Welding SG Iron: The Complete Guide to Ductile Iron Welds

A pump housing cracked at the flange root on a Friday, and the shop’s first instinct was to dial up the amperage and run it like mild steel — the crack that appeared an hour later was twice as long as the original. SG iron rewards patience and punishes shortcuts, and this guide explains exactly why.

What SG Iron Is and Why Welders Keep Running Into It

The phone call usually sounds the same. A pump housing has cracked at the flange root overnight, a gearbox mounting lug has snapped off on a site machine, or a worn bore needs building back up so the part can be machined to size again. The OEM quote is four figures, the lead time is measured in months, and the casting is sitting on your bench right now. Before you strike an arc, you need to know what you are actually holding.

So What Is SG Iron, Exactly?

SG iron is cast iron — not steel, not a clever alloy pretending to be steel — but cast iron that has been treated so its graphite forms as tiny spheres instead of the sharp flakes found in ordinary grey iron. “SG” stands for spheroidal graphite, and you will also hear the same material called nodular iron or ductile iron. All three names mean the same thing.

The governing definition comes from ASTM A536, which describes it as cast iron in which the graphite is substantially spheroidal in shape and essentially free of other graphite forms. That seemingly dry wording is the whole point of the material. When molten iron is treated with magnesium or a cerium alloy before casting, the graphite precipitates as nodules rather than flakes. Those nodules act like tiny crack arresters: a crack running through the metal hits a nodule and blunts instead of propagating. As TWI explains, this nodular graphite gives SG iron significantly higher ductility than flake-graphite grey iron, and that ductility is exactly what improves its weldability.

In plain language: grey iron snaps like a biscuit, SG iron bends and stretches before it lets go. On a service ticket, that difference shows up as tensile strengths commonly in the 400 to 650 MPa range and elongations from around 2% up to 18% depending on grade — figures that grey iron cannot come close to.

Why It Keeps Landing on Your Bench

SG iron is one of the most common engineering castings in the modern world, which is why it turns up in nearly every branch of repair work:

• Pipework and fittings — flanged bodies, valve bodies and large-bore fittings where pressure tightness and strength both matter.
• Automotive castings — engine blocks, cylinder heads, steering knuckles, suspension arms and brake calipers. Many modern vehicles use SG iron specifically where grey iron would be too brittle.
• Pump and compressor housings — probably the single most frequent source of cracked-casting repair jobs, with flange root cracks and fatigue cracks around the volute.
• Gearboxes and drive housings — including worn bearing bores that need building up and re-machining.
• Agricultural and construction equipment — tractor housings, loader components and earthmoving machinery that has been overloaded one time too many.
• Wind and water infrastructure — wind turbine hub and gearbox castings, plus large municipal water valves and hydrants, where a replacement part can have a lead time of a year or more.

The typical failures that send welders into an SG casting are depressingly consistent: flange root cracks from bending loads, broken lifting lugs dropped or overloaded during handling, worn or out-of-round bores from long service, and fatigue cracks that have grown slowly around a bolt hole or rib junction. Each of these is a legitimate candidate for weld repair when the rest of the casting is sound.

What This Guide Covers

Understanding why SG iron behaves the way it does under the arc is what separates a repair that lasts from one that fails. Section by section, we will work through the metallurgy behind grades and microstructure, the four mechanisms that make SG iron welds crack, filler metal selection from pure nickel through nickel-iron and matching SG wire, preheat and interpass temperature numbers you can act on immediately, process selection across stick, TIG, MIG, flux-cored and braze welding, and a full step-by-step repair sequence using a cracked pump housing as the worked example. We will also cover defect diagnosis, the relevant codes and qualification requirements, and the honest question of when welding is the wrong answer and replacement or mechanical repair is safer.

Before you go further, it is worth remembering that welding skill transfers. If you are comfortable running stick electrode work on positional joints, you already have half of what SG iron repair demands — electrode control, short beads and disciplined interpass timing. What changes with SG iron is the thermal thinking, not the hand skills.

SG Iron Metallurgy: Grades, Microstructure and Why It Behaves Differently

“SG iron” covers a surprisingly wide range of materials, and two castings that look identical on the bench can behave completely differently under the arc. One runs happily with light preheat and a nickel rod; the other cracks if you look at it wrong. That difference is metallurgy, and it is readable straight off the drawing.

How SG Iron Is Made: Spheres Instead of Flakes

Grey cast iron and SG iron start from nearly the same melt. According to TWI, molten iron in the range of 3.2–4.5% carbon and 1.8–2.8% silicon is treated with magnesium or cerium additions just before casting. That inoculation step changes the shape of the graphite as the metal solidifies: instead of sharp, continuous flakes that slice through the matrix and make grey iron brittle, the graphite forms rounded nodules.

The nodules matter because they do not act as stress concentrators. Load is carried by the surrounding iron matrix, which is why nodular (ductile) irons combine greater ductility with higher tensile strength than grey cast irons, with matrices available as pearlite, ferrite, or a pearlite-ferrite mix.

The Matrix Decides How You Weld It

This is the part welders skip, and it is the part that matters most. Graphite shape tells you the iron is ductile; the matrix around those nodules tells you how the heat-affected zone will respond to a thermal cycle.

• Ferritic matrix: soft, ductile, the most forgiving under the arc. Ferritic grades tolerate lower preheat, accept a wider heat-input window and machine more easily after welding.
• Pearlitic matrix: stronger and harder, but far more sensitive. Rapid cooling can turn the HAZ hard and brittle, which is exactly where cold cracks appear days later.
• Mixed pearlite-ferrite: the common middle ground, behaving between the two, with preheat decisions that follow the harder phase.

Preheat ranges, interpass limits, filler choice and cooling rate are all driven by how readily the matrix transforms, not by the grade’s tensile number. A high-strength pearlitic casting is, thermally speaking, closer to a low-alloy steel than to a ferritic iron. That is a useful mental shortcut, and one reason electrode selection on these grades deserves the same care covered in our alloy steel welding electrodes guide.

