Design of Mechanical Members and Connections

Introduction and Context

Mechanical design converts loads, motion, environmental conditions, manufacturing capabilities, and safety requirements into a physical component. In Strength of Materials, design is not merely the selection of a dimension that prevents immediate yielding. A satisfactory design must also control deformation, fatigue damage, instability, wear, leakage, vibration, corrosion, cost, and manufacturability.

The principal mechanical members considered in this chapter are shafts, beams, pressure vessels, and springs. Connections—bolted, welded, and bonded—are equally important because many failures originate at joints rather than in the connected members. The design process therefore integrates strength and stiffness requirements with material selection, geometric proportioning, inspection, maintenance, and lifecycle considerations.

A design is acceptable only when the calculated performance remains within specified limits under all relevant load cases. Typical checks include:

• static strength and yielding;
• fracture and fatigue resistance;
• deflection, twist, and alignment;
• buckling or instability;
• local stresses and stress concentrations;
• thermal, corrosive, and environmental effects;
• manufacturability, inspection, repair, and cost.

Learning Objectives

After studying this chapter, a student should be able to:

1. distinguish strength-controlled from stiffness-controlled design;
2. select and apply appropriate equations for shafts, beams, vessels, and springs;
3. design and assess bolted, welded, and bonded joints;
4. compare materials and geometries using safety, cost, manufacturability, and lifecycle criteria;
5. recognize the limitations of elementary formulas and identify when codes, testing, or numerical analysis are required.

1. The Design Basis

1.1 Loads and load cases

Loads may be steady, fluctuating, impact, thermal, residual, or accidental. A design load case combines simultaneous actions that can occur realistically. For example, a rotating shaft may experience torque, bending from a belt, axial thrust, and a transient starting torque. The designer should establish:

• magnitude and direction of each load;
• static or cyclic character;
• number of cycles and design life;
• temperature and environment;
• consequences of failure;
• applicable design code and inspection class.

Load factors and material factors are introduced through either a factor-of-safety approach or a limit-state design approach. In the simple allowable-stress method,

where σ_allow is allowable stress, σ_lim is the relevant limiting material stress (such as yield or ultimate strength), and n is the factor of safety. This relation is appropriate only when the selected limiting stress, load definition, material condition, and failure mode are consistent. For brittle fracture, fatigue, buckling, or pressure-vessel design, a different criterion may govern.

1.2 Strength versus stiffness

Strength concerns prevention of yielding, fracture, excessive permanent deformation, or other ultimate failure. Stiffness concerns displacement, rotation, twist, vibration, or loss of functional alignment. Increasing diameter may improve both strength and stiffness, but not in the same proportion. For a solid circular shaft, torsional strength varies approximately with d³, whereas torsional rigidity varies with d⁴. Consequently, a shaft that is safe against torsional yielding may still be unacceptable because of excessive twist.

1.3 Stress concentration and fatigue

A shoulder, keyway, thread, hole, weld toe, or abrupt change in section disturbs the nominal stress field. The elastic stress-concentration factor is

where σ_max is the local elastic peak stress and σ_nom is the nominal stress based on the unnotched geometry. In fatigue, the effective fatigue concentration is often lower than K_t because of material notch sensitivity; it is represented by a fatigue notch factor K_f. Surface finish, size, residual stress, corrosion, and manufacturing defects must also be considered.

A useful design sequence is: define requirements, identify load cases, select candidate materials and geometry, calculate nominal stresses and deformations, apply concentration and environmental corrections, compare with allowable limits, then review manufacture, inspection, cost, and lifecycle performance.

2. Shafts and Axles

A shaft transmits torque and commonly carries bending loads; an axle primarily supports rotating elements and may transmit little torque. Shafts are therefore commonly checked for combined bending and torsion.

For a solid circular shaft under torque T, the maximum torsional shear stress is

and the angle of twist over length L is

Here d is shaft diameter, J = πd⁴/32 is the polar second moment of area, G is shear modulus, and θ is in radians. These equations assume a prismatic, circular, linearly elastic shaft with uniform torque and Saint-Venant torsion; they are not directly valid near keyways, splines, sharp shoulders, or severe plastic deformation.

For a solid circular shaft under bending moment M,

where σ_b is the outer-fibre bending stress. A preliminary maximum-normal-stress design may combine bending and axial stress algebraically at the critical surface. For ductile materials under combined bending and torsion, the distortion-energy (von Mises) criterion gives

where σ is normal stress and τ is shear stress. The requirement is σ_vm ≤ σ_y/n, with σ_y the yield strength. This criterion is appropriate for ductile, relatively homogeneous materials under static or slowly varying loading. Fatigue requires a separate alternating/mean-stress assessment.

