Joints, Fasteners and Structural Connections

Introduction

Mechanical systems are rarely manufactured as single pieces. Shafts are joined to hubs, frames are assembled from members, covers must be removable, and pressure boundaries must remain leak-free. A connection therefore has two apparently competing purposes: it must transmit forces safely while permitting the required assembly, disassembly, movement, sealing, or manufacturing process.

Connection design is not simply a matter of selecting a bolt or weld from a catalogue. The designer must identify the load path, determine how force is shared among components, evaluate static and fatigue failure modes, control deformation and leakage, and consider inspection, maintenance, corrosion, cost, and safety. This chapter develops the principles used to design and assess threaded fasteners, preloaded bolted joints, riveted and pinned connections, welded and brazed joints, adhesive bonds, and press fits.

Learning Objectives

By the end of this chapter, the student should be able to:

1. Classify mechanical joints and select an appropriate connection technology.
2. Design threaded fasteners for tensile, shear, bearing, and combined loading.
3. Explain preload, joint stiffness, torque–tension behavior, and bolt fatigue.
4. Assess rivets, pins, welds, brazed joints, adhesives, and interference fits.
5. Diagnose failure, leakage, loosening, and maintenance problems.
6. Apply standards, manufacturing constraints, inspection requirements, and safety considerations to connection design.

1. Connection Functions and Load Paths

A joint may be permanent, semi-permanent, or detachable. Welds, brazed joints, structural adhesives, and many riveted joints are normally permanent. Bolts, screws, pins, keys, splines, and press fits may be detachable or serviceable, although a press fit is often intended to remain assembled.

The first design question is: how does the applied load travel from one component to another? A tensile load may pass through a bolt in direct tension, through friction generated by bolt preload, or through bearing between a pin and a hole. A torque may be transmitted by friction, by shear in a key or pin, or by the distributed stresses of a weld or adhesive layer. A sound design makes this load path explicit.

Important connection requirements include:

• strength against yielding, fracture, fatigue, buckling, and pull-out;
• stiffness sufficient to maintain alignment and function;
• sealing against fluid or gas leakage;
• resistance to vibration-induced loosening;
• manufacturability and accessibility;
• inspection and repairability;
• compatibility with temperature, corrosion, and differential expansion.

A connection should be designed for the actual service load spectrum rather than only the largest nominal force. Shock, vibration, thermal cycling, assembly error, and accidental overload can dominate failure.

2. Threaded Fasteners

A threaded fastener consists of a threaded member, such as a bolt or screw, and usually a nut or tapped hole. Important terminology includes nominal diameter, pitch, lead, thread angle, pitch diameter, minor diameter, major diameter, proof strength, and grip length. For a single-start thread, lead equals pitch. Metric threads are commonly designated by nominal diameter and pitch, for example M12 × 1.75.

The tensile stress area, A_t, is used rather than the nominal shank area because the thread root reduces the load-carrying section. For a Unified inch thread, a commonly used approximation is

where d is nominal diameter and n is threads per unit length. For metric threads, manufacturers should be consulted for the standardized tensile-stress area; an engineering approximation is

where d and p are in consistent length units. These approximations apply to conventional 60° threads and should not replace a standard when certification is required.

The average tensile stress in a fastener is

where F_b is bolt tensile force in newtons, A_t is tensile-stress area in square metres, and σ_b is stress in pascals. Direct shear stress, using the resisting shear area A_s, is

The simple formula is appropriate for an approximately uniform shear field. Thread engagement, stress concentration, bending, and uneven load sharing may require a more detailed analysis.

For combined tensile and shear loading, a ductile fastener may be assessed using the von Mises equivalent stress:

This criterion assumes a ductile, isotropic material under a static or quasi-static stress state. A design factor n_d is applied by requiring σ_eq ≤ S_y/n_d, where S_y is yield strength.

Thread Stripping and Engagement

Failure may occur in the external or internal threads before the bolt shank yields. The required engagement length depends on material strength, thread geometry, diameter, and loading. Steel bolts in steel nuts often require an engagement length approximately equal to the nominal diameter for ordinary loading, but this is not a universal rule. Soft metals, plastics, short tapped holes, and high-strength bolts require a calculation or manufacturer specification.

Threads also introduce stress concentrations. Rolled threads generally provide better fatigue performance than cut threads because they produce favorable surface properties and avoid cutting across grain flow. The first engaged thread carries a disproportionately large portion of the load; special nuts, controlled geometry, and adequate engagement improve load distribution.

