10.1.1 BCC, FCC, HCP and BCT lattices

Metals are crystalline solids in which atoms occupy repeating three dimensional lattices. Mechanical, magnetic, diffusional and formability behaviour follows directly from lattice geometry, packing density and available slip systems.

BCC means body centred cubic. A cube with one atom at each corner and one atom at the body centre. Packing efficiency is about 68 percent. Coordination number is 8. Typical metals are alpha iron called ferrite, chromium, molybdenum, tungsten and vanadium. Slip in BCC occurs on several families of planes but none is truly close packed, so the Peierls stress is strongly temperature dependent. The practical result is good room temperature strength combined with a ductile to brittle transition at low temperature. Ferrite is ferromagnetic below the Curie temperature, a property that later governs magnetic particle inspection.

FCC means face centred cubic. A cube with atoms at corners and at the centres of all six faces. Packing efficiency is about 74 percent. Coordination number is 12. Typical metals are gamma iron called austenite, aluminium, copper, nickel and lead. There are four close packed planes of the 111 family, each with three close packed directions, giving twelve equivalent slip systems with low critical resolved shear stress. The practical result is high ductility, high formability and high strain hardening capacity. The large octahedral interstitial voids of FCC dissolve much more carbon than BCC, which explains why austenite is the parent phase for carburising and hardening. Austenite is paramagnetic, hence non magnetic in inspection terms, a classic contrast with ferrite.

HCP means hexagonal close packed. Packing efficiency is about 74 percent, equal to FCC, with coordination number 12. Typical metals are zinc, magnesium, alpha titanium and cobalt. Although packing is dense, the number of easily activated slip systems at room temperature is small, basal slip dominating. The practical result is limited cold formability, strong anisotropy and a tendency to crack in sheet forming. Magnesium sheet and zinc die castings illustrate the point. At elevated temperature additional pyramidal systems activate and formability improves markedly.

BCT means body centred tetragonal. It is a BCC cube stretched along one axis so that height exceeds base edge. The only common engineering example is martensite in quenched steel. When FCC austenite supersaturated with carbon transforms without diffusion, carbon atoms remain trapped in octahedral sites and elongate one cube axis. Axial ratio increases with carbon content. The lattice is highly strained, dislocation motion is severely pinned, hardness is extreme and ductility is minimal. Tempering is therefore mandatory after quenching.

The memory pattern for examinations is direct. Formable non magnetic FCC metals are aluminium, copper, nickel and austenite. Strong temperature sensitive BCC metals include ferrite, chromium and refractory metals. Difficult to cold form HCP metals include magnesium and zinc. Hard brittle BCT material is freshly quenched martensite.

10.1.2 Point, line, planar and volume defects

A perfect lattice is an idealisation. Real crystals contain defects, and every strengthening method works by creating obstacles to dislocation motion.

Vacancy is a point defect formed by a missing atom on a lattice site. Equilibrium concentration rises exponentially with temperature. Vacancies enable substitutional diffusion and assist interstitial diffusion by local lattice relaxation. Carburising, nitriding, homogenisation and creep all depend on vacancy mediated atomic jumps. Quenching from high temperature freezes in excess vacancies, accelerating low temperature ageing.

Interstitial atom is a small atom such as carbon or nitrogen occupying a void between host atoms. In BCC ferrite the voids are small and carbon solubility is limited to about 0.025 percent at 727 C. In FCC austenite the octahedral voids are larger and solubility rises to about 2.11 percent at 1147 C. This factor of nearly one hundred explains the entire heat treatment industry.

Dislocation is a line defect. An edge dislocation is an extra half plane of atoms inserted in the lattice. A screw dislocation is a helical ramp around a dislocation line. Mixed dislocations are the general case. Plastic deformation at low temperature is motion of dislocations on slip planes, not simultaneous sliding of whole planes. Burgers vector measures the lattice closure failure and gives slip direction and magnitude. Increasing dislocation density by cold work raises strength but lowers ductility because dislocations entangle and pile up.

Grain boundary is a planar defect separating crystals of different orientation in a polycrystal. Boundaries obstruct dislocations, raise strength, improve toughness by deflecting cracks, and provide sites for segregation and corrosion. Twin boundary is a special symmetric boundary important in HCP deformation and in annealed FCC metals. Stacking fault is a local error in close packing sequence that controls cross slip and work hardening rate.

Volume defects include voids, porosity, inclusions and cracks. They nucleate fracture and degrade fatigue life. Casting porosity, weld slag and grinding cracks belong in this class.

10.1.3 Hall Petch relation and strengthening logic

Yield strength of a polycrystal rises as grain size is refined according to the Hall Petch equation. In symbols, sigma y equals sigma 0 plus k divided by square root of d, where sigma 0 is friction stress of the lattice, k is a boundary resistance constant and d is mean grain diameter. Fine grains present more boundary area per unit volume. Dislocations pile up at boundaries and a higher applied stress is needed to push slip across. That is why normalising, which refines grains by air cooling from the austenite field, gives higher strength than slow furnace cooled annealing. That is also why grain coarsening during overheating or prolonged high temperature exposure lowers strength and toughness.

The same blocking idea unifies all strengthening routes. Solid solution strengthening uses dissolved atoms such as nickel, manganese and silicon to distort the lattice and drag dislocations. Precipitation and dispersion strengthening use fine carbides and intermetallics such as vanadium carbide and copper rich precipitates in duralumin to pin dislocations. Work hardening raises dislocation density so that dislocations block each other. Phase transformation hardening uses martensite and bainite plates with high interface density and carbon supersaturation. Grain refinement uses boundaries themselves. Every alloying and heat treatment choice in later sections is a selection among these mechanisms.

10.2.1 Phases encountered in steels and cast irons

Ferrite, also called alpha iron, is BCC, soft, ductile, magnetic, with maximum carbon solubility 0.025 percent at 727 C, falling to near zero at room temperature. It provides ductility and weldability in low carbon steels.

Austenite, also called gamma iron, is FCC, non magnetic, tougher and more formable, with maximum carbon solubility 2.11 percent at 1147 C. It is stable above 912 C in pure iron and over a wider field in alloyed steel. All hardening starts by austenitising, meaning heating into this field to dissolve carbon and alloy carbides uniformly.

Delta ferrite is high temperature BCC iron stable near the melting point. It participates in the peritectic reaction and has little direct engineering use, but its presence explains the peritectic horizontal.

Cementite is the intermetallic compound Fe3C with 6.67 percent carbon by mass. It is orthorhombic, very hard, brittle and magnetic below its Curie point. As thin lamellae it strengthens pearlite. As a continuous network on grain boundaries it embrittles hyper eutectoid steels and white cast irons.

Pearlite is not a single phase. It is a lamellar aggregate of ferrite and cementite formed by eutectoid decomposition. The lamellar spacing controls strength. Coarse pearlite from slow furnace cooling is soft and machinable. Fine pearlite from air cooling is stronger. Appearance under the microscope resembles mother of pearl, hence the name.

Ledeburite is the eutectic aggregate of austenite and cementite formed at 4.3 percent carbon and 1147 C. On further cooling the austenite portion converts to pearlite or martensite depending on cooling rate, leaving a matrix rich in cementite. It is characteristic of cast irons.

Martensite is BCT, diffusionless, formed when austenite is quenched fast enough to suppress carbon partitioning. Carbon remains trapped, tetragonality rises, hardness peaks and brittleness is severe. Fresh martensite must be tempered before load bearing service.

Bainite is an aggregate of ferrite laths or plates with fine carbide particles, formed by isothermal holding in the intermediate bay between the pearlite nose and the martensite start temperature. Upper bainite formed at higher temperature is coarser and tougher. Lower bainite formed at lower temperature is finer and harder. Austempering produces bainite directly with an attractive blend of hardness and toughness and with less distortion than quench plus temper.

Retained austenite is untransformed FCC retained at room temperature in high carbon or high alloy steels after incomplete martensitic transformation. It can improve toughness but can later transform under stress with dimensional change. Cryogenic treatment and multiple tempering reduce retained austenite in tools and bearings.

10.2.2 Invariant reactions and the six governing numbers

Only six numbers are needed to reconstruct the industrially relevant part of the diagram. Carbon concentrations 0.025 percent, 0.77 percent, 2.11 percent and 4.3 percent, and temperatures 727 C and 1147 C. A third temperature near 1493 C marks the peritectic, retained for completeness.

Eutectoid reaction at 0.77 percent carbon and 727 C. On slow cooling, solid austenite of eutectoid composition decomposes into solid ferrite plus solid cementite in lamellar pearlite form. The eutectoid temperature is also called the lower critical temperature, denoted A1. All hypo and hyper eutectoid reasoning pivots on this point.

Eutectic reaction at 4.3 percent carbon and 1147 C. Liquid of eutectic composition freezes into solid austenite plus solid cementite as ledeburite. This temperature marks the lowest melting point in the iron carbon system and the boundary between easy melting cast irons and higher melting steels.

Peritectic reaction near 0.16 percent carbon and 1493 C. Liquid plus delta ferrite transforms into austenite on cooling. It has limited shop floor consequence but completes the reaction family and is retained as vocabulary.

Solvus and solidus lines carry the remaining meaning. The A3 line is the boundary between austenite and austenite plus ferrite for hypo eutectoid steels. The Acm line is the boundary between austenite and austenite plus cementite for hyper eutectoid steels. The solubility drop along A3 and Acm on cooling is what precipitates proeutectoid ferrite and proeutectoid cementite respectively.