For a shaft with bending moment M and torque T and no other normal stress, an equivalent-torque form based on von Mises stress is

and d may be found from 16T_e/(πd³) ≤ σ_y/n. An equivalent-bending form can also be developed, but it must not be mixed inconsistently with a different failure criterion.

Worked example: preliminary shaft sizing

A solid steel shaft carries M = 600 N·m and T = 900 N·m. Let σ_y = 300 MPa and n = 2. Ignore stress concentrations and fatigue for this preliminary calculation.

The allowable von Mises stress is 150 MPa. The equivalent torque is

T_e = √(600² + 3×900²) = 1,652 N·m

Thus,

d³ ≥ 16(1,652,000)/(π×150) = 56,042 mm³

and d ≥ 38.3 mm. A practical design would select a larger standard diameter, perhaps 40 or 45 mm, then recheck keyway weakening, shoulders, bearing fits, fatigue, deflection, critical speed, and manufacturing tolerances. The numerical result is not a final design because the stated assumptions exclude important real effects.

Shaft design also requires checks for bearing pressure at fits, key shear and crushing, spline strength, coupling misalignment, and torsional vibration. A hollow shaft can provide lower mass for a specified torsional rigidity, although its manufacture, connection details, and local buckling resistance must be considered.

3. Beams and Structural Members

For a prismatic beam in elastic bending,

where M is bending moment, y is distance from the neutral axis, and I is the second moment of area. The maximum stress occurs at the extreme fibre, y = c. For a rectangular section of width b and height h, I = bh³/12; for a circular section of diameter d, I = πd⁴/64.

The elastic curve is governed by

where E is Young’s modulus and v is transverse deflection. Boundary conditions determine the integration constants. For example, a simply supported beam with a central point load P has maximum deflection

where L is span. The equation assumes small deflection, linear elasticity, a prismatic beam, ideal supports, and a load applied in the beam’s plane of symmetry.

A beam design must check bending stress, shear stress, deflection, local crushing, lateral-torsional buckling, fatigue, and connection forces. A deeper rectangular beam is efficient in bending because I increases with h³, but excessive depth may cause instability, packaging problems, or manufacturing difficulty. In thin-walled sections, shear deformation and local plate buckling may invalidate elementary beam theory.

4. Pressure Vessels

Pressure vessels store energy and can fail catastrophically; applicable codes and qualified inspection are essential. For a thin-walled cylindrical vessel with internal pressure p, internal radius r, and wall thickness t, where t is small relative to r,

is the hoop stress and

is the longitudinal stress. The hoop stress is twice the longitudinal stress and usually governs. These formulas assume a closed, thin, circular cylinder with uniform pressure and negligible local discontinuities. A common thin-wall criterion is t/r ≤ approximately 0.1, but the applicable code may impose a stricter limit.

For a spherical vessel,

A sphere is structurally efficient because membrane stress is uniform and lower than the cylindrical hoop stress for the same p, r, and t. Real vessels require allowances for corrosion, fabrication tolerance, weld efficiency, openings, nozzles, supports, thermal gradients, cyclic pressure, and brittle fracture.

Design example: thin cylindrical vessel

A vessel has internal pressure p = 2 MPa, radius r = 500 mm, allowable membrane stress S = 120 MPa, and weld efficiency η = 0.85. Using the simplified hoop relation with efficiency included,

t = pr/(Sη) = (2×500)/(120×0.85) = 9.80 mm

A code-compliant design would add corrosion and manufacturing allowances and use the code equation rather than adopting 9.80 mm directly. A selected nominal thickness might therefore be 12 mm or greater, depending on allowances and code requirements. Hydrostatic testing, pressure relief, nondestructive examination, and documented material traceability are part of safe design.

5. Springs

Springs store energy and provide force, compliance, positioning, or vibration isolation. Helical compression springs are commonly designed using wire diameter d, mean coil diameter D, active turns N, shear modulus G, and axial load F.

The spring index is C = D/d. The torsional shear stress is corrected for curvature and direct shear by the Wahl factor,

and

The deflection and spring rate are

These relations assume close-coiled helical geometry, small pitch angle, linear elastic wire, and uniform active coils. End coils, buckling, solid height, surge, fatigue, and residual stresses must be checked. Spring design is often fatigue-controlled because springs undergo millions of cycles. Shot peening, suitable surface finish, corrosion protection, and presetting can improve performance.

