Welding Hardened Steel: Preheat, Filler & PWHT Guide

A shop took on a quick repair: a cracked 4140 hydraulic ram, already hardened to about 30 HRC, got a fast root pass with no preheat. The weld looked perfect on Friday and split along the bead by Monday morning, turning a two-hour fix into a scrapped shaft. That gap between ‘looks fine’ and ‘fails later’ is exactly why welding hardened steel deserves a real procedure instead of a guess.

Can You Weld Hardened Steel? The Honest Answer Up Front

That scrapped 4140 ram is the perfect place to start, because it raises the question shops ask constantly: can you actually weld hardened steel? The honest answer is yes. Hardened steel gets welded every day in repair shops, mines, mills, and shipyards. But it can only be welded with deliberate procedure control. There is no “quick bead” version of this job that survives the weekend. The difference between a sound repair and a crack that shows up Monday is never how the weld looks; it is what the heat cycle did to the steel underneath it.

What “Hardened Steel” Actually Means

Hardened steel is steel that has been austenitized (heated until its structure turns to austenite) and then quenched rapidly, usually in oil or water, so the structure transforms into martensite. Most of it is then tempered by reheating to a lower temperature to trade some hardness for toughness. The delivered condition is therefore a quenched-and-tempered structure: high hardness, high tensile strength, martensitic or bainitic grains, low ductility, and high hardenability, meaning the alloy forms hard structures even at moderate cooling rates.

Common examples include 4140 and 4150 chrome-moly shafting, boron steels, tool steels, chrome-moly tubing, high-carbon rail steels, and abrasion-resistant wear plate. The precise condition matters enormously: a part supplied annealed behaves nothing like the same part supplied at 30 HRC, even though the alloy designation on the paperwork is identical.

Why Welding It Is a Different Problem

Welding is a localized melting and rapid-heating process, and hardened steel does not tolerate that casually. When the arc passes, the heat-affected zone (HAZ) right next to the bead is heated above its critical temperature, which means it re-austenitizes. On cooling, that zone reforms fresh, untempered martensite, which is brittle and full of internal strain. At the same time, the shrinkage of the cooling weld metal locks in tensile residual stress, and hydrogen can be present from moisture, coatings, rust, or the consumable itself. Those are the three enemies of every hardened-steel weld:

• Hard martensite in the HAZ gives the crack a brittle path to travel through.
• Tensile residual stress from weld shrinkage and restraint provides the pulling force.
• Hydrogen migrates into the stressed zone and makes the metal lose ductility at the worst possible moment.

Industry guidance frames this as a three-legged stool: a susceptible (hard) microstructure, hydrogen, and tensile stress must all be present for hydrogen-induced cracking to occur. TWI’s technical guidance on hydrogen cracks in steels makes the practical point clearly: remove any one of the three, and cracking stops. That single idea is the foundation of everything in this article, from preheat to filler selection to post-weld heat treatment.

Who This Guide Is For

This guide is written for the people who actually get handed these jobs: fabricators building or repairing alloy assemblies, maintenance techs fixing a broken pin or ram in place, repair shops taking in unknown-condition parts, and experienced DIYers working on motorcycle frames, axle shafts, or drill-press components at home. If you have ever been told “just weld it, it’s only a small crack,” this is for you. For the consumable side of the job, our guide to alloy steel welding electrodes covers rod selection, storage, and handling in detail.

What You Will Get From This Article

If you are searching “welding hardened steel,” your question is almost always practical rather than theoretical: how do I weld this without cracking it? So the rest of this article follows a working sequence you can take to the shop:

1. Understand what the heat cycle does to the steel, starting with the heat-affected zone.
2. Recognize delayed hydrogen cracking as the central danger.
3. Identify the steel, calculate carbon equivalent, and pick filler metal and process.
4. Set preheat and interpass temperature from real numbers, not feel.
5. Control heat input, bead shape, and pass sequence at the arc.
6. Post-weld heat treat, slow cool, and inspect after the delay window.

Each step builds on the one before it, because hardened-steel welding fails at the weakest link rather than on average. Skipping straight to temperature charts without understanding hydrogen is how people end up with procedures they cannot explain or defend.

When Welding Is the Wrong Answer

One thing up front, because good judgment beats good technique: welding is not always the right repair. Machining a replacement from annealed bar, then heat treating it, is usually cheaper and safer for small, cheap, or safety-critical parts. Brazing or braze welding puts far less heat into the base metal and is often the smarter choice for thin sections, cast iron, or parts where distortion cannot be tolerated. And for a case-hardened gear tooth, a worn bushing, or anything where the hardened skin itself is the working surface, weld metal will not restore the original surface properties, so replacement wins. Hot-work, fume, and fire controls also apply to this work; OSHA 1910.252 sets the baseline requirements for welding, cutting, and brazing in the workplace.

With that boundary set, the next section looks at what hardening does to steel at the microstructural level, and why the same alloy can behave completely differently depending on the condition it arrived in.

What Makes Steel “Hardened” — and Why That Changes the Weld

Before jumping into the arc, it’s important to understand that not all steels react the same way to heat. The steel that you’re welding could be soft and ductile one moment, but the rapid heating and cooling from your weld can transform it into something much harder and far more brittle. This is especially true for hardened steels, and it’s the root of most cracking issues that shops run into.

Hardenability and Martensite Formation

The term “hardened steel” refers to steel that has been heat treated to form a martensitic structure. Martensite is an extremely hard, needle-like microstructure that forms when austenite is cooled at a fast enough rate to trap carbon atoms within the iron’s crystal lattice. The ability for a steel to form martensite at a given cooling rate is known as its hardenability.