10.2.3 Hypo eutectoid and hyper eutectoid reasoning

Hypo eutectoid steels contain less than 0.77 percent and more than 0.025 percent carbon. On slow cooling from austenite, proeutectoid ferrite nucleates first on austenite grain boundaries along the A3 line. Remaining austenite enriches in carbon until it reaches 0.77 percent at 727 C, then transforms to pearlite. Room temperature structure is proeutectoid ferrite plus pearlite. Higher carbon within this range means less free ferrite and more pearlite, hence higher strength and hardness and lower ductility. A 0.2 percent steel is mostly ferrite with isolated pearlite colonies and is weldable and drawable. A 0.6 percent steel is roughly half pearlite and suits shafts and springs after heat treatment.

Hyper eutectoid steels contain between 0.77 percent and 2.11 percent carbon. On slow cooling, proeutectoid cementite nucleates first on austenite boundaries along the Acm line. Remaining austenite depletes toward 0.77 percent and then forms pearlite at 727 C. Room temperature structure is proeutectoid cementite network plus pearlite. The continuous brittle cementite network degrades machinability and toughness, so spheroidising annealing is commonly applied to break the network into isolated spheroids before machining.

Beyond 2.11 percent carbon the alloy is a cast iron. Between 2.11 and 4.3 percent it is hypo eutectic. At 4.3 percent it is eutectic. Above 4.3 percent and up to 6.67 percent it is hyper eutectic with primary cementite dendrites plus ledeburite. Melting point falls toward the eutectic, fluidity rises, but ductility falls because of graphite or cementite volume.

Lever rule at just below 727 C gives constituent fractions directly. Pearlite fraction in a hypo eutectoid steel equals alloy carbon minus 0.025 divided by 0.77 minus 0.025. Ferrite fraction is the remainder. In a hyper eutectoid steel the endpoints are 0.77 and 6.67 for cementite accounting, with pearlite as the complementary aggregate. This arithmetic converts diagram memory into microstructure prediction and then into property and treatment selection.

Quick Example — Pearlite Fraction in 0.40 Percent Carbon Steel

Given alloy carbon $C_0 = 0.40$ percent, ferrite endpoint $0.025$ percent, and eutectoid $0.77$ percent just below $727$ C.

Step 1: use $W_{pearlite} = (C_0 - 0.025)/(0.77 - 0.025)$.

Step 2: substitute $W_{pearlite} = (0.40 - 0.025)/0.745 \approx 0.503$.

Step 3: take $W_{ferrite} = 1 - W_{pearlite} \approx 0.497$.

Result: about $50.3$ percent pearlite with about $49.7$ percent proeutectoid ferrite.

Trap: the endpoints apply just below the eutectoid isotherm on slow cooling, not above it and not after rapid quenching.

10.2.4 Reading the diagram as a selection tool

The diagram is a decision map. Low carbon hypo eutectoid compositions give ferrite rich weldable structures. Medium carbon compositions give balanced ferrite pearlite mixtures that respond well to quench and temper. High carbon hyper eutectoid compositions give wear resistant but brittle networks that need spheroidising and careful tempering. Cast iron compositions give fluid melts and compressive strength at the cost of tensile ductility. Austenitising temperature is chosen just above A3 for hypo eutectoid steels and just above A1 for hyper eutectoid steels in order to retain some undissolved carbide for wear resistance while avoiding grain coarsening.

10.3.1 Plain carbon steels by carbon level

Low carbon steels contain less than about 0.25 percent carbon. They are soft, ductile, weldable and drawable. Applications include structural sections, sheet for deep drawing, pipes, wire and general fabrication. They do not harden appreciably by quenching because martensite hardness scales with carbon.

Medium carbon steels contain about 0.25 to 0.60 percent carbon. Strength and hardenability rise, weldability falls and preheat becomes relevant. Applications include shafts, gears, crankshafts, rails, couplings and pressure parts after quench and temper or normalising. They represent the core of refinery rotating equipment steels.

High carbon steels contain about 0.60 to 1.50 percent carbon. They develop very high hardness after quench and temper, hold cutting edges and resist wear, but weldability is poor and cracking risk is high. Applications include springs, piano wire, cutting tools, dies, bearings and wear plates. They demand controlled heating, protective atmosphere to limit decarburisation, and prompt tempering.

10.3.2 Cast iron family with selection reasoning

Iron	Carbon morphology	Mechanical character	Canonical uses and reasoning
Gray cast iron	Graphite flakes in pearlite or ferrite matrix	High compressive strength, superb vibration damping, good machinability, low tensile ductility because flakes act as internal notches	Engine blocks, lathe beds, pump housings, machine bases. Chosen when vibration absorption plus compressive load plus low cost casting matters more than tensile strength
White cast iron	Combined carbon as massive cementite, no free graphite	Extreme hardness, high abrasion resistance, brittle, essentially unmachinable	Mill liners, crusher jaws, shot blasting nozzles, wear surfaces. Chosen when abrasion dominates and machining is not required
Nodular or ductile iron, SG iron	Spheroidal graphite through magnesium treatment	Near steel tensile strength with retained castability, ductile, fatigue resistant	Crankshafts, valves, pipe fittings, refinery pump casings. Chosen when a casting is mandatory but impact and pressure loading demand ductility
Malleable iron	Temper carbon rosettes formed by heat treating white iron	Tough in thin sections, good shock resistance	Brackets, small fittings, agricultural hardware, chain links. Chosen for small thin parts needing toughness without forging cost
Alloy cast iron, including Ni hard and high chromium iron	Carbide stabilised alloy matrix	Superior hot hardness and corrosion resistance	Furnace parts, slurry pumps, refinery wear components in erosive streams

Selection reasoning reduces to three contrasts. Damping plus compressive load plus economy points to gray iron. Ductility plus pressure integrity in a cast shape points to nodular iron. Pure abrasion without machining points to white or high alloy white iron. Weld repair on gray iron is crack sensitive because of graphite films and hard heat affected zones, so preheat and nickel base electrodes are standard practice.

10.3.3 Stainless steels

Ferritic stainless grades are BCC, magnetic, with about 12 to 18 percent chromium and low nickel. They resist oxidation and mild corrosion, cost less than austenitic grades, and suit utensils, trim and furnace supports. They are susceptible to grain coarsening on welding.

Austenitic stainless grades such as 18Cr 8Ni, widely known as 304, are FCC, essentially non magnetic in the annealed state, with the highest general corrosion resistance among common grades. Molybdenum bearing 316 adds pitting resistance in chloride service. Uses include vessels, piping, food and pharmaceutical plant and refinery corrosion service. Work hardening rate is high, so machining needs rigid setups and sharp tools.

Martensitic stainless grades are hardenable by quench and temper, magnetic, with moderate corrosion resistance. Uses include blades, turbine parts, surgical tools and wear resistant shafts.

Duplex grades mix ferrite and austenite in roughly equal proportions, giving high strength plus strong resistance to chloride stress corrosion cracking. Uses include offshore piping, subsea hardware and demanding refinery exchangers.

Precipitation hardening grades combine corrosion resistance with high strength after ageing, used in aerospace fasteners, shafts and precision parts.

10.3.4 Alloying elements with mechanism level effects

Element	Principal effects and mechanism
Nickel	Austenite stabiliser. Lowers critical temperatures, improves low temperature toughness, raises corrosion resistance. With chromium it creates the austenitic stainless field. It does not form strong carbides, so it stays in solution and toughens the matrix
Chromium	Ferrite stabiliser and strong passive film former. Above about 12 percent it confers stainless behaviour through a thin chromium oxide film. It raises hardenability, forms hard carbides, improves oxidation and wear resistance. Core element in stainless steels, tool steels and creep resistant steels
Molybdenum	Raises high temperature strength and creep resistance through solid solution and fine carbide precipitation. Suppresses temper embrittlement. Essential in pressure vessel steels, high speed steels and pitting resistant stainless grades
Vanadium	Very strong carbide and nitride former. Small additions refine austenite grain size, raise hardness and improve wear resistance. Core microalloy in tool steels and high strength low alloy steels
Manganese	Deoxidiser and desulphuriser. It ties up sulphur as manganese sulphide instead of iron sulphide films on grain boundaries, thereby preventing hot shortness. It lowers critical temperature, raises hardenability and strengthens the matrix. Hadfield manganese steel work hardens dramatically under impact
Tungsten	Strong carbide former conferring hot hardness and red hardness. Retains cutting ability at elevated temperature in high speed steels. Also used in filaments and counterweights because of high density
Silicon	Deoxidiser and ferrite strengthener. Raises electrical resistivity for transformer steels and improves oxidation resistance. Excess silicon degrades weldability and surface quality
Sulphur and phosphorus	Normally controlled as impurities. Sulphur improves machinability as manganese sulphide stringers but causes hot shortness when free. Phosphorus strengthens but embrittles, so it is kept low except in weathering steels where a small addition improves atmospheric resistance

The manganese sulphur reaction deserves emphasis. Sulphur segregates as iron sulphide films with low melting point on grain boundaries and causes cracking during hot working. Manganese has stronger affinity for sulphur and forms discrete high melting manganese sulphide particles that are far less harmful and even aid chip breaking. This single reaction explains routine manganese additions to every structural steel.

10.4.1 Bulk heat treatments and what each achieves

Full annealing heats hypo eutectoid steel above A3 and hyper eutectoid steel above A1, holds for homogenisation, then cools very slowly in the furnace. The result is coarse pearlite with maximum softness, maximum ductility, minimum residual stress and best machinability. It also refines a distorted or work hardened structure before further processing.

