Bad MIG Weld Examples: How to Spot, Diagnose, and Fix Them

Table of Contents

Bad MIG Weld Examples: Why the Worst Welds Often Look the Best

It is a moment every welder knows. You strike an arc, lay down a clean, shiny bead, and step back to admire it. Then, with one tap of a chipping hammer, the entire weld peels off the plate. This frustrating experience is one of the most common bad MIG weld examples that beginners and veterans alike will run into. The irony is that some of the weakest welds can be the most visually pleasing.

Gas Metal Arc Welding (MIG/GMAW) is one of the most popular processes in shops today, and for good reason. As noted by TWI, MIG/GMAW is a consumable-electrode process in which the welder has limited control over weld pool penetration independent of depositing weld metal. This tradeoff allows for high travel speeds, but it also means a bead can stack up on top of the base metal without ever bonding to it. To avoid this, it is worth brushing up on the fundamentals with guides such as How to Use a Wire Welder: Complete Beginner’s Guide and Welding Practice: The Complete Guide to Building Real Welding Skills.

Quick Note: A smooth, convex, and spatter-free bead is not proof of strength. Fusion, not cosmetics, is what holds the joint together.

The Short-Circuit Trap

Much of this deception comes from short-circuit transfer, the most common setting for thin materials. According to Lincoln Electric, in short-circuit transfer, the wire touches the pool and the arc shorts 40 to 200 times per second, so metal can melt without fusing to the base plate. This is exactly how a weld can look perfect on the surface while having no root fusion underneath. For those working with different materials or setups, resources like DC Welding Aluminium: How It Works & When to Use It can help tailor settings to avoid this issue.

Discontinuity vs. Defect

Before judging a weld, it is important to know the difference between a discontinuity and a defect. As defined by AWS A3.0, porosity is “cavity-type discontinuities formed in weld metal by gas entrapment during solidification.” A discontinuity is simply an interruption in the weld’s structure, but it only becomes a defect when it exceeds the limits set by a code. For structural steel, AWS D1.1 is one of the most referenced standards. Even environmental factors play a role: AWS D1.1 prohibits GMAW when wind speed near the weld exceeds 5 miles per hour [8 km/h]. These restrictions highlight why proper Welding Facilities: How to Plan & Set Up a Safe Shop matter just as much as technique.

What This Article Will Cover

The goal of this guide is to equip you with the ability to spot, diagnose, and fix bad MIG welds. Over the next sections, we will explore surface flaws, hidden killers, root causes, inspection methods, and proven fixes. Along the way, we will reference practical topics, from Contact Tip Welding: Selection, Maintenance, Wear, and Troubleshooting for GMAW to Good and Bad MIG Welds: How to Tell the Difference.

Above all, remember this: the worst welds do not always look the worst. They often look the best until tested. By learning to read beyond the surface, you can catch these failures before they ever leave your shop.

What Counts as a Bad Weld? Discontinuities, Defects, and Code Standards

Before you can call a weld “bad,” you need the vocabulary inspectors use. Words like “defect” and “discontinuity” get tossed around loosely in a home shop, but they carry very specific meanings on a jobsite governed by a code. Getting the language right is the first step toward reading your own welds like a professional.

Indication, Discontinuity, Defect: Three Different Things

An indication is simply a signal that something might be wrong — a blip on an ultrasonic screen, a shadow on a radiograph, a pinhole glinting on a MIG bead. An indication is a prompt to look closer, not a verdict.

A discontinuity is any interruption in the uniform nature of an item — a void, a crack, a bit of trapped slag, or a spot where the bead never fully bonded. Discontinuities are extremely common; almost every weld contains a few.

The important twist: a discontinuity only becomes a defect when it is severe enough to impair the item’s suitability for its intended service. That’s the heart of the whole system. The same undercut that fails inspection on a seismic brace can be perfectly acceptable on a decorative gate.

Quick Note: “Bad” is never absolute. The same discontinuity can pass or fail depending on the code, the loading, and the service. Inspectors don’t ask “is it ugly?” — they ask “does it exceed the acceptance criteria?”

The AWS B1.11 Checklist of Surface Conditions

The American Welding Society’s AWS B1.11 visual inspection guide lays out the surface conditions an inspector evaluates. It’s the closest thing to a master list of what “bad” actually looks like:

Porosity
Incomplete fusion and incomplete joint penetration
Undercut, underfill, and overlap
Lamination, seams and laps
Cracks
Slag inclusions
Excessive reinforcement, convexity, and concavity
Arc strikes, spatter, and melt-through
Weld size and surface oxidation

Nearly every one of these has a MIG equivalent, and each gets its own treatment later in this article. If you want a head start on telling a sound bead from a suspect one, our guide to good and bad MIG welds walks through the visual tells.

