Carbon Steel TIG Welding: The Complete Guide

A first-year fabricator once spent an entire afternoon chasing porosity in a mild steel bracket, convinced his torch was broken — the real culprit was a greasy plate and a draft from an open shop door. Carbon steel TIG welding rewards patience and punishes small oversights, and understanding why is the difference between a weld that looks like jewelry and one that cracks overnight.

What Is Carbon Steel TIG Welding and Why Does It Matter?

A first-year fabricator once spent an entire afternoon chasing porosity in a mild steel bracket, convinced his torch was broken. The real culprit was a greasy plate and a draft from an open shop door. Carbon steel TIG welding rewards patience and punishes small oversights, and understanding why is the difference between a weld that looks like jewelry and one that cracks overnight.

Let’s answer the principal question first: yes, you can TIG weld carbon steel, and in fact it is one of the most reliable, widely specified combinations in fabrication. Gas tungsten arc welding (GTAW), commonly called TIG, uses a non-consumable tungsten electrode to establish the arc while an externally supplied inert gas shield protects the puddle, and the filler metal is fed in separately by hand. The OSHA welding fact sheet describes this arrangement as producing a superior finish compared with other arc processes, which is exactly why specifiers call it out when appearance, penetration control or X-ray quality matters.

Why fabricators choose TIG on steel

The reasons come down to control. TIG lets you dial amperage precisely with a foot pedal, feed filler one dip at a time, and stop the weld mid-run without leaving slag. The result is a clean, slag-free bead you can inspect visually the moment it cools, low spatter, minimal distortion on thin material, and full access to out-of-position work. For root passes on pipe, thin-wall tubing, pressure vessel tie-ins and precision repair, that control is not a luxury, it is the requirement.

Common carbon and low-alloy base metals

Most everyday structural and pressure-grade steels are TIG candidates. The ones you will actually meet include:

• ASTM A36 — the general structural workhorse: plate, angles, channels and framing.
• ASTM A516-70 — normalized plate for pressure vessels and boilers, often welded in heavier sections where heat control matters.
• ASTM A106 Grade B and C — seamless carbon steel pipe for high-temperature service, where TIG is the standard process for the root pass before a fill pass in another process.
• 1018 and 1020 bar stock — cold-rolled and machined parts, brackets and repair work where a tidy bead is expected.

Filler selection follows the base metal. ER70S-2, the triple-deoxidized rod covered in depth later in this guide, is typically matched to ASTM A36, A285-C, A515-55 and A516-70 (UNS K02600, K02801, K02001 and K02700 respectively), which covers a large share of shop work in one rod size.

Where the process is actually used

You will find carbon steel TIG welding in power plant and refinery piping, pressure vessel fabrication, hydraulic tubing, thin sheet enclosures, ornamental and architectural work, tool and die repair, motorsport and chassis fabrication, and any cosmetic weld that will be left visible. It is also the preferred method for joining sections where backside access is limited and a controlled, fully penetrated root is essential.

It also acts as a gateway process. Once you understand arc length, travel speed and filler rhythm on steel, moving to other materials becomes far less intimidating, whether that means aluminum MIG welding or an alloy steel procedure written to a code. The fundamentals of puddle control transfer; only the settings and consumables change. Your shielding gas comes from a cylinder supplied by a reputable welding gas supplier, and pure argon remains the baseline for everything in this article.

What this guide covers

The rest of this article works through the process in the order you need it: how the arc behaves on carbon steel, which filler rod to load, starting amperage and gas settings by thickness, joint prep and torch technique, preheat and hydrogen control for thicker or harder steels, a defect troubleshooting table, an honest TIG-versus-MIG-versus-stick comparison, fume and safety requirements, and a four-week practice plan with a buying checklist. By the end you should be able to set up a machine, prep a joint and lay a sound bead on A36 plate without guessing.

How the GTAW Process Works on Carbon Steel

Before you strike an arc, it helps to understand what is happening inside that glowing cone of plasma. Gas tungsten arc welding, shortened to GTAW and called TIG (tungsten inert gas) by most fabricators, does something deceptively simple: it uses a non-consumable tungsten electrode to create a stable electric arc, while the filler metal is fed into the puddle by hand from outside the arc. The electrode never melts into the weld. That single design choice is why TIG gives you more control than any other manual process, and why the American Welding Society documents it as a distinct process under its own standard, AWS D17.1 and related GTAW guidance.

Polarity: Why Carbon Steel Runs DCEN

Carbon and low-alloy steels are welded on direct current electrode negative, commonly called DCEN or straight polarity. The torch connects to the negative side of the circuit and the work clamp to the positive side. Electrons flow from the tungsten into the workpiece, concentrating roughly two-thirds of the arc heat at the plate rather than at the electrode. The result is a deep, focused penetration profile with a relatively small, cool heat-affected zone, and a tungsten that runs cool enough to hold its shape through a long root pass.

Alternating current is what you would use on aluminum, because AC’s cleaning action strips the stubborn oxide layer that melts around 3,700°F while the metal beneath it melts at 1,220°F. Steel has no such oxide problem, so AC’s advantages become pure downside here: softer arc, shallower penetration, faster tungsten erosion. If you are curious where AC really earns its keep, our aluminum MIG welding guide covers the other side of that coin. On steel, set the machine to DCEN and leave it there.

