you're a student or beginner: Master Method 1 first. Understand the L₁₀ equation deeply. Know what C, P, and the life exponent mean — not just how to plug numbers in, but what they physically represent. Then work through the equivalent dynamic load calculations for each bearing type until they're second nature.

If you're a working engineer: Start applying Method 2 to your designs immediately. Calculate κ for your current bearing installations. You may discover that some of your "adequately lubricated" bearings are actually running with κ below 1 — and that's a failure waiting to happen.

If you're a manager or decision-maker: Understand the contamination sensitivity table. The difference between "clean" and "contaminated" operations is not a 20% life difference — it's a 6× to 100× life difference. Investing in sealing, filtration, and clean installation practices has an ROI that dwarfs almost any other maintenance expenditure.



Quick Reference Formulas Card

Keep this somewhere accessible:

┌─────────────────────────────────────────────────────────┐ │ BEARING LIFE QUICK REFERENCE │ ├─────────────────────────────────────────────────────────┤ │ │ │ Design Life: L = (60 × N × h) / 10⁶ │ │ │ │ Basic Life: L₁₀ = (C / P)^p │ │ p = 3 (ball) or 10/3 (roller) │ │ │ │ Adjusted Life: Lₙₐ = a₁ × a₂₃ × L₁₀ │ │ │ │ New Life: Lₙₐₐ = a₁ × a_the bearing supplier × L₁₀ │ │ │ │ Viscosity: κ = v / v₁ │ │ dₘ = 0.5 × (d + D) │ │ │ │ Mean Diameter: dₘ = (bore + OD) / 2 │ │ │ │ Static Check: P₀ < C₀ (ball) │ │ P₀ < C₀/1.5 (roller) │ │ │ │ Min Load: Fr > 0.01C (ball) │ │ Fr > 0.02C (roller) │ │ │ │ Hours to Revs: h = (L × 10⁶) / (60 × N) │ │ │ └─────────────────────────────────────────────────────────┘



Final Thought

Every machine you've ever relied on — every car, every elevator, every production line, every wind turbine, every airplane — depends on bearings. They're invisible until they fail. And when they fail, nothing moves.

The difference between the practitioner before and after the incident wasn't talent. It wasn't resources. It wasn't technology.

It was respect for the details.

The loads, the lubricant, the cleanliness, the temperature, the alignment — every variable matters, and the math exists to account for all of them.

Your job isn't to memorize every table and formula. Your job is to understand what drives bearing life, ask the right questions, and use the right method for the right situation.

Get that right, and your machines will run for years. Get it wrong, and someone gets a phone call at 2:47 AM.


What's the most surprising thing you learned about bearing selection from this guide? Have you ever experienced a bearing failure that changed how you approach design? Share your experience — the engineering community learns best from real stories.


Further reading in this series:

  • Chapter 2: Journal Bearings — When Rolling Bearings Aren't the Answer
  • Chapter 3: Belt Drives — Power Transmission Without Gears
  • Chapter 4: Chain Drives — When Belts Aren't Enough
  • Chapter 5: Couplings — Connecting Shafts Without Destroying Them

Context and scope

A single misaligned bearing cost the practitioner's production line 72 hours of downtime and a quarter-million in scrapped product. The root cause wasn't a defective bearing. It wasn't a lubrication failure. It was a .0003-inch shoulder that wasn't square.

This guide covers everything you need to know about bearing fits, squareness, alignment, radial and axial clearance, and handling precautions — the topics that separate reliable machines from chronic maintenance nightmares.



Squareness and Alignment: The Foundation of Bearing Life


Why Squareness Matters More Than You Think

Rolling-contact bearings are made of fully hardened steel. Unlike journal bearings, they do not wear in. They do not self-correct. They do not absorb misalignment gracefully.

When you mount a bearing with an out-of-square shoulder or an off-axis housing face, you are forcing the races into a permanent cock. That cock creates:

  • Wobble — the inner race oscillates axially with every revolution
  • Housing excitation — the wobble drives vibrations across a frequency range from below shaft speed to 100× above it
  • Premature fatigue — uneven loading accelerates spalling on one side of the raceway

Key Insight: Housing response to axial excursions forced by bearing wobble has been identified as a major source of noise and "howl" in small electric motors and rotating equipment. The fix isn't a better bearing — it's a stiffer housing and properly aligned races.