The Codes That Decide Pass or Fail

Acceptance criteria come from the applicable code, and different codes draw very different lines:

AWS D1.1 governs structural steel. Its stance is famously strict: the AWS D1.1 structural welding code states that any crack shall be unacceptable, regardless of size or location.
ASME B31.1 covers power piping in power plants and industrial steam systems, where pressure and temperature service drive the limits.
API 1104 applies to pipeline welding, with acceptance based heavily on radiographic testing.
ISO 5817 defines three quality levels: B (stringent), C (intermediate), and D (moderate). You pick the level based on how the part must perform.

Notice the pattern. A bridge weld might reject a 1 mm undercut that a lawnmower deck would tolerate without a second thought. Porosity allowances, reinforcement height, and crack tolerance all shift with the standard. That context-dependence is why the same visual checklist can produce opposite verdicts on two different jobs. The techniques that prevent these defects, though, are universal — and that’s what the rest of this article covers.

For more context on the defect families inspectors hunt for, the TWI guide to weld defects and imperfections is a solid reference.

With the language settled, let’s start the defect-by-defect tour — beginning with the gas holes that quietly hollow out a bead.

Porosity: The Gas Holes That Weaken a MIG Bead

If there’s one bad MIG weld everybody recognizes on sight, it’s porosity. You see it as a bead freckled with tiny pinholes, as scattered dark dots that survive the wire brush, or as a smooth cap that “pops” open into a hollow crater the instant you hit it with a chipping hammer. Sometimes the holes are elongated tracks called wormholes; sometimes they break the surface as blowholes. Either way, gas got trapped inside the metal as it froze.

How Gas Gets Trapped in the Puddle

Molten steel can hold nitrogen, oxygen, and hydrogen in solution. The catch is that its appetite for those gases collapses as the puddle solidifies. Metal that was happily absorbing gas at 2,700°F suddenly rejects it as it turns solid, and the rejected gas has to go somewhere. If it reaches the surface in time, it escapes. If the weld freezes first, the bubble is locked in — and that void is porosity. Contamination makes it worse because every trace of oil, rust, moisture, or coating adds more gas to the pool. Hydrogen lurking in moisture is the worst offender, because it diffuses fast and rarely escapes in time.

The numbers are sobering. According to the Australian Welding Institute, as little as 1% air entrainment in the shielding gas causes distributed porosity, and more than 1.5% produces gross surface-breaking pores. That’s why even a small draft or a loose fitting shows up as holes instead of a clean bead.

Quick Note: porosity is a shielding problem first. Before you grind, chase, or blame your technique, check gas flow, look for leaks, and feel for a draft.

What Actually Causes the Holes

Poor or turbulent shielding. Too little flow lets the atmosphere in; too much creates turbulence that whips air into the stream. Wire processes generally want about 30–50 CFH, and flow above 60 CFH actually invites the atmosphere into the weld (Executive Filler Metals).
Drafts and wind. A shop fan, an open door, or a breezy bay will strip shielding gas away.
Leaks. A split hose, a loose fitting, or a cracked gun liner pulls air straight into the gas path.
Contaminated base metal. Paint, oil, grease, rust, and especially zinc coatings like galvanizing boil off into the arc.
Excessive contact-tip-to-work distance. Stick the nozzle too far out and the gas has nothing to shield.
Wet or frosted cylinders. Moisture in the gas supply means hydrogen in the weld.

Fixes You Can Apply Today

Start with gas flow: set it in the 30–50 CFH range and confirm it with a flowmeter at the nozzle, not just at the regulator. If it reads fine at the regulator but not at the gun, you have a leak — check the hose, fittings, and solenoid. Keep your wire extension short: a good rule of thumb is to extend the wire no more than 1/2 inch past the nozzle (MillerWelds). That’s the same nozzle discipline that keeps a good Contact Tip Welding setup feeding smooth and clean.

Then attack contamination. Grind or wire-brush the joint to bare metal, remove paint and zinc at least an inch back, and wipe off oil. On cold, thick plate, preheat to help the puddle stay fluid long enough for gas to escape. Outdoors or near fans, use wind screens and increase flow slightly — but never past the point of turbulence.

If you weld without gas, porosity has a different cause: a bad Gasless MIG flux-cored setup or a drag angle that outruns the slag. Either way, learning to see the difference between cosmetic spatter and real gas holes is a core skill — one worth building alongside the basics of Welding Practice.

