Bad TIG Welding: Common Defects, Causes, and How to Fix Them

A fabricator once delivered a set of polished 304 stainless handrails that looked flawless from the shop floor — until the customer’s inspector pulled a borescope inside the tube and found the root of every weld blackened, crusted, and granular like burnt sugar. The welds passed a squint-test from the outside and failed where it mattered, and nobody caught it until the parts were on site. That gap between ‘looks fine’ and ‘actually bad’ is exactly what this guide closes.

What “Bad TIG Welding” Actually Means (And Why It Matters)

Bad TIG welding is any weld that fails the visual, dimensional, or service requirements it was supposed to meet. That is the plain answer to the question people actually type into search: a bad TIG weld is not “a weld that looks a bit ugly.” It is a weld that cannot pass inspection or cannot do its job — and often both at once. Everything else in this guide, from porosity to sugaring, is just a specific way of arriving at that outcome.

The phrase matters because “bad” is vague, and vague is dangerous in a shop. Welding codes are not vague. They draw hard lines between what an inspector is allowed to note, what they are required to measure, and what they must reject outright.

Indication, Discontinuity, Defect: The AWS Distinction

The American Welding Society terminology, as explained in training material from Michigan Tech’s welding defects webinar, separates three words that casual conversation uses interchangeably:

Term	What it means	Is it rejectable?
Indication	Anything detected during inspection — a mark, a signal, a color change. It reports a condition; it does not yet judge it.	No, not on its own
Discontinuity	An interruption in the normal structure of the weld or base metal: porosity, undercut, lack of fusion, a crack. Real, measurable, present.	Not automatically
Defect	A discontinuity that fails to meet the applicable specification or code, or that makes the weld unsuitable for service.	Yes — it must be removed or repaired

The practical consequence is that the same 1 mm undercut can be a non-issue on a non-critical bracket and a mandatory repair on a pressure component. “Bad” is not a property of the weld alone; it is the weld measured against a standard.

Why Bad TIG Welds Cost More Than Bad Welds Elsewhere

TIG (the nonstandard term for gas tungsten arc welding, or GTAW) is selected precisely where appearance and quality both count: food and pharmaceutical tubing, aerospace components, precision thin-sheet fabrication, and decorative architectural work. If a job was quoted as TIG, the customer already signaled that the weld would be visible, inspectable, and accountable. A failed TIG weld is therefore usually also a failed deliverable — the part comes back, the weld map gets marked up, the schedule slips.

TIG also hides nothing. There is no slag covering porosity, no flux buffering contamination, no coating evening out a lumpy bead. The arc is thin, unfiltered, and honest, which is why aerospace TIG welding procedures are built around documentation and technique discipline rather than cleanup. A cosmetic process produces cosmetic accountability: what you see is what you are judged on, and what you cannot see (root fusion, internal cracks) is what fails in service.

What Bad TIG Welding Actually Costs in the Field

• Rejected sanitary welds. In food and pharmaceutical lines, discoloration on stainless is not a blemish, it is metallurgical damage. Any tint means a non-protective oxide layer has formed, so corrosion resistance in that zone has already dropped — the failure is metallurgical, not cosmetic, as The Fabricator explains. Miller Welds’ food-grade stainless guidance treats these defects as product-safety issues, not surface finish issues.
• Fatigue failure at undercut notches. A small undercut is a geometric stress raiser. Under cyclic load it becomes a crack initiation site, which is why undercut is treated as a structural defect rather than a cosmetic ripple, even when it is only a fraction of a millimeter deep.
• Leaking exhaust and pressure systems. Lack of penetration or a crater crack at the end of a bead shows up as an intermittent leak that no amount of external polishing fixes.
• Repairs that crack again. Rewelding over a contaminated or under-penetrated joint without fixing the root cause simply reproduces the defect, often worse, and burns budget twice.

That gap between “looks fine” and “actually bad” is where projects lose money, and it is the gap the rest of this article closes — first by showing why TIG exposes every mistake, then by cataloguing the defects, tracing them to root causes, and turning each one into a prevention routine.

How TIG Welding Works and Why Small Errors Show So Clearly

Before you judge a bad TIG weld, picture what the arc does in the second it passes over the joint. Gas tungsten arc welding is a stripped-down process: an arc, a piece of tungsten, a shield of gas, and a filler rod fed by hand. The American Welding Society’s AWS A3.0M/A3.0 Standard Welding Terms and Definitions lists “TIG” as a nonstandard term for gas tungsten arc welding and files “heliarc” the same way — trade shorthand that stuck even though it named a brand, not the physics.

The Setup: Electrode, Filler, Gas, Arc

• Tungsten electrode. The arc heats the tungsten until it emits electrons; the electrode itself is not consumed. Arc length therefore changes current density and the heat delivered to the puddle, which is why a stretched arc is a documented cause of porosity, spatter, and electrode deterioration.
• Separate filler rod. Metal is added by hand into the leading edge of the pool, so you control heat input and deposited metal independently — a luxury no other arc process offers, and a responsibility.
• Shielding gas. Pure argon covers most steels; argon–helium mixes add penetration on heavier sections. Flow must exclude air without inducing turbulence that sucks atmosphere into the jet. Gas lens nozzles smooth the stream into laminar flow, extending the protected zone well past the cup.
• The arc itself. A thin, intense, unfiltered plasma column doing all the work in a gap you can cover with a fingertip.

The Variables You Actually Control

Five inputs decide almost everything: amperage, arc length, travel speed, torch and work angle, and post-flow. The practical rule of thumb for amperage is roughly one amp per 0.001 inch of thickness, adjusted for position and alloy. Arc length should stay about the diameter of the electrode — closer and you risk touching the puddle; further and the arc spreads, softens, loses penetration, and pulls in air.