The Tolerances You Must Hold

Squareness of end faces and shoulders must be closely controlled. Here are the commercial application alignment tolerances that apply to the vast majority of industrial bearing installations:

Tolerance Item Specification
Housing Face Runout Square to shaft center within .0004 inch/inch of radius (full indicator reading)
Outer Race Face Runout Square to shaft center within .0004 inch/inch of radius (complementary to housing runout — not opposed)
Inner Race Face Runout Square to shaft center within .0003 inch/inch of radius
Cover & Closure Mounting Face Parallelism Parallel within .001 inch
Housing Mounting Face Parallelism Parallel within .001 inch

For precision and preloaded applications: Cut all of the above tolerances in half.


How to Verify Squareness

In applications not controlled by automatic tooling with closely controlled fixtures and bolt torquing mechanisms, races should be checked for squareness by sweeping with a dial indicator mounted against the face of the race while rotating the shaft.

Assembly methods must ensure three things:

  1. Faces are square before the housing cavity is closed
  2. The cover face is square to the shoulder and pulled in evenly
  3. The cover will be located by a face parallel to it when finally seated against the housing


Maximum Shaft Deflection Limits: The Numbers That Save Bearings

Alignment isn't just about the static geometry of the housing. It's about what happens under load when the shaft deflects.

Here are the maximum allowable shaft deflections for different bearing types:

Bearing Type Maximum Deflection (inch per inch)
Well-crowned roller bearings .001
Deep-groove ball bearings .003
Self-aligning ball bearings & spherical/barrel roller bearings Higher limits (designed for misalignment)
All other types (angular contact, tapered, cylindrical roller) .0002
Preloaded ball bearings .0002
Close-clearance tapered bearings .0002
Thrust bearings (most types) .0002

What this means in practice: If your shaft deflects more than .0002 inch per inch under load and you're running tapered roller bearings, you're exceeding the alignment limit — regardless of how perfectly the housing was machined at rest.


The DN Value Threshold

Bearing mounting practice identifies a threshold for alignment sensitivity based on the DN value:

DN=D×NDN = D \times N

Where:

  • DD = bearing bore diameter (mm)
  • NN = shaft speed (RPM)

For DN values of 400,000 or less (medium-speed or slower), medium to light load applications (C/PC/P values of 7 or greater) can endure misalignments comparable to high-capacity precision journal bearings.

Above 400,000 DN, alignment demands increase sharply, and precision mounting practices become non-negotiable.



The Inspection Methods That Actually Matter

For precision and quiet-running applications, you need more than a micrometer:

Inspection Method What It Detects When To Use
Two-point micrometer Size and taper Routine commercial work
Three-point electronic indicator Lobular deviations, waviness, out-of-roundness Any application where vibration matters
Talyrond (continuous recording instrument) Deviations to within a few millionths of an inch Ultra-precise, missile guidance, quiet-running applications

Critical fact: Shaft deformities will be reflected through inner races shrunk onto them. Tight-fit outer races pick up significant deviations in housings. Grinding chatter, lobular out-of-roundness, waviness, and flats of less than .0005 inch from the mean diameter can cause significant roughness.


Roundness Requirements for Needle Roller Bearings

For needle roller bearings (drawn cup type), the requirements are particularly strict because rollers run directly on the shaft surface:

Parameter Tolerance
Mean diameter variation (across radial planes) ≤ .0003 inch (or half the diameter tolerance, if smaller)
Radial deviation from circular form (≤ 1 inch diameter) ≤ .0001 inch
Radial deviation from circular form (> 1 inch diameter) ≤ .0001 × shaft diameter (inches)
Surface finish ≤ 16 micro-inches (0.4 µm) Ra
Housing bore mean diameter variation (NIA type) ≤ .0005 inch (or half tolerance, if smaller)
Housing bore radial deviation (NIA type) ≤ .00025 inch
Housing bore surface finish (NIA type) ≤ 125 micro-inches (3.2 µm) Ra


Bearing Fits: The Science of Getting Tight Where It Counts


Why Fits Exist: Preventing Creep and Slipping

The slipping or creeping of a bearing ring on a rotating shaft — or in a rotating housing — occurs when the fit is too loose. This creep causes:

  • Rapid wear of both shaft and bearing ring when surfaces are dry and highly loaded
  • Fretting corrosion at the interface
  • Loss of radial positioning over time
  • Heat generation and eventual seizure

The fundamental rule of bearing fits is elegant in its simplicity:

The rotating ring gets a press fit. The stationary ring gets a push fit.

