Shafts, Keys, Keyseats, Circlips and Seals: The Complete Shaft Design Checklist

The Complete Shaft Design Checklist

After her two-week crash course, the practitioner created this checklist. She laminated it. She taped it to the wall above her desk. She never got a 3 AM call again.

Phase 1: Load Analysis

• Identify ALL loads: torsional, bending, axial
• Calculate steady loads from each component (gears, sprockets, pulleys, weights)
• Identify shock load sources and magnitudes
• Identify motor starting method (soft-start vs. hard-start) for inertia loads
• Calculate forces at every load point on the shaft

Phase 2: Stress Analysis

• Draw free body diagram showing all forces and reactions
• Calculate bearing reactions
• Draw torque diagram
• Draw shear force diagram
• Draw bending moment diagram (both planes if multi-plane)
• If multi-plane, calculate resultant bending moment using Formula 13
• Identify critical location(s) — where combined stress is maximum

Phase 3: Shaft Sizing

• Apply shock/fatigue factors (K_T and K_M) to get design torque and moment
• Calculate equivalent torque T_E (Formula 3)
• Calculate equivalent moment M_E (Formula 4)
• Determine allowable bending and shear stresses
• Apply 0.75 reduction factor at keyway locations
• Calculate required diameter from BOTH shear and bending formulas
• Use the LARGER diameter (more critical condition)
• Round up to next standard shaft size

Phase 4: Component Selection

• Select standard key size for shaft diameter
• Verify key length for shear and crushing
• Select circlips for axial retention
• Select seals appropriate for speed, temperature, and media
• Verify seal shaft surface finish requirements
• Verify bearing selections can handle the calculated reaction loads

Phase 5: Verification

• Check all stress concentrations (keyways, shoulders, holes)
• Verify factor of safety is adequate
• Review starting torque effects on shaft and key
• Confirm all standard sizes are commercially available

Quick Reference: All Formulas in One Place

Quick Reference: Allowable Stress Rules

Engineering takeaway

the practitioner's shaft didn't fail because of bad steel. It didn't fail because of a defective sprocket. It didn't fail because of a freak accident.

It failed because someone sized it without accounting for the stress concentration at the keyway and the inertia loads from a hard-start motor. The shaft was probably 15–20% undersized at its most critical location. Combined with millions of fatigue cycles and never-checked starting torque, failure wasn't a possibility. It was a mathematical certainty.

The formulas in this guide aren't academic exercises. They're the difference between a shaft that runs for decades and one that screams at 3 AM.

Your Turn: Test Your Understanding

Here's a challenge problem to verify you've internalized this material:

A shaft is driven by a 15 kW motor at 1000 rev/min (hard-start). The shaft carries a single spur gear with PCD 150 mm at the midpoint between two bearings spaced 400 mm apart. The gear meshes horizontally. Shaft material: SAE 1040, UTS 620 MPa, yield 350 MPa. The gear is keyed to the shaft.

Can you determine the required shaft diameter?

Try it. Use the procedure in this guide. Check your work against the formulas. If you get it right — you've just become the engineer who never gets called at 3 AM.

What's the worst shaft failure you've seen — or nearly caused? Drop your war story in the comments. The engineering community learns more from failures than from successes.

Next in the series: Chapter 11 — Rolled Steel Sections: The Structural Backbone of Every Machine Frame

About this series: This is part of an ongoing transformation of the Mechanical Design Data Manual into engaging, actionable engineering education. Each chapter is designed to serve beginners learning the fundamentals, experienced engineers needing a quick reference, and anyone who believes that engineering knowledge should be accessible, memorable, and built to last.

