MEC-280 · Lesson 9 of 12 · 25 min

Mechanical fastening, brazing, soldering and adhesives

The KG-40 has only one welded subassembly. Everything else is held together by other means: the housing halves by twelve M10 bolts, the gears on their shafts by splines and an interference fit, the bearing caps by dowels and screws, the cooler lines by brazed fittings, the sensor board by solder, the cover gasket by a cured-in-place sealant and the nameplate by rivets. Each method suits a different need. Joints that must come apart for service are bolted; joints that must seal are sealed; joints between dissimilar or heat-sensitive materials are brazed, soldered or bonded. This lesson covers how these joining methods work and how to make them reliably, with the bolted housing joint as the main example.

Choosing a joining method

Joining choices turn on a few questions. Must the joint be taken apart, for maintenance or recycling? Bolts, screws, pins and snap fits allow that; welds, rivets, brazes and adhesives generally do not. What loads does it carry: static or cyclic, shear or tension, with or without vibration? Are the materials similar or dissimilar, and can they tolerate heat? Must the joint seal against oil, water or gas? Is it a high-volume joint that must be fast and automated, or a few joints made by hand? How will it be inspected? Often several methods combine: a bonded and riveted aircraft joint, a bolted flange with a gasket, a press fit backed up by a key.

Threaded fasteners

Bolts, screws and studs are the most common removable fasteners. A bolt passes through clearance holes and is tightened with a nut; a cap screw threads into a tapped hole in one of the parts; a stud is threaded into one part and carries a nut. Metric fasteners are specified by thread size and pitch, such as M10 × 1.5, and by property class, such as 8.8 or 10.9; the first number is one hundredth of the tensile strength in MPa and the second is ten times the ratio of yield to tensile strength. Inch fasteners use SAE grades, such as Grade 5 and Grade 8, with threads designated by diameter and threads per inch, such as 3/8-16 UNC.

The key idea of bolted joints is preload. A properly tightened bolt is stretched like a stiff spring, pulling the clamped parts together with a force much larger than the external load. When an external load tries to separate the joint, most of it simply reduces the clamping pressure between the parts; only a fraction, set by the relative stiffness of bolt and joint, adds to the bolt tension. A well-preloaded bolt therefore sees small stress fluctuations under cyclic load and resists fatigue and loosening. An under-tightened bolt lets the joint separate or slip, and then the bolt carries the whole fluctuating load and soon fails.

\[F_i \approx 0.75\, S_p\, A_t \qquad T = K\, F_i\, d\]

For reusable joints, a common target preload is 75 percent of the bolt's proof load, the proof strength \(S_p\) times the tensile stress area \(A_t\). Preload is usually set by tightening torque, using the short-form relation above with the nominal diameter \(d\) and a nut factor \(K\) that lumps together thread and underhead friction. For plain or lightly oiled steel, \(K\) is about 0.2; for zinc-plated or waxed fasteners it can be 0.15 or lower; for dry, rusty threads it can be 0.3. Because friction varies so much, torque control gives a preload scatter of about ±25 percent. Torque-angle tightening, ultrasonic bolt measurement or direct tension indicators give tighter control on critical joints such as cylinder heads and structural connections.

Worked problem

Tightening the housing bolts

The KG-40 housing halves are joined by twelve M10 × 1.5 class 8.8 bolts with zinc flake coating (\(K \approx 0.16\)). For M10 coarse thread \(A_t = 58.0\ \text{mm}^2\) and class 8.8 proof strength is 580 MPa. Find the target preload, the tightening torque and the total clamping force on the joint.

Show solution
  1. Proof load \(= 580(58.0) = 33.6\ \text{kN}\); preload \(F_i = 0.75(33.6) = 25.2\ \text{kN}\).
  2. Torque \(T = 0.16(25\,200)(0.010) = 40.4\ \text{N·m}\).
  3. Total clamp \(= 12(25.2) = 303\ \text{kN}\).
  4. With ±25 percent scatter from torque control, individual preloads range from about 19 to 31 kN.

Answer: The assembly specification calls for 40 N·m with a calibrated nutrunner, tightened in a crisscross sequence in two stages. Using the dry-steel nut factor of 0.2 with these coated bolts would overtighten them by 25 percent.