Not every steel possesses the same hardenability. Plain carbon steels require very rapid quenching to harden, while alloy steels can harden at much slower cooling rates. This is due to the alloying elements, such as chromium and molybdenum, which are found within grades like AISI 4140 and 4150. These chromium-molybdenum steels are through-hardening, meaning their hardenability allows fresh martensite to form throughout the cross section during rapid cooling. This same trait is what makes their coarse-grained heat-affected zone (HAZ) so susceptible to cracking while welding.

Quench and Temper Cycles

The majority of hardened steels that are repaired in the shop are in the quenched and tempered (Q&T) condition. This is a two step process that allows for a balance between strength and toughness. The process consists of:

• Austenitizing: Heating the steel to its critical temperature to form austenite.
• Quenching: Rapidly cooling the part, typically in oil or water, to transform the austenite into martensite.
• Tempering: Reheating the martensitic steel to a lower temperature, then cooling it slowly. This relieves stress and trades a small amount of hardness for a large boost in toughness.

It is this final tempering step that creates a usable part. However, the welding arc completely bypasses this controlled cycle, creating localized zones that can reach austenitizing temperatures and then self-quench from the surrounding cold base metal. This is the main reason that procedure control is needed for any alloy steel welding electrodes application.

Delivered Condition Dictates Your Approach

When welding hardened steel, the “as received” condition of your base metal will dictate your entire game plan. Each condition reacts differently to the heat input from welding:

Delivered Condition	Properties	Welding Considerations
Annealed	Softest, highest ductility, and lowest hardness.	Easiest to weld with the least risk of HAZ hardening.
Normalized	Stronger than annealed with a finer grain structure.	Weldable, however, hardenability must still be considered.
Pre-Hardened	Heat treated to a specific hardness prior to machining.	Very prone to over-hardening in the HAZ. Requires strict heat control.
Quenched and Tempered (Q&T)	High strength with moderate toughness.	High cracking risk without preheat, low heat input, and possible PWHT.
Case-Hardened (Carburized)	Extremely hard, carbon-rich surface with a softer, ductile core.	Never weld directly through the skin. Only weld the core, if at all.

Hardness vs. Toughness and Unpredictable Zones

When welding these materials, you will always be forced to juggle a trade-off: hardness and toughness are inversely related. The heat from welding creates a gradient of temperatures throughout the base metal. This causes some zones to over-harden into fresh martensite, whilst other zones adjacent to them may be softened from the temper being drawn out of the original Q&T structure. These unpredictable shifts create localized stress concentrations, priming the part for failure.

A commonly used guideline for weldability is the carbon equivalent (CE). As stated by industry guidance, a CE below about 0.45 generally indicates steels that are readily weldable with appropriate procedures, whereas a CE above about 0.45 indicates a much greater need for caution regarding preheat and heat input. This, paired with hardenability testing such as the Jominy end-quench test, allows for the prediction of how deep hardening will occur for grade families including 4140/4150, tool steels, chrome-moly, boron steels, high-carbon rail, and wear plate.

The Heat-Affected Zone: Where Hardened Steel Welds Actually Fail

Ask a shop where a hardened steel weld cracked and most people point at the bead. They’re pointing in the wrong place. In the overwhelming majority of hydrogen-induced failures, the weld metal itself is fine; the crack runs through the heat-affected zone (HAZ) — the strip of base metal beside the fusion boundary that never melted, but was cooked hard enough to change its entire microstructure. To weld hardened steel successfully, the HAZ is the territory you actually have to win.

From White Heat to Untouched Metal in a Few Millimetres

Every arc creates a steep temperature gradient. Right at the fusion boundary, base metal briefly sits at or above the melting point. Move outward a fraction of a millimetre and peak temperatures drop to 1,100°C, then 900°C, then 700°C, until the metal is essentially unchanged. That gradient sorts the HAZ into recognizable sub-zones, each with a different structure and a different risk profile:

• Grain-coarsened HAZ (CGHAZ). Peak temperatures above roughly 1,100°C dissolve the original structure and let prior austenite grains grow large. On cooling, those coarse grains transform into hard, brittle martensite in a geometry that resists cracking relief. This is, by a wide margin, the most common crack-initiation region.
• Grain-refined HAZ (FGHAZ). Peaks just above the transformation range (around 900–1,000°C) fully re-austenitize the metal and then refine the grains on cooling. Finer grains mean better toughness, so this band is comparatively benign.
• Intercritical or subcritical HAZ. Lower peaks partially transform or only temper the existing structure. Depending on the original condition, this is where localized softening — the “soft zone” — can form and later fail in service by over-matching the surrounding hard material.

The grain-coarsened region is the danger zone because it combines three unfavorable conditions: coarse grains, a fully hardenable chemistry, and constraint from the surrounding cold metal. TWI’s guidance on hydrogen cracking in steels describes exactly this pattern: cracks initiate in the CGHAZ and propagate either as underbead cracks hugging the fusion boundary, or as toe cracks driven by the tensile residual stress that concentrates where the weld meets the plate.

t8/5: The 800–500°C Clock That Decides Your Microstructure

The key variable isn’t arc energy alone — it’s how fast the HAZ cools through the critical band between about 800°C and 500°C, a parameter metallurgists call t8/5 (time from 800 to 500°C). Austenite transforms into something else inside this window, and which something else depends entirely on the clock:

Cooling behaviour	Dominant transformation product	Result in hardened steel
Very fast t8/5 (thin section, cold part, high travel speed)	Martensite, possibly bainite	Hard, brittle HAZ; high hydrogen crack risk
Moderate t8/5 (adequate preheat, controlled heat input)	Bainite with limited martensite	Manageable hardness, acceptable toughness
Slow t8/5 (generous preheat, higher heat input, warm interpass)	Ferrite and pearlite	Soft HAZ; lower crack risk, some strength loss

This is the whole rationale behind preheat: not “warming the part up so it feels comfortable,” but deliberately stretching t8/5 so soft ferrite and pearlite form instead of martensite. In alloy and hardenable steels this is even more critical than in plain carbon steel, because alloying elements push the transformation curve rightward and let martensite form at slower cooling rates.