Overview

• This chapter covers the mechanical design of shafts — rotating members supported by bearings that transmit torque and power
• Topics include shaft load classification, failure modes, design approaches, stress formulas, load estimation methods, and design procedures for determining minimum shaft diameter
• Also covered: circlip and seal sizing (metric long-life series), and rolled steel section selection for beams and columns
• Key design philosophy: calculate equivalent torque and moment from combined loading, apply shock/fatigue factors, then determine shaft diameter using strength-of-materials formulas
• Two primary design approaches are presented — one based on endurance limit and another based on basic strength of materials with generous safety factors

Key Concepts

• Shaft: A rotating member supported by bearings that transmits torque and power
• Steady Loads: Torsional, bending, and axial loads that occur continuously during operation
• Shock Loads: Intermittent, sudden increases in load (e.g., initial engagement in rolling or pressing operations)
• Inertia Loads: Loads arising from acceleration or deceleration of the shaft and attached equipment
• Equivalent Torque (T_E): A combined measure incorporating both torque and bending moment — used to find shaft diameter from shear stress
• Equivalent Moment (M_E): A combined measure used to find shaft diameter from bending stress
• Shock/Fatigue Factors (K_T, K_M): Multipliers applied to steady torque and moment to account for dynamic loading effects
• Stress Concentration: Localised increase in stress at geometric discontinuities such as keyways, shoulders, grooves, or holes
• Endurance Limit: The stress level below which a material can theoretically sustain an infinite number of load cycles without fatigue failure
• Drive Application Factor (f): A multiplier used to account for the difference between actual transverse shaft load and the simplified calculated value

Shaft Definition and Characteristics

• Definition: A rotating member supported by bearings that transmits torque and power
• Rotation may be continuous, intermittent, uni-directional, or reversing
• Shafts attached to wheels are often called axles
• Usually circular in cross-section — solid or hollow; sometimes square for specific applications
• Typically rigid; flexible shafts (cables) exist for specialised applications but are not covered here
• Usually long relative to diameter
• Commonly made from steel or other metals; non-metallic shafts are sometimes used
• Torque and power are transmitted from an input location (e.g., a prime mover) to an output location (e.g., a driven load)
• Input may come from a motor, engine, or intermediate shaft via gears, belts, or chains
• Output may be transmitted directly to a load or via power transmission devices

Shaft Loads and Stresses

Steady Loads

• Primary design loads that occur relatively continuously during operation
• Three types act on shafts, often simultaneously:

Shock Loads

• Intermittent in nature, causing sudden increases in load
• Common in rolling mills, punching/cropping presses, and similar applications
• Shafts must be designed to withstand shock loads even if they occur only momentarily
• Load-limiting devices (shear pins, overload protection) are often fitted to protect the shaft

Inertia Loads

• Occur during acceleration or deceleration (speed changes)
• Magnitude depends on:
  • Rate of acceleration/deceleration (magnitude)
  • Mass moment of inertia of the shaft and coupled equipment/transmission devices
• Typically occur during start-up and shut-down phases
• For electric motors:
  • Soft-start (current-limiting device fitted): starting torque ≤ 1.5–2× rated load torque
  • Hard-start (no current limiter): starting torque can be 3–5× rated load torque

Shaft Failure Modes

Failure Due to Excessive Load

• Occurs when shaft stress exceeds the yield stress
• Relatively rare due to load-limiting devices (shear pins, overload protection) in most systems
• If overload causes yielding without fracture, the shaft may remain serviceable
• Design goal: prevent stress from exceeding yield stress; permanent deformation = failure

Failure Due to Fatigue

• Most common failure mode for shafts with high revolution counts
• Can occur even when stresses are well below yield point
• Most likely when loads continually fluctuate, especially with stress reversal
• Two mechanisms of stress reversal:
  • Change in load direction: Most common for torsional stress reversal (e.g., vehicle transmission shafts reversing between drive and braking)
  • Rotation of the shaft: Most common for bending stress reversal (e.g., a horizontal shaft with a downward load — top and bottom alternate between tension and compression each half-revolution)

Shaft Design Approaches

Approach 1: Endurance-Based

• Calculate peak loads (including inertia and shock) and stress concentrations as accurately as possible
• Allowable shaft stresses include an allowance for shaft size — larger diameter = lower allowable stress
• For shafts with many revolutions: design to prevent fatigue failure using the endurance limit
• Endurance limit is determined from standardised fatigue tests on polished specimens (typically 8–10 mm diameter)
• A small factor of safety (typically ~1.2) is applied, based on the endurance limit
• Based on relevant national rotating shaft design standards

Approach 2: Strength-of-Materials-Based (Recommended)