Joint design matters as much as the bolt. Stiff, flat clamped parts, long bolts with a thin shank, hardened washers under nuts on soft materials, adequate thread engagement (at least one diameter in steel, about two in aluminum or cast iron) and rolled rather than cut threads all improve a joint. Loosening under vibration happens when the joint slips sideways; enough preload to prevent slip is the real cure, with thread-locking adhesive, prevailing-torque nuts or wedge-locking washers as additional measures.

Rivets, pins, press fits and integral fasteners

Rivets are permanent fasteners deformed in place. Solid rivets are upset with a hammer or squeeze tool, as on aircraft skins and older steel bridges; blind rivets, set from one side by pulling a mandrel, are used where only one side is accessible, as on the KG-40 nameplate. Self-piercing rivets join sheets, including aluminum to steel, without pre-drilled holes and are common in car bodies. Clinching presses sheets together into an interlocking button without any fastener at all.

Pins and dowels locate parts precisely: the KG-40 bearing caps carry two dowels each so that the bores, machined with the caps in place, line up again after disassembly. Press fits and shrink fits join a hub to a shaft by making the hole slightly smaller than the shaft, using the interference fits we met in MEC-260; heating the hub or cooling the shaft eases assembly. The pressure at the interface lets the joint transmit torque and axial force by friction. Integral fasteners, such as snap fits moulded into plastic parts, seams folded in sheet metal and crimped connections, use the part's own shape to hold it together.

Brazing and soldering

Brazing and soldering join metals with a filler that melts at a lower temperature than the base metals, which do not melt. The molten filler is drawn into a narrow joint gap by capillary action, wets the surfaces and solidifies into a strong metallurgical bond. Brazing uses fillers melting above about 450 °C, such as copper, copper-phosphorus, silver and aluminum-silicon alloys; soldering uses fillers below that temperature, mostly tin-based. Because the base metals do not melt, distortion is low, dissimilar metals can be joined and complex assemblies with many joints can be made at once in a furnace.

Joint design for brazing and soldering is different from welding. Joints should be lap or sleeve joints with a small, uniform clearance, typically 0.025 to 0.15 mm for brazing, so that capillary action works; butt joints are weak. A brazed lap joint with an overlap of about three times the thinner member's thickness is often stronger than the member itself. Surfaces must be clean and protected from oxidation during heating, by a flux or a controlled atmosphere or vacuum. The KG-40 oil cooler lines use copper-phosphorus brazed joints, which need no flux on copper, and silver-brazed joints with flux where brass fittings join copper tube. Lead-free tin-silver-copper solder, required in most electronics today, joins the sensor board.

Adhesive bonding

Structural adhesives, mainly epoxies, acrylics, polyurethanes and cyanoacrylates, have transformed joining in cars, aircraft, wind turbine blades and electronics. Bonding spreads load over the whole joint area instead of concentrating it at holes or welds, joins dissimilar materials without galvanic contact or heat, seals at the same time, damps vibration and adds no weight from fasteners. Its weaknesses are sensitivity to surface preparation, limited strength in peel and cleavage, degradation at high temperature or in some environments, cure time, and difficulty of inspection and disassembly.

Adhesives are strongest in shear and compression and weakest in peel. A single-lap shear joint does not carry uniform stress: the adherends stretch, so shear stress concentrates at the ends of the overlap. That is why doubling the overlap length does not double the strength once the overlap is long, and why joints are designed with tapered ends, fillets of adhesive at the edges and enough width. Surface preparation, degreasing, abrading and often a primer or chemical treatment, is the single most important factor in a durable bond; most bond failures in service trace back to it.

\[\bar\tau = \frac{F}{b\, L} \qquad F_{\text{max}} \approx \bar\tau_{\text{allow}}\, b\, L \ \text{(short overlaps only)}\]

Worked problem

Bonding the sight-glass bracket

A 25 mm wide aluminum bracket is bonded to the steel housing cover with a toughened epoxy. Lap shear tests on properly prepared coupons give an average strength of 24 MPa at 12.5 mm overlap. The bracket carries a static load of 1.2 kN in shear. With a design factor of 4 on the coupon strength to cover temperature, ageing and stress concentration, what overlap is needed?

