Cold Lap Weld Defects: Causes, Detection, Code Acceptance & How to Prevent Them

Cold Lap Welds: The Invisible Defect That Fails Perfect-Looking Beads

Your bead looks flawless. The toes are smooth, the profile is textbook, and you are ready to hand the part off. Then the ultrasonic technician runs a probe along the toe, and the screen lights up with an angular indication you never saw coming. That silent rejection is the cold lap weld — the defect that hides in plain sight. Everything on the outside says ship it. Everything on the inside says scrap it.

A cold lap is a micro lack-of-fusion defect at the weld toe that runs more or less parallel to the surface of the parent plate. In plain language, molten filler rolls over steel that never got hot enough to melt, so the two meet, cool, and never truly bond. The result is a planar, buried flaw that behaves like a crack waiting for a reason to open. It can run along a fillet toe, a lap joint edge, or the sidewall of a groove, always parallel to the surface and always easy to miss.

That is what makes cold lap so frustrating. It is not a crater, a pinhole, or a spatter trail you can spot from a meter away. It lives at the fusion boundary, tucked under a toe that looks like it was drawn with a ruler. You can compare your work to a rundown of good and bad MIG welds and still pass every visual box — while a rejection-worthy defect sits quietly in the grain.

Why Cold Lap Matters to More Than One Shop

Fabricators care because a rejected joint means rework, reinspection, and lost schedule. Structural welders care because a cold lap at the toe is a stress riser sitting exactly where cyclic loads concentrate. Automotive and rail builders care because fatigue cracks start at the toe and travel fast. Even hobby welders care, because a pretty bead that never fused is not a weld — it is decoration. In code work, an accepted visual bead means nothing if the fusion line is not there.

The numbers make it concrete. Eliminating a cold lap at the weld toe can improve fatigue life by a factor of roughly 5.4, which is the difference between a joint that survives a service life and one that fails early. And the process most prone to it is the one many beginners reach for first: in GMAW short-circuiting transfer (GMAW-S), AWS D1.1 excludes the process from prequalification because of its high propensity for incomplete sidewall fusion and cold lapping.

So a defect you cannot see can still sink a project. That is why this article digs into the whole picture: what cold lap is and where it lives, the physics of why molten metal meets cold steel and refuses to bond, the three types (spatter, overlap, and spatter-overlap), and the root causes in parameters, technique, joint design, and contamination. From there we cover why it matters for fatigue and code acceptance, why visual inspection fails while UT, phased array, and fatigue testing succeed, and how cold lap compares with overlap, lack of fusion, and incomplete penetration. Finally, we walk through prevention, repair, and practice drills you can run this week. Whether you are chasing certification or just building cleaner habits at the bench, the fix starts with deliberate practice. Build the habit early and your beads will not just look perfect — they will be.

What Exactly Is a Cold Lap Weld? Definition, Terminology, and Where It Lives

A cold lap is a fusion defect wearing the disguise of a geometric one. In shop-floor language, it is a weld that “sat on” the plate instead of biting into it. Molten metal was deposited, it froze, and the profile looks continuous from every angle, yet along that edge no metallurgical bond ever formed between the weld metal and the parent steel.

Standards describe it more clinically. In ISO 6520-1 terminology, the broader group is lack of fusion, reference 401, which subdivides into lack of side-wall fusion (4011), lack of inter-run fusion (4012), and lack of root fusion (4013). Cold laps sit inside that family as micro-lack of fusion defects at the weld toe, oriented more or less parallel to the surface of the parent plate. That parallel orientation is the giveaway: the unbonded plane runs along the edge of the bead rather than across the joint.

Older and trade references use overlapping words for the same phenomenon. Lack of fusion and penetration is also called cold lapping or cold shuts, and occurs when there is no fusion between the weld metal and the surfaces of the base material. The Thermess overview of defect 401 groups these under one umbrella, while AWS D1.1 judges fusion discontinuities by acceptance criteria tied to size, location, and inspection method rather than by nickname.

The Vocabulary, Decoded

• Cold lap / micro-lack of fusion: a tiny, localised lack of bond, usually at the toe.
• Cold shut: a fold where two metal surfaces meet without coalescing, often used interchangeably with cold lap in casting and welding contexts.
• Overlap: excess weld metal that rolls over the toe without fusing; visually similar, mechanically distinct. Our overlap welding defect guide breaks down how the two get confused on the floor.
• Lack of fusion (LoF): the parent category, covering side-wall, inter-run, and root varieties.

