Root Pass Welding: The First Bead That Decides the Weld

Two pipe welders finish the same spool on a Friday afternoon. Both caps look identical — clean ripple pattern, even toes, no undercut. Monday morning, one passes radiography and the other is cut out with a grinder. The difference was laid down in the first ninety seconds of the joint, in a bead nobody sees again: the root pass.

What Is a Root Pass in Welding? The Definition That Actually Matters

Walk through a pipe shop near shift change and you will see a ritual that says more about a welder’s skill than any finished cap: an inspector on a creeper, flashlight in one hand and mirror in the other, reading the inside of a joint that is an hour old. No fill passes yet. Nothing looks finished. And yet the decision on whether that spool ships has, in practical terms, already been made.

That moment exists because of what the root pass is. In plain terms, the root pass is the first weld bead deposited in a joint, made to produce the root bead and achieve fusion at the base of the groove. It is the pass that actually joins the two pieces of metal at their deepest point. Everything after it is filling and finish work on a connection that either exists or does not.

The AWS Definition, Word for Word

The American Welding Society keeps it deliberately short. AWS defines a root pass as “a weld made to produce a root bead” and calls it the first pass in any multi-pass weld. That wording separates two ideas people constantly blur: the pass is the operation, the bead is the metal it leaves behind. It also explains why the term only makes sense in a multi-pass context. A single-pass fillet weld has no root, no fill, and no cap; one bead does every job at once.

For pipe work without backing, the governing reference is ANSI/AWS D10.11, Recommended Practices for Root Pass Welding of Pipe Without Backing. It covers open-root and consumable-insert techniques, joint design, purging, and filler metal selection, which tells you something: the root is a subject big enough to need its own standard.

The Three Jobs the Root Pass Does at Once

A root pass that passes inspection has done three things at once. Miss one and the joint is defective, no matter how good the cap looks.

• Fusion at the base of the groove. The arc must melt both bevel faces and the land, tying the bead into parent metal on both sides. Weld metal that sits in the gap without bonding to the walls is lack of fusion, a rejectable defect, as TWI’s guidance on incomplete root fusion and penetration makes clear.
• Complete joint penetration. CJP means weld metal extends through the full thickness of the joint. Only the root establishes it; fill passes add volume to a connection the root already made, or failed to make.
• A sound root surface. The inside of the bead must be smooth, correctly profiled, and free of notches, internal undercut, or sharp transitions. That is a shape requirement, not a cosmetic one.

Root Pass vs Root Bead vs Hot Pass

Shop talk is loose, and loose talk causes real confusion when a welder reads a WPS or argues with an inspector.

Term	What it actually means
Root pass	The operation: the first pass, made to produce the root bead.
Root bead	The deposit at the base of the joint. Also called the root run in pipe work done as one continuous sequence.
Hot pass	The pass right after the root. Hotter and faster, it burns out slag and tempers the root. It cannot repair a bad root.
Fill passes	Intermediate passes that build thickness toward the surface.
Cap	The final pass, setting contour, toe wetting, and surface appearance.

The distinction that matters: the root is the only one of these that touches both the inside of the pipe and the full wall thickness at once. A spool with a root defect cannot be fixed by stacking more metal on top of it.

Where the Root Sits in a Single-V Joint

In a single-V butt joint, the bevel faces form a V that narrows to the root face, or land, at the bottom. The root pass goes in that narrowest region, bridging the root gap. On pipe, that means working 360 degrees while position shifts from flat at the top through vertical to overhead at the bottom, which is why pipe roots ask more of the hand than plate roots. Our all-position welding rod guide covers how electrode choice changes across those positions.

Inspection comes before burial. The root gets checked from the inside with a light and mirror, or by reading the outside profile where geometry allows. Once the hot pass goes on, that view is gone, and the remaining options are radiography, ultrasonic testing, or cutting the joint out.