Shielding Gas: Pure Argon Is the Baseline

For manual TIG on carbon steel, 100% argon is the standard starting point, and for most thin-to-moderate thickness work it is also the finish line. Argon ionizes easily, giving a smooth, predictable arc that is easy to start and direct, and it welds cleanly without the spatter you get from CO₂ blends. Helium mixes are sometimes used on very heavy sections to push more heat into the joint, and small hydrogen or oxygen additions appear in some specialty European codes, but neither is where a general fabricator should start.

Published manufacturer data sheets typically list argon flow rates between 20 and 30 cubic feet per hour (roughly 8 to 15 liters per minute), depending on cup size, amperage and ambient draft. More flow is not more protection. Excessive gas velocity pulls air in through turbulence and actually worsens shielding, which is exactly the failure that ruined the bracket in our opening story.

Tungsten Selection and Prep

Three types cover nearly all carbon steel TIG work:

• EWTh-2 (2% thoriated, red band) — the classic choice; manufacturer charts pair it with a truncated conical tip for GTAW on steel. It starts easily and resists contamination at high amperage, though it is slightly radioactive in dust form, so grinding needs control.
• EWG 2% lanthanated (blue band) — a popular non-radioactive all-rounder that runs almost identically to thoriated and works on both AC and DC.
• EWCe-2 (2% ceriated, gray band) — excellent for arc starts on thin material and low-amperage repair work.

Cups, Gas Coverage and Adding Filler

The gas cup, or nozzle, simply delivers argon around the arc. A #6 or #7 cup handles most plate work; larger cups, or a gas lens insert, give a wider, slower column of gas that protects the hot puddle longer. That trailing coverage matters on the root pass of a pipe, where the back side of the puddle is still above 700°F and can oxidize seconds after the arc moves on.

Filler rod is added at the leading edge of the puddle, dipped into the joint and pulled back out of the gas envelope when you pause. The rod melts by conduction from the puddle, not by holding it in the arc — resting the rod in the arc is a common beginner habit that causes spatter and, if the rod touches the tungsten, an immediate inclusion. That contaminated tungsten then transfers specks into every subsequent weld until you stop, regrind and start again. If you want to see how far TIG precision can be pushed on the same fundamentals, our look at aerospace TIG welding is a useful reference point.

The through-line is control. Current sets heat, argon sets protection, tungsten geometry sets arc character, and your hand sets everything else. Once those four variables are understood, the settings tables and joint-prep details in the next sections stop feeling like arbitrary numbers and start making sense as a system.

Choosing the Right Filler Metal: ER70S-2, ER70S-3 and ER70S-6

Pick the wrong filler rod and even perfect torch technique will not save the weld. In carbon steel TIG welding, the filler metal decides how forgiving you are of a dirty joint, how the puddle wets out, and whether the finished bead survives an X-ray. Fortunately, the choice is narrower than it looks. Almost every carbon steel GTAW job runs one of three rods from the AWS A5.18 classification: ER70S-2, ER70S-3, or ER70S-6. Here is how to read those codes and pick the right rod.

Reading the AWS A5.18 Classification

The designator looks like a part number, but it is really a specification summary. Work through it left to right:

• E means electrode. GTAW rods are marked ER because they do double duty: electrode in the arc, plus rod added as filler.
• 70 is the minimum specified tensile strength in thousands of pounds per square inch — 70,000 psi, or about 480 MPa.
• S means solid, as opposed to a cored or composite wire.
• -2, -3, or -6 is the deoxidizer package and chemistry variant. This suffix is where the real differences live.

Under AWS A5.18, all three meet the same floor: 480 MPa (70 ksi) tensile strength, 400 MPa (58 ksi) yield strength, and 22% elongation, with impact-rated grades requiring 27 J (about 20 ft-lb) at −30°C. Those minimums are the ticket to entry, not the whole story. Actual as-welded GTAW deposits from ER70S-2 routinely beat them, reaching roughly 530 MPa yield and 590–660 MPa tensile with around 31% elongation. In plain language, the weld metal is usually stronger than the A36 or A516 plate it joins, so the joint is rarely the weak link. Data sheets from Lincoln Electric, Harris Products Group, Pinnacle Alloys and Washington Alloy all confirm this pattern.

ER70S-2: The Triple-Deoxidized Root-Pass Rod

ER70S-2 carries three deoxidizers: aluminum at 0.05–0.15%, titanium at 0.05–0.15%, and zirconium at 0.02–0.12%. Each scavenges oxygen before it can become a gas pore in the finished weld. Aluminum and zirconium are powerful deoxidizers, while titanium also stabilizes nitrogen and ties up residual elements.

That triple-deoxidized chemistry makes ER70S-2 the default for critical root passes, open-root pipe, and any joint that cannot be perfectly cleaned. It tolerates light mill scale and minor contamination better than the other two, and it lays down weld metal reliable enough for radiographic examination. If you are welding pressure pipe, boiler work, or structural joints subject to AWS D1.1-class inspection, start here and stay here unless a written WPS says otherwise. It is also the rod to keep for repairs on components of unknown history, like the greasy brackets and rusty plate that walk into every fabrication shop.

ER70S-3 and ER70S-6: Where They Fit

ER70S-3 drops the aluminum, titanium, and zirconium and relies on milder silicon and manganese deoxidation. The result is a slightly quieter arc and a cleaner-looking bead on immaculately prepared joints, which is why many TIG welders prefer it for cosmetic work on clean, thin sheet. The trade-off is reduced tolerance for contamination, so joint prep matters more.

