What Makes a Good Weld? The Complete Guide to Weld Quality, Defects, and Inspection

What Makes a Weld ‘Good’? Defining Quality Beyond Looks

You just lifted your hood and peeled off your gloves, staring at a bead that could pass for a good weld — or a bad one. It’s shiny, evenly spaced, maybe even a little hypnotic. But is it actually strong, or just pretty? The honest answer is that appearance tells you far less than you’d like. A good weld is one that meets the strength, soundness, and appearance requirements of its specific application, and those three words carry very different weight depending on what you’re building.

A decorative gate and a bridge girder can both have “good” welds, but the bar for each is wildly different. A cosmetic weld on a handrail might only need to look clean and hold a light load. A pressure vessel weld needs full penetration, documented procedures, and inspection records. Quality, in welding, is defined by fitness for purpose — not by how smooth the ripples look. That’s the first lesson behind telling good welds from bad ones.

Discontinuity vs. Defect: The Distinction That Matters

Here’s the concept that separates a beginner from a professional. A discontinuity is any interruption in the normal structure of a weld — porosity, slag, undercut, inclusions, you name it. Every weld contains discontinuities. Some are microscopic, some are visible, and most are harmless.

A defect, on the other hand, is a discontinuity that exceeds the acceptance criteria of the applicable code or standard. This is why the same tiny bit of porosity can be perfectly fine on a lawnmower trailer and grounds for rejection on a structural beam. It’s not the flaw itself that fails the weld; it’s the flaw’s type, size, and location relative to the standard. Inspectors borrow the definition from bodies like TWI’s guidance on weld imperfections, and acceptance limits come straight from documents like AWS D1.1.

Why Weld Quality Actually Matters

• Structural safety. Welds hold up buildings, bridges, pipelines, and pressure vessels. A single defective weld can cause catastrophic failure.
• Code compliance. Contract work often demands certified welders and documented procedures. Non-compliance means rejected work and, sometimes, legal exposure.
• Cost of rework. Repairing a bad weld costs far more than doing it right the first time — grinding, re-welding, re-inspection, and downtime add up fast.
• Reputation. Fabricators live and die on the quality of their work. Word travels.

Scale matters too. The International Labour Organization estimated roughly 11 million welders worldwide back in 2010. Weld quality is a global safety issue, not just a personal one. And it’s a health issue as well: the International Agency for Research on Cancer upgraded welding fumes from Group 2B to Group 1 (carcinogenic to humans) in 2017 in IARC Monograph 118. Responsible welding includes fume control, which is part of planning a safe shop.

Good Welds Are a System, Not a Lucky Bead

Great welders don’t produce good welds by feel alone. They follow a system:

1. Clean base metal — free of rust, oil, paint, and moisture
2. Correct parameters — amperage, voltage, travel speed, gas flow
3. Proper technique — angle, arc length, manipulation
4. Honest self-inspection — knowing what to look for and fixing what isn’t right

Break any one link and the whole chain suffers. Perfect technique with dirty metal still yields porosity. Perfectly clean steel with the wrong polarity still spatters and lacks fusion. This is why good welds come from deliberate, repeatable habits — the kind you build through structured welding practice — not good luck.

This guide is a complete field guide to building those habits. We’ll break down the anatomy of a good weld, the six defects that ruin them, their root causes, how inspectors judge welds by the book, and what “good” looks like across MIG, TIG, stick, and flux-core. We’ll also cover NDT and destructive testing, the mistakes beginners make, and the quick fixes that save a weld.

The Anatomy of a Good Weld: What to Look For

A good weld announces itself through geometry, not gloss. Before you reach for a grinder, your eyes can read several physical signals that separate a structurally sound bead from one that merely looks tidy. Here’s what to examine.

The Visual Signature of a Sound Bead

Run your eyes along the entire length. A quality bead keeps uniform width and consistent ripple spacing from start to finish. When the bead narrows, widens, or the ripples bunch and stretch, your travel speed or arc length drifted, and the weld’s cross-section changed with it.

The toe, where the weld face meets the base metal, tells the truth about fusion. On a good weld, the toes blend smoothly into the parent material in a gentle concave transition. They should never roll over onto unfused metal. That rolled edge is the classic sign of overlap, a defect caused by too-low current or too-slow travel.

Reinforcement height matters too. A bead that sits too flat may lack the throat thickness the joint needs, while one that stands too proud creates a notch-like stress riser at the toe. Aim for a crown that is visible but modest.