Grade Designations: ASTM A536 and EN 1563

Two designation systems show up on drawings. Both are easier to read than they look.

ASTM A536 uses a three-number code: tensile strength in ksi, yield strength in ksi, then elongation in percent. The range runs from 60-40-18 (soft and ductile) up to 120-90-02 (strong, stiff, barely elongating). The higher you climb, the more pearlitic the matrix and the less forgiving the grade during welding.

EN 1563 grades use the EN-GJS prefix followed by tensile strength and elongation as minimums. EN-GJS-400-15 (the former GGG40) shows minimum tensile strength of 400 MPa, yield of 250 MPa, elongation of 15%, and hardness around 130–180 HBS, with a mainly ferritic microstructure.

Grade	Min. tensile	Min. elongation	Typical matrix	Welding character
EN-GJS-400-15 (GGG40)	400 MPa	15%	Mostly ferritic	Most forgiving
ASTM A536 60-40-18	60 ksi (414 MPa)	18%	Ferritic	Wide preheat window
EN-GJS-400-18-LT	400 MPa	18%	Ferritic (low temperature)	Forgiving, toughness grade
ASTM A536 80-55-06	80 ksi (552 MPa)	6%	Pearlite-ferrite	Moderate, control cooling
EN-GJS-700-2	700 MPa	2%	Pearlitic	Poorly forgiving
ASTM A536 120-90-02	120 ksi (827 MPa)	2%	Fully pearlitic	Highest crack risk

Read the elongation column against the last one and the pattern is obvious: ductility in the base metal buys forgiveness under the arc. For the full base-metal specification, see ASTM A536 and the EN equivalent data for grades such as EN-GJS-400.

Identifying the Grade on the Shop Floor

In practice, you identify the grade in this order:

1. The drawing or repair order: most authoritative, and it usually carries the specification and grade callout.
2. Cast markings: heat numbers, foundry logos and grade stamps raised on the body of the casting. Under a wire wheel they usually become legible.
3. The material certificate: if the part came with an EN 10204 3.1 test report, it lists chemistry, mechanical properties and often the nodularity result.
4. A hardness check: a quick portable hardness reading separates soft ferritic stock from hard pearlitic grades when markings are unreadable.

Understanding the matrix is only half the job. The next section explains exactly how heat ruins a good casting, and why those failure mechanisms are the reason every rule in this guide exists.

Why Welding SG Iron Cracks: The Four Failure Mechanisms

SG iron cuts like mild steel, drills like mild steel, and looks forgiving under the hood. Then the arc passes, the metal cools, and a crack appears where nothing was wrong five minutes earlier. That is not bad luck. It is four separate metallurgical mechanisms acting at once, and every rule you have heard about welding SG iron — preheat, slow cool, keep the beads small, do not gouge hot — exists to defeat one of them. Learn the mechanisms instead of memorising the rules, and the rules start making sense.

The starting point is a fact the American Welding Society puts plainly: cast iron is considerably less weldable than steel, because its much higher carbon and silicon content makes it both less ductile and more metallurgically altered when heated. The AWS D11.89 casting-welding guidance (now folded into AWS D11.2) is built around that single sentence. Four mechanisms follow from it.

Mechanism 1: White Iron Formation in the HAZ

Arc heat pushes the graphite near the fusion line above the eutectoid temperature, and the nodules begin dissolving back into solution. The matrix around them becomes high-carbon austenite. If cooling is fast, that carbon cannot re-precipitate as graphite — there is no time and no diffusion path — so it locks in as iron carbide (cementite). The result is a hard, brittle layer of so-called white cast iron sitting right in the heat-affected zone, sometimes a fraction of a millimetre thick, sometimes several millimetres deep depending on heat input and section size.

White iron has almost no ductility. It cannot relieve stress by deforming, so it relieves stress by cracking. This layer is why a weld can look perfect on the surface and still split hours later along the HAZ.

Mechanism 2: Martensite from Rapid Cooling

Below the white-iron layer, the original matrix — ferrite or pearlite in a normal SG casting — also transforms. Rapid cooling through the critical range traps carbon in a body-centred tetragonal lattice: martensite. TWI’s weldability notes on cast irons describe exactly this: fast cooling forms a hard, brittle white-iron and martensitic layer in the HAZ, and preheat works by slowing the cooling rate, suppressing martensite formation and lowering HAZ hardness.

Pearlitic grades (think ASTM A536 80-55-06 or 100-70-03) are the most exposed here, because their matrices already hold carbon in a hard form. Ferritic grades like 65-45-12 and EN-GJS-500-7 give you more margin, but only more margin — not immunity.

Mechanism 3: Hydrogen Embrittlement and Residual Shrinkage Stress

Two actors, one accomplice. The arc atmosphere brings hydrogen in from moisture in electrodes, coatings, oils and humidity. Hydrogen dissolves into the hot metal, diffuses inward as the weld cools, and collects at stress raisers and grain boundaries. Meanwhile, weld metal contracts as it solidifies, and the stiff casting around it refuses to move. The result is high residual tensile stress sitting in a zone that is now hard, brittle and hydrogen-saturated — the textbook recipe for delayed cold cracking.

This is why cracking often happens an hour after the weld, not during it, and why preheating is prescribed not just to control hardness but to dissipate shrinkage stresses and reduce distortion, lessening the likelihood of cracking. Baking electrodes, controlling interpass temperature and slow cooling all target this same mechanism.

Mechanism 4: Carbon Pick-Up (Dilution)

During the weld, carbon from the melted base metal mixes into the weld pool. Even with a nickel or nickel-iron filler, too much dilution pushes the deposit toward a brittle, high-carbon composition — sometimes hard enough to machine like glass, sometimes hard enough to crack. Filler suppliers and the AWS D11.2 guide both recommend controlling dilution and using ductile nickel-iron deposits that can absorb shrinkage stresses without cold cracking.