It helps to know the code tolerance too. AWS D1.1 permits visible piping porosity 1/32 in [1 mm] or greater to total no more than 3/8 in [10 mm] per linear inch and 3/4 in [20 mm] in any 12 in [300 mm]. For fillet welds, frequency is limited to one pore per 4 in [100 mm] with a maximum diameter of 3/32 in [2.5 mm]. Anything beyond that is a defect, not a blemish.

Lack of Fusion and Cold Lap: The Invisible MIG Weld Killers

A lack of fusion defect is the worst kind of bad weld, because it hides. The bead can look smooth, the ripples even, the toes wet in nicely. Yet underneath, the weld metal never bonded to the base metal. One tap with a chipping hammer, or a light load in service, and the joint opens up. These are the defects that pass a visual glance and fail an inspection.

Lack of Fusion vs. Cold Lap: Two Sides of the Same Failure

Lack of fusion means the molten weld metal failed to melt and merge with either the base metal or a previously laid bead. The surfaces touched, but never combined into one. Cold lap is its close cousin: the weld metal rolls over the surface without melting into it, leaving a seam that looks like a weld but behaves like a crack. At the toe it is often labeled overlap, a defect worth studying on its own.

Where These Invisible Defects Hide

Lack of fusion and cold lap prefer tight, hard-to-reach geometry:

Fillet weld toes, where the bead meets the vertical member
Groove sidewalls, especially in thick plate with narrow bevels
Interbead boundaries between passes on multipass welds
The root, where access is limited and heat escapes into the plate

Why MIG Welds Go Cold

A cold weld is usually a cold process. Travel too fast and the arc outruns the melting; travel too slow and the pool piles up ahead of the arc instead of wetting in. According to MillerWelds, lack of fusion is primarily caused by improper gun angle or incorrect travel speed. Both leave the arc aimed at already-molten metal instead of the unmelted joint face. Short-circuit transfer, the standard mode on many small MIG machines, runs at lower heat input and is especially prone to cold lap on thick sections.

Wire feed and voltage set too low starve the pool of energy. Mill scale, rust, and oil make fusion harder because the arc burns through contamination before it reaches clean steel. A worn or wrong-sized contact tip can cause erratic arc starts and inconsistent current, which shows up as poor fusion at the start of a bead.

How to Prevent Lack of Fusion

The fix is heat, angle, and time. Raise voltage and wire feed together so the arc has enough energy to dig in. Hold a 0- to 15-degree gun angle and keep the arc on the leading edge of the weld pool; that pushes molten metal back and drives the arc into virgin metal. On multipass work, weave slightly and pause at each sidewall to let the puddle wet out. Slow down on the root pass, where the plate pulls heat away fastest, and never rush the tie-in at a restart.

Transfer mode matters too. In spray arc transfer the arc never goes out, so cold lapping and lack of fusion are not issues, as Lincoln Electric explains. For thicker steel, switching to spray transfer or to a flux-cored wire that runs hotter is often the simplest cure. Gasless flux-cored welding is a practical upgrade for a small machine that cannot reach spray range.

Quick Note: A weld can be cold even when the settings look right. If your beads sit on top of the plate instead of biting into it, raise voltage in small steps before you touch wire speed.

You Cannot Trust Your Eyes

This is where lack of fusion earns its reputation: it is very difficult to see. A surface that looks tied in can hide a planar flaw with almost no gap. Because it is difficult to detect visually, it must be checked by dye-penetrant, ultrasonic, or bend testing. Radiography often misses it as well, since the crack-like plane gives little contrast. Code bodies take it seriously: TWI notes that ISO 5817 does not permit lack of or incomplete root penetration for Quality Levels B and C. That means a weld can look perfect and still fail a job outright.

Consistent bead-on-plate drills build the muscle memory that prevents cold welds; our guide to welding practice covers the drills worth repeating. For a side-by-side look at strong versus weak beads, see good and bad MIG welds.

Undercut, Overlap, and Poor Bead Profile: Surface Flaws You Can See

These are the defects you can catch without a radiograph. They sit right on the surface of the bead, so a quick look tells you something is wrong. That does not make them cosmetic. Left alone, they quietly shorten the life of a joint. Profile flaws concentrate stress, shrink the throat, and give fatigue cracks a place to start.

Undercut: The Groove Along the Weld Toe

Undercut is a notch melted into the base metal right where the bead meets the plate — the weld toe. Instead of blending smoothly into the parent metal, the edge drops into a groove. You will usually see it on one or both sides of the bead, running the length of the pass.

The usual suspects are too-high welding parameters, inadequate filler metal addition, an incorrect work angle, and too-fast travel speed. The arc burns the parent metal away faster than the filler can fill it back in.

