The Complete Guide to TIG Welding Sheet Metal: Settings, Techniques, and Mistakes to Avoid
The job was supposed to take forty-five minutes. Instead, it ate an entire afternoon, plus a burned-up spool…
READ ARTICLE →You set the machine to DCEN, the setting that welds steel beautifully, struck an arc on a clean piece of aluminum, and waited for the puddle to bloom. Instead, the arc skated across the surface. Black specks floated up, a gray scummy skin formed, and when metal finally moved it sagged beneath that skin like water under ice. It looked nothing like the shiny stack of dimes you expected.
Here’s the frustrating truth: your hand isn’t the problem. The metal is, and so is your polarity.
TIG welding, technically gas tungsten arc welding (GTAW), uses a non-consumable tungsten electrode to strike an arc that melts the base metal while you feed a separate filler rod by hand. It’s the most controllable arc process most shops own, which is exactly why aluminum humbles so many welders. Steel cooperates. Aluminum fights back.
Blame aluminum oxide. The moment bare aluminum meets air, a ceramic skin starts forming, and roughly 90% of its maximum thickness develops within the first 24 hours of exposure.
That skin is stubborn in every way that matters. It melts at roughly 3,600 F (about 1,980–2,072 C), while aluminum itself melts at only about 1,200 F (660 C) — a threefold gap. By the time the oxide finally breaks, the metal underneath is already a runaway puddle.
Three properties make that oxide even nastier:
Aluminum also moves heat away from the weld zone far faster than steel does, so the puddle cools almost as quickly as it forms. That leaves you fighting two battles at once: getting enough heat in to melt metal, while keeping the arc aggressive enough to strip the skin above it.
Quick Note: Because the oxide soaks up oil and moisture, pre-cleaning is not optional. Degrease first, then brush with a stainless-only wire brush, and keep that brush away from steel so you don’t contaminate the aluminum.
Unlike steel, aluminum gives you no glow to read. It shifts from solid to liquid with almost no color change, so a surface that looks cold may already be molten underneath. The oxide skin hides the telltale signs you’d normally watch for, which is why welders burn through aluminum so easily at first.
Since the oxide won’t cooperate, GTAW needs a way to break it down during the weld itself. That’s where polarity comes in. Alternating current swings the arc between two jobs: one half blasts the oxide apart, the other drives heat into the metal. Tuning that balance is the core skill behind TIG welding polarity. Steel lets you ignore most of this. Aluminum forces you to understand it.
This article walks the whole path. We’ll start with DCEN, DCEP, and AC fundamentals, then dig into how electrode-positive cleaning actually works and how balance, frequency, and waveform shape the arc. From there we’ll cover running aluminum on DC, the shop setup and safety basics that make clean welds easier, troubleshooting, and a practice routine to make it stick.
Polarity is simply the direction current travels through the welding circuit. In TIG, that path runs from the power source, through the torch, across the arc, into the workpiece, and back through the ground clamp. Whichever terminal the current leaves from determines what happens at each end, and in GTAW that single detail controls how heat divides between your tungsten and your metal. Flip it, and the arc behaves like a completely different process. Before you touch a single knob, this setting decides how the arc will treat your part.
Direct current electrode negative (DCEN), traditionally called straight polarity, makes the tungsten the negative terminal and the workpiece the positive one. Current flows from torch to work. Here’s the physics that matters: roughly two-thirds of the arc heat is generated at the positive anode, the point the electrons strike. Since DCEN aims that anode at the workpiece, most of the energy lands in your metal instead of your torch. Think of the positive side as the side that soaks up the energy, which is exactly why the work stays positive on thick material.
ESAB notes that DCEN concentrates around 70% of the heat into the workpiece, which gives deep penetration and lets the tungsten run cool and stable. A cool electrode holds its point, which keeps the arc tight and focused. The result is the deep, narrow bead you want on steel and stainless. That’s why DCEN is the default setting for those materials. If you want a deeper walkthrough of how each mode fits different jobs, our guide to TIG welding polarity breaks the options down.
