Manufacturing Processes cover
Interactive book · Semester 3-4

Manufacturing Processes

Size a riser so the casting feeds, find the press tonnage to blank a part, watch a strip bite into the rolls, and see why a pressure vessel is wound at 54.7 degrees

22 chapters, 22 interactives, few words. Every chapter is mapped to the Anna University, VTU, SPPU, Mumbai University and KTU units it covers: patterns and allowances, moulding sand and cores, gating and mould filling time, solidification and riser design, special casting and die casting, flow stress and hot and cold working, rolling, forging, extrusion and wire drawing, blanking, dies and strip layout, bending and springback, deep drawing, gas, arc, resistance, friction and beam welding, the heat-affected zone, injection moulding and extrusion of plastics, composites manufacturing and powder metallurgy.

  • AUME3393Manufacturing Processes
  • VTUBME302Manufacturing Process
  • SPPU202050Manufacturing Processes
  • MUMEC303Production Processes (Modules 1-3 and 6; machining is in the Machining book)
  • KTUPBMET304Manufacturing Processes

By Skill-Lync · Updated 17 Nov 2026 · 22 chapters · 22 interactives · free, no sign-up

Chapter 1

Manufacturing processes, patterns and pattern allowances

Which syllabus covers thisAU ME3393 Unit IVTU BME302 Module 1SPPU 202050 Unit 1MU MEC303 Module 1KTU PBMET304 Module 1

Which family of processes should make a part?

Casting pours liquid metal into a cavity and wins for complex, hollow or very large shapes: engine blocks, pump casings, machine-tool beds. Forming (rolling, forging, extrusion, sheet work) squeezes solid metal and wins on strength and speed for simpler shapes in large numbers. Joining (welding, brazing, adhesives) builds big structures from pieces. Machining removes material for accuracy and finish. Powder metallurgy presses and sinters powder into near-net parts, and additive manufacturing builds layer by layer for low volumes and lattices. Most real parts use two or three: a casting, then machining of the faces that must fit.

What are the steps of sand casting, and what does the pattern do?

Make the pattern and core boxes; ram sand around the pattern in a two-part flask (cope on top, drag below); draw the pattern out; set the cores; close the mould; pour; let it freeze; shake out; cut off gates and risers (fettling); clean, inspect, heat-treat and machine. The pattern is a replica of the casting, made larger by the allowances and carrying core prints that seat the cores. Shops often colour-code patterns, for example black for faces left as cast, red for faces to be machined and yellow for core prints, though practice varies.

What kinds of patterns are there, and what are they made of?

Solid patterns suit simple shapes and a few castings. Split patterns part along the parting line so each half is drawn from its own box. Match plates carry both halves on one plate for machine moulding of small parts in volume; cope-and-drag plates put them on separate plates for big moulds. Loose pieces free an overhang, gated patterns carry their runners, and sweep and skeleton patterns save timber on large one-offs. A follow board backs a thin pattern during ramming. Wood is cheap; aluminium and cast iron survive thousands of rams; plastics resist moisture; wax and polystyrene are melted or burnt out.

Why is a pattern bigger than the part, and by how much?

A casting contracts as it cools to room temperature, so the pattern is made with a contraction (shrink) rule: about 0.8-1.0% larger for cast iron, 1.5-2.0% for steel, 1.0-1.3% for aluminium and 1.5% for brass. The whole casting shrinks like a photo being reduced, holes included. Machining stock adds metal where a face will be cut, so outer sizes grow and bores get smaller. Draft, usually 1-3° on outer faces, lets the pattern leave the sand. Distortion allowance pre-bends shapes that warp; shake allowance is small, negative and has no standard value.

Pattern allowances: shrinkage, machining and draft turn a casting drawing into pattern and core dimensions: interactive loads as you scroll
Predict, then check

A steel casting needs a 50 mm cored bore that is not machined. What diameter should the core be?

Worked examples

Pattern for a cast-iron block with a bore · Grey cast iron block 300 × 180 × 120 mm with a 60 mm through bore; every face and the bore machined, 3 mm stock per face (illustrative); shrinkage 1.0% (upper end of 0.8-1.0%); draft 1° on the outer vertical faces
  1. Add stock: 306 × 186 × 126 mm outside; bore 60 − 6 = 54 mm
  2. Shrink rule × 1.01: pattern 309.06 × 187.86 × 127.26 mm; core Ø 54.54 mm
  3. Draft: each face leans out 127.26 × tan 1° = 2.22 mm
  4. At the parting face the pattern is 313.50 × 192.30 mm

Answer: Pattern 309.1 × 187.9 × 127.3 mm (up to 313.5 × 192.3 at the parting face), core Ø 54.5 mm

Double shrinkage for a metal pattern · A foundry makes a 400 mm steel part from an aluminium match plate, and casts that plate from a wooden master. Shrinkage: steel 2.0%, aluminium 1.3%
  1. The aluminium plate must be 400 × 1.02 = 408.0 mm, so the steel shrinks to 400
  2. The plate is itself a casting, so the wooden master is 408.0 × 1.013 = 413.30 mm

Answer: Wooden master 413.3 mm for a 400 mm steel part

How much metal is machined off · The block of the first example: part volume 300 × 180 × 120 minus the 60 mm bore; the casting is the pattern shape (with draft and the core) after cooling
  1. Part: 61,40,708 mm³
  2. Moulded volume (drafted pattern minus core): 72,32,715 mm³; after 1.0% linear shrinkage it is divided by 1.01³: 70,20,002 mm³
  3. Machined off: 8,79,294 mm³

Answer: About 12.5% of the casting is turned into chips, which is why stock is kept to the minimum the process needs

Chapter 2

Moulding sand, cores, chaplets and moulding machines

Which syllabus covers thisAU ME3393 Unit IVTU BME302 Module 1SPPU 202050 Unit 1MU MEC303 Module 1KTU PBMET304 Module 1

What is moulding sand made of?

Mostly silica (quartz) grains, cheap and refractory. Zircon, olivine and chromite sands cost more but expand less and pull heat faster, so they are used against steel and at hot spots. A binder holds the grains: clay (bentonite) and water make green sand, used moist; baking it gives dry sand; sodium silicate gassed with CO2 sets hard in seconds; resins bind no-bake and shell sands. Additives tune it: coal dust (sea coal) for a smoother skin, cereal binders for strength, wood flour so the mould gives way as the casting shrinks. Core sand needs more strength and must collapse after pouring.

Which sand properties matter, and how are they tested?

Permeability lets steam and gas escape; green strength holds the shape; refractoriness survives the metal's heat; collapsibility lets the casting shrink without tearing; flowability fills corners when rammed. The AFS permeability test blows 2,000 cm³ of air through a rammed specimen 50.8 mm across and 50.8 mm tall and records the pressure and the time, which reduces to PN = 501.3/(p t). Sieve analysis gives the grain fineness number: the percentage retained on each sieve times its multiplier, summed, divided by the total retained. Finer sand gives a smoother casting but lower permeability.

What holds a core in place, and how hard does the metal push it?

A core forms a hole or internal passage. Horizontal, vertical, balanced (one print, counterweighted by its own length), hanging and drop cores differ in how their prints seat in the mould. Liquid metal is much denser than sand, so a submerged core floats: the net lift is V(ρm − ρc)g. For a 100 mm by 500 mm core in liquid iron (6.98 g/cm³) against 1.6 g/cm³ sand, that is about 207 N. Chaplets, small metal supports of a compatible alloy, hold it down and fuse into the casting.

Core buoyancy: the upward force liquid metal puts on a sand core and how many chaplets hold it down: interactive loads as you scroll

How do moulding machines ram the sand?

Jolting lifts the flask and drops it repeatedly, so the sand packs hardest near the pattern, where it matters. Squeezing presses a plate down onto the sand, packing hardest at the top. Jolt-squeeze machines combine both for an even mould and are the workhorse for small and medium boxes. A sand slinger throws sand at high speed from a rotating impeller and suits large moulds and pits. Cores are blown or shot into core boxes with compressed air, then hardened by baking, gassing (CO2 or amine) or self-setting resins.

Predict, then check

A dry sand core of density 1.6 g/cm³ sits in liquid aluminium at 2.375 g/cm³. What happens?

Worked examples

AFS permeability number · A standard specimen (50.8 mm diameter, 50.8 mm tall) passes 2,000 cm³ of air in 1 min 20 s with the manometer at 8 g/cm²
  1. Area A = π × 5.08²/4 = 20.27 cm²; V H/A = 2,000 × 5.08/20.27 = 501.3
  2. t = 80/60 = 1.333 min
  3. PN = 501.3/(8 × 1.333) = 47.0
  4. A finer sand that takes 3 min 10 s at 9 g/cm² gives 501.3/(9 × 3.167) = 17.6

Answer: PN ≈ 47; the finer sand vents less than half as well (18)

Chaplets for a core in iron · Horizontal core 100 mm diameter, 500 mm long between its prints, in liquid iron (6.98 g/cm³); dry sand core 1.6 g/cm³; chaplets rated 100 N each (illustrative); prints not counted on
  1. V = π × 0.1²/4 × 0.5 = 3.927 × 10⁻³ m³
  2. Buoyancy 3.927e-3 × 6,980 × 9.81 = 268.9 N; core weight 61.6 N
  3. Net lift 207.3 N; chaplets ⌈207.3/100⌉ = 3

Answer: 3 chaplets, each carrying about 69 N

The same core in aluminium · As above, but liquid aluminium at 2.375 g/cm³
  1. Net lift = 3.927e-3 × (2,375 − 1,600) × 9.81 = 29.9 N
  2. That is 6.9 times less than in iron

Answer: About 30 N: the prints alone can often hold such a core in an aluminium casting

Chapter 3

Gating system design and mould filling time

Which syllabus covers thisAU ME3393 Unit IVTU BME302 Module 1SPPU 202050 Unit 1MU MEC303 Module 1KTU PBMET304 Module 1

What are the parts of a gating system?

The pouring basin (cup) receives the stream and lets dross float off. The sprue drops the metal to the parting level; a sprue well at its foot absorbs the fall. Runners carry it along the parting plane, and ingates let it into the cavity. The choke is the smallest section and sets the flow rate. Runner extensions and wells trap the first, dirty metal. Gates usually enter at the thickest section. A horn gate curves up into the cavity from below, giving bottom filling for small castings. Steel and cast iron tolerate turbulence; aluminium and magnesium do not.

How do Bernoulli and continuity size a sprue?

A falling stream speeds up: Bernoulli gives v = √(2gh) at a depth h below the basin surface, reduced by a discharge coefficient c (typically 0.8-0.9) for friction and turns. Continuity says the flow Q = A v is the same at every section. So a straight sprue would carry a stream that narrows as it falls; the metal pulls away from the wall and sucks in air and sand gas (aspiration). Tapering the sprue so that A2/A1 = √(h1/h2) keeps it full: with a 40 mm deep basin and the base 250 mm down, the top is 2.5 times the base area.

How long does a mould take to fill?

A top gate pours with a constant head ht above the gate, so the flow is constant and t = V/(Ag c √(2g ht)). A bottom gate fills against rising metal: the head falls from ht to ht − hm, which gives t = (Am/Ag)(2/(c√(2g)))(√ht − √(ht − hm)). A parting-line gate behaves like a top gate for the drag half and a bottom gate for the cope half. Bottom gating fills quietly without splashing; top gating feeds hot metal to the top, which helps the riser. Too slow a fill causes misruns and cold shuts.

Gating design: mould filling time for top, bottom and parting-line gates, gate velocity and sprue taper: interactive loads as you scroll

What are pressurised and unpressurised gating systems?

The gating ratio compares the sprue base, runner and ingate areas. In an unpressurised system the choke is at the sprue base and the areas grow downstream, for example 1:2:2 (typical), so the metal slows down before entering the cavity: preferred for aluminium and magnesium. In a pressurised system the ingates are the smallest, for example 2:1:1 (typical), so the system runs full and the metal enters fast: compact and high-yield, common for iron and steel. Liquid steel has about the kinematic viscosity of water, so gate Reynolds numbers run to tens of thousands: always turbulent.

Predict, then check

One mould, one basin level 300 mm above the cavity floor, a 150 mm tall cavity. Which fills faster, a gate at the cavity top or at its floor?

Worked examples

Top, parting-line and bottom gates on one mould · Cavity 1,000 cm² in plan, 150 mm tall; basin surface 300 mm above the cavity floor; choke 4 cm²; c = 0.85 (typical)
  1. c √(2g) = 0.85 × 4.429 = 3.765
  2. Top gate: head 150 mm, v = 3.765 × √0.15 = 1.458 m/s; t = 0.1 × 0.15/(4 × 10⁻⁴ × 1.458) = 25.7 s
  3. Bottom gate: t = 250 × (2/3.765) × (√0.30 − √0.15) = 21.3 s
  4. Parting-line gate at mid-height (head 225 mm): 22.1 s
  5. Check: summing small steps dh/((Ag/Am) c √(2g(ht − h))) gives 21.30 s for the bottom gate

Answer: Bottom 21.3 s, parting line 22.1 s, top 25.7 s

Tapered sprue and gating ratio · Basin 40 mm deep (illustrative); gate 250 mm below the basin surface; sprue base 4 cm² is the choke; gating ratio 1:2:2 (typical, unpressurised)
  1. Sprue top = 4 × √(250/40) = 10.0 cm²
  2. Velocity at the choke = 3.765 × √0.25 = 1.88 m/s
  3. Runner 8 cm², ingates 8 cm² in total: the metal enters at 1.88 × 4/8 = 0.94 m/s

Answer: Sprue 10 cm² at the top tapering to 4 cm²; ingate velocity halved to about 0.94 m/s

Is the flow in a gate laminar? · Liquid steel, viscosity 6 × 10⁻³ Pa s at 1,550 °C, density about 6,980 kg/m³; square gate 20 × 20 mm; velocity 2.06 m/s (the bottom gate's starting value above)
  1. ν = 6 × 10⁻³/6,980 = 8.60 × 10⁻⁷ m²/s, close to water's
  2. Re = v d/ν = 2.062 × 0.02/8.596e-07 = 47,980
  3. Pipe flow is laminar below about 2,300 and turbulent above 2,900

Answer: Re ≈ 48,000: fully turbulent, so gating design aims to limit velocity and splashing, not to reach laminar flow

Chapter 4

Solidification, Chvorinov's rule, risers and chills

Which syllabus covers thisAU ME3393 Unit IVTU BME302 Module 1SPPU 202050 Unit 1MU MEC303 Module 1KTU PBMET304 Module 1

How does a casting freeze and where does it lose volume?

In three stages: the liquid contracts as it cools to the freezing point, the metal shrinks as it turns solid, and the solid contracts to room temperature (the pattern allowance takes care of that last one). Freezing shrinkage is about 6.6% for aluminium, 4.9% for copper and 2.5-3.0% for low-carbon steel, while grey cast iron can even expand because graphite forms. A pure metal or a narrow-range alloy (under 50 °C) freezes as a skin moving inwards and leaves a pipe; a wide-range alloy (over 110 °C) forms a mushy zone and scattered porosity.

What sets the solidification time?