Undercut matters because it removes base metal exactly where load transfers from plate to weld. That sharp toe becomes a stress concentration and a classic fatigue initiation site. AWS D1.1 puts hard numbers on it. For material less than 1 in [25 mm] thick, undercut is limited to 1/32 in [1 mm], with an exception allowing up to 1/16 in [2 mm] for any accumulated length up to 2 in [50 mm] in any 12 in [300 mm]. For material 1 in [25 mm] or thicker, up to 1/16 in [2 mm] is permitted for any length of weld. In primary members transverse to tensile stress, undercut must be no more than 0.01 in [0.25 mm] deep.

Quick Note: An undercut toe is a built-in crack starter. Even a shallow groove raises local stress and invites fatigue failure under cyclic loading.

Overlap and Poor Profile

Overlap happens when molten metal rolls over the toe of the weld but never fuses to the base metal beneath it. It looks like a smooth ridge lying on top of the plate, and a chipping hammer will often find the edge. It typically comes from too slow a travel speed, too much current for that speed, a work angle that lets the puddle run ahead of the arc, or a sloppy tie-in at a restart. Technically it is a lack of fusion wearing a surface costume — see our full breakdown of the overlap welding defect and how it differs from intentional overlapping welding as a technique.

Poor profile covers a family of shape problems:

Excessive reinforcement (convexity): a tall, fat bead crown. It wastes filler, adds stress concentration at the toes, and can hide cold lap underneath.
Excessive concavity: a dished face that reduces the throat below the design size.
Undersized welds: a bead narrower or thinner than the drawing calls for, so it cannot carry the load.
Underfill: the groove is not filled flush, leaving a depression below the base metal surface.

Every one of these changes the weld’s cross-section. A smaller throat means less load capacity; a concave crater or sharp toe means a stress riser. Underfill and undersize are especially dangerous on fillet welds, where the throat carries the entire shear load. Poor shaping is one of the most common reasons a bead passes the eye test but fails a bend or fatigue check.

AWS D1.1 spells out repair requirements for exactly these conditions. Clauses 5.26.1.1 and 5.26.1.2 address overlap, excessive convexity or reinforcement, excessive concavity, undersize welds, and undercutting. In practice, that means a profile flaw is not an opinion — it is a rejectable condition with a defined fix. If you consistently fight these shapes, slow your travel, watch your work and travel angles, and check your stickout. Clean practice habits pay off here; our guide to welding practice covers the reps that make profile control automatic.

Cracks, Slag Inclusions, Spatter, and Arc Strikes: The Rest of the Rogue’s Gallery

After covering the more common surface and fusion flaws, a few remaining defects can be the most unforgiving for a MIG welder. Cracks, slag inclusions, spatter, and arc strikes may not always be the first things you notice, but each can ruin an otherwise solid bead. Knowing how they form is what separates a rushed repair from a proper fix, and it is a skill that builds with consistent welding practice.

Cracks: The Only Defect With a Zero Tolerance Policy

Out of every flaw in the rogue’s gallery, cracks are by far the most severe. AWS D1.1 prohibits any crack, regardless of size or location, and AWS B1.11 lists the most common types as longitudinal, transverse, throat, crater, toe, and underbead cracks. These are not cosmetic marks, but stress concentrators that can propagate under load.

Cracks fall into two main categories: hot cracks, which form as the weld cools through its solidification range, and cold cracks, which are often hydrogen-induced and can appear hours after welding. Crater cracks are especially common in MIG due to abrupt arc termination, while toe and underbead cracks tend to stem from high restraint or improper preheat. With that said, no crack is safe to leave unaddressed.

Quick Note: AWS D1.1 5.26.1.4 requires the removal and repair of cracks in the weld or base metal. Grinding out the crack entirely before rewelding is non-negotiable.

Slag Inclusions: Rare for MIG, but Still Possible

Slag inclusions are non-metallic solid material trapped within the weld metal. They are normally associated with flux-based processes, such as FCAW, SMAW, and SAW, but TWI notes they can also occur in MIG welding. The Australian Welding Institute adds that slag may become entrapped in cavities or in excessive undercut from preceding weld runs.

This is most relevant when comparing processes. For example, with gasless MIG welding, the flux-cored wire generates its own slag, so proper cleaning between passes and correct polarity for flux core welding are vital. For solid-wire MIG, slag inclusions are far less likely, but poor joint prep can still trap debris. When in doubt, clean thoroughly and inspect multi-pass beads.