The tightness or looseness depends on the service:

  • Shock or vibratory loads → tighter fits than normal
  • Normal service → standard interference/clearance
  • Stationary ring → allowed to creep very slowly so that prolonged stressing of one part of the raceway is avoided

Shaft Fit Selection (Metric Radial Ball & Roller Bearings, ABEC-1/RBEC-1)

The proper shaft fit depends on three factors: rotational condition, load magnitude, and shaft diameter.

Load Classification Reference:

Bearing Type Light Load Normal Load Heavy Load
Ball bearings Up to 0.075C 0.075C to 0.15C Over 0.15C
Cylindrical roller Up to 0.075C 0.075C to 0.2C Over 0.15C
Spherical roller Up to 0.075C 0.070C to 0.25C Over 0.15C

Where CC = Basic Load Rating (per AFBMA-ANSI standards).


Inner Ring Stationary Relative to Load

Condition All Diameters Tolerance Symbol
Must be easily displaceable All sizes g6
Does not need to be easily displaceable All sizes h6

Inner Ring Rotating Relative to Load (or Indeterminate Direction)

Load Ball Bearing Diameter (mm) Cyl. Roller Diameter (mm) Sph. Roller Diameter (mm) Tolerance
Light ≤18 h5
Light >18 to 40 ≤40 ≤40 j6
Light 40–140 40–100 k6
Light 140–320 100–320 m6
Light 320–500 320–500 n6
Light >500 >500 p6
Normal ≤18 j5
Normal >18 to 40 ≤40 ≤40 k5
Normal 40–100 40–65 m5
Normal 100–140 65–100 m6
Normal 140–320 100–140 n6
Normal 320–500 140–280 p6
Normal >500 280–500 r6
Normal >500 r7
Heavy 18–100 k5
Heavy >100 ≤40 ≤40 m5
Heavy 40–65 40–65 m6
Heavy 65–140 65–100 n6
Heavy 140–200 100–140 p6
Heavy 200–500 140–200 r6
Heavy >500 >200 r7
Pure Thrust All diameters All diameters Consult manufacturer j6

Important: For solid steel shafts. Hollow or nonferrous shafts may require tighter fits. When greater accuracy is needed, use j5, k5, and m5 instead of j6, k6, and m6.


Shaft Diameter Deviations from Basic Bore (Inch Design)

Here are the allowable shaft diameter deviations for standard drawn-cup needle roller bearing applications:

Basic Bore (inches) Shaft Rotating/Outer Ring Stationary (ANSI m5) Shaft Stationary/Outer Ring Rotating (ANSI g6)
High / Low High / Low
0.2362–0.3937 +.0005 / +.0002 −.0002 / −.0006
0.3937–0.7087 +.0006 / +.0003 −.0002 / −.0007
0.7087–1.1811 +.0007 / +.0003 −.0003 / −.0008
1.1811–1.9685 +.0008 / +.0004 −.0004 / −.0010
1.9685–3.1496 +.0009 / +.0004 −.0004 / −.0011
3.1496–4.7244 +.0011 / +.0005 −.0005 / −.0013
4.7244–7.0866 +.0013 / +.0006 −.0006 / −.0015
7.0866–9.8425 +.0015 / +.0007 −.0006 / −.0017

Housing Fit Selection (Metric Radial Bearings, ABEC-1/RBEC-1)

Housing fits are driven by the rotational condition of the outer ring, the loading type, and whether the outer ring must be axially displaceable.

Outer Ring Condition Loading Axial Displacement Other Conditions Tolerance
Stationary relative to load Light Must be easily displaceable Heat input through shaft G7
Stationary Light Must be easily displaceable Housing split axially H7
Stationary Light Must be easily displaceable Housing not split H6
Stationary Normal/Heavy Shock with temporary complete unloading Transitional range (K6, K7, M6, M7)

For cast iron or steel housings. Housings of nonferrous alloys (aluminum, magnesium) may require tighter fits.