• Calculate the maximum design load likely under operating conditions
• Apply shock/fatigue factors and a relatively generous factor of safety
• Account for inertia loads and non-uniformity of material properties
• Use basic strength of materials formulas to calculate shaft stresses or diameter
• More fundamental approach; provides better understanding of stresses involved
• Avoids complex formulas with unstated assumptions
• Based on recognised engineering code methodology

Core Stress Formulas

Combined Shear Stress (Formula 1)

$$f_s = \frac{16 \, T_E}{\pi \, d^3}$$

• Used to find shaft diameter from torsional shear stress

Combined Axial Stress (Formula 2)

$$f_s = \frac{32 \, T_E}{\pi \, d^3}$$

• Used to find shaft diameter from bending (axial) stress

Equivalent Torque (Formula 3)

$$T_E = \sqrt{T^2 + M^2}$$

• Combines torque (T) and bending moment (M) into a single equivalent value

Equivalent Moment (Formula 4)

$$M_E = 0.5 \, (T_E + M)$$

• Used for bending stress calculations

Design Torque and Moment (with shock/fatigue factors)

$$T = K_T \cdot T_S \qquad M = K_M \cdot M_S$$

• Where T_S and M_S are the steady torque and moment
• K_T = shock/fatigue factor in torsion
• K_M = shock/fatigue factor in bending

Shock/Fatigue Factor Values

• K_M = 1.5 minimum even for static loads — accounts for bending stress reversal due to shaft rotation (constant load direction and magnitude)
• These factors apply to bending, torsion, or combined bending and torsion (the most common loading)
• For significant axial loads, more complex formulas (e.g., from relevant engineering codes) should be used, including column effects for compression

Allowable Stresses and Factors of Safety

• For steel shafts using the fundamental design approach:
  • Bending (tension or compression): the smaller of 40% f_y or 24% f_ult
  • Torsion (shear): the smaller of 30% f_y or 18% f_ult
• Where:
  • f_y = yield strength
  • f_ult = ultimate tensile strength
• The allowable shear stress is based on the assumption that shear strength ≈ 75% of tensile strength

Stress Concentration at Keyways

• Keyways are one of the most important sources of stress concentration in shafts
• Located where gears, sprockets, or pulleys are fitted — usually the most highly stressed locations
• Design rule of thumb: allowable stresses with a keyway are 75% of allowable stresses without the keyway
• For other stress concentrations (steps, holes), consult relevant engineering design standards and handbooks

Estimating Shaft Loads

. Weight (Gravitational Load)

• Applies when a heavy pulley, flywheel, or similar component is mounted on a non-vertical shaft
• Causes a transverse bending force:

$$F = m \cdot g$$

• Assumed: shaft supported by low-friction bearings (frictional torque negligible)

. Chain Drive

• Chain tension creates a transverse force on the shaft at the sprocket
• Force at sprocket: $F = T_2 + T_1$ — (Formula 5) — tight side + slack side tension
• Torque at sprocket: $T = (T_2 - T_1) \cdot d/2$ — (Formula 6)
• Where d = pitch circle diameter (PCD) of the sprocket (in metres)
• When transmitting power, slack side tension is usually negligible → T_1 ≈ 0

. Belt Drive (Vee or Wedge)

• Same approach as chain drive, but slack side tension is NOT zero (friction-dependent)
• Formulas 5 and 6 apply for belt drives as well
• For parallel belts: F = T_2 + T_1 (correct); for non-parallel belts: use vector sum (but scalar sum errs on the safe side)
• When both T_2 and T_1 are unknown, use one of three methods:

Method (a): Assume Slack Side Tension

• Based on mid-load power at 1000 rev/min, tension ratio 12:1, smallest recommended PCD for belt size

Method (b): Assume Belt Tension Ratio

• Based on 90% of tension ratio at slip point on smaller pulley; wedge angle 38°, friction coefficient 0.3, centre distance = sum of pulley PCDs; centrifugal effects excluded
• Linear interpolation can be used for intermediate values