Carbon pick-up is also why carbon-arc gouging is handled carefully on cast iron: the gouging arc reintroduces carbon and can deepen the white-iron layer you were trying to remove.

The Three Levers You Actually Control

Every mechanism above responds to the same three variables:

• Heat input and amperage — too hot melts more base metal (more dilution), too cold causes lack of fusion and forces you to compensate with extra passes.
• Interpass temperature — holding 150–250 °C between beads keeps the cooling rate slow, suppresses martensite and white iron, and keeps the whole casting out of the danger zone.
• Cooling rate after welding — bury the part in dry sand, cover with a ceramic blanket, and let it fall to room temperature over hours rather than minutes.

One useful edge case: Ni-Hard martensitic and white cast irons are considered generally unweldable because they are too crack-sensitive even with these controls. If you cannot identify the casting and it machines like carbide, that is your answer — braze weld or replace instead.

With the four mechanisms in mind, the next decision is far simpler: which filler metal gives the weld pool the ductility to absorb the stress these mechanisms inevitably create.

Choosing the Right Filler Metal for SG Iron

Pick the wrong rod and no amount of careful preheat will save an SG iron repair. Filler selection sets the ceiling on strength, ductility, machinability and crack resistance. The governing specification is AWS A5.15, Specification for Welding Electrodes and Rods for Cast Iron, which uses a simple naming system: E for electrode, R for rod, Ni or Fe for the alloy system, CI for cast iron. ENiFe-CI is therefore a nickel-iron electrode for cast iron, and RZC a bronze oxy-fuel rod. Learn the letters first.

The Five Families on the Shelf

Filler family	Typical deposit	Strength	Machinability	Best fit
Pure nickel (ENi-CI)	92–97% Ni	Lowest, roughly 240 MPa class	Excellent	Thin grey iron, cosmetic work
Nickel-iron (ENiFe-CI)	Fe-55% Ni balance	~400 MPa tensile, 10–12% elongation	Good with sharp tooling	SG grades, dissimilar joints, heavy repairs
Nickel-copper (ENiCu)	Monel-type, ~55% Ni / ~30% Cu	Moderate, corrosion resistant	Good	Grey iron, non-pressure corrosion service
Mild steel (E70S-6 class)	Carbon steel	High, brittle on dilution	Poor to none	Non-machinable, non-pressure repairs
Stainless (309L/308L)	Austenitic stainless	High	Poor	Niche, only with controlled dilution

Pure Nickel ENi-CI: The Easy-Machining Option

ENi-CI deposits are soft, usually well under 200 HV, and machine almost like parent grey iron, so they suit pump bodies, gearbox covers and blocks with a machined sealing face. The trade-off is strength: specification minimums sit in the 240 MPa (35 ksi) class, far below SG iron’s 400 MPa-plus grades. That is fine in thin grey iron, unacceptable as a structural repair in a heavy nodular casting. Nickel-rich fillers also minimise dilution, as TWI’s cast iron weldability notes explain, because little carbon migrates into a deposit that already has nowhere to put it.

Nickel-Iron ENiFe-CI: The Workhorse

If you keep one consumable for ductile iron, make it ENiFe-CI (the Fe-55 nickel type). A typical deposit lands around 400 MPa tensile, roughly 230 MPa proof stress, with 10–12% elongation and hardness near 170–200 HV. That is close enough to SG grades from ASTM A536 that the weld stops being the weak link, while still machining acceptably. Nickel-iron outperforms pure nickel on strength and suits dissimilar joints to mild steel and pearlitic irons. Most come in all-position profiles, which matters when the crack sits overhead; our all-position welding rod guide covers the technique side. Expect work-hardening: sharp cutters, light cuts, plenty of coolant.

Nickel-Copper, Mild Steel and Stainless

ENiCu (Monel-type) fillers machine nicely and resist corrosion, but they carry more contraction than nickel types and are more prone to hot tearing in restrained sections, so reserve them for grey iron and non-pressure corrosion work. E70S-6 is cheap and strong, but the weld metal and its HAZ pick up carbon and go hard and unmachinable, limiting it to low-duty, non-machinable, non-pressure repairs peened while hot. Stainless (309L) bridges to steel parts, but dilution into a carbon-rich pool can form martensite and crack. For a steel-to-cast-iron transition, our alloy steel welding electrode selection guide covers the steel side.

Matching SG Wire and Cored Options

Matching nodular iron wire and nickel-iron cored wires exist for GMAW/FCAW production repair with high deposition rates. The catch: a matching SG wire behaves like a steel filler in a hardenable casting, demanding preheat, strict interpass control, slow cooling and more carbon pick-up from the base. Most shops get better results with nickel-iron cored wire in spray or pulsed mode on long runs, and solid ENiFe-CI stick electrodes for one-off field work. If you already run flux-cored processes on other materials, the parameter discipline carries over; the consumable does not.

Buttering: The Technique That Makes Any Filler Work

Buttering is laying two or three thin, low-heat-input beads onto the prepared groove faces before you fill, creating a soft buffer that shields the base metal from the main weld’s dilution and shrinkage. For SG grades, keep buttering heat input low: small-diameter electrodes, stringer beads, no weaving, alternating sides, interpass below about 150 °C. Peen each pass lightly while hot to relieve shrinkage stress, then fill with your nickel-iron electrode, still stringering and alternating. It is the single biggest factor in avoiding re-austenitised hard zones along the fusion line, and the practice AWS D11.2/D11.2M recommends for iron casting repairs. One prerequisite: clean, chip-free metal with no graphite smear. If carbon arc gouging was used, grind back that layer until sound metal is exposed; the carbon-rich skin it leaves will poison the buttering bead.

