TIG Welding Problems: 12 Common Defects and How to Fix Them

A fabricator once told me he spent three hours TIG welding a stainless handrail, only to find the back side of every joint glowing with grey, crusty ‘sugaring’ that ruined the corrosion resistance. Nine times out of ten the arc looked perfect from the front — the problem was invisible until the part failed inspection. That is the cruel reality of TIG: the process punishes small mistakes in gas, cleanliness and heat control, and it rarely tells you what went wrong until the weld is already cold.

Why TIG Welding Is So Unforgiving (And Why Problems Happen)

You laid down what looked like a decent bead. The arc was steady, the puddle went where you wanted it, nothing sputtered. Then you wiped the part down and the reality hit: porosity, discoloration, a crater crack, or a lump of grey crud welded permanently into your joint. Whether it was your first TIG weld or your five hundredth, the frustration is identical. TIG (GTAW) has a way of making competent welders feel like beginners again.

If you are reading this, you are trying to answer one question: what went wrong, and how do I fix it? That is what this guide is for. We will work through the twelve most common TIG welding problems one by one, diagnose the real cause behind each, and give you the concrete fix: the setting to change, the habit to break, or the cleaning step to stop skipping.

Why GTAW Shows Every Mistake

To troubleshoot TIG effectively, it helps to understand why this process hides nothing. In gas tungsten arc welding, a non-consumable tungsten electrode carries the arc, the welder feeds filler metal into the puddle by hand, and an externally supplied blanket of inert gas shields the molten pool from the atmosphere. Three things follow from that design.

• There is no flux blanket to hide behind. Stick and flux-cored welding melt a coating that produces slag, and slag covers a multitude of sins. TIG gives you a naked puddle and an unobstructed view. That is great for control, but every ripple, undercut and contamination mark is on permanent display.
• The filler rod is manual and uncontrolled. Your filler hand determines bead shape, fill and reinforcement. Dip it in the puddle, touch it to the hot tungsten, or feed it too slowly, and the defect belongs entirely to you.
• The shielding gas envelope is fragile. A bottle of argon does nothing unless it reaches the puddle at the right rate, through a clean line, out of the right cup, with no draft cutting across it. Disrupt that envelope and the molten metal reacts with oxygen and nitrogen almost instantly, producing porosity, sugaring or heavy oxidation.

Contrast that with MIG welding, where the arc sits under a much larger gas lens or self-shielding flux, the wire feed is mechanized, and the process tolerates a fair amount of operator imprecision. Stick welding is similarly forgiving in some respects: the flux generates its own shielding, and hand speed rather than rod-feeding skill dominates the technique. TIG demands that arc length, travel angle, pedal pressure, filler rhythm and gas coverage all be right at the same moment. When one slips, the weld tells you immediately.

The Five Symptom Families

Most TIG complaints fall into five buckets. Before touching a single dial, identify which family your defect belongs to. That step eliminates most of the

How GTAW Works: The Four Variables Behind Almost Every TIG Problem

Before you can fix a bad TIG weld, you need a working mental model of the process. Gas tungsten arc welding is unusual because almost nothing is automatic: the machine hands you a stable arc, and you supply everything else. Once you understand the four variables that govern the arc — the tungsten electrode, the shielding gas, polarity, and arc length — most tig welding problems sort themselves into predictable categories.

The Tungsten Electrode: The Part That Must Stay Pristine

The tungsten is non-consumable. It carries the current, shapes the arc, and is supposed to outlast the entire job. How you prepare it decides how stable that arc will be. Grind lengthwise, on a dedicated diamond wheel, with the grind marks running parallel to the length of the electrode. This gives the arc a straight channel to follow and keeps ignition crisp.

Grind radially — around the circumference instead of along it — and you get the opposite result. ESAB’s GTAW guide notes that radial grinding encourages a wandering arc and can break off tungsten fragments, which then land in the puddle as inclusions. That is a defect unique to TIG, and it starts at the grinding wheel.

Contamination follows the same logic. Touch the tungsten to the puddle, dab the filler rod against the electrode, or drag it across an aluminum plate, and the arc becomes erratic almost immediately. The fix is never to keep welding. Cut the contaminated tip off, regrind lengthwise, and restart.

Shielding Gas: Delivery, Post-Flow, and What Goes Wrong

Argon is the default for steel, stainless, titanium, and copper. Helium mixes show up on thick material where you want deeper penetration. What you never use is a MIG shielding gas — argon/CO2 blends belong to GMAW, and on TIG they burn the tungsten and poison the puddle.

Gas has two jobs: protect the molten puddle while the arc is on, and protect the cooling crater after it stops. That second job is post-flow. A common rule of thumb is roughly one second of post-flow per 10 amps of current, which keeps the hot crater and the still-glowing tungsten from oxidizing in open air. Shorten post-flow and the crater dulls grey; the tungsten discolours and wears faster.

Flow rate sits in a narrow band. Too low and the puddle sees oxygen; too high and the stream becomes turbulent, sucking ambient air into the shield. The practical range for most TIG work is about 15–20 cfh. Drafts matter just as much — an open shop door or a fan can strip the shield away even at correct flow.

Polarity: DCEN, AC, and the Cleaning Action

Polarity decides where the heat goes and whether the arc cleans. On steel and stainless you run DC electrode negative (DCEN): most of the heat lands in the workpiece, giving clean, deep penetration with a narrow bead. Flip to electrode positive and the tungsten takes the punishment — it overheats, balls excessively, and melts away while the weld stays shallow.