How AWS D1.1 Puts Numbers on HAZ Hardness

AWS D1.1 Annex XI formalizes what experienced welders do intuitively: control HAZ hardness by controlling cooling rate. The annex provides a method for predicting HAZ hardness from carbon equivalent and cooling time, on the premise that if hardness stays below a critical threshold — roughly 350 HV for most structural situations, adjusted for hydrogen level and restraint — hydrogen cracking will not occur. Rather than blindly maximizing preheat, you calculate the minimum preheat that keeps predicted hardness under that threshold: enough heat to avoid martensite, but not so much that you overtemper the base metal or coarsen the HAZ further. Lincoln Electric’s preheat overview walks through the same logic in more practical shop language.

Interpass temperature is the other half of that control loop. On a multi-pass weld, each pass re-heats and re-cools the HAZ beneath it. Holding interpass temperature within the specified range keeps t8/5 consistent across every pass, so the root-adjacent passes aren’t quenched cold while the cap is running on metal that has drifted 150°C above spec. On hardened steel, monitoring interpass temperature isn’t optional paperwork — it’s the mechanism that makes the preheat calculation mean anything. If you’re choosing consumables for these passes, our guide to alloy steel welding electrodes covers which classifications hold up under controlled cooling.

None of this is exotic metallurgy. It’s the practical consequence of recognizing that the arc does something to the steel next to it, that what it does depends on a cooling clock you can measure and control, and that the failures start there — not in the weld you spent all your time watching.

Hydrogen-Induced Cracking: The Delayed Threat You Can’t See

Here is the part of welding hardened steel that catches people out: a weld can look absolutely flawless and still be doomed. Hydrogen-induced cracking, also called cold cracking, delayed cracking, or underbead cracking, happens after the arc is long extinguished, often after the part has been loaded back into service and the repair has already been signed off. Understanding why it happens is the single most important thing you can do before striking an arc on a quenched-and-tempered steel.

The Three Factors That Have to Converge

Hydrogen cracking is rarely caused by one thing. It needs three conditions present at the same time, in the same place, and when all three overlap, cracks form:

• A susceptible microstructure. Hard, brittle martensite in the heat-affected zone is the classic victim. The harder the HAZ, the less strain it can absorb and the more likely it is to crack. This is why alloy steel welding electrodes and correct cooling-rate control matter so much on hardenable grades.
• Hydrogen in the weld. Even a few parts per million of diffusible hydrogen dissolved in the metal can be enough to trigger a crack in a restrained, hardened joint.
• Residual (shrinking) stress. Welding shrinkage in a thick, restrained joint generates tensile stress comparable to the metal’s yield strength. That stress gives the hydrogen a reason to collect and a mechanism to open a crack.

Remove any one of the three and the risk drops dramatically. That is the entire logic behind every hardened-steel welding procedure: dry out the hydrogen, control the cooling rate so the HAZ stays reasonably tough, and manage stress through bead sequencing, preheat, and post-weld heat treatment.

Where the Hydrogen Comes From

Hydrogen is everywhere in a welding environment, and it enters the weld pool as atomic hydrogen, which is small enough to dissolve straight into liquid metal:

• Moisture in electrode coatings and fluxes, the single biggest culprit with stick and flux-cored consumables. This is exactly why low-hydrogen electrodes like the E7018 class must be stored in heated ovens and rebaked per the manufacturer’s schedule.
• Rust, mill scale, condensation, and oil or grease on the joint itself.
• Humid shop or outdoor air, which hydrates fluxes and coatings during welding.
• Cellulose-coated electrodes, such as 6010/6011 types, which generate significant hydrogen from their organic coating. They are useful tools, but a poor pairing with untempered martensite.
• Paints, primers, galvanizing, cadmium plating, and other coatings, which can release hydrogen, and toxic fumes, when burned by the arc.

Once hydrogen dissolves into the molten bead, it diffuses outward as the metal cools. Some escapes at the surface, some migrates into the HAZ, and some gets trapped at microstructural defects like inclusions, dislocations, and grain boundaries. Trapped hydrogen creates internal pressure at exactly the points where hard martensite and tensile stress are already concentrated.

The Delay Window: Why Friday’s Weld Fails Monday

Hydrogen-induced cracking typically occurs below about 600°F (roughly 315°C), and often at 300°F (about 150°C) or lower as the joint continues to cool to room temperature. Timing is the signature: most cracks appear within 48 hours of welding, though in some cases they show up as much as a week later, according to guidance from TWI.

That delay has two practical consequences. First, a visual inspection performed immediately after welding tells you almost nothing. A crack can nucleate under the bead, out of sight, and stay dormant while hydrogen slowly migrates to a stress concentration. Second, parts must be held and re-inspected after the delay window before they are released to service. An underbead crack that reaches service is a rejectable defect that can turn into catastrophic failure of a shaft, ram, or structural member.