Chvorinov's rule: the time grows with the square of the modulus M = V/A, the volume per unit cooling area, t = B M² (the exponent is usually taken as 2, some data fit 1.5-2). B depends on the metal and the mould. A dense metal with a large heat of fusion and a small temperature step into a poorly conducting mould freezes slowly. For iron in dry silica sand, B works out near 3.2 s/mm²; a 100 mm steel cube (M = 16.7 mm) needs about 15 minutes. Thin, spread-out sections freeze first.

How is a riser sized?

A riser must freeze after the casting and hold enough liquid for its shrinkage. The modulus method sets the riser modulus at 1.2 times the casting's, so it freezes 1.44 times later. A cylinder of given volume has the least surface when H = D, so its modulus is then D/6 and D = 7.2 Mc. Then check volume: only part of a riser can be drawn before its own pipe reaches the casting (a plain sand riser gives perhaps 15%, illustrative); thin plates often need a bigger riser or a second one. Caine's method plots the freezing ratio against the volume ratio. Open risers vent to air; blind risers sit inside the mould.

Riser design by Chvorinov's rule: casting and riser modulus, solidification times and the riser size that feeds: interactive loads as you scroll

How do chills, padding and sleeves steer solidification?

Directional solidification means freezing starts far from the riser and moves towards it, so liquid always feeds the freezing front. External chills (iron or copper blocks in the mould wall) pull heat fast from a thick section that would otherwise freeze late and cut off the feed path; internal chills of the same alloy are cast in. Padding adds a taper of extra metal to keep the path open. Insulating and exothermic sleeves keep a riser hot longer; a vendor rates cylindrical sleeves at about 30% feeding efficiency and its mini-risers at 70%.

Predict, then check

A 100 mm cube and a 200 × 200 × 25 mm plate hold the same 1,000 cm³ of steel. Poured into the same sand, which freezes first?

Worked examples

Mould constant from properties · Iron data for steel: liquid density 6,980 kg/m³, heat of fusion 13.81 kJ/mol ÷ 0.055845 kg/mol, melting point 1,811 K; dry sand mould at 25 °C: k = 0.25 W/mK, ρ = 1,600 kg/m³, c = 800 J/kgK; superheat neglected
  1. L = 13,810/0.055845 = 2,47,292 J/kg; ρm L/(Tm − T0) = 6,980 × 2,47,292/1,512.85 = 1.141e+06
  2. π/(4 k ρ c) = π/(4 × 0.25 × 1,600 × 800) = 2.454e-06
  3. B = 3.195e+06 s/m² = 3.20 s/mm² (aluminium 5.40, copper 6.12)
  4. A 100 mm cube: M = 16.67 mm, t = 3.20 × 16.67² = 888 s

Answer: B ≈ 3.2 s/mm²; the cube freezes in about 15 minutes

Riser for a steel cube · Steel cube 150 mm; cylindrical top riser H = D; B = 3.20 s/mm²; volume shrinkage 3.0%; riser efficiency 15% (illustrative)
  1. Mc = 150/6 = 25 mm; Mr = 1.2 × 25 = 30 mm; D = 6 × 30 = 180 mm = H
  2. tc = 3.195 × 25² = 1997 s (33.3 min); tr = 2876 s (47.9 min)
  3. Riser volume π × 0.18³/4 = 4.58 L; shrinkage to feed 0.03 × (3.375 + 4.58) L = 239 cm³
  4. Riser can give 0.15 × 4.58 L = 687 cm³: enough

Answer: A 180 × 180 mm riser; the modulus rule governs

Riser for an aluminium plate · Aluminium plate 300 × 300 × 60 mm; riser H = D; volume shrinkage 6.6%; efficiency 15% (illustrative); B = 5.40 s/mm²
  1. Mc = 54,00,000/2,52,000 = 21.43 mm; modulus rule D = 7.2 × 21.43 = 154 mm
  2. That riser holds 2.88 L and can give 433 cm³, but the plate and riser need 0.066 × (5.4 + 2.88) L = 547 cm³
  3. Volume rule: 0.15 Vr = 0.066 (Vc + Vr), so Vr = 0.066 × 5.4/0.084 = 4.24 L and D = 175 mm
  4. Freezing: plate 41 min, riser 77 min

Answer: D = H ≈ 175 mm, set by volume; a long plate would also need a second riser for feeding distance

Chapter 5

Melting furnaces and special casting: shell, investment, CO2 and centrifugal

Which syllabus covers thisAU ME3393 Unit IVTU BME302 Modules 1, 2SPPU 202050 Unit 1MU MEC303 Module 1KTU PBMET304 Module 1

Which furnace melts which metal?

The cupola is a tall shaft charged from the top with layers of coke, pig iron and scrap, and limestone as flux; air blown in near the bottom burns the coke. Going up, the zones are the well, combustion, reducing, melting and preheating zones, then the stack. It suits grey cast iron, which melts at about 1,130-1,200 °C. Coreless induction furnaces melt iron, steel and non-ferrous alloys cleanly in batches; electric arc furnaces melt steel scrap (carbon steel melts at 1,425-1,540 °C); crucible, pit and resistance furnaces suit aluminium (660 °C) and other low-melting alloys.

How do shell, investment, CO2, ceramic, plaster and vacuum moulds differ from green sand?

Shell moulding drops resin-coated sand onto a heated metal pattern; a thin shell cures, two halves are glued, and the result is a precise, smooth mould. Investment (lost wax) coats a wax pattern with ceramic slurry, melts the wax out and pours into the hot shell: fine detail and no parting line. Ceramic and plaster moulds give smooth faces for non-ferrous parts. The CO2 process mixes sand with sodium silicate and hardens it by blowing carbon dioxide, so no baking is needed. The V-process holds dry, binder-free sand in shape with vacuum under a plastic film.

How fast must the mould spin in true centrifugal casting?

In true centrifugal casting a pipe mould spins about its own axis; metal poured in is flung to the wall and freezes as a tube with no core. The outside comes from the mould, the bore from how much metal is poured. The G factor is the centrifugal force over the weight at the mould bore. A lecture note puts the suitable range at 60-80 for horizontal machines; too low and the metal rains off the top of the bore. A 300 mm pipe needs about 646 rpm for G 70. Lighter slag and oxides drift to the bore, where they are machined off.

Centrifugal casting of a pipe: mould speed for a G factor, metal mass and the pressure on the mould: interactive loads as you scroll

What are semi-centrifugal casting, centrifuging, squeeze, slush and thixocasting?

Semi-centrifugal casting spins a full mould, with a central core where needed, about a vertical axis to make solid, symmetric parts such as wheels; the densest metal is at the rim. Centrifuging places small cavities around a central sprue and spins the whole cluster, so the spin pressurises every cavity. Squeeze casting closes a die on metal while it freezes under high pressure, giving dense parts. Slush casting pours metal out once a thin skin has frozen, leaving hollow ornaments. Thixocasting injects a partly solid, non-dendritic billet that flows like paste. The spinning metal also presses on the mould: p = ρω²(Ro² − Ri²)/2.

Predict, then check

A horizontal machine casts a 300 mm pipe at G 70. To cast a 600 mm pipe at the same G, the mould speed must:

Worked examples

Mould speed and metal for a cast-iron pipe · Horizontal true centrifugal casting of a cast-iron pipe (ρ 7,200 kg/m³): OD 300 mm, wall 10 mm, length 5.5 m, G factor 70 at the mould bore
  1. R = 0.15 m; ω = √(G g/R) = √(70 × 9.81/0.15) = 67.66 rad/s, so N = 60ω/2π = 646 rpm
  2. Metal = ρ π (Ro² − Ri²) L = 7,200 × π × (0.15² − 0.14²) × 5.5 = 361 kg (before any allowance for spill or machining)
  3. Pressure on the mould = ρω²(Ro² − Ri²)/2 = 7,200 × 67.66² × 0.0029/2 = 47.8 kPa
  4. At the bore the metal sees G × Ri/Ro = 70 × 0.14/0.15 = 65.3, still well above 1, so it does not rain

Answer: About 646 rpm, 361 kg of iron, 48 kPa on the mould

A bronze bush on a speed-limited machine · Phosphor bronze (ρ 8,900 kg/m³) bush OD 120 mm, wall 15 mm, length 300 mm, target G 75; the machine tops out at 900 rpm
  1. N for G 75 at R = 0.06 m: (60/2π)√(75 × 9.81/0.06) = 1057 rpm, above the machine limit
  2. At 900 rpm: G = ω²R/g = (94.25)² × 0.06/9.81 = 54.3, below the 60-80 band
  3. At the bore (Ri = 45 mm) G = 54.3 × 45/60 = 40.7: the metal still sticks, but the denser layer is thinner
  4. Metal per bush = 8,900 × π × (0.06² − 0.045²) × 0.3 = 13.21 kg; pressure at G 75 would be 86 kPa

Answer: Needs 1057 rpm; at 900 rpm G is only 54, so use a faster machine or accept a lower G

Why vertical spinning suits short parts · Vertical-axis mould, bore radius 40 mm at the bottom, casting height 200 mm, 600 rpm. The free surface of spinning liquid is a paraboloid: z = ω²(r² − rb²)/(2g)
  1. ω = 2π × 600/60 = 62.83 rad/s
  2. r_top² = rb² + 2gH/ω² = 0.04² + 2 × 9.81 × 0.2/62.83² = 0.0016 + 0.000994
  3. r_top = 50.9 mm: the bore tapers by 10.9 mm in radius over 200 mm
  4. To hold the taper to 1 mm: ω = √(2gH/(0.041² − 0.04²)) = 220 rad/s = 2,102 rpm

Answer: The bore opens from 40 to 51 mm at 600 rpm; about 2,100 rpm keeps it within 1 mm

Chapter 6

Die casting, permanent moulds, continuous casting and casting defects

Which syllabus covers thisAU ME3393 Unit IVTU BME302 Module 2SPPU 202050 Unit 1MU MEC303 Module 1KTU PBMET304 Module 1

What are gravity die, low-pressure die and tilt-pour casting?

A permanent mould is a metal die used thousands of times. In gravity die casting metal is simply poured into a preheated, coated steel or cast-iron die; sand or metal cores make the hollows. Low-pressure die casting pushes metal up a riser tube from a sealed furnace below the die with air at about 0.3-0.5 bar, and raises it towards 1 bar after filling to feed shrinkage: calm filling, little scrap, used for aluminium wheels. Tilt pouring fills the die by slowly rotating it with the cup, so the metal flows in without splashing.

What sets the size of a high-pressure die casting machine?

High-pressure die casting rams metal into a steel die in a fraction of a second at roughly 10-175 MPa. That pressure acts on the whole projected area of cavities, runners, biscuit and overflows, trying to open the die, so the machine's locking force must exceed it with a margin (at least 10 %). Hot-chamber machines keep the injection gooseneck in the melt: fast, for zinc and magnesium. Aluminium attacks the immersed iron parts, so it uses a cold-chamber machine, ladled into a shot sleeve each cycle. HMT's Indian cold-chamber line runs from 80 to 1,100 t.

High-pressure die casting: locking force from projected area and cavity pressure, machine size and fill time: interactive loads as you scroll

How does continuous casting work?

Steel from a ladle flows into a tundish, a buffer box that feeds a short, water-cooled copper mould. The mould oscillates so the thin frozen shell does not stick; below it, water sprays freeze the core while withdrawal rolls pull the strand down and bend it horizontal, and a torch cuts it into slabs, blooms or billets. Output is section × speed × density: a 200 mm square billet strand at 1.5 m/min (illustrative) gives 7,850 × 0.04 × 0.025 = 7.85 kg/s, about 28 t per hour per strand.

Which defects come from gas and from shrinkage?

Blowholes are smooth, rounded cavities from gas trapped by a wet or impermeable mould or poor venting; pinhole porosity comes from gas dissolved in the melt that comes out as it freezes. Remedies: dry sand, more permeability and vents, degassing. Shrinkage cavities are rough, jagged voids in the last region to freeze, often under a riser that is too small or freezes first; centreline shrinkage forms in long thin sections. Remedies: bigger or better-placed risers, chills, padding for directional solidification (chapter 4). Inclusions of sand, slag or oxide come from erosion or poor skimming.

What causes misruns, cold shuts, hot tears, scabs, swells, shifts and fins?

A misrun is a casting that did not fill because the metal froze first; a cold shut is a seam where two streams met too cold to fuse. Both call for hotter metal, faster pouring or better gating. A hot tear is a crack formed while the metal is weak near its solidus and the mould or core resists contraction: use collapsible sand and fillets. Scabs and swells come from sand that spalls or yields under metal pressure; a shift is a step at the parting line from misaligned flasks or cores; fins are thin metal in gaps. Cleaning removes gates, risers and fins; visual, dye-penetrant, radiography and ultrasonic tests inspect.

Predict, then check

Why is aluminium not cast on hot-chamber die casting machines?

Worked examples

Machine size for an aluminium housing · Aluminium housing, projected area 420 cm², one cavity; runners, biscuit and overflows add 30 % (illustrative); cavity pressure 60 MPa (illustrative); safety factor 1.2; machine sizes 80, 125, 160, 250, 400, 630 t (illustrative ladder)
  1. Projected area = 420 × 1.3 = 546 cm² = 54,600 mm²
  2. Opening force = 54,600 mm² × 60 MPa = 32,76,000 N = 3,276 kN
  3. Locking force = 1.2 × 3,276 = 3,931.2 kN = 400.7 t (1 t = 9.81 kN)
  4. Just over 400 t, so the ladder says 630 t. Dropping the intensification pressure to 55 MPa gives 3,603.6 kN = 367 t: a 400 t machine

Answer: About 401 t: a 630 t machine, or 400 t if the process runs at 55 MPa

A four-cavity zinc die: force, gate and shot · Zinc (ρ 7,135 kg/m³, hot chamber) part, projected area 30 cm² per cavity, 4 cavities, 20 MPa (illustrative), wall 1.6 mm; recommended fill time for walls under 1.8 mm: 0.02-0.04 s; gate velocity 25 m/s (zinc range 20-30 m/s); part volume taken as area × wall × 2 (illustrative)
  1. Locking force = 1.2 × 30 × 1.3 × 4 cm² × 100 × 20 MPa = 374.4 kN = 38.2 t: the smallest, 80 t, machine
  2. Cavity volume = 30 × 100 × 1.6 × 2 × 4 = 38,400 mm³
  3. Gate area for a 0.03 s fill = V/(v t) = 38,400/(25,000 × 0.03) = 51.2 mm² in total, 12.8 mm² per cavity
  4. Shot mass with 50 % extra for biscuit, runners and overflows (illustrative) = 7,135 × 0.000038 × 1.5 = 0.41 kg

Answer: 38 t (80 t machine), gate 51 mm², shot about 0.41 kg

Reading a machine table backwards · HMT's H 250D (250 t, cold chamber) lists 1,345 kg/cm² injection pressure over a nominal casting area of 185 cm² with its standard 60 mm shot sleeve. What part area fits at 60 MPa with the 1.2 and 30 % allowances?
  1. Check the table: 1,345 kg/cm² × 185 cm² = 2,48,825 kg = 249 t, the rated locking force: area × pressure is the whole rule
  2. At 60 MPa, 250 t holds 250 × 9.81 × 1,000/60 = 40,875 mm² = 409 cm² of projected area
  3. Divide by the 1.2 safety factor and the 1.3 runner share: 409/(1.2 × 1.3) = 262 cm² of part

Answer: About 262 cm² of part area per shot on a 250 t machine at 60 MPa

Chapter 7

Plastic deformation, flow stress, yield criteria and hot and cold working

Which syllabus covers thisAU ME3393 Unit IIIVTU BME302 Module 3SPPU 202050 Unit 2MU MEC303 Module 3KTU PBMET304 Module 3

Why does forming use true stress and true strain?