Spatter and Arc Strikes: Small Marks With Big Consequences

Spatter is the molten metal droplets that spray around the weld. While often seen as a cosmetic issue, heavy spatter can interfere with coatings, create sharp edges, and increase cleanup time. It is usually a sign of incorrect voltage, wire feed speed, or stick-out. Keeping your contact tip in good condition and dialing in your settings can drastically cut it down.

Arc strikes, on the other hand, are stray arcs that strike the base metal outside of the intended weld zone. These create localized heating and hardening, forming metallurgical notches that act as crack initiation points. AWS D1.1 5.29 addresses arc strikes, and they must be removed and repaired when they occur. This is especially important in code work, as they are treated with the same level of scrutiny as other defects.

Quick Note: Always protect yourself and your surroundings. Stray sparks and spatter reinforce the need for proper protective eyewear for welding and a well-planned welding facility.

Prevention at a Glance

The best way to avoid these flaws is to be proactive. Use the table below as a quick reference for prevention without overcomplicating your process:

Defect	Prevention Tip
Cracks	Proper crater fill, control preheat/interpass, avoid abrupt stops, and remove any existing cracks before rewelding.
Slag Inclusions	Clean between passes, use correct polarity, and maintain proper joint prep, especially for flux-cored wires.
Spatter	Dial in voltage and wire feed, check stick-out, and replace worn contact tips.
Arc Strikes	Keep the torch controlled, avoid dragging the electrode, and grind and repair any stray strikes per code.

These defects round out the catalog, but the takeaway is simple: cracks demand zero tolerance, and the rest are preventable with clean habits and correct settings. Mastering these checks is what will help you consistently produce good and bad MIG welds you can trust.

What Causes Bad MIG Welds? Settings, Technique, Gas, and Equipment

Every defect you spot traces back to one of four buckets: the machine settings, your technique, the shielding and cleanliness of the joint, or the condition of the equipment itself. When a bead goes wrong, stop guessing and work the list top to bottom. Most problems announce themselves in the first few seconds of an arc. Change one variable at a time, because jumping on three knobs at once leaves you unable to tell what actually helped.

Parameter Problems: Voltage, Wire Feed Speed, Travel Speed, Inductance

Voltage sets arc length; wire feed speed sets current. Get them out of balance and the bead tells you immediately. Voltage that is too low usually produces a high amount of spatter, while voltage that is too high makes the arc hard to control and promotes undercut. Crank wire feed speed too high and the wire stubs into the plate, shoving the gun backward; too low and the arc struggles to establish, leaving a high-crowned bead with weak penetration.

Your ears diagnose as well as your eyes. At low amperages in short-arc transfer the arc should give a steady buzz; in spray transfer at high amperages it crackles; a steady hiss means voltage is too high.

What you hear	What it means
Steady buzz (short arc, low amps)	Voltage and wire feed speed balanced
Crackling (spray, high amps)	Healthy spray transfer
Steady hiss	Voltage too high
Popping, sputtering	Voltage too low or a poor ground

Travel speed and inductance finish the picture. Too fast and you get a thin, ropey bead with cold lap; too slow and you pile up a wide, flat bead prone to burn-through. Inductance fine-tunes how fluid the puddle becomes in short-circuit—turn it up for a wetter, flatter bead, down for a crisper, cooler one.

Technique Problems: Angle, Push vs. Pull, CTTWD, Weave

Keep the travel angle at 10–20 degrees from perpendicular (70–80 degrees from horizontal) whether you push or pull, per Executive Filler Metals. Pushing gives a flatter bead and better visibility; pulling drives more heat into the root. Keep the side-to-side angle near zero, or the arc washes one wall and undercuts the other. The bigger culprit is usually contact-tip-to-work distance (CTTWD): hold 3/8 to 1/2 in for short-circuit and 5/8 to 1 in for spray transfer. Stretch it out and voltage drops, spatter rises, and penetration suffers.

Weaving too wide or too fast lowers heat input, which shows up as cold lap and poor root fusion. If you’re new to the gun, a structured reference like our complete beginner’s guide to using a wire welder walks through stance, gun angle, and travel speed in order.

Shielding and Cleanliness: Drafts, Contamination, Coatings, Gas Quality

Gas problems produce porosity almost every time. A draft is the sneakiest cause: AWS D1.1 prohibits GMAW when wind speed near the weld exceeds 5 miles per hour [8 km/h]. Mill scale, oil, paint, and galvanizing all boil off and contaminate the puddle. Flow rate matters too—too low leaves the arc uncovered, too high creates turbulence that sucks in air. Filler with a damaged, moisture-laden surface can add hydrogen and porosity even when everything else is dialed in. Check the wire itself, since it pays to know what MIG welding wire is made of before you blame the machine.