Variable	What it controls	When it’s wrong
Amperage	Heat and penetration	Too cold: cold laps, lack of fusion. Too hot: burn-through, undercut, distortion
Arc length	Current density at the pool	Long arc: porosity, spatter, wandering arc, tungsten damage
Travel speed	Time the pool has to wet out	Too fast: lack of fusion, ropey overlap. Too slow: heavy tint, excess buildup
Work/torch angle	Bead shape and gas coverage	Uneven toes, asymmetrical reinforcement, shielding loss ahead of the pool
Post-flow	Cooling under gas after arc-off	Too short: discolored, degraded tungsten and a dull, oxidized crater

Post-flow is the quiet one. Gas keeps flowing after the arc stops to protect the still-hot tungsten and the cooling weld surface. Cut it short and the electrode discolorates and degrades, contaminating your next arc.

Why TIG Hides Nothing

Stick welding lays a blanket of slag over the bead, and processes that carry their own flux do the same: the flux buffers minor contamination and the covering masks what happened underneath until you chip it away. TIG has no such curtain. The finished weld sits exposed from the moment it solidifies, so every pore, ripple, tungsten fleck, and heat-tinted halo is visible on the outside. Because the arc is thin and unfiltered, small process errors land directly in the puddle with nothing to absorb them. TIG is the cleanest arc process and the least forgiving one at the same time.

AC or DC: Why Polarity Decides the Outcome

On steel and stainless you weld on direct current electrode negative (DCEN): electrons flow from the tungsten into the work, giving deep penetration while keeping the electrode relatively cool. Direct current electrode positive is rarely used alone because it drives so much heat back into the tungsten that the tip melts.

Aluminum changes the equation. Its oxide layer melts around 2,075 °C while the metal beneath softens near 660 °C, so melting the base metal does nothing if the oxide film is still solid on top. That is why aluminum requires alternating current: the electrode-positive half-cycle blasts the oxide apart through “cleaning action,” visible as a frosted white band flanking the bead, while the electrode-negative half drives penetration. Balance the AC too far toward cleaning and you erode the tungsten and overheat the plate; too far toward penetration and the oxide survives, leaving contaminated, lumpy welds.

Master these inputs and the defect catalogue ahead stops looking like random failures. Bad TIG welding is almost always one of these variables drifting out of range, then showing up where nothing can hide it.

The Most Common Bad TIG Weld Defects (Visual Identification Guide)

A bead can look smooth and still be junk. This section is the visual vocabulary inspectors, welders, and customers actually use: each common defect in plain language, the cue your eye catches first, and why it gets flagged as rejectable instead of shrugged off as cosmetic. The familiar industry list of imperfections runs to spatter, porosity, cold lap, undercut, lack of fusion, arc strikes, inclusions and several crack types (Michigan Tech’s welding discontinuities webinar), and TIG gives you nowhere to hide any of them.

Porosity and Blowholes

What it is: gas trapped in the pool and frozen as voids. Recognize it: on the surface, clusters of pinholes or larger blown openings, sometimes stretched into elongated “wormholes”; inside the weld, smooth round cavities you cannot see at all. Why rejectable: voids shrink the load-bearing section, act as stress raisers, and on tubing they become a leak path. Surface pinholes show under raking light, but internal porosity is detected only by radiography or ultrasonics (Welding & Welder’s TIG troubleshooting guide).

Undercut

What it is: the arc melted a groove into the base metal along the weld toe, and the filler never filled it back in. Recognize it: a shadowed crescent notch tracing either edge of the bead, most severe on the toes of a fast, over-ampered weld. Why rejectable: it behaves like a pre-cut crack tip, and fatigue failures commonly start right there; codes place strict limits on its depth (Megmeet’s undercut guide).

Overlap (Cold Lap)

What it is: filler laid across base metal that never actually melted. Recognize it: rounded, bulging ripples sitting proud of the plate, with a visible crease at the toe and no blending into the surface. Why rejectable: the joint is bonded only where the arc wetted the metal, so the effective throat is thinner than the bead suggests.

Lack of Fusion and Lack of Penetration

What it is: lack of fusion means the weld never bonded to the sidewall; lack of penetration means the root was never melted through. Recognize it: fusion gaps often hide behind a pretty cap or show as a cold, unfused land at the joint face, while missing root penetration reads as an unfused line or flat, unbroken root on an open-root joint. Why rejectable: both leave the joint carrying far less strength than the drawing intends, and both usually demand radiography, ultrasonics, or a cut-up to confirm.

Cracks: Hot, Cold, and Crater Cracks

What it is: linear separations in the metal, appearing while the weld solidifies (hot), later under residual stress (cold), or right at the stop-off. Recognize it: fine hairline splits, often along the bead centerline or radiating from the crater’s pinhole. Crater cracks are among the most common beginner TIG defects, caused by pulling the arc without building the crater back up (CK Worldwide’s common TIG mistakes). Why rejectable: cracks are sharp, propagating flaws. Almost no code tolerates them.

Tungsten Inclusions

What it is: fragments of the non-consumable electrode broken off when it touches the puddle or filler rod and embedded in the weld. Recognize it: a rough, bright or discolored speck interrupting otherwise even ripples, usually with an arc that wandered or sputtered. Why rejectable: the brittle inclusion shows on X-ray and counts as a rejectable discontinuity, not a blemish (CK Worldwide’s TIG guide).