The Practical Implication: Control Three Things

Because you cannot see hydrogen or measure residual stress by eye, the defense has to be built into the procedure:

• Use dry low-hydrogen consumables. Stick to H-class or EX-H-class low-hydrogen processes and consumables, store electrodes properly, and keep backup rods in a holding oven.
• Control the heat. Preheat does double duty: as Lincoln Electric explains, it dries the weld area to drive off moisture, removing a hydrogen source at the source, and slows the cooling rate, giving dissolved hydrogen time to diffuse out of the joint before the metal hardens around it.
• Hold the part under observation. Do not release a repaired hardened part the same shift. Keep preheat and PWHT within OSHA 1910.252 best practices in mind, and schedule NDT after the crack window closes.

When preheat and post-weld heat treatment are done correctly, hydrogen diffuses harmlessly out of the joint and hardened-steel welds survive for decades. When they are skipped, the metal essentially cracks itself from the inside, with no warning and no visible sign until it is too late.

Before the Arc: Carbon Equivalent, Filler Metal, and Consumable Prep

Every successful weld on hardened steel starts before the hood drops. The shops that win this battle run the same short decision workflow every time: identify the steel and its delivered condition, calculate carbon equivalent, decide whether the service actually requires matching strength, and only then pick the filler and process. Skip a step and you are guessing with a 3,000-degree guess.

Step 1: Know Exactly What You Have

“Probably 4140” is not a material specification. A part number, mill certificate, or hardness test is. If markings are gone, a portable hardness tester on a flat spot, a spark test, or positive material identification will tell you far more than the shape of the component ever will. Just as important is the delivered condition: annealed, normalized, pre-hardened, quenched and tempered, or case-hardened. A normalized 4140 bar welds almost like mild steel; the same bar at 30 HRC in the quenched-and-tempered condition demands full procedure control. The alloy is identical, the weldability is not.

Step 2: Calculate Carbon Equivalent

Carbon equivalent (CE) condenses the hardenability contribution of every alloying element into one number that tells you how readily the steel will form brittle martensite on cooling. The international IIW formula, also used as the basis for AWS D1.1 preheat tables, is:

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

As a working rule, steels with a CE above about 0.5 generally need preheat, low-hydrogen processes, post-weld heating and slow cooling to weld reliably, as summarized in TWI’s guidance on hydrogen cracks in steels. Hardened 4140 and 4150 typically land well past that line, which is why they crack so readily when welded cold. For carbon and carbon-manganese steels, the European standard EN 1011-2:2001 goes further, providing preheat nomograms where required temperature can reach 200°C (392°F) for thick, high-CE sections. Higher CE plus thicker, more restrained joints equals more heat, more hydrogen control, and more patience.

Step 3: Does the Service Demand Matching Strength?

This question decides your filler philosophy, and the answer surprises people: on a restrained, crack-prone joint, undermatching is often the smarter play. A softer, more ductile weld metal (an E7018-class deposit on 100 ksi-quenched 4140, for example) stretches and yields first, giving hydrogen-induced cracks a ductile path to blunt rather than a rigid one to run through. Overmatching is reserved for cases where the joint must carry the full base-metal load, and even then it invites cracking. Where the core is hard or metallurgically awkward, alloy steel welding electrode selection guides cover the matching, undermatching and buttering choices in detail.

Buttering is the other escape hatch: deposit a layer of ductile, low-hydrogen weld metal on the hard base first, let it temper the surface, then weld the actual joint into that butter layer. It works for dissimilar joints, hard-to-weld cores, and surface-critical repairs.

Step 4: Pick a Low-Hydrogen Process

Hydrogen is the enemy of hardened steel, so process choice is really hydrogen-choice. SMAW with basic low-hydrogen electrodes remains the field standard, while FCAW-G, GMAW with dry, clean gas, and GTAW for critical work all fit when the shielding and consumables are controlled. The chart below compares the usual suspects:

Process / Consumable	Relative Hydrogen Risk	Key Control
SMAW E7018-class (basic)	Low	Rod oven, sealed storage, limited open time
SMAW cellulosic (E6010/E6013)	High	Avoid on restrained hardenable steel
FCAW-G (gas-shielded)	Low–moderate	Dry gas, correct voltage, clean wire
GMAW	Low	Moisture-free liners, dry shielding gas
GTAW (bare filler)	Very low	Clean, dedicated filler and tungsten

Do that homework, confirm your consumable handling matches established hydrogen-control practice, and the arc you strike will already be halfway to a sound, crack-free weld. The remaining half is temperature, which is where preheat comes in.

Preheat and Interpass Temperature: Getting the Numbers Right

Preheat is the single most reliable lever you have when welding hardened steel, and it works for a simple reason: heat slows the cooling rate. A slow-cooling HAZ has time for carbon to diffuse out of the austenite gradually, so instead of snapping straight to hard, brittle martensite it passes through softer products like bainite and pearlite. Preheat also shrinks the temperature gradient across the joint, which lowers residual stress, and it gives dissolved hydrogen an easier path to escape before it can do damage. As Lincolnelectric’s preheat overview puts it, preheat controls the cooling rate and hydrogen escape together, which is exactly what crack-sensitive steel needs.

How to Select the Temperature

Nobody picks a preheat number by feel. It comes from three inputs:

• Carbon equivalent (CE). Higher CE means greater hardenability and more crack risk, so more preheat. The IIW formula (CE = C + Mn/6 + (Cr+Mo+V)/5 + (Ni+Cu)/16) and the AWS D1.1 D1.1-type formula are the common calculators.
• Thickness. Thicker sections act as a bigger heat sink and restrain the weld more, pushing temperature up.
• Restraint and joint geometry. A tight, fully restrained corner joint cracks far more easily than a lightly restrained butt weld on plate.