Equipment Problems: Contact Tip, Liner, Drive Rolls, Cables, Ground

A worn or oversized contact tip makes the wire wander, giving an erratic, spitting arc. A clogged liner drags the wire and causes stutter; drive rolls set too loose slip, too tight they flatten the wire. Kinked cables and a poor work clamp (ground) starve the arc of current, which shows up as weak penetration and random spatter. Long, undersized cables add resistance no setting can fix, and clamping ground to rust or paint is a classic cause of a weak arc. Note that these causes overlap—a worn tip lowers voltage, which mimics a settings problem.

Quick Note: Work the buckets in order: confirm parameters, then clean up technique, then gas and cleanliness, then equipment. Nine times out of ten the defect disappears before you reach the last bucket.

How to Read a Bad MIG Weld: Self-Inspection, Testing, and Diagnosis

Good MIG welders inspect their work constantly, not just once. The best habit you can build is to become your own inspector, because a weld that passes the eye test can still hide flaws that only testing reveals. The trick is sequencing your checks so you catch problems at the cheapest moment — fixing fit-up before you strike an arc costs seconds, while cutting out a bad weld costs hours.

The Three Stages of Visual Examination

AWS B1.11 structures visual examination into three phases: prior to welding, during welding, and after welding. Each stage catches different problems.

Before: Check joint fit-up, cleanliness, wire condition, gas flow, and the hi-lo mismatch between pipe ends.
During: Watch arc stability, listen to the sound, and confirm the puddle wets both toes.
After: Examine bead profile, toe lines, undercut, spatter, and any visible porosity.

Remember the core limitation: visual inspection is confined to exposed surfaces, so subsurface discontinuities may not be detected. A pretty bead is not proof of fusion. For a side-by-side look at what you’re actually reading, see Good and Bad MIG Welds: How to Tell the Difference.

Quick Note: Appearance is not evidence of fusion. A flawless-looking bead can still fail a bend test, so verify before you trust.

Tools of the Trade

AWS B1.11 lists examination aids you should keep within reach: a multipurpose gauge, taper gauge, hi-lo mismatch gauge, fiberscope, ferrite indicator, and tong test ammeter. The multipurpose gauge measures fillet leg size and undercut depth; the taper gauge checks root gaps and weld size; the hi-lo gauge measures internal pipe misalignment. A fiberscope lets you see inside tubes, while a ferrite indicator confirms austenitic stainless isn’t excessively magnetic.

Hold gauges perpendicular to the joint for an accurate read, record what you find, and compare measurements to the code. Undercut depth, bead width, and reinforcement limits are all specified, and “close enough” is not the standard.

Non-Destructive and Destructive Testing

When appearance isn’t enough, test.

Dye penetrant — finds surface-breaking cracks and porosity on non-magnetic metals.
Magnetic particle — detects surface and near-surface flaws in ferromagnetic steel.
Ultrasonic — locates internal lack of fusion and voids.
Radiography — X-ray imaging reveals porosity and inclusions.
Bend testing and cut-and-etch — destructive checks that expose fusion at the root and sidewalls.

Lack of fusion and cold lap are the classic invisible killers — TWI explains why they hide inside beads that look flawless.

A Step-by-Step Diagnostic Flow

Listen to the arc. A steady bacon-frying sizzle means good conditions; popping or crackling points to gas or voltage trouble.
Read the bead profile. A ropey, crowned, or flat bead signals the wrong voltage or travel speed.
Trace the toe lines. Sharp, deep toes mean undercut and weak fusion.
Check spatter. Heavy spatter hints at poor gas coverage or wrong polarity.
Test. When in doubt, cut, bend, or scan it.

Sound is underrated. Experienced welders hear the difference between a healthy spray transfer and the sputter of contaminated shielding gas before they ever look at the joint.

Codes Don’t All Agree

Acceptance criteria shift between standards. AWS D1.1 governs structural steel, D1.1 typically capping undercut tightly on thinner material; ASME B31.1 covers power piping; and API 1104 addresses pipeline girth welds. Three porous “bearded” beads pictured by The Fabricator were all judged acceptable under those three criteria. Yet the limits still differ: ASME B31.1 porosity cannot have dimensions greater than 3/16 in, while API 1104 porosity cannot exceed 1/8 in in diameter.

The lesson: know which standard governs your job before you pass or fail your own work. Diagnosis is a loop, not a verdict — identify the flaw, trace the cause, change one variable, and weld again. Reading your welds critically is a skill, and consistent practice is how you sharpen it.