Below the yield point a metal springs back (elastic); past it, the shape change is permanent (plastic), and forming lives entirely in the plastic range. Engineering stress divides load by the original area and engineering strain uses the original length, both misleading once a part has changed shape a lot. True stress uses the current area; true strain ε = ln(L/L0) adds up correctly over several passes. Plastic flow keeps volume constant, so ε = ln(A0/A) too. Stretching a bar by 50 % is a true strain of ln 1.5 = 0.405, not 0.5.

What is the flow curve, and what is the average flow stress?

Beyond yield, the true stress needed to keep a metal flowing rises as it work-hardens. For most metals σ = Kεⁿ fits well: K is the strength coefficient and n the strain-hardening exponent (annealed low-carbon steel K 530 MPa, n 0.26; copper 325 MPa, 0.54). Forming energy and many force formulas need the average over the strain range, Ȳf = Kεⁿ/(1 + n). In a tensile test the bar necks where dσ/dε = σ, which for this curve is simply ε = n, giving a tensile strength of Knⁿe⁻ⁿ.

Flow curve: true stress-strain, average flow stress for cold working, strain-rate sensitivity for hot working and necking: interactive loads as you scroll

When does a metal yield under combined stresses?

Tresca says yielding starts when the largest shear stress reaches that of a tensile test: σmax − σmin = Y. Von Mises uses distortion energy, which tests match a little better. In plane strain, common in rolling, wide-strip forging and drawing of wide sheet, the metal cannot spread sideways, so the middle stress settles at the mean of the other two and von Mises gives σ1 − σ3 = 2Y/√3 = 1.155Y. Tresca gives Y. That 15.5 % is why plane-strain formulas use 1.155 times the flow stress.

What separates cold, warm and hot working?

The dividing line is recrystallisation, where new strain-free grains form: roughly 0.3-0.4 of the melting point in kelvin for pure metals and 0.5 for alloys. Below it, cold working hardens the metal (flow curve), gives close tolerances and bright finish but needs annealing between heavy passes. Hot working, well above it, keeps the metal soft as it recrystallises, so huge strains are possible, but scale forms and tolerances loosen. Then strain rate matters more than strain: σ = Cε̇ᵐ, with m about 0.01 at a fifth of the melting point and rising towards 1 near melting. Warm working sits between.

Predict, then check

Annealed copper (n 0.54) and aluminium 6061-O (n 0.20) are pulled in tension. Which stretches more before it starts to neck?

Worked examples

Tensile strength from K and n · Annealed low-carbon steel, σ = 530 ε^0.26 MPa
  1. Necking (Considère, dσ/dε = σ) at ε = n = 0.26; a numerical search for the peak engineering stress gives ε = 0.260
  2. True stress there = 530 × 0.26^0.26 = 373 MPa
  3. Engineering stress = true × e^(−ε) = 373 × 0.7711 = 288 MPa: the tensile strength
  4. Uniform elongation = e^0.26 − 1 = 29.7 %

Answer: UTS ≈ 288 MPa, reached at 30 % elongation

Flow stress after heavy cold work · Three annealed metals cold-worked to a true strain of 0.6 (σ = Kεⁿ from the sourced table)
  1. Copper: σf = 325 × 0.6^0.54 = 247 MPa; Ȳf = 247/1.54 = 160 MPa
  2. Low-carbon steel: σf = 530 × 0.6^0.26 = 464 MPa; Ȳf = 464/1.26 = 368 MPa
  3. 304 stainless: σf = 1,275 × 0.6^0.45 = 1,013 MPa; Ȳf = 699 MPa
  4. Stainless needs about 1.9 times the forming force of low-carbon steel for the same strain

Answer: Ȳf: copper 160, steel 368, 304 stainless 699 MPa

Strain-rate sensitivity in hot working · Hot compression tests on low-carbon steel at 1,100 °C (illustrative data): 105 MPa at ε̇ = 1 /s and 132 MPa at 10 /s. A press then squeezes a 50 mm tall billet at 200 mm/s
  1. m = log(132/105)/log(10/1) = 0.099; C = 105 MPa (the stress at ε̇ = 1)
  2. Strain rate at the start = v/h = 200/50 = 4 /s; σ = 105 × 4^0.099 = 121 MPa
  3. T/Tm = (1,100 + 273)/(1,425 + 273) = 0.81; recrystallisation near 0.5 Tm ≈ 576 °C, so this is hot working
  4. Cold, the same steel at ε = 0.5 needs 530 × 0.5^0.26 = 443 MPa, about 3.7 times more

Answer: m ≈ 0.10; flow stress about 121 MPa in the press

Plane strain needs 15.5 % more · A metal with tensile yield 250 MPa (illustrative) is pressed in plane strain: σ3 = 0, no width change so σ2 = σ1/2
  1. Von Mises: √(½[(σ1 − σ1/2)² + (σ1/2)² + σ1²]) = 250 solved for σ1 gives 288.7 MPa
  2. Ratio 288.7/250 = 1.1547 = 2/√3
  3. Tresca: σ1 − σ3 = 250 MPa, no increase
  4. Shear yield: von Mises Y/√3 = 144 MPa, Tresca Y/2 = 125 MPa

Answer: 289 MPa by von Mises (1.155Y), 250 MPa by Tresca

Chapter 8

Rolling: draft, roll force, torque and power

Which syllabus covers thisAU ME3393 Unit IIIVTU BME302 Module 3SPPU 202050 Unit 2MU MEC303 Module 3KTU PBMET304 Module 3

How much can one pass reduce the strip?

Two rolls of radius R squeeze a strip from h0 to hf; the draft is d = h0 − hf. Friction is what pulls the strip into the gap, so it bites only if the roll's grip beats the push-back of the inclined contact: tan α ≤ μ, which for small angles becomes d ≤ μ²R. With μ 0.1 and R 200 mm the largest draft is 2 mm. The contact arc projects to a length L = √(Rd) along the strip. Plastic strain is ε = ln(h0/hf): 4 to 3 mm is 0.288.

How are roll force, torque and power estimated?

The simplest estimate takes the average flow stress over the pass, Ȳf = Kεⁿ/(1 + n), acting over the contact area: F = Ȳf w L. Each roll's torque is the force acting at about half the contact length, T = F L/2, and the power for both rolls is P = 2πN F L/60 with N in rpm. An aluminium 6061-O strip 250 mm wide going from 4 to 3 mm in 200 mm rolls needs about 471 kN and, at 50 rpm, 35 kW. Change the rolls, friction and strip in the interactive.

Flat rolling: maximum draft, contact length, roll force, torque and power with the friction hill and neutral point: interactive loads as you scroll

What are the friction hill and the neutral point?

The strip enters slower than the roll surface and leaves faster, since it gets thinner at constant volume. At one place, the neutral point, the two speeds match. Before it, friction drags the strip in; after it, friction holds it back; both directions pile pressure towards the neutral point, so the roll pressure rises from 1.155Ȳf at entry and exit to a peak: the friction hill. The slab method integrates this balance; in the cold-rolling cases checked for this chapter its force came out 1.26-1.58 times Ȳf w L. Forward slip, exit speed over roll speed minus one, is a few per cent.

Why do mills have so many roll arrangements?

A two-high mill has one pair of rolls, non-reversing or reversing so the bar can go back and forth. A three-high mill passes the stock forward through the lower gap and back through the upper one. Since force grows with √R, small work rolls cut the force, but they bend, so a four-high mill backs two small work rolls with two large rolls, and a cluster mill such as the 20-high Sendzimir backs tiny rolls further for thin, hard strip. A tandem mill places 2 to 18 stands in a row, each taking a draft, with speeds rising stand by stand.

What else is rolled, and what goes wrong?

Shape rolling passes hot bar through grooved rolls in stages to make rails, I-beams and channels. Thread rolling forms screw threads between dies without cutting, so the grain follows the thread. Ring rolling thins and enlarges a pierced ring for bearing races and flanges. Defects: wavy edges when the rolls bend and the edges get longer than the middle, centre buckle for the reverse, zipper cracks in the middle, edge cracks from low ductility at the edges, alligatoring when a slab splits along its middle, and rolled-in scale.

Predict, then check

At the same draft, the roll radius is doubled. The simple roll force F = Ȳf w L becomes:

Worked examples

One cold pass of an aluminium strip · Aluminium 6061-O (K 205 MPa, n 0.20) strip 250 mm wide, 4.0 → 3.0 mm, rolls R 200 mm, μ 0.1 (typical lubricated cold rolling), 50 rpm
  1. Draft 1.0 mm ≤ μ²R = 0.1² × 200 = 2.0 mm: it bites. L = √(200 × 1.0) = 14.14 mm
  2. ε = ln(4/3) = 0.2877; Ȳf = 205 × 0.2877^0.2/1.2 = 133.2 MPa
  3. F = Ȳf w L = 133.2 × 250 × 14.14 = 471 kN; torque per roll = F L/2 = 3.33 kN·m
  4. Power, both rolls = 2π × 50 × 471 kN × 0.01414 m/60 = 34.9 kW
  5. Slab method (S = 1.155Ȳf = 154 MPa): F = 628 kN, 1.33 times; peak pressure 213 MPa; neutral point at 1.30°, forward slip 3.4 %

Answer: F ≈ 471 kN (slab 628 kN), 35 kW for both rolls

Two passes: the flow stress keeps rising · Low-carbon steel strip (K 530 MPa, n 0.26) 300 mm wide, 2.0 mm → 1.0 mm; rolls R 100 mm, μ 0.08
  1. μ²R = 0.08² × 100 = 0.64 mm per pass, so 1.0 mm needs two passes: 2.0 → 1.5 → 1.0 mm (0.5 mm each)
  2. Pass 1: ε from 0 to 0.288; Ȳf = 530 × 0.288^0.26/1.26 = 304 MPa; L = √(100 × 0.5) = 7.07 mm; F = 645 kN
  3. Pass 2 starts work-hardened: ε from 0.288 to ln 2 = 0.693; Ȳf = K(ε2^1.26 − ε1^1.26)/(1.26(ε2 − ε1)) = 438 MPa; F = 929 kN
  4. Treating pass 2 as if the strip were annealed (ε 0 to 0.405) would give 333 MPa, 24 % low

Answer: Two passes; about 645 kN then 929 kN (simple formula)

Smaller rolls, smaller force · The aluminium pass of the first example (4 → 3 mm, 250 mm wide, Ȳf 133.2 MPa, 50 rpm) in rolls of R 100 mm instead of 200 mm
  1. With μ 0.1, μ²R = 1.0 mm: right at the limit. Raise μ to 0.12 (rougher rolls): μ²R = 1.44 mm
  2. L = √(100 × 1.0) = 10.0 mm instead of 14.14 mm
  3. F = 133.2 × 250 × 10 = 333 kN, 0.707 of the 200 mm rolls (1/√2)
  4. Power = 2π × 50 × 333 kN × 0.010 m/60 = 17.4 kW, half, since both F and L shrink by √2

Answer: 333 kN and 17.4 kW, against 471 kN and 34.9 kW

Chapter 9

Forging: open-die and closed-die forging loads and defects

Which syllabus covers thisAU ME3393 Unit IIIVTU BME302 Module 3SPPU 202050 Unit 2MU MEC303 Module 3KTU PBMET304 Module 4

How big a press does open-die upsetting need?

Squash a cylinder between flat dies and its volume stays the same, so as the height h falls the area A = V/h grows and the metal work-hardens. The frictionless load is Yf A, with Yf the flow stress K εⁿ at the current strain ε = ln(h₀/h). Friction at the dies holds the faces back, the billet barrels, and the load rises further: a common estimate multiplies by Kf = 1 + 0.4 μ D/h. The peak load comes at the end of the stroke, so size the press for that.

Open-die upsetting: forging load through the stroke with friction and barrelling, and closed-die load factors: interactive loads as you scroll

What is the friction hill, and why does the slab method give a different factor?

Cut the disc into thin rings and balance the forces on each. Friction pointing towards the centre makes the die pressure climb from Y at the rim to a peak at the axis: p = Y e^(2μ(R − r)/h), the friction hill. Averaged over the face it is close to Y (1 + 2μR/(3h)) when μR/h is small. That is μD/(3h) of extra load against 0.4 μD/h in Kf: the slab result ignores barrelling and assumes uniform sliding, so it sits a little lower. A squat billet (large D/h) feels friction most.

How are impression-die and flashless forging loads estimated?

In an impression die the metal fills a cavity and the excess squirts into a thin gap round the parting line as flash. The flash cools fast and has a long, thin land, so it resists flow and forces the cavity to fill; then it is trimmed off. Loads are estimated empirically as F = C₁ Yf A on the projected area including flash, with C₁ about 8-12 for complex parts with flash and 5-8 for flashless coining, against 1.25-2.5 for plain upsetting. Flashless forging needs an exactly measured billet volume.

What are the basic forging operations, and hammers versus presses?

Open-die work shapes stock step by step: fullering thins a section with convex dies, edging gathers metal into a section, drawing out lengthens a bar, heading upsets the end of a rod (bolt heads), piercing punches a hollow, and coining presses fine detail into a constrained surface. Hammers strike: energy is delivered fast and several blows are used. Presses squeeze: mechanical presses give a fixed stroke at high rate, hydraulic presses a slow, controlled full-force stroke for large or deep parts. Cold forging gives close tolerances and strength from work hardening but needs higher loads.

What goes wrong in forging?

A lap forms when metal folds over on itself, often from too much material or a sharp die radius; a cold shut is two fronts of metal meeting without welding. Incomplete die filling comes from too small a billet, low temperature or poor flash design. Surface cracks appear when the metal is worked too cold or the barrelled free surface is pulled into tension; internal cracks can open in the centre of a heavily drawn-out bar. Scale pits come from oxide pressed into the surface in hot forging. Good preforms, generous radii and the right temperature prevent most of these.

Predict, then check

Two billets of the same steel and volume are upset by the same 50% in height with the same lubricant: one starts tall and thin, the other short and fat. Which needs the higher die pressure at the end?

