Welding Metallurgy Explained: The Science Behind Stronger, Crack-Free Welds

A fabricator once spent $40,000 on a hydraulic ram that cracked 48 hours after it passed visual inspection—no porosity, no undercut, just a hairline running under the bead where nobody thought to look. The welder wasn’t careless; he simply didn’t know that the steel’s chemistry and the cooling rate had quietly conspired to form a brittle martensite layer soaked with hydrogen. That gap between ‘looks good’ and ‘is good’ is exactly what welding metallurgy fills.

Why Welding Metallurgy Decides Whether Your Weld Lives or Fails

The ram cost $40,000, and it cracked 48 hours after the weld passed visual inspection. No porosity, no undercut, no spatter problems, nothing a inspector could point at with a mirror and a flashlight. Just a hairline fracture creeping along the root, underneath a bead that looked textbook perfect. The welder wasn’t careless either. He ran a clean pass, kept a steady arc, and followed the procedure he was given. What he didn’t know was that the steel’s chemistry and the cooling rate had quietly joined forces to form a hard, brittle layer of martensite that was also saturated with dissolved hydrogen. Nothing about that recipe shows up on the surface until the crack decides to announce itself.

That gap between “looks good” and “is good” is the entire reason welding metallurgy exists.

What Welding Metallurgy Actually Means

In plain language, welding metallurgy is the study of how heat, chemistry, and cooling speed reshape a metal’s internal structure while you join it. It’s not about arc length or puddle control, though those matter. It’s about what happens to the atoms after the arc moves on. The puddle you watched freeze in a second or two was a liquid undergoing a phase transformation in real time, and the speed of that transformation decides which microstructure you end up with: soft ferrite, tough pearlite, acicular bainite, or the hard, brittle martensite that ruins careers.

According to ASM International, every fusion weld splits into three distinct metallurgical regions: the weld metal (the pool that melted and re-solidified), the heat-affected zone (base metal that never melted but was transformed by heat), and the unaffected base metal beyond it. Each zone has its own microstructure, its own hardness, and its own failure modes. Most weld failures don’t happen in the bead you can see. They happen in the heat-affected zone, the thin band of steel that nobody photographs.

Why the Cooling Rate Matters More Than the Amperage

Welding is one of the few manufacturing processes where the final mechanical properties are set not by what you dialed in on the machine, but by how fast the metal cooled afterward. Amperage, voltage, and travel speed matter only in that they feed into heat input, and heat input feeds into cooling rate. A weld made at the wrong cooling speed can transform a perfectly good structural steel into something with the toughness of a file. This is why two welders running identical settings on different plate thicknesses can get completely different results: the thick plate acts as a heat sink and pulls heat away faster.

Who Needs This Knowledge, and What They’re Looking For

People land on this topic for very practical reasons. A welder moving from hobby work to pipe welding procedures suddenly has to justify preheat and interpass temperatures to an inspector. A welding student is staring down a CWI exam and a chapter on phase diagrams that reads like a foreign language. A shop supervisor keeps seeing rework on the same joint design and needs to know whether the problem is the electrode selection, the heat input, or the steel itself. Fabricators working with unfamiliar mill certificates want to know what they’re actually buying.

All of those readers need the same foundation: how metal behaves when it’s melted and cooled under field conditions, and which levers they control in real time.

The Invisible Software Under Every Arc

Think of metallurgy as the software running underneath every arc, rod, and procedure you touch. The machine is the hardware, the WPS is the program, but the microstructure is what actually executes. Ignoring it doesn’t make the problem go away; it just moves the failure somewhere you can’t see it, or later than you’re around to catch it. Welding metallurgy also has a human dimension, because fume chemistry is metallurgy too. Welding fume carries manganese, hexavalent chromium, and other heavy metals, and OSHA treats it as a serious occupational hazard, which makes proper gear part of the science rather than an afterthought.

The next section builds the foundation everything else rests on: phases, crystal structures, and what really happens when a weld puddle freezes.

Welding Metallurgy 101: Phases, Crystal Structures and Solidification

Before you can work out why a weld cracks three days later, you need a picture of what is happening inside the steel while the arc runs. A handful of ideas explains almost everything that follows.

Iron’s Two Crystal Structures: BCC and FCC

Steel is mostly iron, and iron atoms can pack themselves in two different arrangements. At lower temperatures they form a body-centered cubic (BCC) structure, with one atom at each corner of a cube and another in the middle. This form is called ferrite. Heat it above roughly 912 °C and the atoms rearrange into a face-centered cubic (FCC) lattice, called austenite, which packs atoms more tightly and can hold far more carbon in solution.

That temperature swap is unremarkable in a heat-treating furnace where parts cool over hours. In welding it is anything but calm. The metal just behind a moving arc swings through those transitions in seconds, then swings back. As the ASM monograph on metallurgy variables in fusion welding describes, the rapid thermal cycle is precisely what gives welding its unique metallurgical character: the same steel can end up with several completely different microstructures within a few millimetres.

The Iron–Carbon Diagram in Plain Language

The iron–carbon diagram is a map of which structure is stable at a given temperature and carbon content. Low-carbon steels sit in ferrite; move right toward higher carbon, or up in temperature, and austenite takes over. The lines show when phase changes begin and end, which is how engineers predict what a weld or HAZ will look like. You do not need to memorize it, only to know that carbon shifts every line and that cooling speed decides whether the steel follows the map or takes a shortcut.