A crater that stops short of full fill invites cracking. Under AWS D1.1, all craters must be filled to the full cross-section of the weld, except the non-effective ends of intermittent fillet welds.

Finally, check size. The leg length on a fillet or the width on a groove weld must match the drawing and the load it carries. A beautiful bead that is undersized is still a rejected weld.

How Each Process Signs Its Work

TIG lays down the tidy stacked-dime pattern most people picture when they hear “good weld” — evenly spaced ripples that stay the same size across the run. That rhythm comes from controlled filler addition and steady travel, and it reflects genuine technique, the kind built through deliberate TIG practice rather than luck.

MIG in spray transfer produces a flatter, smoother bead with fine, tightly spaced ripples and edges that wet deep into the plate. In short-circuit mode the ripples run chunkier, but the rule holds: consistency along the whole length. Unsure whether a bead is sound? A close comparison of good and bad MIG welds makes it obvious.

Stick and flux-cored welds arrive hidden under slag. Chip it off and brush, and the surface underneath should show even, well-spaced ripples with no slag lines trapped along the toes — the same geometry you would judge on gasless flux-cored work.

Profile Rules: What the Code Actually Allows

AWS D1.1 Table 8.1 requires weld profiles to conform to Figure 7.4 and Table 7.9, and it explicitly allows fillet weld faces to be slightly convex, flat, or slightly concave. A flat-faced fillet is acceptable, not a flaw; a mildly concave face can even be ideal on a dynamically loaded joint.

Undercut is where the profile rules get strict. For statically loaded structures, undercut on material less than 1 inch thick must not exceed 1/32 inch (about 0.8 mm). Anything deeper is a reject.

Color Is a Clue, Never a Verdict

Heat tint is a fingerprint of how much heat the metal absorbed. Stainless steel should finish in the straw-to-light-blue range, and a dark, rainbow-tinted surface warns of excess heat and reduced corrosion resistance. Aluminum should stay bright and silver rather than dull grey. If your DC aluminum bead looks smutty, suspect heat and cleaning, not just speed.

But tint only hints at heat input. Two welds can share identical color and wildly different geometry, so color never substitutes for checking width, profile, and toe transition.

A Good Weld Checklist for the Bench

• Uniform width and ripple spacing along the full length
• Toes blend smoothly into the base metal — no undercut, no rolled-over overlap
• Reinforcement visible but modest, not flat, not excessively proud
• Crater filled to the full cross-section
• Weld size matches the drawing and joint requirements
• Slag fully removed before judging stick or flux-core beads
• Heat tint consistent, with no sign of overheating

Print that list, pin it near your bench, and run each weld through it before the next arc. Your eye sharpens fast once you know exactly what it is hunting for.

The Six Defects That Turn Good Welds Bad

Almost every weld that fails inspection fails because of one of six recurring problems. Some are visible to the naked eye, while others hide just beneath a smooth surface until load is applied. Here’s what each defect looks like, what causes it, and the single fix that most often prevents it.

1. Porosity

Porosity shows up as surface pinholes, internal wormholes, or a crater pipe where the arc stops. It forms when nitrogen, oxygen, and hydrogen dissolve into the molten pool and get trapped as the metal solidifies. Much of that hydrogen comes from moisture in electrodes, grease, and oil on the workpiece, and rusty or scaled metal makes it worse. Research from TWI found that as little as 1% air entrainment in the shielding gas produces distributed porosity, while more than 1.5% creates gross, surface-breaking pores. Prevention: clean the joint and verify gas flow before you strike an arc.

2. Lack of Fusion

Lack of fusion appears as cold lap along the toes, missing sidewall fusion, or unfused boundaries between passes. The filler simply didn’t melt into the base metal. Causes include too-low amperage, travel speed that outruns the arc, and the wrong polarity. Getting polarity right matters more than most beginners expect, as our guide to DCEN, DCEP, and AC explains. Left in place, these unfused planes can unzip a joint under cyclic loading, a risk covered well by the Australian Welding Institute. Prevention: raise heat into range, slow down, and aim the arc at the joint root.

3. Incomplete Joint Penetration

Incomplete joint penetration means the weld never reached the root. Root faces stay unfused, leaving a gap that looks fine from above but can’t carry the intended load. It comes from too small a root gap, an excessive root face, insufficient amperage, or an electrode that never reached the bottom. Prevention: check fit-up, set the root gap and land correctly, and confirm heat settings before welding.