Switch to direct current electrode positive (DCEP), or reverse polarity, and the circuit flips. Now the tungsten is positive and the workpiece is negative. The anode, and its two-thirds heat share, sits on your non-consumable electrode. That’s a disaster in slow motion. The tungsten has no way to shed that concentrated energy, so it balls up, sputters, and melts away, contaminating the weld. DCEP in GTAW is rare and mostly a niche for very thin sheet, precisely because the heat focused on the electrode destroys it quickly.
What DCEP does give you is cathode heat and cleaning at the work surface. The positive workpiece sheds electrons and, with them, breaks up surface oxide. Useful on dirty metal, but not worth sacrificing your electrode for on its own. Even dialed way down, DCEP on a standard tungsten won’t survive a long pass.
Quick Note: The anode always takes the heat. That single fact explains why polarity changes everything in TIG, and why the wrong choice damages either your tungsten or your bead.
Alternating current splits the difference by reversing direction many times per second. On the DCEN half-cycle you get workpiece heat and penetration. On the DCEP half-cycle you get oxide cleaning. TIG power supplies alternate the two, giving aluminum the best of both: the cleaning you need and enough penetration to fuse the metal.
That balance between the two half-cycles, often called cleaning width or balance control, is what lets you trade cleaning against penetration. For steel and stainless, DCEN already delivers everything you want, so there’s no reason to alternate. Aluminum’s stubborn oxide layer means cleaning earns its keep, which is why AC becomes the workhorse there instead. For a deeper look at when straight current still fits, see DC welding aluminium. The same DCEN logic governs most stick and wire work too, including steel TIG setups.
| Mode | Electrode | Workpiece | Best For |
|---|---|---|---|
| DCEN (straight) | Negative | Positive | Steel, stainless, deep penetration |
| DCEP (reverse) | Positive | Negative | Thin sheet niche; cleaning, but kills tungsten |
| AC | Alternating | Alternating | Aluminum; cleaning plus penetration |
Get this foundation right and the rest of aluminum TIG stops feeling like a fight. Polarity decides where the heat goes, and where the heat goes decides whether you’re welding or wrecking.
Here’s where the whole TIG aluminum story comes together. On alternating current your polarity isn’t fixed — it flips from electrode positive (EP) to electrode negative (EN) and back again, often dozens of times each second. Each half of that cycle does a completely different job, and together they do what no single DC polarity can: strip away the oxide while still driving heat into the metal.
During the EP half-cycle the tungsten becomes positive and the workpiece becomes negative. That reverses the normal current direction, so electrons leave the aluminum surface and stream toward the tungsten. As they pull away, they mechanically blast apart the tough aluminum oxide layer sitting on top of the metal. That’s the cleaning action, and it happens automatically, right at the leading edge of your puddle.
Then the current flips. On the EN half-cycle the tungsten goes negative, the workpiece goes positive, and the arc dumps intense heat straight into the base metal. That’s the penetration half of the equation. MillerWelds explains the split clearly: the electrode-positive phase breaks up the aluminum oxide while the electrode-negative phase delivers the heat and penetration.
Welders call this cathodic etching. When the workpiece briefly acts as the cathode during the EP phase, the arc’s positive ions accelerate into the surface and physically knock the brittle oxide loose. This matters because aluminum oxide doesn’t melt away the way we’d hope; it fuses at roughly 3,700°F, while the aluminum underneath melts near 1,220°F. The oxide skin is sturdier than the metal it protects, so it has to be broken up mechanically. That’s exactly what the EP half-cycle does, cycle after cycle.
You can see the evidence on every good aluminum weld: a narrow frosted-white band hugging both sides of the bead. Welders call it the etched zone, the etch line, or the cleaning halo. That smut line is the footprint of the EP cycles, proof the oxide was shattered before the filler touched the puddle. A bright weld with no etch band means you’re not cleaning enough; a wide, chalky halo means you’re over-cleaning and giving up penetration.