Five Microstructures You Will Meet in a Weld

Martensite deserves special attention. When austenite cools too fast for carbon to diffuse out of the lattice, the structure transforms by a sudden shear into a distorted, carbon-supersaturated BCC variant: hard, unyielding and prone to cracking. TWI identifies this brittle martensite, combined with hydrogen and residual stress, as the classic recipe for underbead cracking. If you remember one mechanism from this article, make it that one.

Dendritic Solidification and the Segregation Problem

Weld metal does not freeze evenly. It nucleates on the fusion boundary and grows outward in tree-like branches called dendrites, primary arms reaching across the pool and side arms filling in between. Alloying elements and impurities like sulfur and phosphorus are less soluble in solid than in liquid, so the freezing front pushes them ahead and they end up trapped in the last liquid to solidify between the arms. That final liquid is enriched, low-melting and brittle. It collects along the weld centerline, which is exactly why solidification cracking loves that region. It is the same principle that makes thermite rail joints so sensitive to confinement and cooling geometry: the last liquid to freeze decides whether the seam is sound or torn.

Grains, Grain Boundaries and the Cheapest Upgrade in Metallurgy

As the metal cools, atoms group into crystalline regions called grains. The borders where grains meet, grain boundaries, are disordered zones that impede slip, block crack propagation and act as gathering places for impurities. More boundary area per unit volume raises strength and improves toughness at the same time. That is rare in metallurgy, where improving one property usually costs you the other, which is why grain refinement is widely called the cheapest upgrade available. In welding, you get it with lower heat input, so the HAZ is reheated into the austenite range and cooled fast enough to form fine rather than coarse grains, and with fillers carrying grain-refining additions like titanium and niobium. Pair that with proper electrode selection and tight heat control, and the base metal behaves predictably.

Note: Coarse grains are not a defect in themselves, but they reliably reduce toughness. A dull, chunky grain structure on a fracture face means the metal lingered too long at high temperature.

Understand these fundamentals, and you are ready for the real battlefield: what happens to the three zones of a weldment when heat pours through them.

The Three Zones of a Weldment: Weld Metal, HAZ and Base Metal

When a weld fails, everyone blames the bead. The truth is more interesting: in a huge share of structural failures, the weld metal itself was fine. The failure started in a narrow strip of steel next to the weld that never even melted. To understand why, you need to look at a cross-section of any welded joint and see it as three distinct metallurgical territories, each with its own microstructure, its own strengths and its own weaknesses.

Zone 1: The Weld Metal (Fusion Zone)

The fusion zone is the part that actually melted and re-solidified. Its chemistry is a mix of the base metal and the filler metal, and its structure is cast metal: coarse columnar grains that grew directionally inward from the fusion boundary toward the center of the bead, following the direction of heat extraction. Because it solidified from a liquid, the weld metal can hold more dissolved alloying elements than the surrounding steel, and it carries the solidification features discussed earlier, including possible centerline segregation and solidification cracking susceptibility.

Paradoxically, the weld metal is often the best-characterized zone. You chose it. When you flip to electrodes or dial in a MIG setup, you are deliberately selecting that chemistry and its matching mechanical properties. Good practice starts with sound consumable selection.

Zone 2: The Heat-Affected Zone (HAZ)

The HAZ is base metal that never reached its melting point but got hot enough for its microstructure to permanently change. Why does the heat-affected zone matter? Because every metallurgical problem in that strip happens invisibly; you cannot see it, and no amount of visual inspection will reveal it.

Depending on peak temperature, the HAZ subdivides into distinct bands, each with its own microstructure:

• Grain-coarsened HAZ (CGHAZ): Sitting just next to the fusion line, this band saw temperatures well above 1100 °C. Austenite grains grew enormous, and the faster the cooling rate, the harder and more brittle the transformation product becomes: coarse martensite or upper bainite. This band typically shows the lowest toughness and highest hardness in the entire joint.
• Grain-refined HAZ (FGHAZ): Further out, where peak temperatures barely exceeded the austenitizing range, the steel fully recrystallized into fine grains. Fine grains mean better toughness and strength. This band is metallurgically the best zone in the joint, and modern thermomechanically rolled steels are engineered to land here.
• Intercritical HAZ: Heated between the A_c1 and A_c3 temperatures, only part of the microstructure transformed to austenite. Rapid cooling can leave brittle high-carbon martensite islands in a soft ferritic matrix: a classic nucleation site for cracking.
• Subcritical (overtempered) HAZ: Heated below A_c1. Grains did not transform, but tempered microstructures softened. In high-strength low-alloy (HSLA) and quenched-and-tempered steels, this soft band can govern the joint’s overall strength.

Zone 3: The Base Metal

Out beyond the HAZ sits undisturbed base metal with its original mill-supplied microstructure and properties. Ideally, your procedure should never degrade it: the goal is to match the weld and HAZ properties to the plate so the joint behaves as one body.