Show solution
  1. Allowable average shear stress \(= 24/4 = 6.0\ \text{MPa}\).
  2. Required area \(= 1200/6.0 = 200\ \text{mm}^2\); with \(b = 25\ \text{mm}\), \(L = 8.0\ \text{mm}\).
  3. That is shorter than the tested 12.5 mm overlap, so the coupon data apply directly. Use the tested 12.5 mm overlap for margin.
  4. Average stress at 12.5 mm: \(1200/(25 \times 12.5) = 3.8\ \text{MPa}\), a factor of about 6 on coupon strength.

Answer: A 12.5 mm overlap with the joint loaded in shear, not peel, carries the bracket comfortably. The real risk is surface preparation, so the work instruction specifies degreasing, abrasion and primer, and every batch of adhesive is checked with test coupons.

Preload and tightening torque for the housing bolts with several nut factors, the effect of torque scatter, and adhesive overlap sizing.
# Bolted housing joint and bonded bracket for the KG-40
At, Sp, d = 58.0, 580.0, 0.010      # M10 x 1.5 class 8.8: stress area mm^2, proof MPa, diameter m
n_bolts = 12

proof = Sp * At                      # N
Fi = 0.75 * proof
print(f"Proof load {proof/1000:.1f} kN, target preload {Fi/1000:.1f} kN, "
      f"total clamp {n_bolts * Fi / 1000:.0f} kN")

print("\nNut factor  condition                 torque for target (N*m)")
for K, cond in ((0.12, "waxed or molybdenum paste"), (0.16, "zinc flake coating"),
                (0.20, "plain steel, light oil"), (0.30, "dry or rusty threads")):
    print(f"{K:10.2f}  {cond:25s} {K * Fi * d:10.1f}")

T_spec = 40.0                       # N*m specified on the drawing
print(f"\nPreload from {T_spec:.0f} N*m if the actual nut factor differs:")
for K in (0.12, 0.16, 0.20, 0.25):
    F = T_spec / (K * d)
    flag = "over proof!" if F > proof else ("low" if F < 0.6 * proof else "ok")
    print(f"  K = {K:.2f}: preload {F/1000:5.1f} kN ({F/proof*100:4.0f} % of proof) {flag}")

# Bonded bracket: overlap needed
tau_test, design_factor = 24.0, 4.0     # MPa coupon strength, factor for ageing and peaks
b, load = 25.0, 1200.0                  # mm width, N
L_req = load / (tau_test / design_factor * b)
print(f"\nBracket overlap needed: {L_req:.1f} mm (use the tested 12.5 mm)")
print(f"Average stress at 12.5 mm: {load / (b * 12.5):.2f} MPa, "
      f"factor on coupon strength {tau_test / (load / (b * 12.5)):.1f}")
Output
Proof load 33.6 kN, target preload 25.2 kN, total clamp 303 kN

Nut factor  condition                 torque for target (N*m)
      0.12  waxed or molybdenum paste       30.3
      0.16  zinc flake coating              40.4
      0.20  plain steel, light oil          50.5
      0.30  dry or rusty threads            75.7

Preload from 40 N*m if the actual nut factor differs:
  K = 0.12: preload  33.3 kN (  99 % of proof) ok
  K = 0.16: preload  25.0 kN (  74 % of proof) ok
  K = 0.20: preload  20.0 kN (  59 % of proof) low
  K = 0.25: preload  16.0 kN (  48 % of proof) low

Bracket overlap needed: 8.0 mm (use the tested 12.5 mm)
Average stress at 12.5 mm: 3.84 MPa, factor on coupon strength 6.2

Gaskets and sealants

Many joints must also seal. Compressed gaskets of fibre, rubber, or metal fill the surface irregularities between flanges; they need enough bolt preload to seat and enough remaining compression in service. Formed-in-place sealants, typically anaerobic or silicone compounds dispensed by robot along a flange, cure after assembly and can seal machined flanges with very small gaps; the KG-40 cover uses an anaerobic flange sealant applied by a dispensing robot along a programmed path, which replaced a cut gasket and removed a part number. O-rings in machined grooves seal static and slow-moving joints reliably when their groove dimensions follow the manufacturer's guidance for squeeze and fill.