4. Undercut

Undercut is a groove melted into the base metal along the weld toe that the filler never filled back in. It’s typically caused by too-high amperage, excessive travel speed, too-long arc length, or incorrect electrode travel and work angles. The resulting notch concentrates stress exactly where the weld meets the plate. Prevention: shorten the arc, ease off the amperage, and hold a consistent drag angle.

5. Overlap

Overlap, sometimes called cold lap, is a ledge of filler that rolls over the base metal without fusing to it. Under AWS D1.1 it is a zero-tolerance, rejectable defect because the unfused toe acts as a stress riser rather than a joint. It usually stems from travel speed too slow for the current, an electrode angle that lets metal pile up, and too little heat at the edges. It’s easy to mistake for a fat, healthy bead until you study the toes. Our guide to the overlap welding defect covers detection and repair in depth. Prevention: speed up and adjust the travel angle so the toes wet in.

6. Cracks

Cracks are the defect inspectors fear most. Hot cracks form while the metal is still solidifying, often along the centerline of a deep or concave bead. Crater cracks radiate from the end of a weld where the operator pulled out too fast. Hydrogen cold cracks appear hours or days later, driven by hydrogen trapped in hard, high-carbon weld metal. All three are sharp-tipped. Prevention: fill craters, control cooling, use low-hydrogen consumables, and preheat thick or restrained joints.

Why Planar Defects Are More Dangerous

Not all defects are equally serious. Porosity is volumetric: it takes up space, so inspectors can measure it. AWS D1.1 limits the sum of visible porosity 1/32 inch or larger to 3/8 inch in any linear inch of weld. Cracks and lack of fusion are planar. They have almost no volume but end in a sharp tip that concentrates stress and can grow under load, which is why they’re treated as rejectable almost everywhere. A weld can look beautiful and still fail on a planar defect.

Why Good Welds Go Wrong: Root Causes of Defects

Every defect has a cause, and nearly every cause falls into one of four buckets: contamination, wrong parameters, poor technique, or bad joint preparation and fit-up. Diagnose the bucket and you can fix the weld for good. Treat only the symptom, and the defect returns on the very next pass. This is the mindset that separates welders who chase problems from welders who prevent them.

1. Contamination: The Quiet Killer

Oil, rust, paint, moisture, and mill scale all break down under arc heat and release hydrogen and gases straight into the molten pool. The result is porosity, poor fusion, and, worst of all, hydrogen-induced cold cracking that may not surface for hours after the part has cooled. Grind or wire-brush the joint back to bright metal before you strike an arc, and never weld over grease or cutting fluid.

Weldable primers are a subtler trap. TWI notes that these coatings typically must stay below about 20 µm thickness or porosity is likely, especially with processes other than MMA.

2. Wrong Parameters

Settings are the second major cause. Too-low current lets the weld pool bridge the root gap without fully penetrating, leaving lack of fusion hiding beneath a tidy-looking cap. Too-high current or too-low travel speed floods the pool ahead of the arc, producing a wide, convex bead with poor penetration or burn-through. Arc length matters just as much: long arcs run hotter and lose shielding, while short arcs stub and spatter. Matching current to electrode diameter is the starting point, and these 3/16 E7018 amperage settings show what that looks like in practice.

Polarity mistakes are parameters too, which is why understanding TIG welding polarity — DCEN, DCEP, and AC — matters whenever the process calls for it.

3. Poor Technique

Technique is about control. The wrong work angle, a wandering travel path, or heavy weaving that outruns the pool all produce inconsistent fusion and weak, uneven beads. Wide weaves also cool at the edges and fail to tie in, another common source of lack of fusion. Steady travel speed, the correct push or drag angle, and controlled manipulation come from repetition, not talent. If your beads vary from pass to pass, the fix is deliberate, structured welding practice rather than simply burning more rod.

4. Bad Joint Preparation and Fit-Up

Fit-up decides half the outcome before the arc even starts. Gaps set too wide invite burn-through; gaps set too tight prevent penetration. Wrong bevel angles, mismatched edge thickness, and misalignment all create stress concentrators and root defects that no amount of good technique can rescue. Check the joint with a gauge, not your eye, and tack it square and true.

Consumables, Gas, and Hidden Variables

Electrodes and wire are consumables that can sabotage a perfect setup. Low-hydrogen SMAW electrodes such as E7018 should be baked (typically 300–350°C for 1–2 hours) after moisture exposure and stored in a heated rod oven to prevent pickup. Damp electrodes are a direct route to hydrogen cracking, so a well-planned shop keeps a working oven on hand. Shielding gas flow is equally unforgiving: too low and you lose protection, pulling in air and nitrogen; too high and the turbulent stream drags atmosphere into the arc. Clean contact tips and liners keep wire feeding stable, too.