How HAZ Width Tracks Heat Input

HAZ width follows heat input almost linearly. High heat input or slow travel speed delivers a broader thermal pulse; more metal is pushed above the transformation temperatures, producing a wide HAZ with a large, coarse-grained band. Lower heat input shrinks the HAZ and limits grain growth, but push too low and cooling becomes excessively rapid, driving the CGHAZ toward hard martensite. It is a Goldilocks problem: moderate heat input keeps the HAZ narrow without quenching it. Understanding the pipe WPS and preheat rules is a practical way to control this balance on real work.

Multi-pass welding complicates the picture further. Each subsequent pass reheats and tempers the HAZ beneath it, sometimes refining or softening an earlier brittle band, and sometimes creating new intercritical zones. This is the metallurgical logic behind temper-bead welding, covered in a later section.

Remember also that the HAZ is where hazards concentrate beyond metallurgy: heat and spatter exposure near the joint make proper welding PPE and MIG welding safety precautions non-negotiable.

Bottom line: a weld is only as strong as its weakest zone. Learn to read the HAZ, and you have learned to read the joint.

Heat Input, Cooling Rate and the t8/5 Clock: Controlling the Thermal Cycle

Two welders run the same 3/8-inch fillet on the same plate with the same filler. One passes a bend test; the other’s joint cracks open three days later. The arc looked identical to both of them. What differed was the thermal cycle: how much energy went in, and how fast the steel cooled back down. That is the part of welding your eyes cannot catch, and the part metallurgy exists to control.

The Heat Input Formula: Your First Real Dial

Heat input is the energy delivered to the work per unit length of weld. For manual processes, the formula every welder and inspector should know is:

Heat input (kJ/mm) = (Volts × Amps × 60) ÷ (travel speed in mm/min × 1000)

The 60 converts seconds to minutes; the 1000 converts joules to kilojoules. Run 250 amps at 26 volts at 250 mm/min and you land near 1.56 kJ/mm. Drop travel speed to 150 mm/min without touching the machine and you jump to roughly 2.6 kJ/mm, a 65% increase in energy from your settings unchanged. That is why travel speed is the most abused variable in the shop: the weld looks prettier when you go slow, and it is metallurgically worse for it.

Heat input is not a quality score on its own. The same number that is harmless on low-carbon plate can wreck the heat-affected zone of a quenched-and-tempered or microalloyed steel. What matters is chemistry combined with cooling rate, which is why procedures such as pipe welding procedures specify amperage, voltage, travel speed and interpass temperature rather than leaving them to feel.

t8/5: The Clock That Decides Your Microstructure

Metallurgists track cooling with a specific window called t8/5 — the time it takes the weld and heat-affected zone to cool from 800°C to 500°C. This is where austenite decomposes into everything else: ferrite and pearlite on the slow end, bainite in the middle, martensite on the fast end.

• Short t8/5 (a few seconds): rapid quench, hard martensite, high risk of hydrogen-induced cracking.
• Moderate t8/5: bainitic or mixed structures, usually a good balance of toughness and strength.
• Long t8/5 (tens of seconds): coarse ferrite and pearlite, softened HAZ, lower toughness, wider grain growth.

The goal is never “as fast as possible” or “as slow as possible.” It is landing inside the window your procedure was written for.

CCT Diagrams: Turning Guesswork Into a Lookup

Continuous cooling transformation diagrams plot cooling rate horizontally against temperature vertically, with bands showing where each microstructure forms. Give a steel’s CCT chart a cooling time and it tells you what you get: soft and ductile, or hard and brittle. This is the tool that converts welding from folklore into engineering, and it is exactly the kind of variable mapped in ASM’s treatment of metallurgy variables in fusion welding. The same chart explains why thick sections cool faster than thin ones — the plate acts as a heat sink — and why identical settings behave differently at 10°C in winter and 35°C in summer.

Interpass Temperature: The Preheat You Maintain

Interpass temperature is the minimum and maximum temperature of the weldment before the next pass goes down. The minimum keeps cooling slow enough to avoid hard martensite; the maximum stops you from stacking heat and pushing the HAZ into the coarse-grained, over-tempered region. On multi-pass work it is the difference between a controlled thermal cycle and an accidental PWHT. Measure it with an IR thermometer or temperature crayons, not by palm — bare hands reliably misjudge anything above about 60°C.

The Levers You Actually Control

Stringer versus weave deserves its own emphasis because it is the argument welders have in every shop. A wide weave deposits the same metal over a broader area while dwelling longer, so the underlying HAZ grows wider and every later pass re-austenitizes part of the previous one. Stringer beads keep the thermal footprint tight, reduce grain coarsening and give more consistent dilution. Weaving has its place — capping passes, filling a wide joint, out-of-position work — but as a default it trades metallurgy for speed of coverage.

One caveat: the formula assumes a reasonably stable arc. Short-circuit MIG transfers energy in bursts and delivers less net heat than the arithmetic suggests, while spray transfer and stick running a cellulosic electrode run hotter than it implies. If you are setting up a machine from scratch, the baseline values in how to set up a MIG welder are a sound starting point before you begin adjusting travel speed to hit a target kJ/mm.

Treat heat input as a recorded number, not a vibe. Write it down, compare it to the procedure, and the delayed cracks become a lot less mysterious.