Codes classify it in slightly different ways, and the label you see on a report depends on which one governs the job. ISO 6520-1 catalogues it as an imperfection under 401, while AWS D1.1 treats it as a discontinuity to be judged against acceptance criteria. ISO 5817 sets quality levels, and the permitted size of a toe lack of fusion shrinks dramatically as you move from level D to level B.

Where Cold Laps Live

Cold laps cluster in a few predictable places. The weld toe is the classic site, where the bead edge feathers into the plate at a shallow angle. They also form along the fusion line where weld metal meets an unmelted side wall, and at spatter sites, where a droplet of molten metal lands on cold steel and freezes before it can wet in. On a multipass joint they can hide between runs, tucked into a step where a later pass failed to remelt the previous one. Because the toe is where stress concentrates in a loaded joint, this is also the worst possible location for a crack-like flaw to form.

Individual cold laps are very small, typically in the range of 0.01 to 1.5 mm. That scale is exactly why they defeat most non-destructive testing: too thin to return a reliable volumetric signal, too shallow to show up on a radiograph, and too tight to resolve with a standard ultrasonic probe.

That distinction drives everything downstream. A localised micro cold lap may pass fitness-for-service assessment depending on stress state and code, while a continuous toe lack of fusion is a planar defect and a fatigue crack starter. Sharp tie-ins and poor transitions start with technique, and consistent welding practice is what eliminates them. For more on how bead-edge geometry is graded, see how geometric shape imperfections are classified in welding.

The Physics of Cold Lap Formation: Why Molten Metal Meets Cold Steel and Doesn’t Bond

Fusion in welding is not adhesion. For two pieces of steel to become one, the atoms at the interface must sit close enough to share metallic bonds, and that only happens when the parent surface itself reaches melting temperature at the moment the filler arrives. Send molten filler onto a surface that is still cold and solid, and it freezes against it like solder on an unfluxed joint: in contact, but not bonded. That unbonded boundary is the cold lap.

Weld Pool Overflow and Droplet Momentum

Most cold laps start as a geometry problem before they become a chemistry problem. When current is low or travel speed is high, the arc melts less base metal than the bead needs, so the puddle spreads sideways and overflows onto cold steel beside the joint. The filler keeps arriving, but the surface underneath never melts. In GMAW the droplet also carries momentum: it strikes the plate and flattens, yet that mechanical impact transfers almost no extra heat. A worn or misaligned contact tip can throw the arc off to one side, piling more fill over an already cold toe. A molten droplet holds far more heat than an ambient plate can absorb fast enough.

Oxide Films and the Manganese-Silicon Interface

Cold laps are partly or fully filled by oxides, and for common steel consumables the dominant oxides are manganese-silicon oxides formed mainly by oxidation of droplets in flight. These oxides melt at high temperatures and wet steel poorly, so they sit at the interface as a physical barrier. Liquid metal can flow right over that film, but it cannot bond through it. Droplet oxidation is the main source, which means shielding-gas coverage, arc length, and tip condition all change how much oxide you smear into the toe. This is the same wettability problem behind many cases of lack of fusion, and it is why a clean, active weld pool matters so much.

Marangoni Flow, Rayleigh Instability, and Trapped Gas

The molten pool is never static. Temperature gradients across its surface create Marangoni flow, which stirs and spreads the liquid outward. On a properly hot pool that circulation is useful. On a cold one it simply pushes the bead out over steel that will not fuse. Rayleigh instability adds to the mess: a free molten stream or droplet naturally wants to pinch into spheres, producing spatter and jagged, unfused edges. Gas entrapment works the same way. According to published weld-defect research, welding on blasted surfaces increased cold lap formation by entrapment of gas, because the pits hold air that expands under the arc and forms a film between the two metals.

Temperature Governs Whether Bonding Happens

Two temperatures decide the outcome: base metal temperature and droplet temperature. Raise either one and the surface melts sooner, giving atoms the chance to join. Overlap cold lap depth is reduced when welding in a non-oxidising environment, which confirms that oxide films are doing much of the damage. The occurrence frequency of overlap cold laps also decreases with increased preheating temperature of the base metal, a trend reported in welding research on overlap defects. Preheat gives the joint the thermal headroom it needs, and it is one of the most reliable fixes you can apply.