Open Root vs Closed Root

Two families of root exist, and confusing them is a common source of bad welds. Open root means no backing of any kind; the bead is formed entirely by the welder’s arc and, in some processes, by internal gas pressure holding the molten metal in place. That is the demanding version, the one that separates pipe welders from everyone else. Closed root puts something behind the gap: a backing ring, a consumable insert, or permanent backing left in place. Closed roots are more forgiving, but they bring inspection concerns of their own, since the backing can create a crevice or an unfused interface.

It Is Not Only About Penetration

Plenty of welders treat the root as a penetration test and nothing more. Break through the land, and you are done. That view misses two things.

On stainless steel, the root surface is the corrosion-facing surface. It contacts the process fluid, and a rough, oxidized, or undercut root becomes the spot where chloride attack and crevice corrosion start. On cyclic service, in piping subject to vibration or thermal cycling, the root profile is a fatigue-initiation site: a sharp notch at the root toe concentrates stress in a way a smooth, slightly convex bead does not. Either way, geometry nobody will ever see again decides how long the part lasts.

Three Problems Wearing One Name

Think of the root pass as three problems stacked on each other, because each is solved differently. It is a fit-up problem: bevel angle, land thickness, root gap, and alignment decide whether the arc can break through at all, before a single amp is dialed in. It is a heat-input problem: amperage, voltage, and travel speed control whether the keyhole stays stable or drops out as an icicle. And it is a hand-skill problem: no procedure compensates for a welder who cannot hold consistent travel speed through a position change.

The rest of this guide works through those layers in order, from joint preparation to heat control, purging, defect recognition, and the drills that build the skill. Keep the definition above in view, because every technique that follows serves one goal: a first bead that is fused, fully penetrating, and clean on the side nobody sees.

Root Pass vs. Fill vs. Cap: Anatomy of a Multi-Pass Weld

A groove weld is not one weld. It is a stack of welds with different jobs, different heat ranges, and very different consequences when something goes wrong. Understanding that stack is what separates a welder who follows a procedure from one who understands why the procedure exists.

The Pass Sequence, Layer by Layer

A typical open-root pipe procedure stacks four stages: root, hot pass, fill, and cap, with each pass cleaned before the next. That sequence is laid out in resources such as AWS D10.11, and the logic applies to structural groove welds too.

The root pass establishes fusion at the base of the groove and creates the root bead. It is the only pass that reaches the full wall thickness in one shot, and on pipe it is the only pass exposed to both the inside surface and the completed joint. Everything above it is built on what it leaves behind.

The hot pass follows immediately, usually run hotter and faster than the root. Its job is to burn out slag and trapped material, put heat back into the joint, and help temper the root bead and refine its structure. It is not a fill pass. It is a conditioning pass, and the way it is run directly affects the root underneath.

Fill passes do the volume work. They build the joint up to the surface, and they are where heat input and interpass temperature are controlled most deliberately. The cap finishes the weld: correct contour, toes wetted smoothly into the base metal, and enough reinforcement to satisfy the drawing without creating a stress concentration.

Pass Thickness, Bead Shape, and Interpass Temperature

Most procedures cap individual pass thickness, often around 3 mm (1/8 in) for vertical work, with allowances in the flat position. The root and hot pass are usually thinner still. Oversized passes deposit more metal per unit of heat, but they also create a wide, shallow shape that is harder to fuse at the sidewalls.

That is why stringer beads are often preferred over wide weaves for toughness-critical work. A controlled stringer keeps heat input predictable, limits grain growth, and produces a cleaner fusion line. Weaves have their place, but a lazy, wide oscillation in vertical-up welding can leave slag traps and lower notch toughness.

Interpass temperature matters as much as preheat. Carbon steel generally tolerates a higher interpass range, but excessive heat still coarsens the grain structure and increases distortion. Stainless steel needs tighter control to limit sensitization and warping, while duplex stainless is especially sensitive because too much heat can upset the ferrite-austenite balance. The WPS governs the numbers; the point is to measure them rather than guess.