Excess Reinforcement, Arc Strikes, Spatter, and Distortion

Excess reinforcement is a tall, over-built cap with abrupt toes; it looks generous but creates a stress riser and can mask unfused edges, so profile limits apply. Arc strikes are stray arc scars outside the joint, hard spots that can crack later and must be ground smooth. Spatter is frozen droplet contamination, cosmetic on steel but a corrosion starter when it embeds iron on stainless. Distortion is warped, out-of-tolerance geometry, rejected on dimensions rather than on the weld itself.

Heat Tint and Sugaring

Recognize it: straw, gold, brown, blue, gray or black banding beside the bead, and on the root, a blackened, granular, sugar-crust surface. Heat tint is a visible oxide-thickness record, and the “pretty blue” weld is oxidized, chromium-depleted metal (The Fabricator on stainless discoloration); sugaring is that damage pushed to the extreme. Both are rejectable in food, sanitary, and corrosive service.

Defect	Appearance	Typical inspection
Porosity / blowholes	Pinholes, blown holes, wormholes	Visual; RT or UT for internal voids
Undercut	Notch groove at weld toe	Visual; profile gauge
Overlap / cold lap	Bulging ripples with creased toe	Visual, often magnified
Lack of fusion	Unfused land at joint face	RT, UT, or macro section
Lack of penetration	Unfused or flat root line	Visual on open root; RT/UT
Crater / hot / cold cracks	Hairline splits at stop or centerline	Dye penetrant; RT
Tungsten inclusion	Rough bright speck in bead	Radiography (X-ray)
Heat tint / sugaring	Colored oxide band; crusty root	Visual; borescope; color standard

Material matters too: aluminum prep failures show as frosted zones and oxide-driven porosity, so keep the cleaning routine from Aluminum Cleaner for Welding: How to Prep Aluminum Right close at hand before blaming the arc.

Root Causes of Bad TIG Welds: Setup, Cleaning, and Gas Shielding

Every bad TIG weld has a history. Porosity, undercut, sugaring and tungsten inclusions are the symptoms; the actual cause usually happened minutes or hours before the arc was struck, at the cleaning bench, the gas cylinder, or the parameter dial. Trace the defect back far enough and you will almost always land on the same three failures: the joint was not clean, the materials were not dry, or the weld pool was not properly shielded. Get those three right and most of the defect catalogue simply never gets a chance to form.

Contamination: The Silent Cause Behind Porosity

Oil, grease, mill scale, cutting fluid, paint, and moisture on the base metal all decompose in the arc and dump gas straight into the melt. The same happens when filler rod is stored in a damp corner or wiped with a shop rag that has been used on everything else. Research compiled by Michigan Tech lists contamination, loss of shielding gas, improper material cleaning, and poorly stored filler among the documented causes of porosity, with the remedy being thorough cleaning, dry storage, and gas-leak checks, followed by grinding out and rewelding the affected area (Michigan Tech welding discontinuities webinar).

There is a quieter version of this problem too: cross-contamination. Using carbon-steel wire brushes, grinding discs, or pliers on stainless steel leaves embedded iron particles that become the nucleation site for rust, and it ranks among the most common root causes of premature corrosion failure in stainless fabrications (Miller Welds on food-grade stainless defects). The fix is a dedicated toolkit per material, plus a proper degrease-then-abrade sequence, as covered in Aluminum Cleaner for Welding: How to Prep Aluminum Right.

Gas Shielding: Leaks, Flow Rate, and Drafts

The argon blanket is the only thing standing between molten metal and the atmosphere, so anything that disrupts it shows up as discoloration, rough matte gray surface, or porosity. Loose hose connections, a worn torch O-ring, a cracked gas lens, or a cylinder regulator leaking at the fitting all bleed shielding quality without making a sound. Flow rate matters on both ends: too little and the pool is under-protected, too much and the jet becomes turbulent and actually pulls air in. Environmental drafts are a recognized cause of porosity and poor bead appearance because they strip the shielding envelope away from the pool (Common TIG welding defects and prevention). Even a ceiling fan or an open bay door can undo a perfect setup, which is why purge dams, flow meters, and shielding discipline belong in your AllGas Welding Supply: Gases, Services & Safety Guide checklist.

Tungsten, Parameters, and Technique Errors

Electrode deterioration stems from the wrong electrode type or size, poor grinding technique, overheating, and insufficient post-flow gas (CK Worldwide on common TIG mistakes). A contaminated or wrongly ground tungsten wanders, transfers its contamination to the pool, and can break off into the weld as an inclusion.

Parameters drive the geometric defects. Training sources consistently attribute undercut to welding parameters set too high, inadequate filler addition, incorrect work angle, and excessive travel speed (Megmeet on undercut causes). Push the torch too hot “to keep up,” keep the arc long, or rush the joint, and the arc carves the edge away faster than filler can rebuild it.

Root cause	Defects it typically produces
Oil, mill scale, oxide, or moisture on metal/filler	Porosity, blowholes, discoloration, lack of fusion
Gas leak, wrong flow rate, shop draft	Porosity, sugaring, rough gray bead, heavy tint
Excessive amperage or travel speed, too little filler	Undercut, burn-through, excessive reinforcement
Wrong work angle, skipping back purge	Overlap, asymmetric bead, sugared root
Wrong tungsten, bad grind, no post-flow	Tungsten inclusions, arc wander, electrode deterioration

The practical takeaway: before you change your hand, change your setup. Re-clean the joint, check the gas for leaks, verify flow and draft, inspect the tungsten grind, and only then re-strike the arc. In the next section we look at how these same causes play out specifically on stainless steel, where the evidence is written in color.