Worked examples

Cold upsetting a copper terminal slug · Annealed copper (K 325 MPa, n 0.54), D₀ 30 mm, h₀ 45 mm upset to h 18 mm; μ 0.08
  1. Volume V = π × 30² × 45/4 = 31,809 mm³; at h = 18 mm, A = V/h = 1,767.1 mm², D = 47.43 mm
  2. True strain ε = ln(45/18) = 0.9163; flow stress Yf = 325 × 0.9163^0.54 = 310.0 MPa
  3. Kf = 1 + 0.4 × 0.08 × 47.43/18 = 1.0843
  4. F = Kf Yf A = 1.0843 × 310.0 × 1,767.1 = 594 kN (frictionless 548 kN)
  5. Slab method: 1 + 2μR/(3h) = 1.0703, F = 586 kN, about 1.3% below the Kf estimate

Answer: About 594 kN at the end of the stroke

A squat aluminium disc: where friction takes over · Al 1100-O (K 180 MPa, n 0.20), D₀ 60 mm, h₀ 30 mm pressed to h 10 mm; μ 0.05 (good lubricant) or 0.2 (dry)
  1. D = 60 √(30/10) = 103.9 mm, A = 8,482 mm², ε = ln 3 = 1.0986, Yf = 183.4 MPa
  2. μ 0.05: Kf = 1.208, slab linear 1.173, exact friction hill average 1.198: all close; F ≈ 1.88 MN
  3. μ 0.2: 2μR/h = 2.08; Kf = 1.831, slab linear 1.693, exponential hill 2.275
  4. The hill would peak at e^2.08 = 8.0 Yf at the centre; friction cannot exceed the shear yield, so the metal sticks there and the true average lies between the estimates
  5. With Kf: F = 1.831 × 183.4 × 8,482 = 2.85 MN, 52% more than with lubricant

Answer: About 1.88 MN lubricated and 2.85 MN dry: lubrication matters most on squat parts

Press for coining a copper medal · Medal 40 mm in diameter, flashless closed die; the blank reaches a strain of about 0.25 (annealed copper, K 325 MPa, n 0.54); C₁ 5-8
  1. Yf = 325 × 0.25^0.54 = 153.7 MPa
  2. Projected area A = π × 20² = 1,256.6 mm²
  3. F = C₁ Yf A = 5 to 8 × 153.7 × 1,256.6 = 966 to 1546 kN
  4. In tonnes-force: 98 to 158 t, so a 160 t press leaves a margin

Answer: About 966-1546 kN (98-158 t)

Chapter 10

Extrusion, wire drawing and tube drawing

Which syllabus covers thisAU ME3393 Unit IIIVTU BME302 Module 3SPPU 202050 Unit 2MU MEC303 Module 3KTU PBMET304 Module 4

What are direct, indirect, impact and hydrostatic extrusion?

Extrusion pushes a billet through a shaped die, like toothpaste from a tube. In direct (forward) extrusion the ram pushes the billet along a stationary container, so the whole billet slides and container friction adds to the load. In indirect (backward) extrusion the die sits on a hollow ram and moves into the billet, which does not slide: lower force, but the product must pass through the ram. Impact extrusion makes cans and tubes from a slug in one fast blow. Hydrostatic extrusion surrounds the billet with pressurised fluid, removing wall friction. Hot extrusion makes aluminium window and door profiles; cold extrusion makes accurate steel parts.

How is the ram pressure estimated?

The extrusion ratio R = A₀/A_f sets the ideal strain ln R; with no friction the pressure equals the work per unit volume, Ȳf ln R, where Ȳf = K εⁿ/(1 + n). Real dies add friction and redundant shearing, so Johnson's empirical strain ε_x = a + b ln R (about a = 0.8, b = 1.2-1.5) replaces ln R. In direct extrusion the billet also drags along the container: with sticking friction that adds Ȳf × 2L/D₀, which falls as the remaining length L shrinks, so pressure peaks at the start of the stroke.

Extrusion and wire drawing: extrusion ratio, ram pressure, draw stress, die angle and the maximum reduction: interactive loads as you scroll

How is the draw stress in wire and rod drawing found?

In drawing the wire is pulled, not pushed, through a die with an entry, a conical approach of half-angle α, a short parallel bearing and a back relief. The pull per unit exit area is the draw stress. Ideally it is Ȳf ln(A₀/A_f); friction on the cone adds the factor (1 + μ/tan α) and uneven (redundant) deformation adds φ = 0.88 + 0.12 D̄/Lc, where Lc is the contact length. Small die angles raise friction, large ones raise redundant work, so for each reduction and lubricant there is a best die angle.

Why is there a maximum reduction per pass, and how is tube drawn?

The drawn wire must carry the draw stress without yielding, or it simply stretches and breaks. With no friction and no hardening this gives ln(A₀/A_f) = 1: a reduction of 1 − 1/e = 63%. Friction and redundant work lower it; work hardening raises it, so practical passes are about 30-50% and wire is drawn through 4-12 dies in series. Tube is drawn over a fixed mandrel (rod), a floating plug that sits in the die, or with no internal tool (sinking, which controls only the outside). Soap, oil or polymer lubricants and carbide or diamond dies keep friction low.

What defects occur in extrusion and drawing?

Centre burst (chevron cracking) is a string of internal arrow-shaped cracks along the axis, from tension at the centre when the die angle is large and the reduction per pass small; it is invisible from outside. Piping draws oxidised surface and inclusions into the centre at the end of a direct-extrusion billet, so the last part (the butt) is discarded. Surface cracking (fir-tree cracks) comes from too high a temperature, speed or friction. Drawn wire can also show seams, die marks and surface scratches from poor lubrication.

Predict, then check

Ignore friction and work hardening. What is the largest area reduction a single drawing pass can make before the wire itself yields?

Worked examples

Direct extrusion of an aluminium rod · Al 1100-O billet (K 180 MPa, n 0.20), D₀ 80 mm, length 150 mm, extruded cold to a 30 mm rod; Johnson a 0.8, b 1.5; sticking at the container wall
  1. R = (80/30)² = 7.111; ideal strain ln R = 1.9617
  2. Ȳf = 180 × 1.9617^0.2/1.2 = 171.6 MPa; ideal pressure Ȳf ln R = 337 MPa
  3. Johnson strain ε_x = 0.8 + 1.5 × 1.9617 = 3.742; die pressure Ȳf ε_x = 642 MPa
  4. Container friction at the start: Ȳf × 2L/D₀ = 171.6 × 2 × 150/80 = 644 MPa
  5. Start: p = 1286 MPa, F = p × π × 80²/4 = 6.46 MN; near the end (L → 0) F falls to 3.23 MN, the indirect-extrusion load

Answer: Peak ram force about 6.5 MN, half of it container friction

One pass of copper wire drawing · Annealed copper (K 325 MPa, n 0.54), 3.0 mm drawn to 2.5 mm; die half-angle 8°, μ 0.08; exit speed 5 m/s
  1. r = 1 − (2.5/3)² = 30.6%; ε = ln(A₀/A_f) = 0.3646
  2. Ȳf = 325 × 0.3646^0.54/1.54 = 122.4 MPa
  3. D̄ = 2.75 mm, Lc = 0.5/(2 sin 8°) = 1.796 mm, φ = 0.88 + 0.12 × 2.75/1.796 = 1.0637
  4. σd = 122.4 × (1 + 0.08/tan 8°) × 1.0637 × 0.3646 = 122.4 × 1.5692 × 1.0637 × 0.3646 = 74.5 MPa
  5. F = σd × π × 2.5²/4 = 366 N; power = F v = 366 × 5 = 1.83 kW
  6. Check: flow stress of the drawn wire K εⁿ = 188 MPa > 74 MPa, so the pass works

Answer: σd ≈ 74 MPa, pull ≈ 366 N, about 1.8 kW per die

How far can one pass reduce a steel rod? · Annealed low-carbon steel (K 530 MPa, n 0.26); first ideal (μ = 0, φ = 1), then with α = 12° and μ = 0.07
  1. Ideal: Ȳf ε = K εⁿ ε/(1 + n) must stay below the exit flow stress K εⁿ, so ε ≤ 1 + n
  2. ε = 1.26, r = 1 − e^−1.26 = 71.6% (63.2% for a non-hardening metal)
  3. With friction: solve Ȳf (1 + μ/tan α) φ ε = K εⁿ by bisection (φ depends on r through Lc)
  4. Root at r = 61.8% (ε = 0.962), where 1 + μ/tan α = 1.329 and φ = 0.986

Answer: About 62% per pass in this die; work hardening lifts the ideal limit from 63% to 72%

Chapter 11

Sheet metal shearing: blanking, piercing, clearance and cutting force

Which syllabus covers thisAU ME3393 Unit IVVTU BME302 Module 3SPPU 202050 Unit 3MU MEC303 Module 3KTU PBMET304 Module 3

What happens to the sheet as a punch shears it?

As the punch comes down the sheet first bends and is pulled into the gap: this leaves a rounded rollover on the edge. Then the punch pushes metal into the die and the edge is smeared smooth and shiny: the burnished band, wider in soft, ductile metals. Cracks then start at the punch and die edges and, with the right clearance, meet cleanly: the rough, slightly tapered fracture zone. A small burr is left on the side away from the punch. Hard metals show less burnish and more fracture.

How much clearance, and is it taken off the punch or the die?

Clearance c is the gap per side between punch and die, set as c = a t with the allowance a by metal: about 0.045 for soft aluminium, 0.060 for harder aluminium, brass and soft steel, 0.075 for half-hard steel and stainless. Too little doubles the cracks and wears tools; too much gives big burrs and rollover. The tool that makes the finished edge sets the size: blanking keeps the blank, so the die equals the blank and the punch is 2c smaller; piercing keeps the hole, so the punch equals the hole and the die is 2c larger.

Blanking and piercing: clearance, punch and die sizes, cutting force and how shear on the punch lowers it: interactive loads as you scroll

How is the cutting force and press size found?

The punch shears the whole cut length L through the thickness t at once, so F = S t L with S the shear strength of the sheet. Where only the tensile strength is known, S is roughly 0.7 of it (about 0.65 for aluminium, 0.75 for steels: an approximation). A stripper must then pull the sheet off the punch: one empirical estimate is 0.02 L t kN (L, t in mm). With a spring stripper both act on the down stroke, so the press must carry their sum, plus a margin.

Which cutting operations does a press tool do?

Blanking cuts out a part and keeps it; piercing or punching makes a hole and the slug is scrap. Notching cuts a piece from the edge, lancing slits and bends without removing metal (louvres, tabs), trimming cuts flash or excess from a formed part, and parting or cut-off separates parts along a line. Shaving takes a thin second cut off a sheared edge to make it smooth and square. Fine blanking clamps the sheet with a V-ring and a counter-punch and uses a tiny clearance, so the whole edge is burnished in one stroke: used where a clean edge saves machining, such as small gears.

How can the cutting force be reduced?

Grind the punch or die face at a slight angle (shear) so the cut proceeds progressively along the edge instead of all at once: with full shear, about one sheet thickness of rise, the peak force drops by roughly 30%. Shear goes on the die for blanking and on the punch for piercing, so the part that is kept stays flat; double or concave shear keeps the force central. Stepped punches of different lengths, about one thickness apart, let groups of holes cut one after another. Both trade peak force for a slightly longer stroke.

Predict, then check

You blank 50 mm discs from 2 mm steel. Which tool is made exactly 50 mm?

Worked examples

Compound die for a steel washer · Washer OD 40 mm, ID 20 mm from 2 mm SAE 1010 hot-rolled steel (shear strength 35 ksi = 241.3 MPa; soft steel allowance 0.060)
  1. c = 0.060 × 2 = 0.12 mm per side
  2. Outside (blanking): die 40.00 mm, punch 40 − 2 × 0.12 = 39.76 mm
  3. Hole (piercing): punch 20.00 mm, die 20 + 2 × 0.12 = 20.24 mm
  4. Forces: blank S t π 40 = 60.6 kN; pierce S t π 20 = 30.3 kN; both in one stroke 91.0 kN
  5. Stripping 0.02 × (125.7 + 62.8) × 2 = 7.5 kN; press load 98.5 kN = 10.0 t

Answer: Punch 39.76 mm / die 40.00 mm outside, punch 20.00 / die 20.24 mm inside; about 99 kN

A rectangular blank, with and without shear · Blank 120 × 60 mm from 3 mm SAE 1020 cold-drawn steel (shear 43 ksi = 296.5 MPa, tensile 61 ksi; half-hard allowance 0.075)
  1. c = 0.075 × 3 = 0.225 mm; die 120 × 60 mm, punch 119.55 × 59.55 mm
  2. F = S t L = 296.5 × 3 × 360 = 320 kN
  3. Here S/UTS = 43/61 = 0.705, close to the 0.7 rule of thumb
  4. Full shear on the die (rise ≥ 3 mm): about 30% less, 224 kN
  5. Stripping 0.02 × 360 × 3 = 21.6 kN; press load 246 kN = 25.0 t

Answer: 320 kN flat, about 224 kN with full shear; press about 25 t

Same hole, two sheets · A 25 mm hole pierced in 2 mm sheet of 304 stainless (tensile 75 ksi) and of 5052-H32 aluminium (31 ksi); shear estimated as 0.75 and 0.65 of tensile
  1. 304: S ≈ 0.75 × 517.1 = 388 MPa; F = 388 × 2 × π × 25 = 60.9 kN
  2. 5052: S ≈ 0.65 × 213.7 = 139 MPa; F = 21.8 kN
  3. Ratio 2.79: the stainless hole needs almost three times the force
  4. Die openings: 25 + 2 × 0.060 × 2 = 25.24 mm (soft stainless), 25 + 2 × 0.045 × 2 = 25.18 mm (5052); punch 25.00 mm for both

Answer: 61 kN in stainless against 22 kN in aluminium (estimates from tensile strength)

Chapter 12

Presses, progressive and compound dies, centre of pressure and strip layout

Which syllabus covers thisAU ME3393 Unit IVVTU BME302 Module 3SPPU 202050 Unit 3MU MEC303 Module 3KTU PBMET304 Module 3

What kinds of press are used, and what do tonnage, stroke and shut height mean?

Mechanical presses turn a flywheel's rotation into ram motion through a crank (long strokes), an eccentric (short, stiff strokes) or a knuckle joint (huge force near the bottom, for coining and sizing). They are fast and deliver full force only near the bottom of the stroke. Hydraulic presses give full force anywhere in a slow, adjustable stroke, good for deep drawing. Tonnage is the safe force rating; stroke is the ram travel; shut height is the bed-to-ram distance with the stroke down and adjustment up, and the closed die must fit within it.

What are the parts of a die set, and how do simple, compound and progressive dies differ?

A die set has a die shoe and punch holder aligned by guide posts, the punch and die, a stripper that holds the sheet down and strips it off the punch, stock guides, stops and pilots that locate the strip at each stroke. A simple die does one cut per stroke. A compound die blanks and pierces at one station in one stroke, giving accurate concentric features. A progressive die does a sequence at successive stations as the strip advances one pitch per stroke. Combination dies cut and form at one station; transfer dies move separate blanks between stations.

What is the centre of pressure and why must the ram act there?

When several punches cut at once, the resultant cutting force acts at the centre of pressure: the centroid of all the cut edges, each weighted by its length (force per unit length is S t, the same everywhere). A circle acts at its centre, a straight edge at its midpoint. If the press ram's centre line is off this point, the die tilts, clearance closes on one side, punches chip and guide posts wear. Designers place the shank on the centre of pressure, and rearrange stations to bring it near the die centre.

How is a strip layout planned and judged?

Blanks are nested along a strip with a scrap bridge B between them, about 1.25 t for pitches under 64 mm and 1.5 t above, and a front (edge) scrap of about t + 0.015 H on each side, where H is the part width across the strip. The pitch is the part length plus B, and the strip width the part width plus both edge scraps. Utilisation is blank area times parts per pitch over pitch times width; aim for at least 75%. Round blanks gain by staggering two rows; odd shapes by turning or interlocking them.