Aluminum and magnesium need AC, because their oxide layer melts around 3,700°F while the metal beneath melts near 1,220°F. On AC, the electrode-positive half of the cycle blasts the oxide off, and the electrode-negative half drives penetration. If you weld aluminum on DCEN, the arc cannot break that oxide. Filler mixes with partially melted oxide, and the bead comes out grey, lumpy, and contaminated. Anyone switching to aluminum should read our guide to prepping aluminum for welding before striking an arc.

Inverter machines let you tune this. More electrode-positive time means more cleaning — you will see brownish oxidation or “black pepper” flakes lifting out of the puddle. Too much EP and the tungsten balls up and the bead gets over-etched. AC frequency, typically adjustable from 60 to 200 Hz, narrows and focuses the arc as you raise it. Pulsing alternates between peak and background current, limiting total heat input on thin material.

Arc Length and Torch Angle

Arc length is the distance from the tungsten tip to the puddle, and it is the variable welders abuse most. ESAB puts the working range at 1–3 mm. Longer than that and the arc wanders, widens, and pulls in atmosphere — leading directly to oxidation, porosity, and a soft, unfocused bead.

Hold the torch with roughly 10–15° of travel angle. That slight push keeps the puddle visible and lets the gas envelope cover the trailing weld. Push much harder and you blow gas ahead of the arc; drag heavily and you hide the puddle behind the cup.

Matching Current to Tungsten Size

Every tungsten diameter has a current ceiling. A 1/16-inch (1.6 mm) electrode handles roughly 60–80 amps; 3/32-inch (2.4 mm) covers about 80–160 amps; 1/8-inch (3.2 mm) runs 160–250 amps and up. Exceed the ceiling and the tip balls, dips into the puddle, and sheds tungsten into the weld. Undersize the tungsten for the job and the same thing happens faster.

Variable	Correct range	Failure mode when wrong
Arc length	1–3 mm	Wandering, oxidation, porosity
Torch travel angle	10–15°	Poor visibility, weak gas coverage
Polarity	DCEN (steel), AC (aluminum)	Tungsten damage, no cleaning action
Grind direction	Lengthwise	Arc wander, tungsten inclusions
Tungsten size	Matched to amperage	Balling, overheating, contamination

Hold these four variables steady and the troubleshooting in the sections ahead becomes far simpler: every defect points back to one of them. Next, the first and most-searched failure family — gas shielding problems, from porosity to sugaring.

Gas Shielding Problems: Porosity, Contamination and Sugaring

If there is one category of TIG defects that causes more rejected parts than everything else combined, it is gas shielding failures. The shielding gas envelope is the only thing standing between your molten puddle and the oxygen, nitrogen and humidity in the air. Disrupt that envelope for even a second and the weld metal reacts immediately. The result shows up as porosity, the classic TIG problem most people search for first.

Porosity is simply trapped gas: cavities, holes or pinpoint voids left in the weld when gas that was dissolved in the molten metal cannot escape before the puddle solidifies. Some porosity is subsurface and only reveals itself through radiographic testing. Some breaks the surface as tiny pinholes or a sponge-like bead. Either way, it reduces the cross-section of the weld and turns an otherwise attractive TIG bead into a structural question mark.

Every Way Your Gas Shield Can Fail

Porosity is rarely random. Almost always you can trace it back to one of these causes:

– Gas never reached the arc. The cylinder valve is shut, the machine’s gas line is not connected, or the solenoid never opens because of a wiring or pedal fault. The arc runs beautifully on bare argon-depleted air and the weld comes out Swiss cheese.
– Wrong gas mixture. This is the big one. Gas blends designed for GMAW/MIG, typically argon with 15–25% CO₂, must never be used for GTAW. As Miller’s guide to common TIG weld problems notes, CO₂ and oxygen break down in the arc and contaminate the puddle directly, producing porosity no matter how clean the joint is. For steel and stainless, use pure argon. Helium mixes are fine when you deliberately choose them for deeper penetration, but a MIG mix is never acceptable.
– Flow too low. Below roughly 12–15 cfh on a standard #8 cup, the gas envelope is too thin to keep air out, especially with any air movement at all.
– Flow too high. More gas is not more protection. Excessive flow becomes turbulent and actually sucks surrounding air into the shielding envelope. The gas exits the cup in a chaotic swirl instead of a smooth laminar blanket.
– Leaks in the line. Loose fittings, cracked hoses, a damaged O-ring on the torch, or a leaking regulator all bleed gas before it reaches the cup.
– Contaminated or damp supply lines. A brand-new or freshly repaired hose can hold moisture and oil residue inside. Cylinder lines that have taken on water vapor deliver that moisture straight to the puddle.
– Drafts and wind. A fan, an open shop door or a breeze is often enough to strip the shielding away. TIG is far more sensitive to this than MIG or stick, which generate their own protective atmosphere.
– Dirty or damp filler rod. Oil, grease and moisture on the rod vaporize into the puddle. Wipe your rod with a clean, lint-free cloth and store it off the floor.
– Travel speed too fast. Move quickly enough and the gas has no time to do its work, while the trailing edge of the puddle stays hot and unprotected.

For routine work, set your flowmeter in the range Miller recommends of about 15–20 cfh, then adjust to the cup size and conditions. A good habit is the soapy water leak test: mix dish soap and water, brush it over every gas fitting while the cylinder is open and the torch is flowing, and watch for bubbles. It takes two minutes and it eliminates an entire class of mystery defects.