Bouncing between those two roles is also why AC balances heat so well. Using AC distributes heat more evenly between the tungsten and the workpiece, giving you finer control over heat and metal deposition than either DC polarity alone, as ESAB outlines. DCEN burns hot into the work but never breaks the oxide; DCEP cleans beautifully but throws most of its heat into the tungsten, melting the tip and capping your usable amperage.
Quick Note: High frequency does not create the cleaning. It only keeps the arc stable as it crosses zero between half-cycles. Cleaning comes entirely from the electrode-positive portion of the waveform. Mix the two up and your troubleshooting will run in circles.
That’s the core reason pure DC struggles on aluminum, even in the DC welding aluminium workarounds welders sometimes try. Run DCEN and the oxide never breaks down, so the filler mixes with partially melted oxide and your bead comes out contaminated, black, and scummy, as MillerWelds notes. Flip to DCEP and cleaning is aggressive, but the arc concentrates so much heat in the tungsten that you can’t push enough current to fuse real thicknesses. Neither polarity cleans and penetrates at once. Only AC does both — which is why it beats every other current on this metal.
The ratio between those two jobs isn’t fixed, and that’s the whole point. Modern TIG welding polarity machines let you decide how much of each cycle you spend cleaning versus penetrating. Dial in more electrode-positive time and you get a hotter, wider cleaning action for heavily oxidized or dirty stock; lean toward electrode-negative and you trade some cleaning for deeper fusion and a narrower, more focused bead. That one adjustment is what makes AC so tunable, and it’s worth practicing until the etch line tells you exactly what your settings are doing.
Once your machine is running on AC, the next dial that matters is balance. It goes by several names — AC balance, balance control, EN ratio, or E-ratio — but they all describe the same thing: the percentage of each AC cycle spent in electrode negative (EN) versus electrode positive (EP). Turn the knob one way and you get more cleaning; turn it the other and you get more penetration and heat control.
If you want the full picture of how EN and EP differ, our guide to TIG welding polarity: DCEN, DCEP, and AC explained walks through the physics. Here we’ll focus purely on dialing that balance for aluminum.
On AC, the current swaps direction every cycle. The EN half drives heat into the workpiece and gives you penetration. The EP half reverses the flow, lifting electrons off the surface and blasting away the aluminum oxide layer. More EN means deeper, hotter penetration but less cleaning. More EP means a brighter, better-cleaned puddle but less penetration and a lot more heat dumped into the tungsten.
Most modern inverters let you set that split precisely. A typical starting point for aluminum TIG is about 75% EN, then you refine it with test welds on scrap.
Quick Note: Balance only changes the EN/EP split, not your amperage. If you adjust one, expect to re-check the other on scrap before committing to a part.
Push the EN portion too high and you starve the arc of cleaning action. The telltale signs are black specks scattered across the puddle — welders call it “black pepper” — and a brownish oxide skin floating on the molten metal. That smut is aluminum oxide that never got removed, and it kills fusion. The MillerWelds troubleshooting guide flags exactly these symptoms: you get a dull, dirty bead that won’t tie into the base metal.
If you see peppering or that oxide film, you have too little cleaning. The fix is to decrease the balance setting so the arc spends more time in EP.
Drop the EN percentage too far and the opposite problems appear. With so much time in EP, the tungsten takes a beating. It balls excessively, loses its sharp point, and can overheat. The arc also over-etches the aluminum, leaving a wide, frosty white band on either side of the weld. That wide etched zone looks messy, weakens the surface, and signals too much of a good thing. Reducing the EN portion does give you greater cleaning action for heavier oxidation, but it puts more heat on the tungsten. It’s a trade, not a free win.
The right balance depends on how dirty your material is. On a machine without auto-set, dirty or older aluminum may need balance pushed as low as 65% EN to get enough cleaning. Very clean, freshly prepped parts can run closer to 80% EN. That’s the guidance behind MillerWelds’ AC balance control method, and it matches what you’ll feel in the booth.
Increasing the EN portion of the cycle reduces tungsten erosion and balling, lets you run a smaller-diameter tungsten, and narrows the etched zone for a cleaner finish. That’s why clean metal lets you bias toward EN: you don’t need the extra cleaning, so you spend the cycle where it does more for penetration and finish.