Alloying Elements and Carbon Equivalent: How to Read a Mill Test Report

Every roll of steel arrives with a birth certificate. It’s called a Mill Test Report (MTR), and it’s usually a dense printout of element symbols and numbers that most welders skim past in ten seconds. That’s a mistake. Those numbers quietly dictate how hot you preheat, how fast you can travel, which filler you should load, and whether your weldment cracks three days after it passed visual inspection. Welding metallurgy starts, in practice, with learning to read that sheet.

The Elemental Cast: What Each Element Actually Does

Steel is iron with a supporting cast, and every member of that cast changes behavior in the arc:

• Carbon (C) — the primary hardenability driver. More carbon means more martensite on rapid cooling, more hardness, more cracking risk, and lower ductility.
• Manganese (Mn) — a mild strengthener and deoxidizer, and a crucial partner in tying up sulfur. Aim for a healthy Mn:S ratio.
• Silicon (Si) — a deoxidizer that helps produce sound, calm weld metal; too much raises hardness slightly.
• Chromium (Cr), Molybdenum (Mo), Vanadium (V), Niobium (Nb), Titanium (Ti) — hardenability and microalloying elements. In small doses they refine grains and boost high-temperature strength; in larger amounts they push the CCT curve right and make martensite easier to form.
• Nickel (Ni) — improves toughness, especially at low temperature, and is the reason stainless and nickel alloys stay ductile when carbon steels go brittle.
• Sulfur (S) and Phosphorus (P) — the villains. Both lower the melting point of material between grains and promote resolidification (hot) cracking. Sulfur’s damage depends heavily on the manganese-to-sulfur ratio; when Mn:S is low, brittle iron sulfides form along grain boundaries instead of harmless manganese sulfides.
• Copper (Cu) — usually a residual, but above about 0.5% it raises hardenability and can cause hot shortness.

Carbon Equivalent: Turning Chemistry Into a Preheat Number

Because these elements interact, metallurgists compress them into a single index. The most widely used is the IIW carbon equivalent:

CE(IIW) = C + Mn/6 + (Cr + Mo + V)/5 + (Ni + Cu)/15

The rule is simple: higher carbon equivalent means higher hardenability and a greater preheat requirement to avoid hydrogen cracking. A plain A36 structural steel might sit around CE 0.35 and need little or no preheat in thin sections; a quenched-and-tempered pressure vessel steel can exceed CE 0.55 and demand significant preheat plus controlled interpass temperature. The TWI technical knowledge library lays out how these thresholds tie directly to hydrogen crack susceptibility.

Why HSLA Steels Behave Differently

High-strength low-alloy (HSLA) steels get their strength not from carbon but from extremely small additions of vanadium, niobium, and titanium that form carbides and nitrides and pin grain boundaries during rolling. The result is a fine-grained plate that is simultaneously strong and weldable. NIST research made this concrete: low-hydrogen, microalloyed HSLA steels showed markedly lower HAZ cracking susceptibility than high-hardenability HY-100, even though HY-100 is nominally the “stronger” steel. The full NIST IR 5408 report documents these welding metallurgy variables in detail. The lesson is that strength and weldability are not the same axis; chemistry and grain size decide both.

Using CE to Set Preheat in Practice

Here’s the shop-floor workflow:

1. Find the base metal’s CE on the MTR, or calculate it from the element list.
2. Estimate hydrogen level from your consumable (low-hydrogen E7018 handled per E7018 amperage and storage guidance versus cellulosic E6010, which introduces far more hydrogen).
3. Factor in restraint and thickness. Heavy, highly restrained joints crack more easily.
4. Raise preheat as CE, hydrogen, restraint, or thickness increase. AWS D1.1 Table 5.8 is the standard reference mapping CE to minimum preheat temperature for structural steel.

Worked example: a 25 mm (1 in) restrained butt joint in steel with CE 0.48 and low-hydrogen consumables will typically require preheat in the 80–125°C (175–250°F) range, verified with a calibrated contact thermograph or IR thermometer rather than a temperature crayon alone.

Common MTR Mistakes

Three errors show up repeatedly: assuming every “A572” or “S355” plate has the same chemistry when it doesn’t; using a filler that overmatches chemistry and inadvertently raising weld metal CE; and skipping the MTR entirely because the material “welded fine last time.” If your procedure requires traceability under HICC prevention practices, the MTR is your first line of defense.

Understanding chemistry doesn’t just prevent cracking; it tells you exactly which cracks to look for. That’s the subject of the next section.

The Crack Encyclopedia: Hot Cracking, Hydrogen Cracking and Other Metallurgical Failures

Cracks are the one weld defect no code will let you grind out and keep going. Unlike porosity or a little undercut, a crack is a discontinuity with sharp tips that concentrate stress, and it can keep growing every time the part is loaded. What makes them treacherous is that they don’t all look the same, happen at the same time, or share the same cure. Here’s a field guide.

Solidification (Hot) Cracking

Cause: As the weld pool solidifies, low-melting constituents—especially sulfur-rich films—linger at grain boundaries while the shrinking weld is still being pulled by contraction stress. The boundary literally tears open while the metal is still hot. Symptoms: Rough, oxidized fissures along the centerline of the bead or between passes, visible on the crown. Location: Weld metal centerline, craters at the stop, and multi-pass ripple boundaries. Prevention: Keep sulfur low and the manganese-to-sulfur ratio favorable, avoid excessively deep narrow grooves that build a tall centerline seam, use lower heat input to refine grains, and always fill craters. A crater crack is one of the most common beginner mistakes in stick welding practice, and it’s purely technique.