Process annealing is a lower temperature subcritical anneal applied to cold worked low carbon sheet and wire to restore ductility between draws without full phase change.

Spheroidising holds high carbon steel just below A1 for prolonged time to convert lamellar cementite into spheroids. Machinability and formability improve dramatically because sharp lamellae no longer concentrate stress.

Normalising heats above A3 or Acm and cools in still air. Cooling is faster than furnace annealing, so austenite grains transform to finer pearlite with smaller spacing and smaller grain size. Strength, hardness and toughness exceed annealed values. It is the default treatment for structural parts, forgings and weldable fabrications that need uniform properties without quench cracking risk.

Quenching, also called hardening, austenitises fully and then cools faster than the critical rate in water, oil, polymer or gas. Austenite transforms to martensite. Hardness is maximal. Internal stress, distortion and cracking risk are also maximal. Section size matters because only the volume that misses the pearlite nose actually hardens.

Tempering reheats quenched martensite to between about 150 C and 650 C and then cools. Low temperature tempering preserves hardness for tools while relieving quench stress. Medium tempering balances hardness and toughness for springs and fasteners. High temperature tempering produces tempered martensite or sorbite with high toughness for shafts and pressure parts. Untempered martensite in service is a failure waiting to happen because of brittleness and retained stress.

Austempering quenches into a salt or metal bath held above the martensite start temperature, holds isothermally until bainite forms fully, then cools to room temperature. No separate temper is needed. The product is bainitic, hard and tough, with less distortion than direct quenching. It suits springs, clips and small wear parts.

Martempering, also called marquenching, quenches just above the martensite start, holds briefly to equalise temperature between surface and core, then cools through the martensite range. Thermal gradients are reduced, so distortion and quench cracking fall while final hardness remains high after tempering. It suits intricate shapes and alloy tool steels.

10.4.2 TTT and CCT nose logic

A time temperature transformation diagram, often called TTT or S curve, plots isothermal transformation start and finish times of supercooled austenite against holding temperature. A continuous cooling transformation diagram, called CCT, plots the same transformations during continuous cooling and is shifted to longer times and lower temperatures.

Three fields dominate. High temperature holding gives pearlite through diffusion. Intermediate holding gives bainite. Low temperature rapid cooling gives martensite without diffusion. The nose is the temperature of minimum austenite stability, where the start curve lies farthest to the left. Cooling must be fast enough to bypass the nose without intersecting the pearlite start line. If the cooling curve touches the nose, pearlite forms and full hardness is lost. Thick sections cool slower at the core, strike the nose, and remain soft inside while the surface hardens. That is the physical basis of hardenability and of the Jominy end quench test.

Holding just above the nose and cooling slowly yields coarse pearlite. Holding just below the nose yields fine pearlite. Holding in the bay between nose and martensite start yields bainite. Quenching below martensite start without intersection yields martensite. Alloying elements such as chromium, molybdenum, nickel and manganese shift the nose to the right, lower the critical cooling rate and deepen hardening. That rightward shift is the quantitative meaning of higher hardenability.

Practical defects map directly to nose behaviour. Soft spots mean the local cooling rate missed the nose because of a vapour blanket, oil degradation, low agitation or excessive section size. Quench cracks mean cooling was too severe or tempering was omitted. Decarburised skin means furnace atmosphere removed surface carbon, so the surface cannot form high carbon martensite. Distortion means non uniform cooling or asymmetric section.

10.4.3 Case hardening and surface modification

Carburising diffuses carbon into the skin of a low carbon steel at austenitic temperature in solid, liquid or gas media, followed by quench and low temper. The case becomes high carbon martensitic and wear resistant while the core remains low carbon and tough. Gears, camshafts and pinions are classic beneficiaries.

Nitriding diffuses nitrogen into alloy steel at subcritical temperature, forming hard nitrides without quenching. Distortion is minimal, surface hardness is extreme and wear and fatigue resistance rise. Crankshafts, extrusion screws and precision tooling use this route. No post quench is needed because hardening comes from nitride precipitation, not martensite.

Carbonitriding adds carbon and nitrogen together at slightly above the A1 range, giving a harder case than carburising alone with lower distortion. Cyaniding is a related liquid bath variant with stricter environmental controls.

Induction hardening heats a local skin by electromagnetic induction and quenches immediately. Flame hardening uses an oxy fuel flame for the same purpose. Both suit shafts, rails, gear teeth and large parts where only the track needs hardness. Case depth follows power, frequency, traverse speed and quench delay. Higher frequency concentrates heating nearer the surface.

Boronising, chromising and laser or electron beam hardening extend the same idea. The principle remains constant. Create a hard wear resistant skin metallurgically bonded to a tough core, with compressive surface stress where possible to aid fatigue life.

10.5.1 NDT methods with physical basis and selection logic

Method	Physical principle	Detects	Selection logic
Visual testing	Direct and aided optical observation	Surface profile, undercut, overlap, gross cracks, corrosion, misalignment	Always first. Cheapest and fastest. Every weld and casting passes through visual appraisal before costlier methods
Dye penetrant testing	Capillary ingress of coloured or fluorescent liquid into surface breaking discontinuities, followed by developer bleed out	Surface breaking cracks, porosity, laps and seams open to the surface	Chosen for non magnetic materials such as austenitic stainless steel, aluminium and copper alloys, and for field use where power and access are limited. Needs thorough cleaning and dwell control
Magnetic particle testing	Leakage magnetic flux at surface and near surface discontinuities in a magnetised ferromagnetic part attracts iron particles into visible indications	Surface and slightly subsurface cracks, seams, laps in ferromagnetic steels	Chosen for carbon and low alloy steel welds and forgings when speed matters. Invalid on non magnetic metals. Demagnetisation follows inspection of precision parts
Ultrasonic testing	High frequency shear and longitudinal waves reflect from interfaces and flaws. Time of flight gives depth, amplitude gives severity	Internal cracks, laminations, lack of fusion, inclusions, wall thickness loss	Chosen for thick sections, pressure vessels, rails, forgings and in service thickness survey. Needs couplant, calibration blocks and skilled interpretation. Excellent depth localisation
Radiographic testing with X rays or gamma rays	Differential absorption of penetrating radiation by voids and dense inclusions recorded on film or digital detector	Volumetric flaws such as porosity, slag, tungsten inclusions, incomplete penetration	Chosen when a permanent image record is contractually required for vessel and pipeline welds. Radiation safety exclusion zone and film processing discipline are mandatory
Eddy current testing	Alternating coil induces eddy currents in a conductive part. Flaws perturb impedance of the probe circuit	Surface and near surface cracks, wall thinning, sorting of tubes and fasteners	Chosen for non ferromagnetic heat exchanger tubing, condenser tubes and automated high speed sorting. Highly sensitive to lift off and fill factor, so probe control is critical
Acoustic emission and thermography, supplementary	Transient elastic waves from growing defects, and infrared mapping of thermal anomalies	Active crack growth, leaks, delaminations, insulation breakdown	Chosen for monitoring and screening rather than code acceptance. Useful for large area survey before focused ultrasonic or radiographic follow up

Selection reduces to material plus flaw location plus record needs. Magnetic parts with surface cracks point to magnetic particle testing. Non magnetic parts with surface cracks point to dye penetrant testing. Thick pressure retaining sections with internal integrity needs point to ultrasonic testing. Weld volumetric acceptance with archival proof points to radiography. Conductive thin wall tubing in service points to eddy current testing. Refinery shutdown practice follows increasing depth and cost in the order visual, then penetrant or magnetic particle, then ultrasonic or radiographic examination.

10.5.2 Non ferrous alloys with reasoning

Brass is copper zinc alloy. Zinc lowers melting point, improves fluidity, raises strength and retains good corrosion resistance. Cartridge brass with about 70 percent copper and 30 percent zinc combines deep drawability with strength. Plumbing fittings, marine hardware, musical instruments and decorative parts use brass where corrosion resistance plus formability plus attractive finish matter.

Bronze is copper tin alloy, often with phosphorus, aluminium or silicon additions. Phosphor bronze gives hard wear resistant bearings, gears and springs. Aluminium bronze gives high strength marine propellers and pump parts. Silicon bronze gives weldable corrosion resistant hardware. Tin strengthens copper strongly and improves seizure resistance, which is why bronze remains a bearing metal despite higher cost.

Duralumin is aluminium copper magnesium age hardening alloy. Solution treatment followed by ageing precipitates fine copper rich phases that block dislocations. Strength approaches structural steel at a fraction of density, so aircraft skins, spars, truck bodies and stressed panels use it. Corrosion protection by cladding or anodising compensates for lower inherent resistance relative to pure aluminium.

Invar is iron with about 36 percent nickel. Near zero thermal expansion around room temperature arises from magneto volume compensation. Measuring tapes, precision instruments, optical mounts and cryogenic tooling use it where dimensional stability under temperature change is paramount.

Monel is nickel copper alloy with superb resistance to seawater, hydrofluoric acid and alkaline streams. Inconel and related nickel chromium superalloys retain strength and oxidation resistance at gas turbine and furnace temperatures through solid solution and gamma prime precipitation. Refinery high temperature and corrosive streams, flare hardware and heater supports illustrate the selection. Aluminium alloys more broadly serve frames, vessels and transport where strength per mass dominates. Copper alloys serve electrical, thermal and bearing functions where conductivity and seizure resistance dominate. Titanium alloys serve aerospace and chemical plant where strength per mass plus chloride resistance justifies cost.