Here’s the discipline that separates welders who fight their machine from welders who control it: clean your aluminum before you touch the balance dial. Wipe off oils and solvent with a dedicated cleaner, then remove the oxide with a stainless brush reserved for aluminum. Balance is a last line of defense for the oxidation that survives prep, not a substitute for it. If you rely on the arc to do your cleaning, you’ll run low EN settings that ball your tungsten, over-etch your part, and still fight smut.
Good prep plus a sensible balance setting gives you an arc that cleans just enough and penetrates just right. Build that habit through repetition, and settings become instinct. For a structured way to build those skills, see our welding practice guide. And if you’re weighing AC against DC entirely, DC welding aluminium covers when direct current makes sense.
If AC balance decides how much cleaning you get, frequency and waveform decide how that arc behaves moment to moment. Frequency is how many times per second the current flips between electrode positive and electrode negative. Waveform is the shape of that flip. Together they let you shrink or stretch the arc on demand — the difference between a soft, wide puddle and a focused, precise one.
Frequency is measured in hertz (Hz), or cycles per second. A standard 60 Hz arc flips polarity 60 times every second, and each cycle contains one electrode-negative and one electrode-positive half. Modern inverter machines let you push that anywhere from roughly 20 Hz up to 400 Hz or more.
Here’s the practical effect:
For general aluminum work, Miller recommends 100–180 Hz. Crank above that — roughly 120–250 Hz — and the arc narrows into a focused beam. On thin-section U-groove joints, welding resources published by AWS report that pushing frequency near 220 Hz improves fusion where a softer arc struggles to tie in.
The waveform describes how the current crosses from one polarity to the other. Traditional transformer machines output a smooth sine wave, which eases through the zero point slowly. That slow crossing makes the arc less stable and harder to direct.
Square waves changed the game. They cross the zero point almost instantly, so the arc stays lit and controlled rather than flickering. The result is a more stable arc, sharper directional control, and better cleaning at the same settings.
Advanced waveforms go further. Fronius describes an optimized shape built from a rectangle with rounded corners for the negative half-wave and a sine wave for the positive half. That combination delivers the best balance of arc stability and oxide cleaning in most aluminum applications, which is why you’ll find it as the default on many modern inverters.
Frequency isn’t a set-and-forget number — it works hand in hand with material thickness and how fast you move. Higher frequency concentrates heat, so you can travel faster, keep the puddle tight, and limit distortion. Lower frequency spreads heat, so you slow down and let the puddle wet out over a wider area.
As a starting point: thin sheet gets higher frequency, faster travel, and lower amperage; thick plate gets lower frequency, slower travel, and more amps. If the arc feels too harsh or the cleaning zone narrows, ease the frequency back down. If you’re chasing a wandering puddle on thin stock, nudge it up.
Quick Note: Change one variable at a time. Bump frequency, watch the puddle for a few beads, then adjust balance or travel speed. Chasing two settings at once usually just hides what’s actually helping.
Once frequency and waveform click, the arc stops feeling unpredictable. From there, tightening up your hood time is what cements it — a structured welding practice routine makes these settings second nature. And it helps to remember the fundamentals from our guide to TIG polarity, because frequency only refines the polarity relationship you already set up.
Short answer: yes, but with serious caveats. Direct current is not the fast path to pretty aluminum welds, and for most people in most shops it is the wrong tool. Still, “you can’t TIG aluminum on DC” is a myth, and it’s worth understanding exactly why the technique works in niche cases and fails everywhere else.
The core issue is the same stubborn oxide layer that forces aluminum onto AC in the first place. Under DCEN (electrode negative), you get deep, focused penetration and excellent heat input, but the arc provides essentially no cleaning action to break up that oxide. Welding aluminum with DCEN GTAW is possible and does get used, but according to ESAB, it requires a greater level of skill because the oxide forms almost instantaneously and the arc will not scrub it away for you.