Turning Bad Welds Into Good Ones: Prevention, Practice, and Process Choices

Catching a bad weld is useful. Preventing one is better. Almost every defect you have read about in this guide traces back to a controllable variable, and the fix usually costs nothing more than discipline.

Start With a Weld Procedure Specification

A Weld Procedure Specification (WPS) is your recipe: wire type and diameter, amperage, voltage, travel speed, gas mix, and joint design, all written down because the combination was proven to produce sound welds. If your shop follows one, follow it exactly. If it doesn’t, a qualified procedure is the cheapest insurance you can buy. The AWS D1.1 structural welding code and the AWS B1.11 guide for visual inspection both exist for a reason: qualified parameters remove guesswork. Set your machine to the spec, log the settings, and only adjust when you have a documented reason.

Prepare, Tack, and Clean

Most cold lap and porosity are born before the arc even starts. Mill scale, oil, rust, paint, and moisture all sabotage fusion and gas shielding. Grind to bright metal, wipe down with a solvent, and check that joint faces are square and gapped correctly. Tack in a sequence that lets heat draw the parts together rather than pull them apart. A loose, poorly tacked joint moves under the gun, and a moving joint is a fusion defect waiting to happen.

Quick Note: A written WPS, bright metal, and tight tacks prevent more defects than any amount of repair welding.

Build Skill With Deliberate Practice

Settings get you close; technique finishes the job. Consistent travel speed, a steady 10 to 15 degree push angle, and watching the arc sit at the leading edge of the puddle rather than racing ahead are all learned habits. The fastest way to build them is short, focused reps with a scorecard: run beads, cut them open, and grade yourself against the defects you now know how to name. Our guide to building real welding skills through deliberate practice breaks that routine into drills you can run in an afternoon.

When to Change the Process

Sometimes the cure is not better technique but a different process. Short-circuit transfer is prone to cold lap on thick material and out-of-position work. Switching to spray transfer on thicker steel gives a hotter, more penetrating arc that dramatically reduces lack of fusion. If you weld outdoors or on dirty metal, gasless MIG welding with flux-cored wire tolerates wind and contaminants far better than solid wire. For thin sheet and positional welds where heat input must stay low, pulsed MIG delivers fusion without burn-through, and choosing your electrical setup correctly matters too: polarity for flux-cored wire differs from solid-wire MIG.

Control the Fumes at the Source

Prevention is not only about the weld. Welding fume is a real health hazard, and manganese is the constituent that concerns researchers most. OSHA 1910.252 requires local exhaust or general ventilation sufficient to keep toxic fumes, gases, and dusts below their maximum allowable concentrations. The NIOSH welding program recommends controlling total fume to the lowest feasible concentration while meeting the exposure limit for every constituent. For manganese, the NIOSH REL is 1 mg/m³ as a TWA and 3 mg/m³ as a STEL, while the OSHA PEL is 5 mg/m³ as a ceiling. Those numbers matter because studies indicate welders exposed to low manganese levels (under 0.2 mg/m³) still perform more poorly on tests of brain function and motor skills.

To hit those limits, capture fume at the source. A NIOSH health hazard evaluation notes that ACGIH recommends a capture air velocity of 100 to 200 feet per minute at the point where fume is generated. Position the hood close, keep the shop’s air moving, and set up the space correctly from the start using our guide to planning and setting up a safe welding shop.

None of this works without repetition. Run the same joint fifty times, inspect every one, and change a single variable at a time. Deliberate practice with honest feedback is the single biggest fix for bad welds, and it beats buying new equipment every time.

Conclusion: Every Bad MIG Weld Is a Lesson You Can Read

Think back to that smooth, shiny bead that peeled off the plate with a single tap of a chipping hammer. That image is this entire guide in one frame. The weld looked fine; it wasn’t. MIG welds tend to fail quietly, and by the time a chipping hammer, a bend test, or a working load reveals the truth, the part is often already scrap.

The reassuring part? “Bad” MIG welds are rarely mysterious. Nearly every defect you’ll ever produce traces back to one of five root causes: wrong heat, wrong gun angle, wrong travel speed, contamination, or lost shielding gas. And the fixes are usually just as small. Most remedies are simple adjustments to voltage, wire feed speed, travel speed, gun angle, or gas coverage, exactly as the troubleshooting guides from MillerWelds and Lincoln Electric describe.

Run back through the rogue’s gallery. Porosity pocked the bead with gas holes. Lack of fusion and cold lap hid beneath a shiny surface, and an improperly fused weld creates a weak joint that can lead to structural problems in the finished product (Lincoln Electric). Undercut, overlap, and poor bead profile showed up on the outside. Cracks, slag inclusions, spatter, and arc strikes rounded out the tour. And whether any of those is rejectable depends on the governing code, not on how pretty the bead looks.