Preheat, Interpass, and Arc Blow

On thicker or higher-carbon steels, temperature control is what prevents hydrogen-induced cold cracking (HICC). Preheating slows the cooling rate, reduces hydrogen diffusion, and lowers heat-affected-zone hardness, while interpass limits stop the part from overheating between passes. One more curveball: arc blow, the uncontrolled deflection of the arc in ferromagnetic steels, causes lack of fusion and can often be reduced by repositioning the ground clamp or switching to AC.

Judging Welds by the Book: Visual Inspection and Acceptance Criteria

Pretty beads impress customers. Codes impress engineers. When a weld carries structural load, “looks good to me” is a guess, not a verdict. Professional inspectors grade welds against written acceptance criteria, and the most common yardstick in North America is AWS D1.1 Table 8.1. Learn what’s on that table and you can judge your own work before anyone else does.

AWS D1.1 Table 8.1: The Baseline

AWS D1.1 governs structural steel welding, and Table 8.1 lists the visual acceptance criteria for every weld type. It sorts requirements into clear categories — crack prohibition, fusion, crater fill, weld profiles, undercut, and porosity — because each rule targets a flaw that could become a failure point.

The headline rule is absolute: cracks are prohibited. No size limit, no loading exception, no mercy. The table also requires thorough fusion between weld metal and base metal in all loading categories. Welds that fail these checks are rejected regardless of how neat they look.

A few more rules are worth memorizing:

• Cracks: zero tolerance, anywhere on the joint.
• Fusion: complete fusion between weld and base metal.
• Crater fill: craters must be filled to the full cross-section of the weld.
• Profile: free of excessive convexity, concavity, or overlap.
• Undercut: on cyclically loaded primary members, undercut transverse to tensile stress must not exceed 0.01 in (0.25 mm).

Porosity gets its own limits, since trapped gas pockets shrink the load-bearing area of a joint.

Undersized Welds: The Tolerance Rules

Size matters, but codes allow small shortfalls. In statically loaded connections, a fillet weld with a nominal size of 3/16 in or less may be undersized by up to 1/16 in without correction; a 1/4 in weld may run light by up to 3/32 in. That tolerance doesn’t make an undersized weld safe — it simply means a minor shortfall on a static joint won’t automatically trigger repair.

Timing: Why Some Inspections Wait

Visual inspection isn’t always a “weld it and look” job. For quenched and tempered steels — ASTM A514, A517, and A709 Grade 100/100W — inspection must be delayed at least 48 hours after welding. That pause lets delayed hydrogen cracking show itself. Inspect too early and you sign off on a weld that fails overnight.

The Tools of Visual Testing (VT)

AWS CWI practice allows visual inspection to be aided by a flashlight, magnifier, and mirror, and requires suitable measuring tools and gauges. The everyday kit includes:

• Fillet weld gauges — check leg length and throat thickness.
• Undercut and pit gauges — measure depth of undercut and pits.
• Weld size gauges — verify nominal sizes against the drawing.
• Magnifiers and mirrors — reach tight corners and confirm fine detail.

Good light matters as much as good tools, though. Shadows hide porosity and toe cracks that a flashlight beam exposes instantly.

Other Codes: ASME IX and ISO 5817

AWS D1.1 isn’t the only book. ASME Section IX covers procedure and welder qualification for pressure equipment, while ISO 5817 grades imperfections by severity level (B through D), giving fabricators a quality scale rather than a hard pass/fail line. Both share the same logic — cracks prohibited, fusion required, porosity and undercut limited — so once you learn one, the others read quickly.

When Visual Inspection Is Enough — and When It Isn’t

VT can catch surface cracks, poor profiles, incomplete fusion at the toe, and visible porosity. It cannot see inside the weld. Internal lack of fusion, buried porosity, and root cracks need NDT methods like ultrasonic or radiographic testing. Codes decide which joints demand more than a trained eye.

The best welders read the code the way they read a travel speed — as working knowledge. Inspectors don’t have a monopoly on it. Spend time building real welding skills and studying the acceptance tables, and you’ll catch your own defects before a gauge does.