With DCEN, roughly two-thirds of the arc heat goes into the workpiece, which is why penetration and travel speed look so good on thick plate. The trouble is the other side of that coin: no cleaning. The moment you stop removing oxide mechanically, the puddle skins over, black specks float to the surface, and the filler refuses to wet out. You end up fighting the metal instead of welding it.
That makes DCEN a cleaning-management game, not a polarity trick. The weld only works if your mechanical prep is immaculate, and it must happen right before the arc starts. Dedicated DC welding aluminium approaches lean hard on this, accepting the prep burden in exchange for a tighter, hotter arc.
DCEP (electrode positive) flips the physics. Now the welder is the positive electrode, so the arc actually provides cleaning action and breaks up the aluminum oxide. Sounds great until you check the tungsten. On DCEP, most of the arc energy dumps into the electrode tip instead of the plate, and it heats the tungsten past its melting point within seconds. The tip balls up, sputters, and sheds material into the puddle, contaminating the weld while your effective amperage collapses.
Quick Note: This is exactly why AC exists. Alternating current splits the job, giving you electrode-negative half-cycles for penetration and electrode-positive half-cycles for cleaning. See our full breakdown of TIG welding polarity for how those half-cycles trade off.
Because of this, DCEP is never used alone on aluminum. It is not a viable production setup, period. Its cleaning action is real, but the tungsten destruction is fatal to any continuous weld.
DCEN aluminum welding lives in specific, narrow contexts rather than general fabrication:
In these cases, the welder is essentially choosing to trade cleaning convenience for arc intensity, and then paying for it with rigorous surface prep.
Step back and the picture is simple. AC is the standard for TIG welding aluminum and magnesium precisely because it splits the cleaning and penetration duties between half-cycles, which mirrors how CK Worldwide frames the AC-versus-DC decision. DCEN gives you penetration with no cleaning; DCEP gives you cleaning with a self-destructing tungsten. Only AC gives you both in one arc.
So treat DCEN aluminum as an advanced specialty tool, not a shortcut. If you’re still building fundamentals, spend your hours on welding practice with AC before chasing niche DC technique. Mastery there makes the DCEN exceptions easy to recognize, and easy to avoid when they aren’t worth it.
Setting your machine to AC is only half the job. The other half is everything bolted to the torch and sitting on your bench. Get the tungsten, gas, and surface prep right, and the polarity you dialed in has something clean to work with. Skip any one of them and you are back to black specks and a scummy puddle, no matter how good your settings look.
For AC aluminum, the classic choice is pure tungsten (green). It balls up under the alternating current and holds a stable, soft arc. The trade-off is that pure tungsten has a lower current capacity and doesn’t like the higher amperages. On modern inverter machines, ceriated (gray) and lanthanated (blue or gold) tungsten have largely taken over. They keep a stable arc on AC while handling more heat and lasting longer, so if your machine supports them, they’re a smart upgrade.
A slightly balled tip is what you want for AC aluminum. Unlike DC work, where a sharp point focuses the arc, AC needs a rounded end. The ball forms naturally if you start at low amperage and let it build, or you can form it intentionally. Don’t overdo it, though: a big, wobbly ball gives you a wandering arc and a washed-out puddle.
When you grind tungsten, run the stone lengthwise along the electrode rather than across it. The fine longitudinal scratches keep the arc focused and stop it from wandering. As for diameter, a 3/32-inch tungsten is a solid general-purpose starting point, good for up to about 220 amps. Step down to 1/16-inch for thin sheet and go up to 1/8-inch when you’re pushing serious current on thick plate.
| Application | Tungsten | Diameter |
|---|---|---|
| Thin aluminum sheet | Ceriated or lanthanated | 1/16 in |
| General work, up to ~220 A | Ceriated or lanthanated | 3/32 in |
| Thick plate, high amps | Pure or lanthanated | 1/8 in |
Argon is the baseline shielding gas for TIG, and it works well for aluminum on AC. Set your flow rate to 15-20 cfh (MillerWelds). Too low and you lose coverage; too high and the turbulence pulls in air. If you want deeper penetration on thick aluminum, an argon/helium blend runs hotter and helps you move faster, at the cost of a stiffer arc and higher gas expense.