Where wider tolerances are permissible, use P7, N7, M7, K7, J7, and H7 in place of P6, N6, M6, K6, J6, and H6.


Heating Bearings for Assembly

When an interference fit makes assembly difficult, expand the inner ring by heating:

  • Method: Clean oil bath or temperature-controlled furnace
  • Temperature range: 200°F to 250°F (93°C to 121°C)
  • Absolute maximum: 250°F — exceeding this reduces ring hardness
  • Never use this method on prelubricated bearings (heat destroys the lubricant)


Technical challenge

Armed with the lab report, the practitioner went back to his workshop and re-examined everything he thought he knew about bearing installation. He discovered three critical gaps in his process:

  1. He'd never verified housing shoulder squareness — only bore diameter
  2. He'd assumed standard machining was "good enough" — it wasn't, because the housing had been re-bored after a previous failure, introducing lobular errors
  3. He'd never checked mounted clearance — only free-state clearance

Each of these gaps maps to a section of knowledge that most maintenance engineers never receive formal training on.



Radial and Axial Clearance: The Hidden Variable That Kills Bearings


Why Clearance Is a Design Decision, Not an Afterthought

In designing the bearing mounting, a major consideration is to provide running clearances consistent with the requirements of the application. This is more complex than it sounds because:

  • Race fits absorb clearance. Approximately 80% of the actual interference between the race and its seat shows up as a change in race diameter. This reduces the internal bearing clearance.
  • Heavy, stiff housings and solid shafts increase this effect (closer to 100%)
  • Light metal housings (aluminum, magnesium, sheet metal) and tubular shafts reduce it (less closure)
  • Thermal effects add another variable — heat from the shaft or housing changes dimensions during operation

The 80% Rule

Δdrace0.80×Interference Fit\Delta d_{race} \approx 0.80 \times \text{Interference Fit}

This is the starting point for estimating how much internal clearance you'll lose when you press the bearing onto the shaft or into the housing. For critical applications, measure it directly with feeler gauges or dial indicators after assembly.


Temperature Differentials

When the application imposes heat losses through the housing or shaft, or when a temperature differential exists between inner and outer races, allowances must be made in the proper direction:

  • Shaft hotter than housing → inner race expands more → clearance decreases → allow extra initial clearance
  • Housing hotter than shaft → outer race expands more → clearance may increase → may need tighter initial clearance or preload

Ball Bearings vs. Roller Bearings: Different Tolerances for Preload

Bearing Type Preload Tolerance Consequence of Excess Preload
Ball bearings Can run with moderate preloads up to .0005 inch without affecting life or temperature Relatively forgiving
Roller bearings Lesser tolerance for preloading — must be carefully controlled Overheating and self-destruction

Critical warning: Roller bearings must be carefully controlled to avoid overheating and resulting self-destruction. Unlike ball bearings, they have very little margin for error on preload.


What Robs Clearance After Assembly

Even if your interference fit calculations are perfect, these factors can steal your running clearance:

  • Chips or scores under the race seat
  • Race misalignment (cocked during assembly)
  • Shaft or housing denting
  • Housing distortion (from mounting bolt torque or structural loads)
  • End cover off-squareness
  • Mismatch of rotor and housing axial dimensions

Best practice: In all critical applications, check axial and radial clearances with feeler gauges or dial indicators after assembly to confirm mounted clearances fall within design tolerances.


Precision Clearance Control Methods

For applications demanding tight clearance control:

Method Application Notes
Taper-sleeve mountings Adjustable radial/axial clearance Requires skill and manufacturer support
Opposed ball or tapered-roller bearings with shimmed closures Precision spindles, machine tools Shim thickness controls preload
Tapered bore bearings on tapered shafts Heavy-duty, precision Assembly by controlled heating or hydraulic jack
Adapter sleeve with lock-nut Convenient field mounting Lock-nut advances race on tapered sleeve

For tapered bore bearings: Advancement of the inner race can be done by controlled heating (to expand the race) or by hydraulic jack. Removal typically requires hydraulic devices due to heavy interference fits.



Bearing Closures: Keeping the Good In and the Bad Out

Before we continue the practitioner's story, you need to understand the protective systems that keep bearings alive after installation.