Sugaring: The Invisible Stainless Killer

Sugaring is a heavier, more destructive cousin of porosity, and it happens almost exclusively on stainless steel and nickel alloys. When the back side of a weld root is exposed to oxygen while still hot, the chromium in the alloy reacts aggressively and forms a thick, grey, crusty oxide layer that looks exactly like burnt sugar. Corrosion resistance is destroyed along the root, even though the front of the weld may look flawless.

The fix is argon back purging: fill the enclosed side of the joint with argon and let it displace the oxygen before, during and after welding. On open or non-critical joints, reducing amperage and travel speed enough to keep the root from burning through also limits the damage. Related, milder symptoms include straw, blue or purple heat tint on the surface. Light tint is cosmetic; heavy dark discoloration signals excessive heat or gas loss and should prompt you to check flow, travel speed and current. When you are working with corrosion-critical alloys such as Alloy 20, purging is a procedure requirement, not a suggestion.

A Fast Diagnostic Shortcut

| Symptom | Most likely cause | First check |
|—|—|—|
| Pinhole porosity along the whole bead | No gas or wrong gas | Cylinder valve and gas type |
| Scattered porosity on one side | Draft or leak | Wind source, fittings with soapy water |
| Porosity that appears only when speed increases | Insufficient flow | Raise flow into the 15–20 cfh range |
| Crusty grey root on stainless | Oxygen exposure | Back purge setup |
| Straw/blue tint | Heat or gas loss | Amperage, cup size, travel speed |

For more on gas cylinders, regulators and safe handling, see our AllGas welding supply guide. And as CK Worldwide puts it, most shielding problems come down to forgetting that the gas is part of the process, not an accessory to it.

Tungsten Electrode Problems: Inclusions, Contamination and Excessive Consumption

One defect family in GTAW belongs entirely to this process. MIG and stick welders never worry about it, because they don’t run a non-consumable electrode inside the arc. Tungsten inclusions form when particles of the electrode break off, drop into the puddle, and lock into the finished weld. On X-ray they show as bright, dense specks, and they are an automatic reject.

What a Tungsten Inclusion Actually Is

When the tungsten touches the molten pool or the filler rod, a fragment can break away and end up in the weld metal. Tungsten melts at roughly 3,422°C, far above steel, so it does not dissolve into the puddle. It sits there as a hard, brittle inclusion that acts as a crack initiator under load. Troubleshooting guides from Miller and CK Worldwide list the same primary causes every time: accidental contact of the electrode with the puddle or the filler rod, excessive stickout, and wrong or inadequate shielding gas.

The Six Usual Suspects

• Dipping the tungsten. The most common cause by far. Once the tip touches the puddle, the arc changes shape and the contaminated tip must be repaired before you continue.
• Touching the filler rod to the electrode. Easy to do when feeding rod with an unsteady hand or working in a tight joint. The rod melts a nodule onto the tungsten, and that nodule can drop into the weld.
• Excessive electrode extension. If the tungsten sticks out too far beyond the cup, it overheats, wanders, and becomes far more likely to break up or fall into the puddle. Keep stickout to roughly the diameter of the cup or less.
• Too much current for the tungsten diameter. An undersized electrode at high amperage balls up, sputters, and eventually spits particles into the weld. Match electrode size to amperage using the manufacturer’s chart, and remember that ESAB’s process guidance treats electrode selection as a core setup variable, not an afterthought.
• Scratch starting on the workpiece. Dragging the tungsten across the base metal embeds it with material and grinds the tip flat. Many welding codes do not permit scratch starting at all on production work; use a copper strike plate or a high-frequency start instead. This matters especially on controlled applications such as aerospace TIG welding, where electrode discipline is written into the procedure.
• Radial (circumferential) grinding. Grinding marks that run around the tip instead of lengthwise create an arc that spins and wanders, which encourages contamination. Always grind lengthwise, in line with the electrode axis, on a dedicated tungsten grinder or a clean wheel reserved for tungsten only.

How to Fix a Contaminated Tungsten

Never keep welding and hope the arc behaves. Stop, break the arc, and repair the electrode properly:

1. Break the arc and let the torch cool. Clip off the contaminated tip with side cutters, or snap it off at the contaminated zone if the tungsten has cracked.
2. Regard the freshly cut end as contaminated by the cutters. Grind it clean, lengthwise, to the standard pointed or truncated geometry for your application.
3. Re-dress the tip, re-insert the electrode at the correct stickout, and restart. If contamination was severe, replace the tungsten entirely. A contaminated arc wanders, and a wandering arc causes more contamination. It is a feedback loop that only stops when you fix the electrode.

Excessive Tungsten Consumption

Sometimes the electrode doesn’t get contaminated; it just disappears. The tip erodes, rounds off, and requires constant regrinding. The causes are predictable:

• Inadequate gas flow. Too little argon leaves the hot tip unprotected, so it oxidizes and pits. Check your flowmeter and look for hose leaks with soapy water.
• Reverse polarity. On DC, electrode negative puts the heat in the work. Running electrode positive drives roughly two-thirds of the heat into the tungsten, which will destroy it rapidly.
• Wrong tungsten size or type. An undersized electrode, or the wrong alloy for the job, erodes quickly. Ceriated and lanthanated tungstens hold up better at low amperage than pure tungsten.
• Overheating inside the torch. Poor water flow, a kinked line, or running beyond the torch’s rated duty cycle lets heat build in the head. Check that the coolant return line is connected to the power cable lead, not the water-in line.
• Oxidation during cooldown. If post-flow is too short, the still-glowing tip is exposed to air and turns grey and crumbly. A widely used rule of thumb is about 1 second of post-flow per 10 amps of welding current, which keeps gas over the electrode until it has cooled below the oxidation point.