What Each Pass Is Supposed to Do

Pass	Primary purpose	Typical defect if done badly	Inspection method
Root	Fuse bevel faces and land; achieve full joint penetration	Incomplete penetration, suckback, burn-through	Visual inside and out, then RT or UT
Hot pass	Remove slag, temper the root, refine structure	Root pulled thin, slag entrapment, poor fusion	Visual, PT where specified, final RT/UT
Fill	Build wall thickness with controlled heat input	Lack of fusion, porosity, excessive reinforcement	Visual between passes, final RT/UT
Cap	Final contour, toe wetting, corrosion resistance	Undercut, high crown, rough toes	Visual, PT or UT where required

Why a Bad Cap Is Annoying and a Bad Root Is Expensive

A bad cap is usually cosmetic and grindable. A high crown, a rough toe, or slight undercut can often be dressed out without cutting into the joint. A bad root is different. It is buried under every pass that follows, and it is often rejected under radiography or found later as a leak in service. Lack of penetration reduces the effective throat, which lowers both static strength and fatigue life, as ESAB’s defect guidance explains. On thin-wall pipe, the root may be 60 to 70 percent of the wall thickness, so a root defect represents a huge loss of load-bearing section. On heavy wall it is a small fraction, but it is still the most consequential pass in the joint.

Pipe Versus Structural Plate

In pipe, the root pass runs a full 360 degrees and the position changes continuously from flat through vertical to overhead. That is why all-position root work demands such consistent body movement and arc control. In structural plate, the root is often flat or horizontal, and backing is frequently permitted, which removes much of the keyhole challenge but introduces its own fusion issues at the backing interface.

The hot pass is not always mandatory. Some procedures allow it to be omitted, usually on specific thicknesses or processes, but that decision belongs to the WPS and the engineer, not to the welder’s preference. And on any joint, the root should be inspected and cleaned before it disappears under the next layer. Once it is buried, the only way to fix it is to remove metal.

Joint Preparation and Fit-Up: Where Most Root Failures Begin

When a root pass fails, the post-mortem usually starts with the welder’s hand. Travel speed was off, the arc was long, the keyhole got away. Sometimes that is true. More often, the arc was set up to fail before the first electrode was struck. A ragged bevel, an uneven land, or a gap that opens and closes around the pipe circumference turns a capable welder into someone fighting the joint instead of welding it.

Fit-up is not paperwork before the real work. It is the root pass’s foundation. The root has to bridge a precisely shaped gap, fuse two bevel faces and a land, and do it through the full wall thickness. If those dimensions are wrong, no amount of hand skill will reliably compensate.

The Geometry of a Single-V Bevel

A single-V butt joint has four dimensions that decide whether the root can work: bevel angle, root face, root gap, and alignment.

Feature	What it controls	Typical carbon steel pipe starting point
Bevel angle	Access for the arc and room for fusion at the sidewalls	About 37.5° per side, 75° included
Root face (land)	Supports the puddle and resists burn-through	About 1.6 mm (1/16 in.)
Root gap (opening)	Lets the arc penetrate through the land	Roughly 2–3 mm
Alignment	Keeps the gap and bore even around the joint	As close as practical, within the project limit

Those numbers are common starting points for carbon steel pipe, not universal law. The ASME B31.3 framework and your own welding procedure specification govern the actual values on the job. A procedure written for a different wall thickness, position, or process may call for something else entirely.

When the Numbers Drift

Small geometry errors produce very specific root failures:

• Land too thick: the arc cannot punch through, so penetration stops at the land. The result is incomplete penetration with a clean-looking cap.
• Land too thin: the puddle drops through before it can bridge. Burn-through and suckback follow.
• Gap too tight: the arc has nowhere to go. Incomplete penetration is almost guaranteed, especially as the pipe heats and closes.
• Gap too wide: the puddle sags, root reinforcement becomes excessive, and burn-through or suckback becomes likely.
• Excessive mismatch: the gap varies around the circumference. One side runs tight while the other runs wide, so the welder chases the joint instead of holding a stable keyhole.

These are not theoretical concerns. TWI lists an excessively thick root face, too small a root gap, misplaced welds, and low heat input among the causes of incomplete root fusion. Notice how many of those are decided at the fit-up bench.