The practical rule: match the filler to duty, not habit. Machined surface, pressure duty and nodular base metal point at ENiFe-CI with buttering every time. Thin grey iron, cosmetic work and no structural load point at pure nickel. Anything else is a deliberate compromise, documented against the filler specification on the welding procedure.

Preheat, Interpass Temperature and Heat Input: The Numbers That Matter

You can get away with the wrong filler now and then, and a sloppy bead sequence once in a while. Get the thermal numbers wrong on SG iron and no amount of technique will rescue the casting. Preheat, interpass temperature and heat input decide whether the heat-affected zone (HAZ) ends up soft and machinable or hard, brittle and cracked an hour after you put the stinger down.

Preheat Windows: The Numbers Welders Actually Use

The most useful single reference is TWI’s weldability guidance on cast irons, which lists typical preheat values for manual metal arc and MIG welding of nodular (SG) iron. The split is dramatic, and it tracks the matrix rather than anything about the joint:

Cast iron type	Typical preheat (MMA / MIG)	Why
Ferritic nodular (SG) iron	Room temperature to 150 °C	Ferrite has almost no carbon to trap; slow cooling matters less
Pearlitic nodular (SG) iron	200–330 °C	Pearlite plus rapid cooling breeds martensite in the HAZ
Oxy-acetylene (gas fusion), nodular grades	Around 600 °C	The cooler heat source needs a hotter start to avoid a steep gradient

That last row surprises people. Gas welding of cast iron is often sold as the “gentle” option, and the flame is indeed softer than an arc. But a torch delivers heat at a far lower temperature than a 2,000 °C arc, so the surrounding metal stays cold relative to the puddle and the casting has to start hot to keep the cooling rate sane. Found white iron in the next day’s machining pass after an oxy-fuel repair? That is why.

AWS’s Guide for Welding Iron Castings carries its own table of recommended preheat and interpass temperatures, plus electrode manipulation guidance, and it is the document most shops cite when a customer asks for a qualified procedure (AWS D11.2/D11.2M, which carries forward the older D11.89 content covered here by ANSI). TWI gives you typical values; D11.2 gives you values written to be quoted on a procedure sheet. Check both, then defer to the customer’s spec when it is stricter.

Gradual vs. Local Preheat

How you reach temperature matters as much as the target. TWI is explicit: preheat should always be applied gradually, because unequal expansion during localised heating of a complex casting can itself cause cracking. A torch waved at one flange of a pump body makes a hot spot against cold metal, and that gradient alone can open a crack before you strike an arc.

Preferred methods, in rough order of control:

• Furnace or oven preheat — bring the whole casting up at 100–150 °C per hour, soak, and weld from the oven if it fits. Best uniformity, best result.
• Ceramic pad or blanket heaters — good for large castings that will not fit an oven; leave them running during welding to hold interpass temperature.
• Local torch preheat — fine for short, straight cracks in simple geometry, but heat a wide band, not a spot, and verify with a contact or infrared thermometer rather than by colour.

Interpass Control and Heat Input

Interpass temperature is the temperature of the previous bead when you start the next one. Never let it fall below your minimum preheat, and never let it climb far above. Excess interpass temperature is simply extended time in the dangerous zone: long periods near the austenitising range grow a coarse, hard HAZ. Keep beads small, stringer rather than weave, and let the casting sit between passes instead of racing to fill the groove. Low heat input, multiple short passes and a short arc all narrow the HAZ and limit how much graphite dissolves into the pool. Dilution is the enemy; heat input is the throttle.

If you gouge a crack out with carbon arc first, remember that the air stream cools the surface violently and can chill-crack a preheated casting. Re-preheat after gouging, and review Air Arc Welding and Carbon Arc Gouging: Process, Safety, Applications, and Best Practices for its pitfalls. Same logic for grinding the groove: do it warm, or re-heat afterwards.

Peening: Useful Tool, Easy to Overdo

Light peening of each bead while it is still hot (roughly 200–300 °C surface) relieves shrinkage stress in a material that cannot yield much on its own. Peening the final cap is a mistake: it work-hardens the surface, can start cracks, and on thin sections just stretches the metal. Peen lightly with a blunt or rounded tool, and never peen a bead that has dropped below the preheat minimum.

Post-Weld Cooling and Insulation

The job is not finished when the last bead is capped. Cool slowly: a dry ceramic blanket, a vermiculite wrap, or dry sand all work by keeping the cooling rate below the threshold where martensite and retained austenite become problems. Slow cooling also gives hydrogen time to diffuse out before stresses peak, which matters for hydrogen-assisted cracking. Resist the fan and the water spray; if the casting must be moved, move it still wrapped.

For a ferritic SG bracket, room-temperature welding with 100–150 °C of local warmth and a blanket afterwards is usually enough. For a pearlitic pump housing, expect 250 °C preheat held between passes, then a slow insulated cool and ideally a stress-relief soak. Those numbers are not suggestions to be trimmed for convenience. They are the difference between a repair that outlasts the machine and one that comes back through the door two weeks later.

Welding Processes for SG Iron: SMAW, GMAW, GTAW, FCAW and Braze Welding

Once the filler is chosen, the next question is which arc process to strike. SG iron is not welded with some exotic, single-vendor technique. TWI’s technical overview of cast iron weldability lists oxy-acetylene, manual metal arc (MMA), MIG and flux-cored arc welding as usable on cast irons, and all of them carry the same three conditions: low heat input, generous preheat and slow cooling. Change the process and you change how easily those conditions are to hold, not whether they apply.

Stick (SMAW): the forgiving default

Shielded metal arc welding remains the most common choice for cast iron repair, especially in the field. No gas bottle to drag across a plant floor, the electrode flux gives a small insulating blanket over the bead, and the welder controls heat input through rod diameter, amperage and travel speed. It is also the easiest process on which to butter a groove: small-diameter 2.4 or 3 mm nickel or Monel electrodes at modest current lay down a soft, ductile layer that shields later passes from carbon pick-up. For anyone building confidence, our advantages of stick welding guide covers why SMAW still wins in awkward positions.