Strip layout and centre of pressure: pitch, strip width, material utilisation and where the press force acts: interactive loads as you scroll
Predict, then check

A progressive die pierces a 20 mm hole at station 1 and blanks a 40 mm washer one pitch later. Where should the press ram's centre line be?

Worked examples

Strip for a 40 mm washer · Washer OD 40 mm, 1.5 mm sheet; bridge 1.25 t below 64 mm pitch; edge scrap t + 0.015 H
  1. B = 1.25 × 1.5 = 1.875 mm; pitch C = 40 + 1.875 = 41.875 mm (under 64 mm, so 1.25 t holds)
  2. Edge scrap A = 1.5 + 0.015 × 40 = 2.1 mm; strip width W = 40 + 2 × 2.1 = 44.2 mm
  3. Single row: utilisation = (π × 40²/4)/(41.875 × 44.2) = 67.9%
  4. Two staggered rows: centres 41.875 mm apart, rows 36.26 mm apart (C √3/2); W = 40 + 36.26 + 4.2 = 80.46 mm
  5. Utilisation = 2 × 1,256.6/(41.875 × 80.46) = 74.6%

Answer: 67.9% in one row, 74.6% staggered: the double row nearly reaches the 75% goal

Centre of pressure of the washer progressive die · Same washer, ID 20 mm: station 1 pierces the hole (x = 0), station 2 blanks the outside one pitch on (x = 41.875 mm); SAE 1010 steel, shear 241.3 MPa
  1. Cut lengths: pierce π × 20 = 62.83 mm, blank π × 40 = 125.66 mm
  2. x̄ = (62.83 × 0 + 125.66 × 41.875)/(62.83 + 125.66) = 27.92 mm from the pierce punch; ȳ on the strip centreline
  3. Total force = S t ΣL = 241.3 × 1.5 × 188.5 = 68.2 kN, centred at x̄
  4. A compound die cuts both circles in one place: its centre of pressure is the washer centre

Answer: x̄ ≈ 27.9 mm from the piercing punch, two-thirds of the way to the blanking punch

Centre of pressure with two off-centre holes · Plate 80 × 40 mm with 10 mm holes at (20, 20) and (65, 25), measured from the lower left corner; compound die
  1. Outline: 240 mm of cut, acting at the rectangle centre (40, 20)
  2. Each hole: π × 10 = 31.42 mm at its centre
  3. x̄ = (240 × 40 + 31.42 × 20 + 31.42 × 65)/302.83 = 40.52 mm
  4. ȳ = (240 × 20 + 31.42 × 20 + 31.42 × 25)/302.83 = 20.52 mm

Answer: (40.52, 20.52) mm: about half a millimetre off the plate centre in each direction

Chapter 13

Bending: bend allowance, bending force and springback

Which syllabus covers thisAU ME3393 Unit IVVTU BME302 Module 3SPPU 202050 Unit 3MU MEC303 Module 3KTU PBMET304 Module 4

How is sheet bent in a press, and what happens inside the bend?

Bending strains the sheet round a straight axis. In V-bending a V-shaped punch pushes the sheet into a V-die on a press brake: cheap tooling, any angle from obtuse to acute. In edge (wiping) bending a pad clamps the sheet while the punch wipes the overhang down over the die edge, the usual choice for high-volume parts. In air bending the punch stops short and the depth sets the angle; bottoming and coining press the sheet into the die to set it. Inside any bend the outer fibres stretch, the inner fibres compress, and one surface in between, the neutral axis, keeps its length.

How long must the flat blank be before bending?

The neutral axis keeps its length, so the flat blank is the straight lengths plus the arc length of the neutral axis in the bend, the bend allowance. The neutral axis is not at mid-thickness for tight bends: it moves towards the inside, about 0.33t from the inner surface when R < 2t, and sits at 0.5t for larger radii. A 90° bend in 2 mm sheet adds 5.75 mm at R = 3 mm but 11.00 mm at R = 6 mm. Get it wrong and every flange of a bracket comes out short or long.

How much press force does a bend need?

The maximum force is estimated as F = Kbf · UTS · w · t²/D, with w the bend length, t the thickness and D the die opening. Kbf is 1.33 for a V-die and 0.33 for a wiping die, so a V-die needs about four times the force for the same sheet and opening. Force grows with t², so 3 mm sheet needs 2.25 times the force of 2 mm sheet; a wider die opening cuts it. A 1.2 m bend in 3 mm steel over a 24 mm V-die needs about 218 kN.

Why does a bent part open up when the punch lifts, and how is that corrected?

Part of the bend is elastic, and that part recovers when the load comes off: the radius grows and the angle opens. For an elastic-perfectly-plastic sheet Ri/Rf = 4x³ − 3x + 1 with x = Ri·Y/(E·t), so springback rises with R/t and with Y/E. Aluminium 5052-H32 springs back about twice as much as 1010 steel in the same tooling, because its Y/E is nearly double. One lecture deck lists 1-2° for low-carbon steel and 10-15° for spring steel in a 90° bend. Fixes: overbend, bottom or coin the bend, or stretch the sheet while bending so the whole section yields.

Sheet metal bending: bend allowance, flat length, bending force and springback with the overbend needed: interactive loads as you scroll

What limits how tight a bend can be?

The outer surface stretches most. Its engineering strain is e = 1/(2R/t + 1), so it grows quickly as R/t falls, and the bend cracks once the outside runs out of ductility. A rough, conservative estimate equates e to the tensile elongation: 20% for 1010 steel gives R/t ≥ 2.0, 12% for 5052-H32 gives R/t ≥ 3.7. Real limits are often lower because the outer fibre is not necking like a tensile bar. Bend across the rolling direction where possible, deburr sheared edges on the outside of the bend, and choose a softer temper for tight bends.

Predict, then check

The same 2 mm sheet is bent to 90° over a 4 mm punch radius. Which material springs back the most?

Worked examples

Flat blank of an L-bracket · 2 mm aluminium 5052-H32, two legs of 40 mm measured to the bend tangents, one 90° bend; compare inside radius 3 mm and 6 mm
  1. R = 3 mm is below 2t = 4 mm, so k = 0.33
  2. BA = 2π (90/360)(3 + 0.33 × 2) = 5.749 mm; blank = 40 + 40 + 5.749 = 85.75 mm
  3. R = 6 mm is above 2t, so k = 0.5
  4. BA = 2π (90/360)(6 + 0.5 × 2) = 10.996 mm; blank = 91.00 mm

Answer: Cut the blank 85.75 mm long for R = 3 mm, 91.00 mm for R = 6 mm

Press brake force for a cabinet panel · 3 mm 1010 steel (UTS 365 MPa), bend length 1.2 m; V-die with an opening of 8t = 24 mm (illustrative); compare a wiping die
  1. V-die: F = 1.33 × 365 × 1,200 × 3²/24 = 2,18,452 N = 218.5 kN
  2. In tonnes-force: 218.5/9.81 = 22.3 t
  3. Wiping die, same opening: F = 0.33 × 365 × 1,200 × 9/24 = 54.2 kN

Answer: About 218 kN (22 t) on the press brake; 54 kN with a wiping die

Springback and the overbend needed · 1.5 mm aluminium 5052-H32 (Y 193 MPa, E 70.3 GPa) bent over a 15 mm punch radius; the drawing asks for 90°
  1. x = Ri Y/(E t) = 15 × 193/(70,300 × 1.5) = 0.02745
  2. Ri/Rf = 4x³ − 3x + 1 = 0.9177, so Rf = 15/0.9177 = 16.34 mm
  3. The neutral-axis length is kept: αf/αi = (Ri + t/2)/(Rf + t/2) = 15.75/17.09 = 0.9213
  4. Bent to 90° at the tool, the part relaxes to 82.9°
  5. To finish at 90°, bend to 90/0.9213 = 97.7° at the tool

Answer: It springs back 7.1°; overbend to about 97.7°

Chapter 14

Deep drawing, formability, forming limit diagrams and high energy rate forming

Which syllabus covers thisAU ME3393 Unit IVVTU BME302 Module 3SPPU 202050 Units 2, 3MU MEC303 Module 3KTU PBMET304 Module 4

What happens to the metal when a flat blank becomes a cup?

A punch pushes the centre of the blank into a die while a blankholder presses the outer ring, the flange, flat against the die face. The flange is pulled inwards, so its circumference shrinks: it is squeezed in the hoop direction and stretched radially, and it thickens slightly. It bends over the die corner radius, straightens into the wall, and the wall carries the whole punch force down to the bottom of the cup. Punch and die need generous corner radii (sharp ones would pierce the sheet) and a clearance of about 1.1t between them.

How big must the blank be for a given cup?

Thickness changes little in a plain draw, so the blank area equals the surface area of the cup. For a flat-bottomed cylinder of diameter d and height h with a sharp corner, πD²/4 = πd²/4 + πdh gives D = √(d² + 4dh). A bottom corner radius r removes a little area: the exact pieces (a disc, a quarter torus and the wall) give D² = d² + 4dh − 1.72dr − 0.57r². A 60 mm cup 40 mm deep needs a 114.9 mm blank with a sharp corner and 112.6 mm with a 5 mm radius. Add a trimming allowance in practice, because the rim comes out uneven.

How deep can one draw go, and what force does it take?

Three ratios judge a draw. The drawing ratio D/d should stay below about 2.0, equivalently the reduction (D − d)/D below 0.5; beyond that the wall tears before the flange is pulled in. The thickness ratio t/D should stay above 1%, or the flange wrinkles. Deeper cups are made in several draws, with typical limiting reductions of about 50%, 30% and 25% per stage, often with annealing between. The punch force is roughly F = π d t UTS (D/d − 0.7), and the blankholder pressure about 1.5% of the yield strength over the held flange area.

Deep drawing a cup: blank diameter, drawing ratio, number of draws, drawing and blankholder force: interactive loads as you scroll

What goes wrong in deep drawing?

Wrinkling: the flange buckles under hoop compression, worst with thin sheet (small t/D) and too little blankholder force; drawn in, the wrinkles show on the wall. Tearing: an open crack in the wall near the bottom, from too high a drawing ratio, too much holding force or a sharp die corner. Earing: wavy peaks round the rim, because rolled sheet has different properties in different directions; an isotropic sheet would not ear. Scratches come from rough tools or poor lubrication. Ironing, deliberately thinning the wall in a tight clearance, evens the wall and lengthens the cup.

How is formability measured, and what does a forming limit diagram show?

Cup tests stretch a clamped sheet over a ball punch until it cracks: the dome height at fracture (Erichsen, Olsen tests) ranks sheets for stretching. A forming limit diagram is more useful for press shops. A grid of small circles is printed on the blank; after forming they become ellipses, and their major and minor strains are plotted. The forming limit curve, measured for that sheet, separates safe strains below from necking and failure above. Stretching on both axes sits on the right of the plot, drawing (one strain negative) on the left, and the lowest point is near plane strain.

Which other sheet-forming processes are in the syllabus?

Stretch forming wraps a gripped sheet over a die, so the whole section yields and springback is small (aircraft skins). Spinning presses a rotating blank onto a mandrel with a roller, pass by pass. Rubber pad and hydroforming replace one rigid die half with rubber or a fluid-backed diaphragm, which allows deeper draws with less wrinkling. High energy rate forming uses explosives, a spark discharge in water (electrohydraulic) or a magnetic pulse to drive the sheet into a single die within milliseconds. Superplastic alloys are blown into dies slowly at high temperature, and incremental forming builds a shape with a CNC-driven stylus.

Predict, then check

A cup 40 mm in diameter and 100 mm deep (4 mm corner radius) needs a 132 mm blank. How many draws, with typical limits of 50%, 30%, 25%?

Worked examples

Blank size and severity of a cup · Cup: inside diameter 60 mm, height 40 mm, bottom corner radius 5 mm; sheet 1.2 mm
  1. Area by pieces: bottom disc of radius 25 mm, quarter torus (arc 7.854 mm at mean radius 28.183 mm), wall 35 mm high: A = 9,951.6 mm²
  2. D = √(4A/π) = 112.56 mm (sharp-corner formula: 114.89 mm)
  3. DR = 112.56/60 = 1.876 (below 2); reduction = 46.7% (below 50%)
  4. t/D = 1.2/112.6 = 1.07% (just above 1%)

Answer: A 112.6 mm blank; the cup can be drawn in one stage

Punch and blankholder forces · The same cup in aluminium 5052-O (UTS 193 MPa, yield 89.6 MPa), die corner radius 6 mm
  1. F = π × 60 × 1.2 × 193 × (1.876 − 0.7) = 51,342 N
  2. Held flange: outer diameter 112.56 mm, inner 60 + 2.2 × 1.2 + 2 × 6 = 74.64 mm
  3. Fh = 0.015 × 89.6 × π × (112.56² − 74.64²) = 29,977 N
  4. Fh/F = 0.58, above the 'about one third' rule of thumb: the held flange is wide compared with the thin wall

Answer: About 51 kN on the punch and 30 kN on the blankholder

A deep cup in three draws · Cup 40 mm diameter, 100 mm deep, 4 mm corner radius; typical limiting reductions 0.5, 0.3, 0.25
  1. Blank D = √(40² + 4 × 40 × 100 − 1.72 × 40 × 4 − 0.57 × 4²) ≈ 131.6 mm; D/d = 3.29
  2. Draw 1: 131.6 × 0.5 = 65.8 mm; at the same area the cup is 51.1 mm deep
  3. Draw 2: 65.8 × 0.7 = 46.1 mm, still above 40
  4. Draw 3: 46.1 × 0.75 = 34.5 mm would be enough, so draw to 40 mm

Answer: Three draws (with the typical limits); anneal between them if the sheet work-hardens

Chapter 15

Joining processes, gas welding and arc welding

Which syllabus covers thisAU ME3393 Unit IIVTU BME302 Module 4SPPU 202050 Unit 4MU MEC303 Module 2KTU PBMET304 Module 2

How are joining processes classified, and what joints and positions are there?

Fusion welding melts the edges, with or without filler (gas, arc, resistance, beam). Solid-state welding joins without melting, by pressure, friction or diffusion. Brazing and soldering melt only a filler with a lower melting point; adhesive bonding glues; mechanical fastening bolts or rivets, and hybrid joints combine two of these. The five basic joints are butt, lap, tee, corner and edge. Welding positions are flat, horizontal, vertical and overhead, coded 1G, 2G, 3G and 4G for groove welds (F for fillets). Overhead is hardest: the pool must be held up by surface tension and arc force.

How does an oxy-acetylene flame work, and which flame is used when?

Acetylene burns in two stages. In the bright inner cone, C₂H₂ + O₂ → 2CO + H₂, the hottest region at about 3,500 °C. In the outer envelope the CO and H₂ burn with oxygen from the air to CO₂ and H₂O. A neutral flame, with just enough oxygen for the first stage, is the normal welding flame. Excess acetylene gives a reducing (carburising) flame for hardfacing; excess oxygen an oxidising flame, used for some brasses and bronzes. The flame heats a wide area, so only 10-30% of its heat reaches the work. Gas cutting preheats steel, then a jet of oxygen burns it away.

What does an SMAW electrode coating do, and what does E7018 mean?