Put the whole chain together and it reads like this:

1. The arc fails to melt the parent surface.
2. Filler overflows the pool onto cold steel.
3. A manganese-silicon oxide film coats the interface.
4. Droplet momentum and Marangoni flow spread the bead without bonding.
5. Trapped gas and Rayleigh breakup finish the job.

Every one of these mechanisms responds to the same inputs you tune at the machine, so steady welding practice on clean, preheated plate is the fastest way to make cold lap disappear.

The Three Types of Cold Lap: Spatter, Overlap, and Spatter-Overlap

Cold lap doesn’t appear as one uniform flaw. Swedish research on gas metal arc welding (GMAW) and tandem GMAW sorts the defect into three distinct families based on how the unbonded metal arrived: spatter cold lap, overlap cold lap, and spatter-overlap cold lap. Which type you’re looking at tells you a great deal about what went wrong.

The taxonomy comes largely from work at Chalmers University of Technology. Researchers examined how stray metal and irregular weld toes create hidden lack of fusion, and their findings are documented in the published thesis on cold lap formation. The studies centered on GMAW and tandem GMAW, the high-deposition processes common in automated fabrication. Understanding the three types matters because each behaves differently under load and demands a different fix.

Spatter Cold Laps

Spatter cold laps occur when a droplet of molten filler lands on cold base metal beside the pool and freezes there without fusing. The globule sticks, but the interface beneath it stays unbonded. These defects are small, discrete, and notoriously easy to mistake for harmless surface debris. A wire brush removes the visible bump but leaves the lack of fusion untouched. Controlling spatter is the real fix, which is where clean contact-tip maintenance pays off. In tandem GMAW, two arcs share a single pool, so an unbalanced lead and trail arc throws off more stray droplets.

Overlap Cold Laps

Overlap cold laps are different. Rather than a stray droplet, the weld metal itself rolls over the toe without wetting into the base metal. They are usually semi-elliptical at the weld toe and can measure several millimetres in both the longitudinal and transverse directions, making them far larger than spatter defects. They often read as a smooth, tidy toe to the naked eye. For a closer look at how this geometry develops, see our guide to the overlap welding defect.

Spatter-Overlap Cold Laps

The combined type blends both mechanisms. Spatter-overlap cold laps normally appear in a worm-like shape and can sit several millimetres away from the weld toe. The elongated form suggests spatter that was caught, stretched, or rolled into the joint rather than landing as a clean sphere. These can be the hardest to spot because they rarely sit directly on the toe line. Because they sit off the toe, a visual walk-around can miss them entirely.

The RLoF Concept and the 1 mm Threshold

To quantify how severe these defects are, researchers use the ratio of lack of fusion (RLoF) at the spatter/base metal interface, meaning the fraction of the contact area that failed to bond. The number is far from linear. RLoF increases abruptly as spatter diameter falls below a critical value of roughly 1 mm, then jumps to 100 percent and stays there. In other words, the smaller the spatter, the worse the bond. Small spatter freezes fast against cold steel and simply cannot fuse. The threshold explains why tiny spatter is often rejected outright while a larger, partially bonded lump can slip through.

Spatter distance matters too. The farther a droplet travels before landing, the more it cools, and the weaker the bond becomes. Distance also determines where the defect lands relative to the toe, which is why spatter-overlap laps sit well off the weld line.

One field pattern is worth remembering: when cold laps run continuously along an entire toe, the cause is rarely random spatter. Continuous defects usually point to poor surface condition, such as mill scale, oil, or rust, or simply the wrong parameters. A sporadic lap is an accident; a continuous one is a setup problem. Building the habits to catch these early is exactly what structured welding practice develops.

Root Causes of Cold Lap Welds: Parameters, Technique, Joint Design, and Contamination

Cold lap almost always traces back to one problem: not enough heat reached the fusion face before the molten metal froze against it. Settings, torch angle, and joint prep are just different ways of starving the arc or hiding the steel from it. Group the causes into four boxes and diagnosis gets faster.

Heat Input and Parameter Problems

Insufficient heat input sits at the root of nearly every cold lap. That shows up three ways: current too low, travel speed too high, or the arc impinging on the molten pool instead of the joint face. Crank down the amps and the arc no longer has the energy to melt the sidewall, so the puddle solidifies against cold steel.