6. Bolted Connections

Bolted joints should be designed as systems, not merely for bolt tensile strength. A preloaded joint clamps the members; external tensile load first reduces member compression and increases bolt tension. Joint stiffness, preload scatter, embedment, surface roughness, vibration, and thermal expansion affect performance.

For a bolt subjected to direct tensile load P, the nominal tensile stress is

where A_t is the tensile stress area, not the gross shank area for a threaded bolt. For direct shear V,

where A_s is the effective shear area. Combined loading may be assessed with a suitable interaction or von Mises criterion, subject to the design standard.

A preloaded joint should be tightened using a controlled method. Torque control is convenient but has substantial scatter because most tightening torque is lost to friction. Direct tension indicators, turn-of-nut methods, hydraulic tensioning, or calibrated torque-angle methods may provide better control. Threads, bearing surfaces, joint separation, prying action, fatigue, and slip must be considered. Locknuts, prevailing-torque nuts, thread-locking compounds, or mechanical locking devices should not be treated as substitutes for correct preload and joint design.

7. Welded Connections

Welds transfer force through a fusion zone and heat-affected zone, both of which may differ from the parent material. Important failure modes include weld-metal yielding, throat fracture, lack of fusion, porosity, undercut, fatigue cracking at the weld toe, and distortion.

For a fillet weld of leg size s, the effective throat is approximately

For a uniform direct load F distributed over effective throat area A_w,

The exact resistance calculation depends on weld group geometry, eccentricity, loading direction, electrode classification, workmanship, and the governing welding code. A weld is not automatically as strong as the connected plate; weld quality and joint accessibility are design variables.

Good practice includes specifying weld size, length, process, filler classification, position, preheat, post-weld treatment where needed, acceptance criteria, and inspection method. Welding can reduce component count and produce efficient load paths, but it may introduce residual stress and distortion. Fatigue design should focus on weld detail category and stress range rather than relying only on nominal static strength.

8. Bonded Joints

Adhesive joints transfer load through a comparatively large area and can join dissimilar materials while sealing against moisture. Lap joints are prone to nonuniform shear and peel stress, especially near the ends. Joint length, overlap, adhesive thickness, surface preparation, cure cycle, temperature, moisture, and creep govern performance.

A simple average shear stress is

where P is load, b is bond width, and l is overlap length. This is only a preliminary estimate because edge concentrations and peel stresses may dominate. Adhesives are often viscoelastic; allowable stress depends on time and temperature. Surface cleaning, abrasion or chemical treatment, primer, controlled bond-line thickness, and quality verification are essential.

Bonded joints may reduce stress concentration compared with a bolted joint, but inspection of hidden defects is difficult and long-term environmental degradation may be significant. Hybrid bolted-bonded joints can provide fail-safe load transfer, but load sharing and installation sequence must be analyzed.

9. Material, Geometry, and Lifecycle Selection

Material selection should begin with required properties rather than habit. Relevant properties include yield strength, ultimate strength, fatigue strength, fracture toughness, elastic modulus, shear modulus, density, corrosion resistance, thermal stability, weldability, machinability, availability, and recyclability.

Geometry is often more influential than material. Increasing section depth, using a hollow section, adding a fillet, or moving material away from a neutral axis can improve performance without increasing mass substantially. However, optimization must include buckling, local stresses, joining, tooling, inspection, and tolerance control.

A lifecycle comparison should include initial material and manufacturing cost, energy consumption, inspection, downtime, replacement, corrosion protection, disposal, and consequences of failure. A low-cost component that requires frequent replacement may have a higher total cost than a durable alternative.

10. Integrated Design Case

Background

Consider an illustrative design for a belt-driven pump. A steel shaft transmits power from an electric motor, supports a pulley and impeller, and is connected through a bolted coupling. The pump operates continuously in a humid environment.

Problem

The shaft must resist combined bending and torque, maintain acceptable angular twist, and survive cyclic operation. The coupling bolts must resist fluctuating shear and tension. Corrosion and maintenance access influence the final design.

Analysis

The designer first develops torque from power and speed:

where P is transmitted power and ω is angular velocity in radians per second. Belt forces produce a bending moment, and the impeller produces an axial or hydraulic thrust. A von Mises static check is followed by fatigue analysis using alternating and mean stresses. The keyway is included through a local concentration factor, and the selected diameter is checked against torsional rigidity and critical speed.