Joint quality and inspection

Every joining method needs process control. Torque tools are calibrated and their results logged by the assembly system; critical bolted joints are audited by checking breakaway torque or measuring elongation. Rivet setting tools are monitored by force-stroke curves. Brazed joints are leak tested and sectioned on samples. Adhesive bonds are controlled by mix ratio checks, pot life limits, cure temperature records and test coupons made with each batch; nondestructive methods such as ultrasonic inspection find unbonded areas in larger structures. The KG-40 assembly line records the final torque and angle of every housing bolt against the gearbox serial number, so that any field problem can be traced.

Common mistakes

  • MistakeUsing the same tightening torque regardless of coating or lubrication.

    InsteadThe nut factor K varies from about 0.12 to 0.3; torque for a given preload changes in proportion.

  • MistakeTreating a snug bolt as adequately preloaded.

    InsteadSpecify and control a preload, typically 75 % of proof load for reusable joints.

  • MistakeDesigning brazed butt joints.

    InsteadUse lap or sleeve joints with controlled clearance so capillary action fills the joint.

  • MistakeLoading adhesive joints in peel.

    InsteadDesign bonded joints for shear or compression; add features that prevent peel.

  • MistakeAssuming adhesive strength grows in proportion to overlap length.

    InsteadStress concentrates at the overlap ends; long overlaps give diminishing returns.

Summary and key terms

  • Joining choice depends on disassembly, loads, materials, heat tolerance, sealing, volume and inspection.
  • Bolted joints rely on preload: F_i ≈ 0.75 S_p A_t, set by torque T = K F_i d with large friction scatter.
  • Rivets, clinching, pins, press fits and snap fits provide permanent or locating joints without fusion.
  • Brazing (above about 450 °C) and soldering (below) use capillary flow of a lower-melting filler into thin lap joints.
  • Adhesives spread load, join dissimilar materials and seal, but are weak in peel and depend on surface preparation.
  • Gaskets, sealants and O-rings seal joints; every method needs process control and traceable inspection.
Preload
The tension in a tightened bolt and the matching clamp force on the joint.
Proof load
The largest load a bolt can carry without permanent set.
Nut factor
The constant K in T = K F d, combining thread and underhead friction.
Property class
The metric strength designation of a bolt, such as 8.8 or 10.9.
Blind rivet
A rivet set from one side by pulling a mandrel.
Shrink fit
An interference fit assembled by heating the hub or cooling the shaft.
Brazing
Joining with a filler metal melting above about 450 °C without melting the base metals.
Soldering
Joining with a filler metal melting below about 450 °C.
Capillary action
The drawing of liquid filler into a narrow gap by surface tension.
Lap shear strength
The average shear stress at failure in a standard bonded lap joint.
Formed-in-place sealant
A liquid sealant dispensed on a flange that cures after assembly.

Practice problems

Work each problem on paper first, then reveal the answer to check.

  1. Find the 75 % proof preload for an M12 class 10.9 bolt (A_t = 84.3 mm², S_p = 830 MPa).

    Show answer

    F_i = 0.75(830)(84.3) = 52.5 kN.

  2. What torque gives that preload with K = 0.2?

    Show answer

    T = 0.2(52 500)(0.012) = 126 N·m.

  3. Why does a well-preloaded bolt resist fatigue better than a loose one?

    Show answer

    Most of the external load reduces clamp pressure; the bolt sees only a small fraction of the load fluctuation.

  4. Give the typical clearance range for a brazed joint.

    Show answer

    About 0.025 to 0.15 mm.

  5. An adhesive joint 20 mm wide with 15 mm overlap fails at 6.6 kN. Find the average shear strength.

    Show answer

    6600/(20 × 15) = 22 MPa.

  6. In the program, change the nut factor to 0.25 at the same 40 N·m. What preload results?

    Show answer

    F = 40/(0.25 × 0.010) = 16.0 kN, well below the 25.2 kN target.

Lesson quiz · 5 questions

Score 60% or more to complete this lesson.

Take the quiz