Weld too fast and the arc passes over the joint before it can melt the parent metal, depositing metal that looks fused but is only touching. Voltage mismatched to the transfer mode makes it worse. Run spray-transfer voltage on a short-circuit machine, or dial too little voltage into globular transfer, and you get a spattery, unstable arc that cannot wet the toes properly. For the wire-fed side of this, the settings and polarity conventions matter enormously, so it is worth reviewing a dedicated flux-cored welding guide and a wire welder walkthrough before blaming technique.

Short-Circuit MIG on Thick Sections

Short-circuit MIG/MAG on thick sections is the classic cold lap recipe because it combines low heat transfer with high deposition. The process runs cold by design — the arc only exists for a fraction of each cycle — so it deposits a lot of metal without putting much heat into a heavy plate. On thin sheet that is perfect. On half-inch steel, the weld rides on top of a cold wall. The industry’s general review of geometric shape imperfections documents this mechanism.

Technique: Angle, Stick-Out, and Arc Placement

Torch technique decides where the heat lands. Excessive stick-out drops the current (a longer electrical stick-out means lower amperage) and destabilizes the arc. A steep push angle can drive the puddle ahead of the arc, and downhill welding lets gravity pull molten metal onto unmelted walls faster than the arc can fuse them. Oversized weave is a quieter killer: a wide side-to-side motion lets the arc dwell in the middle and skip the toes.

The single most important habit is keeping the arc on the leading edge of the puddle. If the arc rides back into the pool, the molten metal cushions it and absorbs the energy before it ever reaches the joint face. Keep the arc at the front so it is always digging into unmelted parent metal. Contact tip condition factors in too: a worn or oversized bore gives you erratic arc starting and wandering, which nudges you toward riding the pool.

Joint Design and Fit-Up

Geometry hides the arc from the walls. On a wide joint, an arc directed down the centre lets molten weld metal flow and cast against the side walls without melting them. Narrow bevels, tight root gaps, and misalignment all create the same effect: the arc cannot reach the fusion face, so the puddle just pours into the gap and freezes against cold steel. If the arc never touches the wall, no amount of current fixes it.

Contamination Between the Metal and the Arc

Heavy mill scale, rust, or primer on the fusion faces can stop weld metal from fusing with the parent metal, producing overlap and cold lap even when your parameters are perfect. These layers melt at higher temperatures than the steel underneath, so the puddle freezes against a barrier instead of bonding to it. Slag trapped between passes does the same thing on multipass work. Poor cleaning is a root cause hiding in plain sight. The industry’s lack of fusion breakdown covers the mechanism, and it is the same family as an overlap welding defect.

Fix the cause and the defect disappears. Chasing it with a grinder and a fresh pass only hides it until the next inspection.

Why Cold Lap Welds Matter: Fatigue Failure, Stress Concentration, and Code Rejection

A cold lap doesn’t fail loudly. It hides in the toe of a bead that looks flawless, fused to the surface but never genuinely bonded to the parent metal, and quietly rewrites how the entire joint behaves under load. The consequences are what separate a weld that passes a visual glance from one that survives years of service.

Cold Laps Are Sharp Stress Concentrators

Every weld toe concentrates stress to some degree, but a cold lap turns a smooth toe into a sharp notch. Because the unfused interface is planar and crack-like, it acts much like an embedded flaw, and the sharper the geometry, the higher the peak stress that piles up at its tip. This is a classic geometric shape imperfection, and geometry is everything when stress is looking for somewhere to concentrate.

That effect is captured by the stress concentration factor, Kf. In one analysis, a 60-degree flank angle weld containing a cold lap carried an estimated Kf of about 7, compared with roughly 4.2 for the same joint without the defect. Eliminating the cold lap improved fatigue life by at least 5.4 times. A single hidden flaw can shrink a joint’s life to a small fraction of its design intent.

Fatigue Failure: The Real Enemy

Static strength might tolerate a cold lap; cyclic loading usually won’t. Cold lap defects have been identified as important initiation sites of fatigue failure in welded structures, because every load cycle pries the notch tip a little further open until a crack forms, then grows.