ER70S-6 runs higher silicon and manganese. Those elements lower the surface tension of the puddle, so the weld wets and spreads beautifully over mill scale. That makes it a favorite short-circuit and spray-transfer MIG wire, and it works fine under the TIG arc too. The catch: the extra silicon can produce slightly heavier slag and, in some environments, more silicon segregation in the finished bead. It is a good general-purpose rod, but for X-ray-quality TIG root passes, ER70S-2 remains the safer choice.

Property	ER70S-2	ER70S-3	ER70S-6
Deoxidizers	Al, Ti, Zr (triple)	Si, Mn only	Si, Mn (higher)
Typical Mn	0.90–1.40%	0.90–1.40%	1.40–1.85%
Typical Si	0.40–0.70%	0.45–0.75%	0.80–1.15%
Min. tensile / yield	480 MPa / 400 MPa	480 MPa / 400 MPa	480 MPa / 400 MPa
Contamination tolerance	Excellent	Fair	Good
Best TIG use	Root passes, pipe, dirty metal	Cosmetic, clean thin sheet	General fabrication, scaled plate

Rod Diameter vs. Material Thickness

Match filler diameter to the thinnest section you are joining, not the thickest. As a working baseline:

• 0.035 in (0.9 mm) — sheet from 22 ga up to about 1/16 in; also fine for precision repair and thin-wall tube.
• 1/16 in (1.6 mm) — the workhorse size for 1/16 to 3/16 in plate, and the first roll to buy.
• 3/32 in (2.4 mm) — 3/16 to 1/4 in plate, and the standard choice for open-root pipe hot passes.
• 1/8 in (3.2 mm) — 1/4 in and heavier, where the puddle can swallow a smaller rod without waiting.

One last habit: match the rod to the position and joint you are actually welding. A rod that flows perfectly flat will fight you overhead. Filler selection is cheap to change and expensive to get wrong, so when in doubt, run a short test bead on scrap of the same thickness and heat cycle before you strike the arc on the real part.

Dialing In Settings: Amperage, Travel Speed, Gas and Cup Size

Every conversation about carbon steel TIG settings comes back to one number: amperage. Cup size, gas flow and travel speed all support it, but current is what actually melts the puddle. The good news is that published parameter tables and a couple of rules of thumb get you close on almost any thickness; from there, the puddle tells you what to change.

The Amps-Per-Thousandths Rule of Thumb

The classic starting point for TIG welding mild steel flat butt joints is roughly one amp per 0.001 in of thickness at room temperature. A clean 0.125 in (1/8 in) plate starts around 110 to 125 amps; 0.250 in plate starts around 200 to 250 amps before heat sinking. This is a launchpad, not a specification, and three things move it immediately:

• Joint type. An open-root butt joint needs less current than a fillet weld of the same leg size, because a T-joint pulls heat into two plates acting as heat sinks.
• Position. Manufacturer guidance recommends decreasing amperage by about 15% for out-of-position welding, since gravity works against a vertical or overhead puddle.
• Restraint and preheat. A thick, clamped assembly soaks up current; a preheated part already starts hot and needs less from the arc.

A Practical Starting-Parameter Table

These flat-position starting points are for clean, mill-scale-free low-carbon steel with 100% argon shielding and DC electrode negative:

Plate thickness	Filler dia.	Tungsten dia.	Amperage	Cup	Flow (cfh)
1/16 in (1.6 mm)	1/16 in	0.040 or 1/16 in	40–60 A	#5–#6	15
1/8 in (3.2 mm)	3/32 in	3/32 in	80–110 A	#6–#8	15–20
3/16 in (4.8 mm)	3/32 in	3/32 in	110–140 A	#8	15–20
1/4 in (6.4 mm)	1/8 in	3/32 or 1/8 in	140–190 A	#8–#10	20
3/8 in (9.5 mm)	1/8 or 5/32 in	1/8 in	190–260 A	#10–#12	20–25
1/2 in (12.7 mm)	5/32 in	1/8 or 3/16 in	250–320 A	#12	20–25

What the Published Tables Actually Say

Filler metal manufacturers publish current ranges by rod diameter rather than plate thickness, which tells you when to step up a rod size. On 100% argon, published GTAW tables list 1/16 in rod at 50–150 A, 3/32 in at 85–250 A, 1/8 in at 90–280 A and 5/32 in at 120–320 A. A second table ties filler to base metal directly: 1/16 in filler at 100–160 A for 1/16 to 3/32 in base metal, and 1/8 in filler at 150–300 A for 1/4 to 1/2 in plate. Lincoln Electric’s ER70S-2 TIG filler data is a good reference to keep on the bench, and Washington Alloy publishes comparable ER70S-2 ranges.

Notice how wide those bands are. That is not sloppiness; it reflects position, joint design, travel speed and heat sinking. The practical lesson is rod selection: past roughly 180 amps, a 3/32 in rod glows, softens and melts back into the puddle. Move to 1/8 in rod instead of fighting it, and let the bigger rod absorb heat while delivering more metal per dab.