For many hardenable alloy steels, typical preheat falls in the range of 150–300°C (300–570°F). 4140 is commonly quoted anywhere from about 400–750°F, depending on delivered condition, thickness, and filler selection. A normalized or annealed 4140 shaft might only need 200–300°F; the same part already hardened to 30 HRC, welded with a matching filler into a restrained joint, can justify the upper end. Associated Steel’s 4140 welding guidelines are a useful real-world reference for how condition changes the recommendation.

When you need consumable guidance alongside the temperature, our alloy steel welding electrodes guide covers which low-hydrogen classifications pair with these preheat bands.

Where and How to Measure

Guessing with a gloved hand is not measurement. Use calibrated thermocouples with a readout, a surface probe, or temperature-indicating crayons rated for the target range, and replace crayons once they’ve been through a heat cycle.

AWS D1.1 requires the minimum preheat to be established at a distance at least equal to the thickness of the thickest member, and not less than 3 in (75 mm) in all directions from the point of welding. Preheat should be verified by measuring the surface temperature adjacent to the joint, and heating the side opposite to the one being welded is recommended practice, so the full material volume around the joint is heated rather than just a hot skin over a cold core.

Also allow a soak time. Large parts need a few minutes at temperature to equalize, not a quick pass of the torch. Then hold the interpass temperature — the temperature before each subsequent pass — at or slightly below the initial preheat, never above it. Multi-pass welds behave like repeated mini heat cycles; if interpass creeps upward pass after pass, you are effectively running a PWHT cycle you never intended.

Keeping the Band Stable

Treat preheat and interpass as a control band, not a starting line. Watch the interpass with the same thermometer you used for preheat, limit time between passes so the part doesn’t drift down, and if it does drop below the minimum, stop and re-verify before striking the arc again. ESAB’s preheat and PWHT article emphasizes that both the minimum and maximum matter, because exceeding either one quietly degrades the joint. Per OSHA 1910.252, any hot-work operation around these heated assemblies also needs proper fire-prevention controls in place.

Get this one right and the filler choice, technique, and post-weld steps that follow all work with you instead of against you.

Welding Techniques That Work: Heat Input, Bead Shape, and Buttering

Preheat and dry consumables get you to the starting line; technique is what finishes the job. Every habit in this section serves the same two goals: keep the heat-affected zone from cooling into brittle, untempered martensite, and keep hydrogen out. None of it is exotic, but all of it has to be deliberate.

Heat Input: Low Enough to Control, High Enough to Avoid a Quench

Heat input on hardened steel is a balancing act, and both extremes fail. Long arcs, slow travel and wide weaves push too much energy into the joint, coarsening grains in the coarse-grained HAZ and widening the hard, brittle band along the fusion line. Drag a cold bead across preheated-but-cooling steel, though, and you act like a fresh quench: the weld and the adjacent HAZ spike above the transformation temperature and then plunge through it, recreating exactly the martensite the base metal was tempered to avoid.

The practical answer is consistent, moderate input. Hold a short arc length, keep your polarity and voltage settings as the procedure specifies, and maintain steady travel speed rather than surging and stalling. As a rule, small-diameter electrodes, rapid travel, and several small stringer beads are recommended to minimize welding stresses in hardenable steel repair work. Smaller electrodes mean less amperage per pass, which means a narrower, shallower HAZ.

Stringer Beads Instead of Wide Weaves

A weave looks faster and, in a wide face, it ties the toes together nicely. On hardened steel it also dumps heat in one spot and holds it there. Keep the weave width to roughly two to two-and-a-half electrode diameters, and favor stacked stringers. This pays a metallurgical dividend: multi-pass techniques reduce hardness because each subsequent pass re-heats and tempers the previous pass and the underlying HAZ, softening the coarse-grained region. Each bead you lay makes the one beneath it tougher.

Watch the crater at every stop. Fill it, or back up slightly before breaking the arc; crater cracks are a common starting point for a crack that later runs under the bead.

Pass Sequence, Restraint, and Distortion

How you stack passes matters as much as how many. Alternate sides of a double-jointed weld, use back-step or skip sequencing on long runs, and balance the weld metal so shrinkage forces counteract each other instead of accumulating. Deep, narrow grooves magnify restraint, so open the joint angle slightly or add a root pass that can yield. Never accelerate cooling with compressed air, an air line, or a wet rag — the part must walk down in temperature with the preheat still doing its job.

Temper Bead Welding When PWHT Is Off-Limits

When a part can’t go into a furnace for post-weld heat treatment, temper bead (temperbead) welding is the engineered substitute. Beads are placed and sized so that each one tempers the HAZ beneath it, producing a graded, softened structure without heating the whole component. It is recognized as a qualified alternative to post-weld heat treatment; EPRI has published a guideline for developing, qualifying, and implementing temperbead techniques, and ASME Section IX has been revised to support temperbead procedure qualification. You can review the EPRI temperbead guideline directly. The catch: temper bead is a procedure, not a trick. Bead width, overlap, travel speed and layer sequence must be qualified by testing before the first production weld.

Buttering: A Ductile Layer First

Buttering means depositing one or more weld beads onto the hard or metallurgically awkward side of the joint first, then welding the joint into that layer. It’s the standard move for high-carbon cores, tool steels, and dissimilar joints, because it lets you choose a soft, ductile, low-hydrogen buttering layer — and it gives the final weld something forgiving to shrink against. After buttering, the layer can be tempered or machined before the structural pass goes down.

Peening: Useful, but Easy to Misuse

Light, uniform peening of intermediate passes can relieve surface tension and reduce shrinkage stress. But peening a root pass drives strain into the most highly restrained area of the joint, and peening over a hairline HAZ crack simply buries it. Peen only intermediate layers, stop before the cap, and never peen a weld you suspect of cracking — inspect instead.

Process Selection and the Hardfacing Distinction

GTAW gives the tightest heat control and the lowest hydrogen pickup, making it the pick for small, critical, close-tolerance repairs. SMAW with a properly baked low-hydrogen electrode remains the shop-and-field workhorse — see why stick welding still wins in a lot of real jobs. Gas-shielded FCAW works well on thicker sections if the shielding gas is dry and the wire is stored properly.

Finally, don’t confuse structural welding with hardfacing. Hardfacing lays down a wear-resistant, often chromium- or carbon-rich layer on a working surface; it is not part of the load path and its filler chemistry is chosen for abrasion resistance, not joint toughness. Using a hardfacing rod to “strengthen” a structural repair is a category error that ends in a cracked weld.

Post-Weld Heat Treatment, Slow Cooling, and What to Do When PWHT Is Off-Limits

The arc is off, the bead is dressed, and it looks clean. For hardened steel that is not the finish line; it is the start of a second job. PWHT turns a weld that looks correct into one that stays intact, and skipping it is how “perfect” Friday repairs come apart on Monday.

What PWHT Actually Does for a Hardened Steel Weld

PWHT carries out four jobs at once:

• Drives out diffusible hydrogen. Heat gives trapped hydrogen the energy to diffuse to the surface and escape. TWI’s guidance on hydrogen cracking in steels notes that hydrogen keeps migrating through steel for a day or more after welding, which is why delayed cracking appears later. TWI’s hydrogen crack prevention guidance
• Tempers the martensite. The coarse-grained HAZ forms as-hardened martensite. Tempering trades a little hardness for a large gain in toughness and ductility.
• Relieves residual stress. Holding the weld sub-critically lets peak stresses relax instead of summing with service loads.
• Improves overall toughness of weld metal, HAZ and base metal together, for a more uniform structure across the joint.

Timing: Do Not Let the Preheat Drop

For many steels, PWHT may be applied immediately on completion of welding without allowing the preheat temperature to fall, and doing exactly that minimises hydrogen trapping. If the part cools to ambient first, hydrogen is caught in a structure that is now hard and unyielding — the window for cheap, safe removal has closed.

Temperatures, Hold Times, and the One-Hour-Per-Inch Rule

Stress relief and tempering are sub-critical treatments — they stay below the lower transformation temperature. For medium-carbon alloy steels such as 4140 and 4150, typical ranges sit around 595–650°C (1100–1200°F), adjusted to land the finished hardness where the part needs it. Associated Steel’s 4140 welding guidelines give specific temper bands for their grades — a good example of why numbers should come from the material’s own data sheet, not from memory.

The common shop rule of thumb is roughly one hour per inch of thickness, with an absolute minimum of one hour for small parts.

Normalizing: Refining What the Arc Coarsened

When the HAZ is badly coarse or a hard zone cannot be tempered away, normalizing refines coarse weld structures, reduces residual stresses and removes hard zones in the heat-affected zone. Heating above the upper transformation temperature and air cooling produces a fine, uniform ferrite-pearite structure — but it leaves the steel soft and untempered. Highly hardenable steels may need a second treatment after PWHT to temper the hard microstructure produced by the first, so a normalize is normally followed by a controlled re-temper to bring hardness and toughness back into spec.

How You Cool Down Matters Too

Furnace cooling, or at minimum a heavy ceramic blanket, is preferred for thick, hardenable sections. Never quench a finished weld with compressed air, a fan or a cold draft; you are re-creating the original problem. Ambient cooling is acceptable only for small, low-hardenability parts where the cooling rate through the critical range is naturally slow, and even then it should be a deliberate decision, not an accident.

When PWHT Is Off-Limits

Plenty of real repairs cannot go in a furnace: an in-service pump housing, a 300 kg shaft that will not come apart, field work three hours from the shop. The strategy then shifts from fixing the microstructure after the fact to never letting it get bad:

• Higher, verified preheat to slow cooling through the critical range, with interpass control.
• Strictly low-hydrogen consumables, re-baked and protected — our guide to alloy steel welding electrodes covers storage and rebake practice.
• Buttering: laying a ductile weld-metal layer on the hard face first, then joining to that soft layer rather than directly to the hard core.
• Temper bead (temperbead) welding, a controlled technique where each pass’s heat input is calculated to temper the previously deposited bead and HAZ, delivering a PWHT-like result from the welding itself. SMAW’s controllable deposition is why stick welding remains the go-to process for temper bead work.
• Extended holding: after welding, hold the part at roughly 150–200°C (300–400°F) for several hours to give hydrogen a longer, slower escape route. ESAB’s preheat and PWHT overview covers the hydrogen-release rationale and how it interacts with the cooling cycle.

Whatever route you take, schedule magnetic particle or ultrasonic inspection well after the delayed-crack window has passed — heat treatment changes the answer, not the requirement to verify it.

Common Mistakes, Troubleshooting, and How to Catch Cracks Early

Most hardened-steel weld failures aren’t metallurgical mysteries. In hindsight, they trace back to a short list of avoidable shop mistakes, and they’re exactly what gets a repair rejected weeks after the welder has gone home.

The Mistakes That Cause Failures

• Welding a part in unknown condition. “It’s probably 4140” is not a material specification. Annealed, normalized, quenched and tempered, or case-hardened condition each changes the required preheat, filler, and cooling procedure. No identification means guessing.
• Skipping preheat because the job is small. A two-inch crack on a 40-pound shaft still sits in a thick, highly restrained section, and restraint is one of the three converging factors in hydrogen cracking.
• Wet or improperly stored electrodes. Even low-hydrogen consumables pick up moisture from humid air within hours of exposure. Storage and baking discipline is the cheapest hydrogen control in the shop; it’s covered in our guide to alloy steel welding electrodes.
• No interpass temperature control. Letting later passes run hotter than specified progressively enlarges the heat-affected zone and pushes coarse-grained martensite deeper into the base metal.
• Excessive weaving and heat input. Wide weave beads coarsen HAZ grains and build residual stress. Stringer beads keep the thermal cycle tighter and more repeatable.
• Welding through paint, rust, oil, or cadmium plating. Coatings introduce hydrogen, embrittling elements, and porosity at once, and cadmium produces highly toxic fumes. Ventilation and fume control under 29 CFR 1910.252 apply squarely to coated-part repairs.
• Quenching with compressed air. Blowing air or CO₂ on a hot bead to speed up a job is one of the fastest ways to create fresh untempered martensite and a delayed crack.
• Grinding a crack out only partway. Cracks frequently run farther than the visible indication. Stop at the first hint of clean metal and you’ve left the crack tip buried, ready to re-initiate under load.
• Re-welding a rejected repair without re-establishing the procedure. The second attempt starts on a part with fresh thermal history, higher residual stress, and possibly hydrogen already in the metal. Re-establish the procedure from preheat, not from the arc.

A Practical Troubleshooting Flow for a Cracked Weld

When a hardened-steel weld cracks — and if it’s going to, it most often shows up during the delay window described earlier — work the problem in this order rather than jumping straight back to the arc:

1. Locate the crack completely. Clean the area and use dye penetrant or magnetic particle inspection to map the full length. The visible surface break is usually a fraction of the actual crack.
2. Determine the cause before you touch a grinder. Underbead location suggests hydrogen and hardness; toe cracks suggest restraint and stress; porosity points to coatings or damp consumables. Fixing metal without fixing the cause guarantees a repeat.
3. Remove the crack fully. Gouge or grind back to demonstrably sound metal, then re-inspect to confirm the crack tip is gone. Carbon-arc gouging is a common method, and our piece on air arc welding and carbon arc gouging covers the process and its safety requirements.
4. Re-establish preheat exactly as the procedure calls for, measuring rather than estimating, and verify interpass temperature before the root pass goes in.
5. Re-weld with qualified parameters: controlled heat input, stringer beads, low-hydrogen consumables, documented preheat and interpass values.
6. Re-inspect after the delay window. Don’t sign off the same shift. Hold the part, keep it warm if PWHT is called for, and repeat inspection after hydrogen has had time to diffuse.

Catching Cracks Early: The NDT Toolkit

Four methods cover nearly every hardened-steel repair situation:

• Visual inspection catches surface-breaking cracks and workmanship issues, but nothing hidden under the bead.
• Dye penetrant (PT) is inexpensive and highly sensitive to tight surface cracks, including hairline indications the eye misses.
• Magnetic particle (MT) detects surface and near-surface discontinuities in ferromagnetic parts, and works well on shafts and machined surfaces.
• Ultrasonic (UT) reaches deeper flaws, including underbead cracks buried beneath the weld deposit.

The critical timing rule ties directly to delayed cracking: a weld can pass visual inspection on Friday afternoon and present a sharp crack line on Monday. As TWI’s guidance on hydrogen cracks in steels explains, hydrogen needs time to diffuse to stress raisers before it can initiate a crack, so acceptance inspection is repeated after that window closes, not before.

One last discipline point: preheat and slow controlled cooling don’t just manage hydrogen and hardness, they also lower residual stress, the third leg of the cracking tripod. That’s why a procedure that “worked last time” with faster cooling isn’t actually proven. Track what you did — preheat temperature, interpass limits, cooling method, inspection timing — and treat that record as part of the repair, not paperwork afterward. A hardened-steel repair that only looks done is the most expensive kind a shop can produce.

Weld, Machine, Braze, or Replace? A Practical Decision Guide and Final Takeaways

Every repair on hardened steel starts with a question that has nothing to do with amps or travel speed: whether to weld at all. The shop that cracked the 4140 ram never asked it. Ask it, and the choice gets clear fast: each route carries a different level of metallurgical risk, cost, and downtime.

Five Routes, Ranked by Risk

Route	Cost & time	Metallurgical risk	Where it fits
Weld in the hardened condition	Lowest cost, fastest if it works	High: hard martensitic HAZ, delayed cracking	Only with a written procedure, preheat and hydrogen control
Anneal or normalize → weld → reheat treat	Higher cost, needs a furnace	Lowest: no hard HAZ problem to solve	Removable parts where shop heat treatment is available
Braze welding / bronze overlay	Moderate	Low: base stays below its transformation temperature	Buildup, alignment, non-structural joins
Weld overlay / hardfacing	Moderate to high	Moderate: dilution and cracking depend on alloy	Wear surfaces, not load paths
Machine or buy new	High material cost, low risk	None	Safety-critical, small, or cheap-to-replace parts

Welding in the annealed or normalized condition and heat treating afterward is the lowest-risk route, because it eliminates the hard-HAZ problem before the arc is struck. If the part can go to a furnace and come back through quench and temper, or be normalized and machined back to size, that is almost always the smartest money you will spend. A 4140 welding guideline from Associated Steel says it plainly: know the delivered condition, weld in the recommended window, then restore properties with heat treatment.