Repairing a Weld That Already Contains Inclusions

If an inclusion is found during inspection, there is no acceptable repair other than removal. Gouge or grind the entire length of the affected weld out until clean, sound metal is exposed, verify removal by visual and, where required, magnetic particle or dye penetrant inspection, then re-weld with a sound electrode. Inclusions cannot be washed back into the pool, and no amount of subsequent passes will make them disappear. That is why electrode discipline at the bench is far cheaper than repair work at the inspection table.

Aluminum TIG Problems: Oxide, AC Balance and Lack of Fusion

Aluminum is where most TIG welders get humbled, and it comes down to one number: pure aluminum melts at about 1,220°F (660°C), while the oxide skin that forms on its surface within seconds of cleaning does not melt until roughly 3,700°F (2,049°C). The metal underneath can be a liquid lake while the crust floating on top is still solid. Nearly every aluminum TIG complaint — grey dirty beads, black flakes swirling in the puddle, porous surfaces, cold roots on T-joints — traces back to that temperature gap.

Why DCEN Fails on Aluminum

On steel and stainless you would happily run DC electrode negative (DCEN) and enjoy deep, stable penetration. Try that on aluminum and the arc melts the base metal while doing nothing to the oxide film. The puddle then swallows half-melted oxide particles and drags the rest along the joint, producing a grey, crusty bead that is poorly wetted at the toes, often porous, with filler that refuses to flow out smoothly. That contamination also dulls and pits the tungsten, which sends the arc wandering and compounds the problem.

Alternating current fixes this because it performs both jobs in one arc. During the electrode-positive (EP) half-cycles, the arc blasts oxide and contaminants off the surface. During the electrode-negative (EN) half-cycles, it drives heat into the plate for penetration. That is why AC is the standard setting for aluminum and magnesium, while DCEN stays reserved for steel, stainless, copper and titanium. As a practical check, the Miller TIG problem guide lists polarity and balance errors among the most common causes of contaminated, lack-of-fusion aluminum welds.

Reading AC Balance: Cleaning Versus Penetration

AC balance decides how much of each half-cycle goes to cleaning versus penetration, and the arc tells you when you have it wrong:

• Brownish oxidation or “black pepper” flakes in the puddle: not enough cleaning time. Shift the balance toward EP until those specks stop appearing and the puddle turns bright.
• Excessive tungsten balling, a spitting arc, or a chalky etched band beside the bead: too much EP. Pull the balance back toward EN. A correctly set tungsten forms a shallow rounded dome, not a dripping ball, and the area beside the bead stays bright instead of etched. Since settings vary by machine, start at the manufacturer’s default and nudge the balance in small steps while watching the puddle.

The Shiny-Puddle Rule

Miller’s advice on aluminum is simple and worth memorizing: wait until the puddle looks like a shiny dot before you add filler. That bright dot is your visual confirmation that the oxide has broken down. Dab the rod in too early and you are mixing filler with a skin of oxide and whatever oil or moisture was trapped under it, contaminating the bead from the inside. A second rule of thumb is that the filler should be dipped into the leading edge of the puddle, never fed straight into the arc.

Everything upstream of the arc matters even more on aluminum than on steel. Oxide does not wipe off with a rag, and aluminum holds onto hydrogen, so damp rod or a wet surface turns directly into porosity. Use a dedicated stainless wire brush that never touches steel, degrease the joint, and store filler rod clean and dry. The full prep routine is covered in aluminum cleaner for welding: how to prep aluminum right.

Lack of Fusion at the Root of T-Joints

The classic aluminum fillet failure is a bead that looks fine on the outside while the root never fused. The usual suspects:

• Poor fit-up — a gap or mismatch pulls heat away from the root and leaves an unfused line.
• Arc too long — the cone spreads, and heat lands on the visible legs of the T instead of the corner.
• Wrong filler technique — feeding rod into the arc chills the puddle instead of building the joint.
• Heat sinking — heavy, cold sections absorb amperage faster than the machine supplies it.

The fixes are tightening fit-up, keeping the arc roughly 1–3 mm long, starting the arc right at the root and holding it until a puddle forms on both members before adding filler. On inverter machines, raising AC frequency and using pulsing creates a narrower, more focused arc cone that penetrates deeper and gives better directional control in tight corners.

Symptom	Likely cause	First move
Grey, crusty bead	DCEN polarity or contaminated tungsten	Switch to AC, cut off and regrind tungsten
Black flakes in puddle	Too little cleaning (EP) time	Shift balance toward EP, re-clean surface
Unfused root in T-joint	Fit-up gap and long arc	Close the gap, shorten arc, raise AC frequency

And if a job is too large, too thin, or too position-critical for comfortable TIG work, aluminum MIG welding: the complete practical guide covers when the spray transfer alternative is the smarter choice.

Heat and Travel Speed Problems: Undercut, Burn-Through and Poor Penetration

Get your gas right, your tungsten clean and your plate spotless, and you can still ruin a TIG weld with heat alone. The arc gives no audible warning the way a stick electrode does, so you only see the damage once the metal cools: a groove bitten into the toe, a blowout in thin sheet, or a bead that looks fine on top and is barely tacked at the root underneath.

Every heat defect traces back to a small set of variables: amperage, travel speed, arc length, torch angle and how you finish the weld. Change them deliberately and the defects disappear.

Undercut: The Notch That Starts Cracks

Undercut is a groove melted into the base metal along the toe of the weld and left unfilled because the filler never got there. It looks like a small scratch running beside the bead, and it matters far more than its size suggests. An undercut is a stress concentration point, exactly the geometry where fatigue cracks initiate when a part is loaded and unloaded repeatedly.