3. Preloaded Bolted Joints

A preloaded joint is tightened so that the bolt is placed in tension and the clamped members are placed in compression. External tensile service load first reduces the compression in the joint; only part of the service load immediately increases bolt tension. This is the central mechanical advantage of preload.

Let k_b be bolt stiffness and k_m be the combined stiffness of the clamped members. For an external separating load P, the increase in bolt load is

where

is the joint stiffness constant. The decrease in clamping force is

Thus, if the joint is not separated, the bolt experiences F_i + CP, where F_i is initial preload. The member compression becomes F_i − (1 − C)P.

These relations assume linear elastic behavior, concentric loading, constant stiffness, and no separation or slip. Once the members separate, the model changes because the external load is carried mainly by the bolt and contact at the joint is lost.

For a uniform bolt segment, stiffness can be approximated by

where E_b is bolt Young’s modulus, A is the effective tensile area, and L_b is the effective elastic length. For the clamped members, a similar expression k_m = E_m A_m/L_m may be used only when the stress field is approximately uniform. In real joints, the compression cone makes A_m and L_m geometry-dependent; standards and finite-element analysis may be necessary for critical designs.

Preload Selection

The target preload must be high enough to prevent separation or slip, but low enough to avoid yielding, embedment failure, gasket crushing, or thread stripping. A simplified design check is

and, for no separation,

where F_allow,b is the permitted bolt load. In a slip-resistant joint, preload also creates a friction capacity. A simplified estimate for one slip plane is

where μ is the coefficient of friction. With multiple independent slip planes, the capacity is approximately μF_i multiplied by the number of planes, subject to the governing design standard. Surface condition, coating, contamination, and bolt relaxation strongly affect μ.

Torque–Tension Behavior

A frequently used installation relationship is

where T is tightening torque, K is the nut factor, F_i is preload, and d is nominal diameter. K is not a material constant; it includes thread friction, bearing friction, lubrication, coating, and geometry. Consequently, torque control alone can produce substantial preload scatter. Torque–angle methods, tension-indicating washers, direct-tension indicators, or calibrated hydraulic and ultrasonic systems may be preferable for critical joints.

Joint Separation and Leakage

A gasketed joint must maintain sufficient compressive gasket load under operating pressure and thermal cycling. Separation can cause leakage even if the bolt has not fractured. A pressure vessel cover therefore requires analysis of bolt preload, gasket seating stress, flange flexibility, thermal expansion, and pressure-induced separating force. The simplified bolt-joint model is useful for understanding trends but should not replace pressure-vessel codes or gasket manufacturer data.

Fatigue of Bolted Joints

Fatigue resistance is improved when the joint remains clamped. If the bolt load fluctuates between F_min and F_max, the mean and alternating bolt loads are

Corresponding stresses are obtained by dividing by A_t. A Goodman-type assessment for a bolt may be written as

where σ_a and σ_m are alternating and mean stresses, S_e is corrected endurance strength, S_u is ultimate tensile strength, and n_d is design factor. This relation is an engineering approximation for high-cycle fatigue and requires corrections for surface finish, size, reliability, temperature, and stress concentration. Preload mainly raises the mean stress, while joint stiffness controls the alternating component.

Vibration can cause self-loosening when transverse motion repeatedly overcomes friction in the interface. Prevailing-torque nuts, chemical threadlockers, safety wire, lock plates, serrated washers, and positive locking devices may be used, but a locking device does not compensate for inadequate preload or poor joint design.

4. Rivets and Pins

Riveted joints are widely used in aircraft, pressure structures, and sheet-metal assemblies where permanent, inspectable connections are desirable. A rivet may fail in shear, the connected plate may tear across a row of holes, or the plate may fail in bearing. For a rivet of diameter d carrying shear force V, the nominal average shear stress is

where n_s is the number of shear planes. Single-shear and double-shear configurations must not be confused.

For a plate of thickness t with net width b_net, the average tensile stress across the remaining section is

Bearing stress between a hole and fastener is commonly estimated by

These are average stresses and are used with suitable allowable values. Edge distance, pitch, hole quality, eccentricity, and load distribution are important. Rivet holes reduce net section and may initiate fatigue cracks.

Pins are commonly used for locating, hinging, or transmitting shear. A pin should not be treated as a perfect hinge if clearance, bending, and bearing deformation are significant. Retention against axial migration, wear, fretting, and lubrication must be addressed.

5. Welded Connections

Welding joins materials by fusion, pressure, or a combination of heat and pressure. Common processes include shielded metal arc, gas metal arc, gas tungsten arc, submerged arc, laser, and resistance welding. Design must consider the weld metal, heat-affected zone, base metal, residual stress, distortion, and defects.