Don’t forget gas timing. Pre-flow gives the arc a clean pocket of shielding before it starts, and it protects the tungsten from popping at ignition. More importantly, keep the post-flow running 10-15 seconds after you stop the arc (MillerWelds). The tungsten and fresh weld metal are still hot enough to oxidize in open air. Cutting post-flow short is a common reason welds come out dull and gray.
Quick Note: Never use a steel brush on aluminum. A cheap, dedicated stainless brush costs almost nothing and saves you from a weld full of contamination you won’t see until it’s too late.
Aluminum’s oxide layer melts around 3,700°F while the metal underneath melts near 1,220°F. That gap is why surface prep matters so much. A dedicated stainless-steel wire brush is the standard tool. Use one that touches only aluminum. A brush that’s been on steel will transfer contamination and cause discoloration and rust-colored specks in your weld.
For clean stock, brush in one direction, then wipe with acetone to remove oils and residue. On heavily oxidized parts like marine components exposed to saltwater, brushing isn’t enough. You may need a heavy file or carbide burr to cut through the thick oxide before you ever touch it with a brush (MillerWelds). Draw filing along the joint gives you a fresh, bright surface that the cleaning half of the AC cycle can finish off.
Think of it as three legs of a stool: correct polarity, proper gas, and a clean surface. Miss one and the stool tips over. If you’re still building your habits at the bench, a structured approach to welding practice pays off fast. And if you want the polarity side of the equation fully mapped out, our guide to TIG welding polarity breaks down exactly what each setting does.
When an aluminum bead goes wrong, the mistake is almost never mysterious. It usually traces back to a polarity setting, a dirty surface, or an impatient filler hand. The fix starts with a troubleshooting mindset: read the symptom, name the cause, then change one thing at a time and watch how the puddle responds.
The most common error is leaving the machine on DCEN after welding steel. On DCEN, the arc concentrates heat into the work but sends no cleaning action to the electrode-positive half of the cycle, so aluminum’s oxide layer stays put. The result is a scummy, contaminated bead that never wets out properly, exactly what MillerWelds describes when the wrong polarity is used on aluminum.
The opposite slip is selecting DCEP. Full electrode-positive dumps heat into the tungsten, balling and melting it while the workpiece stays cold. If your tungsten mushrooms and the plate barely reacts, check the polarity switch before blaming your technique.
Mixed polarity hides well and deserves its own warning. Using the wrong polarity on aluminum leads to contamination, and a DCEN arc fails to break down the oxide layer at all, producing a contaminated bead. If your bead looks grainy and gray instead of bright and fluid, stop and verify the mode first. There is a narrow place for DC welding aluminium, but standard AC work is not it.
Quick Note: For AC TIG, if arc instability persists, adjust balance toward a higher EN percentage, increase travel speed, or add filler. These three moves settle a wandering arc faster than fiddling with every dial at once.
| Symptom | Likely Cause | Fix |
|---|---|---|
| Black specks or peppering | Too little cleaning (EP) time | Shift balance toward more EP, clean the surface |
| Wide etched zone, bald or balled tungsten | Too much EP | Reduce EP percentage, raise EN |
| Rust-colored discoloration near the weld | Tungsten contaminated with steel, then only reground | Cut the bad end back, don’t just regrind |
| Dull, contaminated bead that won’t wet in | Mixed polarity or uncleaned metal | Confirm AC, degrease, stainless-brush |
| Poor root fusion | Excessive arc length, poor fit-up, low heat | Shorten the arc, clean base metal, tighten the gap |
The puddle tells you when to dip. Adding filler before the oxide is removed, rather than waiting for a shiny-dot puddle, results in weld contamination, according to this TIG basics guide. Give the arc a moment to form a bright, mirror-like pool that reflects light and flows outward. That reflection is your green light to add rod; feeding too early drags oxide and soot straight into the joint.