So work like a good inspector. Verify before you strike the arc, diagnose defects honestly instead of hoping they’ll pass, and test whenever fusion is in doubt. No inspection program gives an absolute guarantee of suitability, but a structured plan delivers commercially realistic confidence, as AWS outlines in VIW-M. And when something is truly bad, correcting a rejectable discontinuity generally means removing it by grinding, gouging, or machining and rewelding, according to The Fabricator.

Quick Note: A weld isn’t good because it looks good. It’s good because it passes the test your job demands.

Then practice deliberately. Repeat one adjustment at a time rather than chasing luck. Keep a log of settings, and revisit the fundamentals when a defect reappears.

Your action checklist before the next bead:

Clean the joint and ground clamp; contamination is a defect waiting to happen.
Confirm polarity, gas flow, and wire feed speed for the material.
Set a comfortable travel angle and steady speed.
Inspect each pass before stacking the next.
Cut, bend, or test when fusion matters.
Fix confirmed defects by removal and rewelding, never by covering them up.

Every bad MIG weld you cut out is data, not failure. Read it honestly, change one variable, and lay the next bead with intent. Explore our related guides on Good and Bad MIG Welds, Welding Practice, and Overlap Welding Defect for deeper dives into technique, diagnosis, and equipment.

Frequently Asked Questions

What does a bad MIG weld look like?

Bad MIG welds usually show one or more of these signs: pin holes or surface porosity, an irregular or ropey bead, deep undercut along the toes, heavy spatter, a convex ‘piled up’ profile, or a weld that sits on top of the metal instead of blending in. Some severe defects, like lack of fusion and cold lap, can look smooth and clean while hiding a weld that never fused.

Why does my MIG weld have tiny holes in it (porosity)?

Porosity is trapped gas in the weld metal, and the most common causes are inadequate or turbulent shielding gas, drafts, leaks in the gun or hose, contaminated base metal, and too much wire stick-out. Check your flow rate (typically 30-50 CFH), keep the wire extension under 1/2 inch, clean the metal, and block drafts above 5 mph.

Can a bad-looking MIG weld still be strong?

Occasionally, yes — codes permit small amounts of porosity, minor undercut, and modest profile irregularities, and appearance alone does not prove weakness. However, cosmetic clues often correlate with real problems like lack of fusion, so any suspicious weld under structural or fatigue loading should be tested before it is trusted.

How can I tell if my MIG weld has lack of fusion?

You usually cannot see lack of fusion from the surface because the bead can look perfectly normal. That is why inspectors use dye penetrant, ultrasonic testing, radiography, or bend tests, and why cold-lap-prone short-circuit welds on thick material should always be verified with a destructive or NDT check.

Is undercut always a defect?

No — undercut only becomes a defect when it exceeds the acceptance criteria of the governing code. AWS D1.1, for example, typically allows up to 1/32 in [1 mm] undercut on material less than 1 inch thick, with tighter limits of 0.01 in for primary members welded transverse to tensile stress.

Can you repair a bad MIG weld?

Yes, in most cases a rejectable discontinuity can be repaired by removing the flawed metal with grinding, gouging, or machining and rewelding to a qualified procedure. Cracks and lack of fusion generally require complete removal down to sound metal rather than a cosmetic cap pass.

Why does my MIG weld look ropey or like a caterpillar?

A ropey, humped bead usually signals too little voltage relative to wire feed speed, travel speed that is too slow or too fast, or an unstable arc. Raising voltage slightly, matching travel speed to the puddle, and holding a consistent 10-20 degree travel angle usually smooths the profile.

What causes heavy spatter in MIG welding?

Excess spatter most often comes from voltage that is too low, too long an arc, incorrect inductance, or a dirty contact tip and liner. It can also result from the wrong polarity or from welding on rusty, oily, or coated steel that disrupts the arc.

Will a bad MIG weld fail a code inspection?

It depends entirely on the governing code and the discontinuity. AWS D1.1 rejects any crack outright but tolerates small, dispersed porosity and limited undercut, while ASME B31.1 and API 1104 set different porosity limits. Knowing your acceptance criteria is the only way to judge whether a weld is truly a defect.

How do I practice to avoid bad MIG welds?

Build consistency by running the same joint repeatedly and changing only one variable at a time — voltage, wire feed speed, travel speed, or angle — so you can hear and see the effect. Keep good arc length, clean metal, and correct gas flow, and cut, bend, or etch samples periodically so you learn what good fusion actually looks like inside.