In shielded metal arc welding a flux-coated rod is both electrode and filler. The coating burns to a shielding gas, forms slag over the bead, stabilises the arc and can add alloy or iron powder. In E7018, E is an electrode, 70 means the weld deposit has at least 70 ksi (483 MPa) tensile strength, 1 means all positions and 8 a basic, low-hydrogen coating with iron powder for a faster deposit; such rods must be kept dry. With DC, the electrode is negative (DCEN, straight polarity) or positive (DCEP, reverse polarity); the choice moves heat between rod and work and changes penetration, and many low-hydrogen rods run DCEP.

How do GTAW, GMAW, FCAW, SAW and the other arc processes differ?

GTAW (TIG) strikes the arc from a tungsten electrode that does not melt, under argon or helium; filler is added separately, giving clean, precise welds on thin sheet. GMAW (MIG/MAG) feeds a wire electrode continuously under gas; metal crosses the arc by short-circuiting, globular, spray or pulsed transfer as current rises. FCAW uses a tubular wire filled with flux. SAW buries the arc under granular flux: high currents, deep welds on thick plate, no visible arc. Plasma arc constricts a tungsten arc through a nozzle, reaching very high temperatures. Electroslag welds thick sections vertically through molten slag; carbon arc and atomic hydrogen are older processes.

How much heat does an arc put into the joint, and how much metal does it melt?

Arc energy is V·I divided by travel speed; heat input multiplies it by a thermal efficiency factor k, 1.0 for SAW, 0.8 for SMAW, GMAW and FCAW and 0.6 for GTAW (relative to SAW, as used in welding procedures). At 24 V, 150 A and 240 mm/min, SMAW puts 0.72 kJ/mm into the joint. Measured arc efficiencies (NPTEL) are 21-48% for GTAW, 66-85% for SMAW and GMAW, 90-99% for SAW. Only part of the heat melts metal (the melting factor f2); the rest spreads into the plate. Melting one mm³ from room temperature needs about Um = 3.33 × 10⁻⁶ Tm² J (Tm in K): 10.3 J/mm³ for steel, 2.8 for aluminium.

Arc welding heat input: voltage, current, travel speed and arc efficiency, and how much metal it melts: interactive loads as you scroll
Predict, then check

SMAW and GTAW run at the same 24 V, 150 A and 4 mm/s. How does the GTAW heat input compare?

Worked examples

Heat input for a procedure sheet · SMAW, 24 V, 150 A, travel speed 240 mm/min
  1. Arc energy = V × I × 60/(1,000 × speed in mm/min) = 24 × 150 × 60/(1,000 × 240) = 0.90 kJ/mm
  2. Heat input = k × arc energy = 0.8 × 0.90 = 0.72 kJ/mm
  3. Same settings with GTAW (k = 0.6): 0.54 kJ/mm

Answer: 0.72 kJ/mm for SMAW

How much steel a MIG arc melts · GMAW on low-carbon steel (Tm 1,760 K), 28 V, 250 A, 6 mm/s; melting factor 0.5 (typical)
  1. Heat into the work q = 0.8 × 28 × 250 = 5,600 W; heat input = 5600/6 = 933 J/mm
  2. Um = 3.33 × 10⁻⁶ × 1,760² = 10.32 J/mm³
  3. Melting rate = 0.5 × 5,600/10.32 = 271 mm³/s
  4. Fused cross-section = 271/6 = 45.2 mm², more than the 32 mm² of an 8 mm fillet

Answer: About 271 mm³/s of steel, a fused section of 45 mm²

Why a gas torch is slow · Oxy-acetylene torch burning 0.5 m³ of acetylene per hour (illustrative); 55 MJ released per m³; heat transfer factor 0.2 (inside the 0.1-0.3 range)
  1. Heat released = 0.5 × 55 × 10⁶/3,600 = 7,639 W
  2. Reaching the work = 0.2 × 7,639 = 1,528 W
  3. Compare the MIG arc above: 5,600 W into the work, 3.7 times as much, on a far smaller spot

Answer: About 1.5 kW reaches the work: gas welding heats slowly and widely, which is why it suits thin sheet and repair

Chapter 16

Resistance welding: spot, seam and projection, and heat balance

Which syllabus covers thisAU ME3393 Unit IIVTU BME302 Module 5SPPU 202050 Unit 4MU MEC303 Module 2KTU PBMET304 Module 2

Where does the heat in resistance welding come from?

Two water-cooled copper electrodes squeeze the sheets and pass a large current through them for a fraction of a second. Heat H = I²Rt is generated wherever there is resistance: in the electrodes, in the sheets, at the electrode-sheet contacts and at the faying surface between the sheets. The electrodes are copper and cooled, so little heat stays there; the faying interface has the highest resistance and melts first, forming a nugget. Typical spot welding uses 5,000-20,000 A at under 10 V for 0.1-0.4 s, on sheets thinner than about 3 mm. No filler, flux or shielding gas is used.

What is a spot-weld schedule, and how big should the nugget be?

A schedule has four stages: squeeze (electrodes close and build force), weld (current on), hold (force kept while the nugget freezes) and off. Weld time is counted in mains cycles: on India's 50 Hz supply one cycle is 20 ms, so 10 cycles is 0.2 s. The usual target nugget diameter is 5√t with t the sheet thickness in mm, so 5 mm for 1 mm sheet, and TWI gives the minimum acceptable as 70-80% of that target. Too little current or time gives a small nugget that fails at the interface; too much gives expulsion of molten metal and deep electrode marks.

Resistance spot welding: I squared R t heat, weld time in cycles, nugget size and the energy actually used: interactive loads as you scroll

What is heat balance, and how is it kept with unequal sheets?

With equal electrodes, cooling is symmetric, so the hottest point is the centre of the stack, not necessarily the faying surface. Weld 1 mm to 2 mm sheet and that centre lies 0.5 mm inside the thicker sheet: the nugget grows there and may barely reach the thin one. Dissimilar metals shift heat too, because the higher-resistivity sheet heats faster. Fixes: a smaller face or a higher-resistance electrode alloy on the thick or more conductive side, a larger face on the thin side, or a projection pressed into the thicker sheet so the current is concentrated where the joint should form.

How do seam, projection and butt resistance welding differ from spot welding?

Seam welding replaces the electrodes with rotating copper wheels: overlapping spots, or a continuous run, give a leak-tight seam on tanks and cans. Projection welding embosses small bumps on one part; they concentrate the current, so several welds form at once, and weld nuts are fixed to sheet this way. Upset butt welding presses two bar ends together and heats them by their contact resistance. Flash butt welding first brings the ends together lightly so they arc and flash, burning off surface contamination, then upsets them hard and expels the molten metal, for rails, chains and rings.

Why does aluminium need much more current than steel?

Heat is I²R, and aluminium sheet has little R: 5052 has a resistivity of 0.0499 µΩ·m against 0.143 for 1010 steel, 2.9 times lower, and its thermal conductivity, 138 against 49.8 W/m·K, carries heat away 2.8 times faster. Its melting energy is lower (2.8 against 10.3 J/mm³), but that does not make up for it. Aluminium is welded with much higher currents in fewer cycles, and its tough oxide film makes contact resistance erratic, so electrode tips wear quickly and need frequent dressing.

Predict, then check

A spot weld comes out undersized. Which change adds more heat: doubling the weld time, or doubling the current?

Worked examples

Heat in a steel spot weld · Two 1 mm low-carbon steel sheets; 10 kA rms for 10 cycles at 50 Hz; electrode face 5 mm (illustrative); contact and interface resistance 60 µΩ (illustrative)
  1. Weld time = 10/50 = 0.2 s
  2. Bulk resistance of the 2 mm stack under a 5 mm face = 1.43 × 10⁻⁷ × 0.002/(π × 0.005²/4) = 14.57 µΩ
  3. R = 14.57 + 60 = 74.57 µΩ
  4. H = I²Rt = (10,000)² × 74.57 × 10⁻⁶ × 0.2 = 1,491 J

Answer: About 1,490 J of heat

How much of that heat makes the nugget · Same weld; target nugget D = 5√t, height 1.6 mm (illustrative), treated as an oblate spheroid; Tm 1,760 K
  1. D = 5√1 = 5 mm; volume = π × 5² × 1.6/6 = 20.94 mm³
  2. Um = 3.33 × 10⁻⁶ × 1,760² = 10.32 J/mm³; melt energy = 216 J
  3. Share = 216/1,491 = 14.5%; the rest heats the sheet round the nugget and the water-cooled electrodes
  4. TWI minimum acceptable nugget: 70% × 5 = 3.5 mm

Answer: Only about 14% of the heat ends up in the 5 mm nugget

Aluminium needs about three times the current · Two 1 mm aluminium 5052 sheets; same face; contact resistance 15 µΩ (illustrative); deliver the same 1,490 J in 4 cycles
  1. Bulk resistance = 4.99 × 10⁻⁸ × 0.002/(π × 0.005²/4) = 5.08 µΩ; R = 20.08 µΩ
  2. Time = 4/50 = 0.08 s
  3. I = √(H/(R t)) = √(1491/(20.08 × 10⁻⁶ × 0.08)) = 30.5 kA

Answer: About 30 kA, three times the steel current, and aluminium's higher conductivity drains heat faster still

Chapter 17

Solid-state and beam welding: friction, friction stir, electron beam, laser and thermit

Which syllabus covers thisAU ME3393 Unit IIVTU BME302 Module 5SPPU not in syllabusMU MEC303 Module 2KTU PBMET304 Module 2

How can two metals be welded without melting either of them?

Atoms bond whenever clean metal surfaces are pushed to within atomic distance. Oxide films and grease keep them apart, so solid-state processes break those films with pressure, plastic flow, heat below the melting point, or time. Forge welding hammers hot parts together; cold (pressure) and roll welding squeeze soft metals such as aluminium and copper at room temperature; diffusion bonding holds polished parts hot under light pressure for hours; explosive welding drives one plate onto another; ultrasonic welding scrubs thin foils and wires together with high-frequency vibration. No fusion zone means no porosity or solidification cracks, and dissimilar metals can often be joined.

How do friction welding and friction stir welding work?

In rotary friction welding one part spins against the other under axial force; rubbing heats the interface to a plastic state, the spin stops and a forging push expels the oxide-laden flash. Inertia welding stores the energy in a flywheel instead of a driven motor. Friction stir welding, commercialised at TWI in 1991, plunges a rotating tool (a broad shoulder with a smaller pin) into a butt joint and traverses it, stirring softened metal across without melting. On the advancing side the tool surface moves with the travel direction; on the retreating side, against it. Aluminium train roof panels and Airbus A380 fuselage panels are friction stir welded.

Why do electron and laser beams make deep, narrow welds with a small HAZ?

What matters is power density, the power reaching the work divided by the spot area. A point source needs about 10 W/mm² just to melt steel; arcs run near 50-100 W/mm²; focused electron and laser beams reach 10⁴-10⁵ W/mm². Under a planar flux the time to melt falls as 1/q², so a beam melts in microseconds while heat has spread only a few micrometres sideways. Near 10⁴ W/mm² metal vaporises and a vapour keyhole carries the beam deep, giving depth-to-width ratios above 10. EBW needs a vacuum and X-ray shielding; LBW (CO2, Nd:YAG, fibre) works in air.

Power density of welding heat sources: oxy-fuel to electron beam, time to melt and when a keyhole forms: interactive loads as you scroll

What is thermit welding, and why is it used to join rails?

Powdered iron oxide and aluminium react: aluminium takes the oxygen, leaving molten iron and alumina slag at up to about 2,500 °C. The mixture releases about 3,956 J per gram. Two rail ends are set with a gap inside a refractory mould, preheated, and the crucible's molten steel is tapped into the gap, fusing both ends; the slag floats on top and is trimmed after cooling. No power supply or heavy machine is needed on site, which is why it suits track work in the field. Its first commercial use was welding tram tracks in Essen in 1899.

Predict, then check

An arc and a laser each deliver 3 kW to a steel plate, the arc over a 7 mm spot and the laser over a 0.4 mm spot. By the planar estimate, how much faster does the laser melt the surface?

Worked examples

Arc versus laser on the same steel plate · 3 kW net into a thick carbon steel plate at 25 °C: an arc over a 7 mm spot and a laser over a 0.4 mm spot (illustrative); steel k = 50 W/m K, ρ = 7,800 kg/m³, c = 475 J/kg K, melting at 1,510 °C (typical)
  1. Arc: q = 3,000/(π × 7²/4) = 78.0 W/mm²; laser: q = 3,000/(π × 0.4²/4) = 23,873 W/mm²
  2. t = π k ρc (Tm − T0)²/(4q²) with k = 0.05 W/mm K, ρc = 0.00370 J/mm³K, ΔT = 1,485 °C
  3. Arc: t = 52.8 ms (Eagar's steel rule (1800/H.I.)² gives 53.3 ms); laser: t = 0.56 µs
  4. Heat spreads √(αt) with α = 13.5 mm²/s: 0.84 mm for the arc, 2.8 µm for the laser

Answer: Arc ≈ 53 ms, laser ≈ 0.56 µs; the laser is in the keyhole range (> 10⁴ W/mm²)

Electron-beam power density · Accelerating voltage 60 kV, beam current 50 mA, 90% of the beam power reaches the work (f1 is 0.8-0.95), focused to a 0.5 mm spot
  1. Beam power E × I = 60,000 × 0.050 = 3,000 W; into the work 0.9 × 3,000 = 2,700 W
  2. Spot area π × 0.5²/4 = 0.1963 mm²
  3. Power density = 2,700/0.1963 = 13,751 W/mm², above 10⁴ W/mm²
  4. Planar time to melt steel ≈ 1.7 µs: the metal vaporises and a keyhole forms

Answer: About 13,751 W/mm²: keyhole mode

How much thermit for a rail joint · A joint needs 10 kg of iron from the reaction (illustrative); Fe2O3 + 2Al → 2Fe + Al2O3; standard atomic masses Fe 55.845, O 15.999, Al 26.982; energy 3,956 J/g of mixture
  1. Molar mass of Fe2O3 = 2 × 55.845 + 3 × 15.999 = 159.687 g/mol, giving 2 × 55.845 = 111.69 g of iron
  2. Aluminium needed = 2 × 26.982/159.687 = 0.338 kg per kg of oxide
  3. Mixture per kg of iron = (159.687 + 53.964)/111.69 = 1.913 kg, so 19.1 kg for 10 kg of iron
  4. Heat released = 19.13 kg × 3.956 MJ/kg = 75.7 MJ

Answer: About 19.1 kg of mixture (14.3 kg oxide + 4.8 kg aluminium), releasing about 76 MJ

Chapter 18

Weld metallurgy, HAZ, distortion, defects, inspection, brazing and soldering

Which syllabus covers thisAU ME3393 Unit IIVTU BME302 Module 5SPPU 202050 Unit 4MU MEC303 Module 2KTU PBMET304 Module 2

What zones does a fusion weld in steel have?

The fusion zone melted and froze as a small casting, with columnar grains growing from the fusion line. Next to it a thin partially melted band. Then the heat-affected zone (HAZ), never melted but heated enough to change: just outside the fusion line the austenite grains grow coarse, further out (just above A3) they refine, then an intercritical band between A1 and A3, and a tempered band below A1 (727 °C). Beyond lies unaffected base metal. The cooling time from 800 to 500 °C (t8/5) decides what forms: short times give hard martensite, long ones softer ferrite and pearlite.