Liquation Cracking

Cause: Similar chemistry, different address. In the HAZ, grain-boundary regions melt at a lower temperature than the grains themselves when sensitizing elements like niobium, titanium, or molybdenum form low-melting eutectics. Symptoms: Short, branching cracks on the HAZ side of the fusion line, often visible only under magnification. Location: HAZ grain boundaries next to the fusion boundary, common in austenitic stainless steels and some nickel alloys. Prevention: Control heat input, avoid stringing passes too slowly along the edge, and control stabilizing-element chemistry.

Hydrogen-Induced (Cold/Delayed) Cracking

This is the killer from our opening story, and it’s the crack that respects no schedule. TWI’s technical guidance and the CWB Group describe the same rule: hydrogen cracking needs four conditions to coincide at once.

• Hydrogen in the steel (damp flux, oily plate, rusty mill scale, or the atmosphere)
• A hard microstructure, typically martensite above roughly 400 HV
• Tensile stress from restraint and contraction
• A susceptible material—higher carbon and alloy content raises risk

Symptoms: Fine, often branched cracks under or beside the bead, sometimes with little surface indication until they open. Location: HAZ, weld root, and toes of heavy passes. Timing: Cracking typically initiates below about 150°C, hours—or even days—after cooling, which is why it hides until the job seems “done.” One 48-hour surprise is all it takes. Prevention: Break the chain by removing any one of the four conditions. Most shops attack hydrogen (dry low-hydrogen consumables, clean joint) and hardness (preheat, correct heat input). Post-weld heat treatment removes both residual stress and diffusible hydrogen, as we’ll cover next.

Reheat Cracking

Cause: Stress-relief heat treatment itself triggers the crack. In creep-resistant steels containing vanadium, niobium, or molybdenum, carbides precipitate in grain interiors during reheating, locking them and forcing strain into the weak grain boundaries. Symptoms: Intergranular cracks in coarse-grained HAZ during or after PWHT. Prevention: Lower heat input to limit grain coarsening, apply temper beads, and follow recommended PWHT temperature bands.

Lamellar Tearing

Cause: Shrinkage strain perpendicular to the plate surface pulls apart flattened, stringer-rich regions. NIST research on weldability ties it to sulfur-rich manganese sulfide inclusions, low through-thickness ductility, and heavy restraint. Symptoms: Stepped, terraced cracks parallel to the plate surface, usually beneath the weld. Location: Thick, restrained T- and corner joints, typically 20 mm and heavier. Prevention: Specify through-thickness-tested (Z-quality) steels, favor butt joints over T-joints where possible, use compressive weld sequences, and keep heat input moderate.

Embrittlement: Slow but Just as Dangerous

Not every failure cracks. Strain aging hardens cold-worked steel on heating; temper embrittlement forms when certain steels creep through 375–575°C during slow cooling; and hydrogen embrittlement of high-strength steels can cut ductility sharply at hydrogen levels too low to form visible cracks. Prevention is metallurgical: pick the right steel grade, control service temperatures, verify hardness limits, and keep proper PPE on hand—see the complete welding PPE guide before grinding and testing begin.

One habit ties them all together: the most metallurgically sophisticated weld in the world fails if the consumable is wrong, the current is wrong, or the technique is wrong. Whether you’re running a simple stick electrode or building a heavy pipe procedure, metallurgy sets the rules and your hands play the game. OSHA’s welding, cutting and brazing guidance makes the same point from the safety side: process control and hazard control are the same discipline seen from two angles.

Filler Metal Metallurgy: Matching vs Overmatching, Low-Hydrogen Consumables and Flux Chemistry

Here’s a question that starts more arguments in welding shops than any other: should the filler metal be stronger than the steel you’re joining, the same strength, or even weaker? The answer shapes HAZ hardness, crack resistance, and the fatigue life of the finished joint. Filler metal metallurgy is where consumable selection stops being a catalog decision and starts being a materials engineering decision.

Matching, Overmatching and Undermatching: Three Philosophies

Matching means the filler’s minimum tensile strength equals the base metal’s minimum, such as an E70XX rod on ASTM A36 structural steel with a 58–80 ksi tensile range. It’s the default for most structural work because the joint properties stay predictable and code reviews are straightforward.

Overmatching steps up a strength class, commonly E80 or E110XX fillers on pipeline and pressure vessel steels. The logic is simple: make the weld the strongest part of the joint so failure gets forced into the base metal where it’s inspectable. But there’s a metallurgical tax. Higher-strength consumables carry more manganese, chromium, and molybdenum, which push the weld and the adjacent HAZ toward harder, more brittle microstructures—especially in the coarse-grained HAZ we covered earlier. As The Fabricator’s coverage of HAZ behavior explains, that’s why pipeline and pressure-vessel procedures pair overmatching fillers with strict preheat and low-hydrogen discipline rather than treating filler strength as free insurance.

Undermatching deliberately uses a softer filler, frequent in aluminum and in multi-pass repairs on high-strength quenched-and-tempered steels. A softer weld metal accommodates strain and keeps the crack path out of a hardened, notch-sensitive HAZ. You accept lower weld strength to protect the more vulnerable base metal region.