The coupling is assessed for bolt tensile area, shear, bearing, preload, and slip. Because vibration may loosen fasteners, the joint uses controlled preload and a positive locking method. A protective coating is selected for the exposed shaft, while bearing fits remain appropriately machined and protected.

Approach and solution

An initial 38 mm calculated shaft diameter is increased to a standard 45 mm diameter after considering keyway fatigue, corrosion allowance, and stiffness. The design uses a generous shoulder radius, a relieved keyway transition, and a replaceable coupling hub. Bolts are selected from a recognized standard and tightened using a documented procedure. Shaft alignment tolerances and inspection intervals are specified.

Outcome

The integrated design is not the lightest possible design, but it provides improved fatigue margin, lower twist, standard procurement, easier maintenance, and reduced risk of unplanned shutdown. The decision demonstrates that a mechanically adequate design is not necessarily the best engineering design unless reliability and lifecycle effects are included.

Learning points

1. Nominal stress calculations are only the first stage.
2. Stiffness and fatigue can govern after static strength is satisfied.
3. Joints require independent analysis and controlled manufacture.
4. Standard sizes and inspectable details often reduce lifecycle cost.
5. Environmental protection and maintenance planning are design requirements, not afterthoughts.

11. Limitations, Codes, and Professional Practice

Elementary formulas assume idealized geometry and loading. Finite-element analysis may be required for nozzles, complex weld groups, contact regions, composite joints, and strong stress gradients. Analysis should be verified by hand calculations, mesh-convergence studies, material data, and testing where consequences are high.

Pressure vessels, structural welds, lifting equipment, rotating machinery, and fasteners are governed by national and international standards. The designer must identify the applicable jurisdiction, edition, material specification, welding qualification, inspection requirement, and documentation rules. Safety factors cannot compensate for an inappropriate failure model or poor quality control.

Ethically, engineers must communicate uncertainty, avoid overstating calculated precision, preserve traceability, and report unsafe conditions. Artificial intelligence can assist with literature search, parameter studies, drawing checks, and anomaly detection, but it cannot replace engineering judgment, validated models, code compliance, or accountable approval. Generated calculations must be independently checked.

12. Best-Practice Design Workflow

1. State function, life, environment, and failure consequences.
2. Define load cases, combinations, tolerances, and applicable codes.
3. Select plausible materials and manufacturing processes.
4. Size members for strength, then check stiffness, fatigue, stability, and wear.
5. Analyze stress concentrations and all connections.
6. Review assembly, inspection, maintenance, repair, and fail-safe behavior.
7. Compare alternatives using mass, cost, lead time, energy, and lifecycle risk.
8. Prototype or test when models are uncertain or failure consequences are severe.
9. Document assumptions, calculations, revisions, and acceptance criteria.

Chapter Summary

Mechanical member design integrates strength, stiffness, stability, fatigue, joining, materials, manufacturing, and lifecycle performance. Shafts are commonly governed by combined bending, torsion, fatigue, and twist. Beams require bending, shear, deflection, and stability checks. Pressure vessels demand membrane-stress analysis supplemented by code rules, corrosion allowances, inspection, and pressure testing. Springs require consideration of corrected shear stress, deflection, solid height, buckling, and fatigue. Bolted joints depend on preload and joint stiffness; welded joints depend on weld geometry, defects, residual stress, and fatigue detail; bonded joints depend strongly on surface preparation, temperature, creep, and environmental durability.

The most reliable designs use realistic load cases, appropriate failure criteria, standard materials and geometries, controlled manufacturing, documented inspection, and independent verification. The final choice should optimize safety and function over the complete lifecycle rather than minimizing initial material cost alone.

Key Terms

Factor of safety: Ratio of a limiting capacity to the design demand.

Allowable stress: Permitted stress after applying the selected safety factors.

Stiffness: Resistance to elastic deformation.

Stress concentration: Local increase in stress caused by a geometric or material discontinuity.

Fatigue: Progressive damage caused by repeated or fluctuating loading.

Von Mises stress: Distortion-energy measure used commonly for ductile-material yielding.

Polar second moment of area: Geometric property governing circular-section torsional rigidity.

Bending stress: Normal stress produced by a bending moment.

Pressure vessel: Enclosed component designed to contain pressure.

Hoop stress: Circumferential membrane stress in a cylindrical vessel.

Spring index: Ratio of mean coil diameter to wire diameter.

Preload: Initial clamping force introduced into a bolted joint.