Shields vs. Seals: Know the Difference

Feature Shield Seal
Contact No contact — definite clearance between shield and inner race Bears against the rotating member
Grease behavior Allows grease exchange between bearing and housing cavity Retains grease within the bearing
Speed limitation Suitable for high speeds Excessive pressure and speed can burn or score the seal
Protection level Moderate — excludes large particles High — excludes moisture, fine dust, and contaminants

Seal Materials and Their Applications

  • Leather seals — wide speed range, but at high speeds the cup must face outward to prevent burning; lubricant must reach the contact area
  • Rubber seals — versatile, available as cartridges that press into housing ends
  • Cork, felt, plastic compositions — specialty applications, lower speed ranges
  • Labyrinth seals — non-contact, frictionless, insensitive to temperature and speed; limit leakage rather than eliminate it; often used as auxiliary protection alongside contact seals

Important: Only light pressure of leather against the shaft should be maintained. At high speeds where abrasive dust is present, arrange the seal with leather cupped outward to lead lubricant into the contact area.



Bearing Fits in Detail: Press Fits, Push Fits, and Everything Between


The Creep Problem

When a bearing ring fits loosely on a rotating shaft or in a rotating housing, it creeps — rotating slowly relative to its seat. This creep:

  • Causes rapid wear of both surfaces when dry and heavily loaded
  • Generates fretting corrosion (red-brown oxide powder at the interface)
  • Leads to progressive loosening and eventual catastrophic movement
  • Creates noise, vibration, and heat

The Correct Fit Philosophy

The stationary ring is intentionally fitted with a slight clearance or light transition fit. This is not a deficiency — it's by design. The stationary ring is allowed to creep very slowly so that prolonged stressing of one part of the raceway is avoided, distributing the load zone over time and extending fatigue life.


The Full Metric Shaft Deviation Table (ANSI/ABMA 7-1995)

For complete engineering reference, here are the allowable deviations of shaft diameter from basic bore diameter:

Basic Bore (mm) Basic Bore (inches) g6 h6 j5 j6 k5 k6 m5 m6 n6 p6 r6 r7
6–10 0.24–0.39 −.005/−.014 0/−.009 0/−.006 +.004/−.002 +.007/−.002 +.007/+.001 +.012/+.006
10–18 0.39–0.71 −.006/−.017 0/−.011 0/−.008 +.005/−.003 +.008/−.003 +.009/+.001 +.015/+.007
18–30 0.71–1.18 −.007/−.020 0/−.013 +.005/−.004 +.009/−.004 +.011/+.002 +.017/+.008
30–50 1.18–1.97 −.009/−.025 0/−.016 +.006/−.005 +.011/−.005 +.013/+.002 +.018/+.002 +.020/+.009 +.025/+.009
50–80 1.97–3.15 −.010/−.029 0/−.019 +.006/−.007 +.012/−.007 +.015/+.002 +.021/+.002 +.024/+.011 +.030/+.011 +.039/+.020
80–120 3.15–4.72 −.012/−.034 0/−.022 +.006/−.009 +.013/−.009 +.018/+.003 +.025/+.003 +.028/+.013 +.035/+.013 +.045/+.023

(All values in mm. For additional sizes beyond 120 mm, consult ANSI/ABMA 7-1995.)


The Full Housing Bore Deviation Table

Basic OD (mm) Basic OD (inches) G7 H7 H6 J7 J6 K6 M6 N6 P6
18–30 0.71–1.18 +.007/+.028 0/+.021 0/+.013 −.009/+.012 −.005/+.008 −.011/+.002 −.017/−.004 −.024/−.011 −.031/−.018
30–50 1.18–1.97 +.009/+.034 0/+.025 0/+.016 −.011/+.014 −.006/+.010 −.013/+.003 −.020/−.004 −.028/−.012 −.037/−.021
50–80 1.97–3.15 +.010/+.040 0/+.030 0/+.019 −.012/+.018 −.006/+.013 −.015/+.004 −.024/−.005 −.033/−.014 −.045/−.026
80–120 3.15–4.72 +.012/+.047 0/+.035 0/+.022 −.013/+.022 −.006/+.016 −.018/+.004 −.028/−.006 −.038/−.016 −.052/−.030
120–180 4.72–7.09 +.014/+.054 0/+.040 0/+.025 −.014/+.026 −.007/+.018 −.021/+.004 −.033/−.008 −.045/−.020 −.061/−.036
180–250 7.09–9.84 +.015/+.061 0/+.046 0/+.029 −.016/+.030 −.007/+.022 −.024/+.005 −.037/−.008 −.051/−.022 −.070/−.041

(All values in mm. Positive values indicate bore larger than basic OD; negative values indicate interference.)