Voltage, Travel Speed and Heat Input

TIG welders rarely think about voltage, but published GTAW data puts a typical arc at roughly 9 to 19 volts depending on arc length and current. Shorter arcs mean lower voltage, less heat input and better shielding. Under AWS D1.1, heat input is calculated from voltage, current, travel speed and a process efficiency factor, which is why slowing down at constant amperage can over-penetrate just as surely as turning the machine up. A balanced travel speed keeps the puddle roughly three to three-and-a-half times the filler rod diameter wide. Too narrow and cold risks lack of fusion; too wide and shiny means slow travel or excessive current.

Cup Size and Gas Flow

Cups from #6 to #8 cover most light and medium work, while #10 to #12 cups give the wider gas envelope needed for hot root passes and large-diameter rod. Bigger cups demand more flow, but more flow is not automatically better: above roughly 20 to 25 cfh on a standard ceramic cup, the shielding gas turns turbulent and pulls air into the stream instead of pushing it away. A gas lens smooths that flow and lets you drop to 12–15 cfh even under a #12 cup.

Fine-tuning mostly happens through the foot pedal: a puddle that refuses to wet out and stays frosty wants a few more amps or a slower travel speed, while one that sags, flattens or blows through wants less current and a quicker hand. Learn to argue with the puddle instead of the machine, and the settings table becomes a confident place to start.

Joint Preparation and Torch Technique That Actually Works

Good carbon steel TIG welding is decided before you strike an arc. Most ugly welds, cold laps and porosity runs trace back to a joint that was never properly cleaned, badly gapped or tacked carelessly. Here is how to set up a joint so the puddle behaves, and how to move the torch once it does.

Start With the Surface: Cleaning Before the Arc

Mill scale, rust, oil, cutting fluid, paint and zinc coatings all contaminate the arc. FHWA bridge welding guidance is blunt about it: contaminated surfaces and heavy mill scale cause porosity and inhibited fusion, which is why structural shops chemically clean or grind bright the areas that will be welded. FHWA steel bridge welding guidance

Deoxidized fillers such as ER70S-2 tolerate dirty joints better than most, and Lincoln even describes ER70S-2 as well suited to single-side, melt-through welding without a protective root shielding gas on the back side. Lincoln Electric ER70S-2 data That tolerance is a margin, not a license. Scale still insulates the puddle from the base metal, and oil still burns into gas. OSHA 1910.252 adds a legal side: welding surfaces must be cleaned of coatings that could create toxic exposure, such as paint and solvent residue.

Bevel Angles, Lands and Fit-Up

For groove welds, joint geometry controls penetration and bead shape. These are typical starting dimensions, always subordinate to a written WPS on code work:

Joint	Bevel per side	Land (face)	Root opening
Butt, thin sheet (up to 3/16 in)	Square edge	None	0 – 1/32 in
Butt, plate 1/4–1/2 in	30–37.5°	1/16 in	1/16 – 1/8 in
Pipe root (6G-style)	37.5° (75° included)	1/16 in	1/8 in
Heavy wall pipe	30–37.5°	1/16–3/32 in	1/8 in per WPS

Grind the land smooth and parallel and remove the sharp burr from cutting. Tack on a file, not by welding on the bevel face, where the tack becomes part of the root. Use four to six tacks for small pipe, more on large diameter, and stagger them. Check the gap at each tack station: a gap that closes as you weld is the fastest route to lack of fusion.

Torch Geometry, Dab Rhythm and Walking the Cup

Keep the tungsten extension (stickout) short, roughly 1/8 to 1/4 in past the cup lip. Long stickout looks convenient but dumps shielding, and the weld goes gray. Work angle stays near perpendicular for butt joints, with a 5–15° trailing drag. The filler rod enters at a shallow angle off to the side and stays out of the arc until you deliberately dab it into the leading edge of the puddle.

Two feeding styles dominate. Dab feeding is the classic rhythm: establish the puddle, dip the rod, lift it slightly, move forward, repeat. It gives precise control of deposit volume, so beginners should learn it first. Continuous feed pushes the rod into the leading edge as the puddle moves, and experienced pipewelders often blend both. On pipe roots, dab feeding with a tight arc lets you watch the joint edges melt through evenly.

Walking the cup rests the nozzle on the bevel walls and rolls the torch in a figure-eight, advancing a set distance per weave. It produces consistent travel speed, even ripple spacing and less hand fatigue on long runs, but it teaches bad habits on open-root sheet work, where the cup cannot rest on anything. Aerospace-grade precision welding expects the same discipline: see Aerospace TIG Welding: Standards, Techniques, and Training for how technique scales up to regulated work.

On hot root passes and long stringers, trailing gas coverage matters. A larger cup, a gas lens or a trailing shield extends argon past the weld zone so the still-glowing metal stays wrapped. That is what keeps the back side of a hot root straw-colored instead of blue-gray.

The Prep Checklist That Prevents 90% of Bad Welds

1. Grind or chemically clean 1 in on either side of the joint back to bright metal.
2. Remove oil, coolant, paint and any zinc coating with solvent, then wipe dry.
3. Cut bevels to the specified angle; grind the land flat and deburr both faces.
4. Check fit-up with a gap gauge; tack with matching filler, four to six tacks.
5. Dress tungsten to the correct grind angle and install a clean, dry cup.
6. Verify gas flow and check for drafts; close doors or shield the arc.
7. Run a coupon on scrap of the same thickness before welding the real part.

None of this is glamorous, but every step directly attacks a specific defect, and it costs minutes compared with cutting out a failed root.