Method (c): Assume a Drive Application Factor

• If slack side tension were zero: $F = 2T/d$
• Actual force is greater because T_1 ≠ 0 → apply factor f:

$$F = \frac{2 \, f \, T}{d}$$

• For vee or wedge belt drives, f is typically taken as 1.5
• For chain drives: if T_1 = 0, then f = 1
• Note: f = 1.5 gives a result ~20% higher than other methods; f = 1.25 gives closer correlation

. Gear Drive

• Force on the shaft = resultant transverse force at the gear tooth contact point
• Three force components:
  • F_t (tangential force): produces the torque; $F_t = 2T/d$ — (Formula 8)
  • F_s (separating/radial force): keeps gears in mesh; acts through gear centrelines
  • F (resultant transverse force): vector sum of F_t and F_s
• Pressure angle (θ): angle between F and F_t — typically 20° unless stated otherwise

Spur Gear Formulas

$$F_s = F_t \cdot \tan\theta \qquad \text{(Formula 9)}$$

$$F = \sqrt{F_t^2 + F_s^2} \qquad \text{(Formula 10)}$$

Helical Gear Formulas

• Helical gears have teeth cut at an angle (helix angle α) to the shaft axis
• Stronger and quieter than spur gears, but produce an additional axial force

$$F_s = \frac{F_t \cdot \tan\theta}{\cos\alpha} \qquad \text{(Formula 11 — separating force)}$$

$$F_a = F_t \cdot \tan\alpha \qquad \text{(Formula 12 — axial force)}$$

• The resultant transverse force is still $F = \sqrt{F_t^2 + F_s^2}$

Design Procedure

1. Estimate all loads acting on the shaft (weight, drive forces, gear forces, etc.)
2. Draw torque, shear force, and bending moment diagrams
   • Shear force diagram is optional but useful to draw before the bending moment diagram
3. Identify the critical location — position of maximum combined stress (usually where torque and bending moment are both at maximum, typically at gear/sprocket/pulley locations)
4. Determine the design torque and moment by applying shock/fatigue factors (K_T, K_M)
5. Calculate equivalent torque (T_E) and equivalent moment (M_E)
6. Calculate allowable stresses (with keyway reduction if applicable)
7. Determine minimum shaft diameter using Formulas 1 and 2
8. Select the closest standard shaft size (round up)

Single-Plane vs Multi-Plane Bending

• Single-plane: all resultant transverse forces act in the same plane → one bending moment diagram needed
• Multi-plane: transverse forces act in different planes (e.g., horizontal and vertical) → draw bending moment diagrams for each plane, then combine:

$$M = \sqrt{M_v^2 + M_h^2} \qquad \text{(Formula 13 — resultant bending moment)}$$

• Where M_v = vertical plane moment, M_h = horizontal plane moment

Design Notes

• Treatment excludes significant axial loads — in most shafts, direct axial stress is small relative to bending and torsional stresses
• Examples use single-diameter shafts; the same principles apply to stepped shafts — each step diameter is determined from the maximum stress at that section
• For stepped shafts: apply a stress-concentration factor at each step (depends on ratio of diameters and internal radius)

Rolled Steel Sections

Overview

• Hot rolled sections are available in standard profiles: universal beams, universal columns, parallel flange channels, equal/unequal angles, and merchant bar (rounds, squares, flats)
• Hot rolled sections have a commercial finish — not suitable for rotating shafts (use bright steel for shafts)
• Available in several grades:

Beam Selection Procedure

1. Determine the maximum bending moment (M) from loading and span
2. Calculate the allowable bending stress using the design factor:
   • e.g., for a design factor of 2 on yield: $f_b = f_y / 2$
3. Calculate the required section modulus: $Z = M / f_b$
4. Select the smallest standard section with Z ≥ required Z from beam tables
5. Check self-weight: recalculate reactions, moment, and Z including beam self-weight
6. Verify the selected section is still adequate