Service loading is what exposes the problem. A single heavy load can sit on a cold lap undetected, but the repeated, smaller cycles of real service — traffic, vibration, thermal expansion — are what turn a notch into a crack. The rail industry learned this the hard way: in thermite weld studies, eliminating cold laps and improving toe geometry improved fatigue life by as much as a factor of 5. The same crack-like logic explains why lack of fusion, another planar discontinuity, drastically reduces joint strength under dynamic or cyclic loading.

Code Rejection: AWS D1.1 Gives No Pass

Codes do not treat this as a judgement call. Under AWS D1.1, both overlap and cold lap are always rejectable — there is no acceptable size threshold and no allowable amount. If an inspector finds one, it requires repair. Compare that with porosity or undercut, where the code sets limits you can work within. The overlap welding defect and cold lap get none, which tells you how seriously a planar, unfused toe is viewed.

Safety, Liability, and the Cost of Late Detection

The real sting is timing. Cold laps defeat visual inspection and often surface only under ultrasonic or radiographic testing — after fabrication, after coating, sometimes after installation. Repair then means gouging, rewelding, re-inspection, and schedule slippage. Rework on a finished, coated, or installed part can cost many times the original welding hour, and on a large fabrication the delay cascades into downstream trades and delivery dates. In a pressure boundary, lifting frame, or structural connection, a late-found cold lap is a safety issue, not just a paperwork one: the failure mode it feeds is sudden, not gradual.

Detection failures also carry liability. When a service failure is traced back to a defective weld, responsibility lands on the fabricator and the inspector who signed it off. The cheapest insurance is consistency: disciplined machine setup, correct contact tip condition, and toe geometry that never creates a notch in the first place. Focused welding practice and, on high-volume work, the repeatability of robotic submerged arc welding both remove the human variability that builds cold laps.

A cold lap is patient. It waits for a fatigue cycle or a probe — and either way, it eventually tells on you.

Detecting Cold Lap Welds: Why Visual Inspection Fails and What Actually Works

Cold lap is the inspector’s nightmare for one reason: the better it looks, the harder it is to catch. This is a buried, planar imperfection — a flat plane of unbonded metal sitting just beneath an otherwise flawless toe. Visual testing to ISO 17637 examines surfaces, and a cold lap barely disturbs the surface at all. A visually perfect bead is entirely compatible with lack of fusion hiding a millimetre underneath.

Why the Eye Cannot Do This Job

Visual inspection is good at surface-breaking flaws: cracks, undercut, an overlap that stands proud of the plate. Cold lap keeps its unbonded plane internal and tightly closed. Even a trained inspector working under magnification and strong lighting sees a smooth, wet-in transition and has no reason to reject the part. That is exactly why a bead can pass every good and bad MIG welds checklist and still fail a volumetric or ultrasonic examination downstream.

Ultrasonic Testing: The Primary Route for Planar Defects

Where the geometry permits access, ultrasonic testing is the leading method for planar imperfections because a planar reflector returns a strong, clean signal. Conventional UT and phased array testing, applied to ISO 17640 and ISO 13588, let a technician steer a beam through the toe region at an angle that will actually intercept an unfused plane. Phased array adds the ability to sweep and focus electronically, so the operator can map the extent and depth of an indication rather than guessing at it.

That matters because cold laps are so small and tight that they have not been successfully detected by conventional non-destructive testing methods with any reliability. UT and PAUT are effective precisely when the beam and the defect plane are favourably oriented — a caveat worth repeating in any inspection plan.

Why Radiography Can Slip Straight Past a Cold Lap

Radiography is excellent for volumetric flaws like porosity and slag, but it struggles with tight, planar ones. A lack-of-fusion plane sends almost no change in density through the beam, so contrast is minimal. Worse, when the plane is oriented unfavourably relative to the radiation path, the film can miss it entirely. A cold lap is not a void with obvious density change; it is often a closed, near-thickness joint. Dependence on RT alone is a common reason field cold laps survive to service.

In-Process Monitoring: Catching the Conditions, Not the Defect

The most promising shift is detecting the conditions that create lack of fusion while the pass is being made. Infrared thermography watches the weld pool and trailing cooling curve; high-speed pool imaging tracks arc behaviour and wetting. Both flag the signatures of low arc energy, high travel speed, and abnormal cooling — the recipe for cold lap. Setting the right parameters first is the cheapest defence; polarity decisions such as TIG welding polarity and wire-feed setup shape whether the toe wets or laps, and these signals expose bad settings immediately.