For a fillet weld of leg size s and length L, the effective throat is

For a uniformly loaded weld group carrying direct force P, the nominal throat shear stress is approximately

where a and L are in consistent units. For a weld group under an eccentric force or moment, direct and torsional components must be combined using the weld-group centroid and polar properties. The simple uniform formula is not sufficient for moment-loaded welds.

Weld failures include throat fracture, toe cracking, lack of fusion, porosity, undercut, lamellar tearing, brittle fracture, and fatigue cracking. Fatigue design is governed strongly by weld geometry and stress concentration; nominal base-metal endurance limits cannot simply be assigned to a weld toe. Welding procedures, preheat, interpass temperature, filler selection, qualification, nondestructive examination, and welder competence are governed by applicable codes.

6. Brazed and Soldered Joints

Brazing uses a filler metal with a liquidus temperature above approximately 450 °C but below the melting temperature of the base materials. Soldering uses a lower-temperature filler. The filler flows through a close-fitting joint by capillary action. Joint strength depends on overlap, clearance, surface preparation, filler compatibility, temperature, and cleanliness.

Brazed joints can join dissimilar metals and thin sections with relatively low distortion. Their limitations include reduced strength at elevated temperature, possible galvanic corrosion, and dependence on joint clearance. The joint should be designed so that the filler is loaded primarily in shear rather than peeled from the interface.

7. Adhesive Joints

Structural adhesives transfer load through an adhesive layer. Lap joints are efficient because they create a large bonding area, but eccentric single-lap joints produce peel and cleavage stresses. Adhesive layers are generally weak in peel and strong in shear when properly prepared.

The average shear stress in a simple lap joint is

where P is transmitted force, b is bond width, and L is overlap length. This is only a preliminary estimate because shear stress is concentrated near overlap ends and depends on adherend stiffness and adhesive shear modulus. A long overlap does not increase strength indefinitely.

Surface preparation, cure temperature, moisture, temperature range, aging, and inspection are decisive. Adhesives can provide sealing, damping, and electrical isolation, but they complicate disassembly and may have limited resistance to sustained load or high temperature. Design values should come from tested joint configurations rather than bulk adhesive data alone.

8. Press Fits and Interference Fits

A press fit is produced when the assembled shaft diameter exceeds the hole diameter by an interference. Assembly may be achieved by force, heating the hub, or cooling the shaft. The resulting contact pressure produces friction for torque and axial-load transmission.

For a simplified friction design, the transferable torque is

or

where μ is interface friction coefficient, p is mean contact pressure, d is nominal diameter, and L is fit length. The corresponding axial force capacity is

These equations assume uniform pressure and friction, neglect end effects, and apply only as preliminary estimates. Actual contact pressure depends on shaft and hub elastic properties, diameters, wall thickness, and interference. Excessive interference can yield the hub or shaft, distort roundness, damage bearings, and make assembly impossible.

Thermal expansion can alter fit pressure. For an approximately free component,

where α is the coefficient of linear thermal expansion and ∆T is temperature change. Differential expansion between hub and shaft must be included during operation.

9. Design Selection and Failure Assessment

Selection should begin with the function rather than the familiar hardware. The following comparisons are useful:

A complete failure assessment should consider:

1. Static strength: yielding, fracture, shear, bearing, pull-out, and thread stripping.
2. Fatigue: stress range, mean stress, surface condition, stress concentration, and number of cycles.
3. Stiffness: joint opening, slip, alignment, vibration, and load redistribution.
4. Environment: corrosion, humidity, temperature, chemical attack, and galvanic couples.
5. Manufacturing: tolerances, weld distortion, hole quality, cure control, and accessibility.
6. Inspection: visual, torque verification, dimensional checks, dye penetrant, magnetic-particle, ultrasonic, or radiographic methods as appropriate.
7. Maintenance: replacement intervals, access, torque procedures, and evidence of fretting or loosening.

Codes and standards are legally or contractually important. Examples include ISO and ASME fastener standards, structural welding codes, pressure-vessel codes, and aerospace or automotive specifications. The governing standard must be identified before final design; generic textbook equations do not override code requirements.

Worked Example: Preloaded Bolted Joint

A steel joint uses a bolt with tensile-stress area A_t = 84.3 mm². The initial preload is F_i = 18 kN. The external separating load is P = 10 kN, and the joint constant is C = 0.20. The allowable bolt load is 30 kN. Determine the bolt load increase, final bolt load, remaining clamping force, and bolt tensile stress.