Joint stiffness: Resistance of connected members and fastener to relative deformation.

Fillet weld: Weld joining surfaces approximately at right angles.

Effective throat: Shortest effective dimension of a fillet weld used in strength calculations.

Adhesive joint: Connection formed by a bonding material between adherends.

Lifecycle cost: Total cost of acquisition, operation, maintenance, and disposal.

Fail-safe design: Design that limits consequences or retains function after a credible failure.

Review Questions

Short Answer Questions

1. Why can stiffness govern the design of a shaft even when its torsional stress is below allowable stress?
2. State the assumptions behind the thin-wall cylindrical pressure-vessel equations.
3. What is the purpose of the Wahl factor in spring design?
4. Explain why preload is important in a bolted joint.
5. Identify two environmental factors that influence bonded-joint durability.

Long Answer Questions

1. Explain a complete design procedure for a shaft subjected to bending, torsion, fatigue, and keyway stress concentration.
2. Compare bolted, welded, and bonded joints in terms of load transfer, inspection, manufacturing, repair, and lifecycle performance.
3. Derive the relationship between shaft diameter and torsional rigidity for a solid circular shaft, and discuss the effect of diameter.
4. Discuss the assumptions and limitations of elementary beam deflection theory.
5. Explain how material selection and geometry should be integrated with cost, manufacturability, and sustainability.

Analytical / Application Questions

1. A solid shaft carries M = 800 N·m and T = 1,000 N·m. If the allowable von Mises stress is 160 MPa, calculate the minimum diameter using the equivalent-torque method, ignoring stress concentrations and fatigue.
2. A simply supported beam has a central load and a specified maximum deflection. Describe how changing material modulus, beam depth, and span affects the required design.
3. Select a joint type for a removable, vibration-prone connection exposed to moisture. Justify the selection and identify the calculations, inspection, and maintenance controls required.

Multiple Choice Questions with Answers

Question 1

For a solid circular shaft under pure torque T, which expression gives the maximum torsional shear stress?

A. 16T/(πd³)

B. 32T/(πd³)

C. 16T/(πd²)

D. 32T/(πd⁴)

Correct Answer: A. 16T/(πd³)

Question 2

For a ductile material under combined normal stress σ and shear stress τ, the von Mises stress is:

A. σ + τ

B. √(σ² + 3τ²)

C. √(σ² + τ²)

D. σ² + 3τ²

Correct Answer: B. √(σ² + 3τ²)

Question 3

In a thin cylindrical pressure vessel, the hoop stress is:

A. pr/(2t)

B. 2pr/t

C. pr/t

D. pt/r

Correct Answer: C. pr/t

Question 4

For a simply supported beam with a central point load, the maximum elastic deflection is:

A. PL³/(48EI)

B. PL³/(24EI)

C. PL²/(48EI)

D. 48EI/(PL³)

Correct Answer: A. PL³/(48EI)

Question 5

The spring index C of a helical spring is defined as:

A. wire diameter divided by active turns

B. mean coil diameter divided by wire diameter

C. spring length divided by wire diameter

D. outside diameter divided by pitch

Correct Answer: B. mean coil diameter divided by wire diameter

Question 6

Why is the tensile stress area used for a threaded bolt?

A. It represents the effective reduced area at the threads.

B. It is always larger than the gross shank area.

C. It includes the nut bearing area only.

D. It applies only to bolts loaded in shear.

Correct Answer: A. It represents the effective reduced area at the threads.

Question 7

The effective throat of an equal-leg fillet weld is approximately:

A. 0.5s

B. 0.577s

C. 0.707s

D. 1.414s

Correct Answer: C. 0.707s

Question 8

Which factor is especially important for long-term adhesive-joint performance?

A. Only the nominal bolt diameter

B. Adhesive creep and environmental exposure

C. The polar moment of the shaft

D. The spring index alone

Correct Answer: B. Adhesive creep and environmental exposure

Question 9

A shaft has M = 600 N·m and T = 800 N·m. Using T_e = √(M² + 3T²), the equivalent torque is approximately:

A. 600 N·m

B. 1,000 N·m

C. 1,526 N·m

D. 2,200 N·m

Correct Answer: C. 1,526 N·m

Question 10

Which design action most directly improves fatigue resistance at a stepped shaft shoulder?

A. Using a sharper corner

B. Increasing stress concentration

C. Providing a generous fillet radius and suitable surface finish

D. Ignoring cyclic loading

Correct Answer: C. Providing a generous fillet radius and suitable surface finish