10.6.1 Process steps from pattern to inspection

Pattern making creates an oversize replica of the part with allowances for shrinkage, draft, machining, distortion and shake. Mould making packs sand or coats a die around the pattern, forms the gating system and risers, sets cores for hollow features, vents gases and applies coatings. Melting prepares liquid metal with controlled chemistry, temperature, deoxidation and degassing. Pouring delivers metal at the correct superheat and rate without turbulence or slag entrainment. Solidification proceeds directionally toward the riser under the influence of chills where needed. Shakeout removes the casting from the mould. Fettling cuts off gates and risers, removes fins, cleans surfaces and straightens distortion. Heat treatment, straightening and machining follow as needed. Inspection closes the loop with visual, dimensional, dye penetrant, magnetic particle, ultrasonic or radiographic examination plus mechanical testing.

10.6.2 Solidification and derivation of Chvorinov rule

Solidification time is governed by heat extraction through the mould wall. Consider a casting of volume V and cooling surface area A. Total heat to remove scales with V through sensible plus latent heat. Rate of heat removal scales with A through the mould heat transfer coefficient. Under the same metal, superheat and mould conditions, freezing time therefore scales with the square of the ratio V divided by A. Formally, solidification time ts equals mould constant Cm multiplied by modulus M squared, where M equals V divided by A. Cm absorbs metal density, latent heat, specific heat, pouring superheat, mould thermal diffusivity and interface resistance. Derivation proceeds by equating heat content to one dimensional conduction into a semi infinite mould, solving for interface motion, then integrating to total freeze time. The square law is the essential result. Doubling the modulus quadruples freezing time.

A cube of side a has volume a cubed and area six a squared, so modulus equals a divided by six. A large cube of 120 mm side has modulus 20 mm. A small cube of 60 mm side has modulus 10 mm. With identical mould constant, freezing time ratio is four to one because modulus ratio squared equals two squared. The slower massive casting dictates feeding design. Plate geometries have lower modulus than compact blocks of equal mass and freeze faster, which is why thin ribs freeze prematurely unless gated and chilled correctly.

Quick Example — Chvorinov Timing at Modulus Ratio Two

Given mould constant $C_m = 3$ s per square mm, $M_1 = 20$ mm, and $M_2 = 10$ mm.

Step 1: use $t_s = C_m M^2$.

Step 2: compute $t_1 = 3 \times 20^2 = 1200$ s.

Step 3: compute $t_2 = 3 \times 10^2 = 300$ s.

Result: freezing times are $1200$ s and $300$ s, a ratio of $4$ to $1$ from a modulus ratio of $2$ to $1$.

Trap: $C_m$ must be identical for both castings, meaning same metal, superheat, and mould.

10.6.3 Riser, chill and directional solidification logic

Riser, also called feeder, is an extra reservoir of liquid metal attached to the casting to compensate solidification shrinkage. Design rule is modulus of riser greater than modulus of casting, commonly by a factor around 1.2 for the neck and body, so that the riser freezes last and remains liquid while the casting solidifies. Top risers feed by gravity and also vent. Side risers feed through a neck and suit flat castings. Blind risers are enclosed and need atmospheric venting. Open risers expose liquid to air and feed more reliably but lose heat faster.

Chill is a metallic insert of steel, copper or graphite placed in the mould wall to extract heat locally and accelerate freezing. External chills sit on the mould face. Internal chills fuse into the casting and must match chemistry to avoid defects. Chills serve two roles. They enforce directional solidification from thin or remote sections toward the riser, and they refine grain structure in heavy bosses that would otherwise freeze slowly and coarsely. Padding adds extra stock to thin sections to smooth thermal gradients toward the feeder.

Directional solidification is the governing doctrine. Freezing should start at the thinnest remote extremity and progress continuously toward the riser without isolated liquid pools. Isolated pools that freeze without feed form centreline shrinkage porosity. Computer thermal simulation modernises the same modulus logic by mapping local freezing times, but the riser last plus chill assisted gradient remains the conceptual core.

10.6.4 Gating system and metal delivery

Gating guides liquid from ladle to cavity with controlled velocity, minimal turbulence, slag separation and correct thermal distribution.

Pouring basin receives the stream, damps turbulence and skims slag. Sprue is a tapered vertical channel. Taper prevents air aspiration as velocity rises under gravity. Runner is a horizontal channel that distributes metal and traps slag through stepped or branched geometry. Gates are narrow entries into the cavity that control velocity and freeze off after filling to aid directional solidification. Strainers, filters and slag traps remove oxides and sand.

Pressurised gating keeps the smallest cross section at the gates, maintaining back pressure and reducing air entrainment, at the cost of higher jet velocity. Unpressurised gating keeps the smallest section at the sprue base, giving gentler filling and better slag retention, preferred for aluminium and magnesium that oxidise readily. Top gating fills fast but erodes mould and traps oxides. Bottom gating fills calmly from below and suits tall steel moulds where smooth rise prevents showering. Parting line gating balances simplicity and control for flat parts.

Pouring temperature and rate close the design. Excess superheat dissolves gases, erodes mould, coarsens grains and increases shrinkage. Insufficient superheat causes misrun and cold shut in thin sections. Pouring rate must keep the sprue full without freezing the runner.

10.6.5 Pattern allowances

Shrinkage allowance scales the pattern up to compensate liquid contraction plus solidification contraction plus solid cooling contraction. Different metals need different shrink rules, with steel needing more than gray iron because graphite precipitation partly offsets contraction.

Draft allowance tapers vertical faces by about one to three degrees so the pattern withdraws without tearing the mould. Internal faces need more draft than external faces.

Machining allowance adds stock on faces that will be machined to remove scale, decarburisation and surface roughness.

Distortion or camber allowance bends the pattern opposite to expected warping in long flat or U shaped castings.

Shake or rapping allowance adds a small negative margin because rapping the pattern to free it enlarges the cavity slightly.

10.6.6 Casting defects with causal mapping

Defect	Appearance	Root cause in process parameters
Shrinkage porosity	Rough internal voids at hot spots and junctions	Riser modulus too small, riser wrongly located, no directional gradient, pouring temperature too high
Cold shut	Linear discontinuity where two streams failed to fuse	Metal too cold, slow pouring, thin sections, poor gating balance
Misrun	Incomplete filling, rounded unfused edges	Low superheat, low pouring rate, excessive wall thinness, high mould heat extraction
Gas porosity and blowholes	Smooth spherical voids, surface blisters	Damp sand, poor venting, undried cores, dissolved hydrogen or nitrogen, excessive binder gas
Sand inclusion and wash	Embedded sand lumps and rough patches	High gating velocity eroding mould, weak ramming, soft sand, missing strainer
Slag inclusion	Non metallic stringers and cloudy machined surfaces	Poor skimming, turbulent gating, unpressurised system wrongly proportioned
Hot tear	Jagged intergranular crack at junctions and restrained arms	Restrained contraction, sharp internal corners without fillet, high sulphur, late shakeout
Mismatch and flash	Step at parting line and thin fins	Worn patterns, loose boxes, misaligned dowels, excessive pouring pressure lifting the cope
Scab and buckle	Rough raised patches from mould face lifting	Thin mould skin overheated, poor permeability, high moisture
Hard skin and chilled edges	Excess white iron at thin edges	Over chilling, low silicon, fast cooling in thin sections

Each defect names its remedy. Enlarge or relocate the riser for shrinkage. Raise superheat and pouring rate and rebalance gates for cold shut. Dry moulds, vent cores and degas melt for gas defects. Reduce gating velocity and strengthen sand for wash. Add fillets, relieve restraint and control sulphur for hot tears.

10.6.7 Special casting processes

Green sand moulding uses moist clay bonded sand for low cost general castings with moderate finish. Dry sand and skin dried moulds improve strength for medium large parts. Shell moulding coats a heated metal pattern with resin bonded sand to form a thin accurate shell, giving fine finish for small steel parts. Investment or lost wax moulding coats a wax pattern with ceramic, melts out the wax and pours metal for precision turbine blades, medical implants and intricate shapes with near net dimensions. Die casting injects non ferrous metal at high pressure into steel dies for thin wall high volume parts with superb finish. Hot chamber suits zinc and magnesium. Cold chamber suits aluminium and copper alloys. Centrifugal casting spins the mould so that centrifugal force feeds dense walls for pipes, liners and rings with clean bore surfaces. Continuous casting freezes slabs and billets in a water cooled mould for mill feed. Carbon dioxide sodium silicate process hardens large moulds rapidly by gassing. Vacuum and low pressure casting reduce turbulence and gas defects for aerospace alloys.

10.7.1 Arc physics and heat input

A welding arc is a sustained plasma column between electrode and workpiece stabilised by ionised shielding gas and metal vapour. Cathode emits electrons, anode receives them, the column reaches several thousand degrees Celsius, and magnetic pinch plus plasma jet transfer heat and droplets. Arc voltage rises with arc length. Arc current controls melting rate and penetration. Travel speed spreads that energy along the joint.

Heat input per unit length H equals voltage V multiplied by current I divided by travel speed v, further scaled by process efficiency where needed. In consistent units, volts times amps divided by millimetres per second gives joules per millimetre. High heat input deepens penetration but widens the heat affected zone, coarsens grains, increases distortion and risks burn through on thin sheet. Low heat input narrows the zone but risks lack of fusion, cold lap and slag entrapment. Procedure qualification therefore brackets a permitted heat input window rather than a single point.