A tungsten that touched the puddle or ran through steel is not saved by a quick regrind. A tungsten contaminated with steel and simply reground can cause rust-colored discoloration near the welds, because the trapped steel vaporizes in the arc. Break off the bad end, regrind it lengthwise on a dedicated wheel, and keep a spare ready so you never hesitate mid-bead.
Wrong gas is another quiet saboteur. AC TIG on aluminum wants straight argon; a C25 mix intended for MIG leaves a sooty, unstable arc and wastes your settings. Check the cylinder label and set flow around 15 to 20 CFH before you chase polarity gremlins.
Lack of fusion at the root is often caused by improper fit-up, excessive arc length, or insufficient heat, and it is rarely a polarity problem. Shorten the arc and clean the base metal, then check your joint gap. On aluminum you want roughly 1/16 to 1/8 inch above the workpiece, so the arc stays tight and the heat goes exactly where you aim it.
When you tighten the arc, watch the balance too. A setting that was stable at a long arc can pepper or etch at a short one. To understand the dials underneath all of this, our guide on TIG welding polarity: DCEN, DCEP, and AC breaks down each mode. Then put in the reps, because welding practice is what turns a checklist into muscle memory. The habit that fixes most aluminum headaches is always the same: start with clean metal, confirm your polarity, and read the puddle before you move.
Knowing why AC wins is one thing. Turning it into a repeatable, clean bead happens only when you build a routine and drill it until your hands stop thinking.
Run this every time, even on scrap. A skipped step shows up immediately in the puddle.
Quick Note: A typical starting point on modern machines is about 75% EN balance. If a weld can be completed without any peppering, the balance setting is likely adequate. Forgetting this quick test is why so many first aluminum beads come out looking dirty.
Watch the puddle, not the tungsten. The molten pool tells you everything: if it looks dull or sluggish, you need more amperage or a cleaner surface. When it turns bright and mirrors the arc, it is ready for filler. Adding filler at the right moment produces the shiny finish associated with good TIG welds; adding it too early or too late leaves a grainy, contaminated bead. Tie that observation to a steady rhythm and your consistency jumps.
Aluminum does not forgive much torch tilt — keep the angle close to 90 degrees, or as near to vertical as possible. Anything more than a slight lean pushes heat sideways and drags the arc off the joint. Control amperage with the pedal as the part heats up instead of charging through at a fixed setting; aluminum soaks heat fast, so what felt cold on the first inch can be near collapse by the last. Travel faster than feels natural.
| Material | Amperage | Balance |
|---|---|---|
| 1/16 in (1.6 mm) | 70–90 A | ~75% EN |
| 1/8 in (3.2 mm) | 120–150 A | ~75% EN |
| 3/16 in (4.8 mm) | 180–220 A | ~70–75% EN |
Write your winning settings in a notebook by thickness and alloy. Modern inverter machines vary widely, and a personal chart beats guessing when you switch materials. Keep a dedicated stainless brush for aluminum only, handle and store filler wire carefully so it stays clean, and wear clean gloves; TIG gloves are made without the oils used in regular leather work gloves, which can contaminate the puddle. A tidy, well-lit bench makes all of this easier, and a properly planned shop pays off in cleaner welds. Safety gear belongs in the routine too, starting with protective eyewear for welding. Then put in the reps; a structured approach to welding practice is what turns these settings into muscle memory.
Polarity is the star of this article, but a perfectly dialed AC balance means nothing if the shop around you is unsafe. Before you chase that last bit of puddle control, lock in the fundamentals that keep you welding for decades.
Aluminum welding produces fume, and the exposure numbers are worth memorizing. OSHA sets a permissible exposure limit (PEL) for aluminum welding fumes at 5 mg/m³ (as Al) over an 8-hour shift, with 15 mg/m³ for total dust, as documented in filler-metal safety data like the Airgas SDS. Aluminum oxide carries the same OSHA PELs — 5 mg/m³ respirable and 15 mg/m³ total dust — while ACGIH recommends a stricter 1 mg/m³ for the respirable fraction.