What “Low Hydrogen” Actually Means

Low hydrogen isn’t a marketing phrase about the weld—it’s about the coating. Cellulose and other organic binders in electrode coatings break down and release hydrogen into the arc. A basic-flux electrode like E7018 uses sodium and calcium carbonates and fluorspar instead, so the deposited weld metal picks up far less diffusible hydrogen. That’s why E7018 rods must be held in controlled holding ovens, typically around 100–120°C (210–250°F), and why rods pulled from a damp cart are treated as suspect on any load-critical joint.

Flux Chemistry: Basic vs Rutile vs Cellulosic

The flux cover does far more than shield the arc. It deoxidizes the pool, shapes the bead, and determines toughness:

• Basic (low-hydrogen) fluxes — E7018 family. Cleaner weld metal with better toughness and crack resistance, but a stiffer slag and harder out-of-position running. If you’re fighting vertical-ups, our overhead stick welding guide covers how rod choice interacts with position work.
• Rutile (TiO₂) fluxes — E6013 family. Beautiful, fluid arcs and easy all-position control, but more hydrogen pickup and lower as-deposited toughness. Great for sheet and cosmetic passes, risky for restrained heavy joints.
• Cellulosic fluxes — E6010/E6011. Deep, forceful penetration that burns through mill scale and-contaminated fit-ups, popular for root passes on pipe. The cellulose content generates substantial hydrogen, so these rods demand preheat and caution on crack-sensitive steels.

Amperage settings matter too; a 3/32-inch E7018 typically runs around 70–110 amps, and a 5/32-inch rod around 160–220 amps (3/32 E7018 amperage settings; 5/32 E7018 settings). Running too cold isn’t just an aesthetic problem—it increases the risk of lack of fusion, a defect covered in our undercut and weld defects guide.

Shielding Gases and Weld Wire Chemistry

For GMAW, gas choice is metallurgy by another name. Straight CO₂ gives deep penetration but a turbulent, high-spatter arc and higher fume generation. Argon/CO₂ blends (75/25 being the shop standard) stabilize the arc and reduce spatter. Adding oxygen boosts wetting on carbon steel; small helium fractions raise heat input for stainless and aluminum. Crucially, the American Welding Society and OSHA both note that gas composition changes both the rate and chemistry of welding fume, particularly manganese content—more on that in our next section. Welders setting up machines for the first time should walk through How to Set Up a MIG Welder before touching gas ratios.

Stainless and duplex fillers add another layer. Matching 308L to 304 isn’t just strength matching—it’s about keeping ferrite in the 3–10 FN range to prevent hot cracking while preserving corrosion resistance. Duplex 2205 fillers are intentionally overalloyed with nickel (about 8–10%) so the weld’s ferrite/austenite balance lands right after cooling, not the base metal’s exact composition.

Choose filler the way a metallurgist would: read the MTR, know your hardness limits, respect hydrogen, and remember that every strength point you gain in the weld can cost you toughness next door.

Preheat, Interpass Control and Post-Weld Heat Treatment: Fixing the Thermal Cycle

You can’t change the chemistry of the steel on your bench. What you change is its thermal history—how fast it heats, how long it stays warm, how slowly it cools. Preheat, interpass control and post-weld heat treatment (PWHT) are the three levers that shape it. Get them right and a crack-prone joint becomes tough; get them wrong and a weld passes inspection today and splits next week.

Why Preheat Does Two Jobs, Not One

Most welders treat preheat as a comfort feature—warm steel welds nicer, the surface is dry. That’s a side effect. The real work is metallurgical, and it splits into two goals.

• Slow the cooling rate. A slow cool keeps the heat-affected zone out of hard, brittle martensite, leaving tougher ferrite, pearlite or bainite instead. A softer HAZ carries less residual stress and tolerates more strain.
• Give hydrogen a way out. Hydrogen dissolves readily in hot weld metal and is trapped as the joint cools. At temperature it stays mobile enough to diffuse away from the fusion line before the weld freezes. Preheat widens that escape window.

That pairing is why the CWB Group treats preheat as both a cooling-rate control and a hydrogen-removal tool. Hydrogen cracks show up hours or days later, running under the bead where no inspection will find them.

This is codified, not optional. AWS D1.1 and ASME BPVC prescribe minimum preheat based on carbon equivalent, material thickness and consumable hydrogen level. That’s why a pipe welding procedure treats preheat as a documented hold point, not a suggestion.

Interpass Temperature: The Ceiling Nobody Watches

Everyone watches the minimum. Almost nobody watches the maximum, and that’s where the money is lost. Interpass temperature is the joint temperature just before the next bead—and it has a ceiling for good reason. Stay too hot and the cooling rate drops too far: grains coarsen, toughness falls, and stainless or duplex steels risk sensitization and brittle intermetallic phases such as sigma. Too cold and you get the opposite: fast cooling, hard microstructure, hydrogen locked in place.

The window is material-specific. Carbon steels tolerate an interpass in the few-hundred-degree range; duplex stainless demands a much lower ceiling to preserve its phase balance. Check the actual joint with a temp crayon or probe after every pass, not the plate three inches away.