The trade-off is deposition rate and the stop-start rhythm of rod changes, which lets the part cool between passes. On thick castings that is a hazard, so hold the interpass temperature inside your preheat window and plan the bead sequence before you strike an arc.

GTAW (TIG): precision on thin sections

TIG gives the finest heat control of any arc process, which matters on thin-walled housings, instrument bodies and small brackets where a stick arc would blow through. The concentrated, adjustable arc lets you hold a small puddle and back off instantly. TIG is used for buttering, for closing the final cap pass, and for repairing cracks near machined surfaces where dilution and distortion must be minimised. It is slower, demands cleaner preparation, and gas coverage cools the weld zone quickly, so preheat and post-arc insulation are still required.

GMAW and FCAW: production and high-deposition repair

MIG suits repeatable shop repairs where jigs and fixtures hold the part. The standard recipe is a neutral gas such as 100% argon with DC reverse polarity giving the best results, small-diameter wire, and short-circuit or pulsed transfer to keep heat input down. Pulsed current is a genuine advantage: it holds a stable transfer at lower average current, shrinking the heat-affected zone. Flux-cored wire adds a slag blanket that slows cooling and suits heavier sections, though self-shielded FCAW runs hotter than most cast iron repairs tolerate.

Process	Where it fits	Watch-out
SMAW	Field repair, thick castings, all-position	Cooling between rod changes
GTAW	Thin sections, buttering, cap passes	Slow; tight cleaning and preheat
GMAW	Production repair, jigged parts	Dilution and heat creep on long runs
FCAW	Heavy sections needing deposition and slag	Self-shielded variants run too hot
Oxy-acetylene powder	Slow repairs on heavy grey/SG castings	Widest HAZ; needs skilled flame control
Braze welding (bronze)	Non-structural build-up, water-facing surfaces	Lower strength; not for loaded welds

Oxy-acetylene powder welding and bronze braze welding

The oxy-acetylene process, often with cast iron powder added to the puddle, is the oldest cast iron technique still in use. A neutral flame with a large, soft heat zone and very slow travel keeps carbon in solution, and the process suits large castings that can be preheated uniformly. It is slow and heavily dependent on operator feel, which is why it has retreated to specialist shops.

Braze welding with bronze filler is a legitimate lower-heat alternative for non-structural repairs because it never melts the parent casting: the filler wets the surface metallurgically rather than fusing with it, so white iron and martensite never form. The joint is softer and weaker than a nickel weld, so it suits build-up, cosmetic restoration, water pump bodies in service and parts where machinability matters more than tensile strength. For torch technique and safety, our air acetylene welding resource covers the fundamentals.

Niche approaches: studs, friction and gouging pitfalls

Where a full groove weld is unnecessary, threaded stud welding, pin repair and friction welding can reattach brackets or replace broken lugs without putting a long fusion line through the casting. Friction welding produces a narrow, rapidly cooled joint and is used for SG iron shafts and automotive components, though it needs purpose-built equipment.

The selection logic is simple: SMAW for the field and the unknown, GTAW for thin and critical, GMAW for repeatable shop work, FCAW for mass, oxy-acetylene for slow heavy sections, and bronze braze welding when the joint does not carry load. Whatever you pick, the process only buys you the chance to control heat; preheat, interpass temperature and slow cooling still decide whether the repair survives.

Step-by-Step: Repair Welding a Cracked SG Iron Casting

Everything that decides whether an SG iron repair survives happens before the arc is struck. The sequence below is the shop-floor order, and it follows the logic in AWS D11.2/D11.2M, Guide for Welding Iron Castings: find the full extent of the damage, remove all of it, control the heat, then rebuild in small, stress-relieved increments.

Step 1: Trace and Arrest the Cracks

Do not trust what you can see. Clean the area and run a dye penetrant over the whole suspected region, because SG iron cracks often branch or run further than the surface suggests. Once the full length is known, grind or stop-drill 3–6 mm holes at each crack tip so it sits inside sound metal. Welding over an un-arrested crack guarantees re-propagation; the residual shrinkage stress will simply pick up where the original crack left off.

Step 2: Excavate the Defective Metal

Remove the cracked volume by grinding or with a tungsten carbide burr, cutting a clean V-groove (roughly 60–70°) down to bright, crack-free metal on every face. Grinding also strips the graphite-rich smut layer that would make the weld metal spongy and unmachinable.

If arc gouging is unavoidable, preheat the local area to about 300 °C first, then gouge and re-grind to remove the affected surface layer (TWI guidance on cast iron weldability). The rapid heating of an air arc can otherwise throw a hard, white-iron skin over the groove, which you then weld straight into. Our Air Arc Welding and Carbon Arc Gouging guide covers the technique and its limits.

Step 3: Clean and Preheat Gradually

Wire-brush or solvent-wipe the groove until no oil, paint or graphite dust remains, then preheat slowly and uniformly. Ferritic grades often accept 100–150 °C locally; pearlitic grades and heavy sections need more, sometimes 250–300 °C over a wider zone. Heat with an oxy-acetylene torch on a sweeping path or put the casting in a preheat oven, and check with temperature-indicating crayons rather than guesswork. The aim is a gentle ramp, not a hot spot.

Step 4: Butter the Groove Surfaces

Run a thin root layer of pure nickel (ENi-CI) or nickel-copper (Monel) stick electrode over the groove walls at modest amperage. This buttering layer acts as a carbon barrier: it limits carbon pick-up from the base metal into the weld pool, which is what causes brittle, carbide-rich weld metal. Chip the slag completely after every pass before another bead goes on top.