Preheat, Interpass Temperature and Hydrogen Control

A quarter-inch bracket in A36 plate almost never needs preheat. Move to 4130 tubing, a 1045 shaft, a one-inch base plate, or a fully restrained joint, and heat management stops being optional. This is where the metallurgy catches up with the technique: a weld can look flawless on the surface and still crack hours later in the heat-affected zone because hydrogen, hardness and stress all showed up at the same time.

Carbon Equivalent: The Number That Predicts Trouble

Steelmakers express how hard a steel will get after welding as a single figure. The common IIW carbon equivalent formula is:

CE = C + Mn/6 + (Cr + Mo + V)/5 + (Ni + Cu)/6

AWS documents also use the Ito-Bessyo Pcm formula, which weights carbon more heavily and adds Nb and B. Whatever your standard calls for, the trend is the same: as carbon content or carbon equivalent rises, the heat-affected zone becomes harder and more brittle, so the required preheat and interpass temperature climb with it. Mild steel below roughly CE 0.40 is forgiving; high-carbon alloys and high-restraint joints are not. The exact minimum preheat and interpass temperatures in AWS D1.1 are selected from tables that account for base metal composition, material thickness, welding process, hydrogen level and joint restraint rather than from a single rule anyone can memorize.

How Hydrogen Cracks Steel in the HAZ

Delayed cracking needs three ingredients together: diffusible hydrogen, a hard susceptible microstructure, and tensile stress. Remove any one of them and the risk drops sharply. Hydrogen sources are everywhere: moisture in the air, oil and grease on the plate, mill scale, primer, galvanizing, and the flux coating on a stick electrode. Once dissolved, hydrogen migrates to the hardest, most stressed region, the HAZ, and can open a crack hours after the weld cools.

Here is where TIG quietly wins. GTAW has no flux and no coating beyond inert gas, so the process is inherently low hydrogen compared with shielded metal arc welding. Stick electrodes are managed with strictly controlled hydrogen categories and baking ovens precisely because their coverings pick up moisture; for hydrogen-sensitive work, low-hydrogen rods like E7018 are the code’s answer, and the wider world of coated consumables is covered in Alloy Steel Welding Electrodes: Selection, Use & Safety. With TIG, your main hydrogen control job is simply keeping the base metal and filler clean and dry. That advantage is real, but it does not excuse skipping preheat on a thick, restrained joint.

Setting Preheat and Interpass Temperatures

Preheat slows the cooling rate, giving hydrogen time to diffuse out before the steel hardens, and it reduces the thermal shock that creates restraint stress. In practice:

• Preheat: measure with a temp stick or infrared thermometer, never by eye. AWS D1.1 requires the base metal to be at least 32 °F (0 °C) before welding, and many jobs require considerably more. Typical targets run from 150–250 °F for moderate-carbon or moderately thick steel up to 400 °F and above for high CE or heavy sections.
• Interpass: keep every subsequent pass within the preheat window. The maximum interpass temperature for ordinary carbon steel commonly lands around 400–600 °F (204–315 °C), with the WPS as the final authority. Overshooting it softens the HAZ; undershooting it reintroduces the cracking risk.
• Heat input: stay inside the range your WPS specifies. TIG’s foot pedal makes this easier than any other process, since you can taper amperage at the crater and between passes without changing electrodes.

The AASHTO and bridge-fabrication guidance in the FHWA steel bridge research publication walks through how these requirements scale with section size and joint restraint.

When Post-Weld Heat Treatment Is Actually Required

PWHT serves one of two purposes: stress relief, typically a subcritical soak in the 1100–1200 °F range, or full heat treatment to restore mechanical properties after heavy welding. It is common on pressure vessels, thick-wall pipe and welded assemblies with high residual stress or hardness limits. But for ordinary carbon steel in normal service it is often unnecessary: filler metal manufacturers’ technical bulletins note that GTAW weld deposits may not require post-weld heat treatment for normal carbon steel service. A well-executed TIG weld with appropriate preheat, controlled interpass temperature and sound technique will frequently meet strength and ductility requirements as-welded. The governing document is always the drawing, code or written WPS, not a general rule of thumb.

Treat preheat the way a pipe welder treats a thermometer: check it, record it, and never guess. Getting this right is what keeps a beautiful-looking carbon steel TIG weld from becoming a Sunday-night failure.

Common Defects and How to Troubleshoot Them

Even experienced TIG welders produce defective welds; the difference is that good welders recognize the defect immediately and know what caused it. Most carbon steel TIG problems trace back to a small set of causes: contaminated base metal, wrong amperage, poor gas coverage, bad torch angle, or a rushed crater. Learning to read the bead is half the skill.