Resistance heating follows Joule law, heat equals I squared R t. Spot and seam welding exploit contact resistance at the faying surface. Nugget size grows with current squared and time, which explains the narrow weld schedule window between undersize nuggets and expulsion.

Gas flame heating follows combustion of acetylene or propane in oxygen. Neutral, carburising and oxidising flames tune chemistry. Flame temperature is lower than arc plasma, heating is diffuse, distortion is wider, but equipment is portable and independent of power supply.

10.7.2 Polarity and shielding logic

Direct current electrode negative, often abbreviated DCEN or straight polarity, connects the electrode to negative and the workpiece to positive. Electrons bombard the workpiece, delivering about two thirds of the heat to the base metal. Penetration is deep and melting rate is high. It is the default for steel stick and TIG welding of steels where penetration matters.

Direct current electrode positive, abbreviated DCEP or reverse polarity, reverses the split. Positive ions bombard the workpiece and break up refractory oxides. Cathodic cleaning suits aluminium and magnesium where tenacious oxide films otherwise block fusion. Penetration is shallower and electrode heating is higher, so tungsten electrodes need larger diameter or AC operation.

Alternating current balances penetration and cleaning on successive half cycles and is standard for aluminium TIG welding. High frequency stabilisation maintains the arc. Shielding gases complete the logic. Argon gives stable smooth arcs for TIG and MIG on aluminium and stainless grades. Helium additions raise heat and penetration for thick copper and aluminium. Active carbon dioxide or argon carbon dioxide mixes stabilise steel MIG arcs and improve wetting at the cost of more spatter and oxidation. Flux blankets and slag systems in SAW and stick electrodes provide gas, deoxidisers and alloy additions plus thermal insulation.

10.7.3 Process comparison

Process	Electrode and shielding	Strengths	Limitations and canonical uses
Shielded metal arc welding, stick	Consumable flux coated electrode, slag plus gas shield	Portable, tolerant of rust and wind, all positions, low setup cost	Slower, slag removal needed, hydrogen control by drying. Field repair, erection, maintenance
Gas tungsten arc welding, TIG	Non consumable tungsten, argon or helium shield, separate filler rod	Finest quality, precise low heat control, superb on thin sheet, stainless grades, aluminium, titanium	Slowest deposition, needs skill and gas cover. Instrument tubing, aerospace welds, root passes on piping
Gas metal arc welding, MIG	Continuous consumable wire, gas shield	Fast, easy to automate, good on medium thickness, all positions with pulsed transfer	Wind sensitive, burn through on very thin sheet without pulse control. Production brackets, automotive, fabrication shops
Submerged arc welding, SAW	Bare wire plus granular flux blanket, arc hidden	Very high deposition, deep penetration, clean operator environment, superb on thick plate	Flat and horizontal positions only, flux handling needed. Pressure vessel longitudinal and circumferential seams, heavy girders
Resistance spot, seam and projection welding	Copper electrodes apply pressure plus short high current pulse, no filler	Very fast sheet joining, no consumables, automation friendly, nugget quality by current and time	Limited to lap sheet, electrode wear, shunting in close spots. Auto body panels, enclosures, wire mesh
Oxy fuel gas welding and cutting	Flame plus filler rod, flux where needed	Portable, no electricity, good for thin sheet, brazing and cutting	Wide heat zone, slower, gas safety discipline. Repair, thin sheet, site brazing
Electroslag and electrogas for vertical thick joints, plasma and laser for precision	Molten slag resistance or focused beam	Single pass very thick sections or narrow deep precision welds	Specialised equipment and joint preparation. Heavy columns, precision sheet, electronics

TIG versus MIG reasoning is compact. Thin sections, reactive metals, cosmetic quality and precise heat control point to TIG. Thicker sections, long production runs and deposition rate point to MIG. Thick pressure vessel seams point to SAW with radiographic or ultrasonic proof. Sheet assemblies in huge numbers point to resistance welding.

10.7.4 Soldering versus brazing and the 450 C line

The filler melting temperature separates the two non fusion joining families. Soldering uses filler melting below 450 C, typically tin lead or lead free tin alloys, with flux to clean surfaces. Joints are leak tight and electrically continuous but mechanically modest. Electronics, plumbing seals and instrument joints are representative.

Brazing uses filler melting above 450 C but below the melting point of the base metals, typically copper zinc, copper phosphorus or silver base alloys. Capillary flow fills a close clearance. Strength is far higher than soldering and approaches base metal in well designed lap joints. Carbide tips, heat exchanger joints, bicycle frames and dissimilar metal joints are representative.

Welding differs fundamentally because it melts the base metals and fuses them, with or without filler of matching composition. Strength can equal plate strength when procedure and inspection are correct. That is why pressure boundaries are welded, not soldered.

10.7.5 Weld defects and distortion control

Porosity arises from damp flux, moist shielding, rust, oil or dissolved gases trapped on freezing. Control lies in drying, cleaning, gas coverage and correct arc length.

Slag inclusion arises from poor interpass cleaning, wrong electrode angle or low current that prevents slag floating. Control lies in chipping, grinding, weaving discipline and adequate current.

Lack of fusion and incomplete penetration arise from low heat input, fast travel, wrong groove angle or misaligned wire. Control lies in raising current, slowing travel within the qualified window and correcting joint geometry.

Undercut is a groove melted at the toe left unfilled because current is too high or travel too fast. Control lies in lowering current, pausing at edges and correcting torch angle.

Hydrogen induced cold cracking appears hours after welding in hardenable steels when diffusible hydrogen, hard martensitic microstructure and restraint combine. Control lies in low hydrogen electrodes, preheat, interpass control, post heat and carbon equivalent management.

Solidification cracking arises from sulphur, phosphorus, deep narrow beads and high restraint. Control lies in chemistry control, wider bead shape and fixture relief.

Distortion arises from non uniform expansion and contraction. Control methods include backstep sequencing, skip and balanced welding about the neutral axis, presetting parts in the opposite direction, rigid clamping and jigs, minimum number of passes consistent with quality, correct joint gap and tack discipline, and post weld straightening or stress relief where code permits. Thin sheet uses copper backing and stitch sequencing to limit heat build up.

10.8.1 Rolling

Rolling compresses slab, bloom or billet between rotating rolls to reduce thickness and elongate length. Draft equals entry thickness minus exit thickness. True strain accumulates over passes. The neutral point is where strip speed equals roll surface speed. Entry side shows backward slip, exit side shows forward slip. Roll force rises with draft, friction, flow stress and contact length. Roll torque follows force times lever arm. Hot rolling breaks cast structure, refines grains and permits large reductions with lower forces. Cold rolling gives bright finish, tight tolerance and strain hardening at the cost of higher forces and annealing needs.

Defects include alligatoring from split flow, edge cracking from non uniform spread, wavy edges and centre buckle from roll bending, and scale pitting from poor descaling. Cambered rolls, bending jacks, tension control and edge trimming maintain flatness. Products range from plate and sheet to rails, beams, seamless tube hollows and foil.

10.8.2 Forging

Forging deforms heated metal under compressive blows or squeezes to refine grains and align grain flow with part contour, giving superior toughness relative to cast equivalents.

Open die forging uses flat or shaped dies with free lateral spread for upsetting, drawing out, fullering, punching and bending of large shafts, rings and custom blanks. Closed die or impression die forging encloses metal in engraved cavities. Flash extrudes into a thin gutter around the parting line. Flash chills and raises pressure, forcing complete die fill and expelling surface scale into the gutter. Flash is later trimmed. Flashless forging uses exact volume control to avoid waste.

Forging load depends on flow stress, friction, projected area and flash geometry. Proper draft, fillets, preform shape and lubricant selection prevent laps, cold shuts and underfill. Warm and cold forging extend the idea to net shape fasteners and gears with superb finish and no scale.

10.8.3 Extrusion

Extrusion pushes a heated billet through a die to create long constant section products. Direct extrusion moves ram and billet together against a stationary die. Friction along the container wall is high, load rises, but handling is simple. Indirect extrusion moves a hollow ram with the die into a stationary billet. Friction is far lower and load falls, but equipment is more complex. Hydrostatic extrusion surrounds the billet with fluid to nearly eliminate friction for brittle materials.

Extrusion ratio equals initial cross sectional area divided by final area and measures severity. Hollow tubes need a mandrel or bridge die to form the bore. Lead for collapsible tubes, aluminium window sections, copper bus bars and steel spline shafts illustrate the range. Defects include pipe or funnel at the tail from oxide inflow, surface cracking from excessive speed and temperature, and centre burst from wrong die angle. Glass lubrication suits hot steel extrusion.

10.8.4 Wire drawing and tube drawing

Drawing pulls rod or tube through a converging die to reduce diameter. The die has entry bell, approach cone, bearing land and back relief. True strain sums logarithmically over passes, permitting schedule design. Small reductions with large die angles promote centre burst, also called chevron or cuppy core, through excessive redundant shear. Large reductions raise drawing stress toward wire breakage. Back tension andarea schedule control stability. Lubrication by soap, lime or phosphate plus good die polish governs finish and die life. Fine copper wire, piano wire, hypodermic tubing and condenser tubes are representative products.

10.8.5 Sheet metal forming with springback, holder force and draw ratio

Blanking, punching, bending, deep drawing, stretch forming and spinning shape sheet without appreciable thickness change except in stretching zones.