Step 5: Fill with Small Stringer Beads

Switch to 3–4 mm nickel or nickel-iron electrodes and fill with short stringer beads, using back-step sequencing so each bead shrinks against a rigid neighbour rather than against the whole casting. Wide weave beads are the classic mistake: they dump too much heat into one place, widen the heat-affected zone and invite a coarse white-iron structure. Keep the interpass temperature inside your preheat window and stick to the manufacturer’s published current range. Peen each bead firmly while it is still hot: peening stretches the bead plastically and offsets shrinkage stress while there is still time. Cold, it does nothing and can crack the weld face.

Step 6: Inspect and Cool Slowly

After the final bead, remove all slag and spatter, dress the surface and re-run penetrant over the weld and HAZ. Bury the casting in dry sand, wrap it in a ceramic fibre blanket or cover it with vermiculite so it cools over hours, not minutes. A casting left on a cold steel bench in a draught often cracks at the toe line after the welder has gone home.

A Worked Example: Cracked Pump Housing

A 90 mm crack at a flange root in a ferritic EN-GJS housing: penetrant first to confirm the ends, 6 mm stop-drills, carbide-burr excavation of a 70° groove, local preheat to 150 °C checked with crayon, ENi-CI buttering, slag removal after every pass, ENiFe-CI stringer fill with peening of each hot bead, a second penetrant check, then a dry sand blanket for slow cooling. Preparation takes longer than the welding, which is normal for stick welding repairs on castings.

When to Stop and Consider Replacement

Some castings should not be saved. Stop and specify replacement when cracks keep reappearing after a sound repair, when the crack crosses a pressure boundary or a fatigue-critical fillet, when the section has thinned from previous gouging, or when the defect sits in a machined fit surface you cannot rebuild without losing tolerance. A repaired joint also rarely regains the fatigue life of the original casting, a point reinforced by Fraunhofer LBF’s NodularWeld research into repair welding of nodular cast iron. A new casting costs money; a failed repair on a pressure housing costs far more.

Common Defects and Troubleshooting SG Iron Welds

Even with the right filler and a controlled preheat, SG iron can hand you a bad day at the bench. The useful part is that most ductile iron weld defects are repeatable and readable: where the crack sits, how the bead surface looks, and how the metal behaves under the cutter all point back to a specific cause. Name the failure and the fix is usually a parameter change, not a change of technique.

First, Read the Crack: HAZ or Weld Metal?

The fastest diagnostic step is looking at where the crack sits relative to the fusion line.

• Cracking parallel to the fusion line, a millimetre or two into the parent metal, is almost always a hard, brittle heat-affected zone layer. It points to too little preheat, an interpass temperature allowed to drift, or a cooling rate that ran away from you. This is the classic cold crack, and it often appears hours after cooling, sometimes with an audible click.
• Cracking down the centre of the bead is a weld-metal problem: shrinkage restraint, too large a bead, or a filler that solidified with too little ductility to absorb the contraction.
• Cracking at the toe or crater is a restart, crater or profile problem, usually fixed by filling craters and washing the toe out rather than changing preheat.

Because hydrogen-assisted cracking in the HAZ can be delayed, TWI’s weldability notes on cast irons recommend leaving welded castings under observation and, where the joint is critical, holding them warm before any load or machining is applied.

Defect, Cause and Fix: A Shop Reference

Symptom	Most likely cause	Fix
Cold crack in the HAZ, parallel to the fusion line	Insufficient preheat, low interpass temperature, or fast cooling on a thick section	Raise preheat toward 200–315 °C, hold interpass minimums, slow-cool under insulation
Hard, unmachinable band along the weld toe	Peak temperature just into the austenite range followed by rapid cooling, forming martensite	Soften with a controlled post-weld thermal cycle or butter the toe; grind and re-inspect
Blown or spongy pores in the weld metal	Contaminated base metal, oil or moisture, or deposition at too low a temperature	Clean thoroughly, verify preheat (at least 200 °C, commonly 315 °C), slow the cooling slightly
Weld metal that will not machine	Excessive carbon pick-up from high dilution, or pure-nickel filler on a high-strength grade	Reduce dilution with stringer beads, switch to a nickel-iron filler such as ENiFe-CI
Graphite degradation at the bond line	Excessive heat input driving carbon into solution	Trim current and travel speed, keep beads small, avoid repeated re-passes
Cracks appearing after machining	Residual stress released when material was removed, or a brittle HAZ left unaddressed	Slow-cool and stress-relieve before cutting; machine lightly and re-penetrant-check

Porosity: The Most Common Cosmetic Failure

Grey and SG cast irons are both susceptible to porosity in the weld metal zone. Slower cooling gives dissolved gases time to escape before solidification, which is why higher line energy and adequate preheat reduce both pores and cooling cracks. The usual culprits are insufficient preheat, a contaminated surface, or a deposition temperature that is too cold, not bad electrode handling alone. If you are chasing bubbles, check the plate temperature with a thermometer before blaming the rod. For contaminated surfaces, the cleaning discipline is the same one used before aluminum cleaner for welding: remove oil, paint and oxide completely, because trapped coating burns straight into the puddle.

Spongy, Unmachinable and Late Cracks

Spongy weld metal that skates a drill bit usually means the deposit picked up too much carbon through high dilution, which is common when wide, hot passes are run to fill fast. Keep beads narrow, keep dilution low, and check filler choice against the base grade: alloy steel welding electrodes and other filler families behave very differently from nickel-irons here, and a pure-nickel rod on a high-strength ductile iron gives machinability but not strength.

Post-machining cracks deserve special suspicion. If a weld passes visual and penetrant checks and then cracks when the cutter engages, residual stress rather than an obvious defect is usually responsible, and the remedy is thermal: proper slow cooling under insulation and, where the component is critical, a post-weld heat treatment cycle described in AWS D11.2/D11.2M.