Column Selection Procedure

1. Determine the effective length (L_e) based on end conditions:
   • Both ends pinned: L_e = L
   • One fixed, one free (cantilever): L_e = 2L
   • One fixed, one pinned: L_e = 0.7L
   • Both ends fixed: L_e = 0.5L
2. Calculate the design critical load = applied load × design factor
3. Calculate the limiting slenderness ratio:

$$\left(\frac{L_e}{r}\right)_{lim} = \sqrt{\frac{2\pi^2 E}{f_y}}$$

4. Select a trial section from column tables; use the smaller radius of gyration (r_y) for buckling analysis
5. Calculate actual $L_e/r$ and compare with limiting value
6. If $L_e/r$ > limiting value → slender column → use Euler formula:

$$F_{cr} = \frac{\pi^2 E A}{(L_e/r)^2}$$

7. Check that $F_{cr}$ ≥ design critical load; iterate if necessary

Comparison Tables

Shaft Design Approaches Compared

Shaft Failure Modes Compared

Belt Load Estimation Methods Compared

Spur vs Helical Gear Forces

Mermaid Diagrams

Shaft Design Process

flowchart TD
    A[Identify All Shaft Loads] --> B[Estimate Load Magnitudes]
    B --> C{Load Type?}
    C -->|Weight| D[F = mg]
    C -->|Chain Drive| E["F = T₂ + T₁ <br/> T = (T₂ - T₁) · d/2"]
    C -->|Belt Drive| F[Use Method a, b, or c]
    C -->|Gear Drive| G["F_t = 2T/d <br/> F_s = F_t · tan θ <br/> F = √(F_t² + F_s²)"]
    D --> H[Draw Shear Force & Bending Moment Diagrams]
    E --> H
    F --> H
    G --> H
    H --> I[Identify Critical Location]
    I --> J["Apply Shock/Fatigue Factors <br/> T = K_T · T_S <br/> M = K_M · M_S"]
    J --> K["Calculate T_E = √(T² + M²) <br/> M_E = 0.5(T_E + M)"]
    K --> L["Calculate Allowable Stresses <br/> Bending: min(0.4·f_y, 0.24·f_ult) <br/> Shear: min(0.3·f_y, 0.18·f_ult)"]
    L --> M{Keyway Present?}
    M -->|Yes| N[Multiply Allowable Stresses × 0.75]
    M -->|No| O[Use Full Allowable Stresses]
    N --> P["Solve for d from: <br/> f_s = 16·T_E / (π·d³) <br/> f = 32·M_E / (π·d³)"]
    O --> P
    P --> Q[Select Larger Diameter <br/> Round Up to Standard Size]

Shaft Load Classification

flowchart LR
    A[Shaft Loads] --> B[Steady Loads]
    A --> C[Shock Loads]
    A --> D[Inertia Loads]
    B --> B1[Torsional]
    B --> B2[Bending]
    B --> B3[Axial]
    C --> C1[Intermittent <br/> Sudden Increase]
    D --> D1[Start-Up / Shut-Down]
    D1 --> D2["Soft-Start: 1.5–2× rated"]
    D1 --> D3["Hard-Start: 3–5× rated"]

Shaft Failure Decision Tree

flowchart TD
    A[Shaft Under Load] --> B{Stress > Yield?}
    B -->|Yes| C[Excessive Load Failure]
    C --> C1[Permanent Deformation]
    C1 --> C2{Shaft Broken?}
    C2 -->|No| C3[May Still Be Serviceable]
    C2 -->|Yes| C4[Replace Shaft]
    B -->|No| D{Cyclic Stress Reversal?}
    D -->|Yes| E{Stress > Endurance Limit?}
    E -->|Yes| F[Fatigue Failure Over Time]
    E -->|No| G[Infinite Life — No Failure]
    D -->|No| G

Multi-Plane Bending Resolution

flowchart TD
    A[Forces on Shaft in Multiple Planes] --> B[Resolve into Vertical & Horizontal Components]
    B --> C[Draw Vertical Plane BM Diagram → M_v]
    B --> D[Draw Horizontal Plane BM Diagram → M_h]
    C --> E["Resultant: M = √(M_v² + M_h²)"]
    D --> E
    E --> F[Proceed with Design Using Resultant M]