Braze welding and hardfacing overlays are legitimate alternatives when base-metal fusion and full strength are not required, since they keep the base metal below its transformation temperature. A bronze build-up on a worn shaft journal, a hardfaced crusher tooth, a braze-repaired tool shank: none needs a fused joint, and none risks forming brittle martensite. Torch-based joining is covered in Air Acetylene Welding: Torch Basics, Brazing & Safety.

Match the Route to the Part

• Field repair of a large shaft: welding or overlay is usually the only practical option. Preheat, low-hydrogen consumables, temper bead technique if PWHT is impossible, and magnetic particle inspection after the delay window.
• Small tool or fixture: braze, weld in the soft condition and reheat treat, or make a new one. Tool steels rarely justify the risk of a direct hard weld.
• Safety-critical component (brake, steering, pressure-retaining): replace it. No weld on a part of unknown history is a defensible engineering decision.

The Complete Workflow, as a Checklist

1. Identify the alloy and its delivered condition. If you cannot, test rather than guess.
2. Decide whether welding is the right route versus braze, overlay, machining, or replacement.
3. Calculate carbon equivalent and set preheat from CE, thickness, and restraint.
4. Choose a low-hydrogen process and filler, and confirm matching versus undermatching. Selection notes live in Alloy Steel Welding Electrodes: Selection, Use & Safety.
5. Bake or dry consumables, clean the joint, and remove coatings, rust, and oil.
6. Preheat uniformly, measure with a calibrated device, and hold interpass temperature.
7. Weld with stringer beads, controlled heat input, and a balanced pass sequence.
8. Cool slowly or go straight into PWHT without letting preheat drop.
9. Inspect after the delayed-crack window by dye penetrant or magnetic particle; use Air Arc Welding and Carbon Arc Gouging to remove cracks to sound metal before re-welding.
10. Re-establish the full procedure before every re-weld of a rejected repair.

Safety: The Part You Cannot Skip

Coated steels, high-alloy grades, and old cadmium-plated surfaces produce fumes no one should breathe, so ventilation and respiratory protection come before the arc. Hot metal, spatter, and burn-through demand normal PPE, and hot work in an occupied building needs a permit and fire watch. OSHA’s welding, cutting, and brazing standard (29 CFR 1910.252) sets the baseline for guarding, ventilation, and fire control; local rules usually go further. Gas and cylinder handling is covered in Airgas Welding Supplies: The Complete Buyer’s Guide.

Final takeaway: yes, hardened steel can be welded, but only with deliberate control of three forces at once, as detailed in TWI’s guidance on hydrogen cracks in steels. Control the cooling rate with preheat, control hydrogen with dry low-hydrogen consumables, and control stress with restraint-aware sequencing and post-weld heat treatment. Do that, and the weld that looks right on Friday is still holding on Monday. For pass control and bead placement, see Advantages of Stick Welding: Why SMAW Still Wins, and keep PPE, fumes, and fire watch current with our AllGas Welding Supply: Gases, Services & Safety Guide.

Frequently Asked Questions

Can you weld hardened steel without cracking it?

Yes, but only with procedure control. The weld must limit heat input, use preheat and interpass temperature, employ low-hydrogen consumables, and follow with slow cooling or post-weld heat treatment so the heat-affected zone does not form hard, brittle untempered martensite.

What preheat temperature is needed for welding hardened steel?

It depends on carbon equivalent, thickness, and restraint, but alloy hardenable steels commonly preheat in the 150–300°C (300–570°F) range, with grades like 4140 often quoted around 400–750°F depending on condition. Verify with temperature-indicating crayons or a thermocouple at the location specified by the procedure.

How long after welding can a hydrogen crack appear?

Hydrogen-induced cracking typically occurs below about 600°F, most often at or below 300°F, within 48 hours of welding and sometimes as late as a week afterward. That is why inspection is repeated after the part has cooled and sat for a period.

What filler metal should I use on hardened steel?

Use a low-hydrogen filler matched to the required service properties — often an E7018-class electrode for general work or a matching alloy filler for 4140/4150 repairs, sometimes with an undermatching, more ductile filler to reduce cracking risk in highly restrained joints. The correct choice follows the carbon equivalent and the required strength of the finished part.

Do you always need post-weld heat treatment?

Not always, but when PWHT is skipped you must compensate with stricter preheat, low-hydrogen consumables, controlled heat input, and sometimes temper bead or buttering techniques. Safety-critical and highly restrained joints should not omit PWHT without a qualified procedure justifying it.

Why did my weld crack under the bead on hardened steel?

Underbead cracking is hydrogen-induced cracking: a hard heat-affected zone, hydrogen from moisture or consumables, and residual tensile stress combined. Typical causes are missing or too-low preheat, damp electrodes, and cooling too fast.

Is it better to weld hardened steel in a heat-treated state or annealed first?

Whenever possible, weld in the annealed or normalized condition and re-heat treat afterward. That eliminates the hard-heat-affected-zone problem, gives the most reliable metallurgy, and is far easier than trying to control hardness with preheat alone.

Can a cracked hardened steel part be repaired by welding?

Yes, if the crack is fully removed by grinding or gouging to sound metal and the repair follows the same preheat, hydrogen control, and post-weld treatment as the original weld. Re-welding without finding the crack root or re-establishing procedure usually causes the crack to return.

When should you not weld hardened steel?

Avoid welding when the part is safety-critical and cannot be post-weld heat treated, when the steel grade or delivered condition cannot be identified, or when the section is highly restrained and prone to cracking. In those cases, machining a replacement, brazing, or weld overlay may be safer and cheaper.