Process by Process: What ‘Good’ Looks Like in MIG, TIG, Stick, and Flux-Core

Every welding process has its own signature bead, rhythm, and common pitfalls. Understanding what “good” looks like for each method helps you spot trouble early and dial in settings with confidence. This section breaks down MIG, TIG, Stick, and Flux-Core, covering technique, polarity, parameters, and the visual cues that separate solid welds from flawed ones.

MIG/GMAW: Controlling Transfer for Clean Fusion

MIG welding is versatile, but the transfer mode changes everything. In short-circuit transfer, the wire touches the puddle repeatedly; at low heat input this often produces cold lap (lack of fusion) because the droplet lacks energy to wet the groove faces. Spray transfer uses higher voltage and wire feed to project fine droplets across the arc, giving deeper penetration and a smooth, flat bead with minimal spatter. Pulsed MIG blends both modes for control on thinner metals.

A good MIG bead is uniform in width and height, with consistent ripple spacing and good tie-in at the toes. Watch for gas flow issues (too low causes porosity, too high causes turbulence) and pay attention to your consumables. A worn or undersized contact tip causes erratic arc and poor weld quality; the tip must match the wire diameter exactly. To see the difference clearly, compare examples in Good and Bad MIG Welds and review fundamentals in How to Use a Wire Welder.

TIG/GTAW: Precision Through Polarity and Control

TIG rewards finesse. For steel, DCEN (electrode negative) is standard because it concentrates heat in the workpiece, producing good penetration and a stable arc. For aluminum, AC is required to break up the oxide layer so the filler can bond properly. Getting polarity right is critical; see TIG Welding Polarity: DCEN, DCEP, and AC Explained and AC or DC for TIG Welding Steel for details. Also note DC welding aluminum considerations.

A good TIG weld shows a consistent stack-of-dimes ripple, clean fusion at toes, no undercut, and minimal discoloration when backpurging is right. Filler rod control matters: add rod smoothly into the leading edge of the puddle without dipping the tungsten. Crater filling with a slope-down at the end prevents crater cracks. These are skills you refine through deliberate welding practice.

Stick/SMAW: Low-Hydrogen Discipline

Stick welding is forgiving outdoors but demands discipline with electrodes. E7018 is a low-hydrogen electrode: proper storage (oven) and handling prevent hydrogen-induced cracking. Amperage is crucial — a 3/16-inch E7018 rod typically runs around 140–180 amps depending on position and joint type. For exact settings, consult 3/16 E7018 Amperage: Settings, Polarity & Technique.

A good Stick bead has even ripples, good slag coverage that chips off cleanly, and solid fusion with no undercut or porosity. Remove slag completely between passes to avoid slag inclusions. If the arc is too cold, you get a high, narrow bead with poor penetration; too hot gives a wide, flat bead with undercut and spatter. Diagnose by reading the puddle and bead shape.

Flux-Core/FCAW: Polarity Makes or Breaks It

Flux-core is often called “gasless MIG,” but the processes differ. Self-shielded flux-cored wire runs on DCEN in most cases, while gas-shielded FCAW typically runs on DCEP — running the wrong polarity produces poor penetration and excessive spatter. This distinction is key; learn more at Polarity for Flux Core Welding: DCEN vs DCEP and Gasless MIG Welding. Also see Is Flux Core the Same as MIG?.

A good flux-core weld shows consistent bead profile, good tie-in, and manageable slag. Self-shielded wires tend to produce more spatter and require attention to travel speed and drag angle. Watch for wormholes/porosity from poor technique or contaminated base metal. Parameter mistakes like wrong polarity or excessive voltage show up as convex beads, undercut, or excessive spatter.

Reading Bead Appearance & Fixes

Diagnose from the bead: a ropey, cold bead signals low heat; a flat, undercut bead means too much heat or wrong travel. Inconsistent width points to travel speed issues; porosity often comes from shielding problems or contamination. Addressing these process-specific mistakes early prevents defects. Whether you’re running MIG, TIG, Stick, or Flux-Core, the goal is the same: consistent fusion, proper profile, and freedom from defects.

The Technique Behind Good Welds: Skills You Can Practice

Weld quality doesn’t come from machine settings alone. Two welders can run identical current, wire, and joint prep and still walk away with completely different beads. The gap is technique: the small, repeatable motions your hands make every second the arc is lit. Best of all, every one of those skills is trainable.

The Four Variables You Control

Travel speed, angle, arc length, and puddle manipulation all sit under your direct control, and each one has a predictable cause-and-effect chain. Learn the chains and you can diagnose a bad bead by what it looks like. Change one at a time and the feedback stays readable.