Step 1: Pre-Installation Shaft and Housing Audit

Before touching a bearing, the practitioner now verifies:

  • Shaft shoulder squareness with a dial indicator (< .0004 inch/inch for commercial, < .0002 for precision)
  • Housing bore roundness with a three-point measurement (not just two-point mic)
  • Surface finish on bearing seats (< 16 µin Ra for shafts, < 125 µin Ra for housings)
  • Fillet radii at shaft shoulders — cross-referenced against bearing corner radius tables to prevent race cocking

Step 2: Correct Fit Verification

the practitioner matches every installation to the fit tables, asking three questions:

  1. Which ring rotates relative to the load? (determines press vs. push fit)
  2. What is the load magnitude? (Light / Normal / Heavy relative to C rating)
  3. What is the shaft diameter? (determines specific tolerance symbol)

Step 3: Post-Assembly Clearance Check

After mounting, the practitioner measures:

  • Radial clearance with feeler gauges
  • Axial clearance (endplay) with dial indicators
  • Running clearance compared to the design engineer's specified tolerance band


Shaft and Housing Design: The Rules That Prevent Failure

These recommendations come from decades of field experience and represent hard-won engineering wisdom:


Shaft Design Rules

  • Avoid more than two bearings on a single shaft — the difficulty of achieving accurate alignment makes three or more bearings problematic. Close-spaced bearings can create extremely heavy loads.
  • Shaft diameter between bearings must resist bending — a weak shaft causes misalignment through deflection, which is the root cause of many "bearing" failures.
  • Undercuts for grinding runout should be as small as possible and end in fillets, never sharp corners — sharp transitions are fatigue initiation sites that lead to shaft breakage.
  • Thread cutting for clamping nuts must be true and square to ensure even pressure on bearing inner ring faces. Never cut threads into the bearing seat area.
  • Shoulders must present sufficient surface area in contact with the bearing face to ensure positive and accurate location.

Housing Design Rules

  • Plenty of metal in wall sections — thin areas deflect the boring tool during finish machining, producing out-of-roundness and taper.
  • Transmit radial loads directly to supporting walls or ribs — diaphragm walls connecting offset housings to main walls deflect unless made thick and well-braced.
  • When two bearings are opposed in separate housings, reinforce with fins or webs to prevent deflection under axial loading.
  • Deep housings with long boring tool overhang tend to produce out-of-roundness and taper — use rigid tooling and light finishing cuts.
  • Avoid overly rough bored housings — metal ridges peen down under load over time, eventually resulting in too loose a fit for the outer ring.

Engineering use and verification

Begin with load paths, motion, interfaces and credible failure modes. Define duty cycle, environment, alignment, lubrication, manufacturing variation and maintenance access before choosing a component. Check static strength, fatigue, stiffness, heat, wear and fastening together because improving one constraint can worsen another. Record assumptions and verify the assembled system, not just catalogue ratings for isolated parts.

  • Confirm scope, assumptions, interfaces and required outcome.
  • Use one controlled unit system and show every conversion.
  • Identify current project, customer and regulatory requirements.
  • Separate source examples from mandatory acceptance criteria.
  • Check calculations, tables and selections by an independent method.
  • Verify safety, maintainability and credible failure modes.
  • Record evidence, revisions, approvals and unresolved limitations.
  • Validate the result under representative operating conditions.

Continue learning

Rolling-Element Bearings: Selection, Fits and Installation: Radial and Axial Clearance ConsiderationsGuide · Machine DesignNEXT LESSON →Rolling-Element Bearings: Selection, Fits and Installation: Soft Metal and Resilient Housings (Aluminum,...Guide · Machine DesignRolling-Element Bearings: Selection, Fits and Installation: Shaft Raceway Tolerance LimitsGuide · Machine DesignRolling-Element Bearings: Selection, Fits and Installation: The Four Bearing Types You Actually Need to KnowGuide · Machine Design