Two rules of thumb help with excavation. Gouging out a crack with a carbon arc generates intense local heat, so follow the air arc welding and carbon arc gouging practice of short, controlled passes and re-check the groove with dye penetrant afterward, because a crack tip left behind is a crack that returns. And where repair is being specified for a critical casting, AWS D11.89, now carried within D11.2, is the document that tells you which defects are acceptable and which must be removed.

Fraunhofer LBF’s NODULARWELD research is a useful reminder that graphite shape changes how cracks propagate in repaired ductile iron, so a weld that looks acceptable on a grey iron casting is not automatically sound on an SG grade. When in doubt, inspect the HAZ first, then the bead, then the geometry.

Codes, Standards, Qualification and Inspection for SG Iron Welds

Shop-floor cast iron repair has always lived in a gray area: everybody does it, almost nobody qualifies it. That changes the moment the casting belongs to a pressure system, a drivetrain, or anything a customer can sue over. At that point the welder’s skill is not the deliverable, a documented, repeatable, inspectable procedure is. This section walks through the documents a professional is expected to reference before striking an arc on SG iron, the qualification route for procedures and welders, and the inspection methods that decide whether the repair ships or gets scrapped.

AWS D11.2/D11.2M: The Working Guide

The central document in North American practice is AWS D11.2/D11.2M, Guide for Welding Iron Castings. It grew out of the older AWS D11.89, renumbered as part of AWS’s standards housekeeping, and it is the document most often cited in repair procedure discussions. The current 2025 edition was updated to include recommended preheat and interpass tables plus explicit weld quality control requirements for welds in iron castings, which matters because preheat is the single biggest variable separating a repair that holds from one that cracks weeks later in service.

The guide’s practical scope covers:

• base metal preparation, including crack excavation and cleaning;
• preheat and interpass control;
• electrode and filler manipulation;
• process-specific guidance, principally for SMAW and GMAW;
• weld quality control requirements; and
• procedure and welder qualification considerations.

Because it is a guide rather than a mandatory specification, D11.2 does not by itself make a weld acceptable. It gives you the technical basis for a written procedure that an inspector, a client or an insurance surveyor will accept. Filler classification for cast-iron electrodes sits separately under AWS A5.15, in the same way that alloy steel electrodes carry their own AWS specification.

Base Metal Specs: ASTM A536 and EN 1563

Before qualification, the base metal itself has to be pinned down. In the United States, ASTM A536 is the reference specification for ductile iron castings, with grades designated by tensile strength, yield strength and elongation, such as 65-45-12 or 80-55-06. European work follows EN 1563 and the EN-GJS naming system, where grades like EN-GJS-400-18 and EN-GJS-600-3 carry the same information in a different format. A repair procedure that does not state the grade, matrix type and section thickness is incomplete: a ferritic EN-GJS-400-18 and a pearlitic EN-GJS-700-2 want different preheat windows, and a procedure that ignores that is a procedure that will eventually fail.

Procedure and Welder Qualification

Qualified procedures for SG iron are typically qualified in Europe to EN ISO 15614-3, the specification for welding procedure qualification of cast iron and other cast base metals. The route looks familiar to anyone who has run a WPS/PQR system: write the procedure, weld a test coupon under the exact parameters, test the coupon, and issue a procedure qualification record. Welder performance qualification follows the same structure, and position is one of the tested variables. Someone qualified flat and horizontal is not automatically qualified overhead on a vertical pump casing wall, so a refresher on how electrodes behave out of position is worth ten minutes before a performance test.

Inspection Methods That Work on Castings

Inspection is where SG iron repair differs most from steel work. Radiography is of limited value on castings because the graphite nodules themselves create a mottled, grainy image that masks small defects. In practice, destructive-free inspection of cast iron repairs relies heavily on dye penetrant and magnetic particle testing, with surface-breaking cracks the primary rejection criterion under most acceptance schedules. Dye penetrant suits non-magnetic castings and complex, pitted surfaces; magnetic particle testing is faster and more sensitive when the casting is reliably magnetic, and it can hint at near-surface discontinuities a penetrant misses. Visual inspection of the finished bead profile, plus the preheat and interpass temperature log, completes the record.

What Fatigue Research Says About Repair Welds

Even a well-executed repair has limits, and the research is refreshingly blunt about them. Fraunhofer LBF’s nodularWeld project tested repair welds in EN-GJS-400-18-LT, EN-GJS-450-18 and EN-GJS-700-2 on 1000 x 600 x 150 mm plates and found that repaired joints are best suited to low-stress or less functionally critical areas of a component. That aligns with the general weldability picture: a cast iron weld, however carefully made, rarely inherits the fatigue performance of the original casting, as TWI’s cast iron weldability notes explain. For a fatigue-loaded part with a stress concentration sitting directly on the repair, replacing the casting is the engineering answer, not running a bigger bead.

Welding vs. Alternatives: When to Weld SG Iron and When Not To

The hardest skill in cast-iron repair is not running the bead. It is deciding whether a bead belongs there at all. Every cracked housing sits on one of five paths: fusion weld repair, braze weld, mechanical reinforcement, stud or pin repair, or replacing the casting outright. Choosing badly costs a second failure, sometimes with the part back in service under load.

The Five Routes, Compared

Use this as a first-pass screen before committing to a procedure; the right column usually decides the job.

Option	Relative cost	Code acceptance	Fatigue life	Primary risk
Fusion weld repair	Low to moderate	Accepted under AWS D11.2 procedures	Reduced in weld zone	HAZ cracking
Braze weld (bronze)	Low	Accepted if temperature and load allow	Lower strength, no HAZ	Insufficient strength or heat resistance
Reinforcing plate / bolted patch	Moderate	Often preferred, certified parts	Good if stiffness managed	Bolt-hole stress concentration
Stud or pin repair	Moderate	Common, local damage	Depends on placement	Edge distance and thread fatigue
Replace the casting	High, sometimes impossible	Full original certification retained	Original design life	Lead time, obsolete parts

When Braze Welding Beats Fusion Welding

Braze welding with a copper-based filler melts the filler but not the parent metal, so the heat-affected zone never reaches the temperatures that form white iron. That removes the most common failure mechanism in fusion-welded SG iron. The trade-off is real: a braze joint cannot match the tensile strength of ASTM A536 base metal and loses capability as service temperature climbs. It suits low-pressure water paths, non-critical brackets, and cosmetic rebuilds on thick sections where fusion welding would demand heavy preheat. Our air acetylene welding guide covers the torch technique and gas setup.