The bolt load increase is

The final bolt load is

The reduction in clamping force is

Therefore, the remaining clamping force is 18 − 8 = 10 kN, so the joint has not separated. The bolt stress is

The final bolt load is below the 30 kN allowable load. The calculation is valid only while the joint remains elastic and closed; it does not assess thread stripping, fatigue, gasket behavior, or installation scatter.

Applied Case: Illustrative Pump-Casing Joint

Background

An industrial centrifugal pump has a bolted casing split. Operators report fluid leakage after several months of vibration. The bolts have not fractured, but inspection shows fretting marks near the flange and inconsistent tightening torque.

Problem Analysis

The original design used a nominal torque value without specifying lubrication or a tightening sequence. Because friction varied, actual preload varied widely. Some bolts had insufficient preload, allowing transverse micro-slip. Others were highly tightened and approached their allowable load. Thermal cycling and pressure pulses periodically reduced contact force, opening microscopic leakage paths through the gasket.

Approach and Solution

The engineering team first measured bolt material, tensile-stress area, flange stiffness, gasket seating requirements, and operating pressure. A joint model was used to estimate the stiffness constant and the minimum preload needed to prevent separation. The bolts were replaced with certified fasteners, and the joint was redesigned with a controlled cross-pattern tightening sequence. Lubricant and its friction coefficient were specified. Calibrated tension-control equipment was introduced, and flange flatness and surface finish were inspected. The gasket was selected using the applicable pressure-equipment standard, and a retightening policy was prohibited unless supported by the gasket manufacturer.

Outcome

After installation, preload scatter was reduced, flange slip marks disappeared, and leakage did not recur during the monitored operating period. The result came not from simply increasing torque, but from controlling the complete load path, joint stiffness, gasket compression, and installation process.

Learning Points

• Torque is an indirect and friction-sensitive measure of preload.
• A bolt can survive while the joint still fails by separation or leakage.
• Maintenance procedures are part of the engineering design.
• Inspection and traceability are essential for safety-critical connections.

Contemporary Practice, Ethics, and Safety

Modern design increasingly uses digital torque records, electronic fastener databases, finite-element weld-group analysis, and digital inspection records. Sensors and machine-learning systems may detect abnormal vibration or preload loss, but they should support—not replace—validated engineering judgment and physical inspection. Training data can be incomplete, and an apparently normal signal does not prove structural integrity.

Ethically, designers must not conceal uncertainty in friction, material properties, fatigue life, or inspection reliability. Safety factors should reflect consequences of failure, not merely minimum cost. Traceability of material certificates, weld procedures, inspection results, and torque records is particularly important in lifting equipment, pressure systems, transportation, and medical machinery.

Sustainable design considers disassembly, repair, recyclability, and avoidance of unnecessary mixed-material joints. Adhesives and permanent welds may reduce part count but complicate end-of-life separation. Corrosion prevention, replaceable fasteners, and standardized hardware can reduce lifecycle impact.

Chapter Summary

Mechanical connections transmit load, maintain stiffness, control leakage, and determine whether a machine can be assembled and maintained safely. Threaded fasteners must be checked for tensile, shear, combined stress, thread stripping, and fatigue. In preloaded joints, the joint stiffness constant determines how external load is divided between bolt-load increase and loss of member compression. Adequate preload prevents separation and slip, but torque is an uncertain proxy because friction varies.

Rivets and pins require checks for shear, net-section tension, bearing, bending, and wear. Welds require attention to effective throat, weld-group loading, heat-affected zones, defects, residual stress, and fatigue. Brazed joints depend on capillary flow, clearance, cleanliness, and temperature. Adhesive joints provide distributed load transfer and sealing but are sensitive to peel, surface preparation, aging, and temperature. Press fits transmit torque and axial force through friction generated by interference, while excessive interference can cause yielding or assembly damage.

Good connection design integrates mechanics, materials, manufacturing, inspection, maintenance, standards, environmental conditions, and human factors. The governing design is the complete joint system, not an isolated bolt, weld, or adhesive specification.

Key Terms

Preload: Initial tensile force in a fastener before service loading.

Joint stiffness constant: Fraction of an external separating load that increases bolt load.

Tensile-stress area: Effective threaded area used to calculate bolt tensile stress.

Proof strength: Stress a fastener can withstand under a specified proof test without unacceptable permanent deformation.

Thread stripping: Shear failure of internal or external thread engagement.

Joint separation: Loss of contact compression between clamped members.

Slip-resistant joint: Joint designed to transmit shear through friction rather than bearing.

Fatigue: Progressive failure caused by repeated or fluctuating stress.