The typical causes are:

• Excessive current, which makes the puddle too fluid to stay where you put it.
• Travel speed too fast, so the filler cannot keep up with the melting arc.
• Arc too long, which spreads heat sideways instead of focusing it.
• Abrupt weaving or whipping, where the arc dwells on the toes, melts them, then leaves before filler covers them.
• Poor electrode manipulation, such as too steep a travel angle dragging the pool backward.

The AWS D1.1 Structural Welding Code names excessive welding current, excessive travel speed and poor electrode manipulation as the primary causes of undercut, and its remedies are just as direct: reduce current, reduce travel speed and train the operator. For static connections under 1 in thick, undercut is generally allowed to run no deeper than 1/32 in (1 mm). On cyclically loaded primary tension members the limit drops to roughly 0.01 in, because fatigue cracks grow from exactly those notches. Our Aerospace TIG Welding: Standards, Techniques, and Training guide shows how much tighter the scrutiny gets.

Burn-Through on Thin Sheet

Burn-through is the opposite problem: too much heat for the section, opening a hole instead of a puddle. It happens on thin material, at starts and stops where heat accumulates, and in inside corners where two walls feed heat into the same spot. Slow travel, a long arc and a stationary torch all make it worse.

The fixes are as much mechanical as electrical: drop the amperage, keep the arc moving, and use the foot pedal to back off as you approach a corner or as the part heats up. A copper backing bar or chill block clamped behind the joint pulls heat away, which is a big help on thin aluminum and stainless. Pulsed TIG helps too, with peak current driving penetration and background current letting the puddle partially freeze. On thin-aluminum hull and tank work this matters most; see Aluminum Boat Welding: The Complete Guide for Strong Hull Joints for joint-specific advice.

Lack of Penetration, Cold Lap and Concave Beads

Too little heat produces its own family of defects. Lack of penetration at the root comes from amperage that is too low, travel too fast, a long unfocused arc, a root opening that is too tight, or filler dumped into the puddle ahead of the arc. Overlap, or cold lap, is filler riding over un-fused base metal instead of wetting into it, and it usually traces to the same root causes: not enough heat, or the rod tip straying outside the puddle’s leading edge.

Both respond to the same corrections: raise amperage or slow down, shorten the arc, keep the rod tip inside the leading edge of the puddle, and clean the surface properly. Oxide and oil stop the puddle from wetting, which is why aluminum prep deserves its own routine — our Aluminum Cleaner for Welding: How to Prep Aluminum Right walkthrough covers it step by step.

Concave and underfilled beads come from pulling the arc away too quickly at the stop, or not feeding enough filler into the crater — a finish CK Worldwide lists among the most common TIG mistakes. Besides looking weak, an underfilled fillet fails on measurement: AWS D1.1 allows fillet welds to be undersized by no more than 10% of the weld length, so a thin stretch along part of the bead can condemn the entire joint.

A Practical Way to Set Your Heat

Do not guess at the dial. Cut offcuts from the same material, thickness and joint geometry, and adjust the current until the puddle is fluid but controllable. For mild steel, a common starting point is roughly 1 amp per 0.01 in (0.25 mm) of thickness in flat position, then correct for your situation: reduce for vertical or overhead work or for heavy fixtures that sink heat, increase for aluminum, which sheds heat fast. Miller’s guide to common TIG weld problems gives the same order of operations: set the heat for the joint, not the dial.

Keep a card by the machine with the settings that worked. Heat control is a feel skill, built by changing one variable at a time — never two.

Cracking: Crater Cracks, Hot Cracking and Hydrogen Cracking

Every other defect in this article has some gray area: a little porosity in a non-structural bracket, a slightly lumpy bead, minor undercut you can dress out with a file. Cracks are the exception. AWS D1.1 says so in Table 8.1: cracks carry an absolute prohibition, meaning any crack is rejectable regardless of size or location, for both static and cyclic connections.

Crater Cracks: The Crack You Create at the End of Every Pass

Crater cracking is the most common crack a TIG welder produces, and it happens in the last half-second of the weld. Lift the torch off abruptly while the arc is hot, or pull the filler rod away before the pool is fed out, and you leave a small, rapidly cooling puddle shrinking against a solidifying bead with no added metal to relieve the tension. The result is a star-shaped tear or a narrow finger crack sitting in the crater, often at the end of a bead that otherwise looks fine.

Three habits eliminate almost all crater cracks:

1. Use down-slope current. Ease off the pedal (or set the down-slope timer) so current decays over a second or two rather than cutting instantly, and keep feeding filler until the arc is gone.
2. Weld back into the crater. Reverse direction at the joint edge and run back over the crater while it is still fluid, tying it into the previous bead. This works even without down-slope.
3. Hold position through post-flow. Keep the tungsten in place until post-flow protects the hot crater, so it does not oxidize and embrittle while it cools.

Hot Cracking and Centerline Cracking

Hot cracking, also called solidification cracking, happens while the weld metal is still freezing. As the bead solidifies it shrinks, and if a thin film of liquid metal remains along the center, that film pulls apart and leaves a continuous crack down the middle. It is most likely in a single-pass bead that is narrow and deep, on a joint with too wide a root opening, or at a travel speed so fast the bead never gains width to reinforce itself. Crater cracks are the same mechanism.