Springback is elastic recovery after unloading. The bent sheet opens by a few degrees because the outer fibres that were stretched elastically contract. Compensation uses overbending, coining the bend radius, bottoming the punch or applying tension. High strength steels and aluminium show larger springback because yield to modulus ratio is higher.

Blank holder force suppresses wrinkling in the flange during deep drawing by clamping the sheet against the die. Insufficient force lets the compressive hoop stress buckle the flange into wrinkles. Excessive force prevents radial feeding, raises wall tension and tears the cup at the punch radius. Lubrication, draw bead geometry and punch nose radius tune the window between wrinkling and tearing.

Limiting draw ratio is the maximum blank diameter divided by punch diameter that can be drawn in one step without failure, commonly near 2.0 for ductile sheet in the first draw and lower in redraws. Earing from planar anisotropy, orange peel from coarse grains and stretcher strains from yield point elongation are additional sheet phenomena controlled by texture, grain size and temper rolling.

10.8.6 Powder metallurgy

Powder metallurgy blends elemental or prealloyed powders, compacts them in a die to green strength, sinters below the melting point to bond particles by diffusion, then sizes, coins, infiltrates or heat treats as needed. Porosity is controlled, not eliminated. Oil impregnated porous bronze bearings hold lubricant in pores and run without external oil. Cemented carbide tips sinter tungsten carbide in cobalt binder to extreme hardness. Filters, friction parts, soft magnetic cores and small complex gears exploit net shape capability without machining.

Advantages are material utilisation, composition control, fine microstructure and suitability for refractory and immiscible systems. Limitations are size limits from press capacity, lower impact toughness than forged equivalents because of residual porosity, and powder cost and handling discipline including explosion and inhalation controls.

10.9.1 Merchant circle and shear angle

Orthogonal cutting idealises the tool as a wedge with rake angle alpha removing a layer of uncut thickness t0 to form a chip of thickness tc. Chip thickness ratio r equals t0 divided by tc and is less than unity because the chip thickens in shear. Shear occurs on a plane at shear angle phi to the cutting direction. From chip geometry, tan phi equals r cos alpha divided by one minus r sin alpha. Larger rake and thinner chips raise phi, shorten the shear plane, lower cutting force and improve finish, at the cost of weaker tool edge.

Merchant force circle closes shear force, normal shear force, friction force, normal friction force and resultant. Friction angle beta follows from the ratio of friction to normal force on the rake face. Minimum power shear angle approximates phi equals 45 degrees plus alpha over two minus beta over two. Lower friction through coatings and cutting fluids raises phi and lowers power.

Chip forms complete the picture. Continuous chips from ductile metals at high speed with good lubrication give fine finish. Discontinuous segmented chips from brittle metals such as gray cast iron break readily and protect the tool but leave matte surfaces. Continuous chips with built up edge arise when workpiece welds to the rake face at moderate speed and then tears off periodically, degrading finish. Raising speed, raising rake, applying coolant and using coated carbides suppress built up edge.

Oblique cutting adds an inclination angle and a side cutting edge angle, turning the chip flow three dimensional and distributing wear, which is why production turning tools are oblique rather than strictly orthogonal.

10.9.2 Taylor tool life with exponent reasoning

Cutting speed dominates tool life through Taylor equation V times T raised to n equals C, where V is cutting speed, T is life to a defined wear land, n is the Taylor exponent and C is a constant for a given tool work pair and feed and depth. In logarithmic form the line slope is minus one over n, so small n means steep life fall with speed.

Typical n ranges explain tool selection. High speed steel shows n about 0.10 to 0.15, meaning life is extremely sensitive to speed and economical speeds are modest. Cemented carbide shows n about 0.20 to 0.30, permitting far higher speeds for the same life. Ceramic shows n about 0.40 to 0.60, cubic boron nitride higher still, diamond highest for non ferrous work. Higher n flattens the life curve and justifies the higher tool cost through productivity.

A direct consequence is that doubling speed with n equal to 0.25 divides life by sixteen because life scales with speed to the power minus one over n. Rapid wear complaints after a seemingly small speed increase are therefore arithmetic, not random variation. Feed and depth enter through extended Taylor forms, but speed remains the dominant lever.

Tool wear modes include flank wear from abrasion measured as VB land, crater wear on the rake from diffusion and adhesion, notch wear at the depth line, plastic deformation at high temperature and chipping or fracture from overload. Taylor life is normally defined at a flank wear limit such as 0.3 mm for finishing and higher for roughing.

10.9.3 Tool geometry and tool materials

Single point tool geometry balances chip flow, strength, heat and finish. Back rake and side rake control chip direction and cutting force. Larger positive rake lowers force but weakens the edge. Clearance angles prevent rubbing on the freshly cut surface. Excess clearance weakens the edge and invites chatter. Side cutting edge angle distributes load and thins the chip. End cutting edge angle clears the trailing edge. Nose radius strengthens the tip, improves finish by reducing feed marks and dissipates heat, but too large a radius invites vibration on slender parts.

Tool material ladder climbs hardness and hot hardness at the cost of toughness and price. Plain carbon tool steel suits low speed hand tools. High speed steel retains hardness to about 600 C and tolerates shock, suiting drills, taps and formed tools. Cemented carbide combines tungsten carbide grains in cobalt binder and dominates production turning and milling. Cermets blend ceramic and metal for fine finishing of steels. Ceramics including alumina and silicon nitride permit very high speed finishing of cast iron and hardened steel but are brittle. Cubic boron nitride machines hardened steels and white iron where carbide fails. Diamond, either polycrystalline or single crystal, machines aluminium, copper, composites and ceramics with mirror finish but must never cut steel because carbon dissolves in iron at cutting temperature and wear becomes catastrophic.

Coatings extend the ladder. Titanium nitride, titanium carbonitride, aluminium oxide and diamond like carbon lower friction, block diffusion and add thermal barrier function. Multilayer and nanocomposite coatings tune toughness and hot hardness for specific work materials.

10.9.4 Machining economics

Two optimum speeds govern practice. Minimum cost speed minimises total cost per part, summing machining cost, tool cost amortised over life and tool change cost including downtime. Maximum production speed minimises time per part, ignoring tool cost. Maximum production speed always exceeds minimum cost speed because it trades extra tool consumption for saved machine time. High overhead shops with expensive machines run nearer maximum production. Job shops with low machine rates run nearer minimum cost. Feed and depth are raised first for economy because they increase removal rate with less life penalty than speed. Speed is set last to fill the remaining tool life budget.

10.9.5 Grinding and finishing

Grinding removes metal with bonded abrasive grains at very high speed and small chip load. G ratio equals stock removed divided by wheel wear and measures wheel economy. Hard work needs a soft grade wheel so dull grains release and fresh sharp grains expose, a self sharpening rule that reverses novice intuition. Hard grade wheels suit soft gummy work where grain retention matters. Dressing fractures or exposes grains to restore cutting ability. Truing reshapes the wheel profile concentric to the spindle. Creep feed grinding takes deep slow cuts with continuous dressing for turbine slots and form profiles. Honing, lapping, superfinishing and burnishing follow grinding to improve form, finish and bearing ratio. Grinding burns, tensile residual stress and surface cracking are controlled by gentle parameters, sharp dressed wheels, adequate coolant and spark out passes.

10.9.6 Non traditional processes with wear logic

Process	Removal mechanism	Tool and wear behaviour	Best domain
Electric discharge machining	Repeated sparks in dielectric melt and vaporise microscopic craters	Copper or graphite tool wears by same sparks. Wear ratio of work to tool matters in sizing electrodes. Orbiting and multiple electrodes manage wear	Hardened dies, blind cavities, sharp internal corners, wire cut profiles in tool steel and carbide
Electrochemical machining	Anodic dissolution in electrolyte under high current at small gap	Cathode tool ideally suffers zero wear because removal is ionic, not mechanical or thermal. Electrolyte management dominates	Burr free complex aerospace shapes, blisks, deep slots, parts demanding zero residual stress and zero recast
Ultrasonic machining	Abrasive slurry hammered by sonotrode at ultrasonic frequency micro chips brittle work	Sonotrode and abrasive both wear. Tool wear is significant and profile compensation is needed	Glass, ceramics, quartz, hard brittle stones and semiconductor features
Abrasive jet machining	Gas borne abrasive jet erodes by micro cutting	Nozzle wears by same jet and needs periodic replacement. Abrasive consumption dominates cost	Deburring, cutting thin brittle sheet, cleaning and frosting without heat
Laser beam machining	Focused photons melt and vaporise with assist gas ejection	No contact, hence zero mechanical tool wear. Lens and nozzle maintenance replaces tool wear	Micro holes, fine cutting, marking, thin sheet profiling with narrow kerf
Electron beam machining	Focused electrons in vacuum melt and vaporise	No contact, hence zero mechanical wear. Vacuum and X ray shielding dominate	Deep narrow welds and holes in reactive and refractory metals
Water jet and abrasive water jet	High pressure jet plus abrasive erosion	No thermal damage. Focusing tube wears	Thick plate, composites, stone and food where heat must be avoided
Chemical and photochemical machining	Controlled etching through masks	No tool contact, etchant consumption and mask control dominate	Thin shims, meshes, printed features, burr free light parts

Examination logic is direct. Processes with spark or abrasive contact wear the tool, namely EDM, USM, AJM and water jet focusing tubes. Processes with ionic or photonic removal and no contact show essentially zero mechanical tool wear, namely ECM, laser and electron beam. ECM is the canonical zero wear contact process because the cathode does not participate electrochemically.