Ventilation and local exhaust are the primary controls under OSHA’s hierarchy of controls; respirators are a secondary layer. AWS F3.2M/F3.2, Ventilation Guide for Weld Fume, is the recommended practice for designing that airflow, and ANSI Z49.1, Safety in Welding and Cutting, is the consensus standard that underpins most shop rules. If you’re outfitting a booth from scratch, our guide to planning a safe shop walks through layout, airflow, and extraction.
Quick Note: Aim your exhaust so the plume is pulled across the work and away from your breathing zone. A small downdraft table or fume extractor near the arc beats a shop fan every time.
AC TIG throws intense light and UV, and bare aluminum reflects it back at you. Protect yourself with:
Our breakdown of protective eyewear for welding covers lens shades in detail. On the electrical side, remove conductive jewelry, insulate your body from the work, inspect torch cables and grounds for damage, and never weld in wet conditions. Open-circuit voltage is real, and a sweaty glove is a poor insulator.
Everything in this article traces back to a single fact: aluminum’s tenacious oxide layer. That’s why you run AC, set balance to match the material’s condition — more cleaning for cast or heavily oxidized parts, less for clean, new sheet — keep the arc tight, and clean religiously before, during, and between passes. Polarity is not an arbitrary switch. It’s a decision that flows directly from how the metal behaves, and once that clicks, the settings stop feeling random.
Materials tie into this too. Match your filler to the base alloy, keep it sealed and dry, and work on stock that’s been cleaned, not weathered. Good inputs make polarity tuning predictable; bad ones make every setting feel wrong.
Your next step is simple. Grab a scrap plate, set the machine to AC, dial in balance, and run bead after bead until the puddle behaves the way you predict. Deliberate repetition is the heart of real welding practice, and it turns a confusing control panel into second nature. For a broader refresher on current types, revisit our TIG welding polarity guide, or read up on DC welding aluminium if your work calls for it. Master the polarity, and aluminum stops fighting you.
Use alternating current (AC). AC alternates between electrode positive, which blasts away the aluminum oxide layer, and electrode negative, which drives heat into the base metal for penetration. Neither DCEN nor DCEP alone can clean and penetrate aluminum without causing contamination or destroying the tungsten.
Yes, it is possible, and some shops do it on thicker material or with helium-rich gas for deeper penetration. However, DCEN provides no oxide cleaning action, so you must clean the aluminum thoroughly and mechanically, and the process demands more skill and produces less reliable, dirtier puddles than AC.
In DCEP the electrode is positive, and roughly two-thirds of the arc heat concentrates at the positive anode. That heat balls and melts the tungsten almost instantly, contaminating the weld. DCEP provides good cleaning but almost no usable penetration in GTAW, so it is only a niche choice for very thin sheet.
A common starting point on modern inverter machines is about 75% EN. Dirty or heavily oxidized aluminum may need a lower setting (more EP cleaning), around 65%, while very clean parts can tolerate 80% or higher. Always test weld on scrap of the same material before committing to the real joint.
Black specks, or peppering, usually mean insufficient cleaning action, so decrease the balance setting to spend more time in the EP half-cycle. It can also come from contaminated gas, a dirty filler rod, or adding filler before the puddle becomes a shiny dot. Fix prep first, then adjust balance.
A general starting range is about 100-180 Hz. Lower frequencies give a wider, softer arc and a broader bead, while higher frequencies (up to roughly 250 Hz) produce a tighter, more focused arc that is better for thin material and precise work. Increase frequency on thin sections to concentrate heat.
Yes. Pure (green) tungsten has traditionally been used for AC aluminum because it balls cleanly, but ceriated (gray) and lanthanated (blue or gold) tungstens are now common on inverter machines and work well on AC. Prepare a slightly balled tip for AC instead of the sharp cone used for DCEN on steel.
A dull, scummy, or oxidized-looking bead usually means the oxide layer was not fully removed or the shielding was inadequate. Check that your machine is actually in AC, increase EP cleaning (lower balance), verify 15-20 cfh gas flow, keep a short arc, and make sure post-flow is protecting the cooling weld.