Rivet: Permanent fastener plastically formed to clamp or join components.

Bearing stress: Contact stress between a fastener and the wall of a hole.

Effective throat: Minimum design thickness of a fillet weld.

Heat-affected zone: Base-metal region altered thermally by welding.

Brazing: Joining with a filler metal melting above approximately 450 °C but below the base-metal melting temperature.

Adhesive joint: Connection in which a bonding material transfers load between adherends.

Peel stress: Stress tending to separate an adhesive joint normal to its bond plane.

Press fit: Assembly using interference between mating cylindrical surfaces.

Interference: Positive difference between shaft and hole dimensions before assembly.

Fretting: Surface damage caused by small-amplitude oscillatory motion under contact pressure.

Joint efficiency: Useful load capacity of a connection relative to the connected material or reference section.

Review Questions

Short Answer Questions

1. Why is tensile-stress area used instead of nominal bolt area?
2. What physical meaning does the joint stiffness constant have?
3. State two common failure modes of a riveted connection.
4. Why are adhesive lap joints vulnerable to peel near their ends?
5. What is the purpose of preload in a bolted joint?

Long Answer Questions

1. Explain the elastic model of a preloaded bolted joint and derive the expressions for bolt-load increase and loss of clamping force.
2. Compare bolted, welded, brazed, adhesive, and press-fit connections for a machine casing that requires strength, sealing, and periodic maintenance.
3. Discuss the major static and fatigue failure modes of a fillet-welded connection.
4. Explain why torque-controlled tightening produces variable preload and describe methods for improving preload control.
5. Develop a systematic procedure for selecting and validating a connection in a safety-critical machine.

Analytical / Application Questions

1. A joint has F_i = 24 kN, P = 15 kN, and C = 0.25. Determine the final bolt load and remaining clamping force. State whether separation occurs under the simplified elastic model.
2. A single-lap adhesive joint carries 6 kN over a width of 40 mm and overlap length of 60 mm. Calculate the average adhesive shear stress and explain why the actual peak stress will not equal this average.
3. A press fit must transmit a specified torque under uncertain friction and temperature variation. Identify the data required, calculate a preliminary fit capacity using the friction model, and recommend checks needed before production approval.

Multiple Choice Questions with Answers

Question 1

In a preloaded bolted joint, what fraction of an external separating load P increases bolt load in the linear elastic model?

A. (1 − C)P

B. CP

C. P/C

D. F_iP

Correct Answer: B. CP

Question 2

Which area is normally used to calculate tensile stress in a threaded fastener?

A. Nominal shank area only

B. Washer bearing area

C. Tensile-stress area

D. Thread pitch area

Correct Answer: C. Tensile-stress area

Question 3

A bolt joint has C = 0.20 and an external separating load of 12 kN. What is the increase in bolt load?

A. 2.4 kN

B. 9.6 kN

C. 12.2 kN

D. 60 kN

Correct Answer: A. 2.4 kN

Question 4

Which condition most directly promotes vibration-induced self-loosening?

A. Adequate preload with no transverse motion

B. Transverse micro-slip at the joint interface

C. A large bolt tensile-stress area

D. A polished bolt head

Correct Answer: B. Transverse micro-slip at the joint interface

Question 5

For a fillet weld with leg size s, the idealized effective throat is approximately:

A. s/2

B. s

C. 0.707s

D. 1.414s

Correct Answer: C. 0.707s

Question 6

Which failure mode is especially important in a riveted sheet connection?

A. Plate net-section tension

B. Adhesive peel

C. Hub yielding from interference

D. Gasket extrusion only

Correct Answer: A. Plate net-section tension

Question 7

A simplified slip capacity for one friction interface in a preloaded joint is:

A. μF_i

B. F_i/μ

C. μ/P

D. P + F_i

Correct Answer: A. μF_i

Question 8

Why does torque not uniquely determine bolt preload?

A. Bolt diameter has no effect on torque

B. Friction varies in threads and under the bearing surface

C. Preload is independent of tightening

D. Torque can act only in shear

Correct Answer: B. Friction varies in threads and under the bearing surface

Question 9

In a simple adhesive lap-joint estimate, average shear stress equals transmitted force divided by:

A. Bolt tensile area

B. Weld throat area

C. Bond width multiplied by overlap length

D. Interference diameter

Correct Answer: C. Bond width multiplied by overlap length

Question 10

What is the primary mechanism by which a press fit transmits torque?

A. Magnetic attraction

B. Friction generated by contact pressure

C. Thread engagement

D. Adhesive curing

Correct Answer: B. Friction generated by contact pressure