Bad TIG Welding on Stainless Steel: Heat Tint, Sugaring, and Corrosion Loss

Stainless steel punishes sloppy TIG work harder than any other common weld metal, because the damage shows up on the surface as color and hides underneath as lost corrosion resistance. A stainless bead can look bright, flat, and perfectly rippled and still be metallurgically ruined. Everything comes down to one variable: how much oxygen touched the metal while it was hot.

Heat Tint: The Oxide Layer You Can Read

Heat tint is a thin oxide film that grows on stainless whenever the surface is heated in the presence of oxygen. Its thickness depends on only two things: the temperature reached and how long the metal sat there with oxygen available. Because the film refracts light differently as it thickens, the resulting color is a visible record of oxide thickness and of how much chromium was pulled out of the surface to build it.

That is the part beginners miss. Chromium is what makes stainless “stainless,” and forming the tint consumes chromium from the layer directly beneath it. Once that subsurface chromium drops below roughly the 10–12 percent threshold, the passive layer can no longer reform. As The Fabricator explains, any visible discoloration next to a stainless weld indicates reduced corrosion resistance, and the color tells you how much oxygen and heat were involved.

Weld color	What it signals	Practical verdict
Silver / bright	Negligible oxide, shielding held	Acceptable for most service
Straw to gold	Light oxide, minor chromium draw	Often tolerable; check the spec
Brown to blue	Thicker oxide, measurable chromium loss	Corrosion resistance is already reduced
Gray to dull black	Heavy oxide, severely depleted surface	Reject; clean, re-passivate, or remove

The “pretty blue” weld is the classic trap. Blue looks attractive, and plenty of shops ship it proudly, but blue means the oxide has grown thick enough to interfere with the passive layer. In chloride service or food handling, that band becomes the spot that pits first.

Sugaring: When the Backside Burns

Sugaring is the extreme version of the same reaction. On a root pass with full penetration, the back side of the joint is molten or near-molten while facing raw atmosphere. Chromium reacts aggressively with oxygen and forms a dull gray, granular crust that looks exactly like burnt sugar. Miller Welds notes that sugaring destroys the passive chromium layer on the root and leaves a rough, porous surface that traps product residue, which is why a sugared weld is an automatic rejection in food-grade and sanitary systems even when the front side looks perfect.

Back Purging Done Properly

Purging replaces the oxygen inside the tube or joint with an inert gas, usually argon, before the arc is struck. Getting it right is a measurement job, not a guess. Industry practice is to reach a residual oxygen level below about 50–100 ppm before arc initiation and to confirm it with a calibrated oxygen analyzer rather than counting minutes of flow. For regulators, hose, and analyzers, suppliers like AllGas Welding Supply carry the components needed to build a repeatable setup.

Good purge practice also includes:

• Purge dams — foam or water-soluble barriers that limit the volume to be purged so you reach target ppm in reasonable time.
• Ventilation strategy — inlet and outlet vents placed so fresh argon sweeps the whole joint instead of short-circuiting around it.
• Protection past the arc — argon post-flow on the back side must continue until the root cools below roughly 400°C, because oxidation resumes the moment shielding stops on still-hot metal.
• Leak checks — a pinhole in the purge line reintroduces oxygen continuously and produces a discolored root in an otherwise “correct” setup.

The MTM research on purge gas and austenitic stainless GTAW welds shows how strongly oxygen contamination in the purge path affects root quality, which is why purged welds are also specified wherever cryogenic toughness matters.

Standards: What Inspectors Actually Check

AWS D18.2 is the standard visual guide for this work: it correlates discrete discoloration levels inside austenitic stainless tube with purge-gas oxygen content in parts per million, giving shops a common language instead of an argument over “looks fine.” Sanitary codes in the ASME BPE family go further and require heat tint on product-contact surfaces to be removed, by pickling and passivation, electropolishing, or mechanical polishing, because tint is classified as a surface defect rather than a cosmetic preference.

Two more numbers keep stainless welds inside spec: hold interpass temperature below about 150°C, and for duplex grades keep heat input inside the commonly quoted 0.5–2.5 kJ/mm window to avoid chromium nitride precipitation. Exceed either one and the same pass comes out discolored, softened, and corrosion-prone.

Stainless rarely fails because of poor bead geometry. It fails from oxygen, heat, and time. Control those three, and the color, the root, and the service life take care of themselves.

Bad TIG Welds on Aluminum and Thin Sheet: Burn-Through, Oxide, and Distortion

Aluminum is where good steel habits go to die. A welder who runs clean beads on mild steel can open an aluminum joint and get a dirty, frosted, crater-cracked mess. Aluminum hides a chemical trap that steel doesn’t, and thin sheet leaves no margin for error.

The Oxide Layer: Aluminum’s Built-In Contaminant

Every piece of aluminum carries a thin skin of aluminum oxide that forms the moment the metal touches air. The problem is one number pair: aluminum oxide doesn’t melt until roughly 2,075°C, while the aluminum underneath melts near 660°C. Weld before cleaning and you melt base metal under a floating sheet of still-solid oxide. The result: a dull pool, stringy beads, and black oxide inclusions that reject under magnification.

This is why surface prep matters more on aluminum than on any other common metal. Degrease to remove oils, then abrade with clean, dedicated stainless brushes or sanding discs immediately before welding; acetone alone does not cut oxide. Our guide on aluminum cleaner for welding covers the full prep sequence and why “clean” means abraded within minutes of striking the arc.

Frosting, Black Snow, and Aluminum Sugaring

Two visual signs tell you the shielding was wrong. A milky white “frosted” zone around the bead is fresh oxide that grew because nothing protected the cleaned surface afterward. Light frost is normal and proves the AC cleaning worked; a chalky white band means excess gas turbulence, too much cleaning action, or a torch held too far away.

The backside equivalent is black oxidation or “sugaring” on aluminum, a sooty, crumbly deposit where the root saw air. It destroys root strength and corrosion resistance and must be ground out; back purging with argon, using the same dam-and-flow setup you would use for stainless, prevents it.

Burn-Through on Thin Sheet

Thin sheet has little heat sink at the joint, and although aluminum’s conductivity pulls heat sideways fast, the melt happens so quickly the arc blows through before it spreads. A bead that suddenly drops a glowing ball of metal into the joint is classic burn-through. Typical causes are amperage too high, travel too slow in one spot, arc length too long, and tacks spaced so widely the gap opens as the joint heats.

Filler Mismatch: 4043 vs 5356

Wrong filler produces cracking that looks like a technique problem but isn’t. ER4043 is the forgiving choice: higher silicon gives a fluid puddle and resists solidification cracking on heat-treatable alloys like 6061. ER5356 carries more strength and holds color better if the part will be anodized, but it is stiffer, less fluid, and more crack-sensitive in restrained joints. Weld a 6061 fixture with 5356 under high restraint and the bead splits at the crater. See our aluminum MIG welding guide for the wider filler landscape.

Distortion: The Defect That Moves the Part

Distortion is one of the most frequently cited weld defects in industry references, including Michigan Technological University’s welding discontinuities training. Poor clamping, uneven heating, and bad sequencing cause it, and the risk is amplified on aluminum and stainless because both distribute heat differently than carbon steel, concentrating expansion near the joint.

Controlling it comes down to discipline:

• Fixturing: rigid fixtures, strong tacks, and weld placement that pulls the part back toward square.
• Heat sinking: chill bars, oversized backup bars, or wetted backings that absorb excess energy.
• Backstep sequencing: short segments welded opposite the direction of travel to balance shrinkage.
• Pulse settings: pulsed TIG cuts average heat input while holding penetration, reducing distortion and burn-through.
• Travel discipline: steady travel with consistent arc length; hesitating to check dumps heat in one spot.

Those same habits govern thin-section work, from alloy wheel repair to aluminum boat welding, where heat control decides whether a hull stays fair.

Aluminum punishes craters harder than steel: every crater crack starts with an arc-off over a shrinking molten pool. Back the amperage off gradually or add two or three filler dabs as you step out, every time. The CK Worldwide guide to common TIG mistakes lists crater failure among the most frequent avoidable rejections, near the top for aluminum.

Clean the oxide right before welding, shield the backside, control heat with clamps, chill bars, and pulse, match the filler to the alloy, and never arc off without filling the crater. Skip one, and a bead that looked fine from six feet away fails its first real inspection.

Beginner TIG Mistakes That Ruin Welds (and the Habits That Fix Them)

Almost every bad TIG weld a beginner produces comes from the hands, not the hardware. CK Worldwide lists porosity, poor gas coverage, and weld discoloration among the most common mistakes new TIG welders make, and each one traces back to a repeatable habit you can correct in a single practice session. Better defaults, not more talent.

Arc Length Drift and the Speed Trap

Two errors start most ruined welds: inconsistent arc length and moving too fast. TIG has no slag or flux to buffer you, so when the torch drifts an eighth of an inch off the joint, gas coverage thins, the arc wanders, and you get undercut on the toes or a bead that never ties into the root. Speed does the same damage from the other side: the puddle never catches up, so filler sits on top instead of wetting in, leaving cold lap and lack of fusion.

The corrective habit: hold a comfortable arc length, roughly the diameter of the tungsten tip, and drive travel with your wrist, not your shoulder. The drill: run beads on flat plate with no filler, judging only two things — steady arc sound and a consistent, mirror-bright bead width. Keep both steady for six inches, then add the rod.

Dipped Tungsten: Stop, Don’t Rip

Touching the tungsten to the puddle is a rite of passage. The reaction that follows is the real mistake: yanking the electrode free under live arc. Stop the current first. If you don’t, the tip breaks off and gets buried in the bead as a tungsten inclusion, which contaminates the weld and usually demands a grind-out and restart. Tearing free under arc also leaves an arc strike — a stray mark outside the joint that scores the base metal and seeds fatigue cracking.

The corrective habit: release the trigger or foot pedal the instant you dip, let the joint cool a second, then regrind the contaminated tip before resuming. The drill: dip deliberately on scrap and recover cleanly until stopping the arc first becomes automatic.

Craters, Stop-Offs, and Shrinkage Cracks

Most beginner cracks aren’t mid-run failures; they’re crater cracks at the stop-off. As the puddle solidifies, shrinkage pulls the last metal apart when the crater wasn’t fed full. Back-feed it with filler, or ramp current down using the machine’s crater timer. Shrinkage cracks starting in unfilled stop-offs are rejectable every time.

Tungsten Grinding Direction

Tungsten ground with random or circular facets throws the arc around: it starts on one high point and wanders, showing up as an erratic puddle and an off-center bead. Grind lengthwise, with longitudinal marks aligned to the direction of the arc, on a dedicated diamond wheel used for nothing else. Keep the grind diameter matched to your tungsten size, and don’t let the tip ball up on contaminated steel.

Filler Hygiene and Post-Flow

Dirty filler rod is a silent porosity generator. Oil, oxidation, and skin residue transfer straight into the puddle, and since TIG offers no flux cover-up, the blowholes are obvious. Wipe rod with clean solvent, store it off the floor, and cut off any end that was dropped or welded over. Skipping post-flow is the twin error: the finished weld is still hot enough to react with air, so starts come out discolored and oxidized. Hold the torch over the crater until the post-flow stops. On aluminum, aluminum cleaning before welding removes the oxide layer that otherwise wrecks the bead.

Mistake	Habit that fixes it	Practice drill
Running amperage too hot “to keep up”	Set heat for the thickest joint, adjust travel	Weld one joint at three amperages; note which holds your rhythm
Inconsistent dab or rod feed	Dip on a steady count, advance rod only when fluid	One dab per half-inch, counted aloud
Welding over contamination	Clean first, always	Mark dirty coupons and refuse to strike an arc until prepped
Skipping post-flow	Hold position until gas stops	Count to three after every stop-off

Note: On thin-wall tubing and chromoly tube joints, low-amp start capability matters more than raw output. High-frequency starts below roughly 5 amps let you establish the puddle without blowing through the wall, which is why builders of thin structures prize machines with good low-end control.

Work these corrections in order and results come fast. For structured feedback on technique, aerospace TIG welding standards and training material are a useful benchmark, since that work demands the same discipline at smaller scale. Fix the habit, and the defect disappears with it.

How to Inspect and Judge a Bad TIG Weld

Judging a TIG weld is not a matter of squinting at the bead and saying it looks nice. Real weld inspection is a sequence of checks with defined acceptance criteria, and it starts long before anyone strikes an arc. The National Board’s basic weld inspection guidance puts it plainly: the inspector examines joint preparation, bevel angles, root gaps, fit-up, and the limits written into the welding procedure before a single pass is deposited. A perfect-looking cap over a root gap that drifted out of tolerance is still a bad weld.

The Three-Stage Visual Check

Visual inspection (VT) is the first and most frequently used method, and it happens in three stages:

• Pre-weld: cleanliness of the joint and filler, bevel geometry, root opening, misalignment, and whether the correct filler alloy and tungsten are staged at the machine.
• In-process: watching arc length, travel speed, filler dip rhythm, shielding coverage, and interpass temperature. Catching a problem here costs a minute; catching it after the fact costs a grind-out.
• Post-weld: bead profile, undercut, overlap, porosity open to the surface, crater cracks, heat tint, spatter, arc strikes, and dimensional checks against the drawing.

For the post-weld pass, inspectors reach for a weld profile gauge or profilometer to measure reinforcement height and leg sizes, a magnifier or borescope for tight areas, and calibrated rulers for length and location. Undercut that feels like a fingernail catch is measured, not estimated.

When the Eye Is Not Enough: NDE Selection

Non-destructive examination is chosen based on where the suspected defect lives and how it is oriented. Subsurface porosity, lack of penetration, and tungsten inclusions require radiographic (RT) or ultrasonic (UT) examination to detect. Surface-breaking cracks, including the crater cracks TIG welders leave when they lift off without filling the stop, are the domain of dye penetrant testing (PT), which can reveal a hairline crack that visual inspection misses under tint or spatter. Magnetic particle testing (MT) applies to ferromagnetic materials, and eddy current testing handles thin non-magnetic surfaces.

Defect type	Typical method	What it reveals
Surface cracks, crater cracks	Dye penetrant (PT)	Breaks hidden by tint or profile
Internal porosity, tungsten inclusions	Radiography (RT), ultrasonics (UT)	Volumetric and subsurface flaws
Undercut, reinforcement, leg size	Visual + profile gauge	Dimensional nonconformance
Root fusion and penetration	RT, borescope, or destructive section	Backside completeness

Acceptance Criteria, Weld Symbols, and the WPS

A weld is not “bad” because someone dislikes it; it is bad because it violates an acceptance criterion. Those criteria come from the applicable code, the drawings and their AWS-standard weld symbols, and the qualified welding procedure specification (WPS) with its supporting procedure qualification record (PQR). The WPS fixes the variables: material group, filler classification, polarity, amperage range, heat input limits, preheat, shielding and purge gas, and position. Straying outside those limits is a procedural rejection even if the bead looks attractive. AWS’s Welding Inspection Handbook treats dimensional inspection, NDE methods, welding metallurgy, destructive testing, and inspector qualification as one integrated quality framework rather than isolated skills.

Witness Welding You Have Not Seen Before

There is one duty that catches inexperienced inspectors out: if you are unfamiliar with a shop’s procedures, equipment, or welders, you have both the right and the responsibility to witness sample or production welds before accepting them. You cannot judge consistency from one finished coupon. Watching the run tells you whether the WPS is actually being followed, whether purge gas is flowing before the arc, and whether post-flow is happening at all. This is the same principle taught in formal certification programs, such as an aluminium welding course, where instructors evaluate the process and not just the finished bead.

Note: Document what you see with photographs, measurements, and the specific acceptance clause violated. “Looks bad” is an opinion; “undercut 0.8 mm exceeds the 0.5 mm limit per the governing code” is an inspection report.

Handled this way, inspection stops being a blame exercise and becomes what it should be: the feedback loop that tells the welder exactly which variable to correct on the next joint.

How to Fix and Repair Bad TIG Welds the Right Way

A repair weld gets more scrutiny than the original, not less. The metal is already heat-cycled, the geometry is already compromised, and the person who authorized the repair now knows exactly where to look. That is why “just run another bead over it” is the most expensive sentence in a fabrication shop. A proper TIG repair is a controlled procedure, and it does not start at the arc.