Destructive Verification: Macroetch and Metallography

When you need ground truth, you cut the weld open. Macrosection and metallographic examination reveal the unbonded toe line directly and are the reference standard for proving a technique. Welders build that eye by cutting coupons deliberately — a routine covered well in welding practice guides — because seeing dozens of lap lines teaches you what caused them.

One important caveat: cosmetic fixes are not cures. Even TIG dressing, which re-melts the toe and improves the profile, does not fully eliminate cold laps. It can bury a planar defect deeper rather than remove it. For more on how these related geometric flaws present, see our breakdown of the overlap welding defect.

Cold Lap vs Overlap vs Lack of Fusion vs Incomplete Penetration: Clearing Up the Confusion

Four names get thrown around interchangeably on the shop floor, and that sloppiness causes real problems. A repair crew grinds at a cold lap, the inspector calls it lack of fusion, the engineering report lists incomplete penetration, and nobody agrees on whether the part ships. They are related, but not identical, and your welding code treats them differently.

Here is the clean version of each.

Cold Lap and Overlap: Same Defect, Two Names

An overlap is an extension of weld metal beyond the weld toe, not fused to the base metal. Picture molten metal rolling out over the parent plate like a wave freezing in place; it sits there, smooth and convincing, but never bonds underneath. A cold lap has the same appearance and the same consequence. It is weld metal laid over cold steel without achieving fusion, leaving a notch at the toe.

The terminology drifts because overlap describes geometry (metal beyond the toe) while cold lap describes the mechanism (cold metal, no bond). In practice inspectors use the words together. For a deeper walkthrough of causes and fixes, see our guide to the overlap welding defect. The concept worth separating is technique from defect, which we unpack in overlapping welding: technique vs. defect.

Lack of Fusion: The Broader Family

Lack of fusion (LOF) is the umbrella condition. It means weld metal failed to fuse to base metal or previously deposited weld metal anywhere along a sidewall, root, or toe. Overlap and cold lap are specific surface presentations of LOF. Classic LOF hides deeper, along the fusion line, and is a planar discontinuity. Therness classifies lack of fusion as one of the most damaging weld defects because the sharp, unfused interface concentrates stress. TWI’s review of geometric shape imperfections groups overlap and cold lap as surface imperfections, distinct from internal fusion problems.

Incomplete Joint Penetration: A Different Animal

Incomplete joint penetration (IJP) is not a surface fusion problem at all. It is weld metal that fails to extend through the full thickness of a joint. In a complete joint penetration (CJP) groove weld, the root never reaches the far side of the groove. IJP creates a planar discontinuity with a sharp tip running along the root, and it is always rejectable in a CJP weld, no matter how small. It is a design-intent failure, not a technique slip.

Acceptance Criteria: Why the Code You Weld To Changes Everything

This is where the confusion turns into money. A discontinuity becomes a defect only when it exceeds the governing code’s acceptance limit. Same flaw, different verdict.

Under structural steel codes like AWS D1.1 and D1.3, overlap is typically rejectable. The reasoning is straightforward: an unfused toe is a stress riser, and structural welds live and die by fatigue performance. Now shift to sheet metal under AWS D9.1. There, overlap may be acceptable if the toe is visible and fusion to the base metal can be verified. That leniency reflects sheet metal’s different service demands.

International codes sit in between. BS EN ISO 5817 permits only short overlap lengths at moderate quality level D, tightening further at higher quality levels. The takeaway is the same across all of them: the flaw does not change, the standard does.

For a full discussion of how quality criteria drive acceptance, see welding discontinuity vs defect. And if you want to keep these boundaries straight on the job, build the habit through deliberate welding practice. Incomplete penetration and cold lap often get lumped together in training, but root geometry is where they part company.

Prevention and Repair: Practical Strategies to Eliminate Cold Lap Welds

Cold lap is rarely random. It is the predictable result of an arc that never quite reached the joint face, a bead that bridged instead of bonded, or a surface too dirty to fuse. Preventing it comes down to controlling four things: heat input, aim, geometry, and cleanliness. Repairing it means removing everything that didn’t bond — not covering the flaw with a cosmetic cap.