Mechanical and Stud Repairs: When Certification Rules

On pressure-retaining vessels, pump casings handling hazardous media, and any safety-critical casting, welding often voids the original certification. In those cases a reinforcing plate, a pinned patch, or a bolted cover may be preferable: the repair stays inspectable and the part’s documented status survives. This is also where stick welding advantages matter indirectly: field-fit stud and pin work still depends on that same controlled arc discipline.

Excavation starts with sound removal of damaged material, and air arc gouging remains the practical tool when grinding cannot reach the root: gouge, then grind to bright metal.

What the Fatigue Research Actually Says

Fraunhofer LBF’s NodularWeld project examined repair welding of nodular (ductile) cast iron and concluded that repair-welded castings are primarily suitable for low-stress regions or less functionally relevant areas, precisely because fatigue strength in the weld zone is reduced (Fraunhofer LBF). Read that before certifying a crankcase for cyclic duty: a weld that holds statically can still be the weakest link under vibration, and peening does not restore the original fatigue curve.

The same caution runs through AWS D11.2/D11.2M, which frames cast-iron welding as procedure-driven rather than improvisation, and through TWI’s weldability notes on cast irons.

Final Pre-Weld Checklist

1. Confirm the casting grade from the drawing or certificate; a misidentified pearlitic grade welded on a ferritic procedure is a classic failure route.
2. Confirm duty: pressure, temperature, cyclic load, and whether the part is safety-certified.
3. Choose the route using the table above and document why welding was selected on a certified part.
4. Verify filler classification against AWS A5.15 and check low-hydrogen storage and baking.
5. Set preheat and interpass windows for the actual matrix, with calibrated temperature indication.
6. Plan slow cooling: blanket, dry sand, or a furnace ramp, with no drafts.
7. Schedule inspection (visual, then dye penetrant or magnetic particle) and record heat numbers, temperatures, and operator for the job file.

Safety Notes for Preheat and Post-Weld Work

Preheat changes the hazard profile of the job. Heat-resistant gloves and sleeves are mandatory, and temperature must be measured with temp sticks or a calibrated infrared thermometer rather than guessed, with heating rates slow enough to avoid thermal-shock cracking in a cold, brittle section. Ventilation matters equally: preheated cast iron releases more fumes than a cold weld. Confirm shielding gas and consumables come from a reputable welding supply source with traceable certificates.

Takeaways for Welders and Shop Foremen

SG iron can be welded, but the decision deserves as much rigour as the execution. Weld when the section is moderate, the duty is forgiving, the grade is confirmed, and a written procedure exists. Braze weld when strength and temperature allow and cracking risk outweighs load. Bolt, pin, or stud when certification must be preserved. Replace when the crack sits in a high-stress, safety-critical location or the casting shows prior damage, repeated repair, or graphitisation.

The pattern behind every successful repair is the same: know the grade, control the heat, choose the filler deliberately, cool slowly, inspect honestly, and know when to stop. That discipline, not amperage, is what keeps a repaired casting alive.

Frequently Asked Questions

Can you weld SG iron (ductile/nodular) cast iron?

Yes. SG iron is weldable with proper technique — controlled preheat, low heat input, nickel or nickel-iron filler and slow cooling. It behaves far better than grey iron thanks to its spheroidal graphite, but a careless procedure will still crack it.

What preheat temperature is needed for welding SG iron?

For ferritic nodular grades, TWI’s typical guidance is room temperature to 150 °C for MMA/MIG; pearlitic nodular grades need 200–330 °C. Always apply preheat gradually and hold interpass temperature within the procedure’s range.

Which filler rod is best for SG iron?

AWS A5.15 ENiFe-CI (about 55% nickel, balance iron) is the most versatile choice for nodular and ductile irons, offering roughly 400 MPa tensile strength with 10–12% elongation and good machinability. Pure ENi-CI is softer and better for light grey-iron work.

Why does my SG iron weld keep cracking along the edge?

Cracking along the fusion line usually means the heat-affected zone cooled too fast and formed a hard white-iron or martensitic layer. Increase preheat, reduce heat input swings, use small stringer beads, peen while hot, and cool the casting slowly under insulation.

Is MIG welding SG iron acceptable?

Yes — TWI lists MIG/FCAW among usable processes for cast irons, typically with a neutral gas such as 100% argon and DC reverse polarity, using low heat input, preheat and slow cooling. Pulsed MIG helps keep heat input down on thin castings.

Do I need to preheat if the casting is small and thick?

Often yes, even on small parts, because mass increases constraint and shrinkage stress. Follow the grade-specific table — ferritic SG can sometimes be welded from room temperature, but pearlitic or heavily restrained parts should be preheated to 200–330 °C.

Should welded cast iron be machined afterwards?

Machinability depends on filler choice — nickel-iron ENiFe-CI deposits machine reasonably well, pure nickel even better. Allow the part to cool slowly first, and be aware that hardened HAZ areas near the weld can destroy tooling if preheat was insufficient.

Is a welded SG iron repair as strong as the original casting?

Not in fatigue. Fraunhofer LBF research on repair-welded nodular iron found lower strength in the weld zone, so repaired castings are best suited to low-stress or non-critical areas. For pressure-retaining or safety-critical parts, replacement or engineered reinforcement may be safer.