The countermeasures are straightforward:

• Run a wider, flatter bead so shrinking metal has cross-section to carry the load.
• Tighten an oversized root opening instead of bridging it with a long, thin keyhole bead.
• Slow down slightly and let the pool wet out on both toes.
• Preheat where material or joint geometry dictates, so cooling is less severe.
• Always fill the crater, and use a cleaner, well-matched filler alloy. For crack-sensitive, high-alloy work, follow procedures and matching consumables such as these Alloy 20 welding procedures.

Restraint is the hidden amplifier: a heavily boxed-in joint tacked hard on every side has nowhere to shrink, so the stress lands in the hot metal itself.

Cold Cracking After the Weld Cools

Cold cracking (post-weld or hydrogen cracking) is the nastiest family because it can appear hours after the weld is done. It takes three ingredients acting together: excessive joint restraint, rapid cooling, and hydrogen in the weld zone. Hydrogen sources include damp or oily base metal, wet filler rod, contaminated shielding gas lines, and coatings or mill scale that break down in the arc. As the metal cools, hydrogen migrates to stressed regions and the restrained joint tears.

The standard countermeasures:

• Preheat the joint and control interpass temperature so cooling is slower.
• Eliminate hydrogen sources: clean, dry shielding gas, dry filler, degreased and descaled base metal.
• Change the joint design to reduce restraint: fewer heavy tacks, a root face rather than a knife edge, room for shrinkage.
• Avoid underbead notches and poor fusion profiles that act as stress raisers where hydrogen collects.

TIG has a natural advantage here: with no flux coating to introduce moisture, a properly set up GTAW arc is inherently low-hydrogen. That advantage vanishes the moment the gas line, filler or plate is dirty.

The Three TIG Cracks at a Glance

Crack type	When it forms	Primary trigger	Fix
Crater crack	Arc termination	Abrupt arc stop, filler pulled early	Down-slope, keep feeding filler, or weld back in
Hot / centerline crack	During solidification	Narrow bead, wide root, fast travel, restraint	Wider bead, tighter root, preheat, matched filler
Cold / hydrogen crack	Hours after welding	Restraint, rapid cooling, hydrogen	Preheat, clean dry gas and filler, lower-restraint joint

Because cracks are non-negotiable under the code, inspection is strict everywhere, from structural steel to aerospace TIG welding, where a single cracked bead scraps the part. See a crack, stop, cut it out, find the cause, and re-weld. Grinding it flush and hoping it was only surface deep is never an option.

Equipment and Setup Mistakes That Cause Bad Welds

Before you blame your hand, rule out the machine. A surprising share of bad TIG welds traces back to the rig itself: a regulator at the wrong pressure, a cup too small for the job, a torch running hotter than it is rated for. These faults repeat on every arc, so no amount of practice fixes them. As CK Worldwide’s breakdown of common TIG mistakes puts it, most persistent arc problems are setup problems wearing a technique disguise.

Gas Flow Blockages, Leaks and Wrong Regulator Settings

Gas faults start at the regulator. TIG flowmeters are typically set at about 35 psi with a low flow rate, while MIG flowmeters run at 65 psi or higher. Bolt a MIG regulator onto a TIG bottle and you get a weak, wandering shield or a blast of gas that churns the air into turbulence; either way the weld picks up porosity. Each process needs its own flow control, as this Airgas Welding Equipment Guide explains.

Practical flow sits around 15–20 cfh for open-joint work: lower for small cups, higher for large cups and out-of-position welding. Too little flow leaves the puddle exposed; too much pulls air in from the sides. Soapy water on every connection from cylinder to torch head shows leaks as bubbles, and that hidden leak becomes porosity you cannot see until the part is cold. Check the line too: moisture or oil inside a hose, a failed solenoid or a kinked liner will starve the arc of gas.

Arc Blow: When the Magnetic Field Pushes the Arc

Arc blow comes from the magnetic field induced by the DC weld current. The arc is dragged sideways, usually near the end of a run or where geometry is uneven, and a tidy bead wanders or traps oxide at the toes. Correct it in this order: shorten the arc, reduce the current, reposition the ground clamp to balance the field, and where the material allows, switch to AC with high frequency, which alternates the field and cancels the pull. Miller’s common TIG weld problems guide lists arc wander among the most frequent setup-driven complaints.

Cups, Stickout, Collets and Torch Cooling

• Excessive stickout pushes the nozzle off the puddle, weakening coverage and inviting contamination. Keep extension as short as the joint allows.
• Cups too small for the amperage give a shield that dies before it reaches the trailing edge of the weld.
• Twisted split collets clamp unevenly, shorten collet life and let the tungsten slip or sit off-centre in the cup.
• Cracked or wrongly sized cups leak gas and let the tungsten sit off-axis, distorting the arc.
• Cooling faults: on a water-cooled torch the coolant return must feed the power cable lead, not the gas fitting. A misrouted line cooks the head within minutes.
• Rated capacity: every torch has an amperage and duty-cycle rating. Exceed it and you will burn through hoses, collets and cups in short order.

Cleaning Is a Setup Step, Not an Afterthought

Base metal preparation belongs in the setup checklist. Welding hot-rolled mild steel without removing mill scale first is a documented cause of contamination, and oil, paint or damp filler rod do the same job just as efficiently. Wipe with acetone and brush back to bright metal, then store filler rod off the floor. Aluminum raises the stakes: dedicated stainless brushes never used on steel, plus a solvent wipe. Our Aluminum Cleaner for Welding guide covers the prep routine in detail.