Step 1: Find the Root Cause Before You Touch the Metal

Grinding out a defective area and rewelding it is the documented repair for porosity, and Michigan Tech’s training material describes exactly that sequence: remove the affected area back to sound metal, then reweld. The condition attached is the entire point: the cause must be identified and corrected first. If the torch line leaked, the base metal was oily, or the filler rod was damp, the replacement weld fails the same way, and you have spent an hour reproducing the problem.

Three questions before the grinder comes out:

• What is the defect? Porosity, undercut, lack of fusion, crater cracks and heat tint each call for different repair methods.
• Why did it happen? Contamination, wrong gas flow, bad tungsten prep, incorrect amperage, rushed travel speed, or a purge that was never really sealed.
• Can the fix actually be reached? Repairing a draft problem is pointless if the purge dam still leaks or the electrode is still contaminated.

Step 2: Excavate to Sound Metal, Then Blend the Edges

Remove the defect completely. Porosity and lack of fusion must be ground out until you reach dense, clean metal with no remaining indication. Cracks need extra margin: grind past the visible crack in both directions, because they routinely travel further than the surface shows, and feather the edges so the replacement pass ties in without creating a fresh stress riser. A dye penetrant check after grinding confirms removal was complete before you commit to the reweld.

Undercut is the gentler case. The standard repair is light grinding to form a uniform groove, followed by a pass that adds enough filler to restore the original profile, as outlined in the practical guidance on undercut repair. The target is a smooth transition, not a proud lump that walks you into an excess-reinforcement rejection.

Step 3: Re-Clean, Then Re-Weld to Procedure

Treat the repaired area like a brand-new joint: degrease, abrade with dedicated tooling, remove grinding dust, and verify gas coverage before striking the arc. Weld the replacement pass to the same procedure as the original, meaning the same filler, parameters and shielding. Aluminum is even less forgiving on surface prep, and our aluminum welding prep guide covers that cleaning discipline properly.

Set a repair limit too. Most specifications allow a defined number of repair cycles at one location, commonly two, after which the part is re-evaluated or scrapped. Repeated thermal cycling in the same spot is how a repair turns into a crack.

Heat Tint Removal: The Expensive Cleanup

Stainless repairs carry an extra cost center. Heat tint is chromium-depleted oxide, not a stain sitting on the surface, and it cannot be scrubbed or passivated away. The metal beneath the tint is genuinely less corrosion-resistant than the base alloy, which is why food-grade work treats heavy discoloration as rejectable, per food-grade stainless welding practice. Removal requires pickling paste, electropolishing, or mechanical polishing with tooling that has never touched carbon steel. And as a practical rule, removing heat tint costs several times more than preventing it with correct purge gas and travel speed, as stainless discoloration analysis makes clear.

When Repair Is Not Allowed: Cut Out and Replace

Every repair has a wall thickness floor. On pressure parts, if grinding the defect out would leave the wall below its minimum required thickness, cutting the section out and replacing it is the only code-compliant option. The same logic applies to thin-wall tubing, members in fatigue service, and aluminum components where grinding can quickly eat into a section with little material to spare, such as the alloy wheel repair scenarios where geometry leaves no margin.

Base metal damage changes the calculus too. Arc strikes in a hardened zone, gouges from careless grinding, or contamination burned deep into the surface are not things a cover pass fixes. In those cases replacement is not the last resort; it is the correct first decision.

Knowing When a Bad Weld Must Be Scrapped

Sometimes the honest answer is that the part does not get saved. Multiple repair cycles in one zone, widespread porosity suggesting a systemic gas problem, cracking that recurs after a technically correct repair, or any defect whose failure would carry unacceptable consequences all point to scrapping rather than another attempt. A scrapped part costs material and labor; a failed repair that ships costs reputation, warranty claims, and potentially someone’s safety.

The through-line is simple: diagnose, excavate, verify, clean, reweld, inspect. Skipping any step converts a one-time fix into a recurring problem, and the best repair remains the one you never had to make.

Prevention Playbook: Settings, Routines, and Practice Drills That Eliminate Bad Welds

By now you know what bad TIG welding looks like and why it happens. Almost every defect in this guide traces back to a small set of controllable variables. Fix the system and the welds follow. This section turns everything above into a shop routine you can run before, during, and after every job.

The Pre-Weld Cleaning Protocol

Degrease first, then abrade. Hit an oily or oily-dusty surface with a stainless brush or an abrasive pad and you don’t remove the contamination, you grind it fresh into the metal and spread it across a wider area. That is where it later surfaces as porosity or a black smear in the puddle. Wipe the joint and the surrounding area with a solvent-rated cleaner, let it dry, and only then scuff with a dedicated abrasive.

Dedicated is the operative word. Keep separate brushes, pads, and grinding wheels for stainless steel, carbon steel, and aluminum, and label them. A carbon-steel brush used on stainless leaves embedded iron particles that rust later; a brush that touched mild steel on aluminum just planted a new corrosion cell. Store filler rod in closed, dry storage, because a rod rolled across a greasy shop floor or left out in humidity is a hidden source of gas porosity later. The same discipline applies to the aluminum cleaning routine, where oxide removal is the entire job.

Gas System Checks

Shielding gas problems cause a disproportionate share of ugly welds, and they are cheap to prevent. Do a quick check every session:

• Leak check. Set the regulator to working pressure, close the cylinder valve, and watch the gauge. A dropping needle means a leak somewhere in the hose or fitting chain.
• Flow rate. Metered at the torch with the arc off, most TIG work sits around 15–25 CFH of argon for a #5 or #6 cup. More is not better; turbulence pulls air in.
• Draft control. Even a light shop breeze defeats argon. Block the airflow or move the job, and use a larger cup or a trailing shield when the part geometry allows it.
• Purge setups. For stainless root passes, use purge dams, verify with an oxygen meter rather than guessing, and, critically, keep the purge flowing after the arc stops. The root stays hot enough to oxidize for minutes, so ending protection early is a documented cause of sugaring. Research on austenitic stainless GTAW shows purge quality directly affects root appearance and toughness, making that extra minute of gas cheap insurance (MTM Inc. purge gas study).