What is the carbon equivalent, and when does a steel need preheat?

Hydrogen cracking needs three things at once: hydrogen from damp flux or grease, a hard HAZ, and stress. The carbon equivalent sums how strongly each element hardens the HAZ, carbon counting fully, manganese a sixth. TWI's guidance: below CE 0.40 a steel is generally safe with low-hydrogen practice; from 0.40 to 0.50 thin sections need nothing but thick ones need low preheat; above 0.50 preheat, low-hydrogen consumables and post-weld heating. Thick, high-CE steels may need about 200 °C. Preheat slows cooling, so hydrogen escapes and the HAZ is softer.

Weldability and the heat-affected zone: carbon equivalent, preheat, peak temperature and HAZ width: interactive loads as you scroll

How far does the HAZ reach, and how fast does it cool?

Each point beside the weld heats to a peak and cools. For a single pass through a thin plate, Adams' equation gives the peak at distance Y from the fusion line: it falls from the melting point at Y = 0, and the HAZ ends where the peak drops to A1. Its width grows in proportion to the net heat input per unit length and shrinks with thickness and ρc. For cooling, EN 1011-2 gives t8/5 for 3-D heat flow (thick plate, independent of thickness) and 2-D flow (thin plate, ∝ Q²/d²); the larger of the two applies. Preheat both widens the HAZ and lengthens t8/5.

Why do welds distort and leave residual stress, and how is it controlled?

The hot weld expands but is held by cold metal around it, so it yields in compression; on cooling it wants to shrink more than its surroundings allow. The result is tensile residual stress near yield along the weld, balanced by compression further out, and distortion: transverse shrinkage across the joint, longitudinal shrinkage and bowing along it, and angular distortion when a V-groove has more metal at the top. Remedies: smallest adequate weld, double-V and balanced sequences, back-step and intermittent welding, presetting parts at an angle, clamping, and stress-relief heat treatment.

What are the common weld defects, and how are they found?

Porosity: gas trapped by damp consumables or poor shielding. Slag inclusion: slag left between passes. Lack of fusion: weld metal not bonded to the side wall or previous pass. Incomplete penetration: root not fused. Undercut: a groove melted into the plate at the weld toe by too much current or speed. Cracks: hot (solidification) cracks in the weld, cold (hydrogen) cracks in the HAZ hours later. Spatter: droplets thrown out. Inspection runs from visual checks, to dye penetrant for surface cracks, magnetic particle for near-surface flaws in ferromagnetic steel, ultrasonic and radiography for buried ones.

How do brazing, braze welding and soldering differ from welding?

None of them melts the base metal: a lower-melting filler wets the surfaces and bonds to them. The American Welding Society calls it brazing when the filler melts above 450 °C, and soldering below; soft solders melt below about 400 °C (63/37 tin-lead at 183 °C, lead-free tin-silver-copper at about 217-220 °C). Brazing and soldering draw the filler into a close-fitting gap by capillary action, so clearance matters. Braze welding builds a fillet with a bronze rod and no capillary flow. Adhesive bonding joins with a polymer instead, spreading load over the whole overlap.

Predict, then check

On a thin plate welded in one pass, you double the net heat input and keep everything else the same. What happens to the HAZ width predicted by Adams' equation?

Worked examples

Carbon equivalent of a structural plate · Mill certificate (illustrative): C 0.18, Mn 1.35, Cr 0.15, Mo 0.05, V 0.02, Ni 0.20, Cu 0.25 wt%
  1. CE = C + Mn/6 + (Cr + Mo + V)/5 + (Ni + Cu)/15
  2. = 0.18 + 1.35/6 + 0.22/5 + 0.45/15 = 0.18 + 0.2250 + 0.0440 + 0.0300
  3. = 0.479: in the 0.46-0.50 band, rated 'fair' in the IIW weldability table
  4. TWI band 0.40-0.50: thin sections without preheat, thicker sections need low preheat and low-hydrogen electrodes

Answer: CE ≈ 0.48; preheat thick sections and use baked low-hydrogen consumables

HAZ width with and without preheat · 10 mm steel plate, single full-penetration pass, net heat input 1.2 kJ/mm (illustrative); ρc = 3.705 J/cm³K (7,800 × 475), Tm = 1,510 °C (typical), HAZ edge at A1 = 727 °C
  1. Adams: Y = (1/(Tp − T0) − 1/(Tm − T0)) × H/(4.13 ρc d), with 4.13 ρc d = 0.15312 J/mm²K
  2. Plate at 20 °C: Y = (1/707 − 1/1,490) × 1,200/0.15312 = 5.83 mm
  3. Preheated to 150 °C: Y = (1/577 − 1/1,360) × 1,200/0.15312 = 7.82 mm
  4. Check: 2 mm from the fusion line the peak is 1100 °C (plate at 20 °C), well inside the HAZ

Answer: HAZ ≈ 5.8 mm cold, ≈ 7.8 mm with 150 °C preheat

Cooling time t8/5 for a 12 mm plate · Effective heat input Q = 1.5 kJ/mm, bead on a 12 mm plate (F2 = F3 = 1), plate at 20 °C and then preheated to 150 °C
  1. 3-D: t = (6700 − 5T0) Q (1/(500 − T0) − 1/(800 − T0)) = 7.9 s at 20 °C
  2. 2-D: t = (4300 − 4.3T0) × 10⁵ (Q/d)² (1/(500 − T0)² − 1/(800 − T0)²) = 17.8 s at 20 °C
  3. The larger value is valid: 2-D, because 12 mm is below the transition thickness 18.0 mm
  4. At 150 °C: 3-D 11.8 s, 2-D 33.1 s, so t8/5 = 33.1 s

Answer: t8/5 ≈ 18 s cold, ≈ 33 s with preheat: slower cooling, softer HAZ

Chapter 19

Plastics and injection moulding

Which syllabus covers thisAU ME3393 Unit VVTU not in syllabusSPPU 202050 Unit 5MU MEC303 Module 6KTU not in syllabus

What kinds of polymers are moulded, and how do they differ?

Thermoplastics (PP, ABS, nylon, PC) soften on heating and harden on cooling, again and again, so they are injection moulded and recycled. Thermosets (phenolics, epoxies) cross-link irreversibly as they cure and are compression or transfer moulded. Elastomers are lightly cross-linked and rubbery. Among thermoplastics, amorphous ones (ABS, PC) soften gradually above their glass transition, while semicrystalline ones (PP, nylon) have a sharp melting point and pack into crystals as they freeze, so they shrink more: typically 1-2% for PP against 0.4-0.8% for ABS.

What happens in one injection moulding cycle, and which step takes longest?

The mould closes and clamps. The screw drives forward as a plunger and fills the cavity with melt in a short burst. It then holds pressure, packing in extra melt as the part shrinks, until the gate freezes. Meanwhile the screw rotates back to plasticise the next shot. The part cools against the water-cooled mould until its core is stiff enough to eject, the mould opens, ejector pins push the part out and the mould closes again. Cooling usually dominates, and it grows with the square of wall thickness: thin, uniform walls make fast cycles.

Injection moulding: clamp force, cooling time, cycle time and parts per hour for common polymers: interactive loads as you scroll

What are the parts of an injection moulding machine, and how big a clamp is needed?

Granules fall from the hopper into a barrel with heater bands, where a rotating, reciprocating screw melts them mostly by shear and carries the melt forward; the nozzle seats against the mould's sprue bushing. Older plunger (ram) machines just push pellets past a heated torpedo, with poorer mixing and temperature control. The clamp unit, toggle or hydraulic, must hold the mould shut against the cavity pressure acting on the projected area of all parts and runners, plus a margin of 20-30%. Machines are rated by this clamp force, most between 50 and 250 tonnes.

What are sprues, runners and gates, and why use a hot runner?

Melt enters through the sprue, flows along runners to each cavity and passes through a small gate into the part. The gate is narrow so it freezes first, sealing the cavity after packing, and so the runner snaps off cleanly. Balanced runners fill every cavity at the same time. In a cold-runner mould the runner freezes and is ejected with the parts every cycle, then reground. A hot-runner mould keeps the runner molten in a heated manifold, saving that plastic and often shortening the cycle, at a higher tool cost. Gate position decides where weld lines and air traps fall.

What are the common moulding defects, and what causes them?

Short shot: the cavity does not fill, from too little shot, low pressure, a cold melt or blocked vents. Flash: a thin fin at the parting line when cavity pressure overcomes the clamp or the mould is worn. Sink marks: dimples over thick sections that shrink after the skin has frozen, cured by more packing or thinner walls. Warpage: uneven shrinkage from varying wall thickness, uneven mould cooling or fibre orientation. Weld lines: weak, visible seams where two flow fronts meet. Burn marks come from air trapped and compressed at the end of fill.

Predict, then check

A part's wall is thickened from 2 mm to 3 mm, same polymer and temperatures. What happens to the cooling time?

Worked examples

Clamp force for a two-cavity PP lid · Each lid projects 150 cm² onto the parting plane, 2 cavities, runner adds 10% (illustrative); PP cavity pressure 250-400 kg/cm² (typical), use 325
  1. Projected area = 150 × 2 × 1.10 = 330 cm²
  2. Force = 330 cm² × 325 kg/cm² = 1,07,250 kgf = 107.2 t (1,052 kN; 325 kg/cm² = 31.9 MPa)
  3. Add a 20% margin: 107.2 × 1.2 = 128.7 t

Answer: About 107 t, so choose a machine of at least 129 t

Cooling time of an ABS housing wall · ABS, wall s = 2.5 mm, melt 230 °C, mould 50 °C, ejection 110 °C, α = 0.12 mm²/s (typical values)
  1. s²/(π²α) = 6.25/(9.8696 × 0.12) = 5.277 s
  2. ln[(4/π)(230 − 50)/(110 − 50)] = ln(3.820) = 1.3402
  3. t = 5.277 × 1.3402 = 7.07 s
  4. At 2.0 mm it would be 4.53 s: (2.5/2)² = 1.5625 times less wall, 1.5625 times less time

Answer: About 7.1 s of cooling

Cycle time and output for the PP lid · PP lid of example 1 with a 2.0 mm wall, melt 250 °C, mould 50 °C, ejection 105 °C, α = 0.09 mm²/s; fill 1 s and open-eject-close 3 s (illustrative); hold overlaps cooling
  1. Cooling: (4/(π² × 0.09)) × ln[(4/π)(200/55)] = 4.503 × 1.5325 = 6.90 s
  2. Cycle = 1 + 6.90 + 3 = 10.90 s
  3. Output = 3,600/10.90 × 2 cavities = 660 lids per hour

Answer: Cycle ≈ 10.9 s, about 660 lids per hour

Sizing the cavity for shrinkage · An 80.00 mm long part; linear shrinkage PC 0.5-0.7% (use 0.6%), PP 1.0-2.0% (use 1.5%), shrinkage taken as a fraction of the cavity size
  1. PC: cavity = 80.00/(1 − 0.006) = 80.48 mm
  2. PP: cavity = 80.00/(1 − 0.015) = 81.22 mm
  3. The two cavities differ by 0.74 mm, so a tool cut for PC moulds a PP part about 0.72 mm short

Answer: PC cavity ≈ 80.48 mm, PP cavity ≈ 81.22 mm

Chapter 20

Plastic extrusion, blow, rotational, compression and transfer moulding and thermoforming

Which syllabus covers thisAU ME3393 Unit VVTU not in syllabusSPPU 202050 Unit 5MU MEC303 Module 6KTU not in syllabus

What are the parts and zones of a plastics extruder?

Pellets fall from a hopper into a heated barrel, where a rotating screw carries them forward. The feed zone has deep flights and conveys solid pellets. In the compression zone the screw root grows, squeezing the pellet bed against the hot barrel so it melts, mostly by shear heating. The shallow metering zone pumps a uniform melt at steady pressure. A breaker plate and screen pack filter the melt and straighten its spiral motion before the die shapes it. Screws are described by diameter D and length-to-diameter ratio, typically 20 to 30.

How much does the screw pump, and where does it settle against a die?

Unroll the metering channel: the barrel slides over it and drags melt down the channel (drag flow, proportional to speed N and depth dc). The pressure the die demands pushes melt back (pressure flow, proportional to dc³/η). So output falls in a straight line from Qmax at open discharge to zero at pmax, the closed-discharge pressure. The die passes Q = Ks p, a line through the origin, and the operating point is where they cross. With one Newtonian viscosity, η cancels from the output: a stiffer melt raises head pressure, not flow.

Plastic extruder and die characteristics: drag flow, back pressure flow and the operating point: interactive loads as you scroll

How are pipe, sheet, film and cable coating extruded?

Change the die and the line after it. Pipe: an annular die with a mandrel, then a vacuum sizing sleeve and a water bath. Sheet: a wide coat-hanger die and chill rolls; calendering instead squeezes melt between heated rolls, as for PVC sheet. Blown film: an annular die extrudes a tube that air inflates into a bubble. The blow-up ratio is bubble diameter over die diameter, and hauling off faster than the melt leaves the die thins it further (draw-down). Cable: a crosshead die wraps melt round a moving wire.

How do blow moulding and rotational moulding make hollow parts?

Extrusion blow moulding drops a tube (the parison) between two mould halves, pinches it shut and blows it against the cold mould: bottles, cans, fuel tanks, each with a pinch-off scar. Injection blow moulding first moulds a preform with a finished neck, then blows it; stretch blow moulding of PET also stretches it lengthwise with a rod for strength and clarity. Rotational moulding loads powder into a closed metal mould that turns slowly about two axes inside an oven, coating the walls, then cools: large, seamless, low-stress water tanks with cheap tooling but long cycles.

What are compression moulding, transfer moulding and thermoforming used for?

Thermosets cure instead of melting, so they are shaped hot and held until set. Compression moulding puts a weighed charge (powder, preform or sheet moulding compound) in a heated open mould and the press closes to shape and cure it: switchgear parts, melamine tableware. Transfer moulding heats the charge in a pot first, then a plunger forces it through sprues into a closed cavity, so delicate inserts survive: encapsulated electronics. Thermoforming softens a thermoplastic sheet and draws it onto a mould by vacuum, air pressure or a plug: trays, blister packs, refrigerator liners. Deeper draws mean thinner walls.

Predict, then check

An extruder feeds a die with a very small hole. You fit a screw whose metering flights are twice as deep, everything else the same. What happens to the output?