Beam Selection Flowchart

flowchart TD
    A[Given: Span, Loading, Grade, Design Factor] --> B[Calculate Max Bending Moment M]
    B --> C["Allowable Stress f_b = f_y / Design Factor"]
    C --> D["Required Z = M / f_b"]
    D --> E[Select Smallest Section with Z ≥ Required]
    E --> F[Check Self-Weight]
    F --> G{Z Still Adequate?}
    G -->|Yes| H[Section Confirmed]
    G -->|No| I[Select Next Larger Section]
    I --> F

Key Terms Glossary

Shaft Design — Must-Know Formulas

• Combined shear stress: $f_s = 16 T_E / (\pi d^3)$
• Combined bending stress: $f = 32 M_E / (\pi d^3)$
• Equivalent torque: $T_E = \sqrt{T^2 + M^2}$
• Equivalent moment: $M_E = 0.5(T_E + M)$
• Design torque: $T = K_T \cdot T_S$
• Design moment: $M = K_M \cdot M_S$

Allowable Stresses (Steel Shafts)

• Bending: smaller of 40% f_y or 24% f_ult
• Shear: smaller of 30% f_y or 18% f_ult
• With keyway: multiply both by 0.75

Shock/Fatigue Factors — Quick Reference

• Static/gradual: K_T = 1.0, K_M = 1.5
• Sudden + minor shock: K_T = 1.0–1.5, K_M = 1.5–2.0
• Sudden + heavy shock: K_T = 1.5–3.0, K_M = 2.0–3.0

Shaft Load Formulas

• Weight: F = mg
• Chain/belt drive force: F = T_2 + T_1
• Chain/belt torque: T = (T_2 - T_1) · d/2
• Belt drive with factor: F = 2fT/d (f ≈ 1.5 for belt drives)
• Gear tangential force: F_t = 2T/d
• Spur gear separating force: F_s = F_t · tan θ
• Helical gear separating force: F_s = F_t · tan θ / cos α
• Helical gear axial force: F_a = F_t · tan α
• Resultant gear force: F = √(F_t² + F_s²)
• Multi-plane resultant moment: M = √(M_v² + M_h²)

Column Design — Quick Reference

• Limiting slenderness ratio: $(L_e/r)_{lim} = \sqrt{2\pi^2 E / f_y}$
• Euler critical load: $F_{cr} = \pi^2 E A / (L_e/r)^2$
• Use the smaller radius of gyration for buckling checks
• If $L_e/r$ > limiting value → slender → Euler applies

Key Design Reminders

• K_M is never less than 1.5 (even for static loads) due to bending stress reversal from rotation
• Fatigue is the most common shaft failure mode — not overload
• Soft-start motors: 1.5–2× rated torque; hard-start: 3–5× rated torque
• Shaft diameter is determined by the more critical of shear stress and bending stress — check both
• For stepped shafts: apply stress concentration factors at each step
• Hot rolled sections → not for rotating shafts; use bright steel instead

The Scene: A Conveyor That Wouldn't Stop Breaking

the practitioner had been the lead maintenance engineer at a mid-sized cement plant for six years. He was sharp, methodical, and rarely stumped. But on a Tuesday morning in March, he stood staring at a shattered shaft — the third one in eight months — scattered across the floor of the main conveyor hall.

The cost wasn't just the shaft itself. Every failure shut down the entire production line. Replacement parts had to be shipped internationally. Overtime wages for emergency crews. Angry clients waiting on delayed shipments. The plant manager was breathing down his neck.

"Just get a bigger shaft," his supervisor kept saying.

But the practitioner had a gut feeling. The problem wasn't the size of the shaft. It was everything around it — the wrong key, poor seal selection, and a complete misunderstanding of the forces at play.

If you've ever designed, maintained, or troubleshot a rotating system — this story is for you. Because the lessons the practitioner learned the hard way are the same ones that separate costly guesswork from confident, reliable engineering.

What a Shaft Actually Does (And Why Most People Get It Wrong)

Before the practitioner could fix the problem, he had to go back to basics. And that's where most engineers — beginners and veterans alike — go wrong. They treat a shaft as a simple spinning rod.

A shaft is a rotating member, supported by bearings, that transmits torque and power.

That single sentence carries more weight than most people realize. Let's break down what's actually happening inside that steel cylinder: