Preheat Cast Iron for Welding: Temperatures, Methods & Mistakes

The antique bench vise had been in the family for three generations, and the new crack in its jaw looked like a five-minute job. Twenty minutes later, a sharp ‘ping’ from the cooling weld turned a repairable casting into two broken pieces. The welder had done almost everything right, except one thing: he never preheated the iron.

Why Preheat Cast Iron at All? The Real Answer Behind the Cracks

The bench vise from the opening story did not fail because the welder lacked hand skills. It failed because the casting cooled too quickly. The bead may have looked sound for a few seconds, but the metal around it was already transforming into something hard, brittle, and ready to crack. Preheat is the part of the job that controls that transformation.

The real answer to “why preheat cast iron?” is not that heat magically makes a difficult alloy weldable. It is that preheat slows the cooling rate of the weld and the surrounding heat-affected zone. That slower cooling gives carbon time to stay in graphite form instead of forming hard white iron and martensite, and it lets contraction stresses relax gradually rather than snap the casting.

Two Ways a Cast Iron Weld Fails

The first failure is metallurgical. Cast iron contains roughly 1.7% to 4.5% carbon, far more than weldable structural steels. When the heat-affected zone is heated above the critical range and then cools quickly, that carbon does not have time to diffuse back into graphite. Instead, it forms white cast iron and martensite, both of which are extremely hard and brittle. Lincoln Electric describes a critical temperature in most cast iron of about 1450°F, where the conditions that lead to cracking begin. A preheated part narrows the temperature gap between the weld pool and the base metal, so the heat-affected zone cools at a gentler rate.

The second failure is mechanical. As the weld solidifies and cools, it shrinks. The surrounding casting resists that movement because it is stiff, thick, and often complex in shape. The heat-affected zone sits between the shrinking weld and the unyielding base metal, and it has just been made brittle by the thermal cycle. Something has to give, and it is usually a crack at the edge of the weld or just outside the heat-affected zone.

Preheat attacks both problems at once. A warm casting does not pull heat out of the weld as aggressively, so the weld and the heat-affected zone cool together instead of the weld chilling against cold iron. That gentler thermal gradient reduces the peak stress and gives the metal more time to accommodate contraction. In practical terms, preheat turns a sudden, violent cooling event into a slower, more forgiving one.

What Slow Cooling Actually Changes

Cooling rate is the hidden variable in every cast iron repair. Fast cooling favors hard, brittle phases. Slow cooling moves the heat-affected zone closer to equilibrium, where carbon can stay in graphite form and the matrix is less likely to transform into martensite. That is why a preheated casting can survive a repair that would crack a cold one, even when the welder uses the same filler and the same technique.

There is also a thermal shock element. An arc struck on a cold, thick casting creates an abrupt local expansion. Cast iron is strong in compression but weak in tension, and it has little ductility to absorb that mismatch. Preheating raises the starting temperature, so the arc does not hit the casting like a thermal hammer. The difference is especially noticeable on edges, corners, and thin sections that heat and cool faster than the bulk of the part.

• Less white iron: slower cooling gives carbon a chance to stay as graphite.
• Less martensite: the heat-affected zone is less likely to harden into a brittle band.
• Lower stress: the weld and base metal contract over a smaller temperature range.
• Fewer thermal shocks: the arc starts on warm metal, not cold metal.

None of this makes cast iron as forgiving as mild steel. Graphite flakes in gray iron still act like tiny internal notches, and a badly restrained repair can crack even with preheat. What preheat does is remove the most common, most preventable cause: cooling too fast and shrinking too hard. That is why the bench vise cracked, and that is why preheat belongs at the start of the job, not as an afterthought.

Cast Iron Metallurgy for Welders: Graphite, White Iron, and the Heat-Affected Zone

You do not need a metallurgy degree to repair cast iron well, but you do need to understand three things: how the carbon is sitting inside the iron, what heat does to that carbon, and what the arc drags out of the base metal and into the weld pool. Get those three straight and most of the confusing advice about preheat, filler choice, and peening suddenly makes sense.

The Cast Iron Family: Five Cousins, Five Personalities

“Cast iron” is not one material. It is a family of alloys, all roughly 2 to 4 percent carbon plus silicon, but with carbon arranged in very different shapes. The shape matters more than the chemistry number on the spec sheet.

• Gray iron holds its carbon as thin graphite flakes scattered through an iron matrix. It is cheap, machines beautifully, damps vibration, and is the classic material for vises, manifolds, machine bases, and old pump housings.
• Ductile (nodular) iron gets a magnesium or cerium treatment that makes the graphite grow as rounded nodules instead of flakes. That single change gives it markedly more ductility and toughness. Most ductile iron you meet falls under ASTM A536 territory, and it is the cast iron grade most often welded successfully.
• Malleable iron starts as white iron and is heat treated until the carbon clumps into irregular rosettes. It is tough for its thickness but usually found in small, thin fittings.
• Compacted graphite iron (CGI) sits between gray and ductile, with short, connected, worm-like graphite. It is common in modern diesel engine blocks and is notoriously fussy to weld.
• White iron has almost no free graphite at all. Its carbon is locked up as hard iron carbide, making it abrasion-resistant and nearly unweldable.

Why Graphite Flakes Are the Real Problem

Graphite carries virtually no strength. Picture a gray iron casting as a steel-like matrix riddled with thousands of tiny internal cracks. Those flakes concentrate stress at their sharp tips, so when the part is loaded in tension, fracture starts almost immediately and propagates with very little plastic strain. That is why gray iron has almost no ductility and why it snaps instead of bending. It also explains why a crack in gray iron tends to run: once it finds a flake, it has a ready-made path.

The graphite is not all bad news. It lubricates machining, absorbs vibration, and helps chips break cleanly. But for welding, it means the base metal offers you almost no forgiveness when contraction stress arrives.

What Happens in the Heat-Affected Zone

Now consider what the arc does. Right beside the weld bead sits a narrow band of base metal that gets hot but does not melt: the heat-affected zone, or HAZ. In steel, this zone is usually manageable. In cast iron, it is where repairs die.

Cool the HAZ quickly and the carbon has no time to diffuse back into graphite. Instead it stays trapped as iron carbide, producing a hard, brittle white cast iron structure, and the rapid quench can also form martensite. Both are glass-hard and crack-prone. Lincoln Electric’s cast iron guidance is blunt about the consequence: the weld becomes unsuitable where service stresses are significant, and the brittle zone can crack on its own as it cools.

Slow the cooling and the story changes. With preheat holding the whole area warm and insulation slowing the drop afterward, carbon keeps enough thermal energy to stay as graphite rather than re-forming as carbide. The HAZ still is not as good as new metal, but it is far less likely to crack. That is the entire metallurgical case for preheat, in one paragraph.

Dilution and Carbon Pickup

The weld pool is a mixture. Some of it is filler metal; some is base metal melted off the sidewalls. That fraction of base metal is called dilution, and cast iron base metal is loaded with carbon. Melt too much of it into a steel electrode deposit and you have effectively created a high-carbon steel bead that air-hardens into something closer to a file than a weld.

This is why nickel-base fillers dominate cast iron repair. Nickel does not form the same brittle carbides, so it tolerates carbon pickup and stays relatively ductile. It is also why welders keep current low, use small electrodes, and run short beads: every one of those choices reduces how much base metal gets dragged into the pool. The same logic of matching consumable to parent metal applies to alloy steel welding electrodes, just with different chemistry at stake.

Identify Before You Heat

Before any preheat, confirm what you are holding. Look for casting marks and grade codes, check a chip from a hidden edge (gray iron gives fine, powdery chips; ductile iron gives tougher, curly ones), and compare a fresh fracture surface. Gray iron breaks with a dull, dark gray, granular look. White iron shows a bright, almost silvery fracture and often rings hard under a file. Ductile iron fractures with a tougher, less uniform appearance.

White iron and heavily chilled castings should send you toward brazing or replacement, not arc welding. The same caution applies to thin, complex, highly restrained parts. Knowing which cousin you have is the difference between a repair and a pile of scrap, and it is the step most rushed jobs skip entirely.

Preheat Temperature Charts: What to Use for Gray, Ductile, and Malleable Iron

Ask ten welders for a cast iron preheat number and you will get ten answers. Ask a code book and you will get a range wide enough to drive a truck through. Both are normal. The useful question is not “what is the preheat for cast iron?” but “what is the right preheat for this iron, this section, this restraint, and this process?”

The published starting points

Lincoln Electric’s cast iron welding guidelines give a broad preheat range of 500–1200°F, with a practical ceiling around 1400°F. Push much beyond that ceiling and the casting enters the critical temperature range, where cracking conditions become far more likely.

TWI’s guidance for spheroidal graphite iron is more specific. It advises 100–200°C (212–392°F) for thicker and more complex ferritic grades, at least 250°C (482°F) for normalized, pearlitic, or hardened-and-tempered grades, and 320–330°C (608–626°F) where high welding stresses are expected.

AWS D11.2, the structural welding code for cast irons, reflects how much the process changes the answer: AWS guidance cited in industry discussion gives a preheat and interpass range of 70–600°F for arc-welded gray iron, but 800–1200°F for oxyacetylene welding. Filler suppliers commonly land in the middle with 300–600°F (149–315°C), depending on iron type and mass.

Iron / condition	Typical preheat band	What pushes you higher
Gray iron, light and unrestrained	300–500°F (149–260°C)	Thick sections, long cracks, high restraint
Gray iron, heavy or restrained	500–900°F (260–482°C)	Mass, complex shape, limited expansion
Ductile iron, ferritic	212–392°F (100–200°C)	Thickness above roughly 1 in, complex geometry
Ductile iron, pearlitic or Q&T	At least 482°F (250°C)	Hardened condition, high joint stress
Malleable iron, ferritic	300–600°F (149–315°C), typical supplier band	Pearlitic grade, thick flanges, rigid assemblies
High-stress ductile joint	608–626°F (320–330°C)	Restraint, thick sections, critical service
Oxyacetylene welding, gray iron	800–1200°F (427–649°C)	Slower process needs a hotter starting point

Why the ranges are so wide

Cast iron is a family, not a single metal. Gray iron’s graphite flakes, ductile iron’s nodules, malleable iron’s temper carbon, and white iron’s carbides all respond differently to heat. Add section thickness, casting mass, joint restraint, carbon equivalent, filler choice, and process, and a single number becomes meaningless.

Process matters as much as metallurgy. Oxyacetylene delivers heat slowly and keeps the weld pool fluid longer, so its published band sits higher. Arc welding is faster and more localized, so the low end can start at room temperature for lightly restrained gray iron. That does not make 70°F a good idea for a 500 lb frame with a crack running through a thick flange.

To narrow the chart, work through five questions: What iron is it? How thick is the section? How restrained is the joint? Which process and filler are you using? What does the manufacturer or code require? A 400 lb gray iron housing with a 1 in section and a crack between two rigid bosses belongs near 500–600°F, not at the bottom of the AWS range.

Interpass temperature is not optional

Preheat gets all the attention, but interpass temperature is where most home-shop repairs quietly fall apart. The number you set at the start has to hold through every pass. Let the part cool between beads and you reintroduce the thermal shock you paid good money to avoid. A quick touch with a temperature stick before each pass keeps you honest, and a firebrick blanket or a second torch helps hold the heat on smaller castings that lose it fast.

How to Preheat Cast Iron: Furnaces, Torches, Resistance Heaters, and Improvised Methods

Choosing a preheat temperature is the easy part. Getting the entire casting to that temperature evenly, without cracking it before you strike an arc, is where the real work happens. A torch waved over a joint for two minutes is not preheat. It is a thermal shock waiting to happen.

Complete, Local, and Indirect Preheat: Pick the Right Scope

Complete preheat brings the whole casting up to temperature. It is the safest approach for intricate castings with widely varying section thickness, such as cylinder blocks, manifolds, and pump housings. The entire part expands and contracts together, and the preheat temperature must be maintained throughout welding, not just at the start.

Local preheat heats a defined zone around the repair. It suits small, unconstrained parts or a crack in a thick, simple section. But local heating of a small area on a large, complex casting can itself cause cracking through unequal expansion. The heated spot grows while the surrounding iron stays cold, and the restraint can tear the casting. Local preheat must always be gradual, with a generous heated zone rather than a tight hotspot.

Indirect preheat goes a step further. In addition to the local preheat, it brings other critical parts of the job to about 200°C (392°F) so they contract with the weld and reduce contraction stresses. This is useful on frames, spokes, and long slender castings where one area moves against another.

The Heating Methods That Actually Work

A full furnace is the gold standard. The casting is charged cold, ramped slowly, soaked until the core matches the surface, welded while hot, and then cooled under control. An oven or heat-treatment furnace gives you that same controlled ramp and soak, provided it is large enough and rated for the part. Never force a cold casting into a hot chamber.

Resistance and ceramic pad heaters are the practical shop-floor option. Flexible pads clamp or wrap around the casting, and a controller holds the setpoint. They are cleaner than a torch and far more even. Choose a controller with a thermocouple input rather than trusting the dial. Good temperature control equipment matters more than the heater itself, and a reliable supplier such as Airgas welding equipment can help match pads and controllers to the job.

Oxy-fuel torches are convenient but uneven. A single flame puts intense heat into one spot while the rest of the casting lags behind. On a large casting, that gradient is exactly what causes cracking. Use a large, soft, multi-flame head, keep it moving, and heat both sides where possible. For oxyfuel welding of cast iron, the part generally must be preheated to a dull red heat of approximately 650°C (1,200°F) and protected from drafts during welding, as outlined in these cast iron welding procedures.

Improvised brick enclosures with gas jets or slowly burning charcoal are a traditional way to preheat a casting evenly. The enclosure traps heat, reduces drafts, and lets the part soak. It is slower and dirtier than a furnace, but it works well for one-off repairs. Induction heating is fast and precise, with coils shaped to the joint, though it is less common in general repair shops.

Method	Best for	Main risk
Full furnace	Complex castings, production repair	Size limits, slow cycle
Oven or heat-treat furnace	Controlled ramp and soak	Chamber capacity
Ceramic pads	Local and complete preheat	Poor contact, uneven zones
Oxy-fuel torch	Small parts, field work	Hotspots and thermal shock
Brick and charcoal	One-off, low-budget repairs	Smoke, uneven heat, fire risk

Heat Soak, Support, and Expansion

A surface reading does not mean the casting is hot through. Thick sections need dwell time so the core catches up with the skin. Many shops allow roughly one hour of soak per inch of thickest section, with a one-hour minimum, then confirm with a contact thermocouple. Rushing this step produces a part that reads hot on the outside and cracks from the inside.

Support the casting so it can move. Set it on clean blocks or rollers, remove rigid clamps and fixtures, and leave space for expansion. A part pinned in place cannot grow, so it relieves the stress by cracking instead.

Setting Up a Preheat: A Practical Sequence

1. Identify the iron and confirm the part is weldable before applying heat.
2. Clean the casting and remove paint, oil, and casting skin from the repair zone.
3. Decide complete, local, or indirect preheat based on mass, shape, and restraint.
4. Support the part freely and remove anything that restricts expansion.
5. Apply heat slowly and evenly, heating the thick sections first.
6. Soak until the core reaches temperature, then verify with a contact probe.
7. Weld while holding the preheat and interpass temperature within the qualified range.
8. Cool under insulation or in the furnace, never in a draft.

The method matters less than the thermal behavior it produces: slow, uniform heating, a genuine soak, and a controlled cool. Get those right and even a brick pile and a charcoal bed can outperform a careless torch on a valuable casting.

Measuring and Holding Preheat: Thermocouples, Crayons, and the Mistakes That Fool You

You can own the best furnace in the county and still ruin a casting, because preheat only works if you actually know the temperature. Guessing is where most repairs quietly go wrong. The iron may look hot, the torch flame may be roaring, and the surface may be glowing a promising dull red — and yet the mass behind the joint is still 200°F while the skin reads 700°F. That gap is the difference between a weld that survives cooling and one that cracks overnight.

AWS guidance is blunt about this: preheat and interpass temperatures should be closely monitored with pyrometers, temperature-sensitive crayons, or thermocouples so the process stays under real control rather than hopeful estimation (AWS Welding Handbook, Cast Irons). Verification is not paperwork. It is the part of the job that decides whether the thermal plan you chose in the previous sections actually happens at the joint.

The Four Ways to Read Cast Iron Temperature

Each method has a personality, and each one lies in a different way.

• Thermocouple contact meters. A surface probe or a clamped bead gives the most trustworthy reading on rough cast iron, because it measures actual contact temperature instead of radiation. The catch is contact quality: a probe touching scale, rust, or a curved boss reads low. Clamp or tack the bead near the joint and let it sit long enough to equalize.
• Infrared pyrometers. Fast and convenient, and easy to misread. Infrared thermometers can misread rough, oxidized, or low-emissivity cast surfaces, so contact measurement or an emissivity correction is preferable. Shiny machined pads, scale, and mill finish all change how much energy the sensor sees. Aim at a dull, consistent, soot-darkened patch and treat the number as a trend, not gospel.
• Temperature-indicating crayons and pellets. These melt at a rated temperature and are excellent for verifying that a surface has reached a threshold. Mark the thick section before heating and watch for the mark to liquefy. They do not tell you how hot something is beyond that threshold, and they leave residue that must be ground off before welding.
• Heat tint. Straw, brown, purple, blue. Useful as a last resort, useless as a specification, because tint depends on time as much as temperature and varies with surface condition and alloy.

Where to Measure, and How Many Readings

One reading in one spot is not a preheat. Take at least three, and place them deliberately:

1. Beside the joint, within an inch or two of the weld preparation.
2. On the thickest section of the casting, which heats slowest and holds the heat longest.
3. On the opposite side of the weld, to prove the heat has travelled through rather than sitting in the skin.

On a big frame or a housing with several arms, add readings at each extremity. A casting is a heat sink with opinions, and the thin web you measured first will be cooler than the boss you are about to weld.

Heat Soak: The Part Everyone Rushes

Preheat is not a surface event. Thick sections need dwell time at temperature so the interior catches up with the outside. If your meter says 500°F on the skin and you strike an arc thirty seconds later, the weld is effectively being made on a cold casting with a hot jacket. Give the part time to equalize, then re-check before welding.

There is also an upper limit worth respecting. A casting should not be held for long periods at roughly 1450°F, because that is where cracking conditions develop. That is one reason the numbers in the preheat charts stay well below the ceiling — and one reason heat-tint judgement and careless torch work are so dangerous.

Temperature control also has a pre-weld dimension. Uneven or localized heating of a restrained casting creates its own thermal stresses before the arc is even struck. That is why the torch method needs patience: bring the surrounding area up gradually rather than parking a flame on one spot and calling it preheat. The same discipline applies when you switch processes; the principles for holding an even thermal field are not unlike those covered in our guide to air arc welding and carbon arc gouging, where heat input and restraint also shape the outcome.

Monitoring Interpass and Re-Heating Between Passes

The starting preheat is only the first checkpoint. Every subsequent pass adds heat, and a multipass repair can drift above the planned interpass temperature quickly, especially on a small casting. Check between passes, and let the part cool in a controlled way if it climbs too high rather than pushing on.

On the other side of the coin, a casting that sits in a cold shop between passes will lose temperature fast. Re-heat before resuming, and note the reading each time. Keep the interpass window as steady as you can: a sawtooth of hot and cold passes is a stress generator.

What Steals Your Heat

Shops are not ovens. Cold concrete floors pull heat out of a casting through any contact point. Steel tables, fixtures, clamps, and chains act as fins. Open doors, fans, and extraction hoods move air across the surface and cool it unevenly. Set the casting on insulating blocks or firebrick, shield it from drafts, and remove unnecessary clamping where you safely can. Then measure again — because your carefully reached preheat may have been gone before the first bead.

A Simple Record Worth Keeping

Write it down, even on a scrap of paper taped to the bench: iron type, target preheat, measured readings with locations and times, interpass readings, filler and electrode size, and how the part was cooled. It takes a minute, and it turns a one-off guess into something repeatable. If a repair fails in service, that note is the only honest record of what the iron actually experienced — and it is the difference between learning something and starting over blind.

Welding Cast Iron With Preheat: Filler Metals, Current Settings, and Bead Technique

Once the casting is hot, soaked, and verified, the repair comes down to three things: what you put in the groove, how much heat you push through it, and how you lay the bead down. Get the preheat right and then run a fat, hot, weaving bead with the wrong filler, and you will undo every bit of the thermal work you just paid for.

The AWS A5.15 Filler Groups and What Each One Buys You

In North America, cast iron consumables are classified under AWS A5.15, which covers rods and electrodes for oxyfuel gas welding, GMAW, FCAW, and SMAW of gray, malleable, nodular, compacted graphite, and certain alloy cast irons. The classification letters tell you the deposit chemistry at a glance.

Group	Typical deposit	Preheat appetite	Machinable?
ENi-CI	Commercially pure nickel (~95% Ni)	Low to moderate	Yes, easily
ENiFe-CI	Nickel-iron, roughly 55% Ni	Low to moderate	Yes, slightly harder
ENiCu-B	Nickel-copper (Monel type)	Low	Yes
ECi	Cast iron core, matches base metal	Very high	Yes, but demands heat
Steel (E7018 etc.)	Mild steel	Moderate to high	No, hard fusion zone
Bronze (ECuAl-A2, Si-bronze)	Copper alloy, braze welding	Low	Yes

The ENi group is the workhorse. Nickel electrodes produce a soft, ductile, machinable bead and fusion zone, with some sacrifice in strength and a noticeably higher price per rod. Lincoln Electric notes that on larger castings these nickel electrodes can be run without preheat, but only where the loads are light. ENiFe-CI is the usual pick for ductile iron and for joints that need a bit more tensile strength; TWI’s guidance on MMA welding of spheroidal graphite iron walks through the ferritic versus pearlitic grades, because the pearlitic ones want more heat and more care.

ECi-group electrodes with a cast iron core give weld metal closest to the original casting, which sounds ideal until you read the fine print: they should not be used unless the part can be furnace-heated or given an extremely high, red-hot preheat. This is the filler for a shop with a heat-treat oven, not for a driveway repair. Steel electrodes are cheap and strong but leave a hard, crack-prone fusion zone and a deposit you will not machine without carbide and patience. Bronze braze welding sits outside fusion welding entirely. It never melts the casting, keeps dilution near zero, and remains the honest answer for many antique and decorative repairs. Ampo’s cast iron welding guide is a useful cross-check on which alloy suits which iron.

Current Settings, Polarity, and the Case for Running Cold

With preheat in the part, it is tempting to open the machine up. Do the opposite. Lincoln Electric’s cast iron guidelines recommend low current to minimize admixture and residual stress, and restricting welds to roughly 1-inch segments where necessary to prevent stress buildup and cracking. Dilution is the enemy: too much current drags carbon-rich base metal into the pool, and the deposit stops behaving like nickel.

Practical starting points for SMAW on preheated iron:

• Polarity: DCEP (electrode positive) for most nickel and nickel-iron electrodes. Follow the manufacturer’s data sheet; some bronze and specialty rods prefer DCEN.
• Amperage: run at or slightly below the low end of the manufacturer’s range for the rod diameter. If the arc wanders or stubs, the iron is probably not hot enough, not the machine.
• Electrode size: small diameter, typically 3/32 in or 1/8 in. Thin rods at low current give you control the big stuff cannot.
• Arc length: short. A tight arc keeps the heat where you want it and shields the pool properly.

If you are running SMAW, it is worth revisiting why the process still dominates repair work like this — see advantages of stick welding for the practical reasons.

Bead Technique: Short, Skipped, and Peened

Technique is where preheat and filler either pay off or get wasted.

1. Stringer beads, not wide weaves. A narrow bead limits the heat-affected zone and the shrinkage volume.
2. 1 to 3 inch segments. Start short; lengthen only if the part is heavy and well supported.
3. Backstep and skip welding. Run a bead, skip ahead, come back. Alternate sides of the joint so contraction pulls against itself instead of into one long seam.
4. Controlled wandering. Spread the passes around the casting so no single area stays at welding temperature for long.
5. Peen immediately. Each bead gets peened right after deposition with a blunt tool — a rounded chipping hammer, a ball-peen, a needle gun on low setting. Peening spreads the weld metal and counteracts contraction stresses. Blunt is critical; a sharp tool tears the hot surface. Never peen the root pass of a joint that must be leak-tight, and never peen the final pass on a machined surface.

Joint Preparation: Where Most Porosity Is Born

Cracks must be removed completely, not capped. Chase them to solid metal with a burr or a carbide bit, open the groove into a U or a wide V, and take out the casting skin along the walls — that skin carries sand, oxide, and years of oil. Air arc gouging is often the better tool for crack removal and for excavating contaminated metal, because grinding can smear graphite and impurities along the groove wall instead of cutting them away.

Degrease before you strike an arc, and again after heating, because hot castings sweat oil out of porosity. Preheating burns off paint and oil, which is convenient for you and unpleasant for your lungs. Any contaminant that survives becomes porosity in the deposit.

Buttering: The Trick for Dirty or Restrained Joints

When the casting is heavily oil-soaked, or the joint is so restrained that a full fusion pass would tear itself apart, butter the faces first. Lay a thin layer of nickel or nickel-iron on each prepared face, peen it, let it cool, clean it, then weld against the buttered layer. The butter pass dilutes into contaminated base metal where it can afford to, and the closing passes then run on clean nickel instead of on oil and graphite. It costs an extra hour and saves a lot of castings.

Get the filler, the amperage, and the bead sequence right and the casting has a genuine chance. Get them wrong and the preheat just delays the ping.

The No-Preheat Route: When Cold Repair Makes More Sense

Preheat is not always the answer. Sometimes it is physically impossible, sometimes it is destructive, and sometimes it is simply the wrong tool for a thin-section casting that would distort the moment you brought a torch near it. The cold route exists for those jobs, and it deserves a fair hearing instead of the dismissive “you have to preheat cast iron” line that circulates in shop talk.

But the cold route is not easier. It is a different discipline, and it demands more patience, not less.

When Preheat Is Genuinely Off the Table

Lincoln Electric’s guidance on cast iron welding is clear that the cold approach has a legitimate place. Some castings cannot be safely heated to 500°F or more without collateral damage. The list is longer than most welders expect:

• Large installed castings — machine bases, press frames, and structural housings that cannot be moved to a furnace and cannot be uniformly heated in place.
• Engine blocks still in the chassis — seals, gaskets, sensors, wiring, and fuel systems all have temperature limits far below any useful preheat.
• Parts with bearings, bushings, or seals pressed into them. Even 400°F can soften or deform some elastomers and lubricants.
• Painted, coated, or plated surfaces that will blister, discolor, or release fumes.
• Machined surfaces — bearing bores, dovetails, and precision ways that cannot tolerate oxidation scale or thermal movement.
• Thin sections where heat input alone risks warping or cracking from the temperature differential itself.

In these cases, forcing a preheat does more harm than good. The cold route becomes the responsible choice.

The Cold Technique, Step by Step

The governing idea is to keep the casting cool but not cold. According to Lincoln Electric’s cast iron guidelines, raising the casting to roughly 100°F is helpful because it knocks off surface moisture and takes the edge off thermal shock, but you should never heat it so hot that you cannot place your bare hand on it.

That hand test is not a joke. It is a practical, repeatable field check that keeps you in the safe window. If you can hold your palm against the casting near the weld zone, you are in the right range. If you pull away, you have gone too far.

A quick warm-up can come from a heat gun, a hot-water wipe, or, on an engine block, simply running the engine briefly. The goal is a casting at skin temperature, nothing more.

From there, the weld itself does the work:

1. Use nickel-based filler. Nickel-iron (ENiFe-CI) or pure nickel (ENi-CI) deposits are ductile and tolerate the carbon pickup that comes with dilution. These are covered in detail in the filler section of AWS A5.15.
2. Keep current low. Small diameter electrodes, low amperage, short arc. You are trying to minimize penetration into the brittle zone and limit dilution.
3. Run beads about 1 inch long. Lincoln’s no-preheat guidance is explicit on this length. Short beads limit heat input and residual stress.
4. Peen immediately. Right behind the bead, while it is still hot, with a blunt tool. Peening stretches the cooling weld metal and relieves contraction stress.
5. Fill every crater. A star crater is a built-in crack starter. Never leave one open.
6. Let each area cool before returning. No water. No compressed air. Let it sit.

That last point bears repeating, because it is the most commonly violated rule in cold repair. Lincoln advises that welds in the no-preheat approach be allowed to cool naturally and never accelerated with water or compressed air. Quenching a nickel deposit on cast iron is a reliable way to crack the casting you just saved.

Controlled Wandering and Backstep

Because you are working on a cold casting, you cannot rely on stored heat to smooth out thermal gradients. You have to spread the heat yourself.

Controlled wandering means moving around the joint after each short bead: weld an inch at one end, move to a different area, weld another inch, and keep rotating. No single spot ever gets enough concentrated heat to build a steep gradient.

Backstep means running each bead opposite to the direction of overall progression, so the weld you just laid preheats the start of the next one slightly while the finished end cools behind you.

Together these two habits do much of the work that preheat would otherwise do, just on a smaller, local scale. If you have spent time with thin-material distortion control in aluminum, the mindset is familiar; the practical side is covered in the aluminum MIG welding guide.

The Trade-Off You Must Accept

Here is where honesty matters. Even with nickel filler, the heat-affected zone of a cold-welded casting stays hard and brittle. The rapid cooling at the edge of the weld still produces white iron and martensite. Preheat plus slow cooling is what prevents that; the cold route does not.

So the strategy shifts. You are not trying to eliminate the brittle zone. You are trying to make it as small as possible, and you are relying on the ductile nickel weld metal to yield and absorb strain rather than transmitting it into the brittle base metal.

That has practical consequences for how you inspect and how you set expectations. The repair can be strong and serviceable while still having a hard underbead. It is not the same quality as a properly preheated and slow-cooled job, and you should say so.

Any post-weld machining will also be affected, since that hard zone will fight a cutting tool. That is one more reason to know what you are committing to before you strike an arc.

Don’t Change Horses Mid-Repair

This is the rule that turns a mediocre repair into a failed one. Lincoln’s guidance is explicit: once you select a hot or cold method, stick with it, because switching mid-repair causes problems.

The failure looks like this. A welder starts cold, gets a small crack, decides the job “needs more heat,” and torches the area to 400°F. Now the part has a partially preheated zone, a cold zone, a hard heat-affected zone from the first beads, and a thermal gradient running through all of it. The contraction stresses from that mixed condition are exactly what cracks castings.

The same applies in reverse. Start hot, then get impatient and stop blanketing between passes, and you have created a differential cooling pattern that no amount of peening will rescue.

Cold Repair Is a Legitimate Choice

The cold route is not the amateur’s shortcut. For an installed engine block, a bearing-laden housing, or a thin casting that would distort under preheat, it is the only sensible option. It requires nickel filler, low current, one-inch beads, immediate peening, full craters, patient cooling, and the discipline to stay cold once you have committed.

What it cannot deliver is a soft, machinable heat-affected zone. Accept that, keep the zone small, let the weld metal do the yielding, and you will get a repair that holds. Pretend otherwise, and you will be the welder listening for a ping.

Post-Weld Cooling and Stress Relief: The Half of the Job Most People Skip

The arc stops, the slag chips away, and most welders start packing up. That is the moment a cast iron repair is most vulnerable. The weld itself may look perfect while the metal around it is contracting, pulling against a rigid casting that has no give. Cooling is not a cooldown period. It is the second half of the thermal process, and rushing it can undo every careful minute you spent preheating.

Cast iron cracks because of stress and because of microstructure. Slow cooling addresses both. It gives the heat-affected zone time to soften instead of hardening into brittle white iron or martensite, and it lets the weld and base metal contract gradually rather than fighting each other. Lincoln Electric’s cast iron welding guidelines treat controlled cooling as part of the procedure, not an optional courtesy.

Slow Cooling Methods That Actually Work

The goal is simple: reduce the cooling rate without introducing a draft, a cold surface, or a shock. Wrapping the casting in an insulating blanket or burying it in dry sand slows cooling and reduces cracking tendency. For small parts, a bucket of dry sand, vermiculite, or lime works well because these materials hold heat and block air movement. For larger or critical castings, a furnace is better because the temperature can be ramped down at a controlled rate.

Method	Best for	Practical note
Insulating blanket	Small to medium castings	Wrap completely; avoid gaps and drafts
Dry sand or vermiculite	Irregular shapes and small repairs	Bury the whole part, not just the weld
Lime	Shop-floor cooling of compact parts	Keep dry; moisture causes steam and shock
Furnace cooling	Critical or code work	Control the ramp-down rate and record it

Place the part on an insulating surface as well. A cold steel table will wick heat out of the bottom of a casting while the top cools slowly, which produces exactly the uneven contraction you are trying to avoid.

Post-Weld Heat Treatment: Stress Relief vs. Annealing

Slow cooling is the minimum. Post-weld heat treatment (PWHT) is the next level, and it is recommended whenever it is feasible. Stress relieving at about 1150°F (620°C), followed by furnace cooling to at least 700°F (370°C) before the part is exposed to air, relaxes residual stresses and reduces the risk of delayed cracking. TWI states that the only post-weld heat treatment generally applicable after fabrication is a stress relief at 600–620°C followed by slow cooling, which aligns closely with that practice.

Full annealing at about 1650°F (900°C) is sometimes used for maximum softening of the weld zone or more complete stress relief. The trade-off is real: annealing lowers the as-cast tensile strength of all but the softest irons. Do not promise a full-strength restoration after an annealing cycle, and do not anneal a part that carries structural load without engineering approval.

When PWHT Is Not Possible

Many repairs cannot go into a furnace. An engine block still in a frame, a machine base with bearings and seals, or a painted casting with machined surfaces may not tolerate 1150°F. In those cases, use the best available slow-cooling method, document what was done, and tell the customer plainly that the repair has a harder heat-affected zone and a higher risk of future cracking. A written note on the work order protects both of you.

Inspecting the Finished Repair

Inspect after the part has cooled to room temperature, not while it is warm. Start with a visual check for cracks, undercut, and incomplete fusion. Dye penetrant is useful for surface-breaking flaws on non-porous surfaces, while magnetic particle inspection works well on ferromagnetic gray and ductile iron. A hammer ring test can reveal a dull thud where a crack interrupts vibration, but it is a screening tool, not proof. Pressure-containing castings should be hydrostatically or pneumatically tested according to the relevant procedure. AWS D11.2, the guide for welding cast iron, is a sensible reference when the repair needs a formal acceptance basis.

Machining, Hardness, and One Last Check

If the part must be machined after welding, check hardness near the weld before committing a cutter. Nickel-based deposits machine reasonably well, but a hard heat-affected zone can dull tooling and leave a poor surface. A small test coupon welded alongside the repair, using the same filler and cooling practice, is cheap insurance. For stick welding work, the fundamentals in this SMAW guide apply to cast iron repairs too, but the cooling rule is stricter: keep the part warm, cover it, and walk away.

Note: The weld is finished when the casting reaches room temperature without a crack. Until then, the job is still in progress.

Common Mistakes, Troubleshooting, and Safety on Cast Iron Repairs

Almost every failed cast iron repair can be traced back to one of a handful of avoidable errors. The good news is that each error has a signature: a specific way it shows up in the finished weld. Learn to read those signatures and you can diagnose a bad bead in seconds instead of guessing.

The Mistakes That Sink Cast Iron Repairs

Start with the big one: welding a dirty casting. Cast iron is porous and has been soaking up oil, coolant, and paint for decades. Heat drives those contaminants out of the pores and straight into the weld pool, and you get porosity that looks like a sponge. Grind off the casting skin, degrease with a proper solvent, and remember that a crack rarely ends where you can see it.

The rest of the lineup looks like this:

• Wrong filler for the iron type. Nickel-iron works for most gray and ductile repairs, but a steel electrode on a restrained gray iron casting is a crack waiting to happen.
• Too much current. High amperage means deep dilution, more carbon pickup, and a wider brittle heat-affected zone.
• Long continuous beads. Every inch of continuous bead adds contraction stress. Short segments and backstepping spread that stress out.
• No peening. Peening immediately behind the bead, with a blunt tool, relieves the shrinkage stress while the metal is still plastic.
• Quenching or fast cooling. Water, compressed air, or pulling the blanket off early all promote white iron and martensite.
• Skipping preheat on a thick section. Mass wicks heat away faster than the arc can put it in. A 200 lb housing needs preheat; a thin bracket may not.
• Guessing temperature. An IR gun on a rough, oxidized casting can be off by hundreds of degrees. Use a contact thermocouple or temperature crayons.
• Welding a part that should have been brazed or replaced. Bronze braze welding, explained in this guide to air acetylene welding and torch work, often outperforms fusion welding on thin, intricate, or badly contaminated castings.
• Misidentifying white iron. White iron is extremely hard and essentially unweldable. If a file skates off the surface and the part came from a wear application, stop before you strike an arc.

Symptom-to-Cause Table

Symptom	Most Likely Cause	Fix
Porosity, wormholes	Oil, paint, or moisture in the casting; dirty filler	Grind to bare metal, solvent clean, bake out at low heat
Underbead cracking (often hours later)	Insufficient preheat, fast cooling, high restraint	Raise preheat, slow the cooling, use nickel-iron filler
Weld metal cracking	Wrong filler, long beads, no peening	Switch to ENiFe-CI, shorten segments, peen immediately
Hard, chattering machining	White iron or martensite in the HAZ	Higher preheat and slower cooling, or anneal after welding
Spatter, unstable arc	Current too low for the electrode, or contaminated surface	Adjust within the low range, clean the joint
Distortion	Localized heating, no support, continuous beads	Uniform heat, pre-support, backstep and skip weld

Safety: Fume, Heat, and the Invisible-Hot Problem

Cast iron repairs combine two fume hazards. Nickel-bearing electrodes produce nickel fume, and hot castings burn off oil, paint, and old coolant. According to OSHA’s welding safety guidance, prolonged exposure to welding fume may cause lung damage and cancers of the lung, larynx, and urinary tract, and manganese fume can cause Parkinson’s-like symptoms.

Ventilation is not optional. OSHA 1910.252 requires mechanical ventilation when you weld in a space of less than 10,000 cubic feet per welder, in rooms with ceilings under 16 feet, or where partitions obstruct cross ventilation. The minimum general ventilation rate is 2,000 cubic feet per minute per welder unless local exhaust hoods or approved airline respirators are used. Local exhaust hoods must hold a capture velocity of 100 linear feet per minute in the welding zone at their most remote working distance.

For nickel and stainless-bearing filler, add the hexavalent chromium rule to your checklist: OSHA’s permissible exposure limit is 5 micrograms per cubic meter as an 8-hour time-weighted average. In practice that means extraction at the arc, not a fan blowing the plume across the shop.

Then there is the thermal hazard, which catches experienced people. Cast iron at 400–600°F looks essentially the same as cold cast iron. There is no glow, no shimmer, no warning. Treat every recently heated part as hot, mark it clearly, and keep gloves and handling tongs within reach. Blankets, ceramic heaters, and torches all need a fire watch and clearance from combustibles. Wear a face shield when grinding, and never lean on a casting you just pulled from the oven.

If you take away one thing from this section, make it this: the mistakes cluster around heat and cleanliness, and the injuries cluster around heat you cannot see. Control both deliberately, and cast iron repair becomes predictable instead of a coin flip.

A Practical Decision Framework, Key Takeaways, and Final Word

Everything in this guide comes down to a sequence of decisions. Made in order, they turn a frightening repair into a controlled thermal process. Made out of order, they turn a bench vise into scrap.

The Cast Iron Repair Decision Tree

1. Identify the iron. Gray, ductile, malleable, compacted graphite, or white? Spark test, fracture appearance, casting marks, and a magnet tell you plenty before any heat goes on. If it is white iron, stop.
2. Assess the part. Mass, section thickness, restraint, contamination, and what the casting does for a living. A decorative bracket tolerates risk a lifting lug never will.
3. Pick the route. Hot route: high preheat, slow cooling, a ductile heat-affected zone. Cold route: low heat input, a small brittle zone, nickel filler letting ductile weld metal yield instead of the casting.
4. Choose filler and process. ENiFe-CI or ENi-CI for most arc work, bronze for braze welding, steel only when machinability does not matter. Stick is the practical field default, per TWI’s MMA/SMAW guidance on nodular iron.
5. Set preheat and interpass. A workable general starting window for arc welding is roughly 500–1200°F, never exceeding about 1400°F, with the exact figure set by iron type, mass, and restraint. Lincoln Electric’s cast iron guidelines cover the same band, while AWS D11.2 allows a far wider 70–600°F window. Narrow whichever you use to your job.
6. Prepare the joint. Gouge or grind the crack to sound metal, remove casting skin, degrease, and dry. Contamination causes most of the porosity blamed on the rod.
7. Weld in short segments. One-inch beads on thin work, up to three inches on heavy sections, backstepped and skip-welded, stringers only, peening immediately behind the bead with a blunt tool.
8. Cool slowly. Blanket, bury in dry sand or vermiculite, or furnace cool. Never quench, never blow air.
9. Inspect and document. Visual, then dye penetrant or magnetic particle where access allows, plus a ring test on hollow castings.