Prevention Starts With Real Numbers in the WPS

The written procedure is your first line of defence. Instead of trusting the nominal figures on a machine dial, compute arc energy from measured current, voltage, and travel speed for every pass. Settings drift, and a travel speed that creeps up can quietly drop heat input below the point where fusion happens. Mark a known length, time the pass, and log the result so the numbers stay honest shift after shift.

Aim, Angle, and Stick-Out

Keep the arc on the leading edge of the pool, aimed at the joint face — never buried in molten metal. When the arc sits in the puddle, the base metal never gets hot enough to fuse, and you get that smooth, deceptive toe. Angle the torch so the arc pushes into the root and sidewalls, and mind your stick-out and contact-tip condition, because excessive stick-out reduces current and destabilises the arc. Accurate aim is a skill, and it improves fastest with deliberate welding practice rather than hours of repetition on scrap.

Joint Design and the Split-Bead Technique

Narrower joints with adequate side-wall access let the arc reach the fusion faces instead of bridging the gap. On thick material, use a split-bead technique after the root passes: run stringers down one side, then the other, and fill between them. Large beads that bridge the entire gap in a single pass are the classic cold lap generator, because the arc never touches the sidewalls and molten metal simply lays over cold steel.

Cleanliness, Preheat, and Arc Stability

Enforce interpass cleaning as a hold point. A residue film — oil, mill scale, slag, or oxide — prevents metallurgical bonding even when thermal conditions are correct. Preheat the workpiece, weld in a non-oxidising environment where possible, and keep the process as stable as you can, since lack of fusion is suppressed by a steady arc on clean, warm metal.

Repair: Excavate, Reweld, Re-Inspect

A cold lap cannot be fixed by extra passes on top. Excavate the defect fully, remove the unfused toe with a grinder or arc gouging until you reach sound metal, then re-weld the cavity to a qualified procedure. Re-inspect using the same method that found it — usually ultrasonic or phased-array — because a geometric imperfection this tight against the toe still hides from the naked eye after repair.

Common Mistakes When Chasing Cold Laps

Welders trying to fix cold lap often make it worse. Watch for these traps:

• Cranking amperage without correcting aim or travel speed
• Adding a cover pass over a toe that never bonded
• Skipping preheat on thick or high-carbon sections
• Leaving interpass residue in place between passes
• Signing off after a visual check that missed the defect the first time

Fix the cause, not the symptom. Correct heat input, honest aim, sensible bead sequence, and disciplined cleaning will keep cold lap out of the joint — and when it does slip through, excavate it properly instead of hiding it.

Field Tips, Practice Drills, and Final Takeaways on Cold Lap Welds

Cold lap is a heat-and-technique defect, not bad luck. Every rejected toe is a process clue: the puddle was too cold, the arc too far ahead, or the joint too tight to let molten metal wet the sidewall. Build habits that remove those variables and the defect becomes rare instead of routine.

Run-Time Discipline on the Floor

• Own your heat input. Lock travel speed, amperage, and wire feed into a repeatable window before chasing cosmetics.
• Watch the puddle leading edge. If it races ahead of the arc and washes across unmelted steel, you are overlapping, not fusing.
• Re-verify torch alignment after every tip change. On robotic cells, a new contact tip that seats a millimetre off throws the arc off the joint line for the next hundred parts.
• Fix the fit-up, not just the welder. Repeated lack-of-fusion occurrences point to a design or fit-up problem, not only an operator problem. A slightly wider bevel is cheaper than an ultrasonic testing repair.
• Catch it live. In production, real-time thermal monitoring flags a cold joint during the weld rather than after the customer’s ultrasonic test.

A Self-Check Routine You Can Run Tomorrow

1. Before welding: confirm polarity and settings, inspect the tip and nozzle, and measure the joint gap.
2. During the pass: watch the puddle wet both sidewalls; pause if the leading edge outruns the arc.
3. After cooling: examine the toes under good light. A sharp, unfused line beside a smooth bead is a flag, not a finish.
4. End of shift: log any parameter or fit-up change so patterns surface before parts ship.

What the Codes and the Data Tell You

Codes treat cold lap as a fusion discontinuity that must meet acceptance limits under AWS D1.1, while visual inspection alone routinely misses it because the surface looks perfect. Ultrasonic testing catches the planar indication that hides beneath. The consequence is fatigue: cold lap acts like a crack starter. Improving the toe by reducing flank angle and increasing toe radius lowers stress concentration and boosts fatigue life (Chalmers research). And because cold lap cannot be avoided completely in GMAW, minimising its number and size is the practical target.