10.10.1 Least count and length instruments

Least count is the smallest measurement resolvable by an instrument and equals one main scale division divided by number of vernier or drum divisions. Vernier calliper with 1 mm main divisions and 50 vernier divisions where 50 vernier divisions span 49 mm gives least count 0.02 mm. Micrometer with 0.5 mm pitch and 50 thimble divisions gives least count 0.01 mm. Dial indicators amplify spindle motion through gearing for comparative measurement of runout and deflection. Slip gauge blocks provide end standards for calibration by wringing into stacks. Height gauges, depth gauges, bore gauges and air gauges extend the same least count idea to specific geometries. Abbe principle states that measuring axis should be collinear with scale axis to avoid cosine error, which is why micrometers are inherently more accurate than callipers for the same least count.

10.10.2 Angular measurement and sine bar principle

Sine bar converts length into angle through trigonometry. A hardened bar with two rollers of equal diameter spaced at gauge length L rests on a datum. Slip gauges of total height h raise one roller. The bar tilts at angle theta where sin theta equals h divided by L. Standard lengths are 100 mm and 200 mm. Accuracy is highest below about 45 degrees because sine sensitivity falls at high angles and stack error magnifies. Rollers must sit on a clean datum without clamping distortion. Sine centres, sine tables, bevel protractors, autocollimators and angle deckers extend angular capability to tapers, threads and machine alignment.

10.10.3 Fits, tolerances and gauge philosophy

A fit is the assembly relation between hole and shaft. Clearance fit always leaves space for sliding and lubrication, exemplified by H7 g6 for smooth sliding spindles. Transition fit may leave clearance or interference in different assemblies, exemplified by H7 k6 for located keys that tap home. Interference fit always overlaps and needs pressing or thermal assembly, exemplified by H7 p6 for gear hubs and bearing seats.

Hole basis system keeps hole as basic with lower deviation zero, denoted H, and varies shaft letters to obtain clearance through interference. Shaft basis reverses the roles and suits long shafting with varied accessories. Tolerance grade number sets magnitude, lower numbers meaning finer tolerance and higher cost. IT7 and IT6 are typical for precise running fits. Surface finish, form tolerance and temperature discipline accompany tight grades because thermal expansion can swallow small clearances.

Taylor gauge principle governs limit gauges. GO gauge checks maximum material condition and full form assemblability, so it must enter or pass over the feature. NO GO gauge checks minimum material condition at a single element, so it must not enter. GO gauges are full form plugs and rings. NO GO gauges are segmental or pin checks. This asymmetry prevents acceptance of parts that assemble locally but fail functionally.

10.10.4 Control charts and operating characteristic idea

Statistical quality control separates chance variation from assignable causes. Variables data uses measurements. X bar chart tracks subgroup mean and detects shifts in process centring. R chart or S chart tracks range or standard deviation and detects changes in spread. They are used together because mean can stay constant while spread explodes. Attributes data uses counts. p chart tracks fraction defective with variable sample size. np chart tracks number defective with constant sample size. c chart tracks defects per unit such as pits per casting. u chart tracks defects per unit with variable inspection area.

Control limits are set from a stable process, commonly at plus minus three standard errors, distinct from specification limits set by design. A point outside control limits, a run of points on one side, or a trend signals an assignable cause demanding stoppage and correction. Capability indices Cp and Cpk then compare natural spread and centring against specification width to judge whether a stable process is also adequate.

Operating characteristic curve plots probability of lot acceptance against true fraction defective for a sampling plan. A steep curve discriminates sharply between good and bad lots. Producer risk alpha is the chance of rejecting a good lot at acceptable quality level. Consumer risk beta is the chance of accepting a bad lot at limiting quality level. Tightened sampling, larger samples and smaller acceptance numbers steepen the curve but raise inspection cost.

10.11.1 Work study and method analysis

Method study improves how work is done through record, examine, develop, install and maintain. Charts include operation process, flow process, man machine, two handed and multiple activity diagrams. Gilbreth Therbligs decompose hand motion into search, select, grasp, transport, position, assemble, use, release and delay elements, exposing wasteful reaches and holds. Principles of motion economy favour simultaneous symmetric motions, gravity feeding, jigs over holding, and workplace layout within normal reach envelopes.

Time study sets standard time. Observed time averaged over cycles is levelled by performance rating to normal time. Allowances for personal needs, fatigue and unavoidable delays are added to give standard time used for planning, costing and incentives. Work sampling estimates delay proportions by random observations without stopwatch pressure. Predetermined motion time systems synthesise times from tabulated Therbligs for new jobs. Ergonomics, illumination, noise and safety bound every productivity gain.

10.11.2 Plant layout and material handling

Product or line layout arranges machines in process sequence for high volume low variety flow such as refineries, bottling and automotive lines. Throughput is high and unit cost low, but breakdown of one station stalls the line and variety is costly.

Process or functional layout groups similar machines such as lathes, mills and welders into departments for job shops with high variety. Routing is flexible but material travel is long and scheduling is complex.

Fixed position layout keeps the product stationary and brings men and machines to it. Ships, pressure vessels, furnaces and refinery shutdown jobs illustrate the model.

Cellular or group technology layout forms families of similar parts and dedicates a cell with needed machines, blending flow efficiency with variety capability. Line balancing then distributes work elements among stations to minimise idle time and maximise efficiency for a given cycle time.

Material handling doctrine is shortest distance, unit load, gravity feed, mechanised lift and safe stacking. Conveyors suit steady flow, forklifts suit flexible lots, cranes suit heavy lifts, pipelines suit fluids. Poor handling adds cost without adding value, so layout and handling are designed together.

10.11.3 Inventory and derivation of EOQ

Inventory decouples supply from demand but ties up capital, space and risk. Annual relevant cost for the basic lot size model sums ordering cost and carrying cost. Let D be annual demand, S ordering cost per order, H holding cost per unit per year, Q order quantity. Number of orders per year equals D divided by Q. Average inventory under instant replenishment and no stockouts equals Q divided by two. Total cost TC equals D divided by Q times S plus Q divided by two times H. Differentiate with respect to Q and set to zero. Minus D S divided by Q squared plus H divided by two equals zero. Rearranging gives Q squared equals two D S divided by H. Optimal quantity Q star equals square root of two D S divided by H. Second derivative is positive, confirming a minimum. At optimum, ordering cost equals holding cost, so total minimum cost equals square root of two D S H. Number of orders equals D divided by Q star. Cycle time equals Q star divided by D.

Assumptions are constant known demand, fixed ordering cost, linear holding cost, instant replenishment, no quantity discounts and no stockouts. Relaxing each assumption creates a variant. Finite replenishment rate gives economic production quantity with a correction for usage during production. Planned backorders lower optimal lot size at the cost of shortage penalty. Quantity discounts shift the optimum to price break evaluation by comparing total cost at EOQ and at break points.

Reorder point triggers replenishment so stock arrives exactly when needed. With constant demand rate d and lead time L, reorder point equals d times L plus safety stock. Safety stock covers demand variability and lead time variability for a chosen service level. Higher service level needs larger safety multiple of demand standard deviation over lead time.

Quick Example — EOQ Near 346 Units

Given $D = 6000$ units per year, $S = 200$ per order, and $H = 20$ per unit per year.

Step 1: use $Q^{\star} = \sqrt{2DS/H}$.

Step 2: substitute $Q^{\star} = \sqrt{120000} \approx 346.4$ units.

Step 3: check $TC = DS/Q + HQ/2$, giving about $6928$ at $346$, versus $7000$ at $300$ and $7000$ at $400$.

Result: the optimum is about $346$ units with minimum cost about $6928$ per year.

Trap: the bracket check must compare full cost $TC$, not ordering cost alone, and rounding $Q$ keeps cost near flat at the minimum.

10.11.4 ABC, VED, FSN and related classifications

ABC analysis sorts items by annual usage value, meaning annual demand multiplied by unit price. A items are few in number but dominate value and receive tight control, frequent review, accurate records and low safety cover. B items receive normal control. C items are numerous but trivial in value and receive simple bulk policies with higher safety cover to avoid stockouts of cheap parts. The Pareto pattern makes this segregation highly economical.

VED analysis sorts spares by criticality. Vital items stop the plant and need assured availability, often through redundancy or local stock. Essential items degrade performance and need moderate cover. Desirable items cause minor inconvenience and can be ordered on need. Refinery maintenance spares rely heavily on VED logic.

FSN analysis sorts by movement speed into fast, slow and non moving items, guiding layout, review frequency and obsolescence action. HML sorts by unit price. SDE sorts by supply difficulty and lead time reliability. XYZ sorts by demand variability. Combined matrices such as ABC VED balance value against criticality so that cheap but vital seals are not starved while expensive but desirable trim is not overstocked.

10.11.5 Queuing models and M M 1 meanings

Queuing theory analyses waiting lines at counters, machines, berths and repair bays. Kendall notation A slash B slash c describes arrival process, service process and number of servers. M denotes memoryless exponential inter arrival or service times, equivalent to Poisson arrivals. D denotes deterministic fixed times. G denotes general distribution. M M 1 therefore means single server with Poisson arrivals at rate lambda and exponential service at rate mu. Traffic intensity rho equals lambda divided by mu and must be below unity for steady state, otherwise the queue grows without bound.

Standard steady state results for M M 1 are compact. Mean number in system L equals rho divided by one minus rho. Mean number waiting in queue Lq equals rho squared divided by one minus rho. Mean time in system W equals L divided by lambda, which simplifies to one divided by mu minus lambda. Mean waiting in queue Wq equals Lq divided by lambda, which simplifies to rho divided by mu minus lambda. Probability of empty system equals one minus rho. Probability of n in system equals one minus rho times rho to the power n.