The One-Page Preheat Checklist

Step	Check
Identification	Iron type confirmed; white iron ruled out; casting marks photographed
Condition	Crack removed to sound metal; oil, paint, and skin ground away; part dry
Setup	Part supported so it can expand; insulation ready; fire watch briefed
Heating	Slow, uniform rise; thick sections soaked; no single-spot torching
Verification	Pyrometer or thermocouple reading near the joint and on the thickest section
Welding	Interpass in range; short segments; peening behind the bead; craters filled
Cooling	Blanketed or buried; no drafts; no quenching
Records	Preheat, interpass, and post-weld temperatures logged with date

Numbers Worth Remembering

• 500–1200°F: general preheat starting window for arc welding cast iron.
• 1400°F: practical ceiling before you risk damaging the casting.
• 800–1200°F: typical oxyacetylene preheat range, per AWS.
• 1150°F: common post-weld stress relief, furnace-cooled to at least 700°F.
• 1650°F: full annealing to soften the weld zone, at the cost of as-cast tensile strength.
• 100°F: target for the cold route — warm to the touch, never hot.

Logging those numbers creates the traceability that separates a professional repair from a guess. Two lines in a notebook beat a confident nod every time.

Revisiting the Vise

The vise failed because heat left the weld too fast. A proper repair would have meant identifying the casting, grinding the crack out to sound metal, preheating the jaw to a temperature matched to its mass and restraint, welding in short nickel-iron beads with peening, then burying the whole thing in dry sand overnight. Instead the weld cooled in open shop air, the heat-affected zone flashed to white iron and martensite, contraction stresses stacked up, and the casting split. Same welder, same rod, same machine — a different thermal story.

When to Walk Away

Some cast irons are effectively unweldable, and for safety-critical or highly restrained parts a replacement fabrication or brazed repair is the responsible choice. Walk away from unknown white iron, from structural or lifting duty without engineering sign-off, from thin high-restraint sections where post-weld heat treatment is impossible, and from any job where new steel is cheaper and safer than the repair. Saying no is a professional answer, not a failure.

The Final Word

Cast iron repair is not really a welding problem. It is a controlled thermal process that happens to involve an arc. Get the heat right going in, get the heat right coming out, and the metal does what you ask. Get it wrong, and no filler in the catalog will save you. If you work other materials too, our guide to the advantages of stick welding covers the process control habits that carry straight over.

Frequently Asked Questions

What temperature should I preheat cast iron to for welding?

A practical starting window for arc welding is about 500-1200°F, depending on the iron type, casting mass, section thickness, and restraint. Do not exceed roughly 1400°F, since that enters the critical temperature range where cracking conditions develop. Always verify with a pyrometer or temperature-indicating crayon rather than judging by color.

Can you weld cast iron without preheating it?

Yes, but only with the right technique and expectations. Using nickel-base filler, low current, short 1-inch beads, immediate peening, and slow air cooling keeps the brittle heat-affected zone small so the ductile weld can yield. The heat-affected zone will still be hard and brittle, so this route suits low-stress repairs rather than safety-critical parts.

Do I need to preheat ductile or nodular iron the same way as gray iron?

Not identically. Ductile iron is more forgiving because its graphite is nodular rather than flake form, but thick or complex ferritic grades still benefit from roughly 100-200°C preheat. Pearlitic, normalized, or hardened-and-tempered grades may need at least 250°C, and highly restrained joints can call for 320-330°C.

Why does my cast iron weld keep cracking after I finish?

Cracking after welding usually comes from cooling too fast or from restraint. Rapid cooling forms hard white iron and martensite in the heat-affected zone, and contraction stresses then tear it. Slow the whole cycle down: adequate preheat, maintained interpass temperature, short peened beads, and cooling under a blanket or in dry sand.

What filler metal should I use for preheated cast iron?

Nickel and nickel-iron electrodes (ENi-CI and ENiFe-CI) are the common choice because they deposit ductile, machinable weld metal that tolerates carbon pickup. Cast iron (ECi) electrodes match the base metal best but generally require furnace heating or a very high red-hot preheat. Steel electrodes are cheaper and stronger but give a hard, non-machinable weld.

How long should cast iron cool after welding?

Cool it as slowly as practically possible, ideally in still air under an insulating blanket, in dry sand, or in vermiculite. Do not uncover the part until it is below roughly 60°C (140°F), and never quench it or blow compressed air on it. When a furnace is available, controlled furnace cooling after stress relief is the best option.

How do I know if a casting is gray iron, ductile iron, or white iron?

Look for casting marks and specifications such as ASTM A48 for gray iron or ASTM A536 for ductile iron, and check whether the part is magnetic and how a fresh fracture looks. White iron is extremely hard and essentially unweldable, so a file or hardness test on an inconspicuous spot is a worthwhile check before committing to a repair.

Is it safe to preheat a casting with an oil or paint contamination?

Not without care. Heating oily or painted castings drives off fumes and can ignite residue, and some coatings contain hazardous metals. Strip coatings, degrease thoroughly, and work with ventilation or local exhaust; OSHA requires mechanical ventilation in small spaces, low-ceiling rooms, and obstructed areas.

When should I refuse to weld a cast iron part?

Walk away from suspected white iron, unknown alloys, safety-critical castings with no qualified procedure, and highly restrained thin sections where post-weld heat treatment is impossible. In those cases brazing, metal stitching, or fabricating a steel replacement is usually cheaper and far safer than a weld that fails in service.