The Defect Diagnostic Table

Defect	Likely Causes	Fix
Porosity (pinholes)	Oil, grease, paint, or rust on the plate; draft from doors or fans; leaking torch valve or loose gas fitting; insufficient flow	Solvent-clean and wire-brush, move away from drafts, leak-test the torch with soapy water, verify 15–20 CFH argon
Tungsten inclusions	Tungsten touched the puddle; rod struck the electrode; a clipped tungsten fell into the joint	Stop, grind the contaminated end back into fresh metal, re-sharpen, re-weld the area
Lack of fusion / cold lap	Amperage too low; travel too fast; work angle too steep; filler piled on the surface instead of dabbed in	Increase amps 5–10 at a time, slow down, aim the arc at the joint line, feed filler into the puddle, not cold metal
Undercut	Amps too high; travel too fast; weaving without letting the puddle fill; wrong torch angle	Reduce amperage or speed, pause at the toes, add filler at the edges, keep a 10–15° drag angle
Crater cracks	Torch pulled away without filling the crater; stopping on a restart point; too-high termination current	Taper current with the pedal while feeding rod until the crater is slightly convex, never just yank the arc
Burn-through	Amperage too high; travel too slow; dwelling in one spot; thin edge without support	Drop 5–10 amps, speed up, use a copper backing bar as a heat sink, or pre-bend edges for tighter fit-up
Distortion	Unbalanced heat input; skip welding without staggering; heavy tacking; too much heat in thin sheet	Use stitch sequences, clamp to a fixture, pre-set a slight bow, alternate sides, keep interpass time short
Discoloration (straw-to-black oxide)	Too little shielding gas; contaminated gas line; arc extinguished before postflow ends; coverage too narrow for the bead	Increase flow modestly (turbulence pulls air in), upgrade to a gas lens, extend postflow
Worm tracks	Coated or galvanized steel, rust, or oil trapped under the arc releasing gas during solidification	Strip coatings with a grinder or appropriate chemical cleaner, weld on bare metal only

Why Cleaning Still Matters with Triple-Deoxidized Rod

It is tempting to assume ER70S-2 will burn through anything because its aluminium, titanium, and zirconium scavengers handle dissolved oxygen. They do, but they cannot neutralise petroleum, paint, or zinc. Hydrocarbons crack under the arc and release hydrogen and carbon straight into the puddle, producing porosity and worm tracks, while galvanized coatings vaporise and create both fume and fusion problems. Grind mill scale back about an inch from the joint on critical work, and wipe with acetone if the plate has ever been oiled.

Inspection: What Your Eyes Can and Cannot See

Visual inspection is the first and fastest quality gate, but it has hard limits. AWS D1.1 visual acceptance criteria address profile problems such as undercut, weld width variation, and crater cracking, yet a bead that looks perfect can still hide a fusion defect or internal porosity below the surface. Under D1.1, incomplete fusion and internal porosity may be revealed by radiographic (RT) or ultrasonic (UT) examination, and visual acceptance alone is never sufficient where the engineer has specified those methods.

A practical inspection routine for shop work:

1. During welding: watch the puddle, not the arc. A healthy puddle is bright and fluid; dark and stiff means too cold, wandering and overly fluid means too hot.
2. Immediately after: inspect stop and restart points, where crater cracks and lack of fusion hide.
3. After cooling: brush the bead, check undercut with a visual comparator or small mirror, and look for clean wetting at the toes.
4. For code work: hand it to a certified inspector and the specified RT/UT procedure. Never self-certify a pressure or structural weld.

One last diagnostic tip: if you make the same defect repeatedly across different joints, the problem is rarely the joint, it is your setup. Amperage 10 amps low produces cold lap on out-of-position welds just as readily as flat ones, and a gas leak follows you from plate to pipe. Fix the root cause before you change technique, and your scrap rate drops fast.

TIG vs MIG vs Stick on Carbon Steel: An Honest Comparison

Ask ten fabricators which process is best for carbon steel and you’ll get ten answers, each defended like a family recipe. The honest answer is that GTAW, GMAW and SMAW are three different tools solving three different problems on the same base metal. Choose wrong and you don’t just lose time; you can lose penetration, toughness, or a passing inspection.

The Three Processes Side by Side

Factor	TIG (GTAW)	MIG (GMAW)	Stick (SMAW)
Penetration	Deep, narrow, precisely controlled	Moderate, smooth profile	Deep, heavy on plate
Travel speed	Slowest	Fastest on thin material	Moderate
Operating cost	Highest: argon, tungsten, cut rod	Low cost per pound deposited	Cheap machine, some rod waste
Position work	All positions, pedal-controlled	All positions with the right mode	Best all-position capability
Common defects	Tungsten inclusions, gas-starved porosity	Porosity, cold lap, feed hiccups	Slag inclusions, lack of fusion
Skill curve	Steepest	Fastest to learn	Moderate
Cleanliness needed	Strict	Tolerates light mill scale	Tolerates rough conditions
Hydrogen risk	Inherently low	Low	Low only with low-hydrogen rods
Cosmetic result	Best	Good when tuned	Roughest bead
Best fit	Root passes, thin wall, precision repair	Sheet, production fabrication	Structural, field, heavy plate

When 75/25 Short-Circuit MIG Beats TIG

On carbon steel sheet under 1/8 in, GMAW in short-circuit transfer running a 75% argon / 25% CO2 mix is the recommended shielding gas blend for high travel speed with minimal spatter and minimal distortion. A typical ER70S-2 short-circuit program on 0.035 in wire lands around 140–200 A at 16–20 V, and the wire feed speed sets a pace the operator can sustain all day.

The heat input per inch stays low because the wire shorts to the work thousands of times a second instead of holding a continuous arc, so 16-gauge panels stay flat instead of waving. On a run of brackets or gussets, a comfortable MIG operator will outrun a TIG operator by a wide margin while still producing a bead that needs little cleanup. ER70S-6 wire is the pragmatic choice when the stock has light mill scale, since its extra deoxidizers tolerate the contamination that would put tungsten and porosity into a TIG weld.