Managerial reading is direct. Utilisation near unity inflates queues nonlinearly. At 80 percent utilisation mean number in system is four and mean waiting is four times service time. Relief comes from adding a server, cutting mean service time, smoothing arrivals by appointment or triage, or pooling separate queues into one. M M c, M G 1 with Pollaczek Khinchine correction and finite population variants extend the same logic to multiserver, general service and closed repair shops.

10.12.1 Linear programming methods tour

Linear programming maximises or minimises a linear objective subject to linear constraints and non negativity. Graphical method solves two variable problems by plotting constraints, identifying the convex feasible polygon and sliding the objective line to the extreme corner. Corner point theorem guarantees an optimum at a vertex when one exists.

Simplex method pivots algebraically among basic feasible solutions. Slack and surplus variables convert inequalities to equalities. The entering variable is the one with most favourable reduced cost, Cj minus Zj for maximisation. The leaving variable follows the minimum ratio test to preserve feasibility. Iterations continue until all reduced costs satisfy optimality. Degeneracy, unboundedness, infeasibility and alternate optima are diagnosed from tableau patterns.

Duality associates every primal with a mirror dual. Primal maximisation with less than constraints becomes dual minimisation with greater than constraints. Weak duality bounds objective values. Strong duality equates optima when both are feasible. Dual variables are shadow prices measuring marginal value of each resource, central to sensitivity analysis and to economic interpretation of refinery blending and capacity decisions.

Transportation problems ship from sources to destinations at minimum cost. Starting solutions use North West corner for simplicity, Least Cost for greediness and Vogel approximation for penalty awareness, with Vogel generally nearest optimal. Optimality test uses MODI or UV method with occupied cell potentials and evaluates unoccupied cell opportunity costs. Stepping stone reallocation improves the plan until all opportunity costs are non negative.

Assignment problems allocate one job to one worker or machine to minimise total cost or time. Hungarian method subtracts row minima then column minima, covers zeros with minimum lines, adjusts uncovered elements and iterates until a full zero assignment is possible. Dummy rows or columns balance rectangular problems. Maximisation variants convert by opportunity loss. Travelling salesman and sequencing variants extend the same combinatorial core.

10.12.2 Forecasting logic and error control

Simple moving average smooths noise by averaging the last N demands. Larger N gives stability but lags trends and turning points. Weighted moving average gives more weight to recent periods while retaining smoothing. Exponential smoothing updates recursively through new forecast equals alpha times latest demand plus one minus alpha times old forecast. Large alpha near unity chases recent demand, responding fast but passing noise through. Small alpha near zero is stable but sluggish on regime change. Trend corrected Holt method adds slope smoothing. Seasonal Holt Winters adds seasonal indices for refinery fuel demand, festival loads and climatic cycles.

Error measurement guides selection. Mean absolute deviation gives robust average error. Mean squared error penalises large misses. Mean absolute percentage error scales across items. Bias is tracked by cumulative sum of errors and by tracking signal equal to cumulative error divided by mean absolute deviation. Tracking signal beyond about plus minus four indicates systematic bias and demands model revision or parameter retuning. Qualitative Delphi, market survey and analogy methods complement time series when history is absent, as in new grade launches.

10.12.3 MRP versus EOQ and JIT Kanban

EOQ suits independent steady demand where each item can be lot sized alone. Material requirements planning suits dependent lumpy demand where need for components derives from parent schedules. MRP explodes the bill of materials through the master production schedule, adjusts for inventory on hand and scheduled receipts, offsets by lead times and generates planned order releases by lot sizing rules such as lot for lot, EOQ or periodic order quantity. Capacity requirements planning then checks load against work centres. Success needs accurate bills, inventory records, lead times and schedule discipline. Nervousness from frequent rescheduling is dampened by freezing near horizons.

Just in time pulls small lots only when consumed downstream, minimising work in process and exposing defects immediately. Kanban cards or containers authorise production and movement. Number of kanbans equals demand during lead time plus safety stock divided by container size. Single minute exchange of dies, total productive maintenance, poka yoke mistake proofing, levelled heijunka schedules and close supplier partnerships make pull feasible. EOQ push and JIT pull are therefore complements separated by demand character. Stable independent spares suit EOQ. Assembly dependent parts with reliable suppliers and short setups suit JIT.

10.12.4 PERT versus CPM with time statistics and crash logic

PERT, programme evaluation and review technique, handles uncertain one off work with three estimates per activity. Optimistic time o assumes everything goes well. Most likely time m assumes normal conditions. Pessimistic time p assumes major difficulties. Expected time te equals o plus four m plus p divided by six, weighting the likely outcome most heavily under a beta distribution assumption. Standard deviation sigma equals p minus o divided by six, so variance equals sigma squared. Project expected duration Te is the sum of te over critical path activities. Project variance is the sum of critical variances only, non critical slack absorbing local variation. Project standard deviation is the square root of that sum. Probability of meeting a scheduled date Ts follows the standard normal variate Z equals Ts minus Te divided by project standard deviation. This machinery suits research, shutdown discovery work and first of a kind jobs.

CPM, critical path method, uses single deterministic durations and focuses on cost time tradeoff. Network analysis computes earliest start and finish by forward pass and latest start and finish by backward pass. Total float is latest minus earliest. Critical path is the longest continuous chain with zero float and sets project duration. Free float and interfering float refine scheduling flexibility.

Crashing shortens selected activities by adding resources, overtime or parallel crews. Crash slope equals crash cost minus normal cost divided by normal time minus crash time, giving cost per day saved. Economic crashing shortens the cheapest critical activity first, recomputes the network because a new path may become critical, and stops when marginal saving from earlier completion no longer exceeds marginal crash cost or when no further shortening is technically possible. Refinery shutdowns are textbook CPM crash settings where each day saved restores production margin, while emergent discovery scope retains PERT style uncertainty. Resource levelling, least cost scheduling and time scaled networks complete the project toolkit.

Quick Example — PERT Expected Time and Spread

Given $o = 3$ days, $m = 6$ days, and $p = 12$ days.

Step 1: use $t_e = (o + 4m + p)/6$.

Step 2: substitute $t_e = (3 + 24 + 12)/6 = 6.5$ days.

Step 3: use $\sigma = (p - o)/6 = 1.5$ days, so variance $\sigma^2 = 2.25$.

Result: expected duration $6.5$ days with standard deviation $1.5$ days.

Trap: only critical path variances add to project variance, never the sum over all activities.

• Lattice geometry governs behaviour. Close packed FCC dissolves carbon, deforms readily and appears non magnetic as austenite. BCC ferrite is magnetic and temperature sensitive. HCP offers few easy slip systems and resists cold forming. Strained BCT martensite is hard and brittle. Vacancies enable diffusion, dislocations carry plasticity, boundaries block slip through the Hall Petch relation.
• Six iron carbon numbers organise steels and irons. Concentrations 0.025, 0.77, 2.11 and 4.3 percent with temperatures 727 C and 1147 C locate ferrite, austenite, cementite, pearlite, ledeburite, martensite and bainite. Hypo eutectoid structures blend ferrite with pearlite. Hyper eutectoid structures add grain boundary cementite. Cast irons extend beyond 2.11 percent toward the eutectic for fluidity and damping at the cost of ductility.
• Alloying and heat treatment tune the as cast or as rolled structure. Nickel toughens and stabilises austenite. Chromium builds passive films and carbides. Molybdenum guards high temperature strength. Vanadium refines grains. Manganese neutralises sulphur and adds hardenability. Tungsten confers hot hardness. Annealing softens, normalising refines, quenching hardens to martensite, tempering restores toughness, austempering yields bainite, case methods harden skins on tough cores. The TTT nose dictates whether a given cooling rate secures martensite or loses hardness to pearlite.
• Inspection follows material and flaw physics. Visual appraisal comes first. Magnetic particle suits ferromagnetic surface cracks. Dye penetrant suits non magnetic surface cracks. Ultrasonic maps interior and thickness. Radiography archives volumetric weld quality. Eddy current surveys conductive tubing. Non ferrous selection follows function, with brass for corrosion plus formability, bronze for bearings, duralumin for airframes, invar for dimensional stability and nickel alloys for hot corrosive service.
• Shape processes obey thermal and deformation logic. Chvorinov square law with riser last and chill assisted gradients controls casting soundness. Gating balances calm filling against slag and air. Heat input V I over v with polarity control governs weld penetration and cleaning. Solder stays below 450 C filler melting, braze stays above it without melting base metal, welding fuses base metal. Rolling, forging, extrusion, drawing and sheet forming each carry signature defects and remedies, from springback and holder force to draw ratio limits. Powder metallurgy bonds below melting for net shape porous or hard parts.
• Removal, measurement and planning close the loop. Merchant tan phi relation and Taylor V T to n law with material dependent exponent govern cutting and tool life economics. Grinding grade logic and non traditional wear behaviour separate spark and abrasive wear from zero wear ionic and photonic removal. Least count, sine bar trigonometry, fit classes, GO NO GO philosophy, control chart selection and operating characteristic curves assure dimensions. Work study, layout choice, EOQ derivation with ABC VED FSN control, M M 1 queuing relations, linear programming tour, forecasting discipline, MRP versus EOQ versus JIT pull, and PERT te sigma statistics with CPM crash slopes schedule the enterprise in which all hardware is made and maintained.