Post-Weld Heat Treatment: Stress Relief and Tempering

PWHT complements preheat with two jobs. It tempers any martensite that formed, trading hardness for toughness, and relaxes residual stress, cutting the risk of environmentally assisted cracking. The TWI job knowledge library puts it the same way: heat treatment changes what the microstructure actually is.

Typical carbon-steel stress relief runs about 1,100–1,250°F, held roughly one hour per inch of thickness, with controlled heat and cool rates and thermocouples on the part rather than the furnace wall. Two cautions:

• PWHT will not rescue a bad weld. It reduces stress; it doesn’t fill lack of fusion.
• On quenched-and-tempered steels, a wrong PWHT temperature can soften the base metal below spec. Check the code and mill cert before you dial in.

Temper-Bead: The Field Alternative When You Can’t PWHT

Sometimes a furnace isn’t practical—field repairs, large structures, in-service work. Temper-bead is the workaround: a controlled second pass deposited over the first tempers the coarse-grained HAZ underneath. Each pass tempers the one below, leaving a fine, tough structure beneath the surface.

The catch: the final pass is never tempered, so it must land where it does the least harm—ideally a low-stress region. Temper-bead demands disciplined bead placement and steady heat input, but in qualified procedures it delivers real toughness gains without a furnace.

Measuring Temperature Without Guesswork

Temperature control is only as good as the instrument behind it:

Measure behind the arc, and remember hot metal stays hot. Handling preheated parts without proper gloves and eye protection is how welders get burned; the welding PPE guide covers the basics for hot work.

Then there’s the classic mistake: skipping preheat on “mild steel” because it welds fine. It often does—at the surface. The same steel, welded cold at heavy thickness with a hydrogen-bearing consumable, can crack two days later with no warning. A 3/32 E7018 running beautifully at 90 amps can still leave a hardened HAZ underneath if the joint never got warm. Preheat is cheap insurance; the crack is not.

Distortion, Residual Stress, Fatigue and Fume Safety: Where Metallurgy Meets the Real World

Throughout the welding process, the visible bead is only half of the battle. The invisible side of your work lies within the metal itself. As we mentioned before, residual stress is a metallurgical legacy: faster cooling and higher restraint mean more stored stress and greater distortion risk. This is where the science of microstructure directly collides with the real world of shop floors, inspection bays, and most importantly, your personal safety.

Residual Stress and Distortion: The Metal Wants to Move

When heat is applied to a joint, the metal expands. As it rapidly cools, it contracts. However, due to the restraint of the surrounding base metal, this contraction is unable to freely occur. This leads to residual stress being locked into the weldment, causing the entire part to pull and twist.

To combat this, experienced welders utilize a few metallurgically sound distortion control methods:

• Tack Welding Strategically: Utilizing proper tacks, like those taught in a MIG tack weld guide, helps hold parts in place while evenly distributing stress throughout the joint.
• Balanced Welding: Welding from the center outward or alternating sides prevents one side from pulling the part out of alignment.
• Controlling Heat Input: By utilizing proper settings for your process, such as the 3/32 E7018 amperage or 5/32 E7018 amperage, you can slow your cooling rate, drastically reducing the amount of stored stress.
• Preplanning Your Sequence: For larger fabrications like those found within pipe welding procedures, a well thought out pass sequence is crucial to prevent warping.

Fatigue Life: The Silent Failure

Residual stress is more than just a distortion issue, it’s a fatigue issue. Cyclic loading, or repeated stress on a part, will cause cracks to initiate at points of high tensile residual stress. These cracks propagate slowly over time, oftentimes failing with no warning.

This is why proper inspection is vital for critical weldments. Nondestructive testing paired with correctly engineered procedures helps combat these failures. Resources from the American Welding Society (AWS) are industry standard when qualifying these joints. No matter if you’re performing overhead stick welding or TIG welding sheet metal, mitigating stress is just as important as your bead appearance.

Fume Safety: The Invisible Threat

While stress affects the metal, welding fumes affect the welder. The extreme temperatures of the arc create a cocktail of metallic oxides and gases, making proper protection non-negotiable. This is exactly why utilizing the correct welding PPE along with shop ventilation is a must for your career longevity.

Stainless steel fumes are far more toxic than carbon steel fumes due to their chromium content. However, both can be hazardous without controls. For this reason, the OSHA Welding Hazards Fact Sheet stresses the importance of local exhaust ventilation, fume extraction, and supplied air when engineering controls cannot reach these limits.

Practical Safety Habits For The Shop

Metallurgy may be the root of these hazards, however prevention lies in your daily habits. Whether you’re learning the ropes or are a veteran, these practices will protect you for the long haul:

• Engineer the Airflow: Utilize fume extraction guns or downdraft tables, especially when welding alloys. This is heavily stressed within MIG welding safety precautions.
• Dress For Defense: Cover all exposed skin. As stated within our welding PPE guide, this prevents not only fumes, but also the weld burns on arms that plague unprotected welders.
• Know Your Consumables: Flux chemistry directly affects fume generation. Be conscious of your electrode selection, such as those found in our breakdown of electrodes in welding, to avoid unnecessary exposure.
• Stay Educated: Continuously reference up to date standards from organizations like NIOSH and OSHA to ensure your shop is always compliant.