• Travel speed: too fast for the welding current and the arc outruns the molten metal, raising the risk of lack of fusion; too slow and the weld pool floods ahead of the arc, producing non-uniform penetration and piled-up reinforcement.
• Angle: an incorrect electrode or gun angle prevents adequate fusion of the joint sidewall even when your amperage and travel speed are correct. Work angle aims the arc at the joint; travel angle tilts it along the direction of motion.
• Arc length: too long an arc increases the risk of preferential melting along one side of the joint and causes shallow penetration. Keep it tight and stable, and the arc concentrates heat where you want it.

Weaving, Dwell, and Reading the Puddle

On multi-pass or open-root work, weaving with a slight dwell at the joint sidewalls lets the weld pool wash into the parent metal and greatly improves sidewall fusion. The dwell is the point of the weave, not the sweep. Pause a beat at each toe, then move.

The puddle is your feedback display. Watch its shape, size, and how it wets the plate edges. A puddle that sits high and sluggish is usually too cold or too fast. One that runs liquid and sagging is too hot or too slow. Reading that signal in real time is what separates a welder who reacts from one who just hopes. Adjust as you go, not after the pass is over.

Body Positioning and Stability

Technique collapses without stability. Brace your forearms, rest your elbows, and lock your wrists so the only movement is intentional. Position yourself so you can see the joint and the puddle, not so you’re reaching or twisting. When the arc drifts, it’s usually your body, not your hands. Set up the welding facility so your table height and lighting let you weld comfortably, and protect your eyes with proper protective eyewear for welding.

Practice That Actually Builds Skill

Real skill is consistency: repeating the same good weld, the same way, over and over. That takes deliberate practice, not just seat time.

1. Beads on plate. Run long, straight stringer beads at a fixed travel speed and arc length. Aim for uniform width, height, and ripple spacing.
2. Timed exercises. Weld a set length in a set number of seconds. Vary only one variable at a time and note the result.
3. Coupons. Move to T-joints and butt joints, then cut, etch, and inspect them.

Before you grind a bead away, critique it. Compare toe lines, spatter, and bead profile, then write down what you’d change. That self-review loop is where improvement actually happens. It’s the same mindset covered in depth in our welding practice guide, which walks through drills and progression.

Consistency is the real definition of skill. Anyone can land one beautiful bead; a professional lands the same bead fifty times in a row, at the same speed, with the same profile. Chasing that repeatability, not a single perfect pass, is what makes good welds a habit.

Beyond the Eye: NDT and Destructive Testing of Welds

Visual inspection is fast, cheap, and surprisingly effective — but it only ever sees the surface. Discontinuities like lack of fusion, incomplete penetration, and internal porosity can hide beneath a flawless-looking bead. A weld can pass the eye test and still fail in service. For structural, pressure, and cyclically loaded work, codes demand more: methods that look inside the metal without cutting it apart.

Non-Destructive Testing Methods

Radiographic testing (RT) fires X-rays or gamma rays through the weld onto film or a digital detector. Dense material blocks the beam; voids let it through, appearing as dark spots on the image. RT is normally more effective at detecting and characterizing porosity, especially in thicker sections. Its blind spot is orientation: planar defects like lack of fusion may give weak or no indication when the beam isn’t perpendicular to the defect plane. That limitation is one reason codes often pair RT with UT on critical joints.

Ultrasonic testing (UT) sends high-frequency sound waves into the weld and listens for returning echoes. It’s the most reliable method for detecting planar defects such as lack of fusion, which can slip past radiography entirely. Technicians use angled shear-wave probes to steer the beam through the joint, so defects that run parallel to the surface still reflect sound back. UT also needs access to only one side and handles heavy sections well, which makes it a workhorse on thick-walled pipe and heavy structural steel.

Magnetic particle (MT) and dye penetrant (PT) testing catch surface-breaking discontinuities — toe cracks, crater cracks, and other flaws visual inspection can miss. MT suits ferromagnetic steel; PT suits non-magnetic materials like stainless and aluminum, drawing penetrant into tight cracks that bleed back out as a visible indication. Both are quick and inexpensive, which is why they’re common first-line checks on site.

Destructive Testing: Qualifying Welders and Procedures

Non-destructive methods prove a weld is acceptable right now; destructive tests prove a welder and a procedure can reliably produce acceptable welds. The coupon is sacrificed so the joint’s real behavior can be measured under controlled conditions, rather than guessed at from an image or an echo.

• Guided bend tests bend the coupon around a former to stretch the root and face, exposing lack of fusion or poor ductility.
• Tensile tests pull the specimen apart to measure strength.
• Macro-etch examination cuts, polishes, and acid-etches a cross-section to reveal penetration depth and internal defects.
• Nick-break tests notch and break the coupon to expose fusion flaws along the root.

Under AWS D1.1, welders must be qualified for the process, position, and procedure they actually use — typically verified through guided bend and visual testing of test coupons.

If you’re a hobbyist welding a gate or a trailer repair, you don’t need an X-ray budget. Disciplined visual inspection plus bend-testing your own practice coupons is a realistic, hands-on quality check.

Common Mistakes Beginners Make (and the Quick Fixes)

Almost every bad weld traces back to a handful of habits. Each mistake has a tell, and most fixes take seconds. Bookmark this troubleshooting table, then read the notes for the field checks that catch problems early.

Two Field Tests Worth Memorizing

The fingernail test. Run a fingernail along the weld toe, where the bead meets the base metal. If your nail catches in a groove, you have undercut. Some undercut on a hot bead’s edges is normal, but a distinct notch can blow past code limits — Table 8.1 in AWS D1.1 caps undercut depth by material thickness. If it snags, ease off the amperage, tighten your travel angle, and add filler to wash the toes in.

The ring test. Tap a finished weld or the part itself with a light hammer and listen. A sound joint rings clear; a dull thud hints at a crack, lack of fusion, or a cold lap under a tidy surface. It’s a screening tool, not a verdict.

The Fixes That Matter Most

Cleaning is the cheapest insurance you have. Rust, oil, paint, and moisture on joint faces are among the most common causes of porosity, slag inclusions, and lack of fusion, and grinding to bright metal knocks out most of them in seconds. TWI’s guide to weld porosity confirms contamination acts like a direct fuel line for gas defects.

Gas flow is the sneaky one. Too much shielding gas creates turbulence that entrains air and causes porosity, and a leak in the gas line is a frequent hidden cause. Set flow at the low end of the manufacturer’s range and bubble-test the hose.

Never walk away from an unfilled crater. It is a prime site for crater cracks and shrinkage pipes. Use a slope-down or crater-fill function on your machine, or add filler as you stop so the crater fills to full cross-section.

Whipping the arc too far ahead of the puddle spreads heat and invites fusion flaws — the same root problem behind overlap, a defect that hides under a rounded edge. Check your consumables too: a worn contact tip or clogged liner causes erratic arcs and porosity, which our contact tip troubleshooting guide covers in detail. For a visual gut-check, see good and bad MIG welds.

Conclusion: Making Good Welds a Habit

By now the thread running through this guide should be obvious. A good weld is never a lucky accident. It is the predictable result of a repeatable system, not a burst of talent. The welders who consistently turn out sound beads are not gifted, they simply run the same disciplined loop on every joint.

A Routine You Can Repeat

1. Prepare the joint. Clean metal, correct fit-up, proper gap, solid tacks. Most defects are born long before the arc strikes.
2. Set from data, not memory. Take parameters from the WPS or the manufacturer’s chart. Polarity feeds into that choice, whether you’re dialing in TIG polarity or checking DCEN vs DCEP for flux core.
3. Weld with disciplined technique. Steady travel speed, consistent work angle, controlled arc length. This is the skill you build through deliberate welding practice.
4. Inspect honestly. Read your own bead the way an inspector would. Don’t grade on effort, grade against the acceptance criteria.
5. Fix the root cause. When a defect shows up, ask why. A worn contact tip or the wrong polarity is the real problem, not the ugly bead you’re about to grind out.

Notice the pattern: every step feeds the next. Good prep makes good technique easier, good technique makes inspection boring, and honest inspection feeds better prep. Break one link and defects creep in. Fix the system and “good” becomes your default.

Good Welds Come From Healthy Welders

None of this matters if you can’t weld tomorrow. Welding fume is a real health hazard, and OSHA recommends local exhaust ventilation, positioning yourself upwind of the plume, and respiratory protection when ventilation can’t bring exposure down to safe levels. NIOSH sets a recommended exposure limit of 1 mg/m³ as a TWA for manganese in welding fume, and neurological effects have been reported in studies below 0.2 mg/m³. A WHO/ILO systematic review and meta-analysis estimated that occupational welding fume exposure raises lung cancer incidence risk by about 48% (RR 1.48, 95% CI 1.29–1.70).

So invest in the unglamorous parts. Set up a safe shop with genuine extraction, and wear proper protective eyewear and PPE every time, not only when you remember.

Then practice, and self-inspect every bead. The gap between a welder who guesses and one who knows closes with each honest assessment. Log what worked, study what failed, and the routine hardens into a habit. Do that, and good welds stop being something you hope for and become something you expect.

Frequently Asked Questions

What does a good weld look like?

A good weld has a uniform bead width, evenly spaced ripples, and toes that blend smoothly into the base metal with no undercut, overlap, or visible porosity. The crater is filled, the reinforcement is consistent, and the weld meets the size and profile requirements of the applicable code or drawing.

How can you tell if a weld is strong without testing it?

You can’t guarantee strength from looks alone, but a weld that is free of cracks, lack of fusion, undercut, and excessive porosity and that has full penetration is far more likely to be sound. For critical joints, strength is verified with bend tests, tensile tests, or NDT such as radiography and ultrasonic testing.

What is the difference between a discontinuity and a defect?

A discontinuity is any interruption in the typical structure of a weld, such as a small pore or a slight ripple — nearly all welds have some. A defect is a discontinuity whose type, size, or location exceeds the acceptance criteria of the governing code, making the weld rejectable.

What are the most common welding defects?

The most common defects are porosity, lack of fusion, incomplete joint penetration, undercut, overlap, slag inclusions, and cracks. Porosity and undercut are usually caused by contamination or wrong parameters, while cracks and lack of fusion are the most dangerous because they act as sharp stress risers.

Is porosity always a rejectable defect?

No. Minor porosity is acceptable up to the limits in the governing code — for example, AWS D1.1 allows visible porosity totaling no more than 3/8 inch per linear inch in statically loaded welds. However, cracks and lack of fusion have zero tolerance, and porosity in critical joints like transverse tension butt welds is not permitted.

What causes undercut and how do you fix it?

Undercut is a groove melted into the base metal at the weld toe, usually from too-high amperage, too-fast travel speed, too-long arc length, or an incorrect electrode angle. Fix it by lowering heat, slowing down, shortening the arc, and directing filler into the toes; minor undercut can be repaired with a light corrective bead if the code permits.

Why do my MIG welds have holes in them?

Holes are porosity caused by gas trapped in the solidifying weld pool. Common culprits are contaminated base metal (oil, rust, paint), moisture in the wire or gas line leaks, and shielding problems — either too little gas flow or too much flow creating turbulence that pulls in air.

Do I need NDT on every weld I make?

No. For most non-critical hobby and light fabrication work, careful visual inspection plus destructive testing of practice coupons (like bend tests) is a realistic quality check. NDT such as radiography or ultrasonic testing is required by code for critical welds in pressure vessels, bridges, and cyclically loaded structures.

How do I know what amperage to use for a good weld?

Start from the electrode or wire manufacturer’s recommended range for the material thickness and joint, then fine-tune by reading the puddle and bead. Too low gives poor penetration and a ropey bead; too high causes undercut, spatter, and burn-through. A qualified welding procedure specification (WPS) locks in the correct range for production work.

What is the most important habit for producing good welds?

Cleaning the joint before welding. Contamination — oil, grease, rust, paint, and moisture — is the root cause of porosity, lack of fusion, and slag inclusions, and no amount of skill can overcome a dirty joint. The second most important habit is honestly inspecting every bead and fixing the root cause instead of grinding and re-running.

Authoritative Sources & Further Reading

• https://pubs.aws.org/Download_PDFS/D1.1-D1.1M-2025WebPV.pdf
• https://open-exam-prep.com/study-guides/aws-cwi-endorsement-d1-1/visual-inspection-clause8/table-8-1-visual-acceptance
• https://www.twi-global.com/technical-knowledge/job-knowledge/defects-imperfections-in-welds-porosity-042
• https://welding.org.au/knowledge/fabrication-defects-fusion-and-porosity/
• https://www.kobelco.co.jp/english/products/welding/pdf/2015_KOBELCO_Defect.pdf
• https://www.osha.gov/sites/default/files/publications/OSHA_FS-3647_WELDING.pdf
• https://www.cdc.gov/niosh/welding/about/index.html
• https://stacks.cdc.gov/view/cdc/230050/cdc_230050_DS1.pdf
• https://www.frontiersin.org/journals/public-health/articles/10.3389/fpubh.2022.990547/full
• https://pmc.ncbi.nlm.nih.gov/articles/PMC10956262/