MIG’s weakness shows up in the root of a joint and overhead. Short-circuit transfer can wet into the sidewalls instead of digging in, so critical roots and heavy plate often get handed to TIG or stick instead.

Where Stick Still Wins

Take the machine to the beam and you’ll understand why SMAW refuses to die. There is no gas cylinder to roll, no liner to clog and no wire feed to tune, and the process keeps working in wind, light rain and positions where a gas lens cannot be held steady. A 1/8 in E6010 burns through mill scale and rust for a root pass on pipe, while E7018 gives a clean, tough bead for structural work.

Hydrogen is the one place stick demands respect. The American Welding Society structural code treats electrode storage as a control on hydrogen cracking, so low-hydrogen rods stay in a rod oven or holding oven until the moment they are struck. Use a generic E6011 on a restrained joint and you have quietly raised the cracking risk that TIG and MIG never introduced.

Where TIG Is the Only Real Answer

Open-root pipe, thin-wall tubing and visible beads belong to GTAW. Nothing else gives the same pedal-controlled regulation of heat, filler timing and puddle shape, which is why precision repairs on machined parts go to TIG even when the shop owns three MIG machines. Pure argon keeps the tungsten happy and the carbon steel weld virtually spatter-free.

The honest comparison ends where every shop floor argument ends: match the process to the joint, not to the machine that happens to be free. Root pass and thin wall go to TIG, sheet production goes to MIG, and structural field work goes to stick. On carbon steel, all three are correct answers to different questions.

Safety, Fumes and Ventilation When TIG Welding Carbon Steel

TIG welding has a reputation as the cleanest arc process, and on carbon steel it largely earns it. There is no flux coating burning, no spatter flying across the bay, and far less smoke than stick welding. That clean image is also a little dangerous, because it lulls people into skipping the ventilation, the respirator, and the pre-job hazard check. TIG produces fewer fumes, not zero fumes, and the fumes it does produce contain some genuinely unfriendly ingredients.

What Is Actually in the Fume

Welding fume is not one thing. It is a fine particulate cloud built from vaporized base metal, filler metal, coatings, and atmospheric gases. On carbon steel, the dominant components are iron oxide, manganese, and silica from the steel itself, along with trace amounts of other alloying elements depending on the material. OSHA’s welding safety guidance is blunt about the stakes: prolonged exposure to welding fume may cause lung damage and cancers of the lung, larynx, and urinary tract.

Manganese deserves special attention because it is the hazard TIG welders underestimate most. It is present in virtually every carbon steel and in common fillers, and long-term overexposure has been linked to a cluster of neurological symptoms resembling Parkinson’s disease, including tremor, slowed movement, and coordination problems. NIOSH maintains the current exposure limits for welding-related chemicals, and anyone running a TIG rig professionally should know where manganese sits on that list. The research base keeps growing; a recent review of welding fume exposure and neurological and respiratory outcomes is a good starting point if you want the science rather than the shop folklore.

Carbon monoxide is the other invisible one. CO forms when the arc breaks down shielding and atmospheric gases, and it gives you no warning beyond a headache and a slight drowsiness that impairs judgment before you notice anything is wrong. OSHA’s permissible exposure limit for carbon monoxide is 50 ppm as an 8-hour time-weighted average. In a poorly ventilated shop with several arcs running, hitting that number is easier than most people think.

The Scaffolding: OSHA, ANSI and You

The governing rule for general welding safety is OSHA 1910.252, the respiratory protection provision of the welding, cutting, and brazing standard. It requires that welding be done in the minimum enclosed space necessary, that ventilation be sufficient to keep fume and gas concentrations below the applicable limits, and it explicitly cites ANSI Z49.1, Safety in Welding, Cutting and Allied Processes, as the reference for the technical details. If your shop has a written ventilation procedure, Z49.1 is almost certainly where it came from. Chemical hazards also fall under OSHA’s HazCom standard, which means the Safety Data Sheets for your filler, degreaser, and coatings should be reachable in the work area, not locked in an office.

PPE and the Ventilation Hierarchy

Start with the eyes and skin. The TIG arc emits intense ultraviolet radiation that causes arc burn in minutes and, with repeated exposure, permanent corneal damage. A shade 10 to 13 filter lens in a well-sealed helmet is the normal range for carbon steel TIG work, with lighter end of the range suited to lower amperage. Cover every inch of skin with leather or flame-resistant clothing: a collar, sleeves, gauntlets, and no cuffs that catch sparks. Even a spatter-free process throws enough UV to give you a sunburn across any exposed forearm.

For respiratory protection, use the hierarchy that applies to every industrial hazard. First, engineering controls: position yourself upwind of the fume plume, use local exhaust ventilation or a fume extractor with the capture hood as close to the arc as your torch movement allows, and increase general dilution ventilation. Second, administrative controls: rotate

Practical Practice Plan, Buying Checklist and Conclusion

Knowing the theory of carbon steel TIG welding and laying a clean bead are two different things. The good news is that GTAW is one of the most forgiving processes to learn on, because nothing happens unless you say so: the arc sits where you put it, the puddle waits for you, and the foot pedal lets you dial heat up and down mid-joint. Here is a plan that takes a beginner from first arc to a weld worth inspecting.

A Four-Week Practice Progression

Practice on 3 in to 6 in offcuts of 16-gauge and 1/8 in mild steel plate. Buy a marker and a cheap grinder; the schedule matters less than the reps.