The Practical Playbook: A Welder’s Action Plan, Common Mistakes and the Future of Welding Metallurgy

Metallurgy does not care whether you are running a ten-thousand-dollar robotic cell or a buzz box in the driveway. The iron–carbon diagram obeys the same rules either way, which is why everything in this article compresses into a short, repeatable routine you run before, during and after every weld. The science is complicated. The habit is not.

The Six-Step Pre-Weld Checklist

Run this sequence on every job, whether it is a pressure pipe weld or a weekend bracket built in the garage:

1. Read the mill test report. Confirm grade, heat number, chemistry and thickness against the drawing. Never assume the plate in the rack is the plate on the print.
2. Calculate carbon equivalent. Run the IIW formula or Pcm on the actual chemistry. That number tells you how hard the heat-affected zone wants to get and how much preheat you owe it.
3. Set preheat and verify it with a contact probe or calibrated infrared thermometer, not a guess, then log the temperature.
4. Control heat input. Set voltage, wire feed and travel speed from the qualified procedure, and check parameters on the first pass rather than after the cap. A proper MIG setup is where that control starts.
5. Store low-hydrogen consumables dry. Keep rods in a holding oven at the manufacturer’s temperature and follow electrode storage guidance — re-exposure to humid air is the cheapest way to buy a hydrogen crack.
6. Verify after cooldown. Hydrogen cracking is delayed: a weld that looks perfect at 200 °C can still open a day later, exactly like the hydraulic ram in our opening story. TWI’s guide to identifying hydrogen cracks is worth reading before you sign anything off.

Common Metallurgical Mistakes and the Fix

Notice the pattern: every fix costs a little time and a little money up front, while every failure costs demolition, rework and reputation. Preheat and rod storage are cheap insurance compared to cutting out a cracked assembly.

What’s Next: Advanced Steels, Automation and Computational Metallurgy

The steels getting designed today push welding metallurgy into genuinely new territory. Third-generation advanced high-strength steels rely on carefully tuned mixtures of martensite, retained austenite and bainite for their strength, which means the heat of a weld can undo the very heat treatment the mill performed. Quenching and tempering alloys, thermal-mechanical processed plate and nickel-heavy low-temperature steels all demand a written procedure rather than inherited shop folklore.

Automation amplifies both the good and the bad: a robot repeats a perfect parameter set a thousand times, but it will also repeat a flawed one a thousand times. And computational welding metallurgy is changing the front end of design. Thermal and metallurgical models now let a shop predict HAZ peak temperature, cooling time and resulting microstructure before a single bead is laid, so procedures can be qualified virtually and trial welds cut from months to days.

Career Paths: Where This Knowledge Takes You

Shops moving toward advanced steels and automation need people who can read both the arc and the paperwork. The AWS Certified Welding Inspector credential remains the most recognizable doorway, supported by NDT Level II certification, WPS/PQR authoring and metallurgy coursework. On the other side sit materials engineers and welding metallurgists who run failure analysis, write procedures and interpret microstructures under a microscope.

That is the whole philosophy in one line: read the certificate, do the arithmetic, control the cooling, and verify after the metal goes cold. The weld you cannot see is the weld you should be thinking about hardest.

Frequently Asked Questions

What is welding metallurgy in simple terms?

Welding metallurgy is the study of how heat and chemistry change a metal’s internal microstructure during welding. It explains why the weld, the heat-affected zone and the base metal all end up with different properties even though they look like one continuous piece.

Why does the heat-affected zone matter?

The HAZ never melts, but its microstructure is rewritten by the thermal cycle, often making it the hardest and least tough region of the joint. Most cracking failures initiate in the coarse-grained HAZ, so controlling HAZ properties is central to weld integrity.

What is carbon equivalent and why do welders calculate it?

Carbon equivalent (CE) is a formula that combines carbon with other alloying elements to estimate hardenability. The common IIW formula is CE = C + Mn/6 + (Cr+Mo+V)/5 + (Ni+Cu)/15, and higher values signal a greater need for preheat to prevent hydrogen cracking.

How long after welding can hydrogen cracking appear?

Hydrogen-induced cracking is delayed and can show up hours or even days after welding, which is why it is sometimes called cold or delayed cracking. That delay is why destructive or NDT verification matters before a joint goes into service.

Does heat input really affect weld quality?

Yes. Heat input controls the cooling rate, which controls which microstructure forms in the HAZ and weld metal. Too much heat widens the HAZ and softens it; too little can produce hard, crack-prone martensite in susceptible steels.

What is the difference between hot cracking and cold cracking?

Hot cracking happens during or immediately after solidification, usually along dendrite boundaries enriched in segregated impurities. Cold (hydrogen) cracking happens later, driven by hydrogen, hard microstructure, and tensile stress acting together.

How does welding metallurgy affect welder safety?

The same arc chemistry that forms your microstructure also generates fume containing manganese, hexavalent chromium and other metals. OSHA caps Cr(VI) at 5 µg/m³ over eight hours and NIOSH recommends 1 mg/m³ for manganese, so ventilation and respiratory protection are part of metallurgical practice.

What is the best way to learn welding metallurgy?

Start with authoritative references like AWS, ASM International and TWI technical knowledge, pair them with your procedure qualification records, and reinforce it with hands-on macrosection and hardness testing. Certification paths such as AWS CWI also formalize the knowledge.