Symptom-to-Check Quick Table

Symptom	First check
Porosity in every weld	Regulator pressure (35 psi), leaks, cup size
Arc wanders or darts	Stickout length, tungsten grind, arc blow
Tungsten discolours or balls up	Collet fit, coolant routing, current vs tungsten size
Cup cracks or scorches	Cup size, tungsten position, torch duty rating
Grey, crusty discoloration	Wind draft, gas leaks, mill scale and oil
Erratic arc after a few minutes	Torch overheating, coolant line

Run through this list before regrinding a tungsten or changing settings; five minutes of checking usually saves an afternoon of grinding out rejects.

Safety and Health Problems Around TIG Welding

Not every TIG welding problem shows up in the bead. Some show up in your eyes at 2 a.m. or in a lung capacity test years later. Troubleshooting GTAW isn’t only metallurgical: the arc itself creates hazards that are easy to ignore precisely because TIG looks so clean and quiet compared to stick welding.

What the Arc Does to Air and Skin

The GTAW arc produces intense ultraviolet radiation. In open air, that UV converts oxygen into ozone and nitrogen oxides, gases that irritate the eyes and the respiratory tract even at low concentrations. You smell that sharp, electric odor after a long bench session for a reason. Direct exposure also causes photokeratitis — arc eye — the sandpaper-in-your-eyes feeling from an unprotected glance at the arc, plus ordinary thermal burns from hot metal and spatter-free as TIG is.

Fume is the second concern. TIG fume levels are generally lower than stick or flux-cored processes, but they are not zero. Welding coated, plated or alloyed metals can release zinc, chromium, nickel or manganese compounds. OSHA’s welding hazard guidance warns that acute exposure can cause eye, nose and throat irritation, dizziness and nausea, while prolonged exposure may cause lung damage and cancers of the lung, larynx and urinary tract. Manganese in particular is associated with nervous-system effects, a concern addressed by NIOSH’s manganese profile.

The Silent Hazard: Shielding Gas Displaces Oxygen

Argon is heavier than air and helium is lighter, but both share a dangerous property: they don’t support life. OSHA and HSE guidance is blunt that argon, helium and carbon dioxide displace oxygen and can cause suffocation, particularly in confined or enclosed spaces. A leaking cylinder line in a tank, a pit or a poorly ventilated bathroom renovation can pool gas invisibly before you notice lightheadedness — and by then your judgment is already impaired.

The Control Hierarchy Regulators Actually Recommend

Welding standards don’t start with a respirator — they work down a hierarchy. First, process selection: choose the process and consumables that generate the least hazard for the job. Second, local exhaust ventilation (LEV), which captures fume at the source. Third, general ventilation to dilute what escapes. Only then, respiratory protection, as the last layer rather than the first line of defense.

AWS Z49.1, Safety in Welding, Cutting and Allied Processes, sets the ventilation and air-cleaning criteria U.S. shops build their programs around, including minimum airflow and capture-velocity requirements for LEV hoods and fume extractors. The UK’s HSE welding fume guidance adds a practical rule of thumb: sporadic, low-intensity TIG welding can generally be controlled with good general ventilation, while regular or high-intensity TIG requires local exhaust ventilation. If you weld stainless every day at the bench, a gooseneck extractor positioned at the arc is not optional equipment.

Where does that leave the hobbyist or small shop? Reasonably, here:

• Sporadic TIG on clean steel: general ventilation plus an open door or fan across the workspace is usually sufficient, per HSE’s low-intensity guidance.
• Regular TIG, stainless, aluminum or coated material: LEV with capture at the arc, as required by AWS Z49.1 criteria.
• Confined, coated or high-alloy work: LEV, atmosphere monitoring, and respiratory protection selected for the specific contaminant.

Aluminum work deserves a mention because prep chemicals and cleaning routines add their own exposure. If you’re brushing and degreasing before every pass, read our guide on aluminum cleaner for welding to make sure your solvent choices aren’t contributing to the fume load. The same logic applies across processes: whether you run aluminum flux core welding or GTAW, the ventilation hierarchy stays the same even as the fume chemistry changes.

Quick Safety Checklist Before You Strike an Arc

Hazard	Control
UV / arc eye	Correct shade lens, skin covered, bystanders shielded
Ozone and NOx	General ventilation; LEV for regular work
Metal fume	Capture at arc, respiratory protection as last resort
Gas asphyxiation	Leak checks, ventilation, O₂ monitoring in enclosed spaces
Electric shock	Dry gloves, insulated torch, machine grounding, correct polarity setup
Burns	Leather gloves, sleeves, no hot parts left on the bench

Fixing a porosity or tungsten inclusion problem starts with the arc; preventing a health problem starts before it. Cleanliness, correct gas and disciplined setup protect the weld and the welder at the same time — and the troubleshooting routine in the next section is built on that same habit of checking one variable at a time.

A Step-by-Step TIG Troubleshooting Routine, Inspection Limits and Final Checklist

You have matched your bead to a picture and still have a bad weld on the bench. Knowing every TIG problem by name does not fix one; a routine run the same way every time does, until the cause shows itself.

The Five-Step Diagnostic Routine

1. Name the visual symptom before you touch a dial: bead, toes, crater, back side, tungsten tip. Holes? Discoloration? A rough surface? A crack? An arc that will not behave?
2. Classify it into one family: gas, heat, cleanliness, electrode or technique. Most symptoms belong to one only.
3. Change one variable only. One dial, one rod, one setting. Change two and you learn nothing, because you never know which change worked.
4. Test on scrap of the same thickness and alloy. Thicker plate tells you nothing about the part in front of you.
5. Record what worked — amps, tungsten size, cup, gas flow, travel speed, post-flow. Your notes become your settings library.