Tungsten Prep Routine

Grind tungsten on a dedicated, contamination-free wheel, always lengthwise so the grind facets run parallel to the electrode axis. Radial or cross grinding makes the arc wander, and wandering means inconsistency. Re-sharpen whenever the tip has melted or been dipped, and remember that gas lens setups let you pull the tungsten further out for tight corners.

Parameter Starting Points

A practical starting rule for TIG amperage is roughly 1 amp per 0.001 inch (1 amp per 0.025 mm) of material thickness on steel, trimmed downward for thin sheet, overhead positions, and high-conductivity alloys like copper. Aluminum needs a bump above the rule because the oxide layer eats heat before the puddle forms. Write your proven settings down per material and thickness; a parameter sheet is the fastest teacher in the shop. Frameworks like AWS D1.6/D1.6M for stainless, ASME Section IX, ASME B31.3, ISO 5817, and ASTM A967 supply the formal acceptance and cleaning criteria behind those numbers.

Variable	Starting Discipline	Defect It Prevents
Amperage	~1 amp per 0.001″ thickness, trim down for thin/overhead	Burn-through, undercut, distortion
Arc length	Keep close; roughly electrode diameter or less	Porosity, wandering, loss of shielding
Travel speed	Slow enough to wet the edges fully	Lack of fusion, cold lap, overlap
Crater fill	Back off current or add 2–3 rod dips before stopping	Crater cracks
Post-flow	5–10 seconds of gas after arc-out	Tungsten oxidation, hot root discoloration
Interpass temp	Monitor and let the part cool between passes	Distortion, hot cracking, heat tint
Tacking/fixturing	Enough tacks to hold shrinkage forces	Distortion, gap opening, misalignment

A Practice Progression for Beginners

Don’t practice on the job. Run a deliberate ladder: straight beads on flat plate, then curved and corner beads, then fillet and lap joints, then tube butt joints with a purge, then thin sheet and out-of-position work. Each rung adds one variable. If you are branching into aluminum specifically, our aluminium welding course overview maps how structured training compresses that curve. The habit that ties it together is the one every disciplined welder shares: stop after each pass, look at what you made, and correct the cause before you strike the next arc.

Conclusion: Bad Welds Are a Systems Failure, Not a Talent Failure

Porosity, undercut, sugaring, tungsten inclusions, crater cracks, none of them are mysteries. They are the predictable output of a specific, identifiable breakdown in cleaning, shielding, parameters, or technique. Bad TIG welding is almost always a preventable systems failure, not a talent failure. Clean the metal properly, verify your gas, set proven parameters, control arc length and travel speed, fill your craters, and inspect as you go. Do that consistently and the borescope will find the same smooth, sound root every time.

Frequently Asked Questions

What is the most common cause of bad TIG welds?

Contamination and poor shielding are the two dominant root causes. Oil, oxide, moisture, or dirty filler rod on the joint, combined with inadequate or disrupted shielding gas, produces most porosity, discoloration, and weld rejection cases beginners and shops encounter.

Why does my TIG weld have little holes in it?

Those holes are porosity — trapped gas pockets that formed as the pool solidified. They are typically caused by surface contamination, moisture, gas leaks, low or disturbed gas flow, or drafts stripping the shielding envelope away from the puddle.

Why is my TIG weld blue or purple on stainless steel?

That color is heat tint, a thickening chromium-oxide layer that records how much oxygen hit the metal and how hot it got. Any visible tint means the surface’s corrosion resistance has been reduced, so most sanitary specifications require it to be removed.

What is sugaring in TIG welding?

Sugaring is heavy, rough, granular oxidation on the root side of a full-penetration stainless weld caused by air reaching the hot metal. It destroys the passive layer and can only be prevented with proper back purging — once formed, it usually has to be cut out and rewelded.

How do I stop getting tungsten in my weld?

Tungsten inclusions come from touching the electrode to the pool or dipping the filler into the arc. Keep a tight but non-contact arc, feed the rod away from the tip, and if you do contaminate the tungsten, stop and regrind it rather than trying to weld through it.

Can a bad TIG weld be repaired, or does it need to be cut out?

Shape defects like undercut and shallow porosity can be ground out to sound metal and rewelded once the root cause is fixed. Cracks and deep internal porosity, or repairs that would thin a wall below its minimum thickness, require cutting out the section and replacing it.

Why do my TIG welds crack at the end?

Those are crater cracks, caused by shutting off the arc while the crater is still shallow. Fill the crater with filler before breaking the arc, or use the machine’s downslope/crater current setting to let the pool solidify slowly and completely.

Is a colorful TIG weld a strong weld?

No. On stainless and titanium, color is a record of oxidation, not penetration. A rainbow or blue bead may be fully fused, but the oxide it carries is nonprotective and less corrosion-resistant than the base metal, so it often fails inspection even when it looks attractive.

How can I practice to stop making bad TIG welds?

Run consistent beads on scrap of the same alloy and thickness you actually build with, one variable at a time: arc length, travel speed, then filler timing. Clean every joint the same way every time — repetition of a fixed setup routine fixes more defects than raw arc hours.