• Week 1 — Beads on plate. Run straight stringer beads at roughly 60–80 A on 1/8 in plate with 1/16 in ER70S-2. No filler at all for the first hour: just fuse the base metal and practice pedal control. The goal is a puddle that stays 2–3 times the rod diameter wide and never wanders.
• Week 2 — Butt joints. Tack both ends, check for wind-up in the plate, then run a single pass with filler at about 65–80 A. Dip the rod to the leading edge of the puddle, add a dab, and move. When a weld comes out too flat, add heat or slow down; when it piles up and looks like a caterpillar, add amps or travel faster. Keep the tungsten trimmed to a sharp point and stop to regrind whenever it touches the plate.
• Week 3 — Fillets and laps. Clamp two coupons into a T and weld at 45°, then repeat as a lap joint. Fillets expose bad torch angle quickly, because a steeper work angle pulls the puddle into the vertical plate and leaves the root unfused. Add a gas lens here if you bought one; it extends coverage and lets you slow down without boiling the puddle.
• Week 4 — Open root. Move up to 2 in Schedule 40 pipe or a 1/4 in open-root plate joint with a 1/16 in gap. Purge the back side with argon, run the root cold enough that it does not fall through, then cap it. This is the first week where a complete TIG welding settings chart matters more than feel, because small changes in amperage decide whether the root sinks or stands up.

Buying Checklist

Spend money in this order, and skip anything that does not appear on the list until your basics are covered:

• Machine. DC-only TIG is enough for carbon steel and usually cheaper than a full AC/DC unit. Look for high-frequency start, a foot pedal in the box, and roughly 150–200 A output with a real duty-cycle rating. A beginner-focused TIG welder guide explains how to read those numbers.
• Torch. A 150 A air-cooled 17-style torch handles anything you will practice on for months. Go water-cooled only if you weld continuously on thicker stock.
• Consumables. Buy 1/16 in and 3/32 in 2% lanthanated tungsten, matching collets, a standard cup, and one #8 gas lens. Lanthanated handles steel as well as the classic red thoriated type and is non-radioactive to grind.
• Filler. ER70S-2 in .045 in and 1/16 in covers most light fabrication; ER70S-6 runs wetter and faster if your fit-up is sloppy. Full details live in our filler metal selection guide.
• Gas. 100% welding-grade argon on a CGA 580 regulator, 15–20 CFH, more with a gas lens.
• Safety and shop gear. Auto-darkening helmet in the #10–12 range, leathers, gloves, ear and eye protection, and real ventilation. OSHA’s welding, cutting and brazing page sets the baseline for fume control. Add a grinder, flap discs, a stainless wire brush dedicated to steel, magnets, and clamps.

Conclusion

Carbon steel TIG rewards patience more than money. A modest DC machine, a box of ER70S-2, and four disciplined weeks of plate practice will outperform an expensive setup that collects dust. Work through beads, butt joints, fillets, and open roots; photograph each result and compare it to the previous week. When your beads stay flat, wetted at both toes, and consistent from start to stop, you are no longer learning the process — you are fabricating with it, and every other material becomes a variation on what you already know.

Frequently Asked Questions

What is the best filler rod for TIG welding carbon steel?

AWS ER70S-2 is the default choice for GTAW on carbon steel. Its triple-deoxidizer package (aluminum, titanium, zirconium) tolerates minor surface contamination and produces sound, radiographic-quality welds, especially on root passes and pipe.

What gas should I use for carbon steel TIG welding?

100% argon is the standard shielding gas for GTAW on carbon steel, typically at 15-25 CFH depending on cup size and draft. Helium or argon-helium mixes can increase penetration on heavy sections but are not needed for general shop work.

What polarity does carbon steel TIG welding use?

Carbon steel TIG runs on DC electrode negative (DCEN, or straight polarity), which concentrates heat in the workpiece and gives deeper penetration. AC is used for aluminum and magnesium, not for carbon steel.

How do I know what amperage to use?

A common starting point is roughly one amp per 0.001 inch of material thickness, then adjust by watching the puddle. Published data sheets give ranges such as 50-150 A for 1/16 in rod and 85-250 A for 3/32 in rod; reduce amperage about 15% for out-of-position welding.

Do I need to preheat mild steel before TIG welding?

For thin, low-carbon steel like A36 or 1018 in a low-restraint joint, preheat is usually unnecessary. Thicker sections, higher-carbon or alloy steels, and highly restrained joints need preheat based on carbon equivalent and the applicable code, such as AWS D1.1.

Why am I getting porosity in carbon steel TIG welds?

Porosity almost always comes from contamination or lost shielding — oil, grease, paint, moisture, heavy mill scale, or a draft blowing argon away from the puddle. Clean the joint, check your gas line and flow, and shield the arc from air movement.

Is TIG welding carbon steel stronger than MIG welding?

With the same filler classification, both processes can meet the same minimum specification — AWS A5.18 ER70S-2 requires 70 ksi tensile strength either way. TIG typically produces cleaner, tougher welds with fewer defects, while MIG is faster and cheaper in production.

Is TIG welding carbon steel safe?

TIG produces less fume than many processes, but it still generates metal fume and gases including carbon monoxide, and argon can displace oxygen in confined spaces. Use ventilation, position yourself upwind, wear proper eye protection and follow OSHA ventilation requirements.