Worked examples

Operating point of an HDPE rod extruder · Screw D = 60 mm, square pitch (A = 17.66°), metering depth 3.5 mm, metering length 480 mm, 50 rpm; HDPE at 220 °C, power law K = 1,595 Pa sⁿ, n = 0.74 (measured data); rod die 8 mm diameter, 30 mm land; one Newtonian η at the channel shear rate
  1. Channel shear rate πDN/dc = π × 0.060 × (50/60)/0.0035 = 44.9 s⁻¹, so η = 1,595 × 44.9^(0.74 − 1) = 593 Pa s
  2. Qmax = ½π² D² N dc sin A cos A = 14.98 cm³/s
  3. pmax = 6πDNηL cot A/dc² = 68.8 MPa
  4. Die: Ks = πDd⁴/(128 η Ld) = 5.649 × 10⁻¹² m³/(s Pa)
  5. Crossing: p = Qmax/(Ks + Qmax/pmax) = 2.55 MPa, Q = Ks p = 14.42 cm³/s
  6. Back-flow is only 3.7% of the drag flow: this die is far from restrictive

Answer: About 14.4 cm³/s at a head pressure of 2.6 MPa

Blow-up and draw-down in blown film · Illustrative line: annular die 150 mm diameter with a 1.2 mm gap; bubble 450 mm diameter; film 40 μm thick
  1. Blow-up ratio = 450/150 = 3
  2. Lay-flat width of the collapsed tube = π × 450/2 = 707 mm
  3. Volume constancy: gap × π Ddie × v_die = t × π Dbubble × v_haul, so v_haul/v_die = 1.2 × 150/(0.040 × 450) = 10
  4. The film is stretched 3 times round the bubble and 10 times along it; the ratio of the two sets the balance of strength in the two directions

Answer: BUR 3, lay-flat 707 mm, haul-off 10 times the die exit speed

Wall thinning in a vacuum-formed cup · Illustrative: a 2.0 mm sheet clamped over a 100 mm diameter opening is drawn into a cylindrical female mould 50 mm deep
  1. Sheet area inside the clamp = π × 100²/4 = 7,854 mm²
  2. Formed area = base + wall = 7,854 + π × 100 × 50 = 23,562 mm²
  3. Volume is conserved, so average thickness = 2.0 × 7,854/23,562 = 0.67 mm (area ratio 3)
  4. The sheet touches the wall first and freezes there, so the base corners end up thinner than this average: plug assist or a male mould redistributes it

Answer: Average wall about 0.67 mm, a third of the sheet

Chapter 21

Manufacturing of composites: lay-up, filament winding, RTM and pultrusion

Which syllabus covers thisAU not in syllabusVTU not in syllabusSPPU 202050 Unit 6MU not in syllabusKTU not in syllabus

What is a composite, and why make one?

A composite combines a reinforcement, usually strong, stiff fibres, with a matrix that holds them in place, passes load between them and protects them. Polymer-matrix composites (glass or carbon in epoxy or polyester) are the most common; metal-matrix composites put ceramic particles or fibres in aluminium; ceramic-matrix and carbon-carbon composites survive temperatures that would melt metals. The gain is strength and stiffness per kilogram along the fibres. How stiffness splits between fibre and matrix is the rule of mixtures, which has its own interactive in the Engineering Materials and Metallurgy book.

How do glass, carbon and aramid fibres compare, and in what forms are they supplied?

From data sheets: E-glass about 3,445 MPa, 76 GPa, 2.58 g/cm³, cheap and the default; carbon T700S 4,900 MPa, 230 GPa, 1.80 g/cm³; aramid Kevlar 49 3,600 MPa (resin-impregnated strand), 124 GPa, 1.44 g/cm³, tough but weak in compression. Per kilogram, T700S is 2.0 times as strong as E-glass and 4.3 times as stiff. Fibres come as continuous roving for winding and pultrusion, chopped-strand mat, woven fabric, and prepreg: fabric or tape already wetted with part-cured resin and kept frozen until use.

How do hand lay-up, spray-up, vacuum bagging and autoclave curing differ?

Hand lay-up: gel coat on an open mould, then layers of mat or fabric wetted with resin and rolled to drive out air. Tooling is cheap and size is unlimited (boat hulls, water tanks), but quality depends on the worker. Spray-up: a gun chops roving and sprays it with resin, faster but less controlled. Vacuum bagging seals a film over the wet lay-up and pulls a vacuum, pressing the plies together with up to one atmosphere and drawing off excess resin. Prepreg parts are bagged and cured in an autoclave, a heated pressure vessel, giving the low-void laminates aircraft need.

How does filament winding work, and why 54.7 degrees?

Resin-wet roving is laid on a rotating mandrel by a carriage moving along it: hoop windings near 90°, helical windings at ±α, polar windings round the ends. For pipes and pressure vessels, netting analysis lets only the fibres carry load. A closed cylinder's hoop resultant pR is twice its axial pR/2, and a band at angle α carries cos²α of its strength axially and sin²α round the hoop. Matching 2 to 1 needs tan²α = 2, α = 54.7°. Real vessels wind low-angle helicals to pass round the domes and add hoop layers; the total fibre needed comes out the same.

Filament winding a pressure vessel: netting analysis, the 54.7 degree winding angle and the fibre thickness: interactive loads as you scroll

What are RTM, vacuum infusion, SMC compression moulding and pultrusion?

Resin transfer moulding (RTM) places a dry fibre preform in a closed, matched mould and pumps resin in under pressure: a good finish on both faces at medium volumes. Vacuum infusion uses one rigid mould and a bag, and the vacuum pulls resin through the dry stack: large parts such as wind-turbine blades. Sheet moulding compound (chopped glass in thickened polyester) is compression moulded in heated steel dies in minutes: vehicle body panels. Pultrusion pulls rovings through a resin bath and a heated die, curing a constant section continuously: rods, gratings, ladder rails, rebar.

How are metal-matrix, ceramic-matrix and carbon-carbon composites made?

Metal-matrix: stir casting mixes ceramic particles (silicon carbide, alumina) into molten aluminium and casts it, the cheapest route; the powder route blends metal and ceramic powders and sinters them; squeeze infiltration forces molten metal into a porous fibre preform under pressure. Ceramic-matrix: chemical vapour infiltration (CVI) deposits matrix from a gas inside a fibre preform, slowly; polymer infiltration and pyrolysis (PIP) soaks the preform in a ceramic-forming polymer and bakes it, several times over. Carbon-carbon starts as carbon fibre in resin, then is carbonised and re-densified: aircraft brake discs and rocket nozzles.

Predict, then check

A closed cylindrical vessel is wound with ±α helical layers only, and only the fibres carry load. Which angle needs the least fibre?

Worked examples

Fibre thickness of a helical-wound cylinder · Cylinder radius 150 mm, working pressure 20 MPa, burst pressure taken as 3 × working (illustrative ratio) = 60 MPa; helical layers only at the optimum angle; netting analysis with full fibre strength
  1. Optimum angle: tan²α = 2, α = 54.7°; fibre thickness t = 3pR/(2σ)
  2. Carbon T700S (4,900 MPa): t = 1.5 × 60 × 150/4,900 = 2.76 mm; at 60% fibre volume the laminate is about 4.6 mm
  3. E-glass (3,445 MPa): t = 1.5 × 60 × 150/3,445 = 3.92 mm
  4. Fibre mass per metre = 2πRtρ: carbon 4.67 kg/m (ρ 1.80), glass 9.53 kg/m (ρ 2.58)

Answer: Carbon 2.76 mm and 4.7 kg/m of fibre; glass 3.92 mm and 9.5 kg/m, about twice the mass

Low-angle helicals plus hoop layers · Same cylinder and 60 MPa burst pressure, carbon T700S; helicals wound at ±15° so they pass round the domes; hoop layers make up the rest
  1. Geodesic (Clairaut) winding, R sin α = r0: the helicals turn round a polar opening of radius 150 × sin 15° = 38.8 mm
  2. Axial: t_h = pR/(2σ cos²α) = 60 × 150/(2 × 4,900 × cos²15°) = 0.984 mm
  3. Hoop: t_90 = pR/σ − t_h sin²α = 1.8367 − 0.0659 = 1.771 mm
  4. Total 2.755 mm, identical to the 54.7° design; helicals alone at 15° would need 27.4 mm because the hoop direction governs

Answer: Helical 0.98 mm + hoop 1.77 mm = 2.76 mm of fibre

Rovings for a pultruded carbon rod · Rod 10 mm diameter, target fibre volume 60%; T700S 12K roving: yield 800 g per 1,000 m, density 1.80 g/cm³, 7 μm filaments
  1. Fibre cross-section per roving = mass per metre ÷ density = 0.800/1.80 = 0.4444 mm²
  2. Rod area = π × 10²/4 = 78.54 mm²; rovings needed = 0.60 × 78.54/0.4444 = 106.0
  3. With 106 rovings, Vf = 106 × 0.4444/78.54 = 60.0%
  4. Check from the filaments: 12,000 × π × (0.0035 mm)² = 0.462 mm², within 4% of the yield-based value

Answer: 106 rovings of 12K T700S give Vf ≈ 60%

Chapter 22

Powder metallurgy: powders, compaction and sintering

Which syllabus covers thisAU not in syllabusVTU not in syllabusSPPU not in syllabusMU MEC303 Module 6KTU PBMET304 Module 1

What is powder metallurgy, and when is it the right choice?

Powder metallurgy (PM) presses metal powder into shape in a die, then sinters it below the melting point so the particles bond. The steps: make the powder, blend it with lubricant and alloying additions, compact, sinter, and finish if needed. It wins for many identical small parts made to near-net shape with little scrap: gears, cams, oil-pump rotors, self-lubricating bushes. It also makes what melting cannot: controlled porosity, and tungsten carbide-cobalt cutting tools. Limits: dies need volume to pay off, parts must eject from the die, and the remaining pores lower strength.

How are metal powders made and characterised?

Water atomisation breaks a molten stream with high-pressure water jets into irregular particles; gas atomisation gives spheres. Reducing iron oxide gives porous sponge iron; electrolysis gives pure, dendritic copper; mechanical alloying ball-mills powders together. Powders are characterised by particle size (sieve analysis), shape, flow rate (seconds for 50 g to run through a Hall funnel) and apparent density (loose fill). Höganäs data: water-atomised ASC100.29 2.98 g/cm³ and 25 s; sponge NC100.24 2.45 g/cm³ and 32 s. The atomised powder packs and compresses better; the sponge's ragged particles interlock for higher green strength.

What happens to the powder in the die, and how much pressure does it need?

As the punch moves, particles first slide and rearrange, then deform plastically and cold-weld at their contacts. Density climbs quickly, then ever more slowly towards solid iron, 7.874 g/cm³: Höganäs ASC100.29 reaches 6.77, 7.20 and 7.45 g/cm³ at 400, 600 and 800 MPa. The Heckel equation turns such data into a straight line. Friction at the die wall makes pressure fall with depth, so a single-action compact is densest under the moving punch; double action presses from both ends. Press force is pressure times projected area, and the die is filled about 2.4 times deeper than the part.

Powder metallurgy compaction: pressure versus green density, press tonnage, fill height and sintering shrinkage: interactive loads as you scroll

What happens during sintering?

The green compact is heated in a protective atmosphere well below melting; Höganäs quote 1,120 °C for 30 minutes in 90/10 nitrogen-hydrogen for iron parts, about three quarters of iron's melting point in kelvin (1,393 K against 1,811 K). The lubricant burns off first. Atoms then diffuse to the contacts and grow necks between particles; pores round off and shrink, so strength rises and the part may shrink slightly. Liquid-phase sintering (copper in iron, cobalt in tungsten carbide) is faster. The same shrinkage arithmetic appears in the Additive Manufacturing book's sintering interactive.

What secondary operations finish PM parts?

Sizing re-presses a sintered part in a die to tighten its dimensions; coining re-presses harder to raise density and sharpen detail. Infiltration wicks molten copper into the pores of an iron part to raise strength. Impregnation fills the connected pores with oil under vacuum, so a self-lubricating bush feeds oil to its shaft as it warms. Parts can also be resin-impregnated before plating, steam-treated for wear and corrosion resistance, heat-treated, or machined where the die cannot form a feature, such as a cross hole or an undercut.

Predict, then check

Two iron powders are pressed at the same 600 MPa: water-atomised ASC100.29 and sponge NC100.24. Which compact comes out denser?

Worked examples

Press force and green density of a bush · Bush 30 mm outside, 18 mm bore, 25 mm high, ASC100.29 pressed at 600 MPa in a lubricated die (green density 7.20 g/cm³ from the data sheet); iron 7.874 g/cm³
  1. Projected area = π/4 × (30² − 18²) = 452.4 mm²
  2. Press force = 600 × 452.4 = 2,71,434 N = 271 kN, about 27.7 tonnes
  3. Relative density = 7.20/7.874 = 91.4%, so 8.6% is pore space; mass = 7.20 × 452.4 × 25/1,000 = 81.4 g
  4. A Heckel line through the three data points (K = 2.39 × 10⁻³ /MPa, A = 1.013) predicts 7.34 g/cm³ at 700 MPa: 100 MPa more buys only 0.14 g/cm³

Answer: About 271 kN (28 t); 91% dense, 81 g

How deep must the die be filled? · The same 25 mm bush; ASC100.29 (apparent density 2.98, 7.20 g/cm³ at 600 MPa) or sponge NC100.24 (apparent density 2.45, 6.79 g/cm³ at 500 MPa)
  1. Mass is conserved: ρ_apparent × A × H_fill = ρ_green × A × H, so H_fill = H × ρ_green/ρ_apparent
  2. ASC100.29: compression ratio 7.20/2.98 = 2.42, fill height 25 × 2.42 = 60.4 mm
  3. NC100.24: compression ratio 6.79/2.45 = 2.77, fill height 69.3 mm
  4. The die and the press stroke must accommodate the fill, not the part: the sponge powder needs a die about 9 mm deeper here

Answer: 60 mm of atomised powder or 69 mm of sponge powder for a 25 mm part

Sintering shrinkage of a stainless steel part · Höganäs 316LHC: green density 6.55 g/cm³ at 600 MPa with 1% wax; sintered at 1,250 °C in hydrogen to 6.80 g/cm³; measured diametral change green to sintered −1.39%
  1. Isotropic shrinkage with no mass loss: linear shrinkage = 1 − (ρ_green/ρ_sintered)^⅓
  2. Ignoring the wax: 1 − (6.55/6.80)^⅓ = 1.24%
  3. The 1% wax burns off, so the metal-only green density is 6.55 × 0.99: 1 − (6.485/6.80)^⅓ = 1.57%
  4. The measured 1.39% lies between: real parts shrink a little differently in diameter and height, so dies are sized from trials

Answer: Estimate 1.2-1.6% linear against 1.39% measured

Keep going

Recommended reading

  • Fundamentals of Modern Manufacturing, M. P. Groover
  • Manufacturing Engineering and Technology, S. Kalpakjian, S. R. Schmid
  • Manufacturing Technology Vol. 1 and 2, P. N. Rao
  • Principles of Metal Casting, R. W. Heine, C. R. Loper, P. C. Rosenthal
  • Manufacturing Science, A. Ghosh, A. K. Mallik
  • Production Technology, R. K. Jain

How this book was made

Chapters follow the official syllabi listed above. Every interactive's formulas and every worked-example number were checked by script. Explanations, animations and examples are original to Skill-Lync. Topics informed by MIT OpenCourseWare 2.008 (Design and Manufacturing II), CC BY-NC-SA. All explanations, animations and examples are original to Skill-Lync; MIT does not endorse this page.

From the exam to a core job

Talk to a counsellor.

Ask how manufacturing and design engineers use manufacturing processes on real projects, and what the Post Graduate Program in Manufacturing Design covers.

Talk to a career counsellor

Free 20-minute call. We'll map the right course to where you want to work.

By submitting you agree to our Terms and Privacy Policy.

Talk to a career counsellor

Talk to a career counsellor

Free 20-minute call. We'll map the right course to where you want to work.

By submitting you agree to our Terms and Privacy Policy.