Quick Reference: Symptom, Cause, Fix

Symptom	Most likely cause	First fix
Pinholes or pitting on the bead	Gas shielding lost, draft, or damp filler	Confirm gas is on, set 15–20 cfh, soapy-water leak test, clean the rod
Grey streaks, erratic arc, spatter	Tungsten dipped in the puddle or touched by filler	Cut the contaminated tip off and regrind lengthwise
Crusty grey backside on stainless	No argon back purge, excessive heat	Purge the back side and reduce amperage
Crack at the stop point	Arc ended abruptly, crater left unfilled	Down-slope while still feeding filler into the crater
Groove melted into the toe	Too much current, travel too fast, arc too long	Drop amps, slow travel, shorten the arc
No fusion at the root	Too little heat, poor fit-up, long arc	Raise amps, tighten fit-up, shorten arc length
Arc wanders off the joint	Contaminated or badly ground tungsten, arc too long	Grind lengthwise, reduce stickout, shorten arc
Tungsten erodes or mushrooms fast	Wrong polarity, tungsten too small, post-flow too short	Check polarity, upsize tungsten, extend post-flow

How a Weld Is Actually Judged

Under AWS D1.1, every weld is visually inspected, with acceptance judged against Table 8.1 covering cracks, fusion, craters, weld profile, size, undercut and piping porosity. Visual inspection covers 100% of the weld surface; radiographic or ultrasonic testing is only required when the contract or specification calls for it.

Cracks get zero tolerance: any crack is rejectable. Complete-joint-penetration groove welds in butt joints loaded transverse to the weld allow no visible piping porosity under static loading. AWS attributes porosity to hydrogen, nitrogen or oxygen in the welding atmosphere, dirty metal, improper arc length and a high solidification rate — four things you control at the torch.

Inspection culture is tighter in some industries than others. For precision work, read Aerospace TIG Welding: Standards, Techniques, and Training, where acceptance criteria go far beyond structural work.

Prevention Priorities: The Final Checklist

1. Cleanliness. Wipe base metal and filler; remove oil, oxide and mill scale. Aluminum prep is its own discipline — see Aluminum Cleaner for Welding: How to Prep Aluminum Right.
2. Correct gas. Right gas, right flow, no leaks, no drafts, purge where required.
3. Correct polarity. DCEN on steel and stainless, AC on aluminum and magnesium.
4. Arc length. Roughly 1–3 mm, closer on low amperage.
5. Heat control. Match current to thickness, test on offcuts, use the pedal.
6. Crater fill. Down-slope plus filler until the pool solidifies.

That last item is a code requirement, not a style preference, and crater cracks rank among the most common TIG problems shop welders face.

Every defect here shares one pattern: a small, invisible lapse in gas, cleanliness, arc length, heat or technique becomes a visible flaw once the metal solidifies. Run the routine with the table beside your machine, and you will know which dial to move — and why.

Frequently Asked Questions

Why does my TIG weld have small holes in it?

Those holes are porosity, caused by gas trapped as the puddle solidifies. Common causes are shielding gas not flowing or flowing at the wrong rate, a hose leak, drafts, dirty or oily base metal, damp filler rod, or travelling too fast for the gas to escape. Clean the joint, check for leaks with soapy water, set flow to roughly 15–20 cfh and slow your travel speed.

Why does my tungsten keep sticking to the weld pool?

The electrode is touching the molten puddle or the filler rod, usually because the arc is too long, the current is too low for the tungsten size, or your hand is fatigued. Shorten the arc, match tungsten diameter to amperage, trim excessive stickout, and grind a fresh tip before restarting — never weld on a contaminated electrode.

What is sugaring on stainless steel TIG welds?

Sugaring is heavy oxidation on the root side of a stainless weld that forms when hot metal is exposed to oxygen. Prevent it by back-purging the joint with argon and reducing amperage, and by making sure purge gas has enough time to displace air before striking the arc.

Why are my aluminum TIG welds dirty or black?

You are most likely on the wrong polarity or have your AC balance set too far toward penetration. Aluminum needs AC so the electrode-positive half breaks up the oxide layer. If you see brown oxide or black-pepper flakes in the puddle, increase cleaning (EP) time, use a dedicated stainless brush, and wait for a shiny puddle before adding filler.

How do I stop crater cracks at the end of a TIG weld?

Keep feeding filler while you back off the current — use the down-slope or foot pedal setting, then weld back into the crater or add a final small dab before breaking the arc. Abruptly stopping the arc or pulling the rod away early is the usual cause, and unfilled craters are rejectable under AWS D1.1.

How much shielding gas flow does TIG welding need?

Most GTAW work runs about 15–20 cfh on 100% argon with a standard cup. More flow is not better: excessive flow creates turbulence that draws air into the arc. If coverage still looks poor, use a larger cup or gas lens rather than simply opening the regulator further, and always work away from drafts.

Is TIG welding fume dangerous?

TIG produces less fume than stick or flux-cored processes, but it still generates ozone and nitrogen oxides from arc radiation, plus metal fumes when welding coated or alloyed metals. OSHA notes that acute exposure can cause eye, nose and throat irritation, dizziness and nausea, and that prolonged exposure may cause lung damage and cancer, so ventilation or respiratory protection is still required.

When is a TIG weld considered rejectable?

Under AWS D1.1 visual acceptance criteria, any crack is unacceptable regardless of size, craters must be filled, incomplete fusion at the surface is not allowed, undercut is limited (generally 1/32 in for static joints under 1 in thick), and visible piping porosity is tightly limited. Visual inspection is required on all welds; radiographic or ultrasonic testing is added only when the contract specifies it.