Build the Skill Before the Deadline

Set up a dedicated booth first; a clean, stable bench removes half the variables, as covered in our guide to welding facilities. Then drill deliberately: run beads at three travel speeds on scrap, change one variable at a time, and break the coupons in a press to see fusion up close. That feedback loop beats guessing, and it is exactly how welding practice turns puddle-reading into instinct.

Cold lap hides in plain sight, but it never survives good heat control and honest technique. Verify heat, watch the toe, and inspect before the customer does. Run one self-check drill this week, break the coupon, and let the metal show you the truth.

Frequently Asked Questions

What is a cold lap weld?

A cold lap is a small lack-of-fusion defect at the weld toe that runs roughly parallel to the surface of the parent plate. It forms when molten weld metal flows onto base metal that never reached melting temperature, so the two are in contact but not metallurgically bonded. In ISO 6520-1 it is classified as a type of lack of fusion (reference 401).

Can you see a cold lap with the naked eye?

Generally no. Cold laps are planar and usually buried beneath a bead that can look perfectly acceptable, which is why visual inspection alone cannot certify against them. The irregularity at the toe may hint at an overlap cold lap, but confirmation normally requires ultrasonic or metallographic examination.

Is cold lap rejectable under AWS D1.1?

Yes. AWS D1.1 treats overlap and cold lap as always rejectable discontinuities with no acceptable threshold, requiring repair. Some other codes, such as AWS D9.1 for sheet metal, are more lenient if the weld toe is visible and fusion to the base metal can be verified.

What causes cold lap in MIG welding?

The usual causes are insufficient heat input, travel speed that is too high, and the arc riding on the weld pool instead of the leading edge. Short-circuit transfer on thick sections is a classic recipe, and contamination such as mill scale, rust, or primer also prevents proper fusion.

Why are cold laps dangerous?

Cold laps behave like sharp planar cracks and concentrate stress at the weld toe, making them prime initiation sites for fatigue failure. Research on rail thermite welds showed that eliminating cold laps could improve fatigue life by roughly a factor of five or more.

How do you detect cold lap welds?

Ultrasonic testing, especially phased array, is the preferred method because planar reflectors produce strong signals. Radiography can miss a tight cold lap entirely. In-process thermal monitoring can flag the cold-joint conditions while the weld is still being made.

Can cold lap welds be repaired?

Yes, but repair requires removing the unfused material entirely, not simply capping over it. The affected toe or fusion face must be excavated to sound metal, rewelded using a qualified procedure with adequate heat input, and then re-inspected.

Is cold lap the same as incomplete fusion?

They are closely related. Cold lap is generally a micro-level lack of fusion at the weld toe, while incomplete fusion is the broader category covering any failure of weld metal to fuse with the base metal or a previous pass. In practice, the terms are often used interchangeably on the shop floor.

Does preheating prevent cold laps?

Preheating helps significantly, particularly for larger spatter-generated cold laps and overlap cold laps. Higher base metal preheat temperature reduces the frequency of overlap cold laps by helping the parent surface reach fusion temperature when metal arrives.

Authoritative Sources & Further Reading

• https://www.twi-global.com/technical-knowledge/job-knowledge/a-general-review-of-geometric-shape-imperfections-types-and-causes-part-1-067
• https://www.diva-portal.org/smash/get/diva2:648916/FULLTEXT01.pdf
• https://research.chalmers.se/en/publication/163405
• https://pubs.aws.org/Download_PDFS/D1.1-D1.1M-2025WebPV.pdf
• https://onlinepubs.trb.org/onlinepubs/archive/studies/idea/finalreports/highspeedrail/hsr-24final_report.pdf
• https://www.scientific.net/AMR.875-877.1421
• https://www.therness.com/defects/401-lack-of-fusion/
• https://www.therness.com/blog/lack-of-fusion-welding-causes-detection-acceptance-criteria/
• https://wpswelding.com/blog/weld-discontinuity-vs-defect-aws-d1-1
• https://materialwelding.com/lack-of-fusion-in-welding-causes-remedies/
• https://app.aws.org/forum/topic_show.pl?tid=38515