Cut, Bevel, and Clean for the Root

Machining gives the most consistent bevel. Grinding can work when the operator keeps the angle and land uniform. Thermal cutting is acceptable only if the oxidized edge is removed afterward. A bevel that looks shiny at a glance can still carry a dark, glassy layer that prevents fusion.

Grind bevel faces to bright metal, and clean at least 25 mm back from the joint. That removes mill scale, rust, and thermal oxidation that can cause lack of fusion at the root. Then wipe the faces with a suitable solvent to remove cutting oil, grease, and marker residue. Let the solvent flash off before welding, and do not handle the cleaned faces with oily gloves.

Tacks, Braces, and Purging Dams

Tacks are part of the root, whether welders treat them that way or not. Place them evenly around the joint, avoid the hardest access position where possible, and feather both ends to a gentle taper. A tack that sits as a cold lump on the bevel face will not fuse when the root passes over it. Its ends also cool quickly under restraint, which makes them a common starting point for cracks.

Bridge tacks, strongbacks, and pipe clamps hold the gap against welding distortion. Use them deliberately, not as an excuse to force a bad fit. On purged work, remember that dams block internal access. Install them before the joint is closed, leave a vent path, and confirm they can be removed cleanly after welding.

A Fit-Up Checklist That Earns Its Keep

1. Read the WPS and confirm the bevel angle, land, and gap it requires.
2. Check the bevel angle and land at several points, not just one.
3. Measure the gap at four points around a pipe; look for a gap that closes or opens.
4. Check alignment and high-low mismatch against the project limit.
5. Verify bevel faces are bright for at least 25 mm back and free of oil.
6. Feather tacks and confirm they are fused, not sitting cold.
7. Install purge dams and confirm ventilation or venting.
8. Brace the joint against distortion before the root is struck.

Inspection Starts Before the Arc

Under ASME B31.3, in-process examination can require inspection of joint preparation, fit-up, clearance, alignment, and the condition of the root pass after cleaning. That matters because once the hot pass goes on, the root is buried. An inspector can no longer see whether the land was uniform or the gap was true. The time to catch a fit-up problem is before it becomes a radiograph rejection.

Electrode selection gets plenty

Root Pass Techniques by Process: SMAW, GTAW, GMAW, and FCAW

Every process can put metal in a groove. Only some can put it at the bottom of the groove and still come back clean on a radiograph. The difference is not shop tradition, it is how each arc behaves when there is nothing underneath but open air. Here is how the four common processes handle an open root, and what each one demands from your hands.

Much of the shared practice comes from ANSI/AWS D10.11, which covers root pass practice for GTAW, GMAW, and SMAW on open-root and consumable-insert pipe joints, including gas purging.

SMAW (Stick): The Pipe Welder’s Default

For open-root carbon steel pipe, cellulosic electrodes such as E6010 remain the standard choice. The cellulose coating produces a forceful, digging arc and a fast-freezing slag, so you can push through the land and still have the bead set before gravity pulls it out. E7018 is used where the procedure allows it, but low-hydrogen electrodes dig less and punish loose fit-up.

The classic motion is whip-and-pause: a short forward flick to open the keyhole, then a decisive return, pausing just long enough to tie both bevel faces before moving again. Hold the electrode close to perpendicular with the work angle bisecting the joint, keep the arc tight, and travel at a rate the puddle can actually follow.

Why do welders call 6010 unforgiving? Because it reacts to gap. A joint that varies by a millimetre around the circumference swings from “will not break through” to “keyhole drops out” with no change in your hand. That is a fit-up conversation more than a technique one. For rod selection across positions, see the all-position welding rod guide, and for the case that stick still makes, read why SMAW still wins.

GTAW (TIG): Cleanest Root, Least Margin for Sloppiness

GTAW gives the cleanest root available: no slag, no spatter, precise heat control, and a bead that photographs well on stainless, thin wall, and critical service where the inside surface matters as much as the outside.

Two decisions come first: autogenous (fused without filler) or with filler wire. Autogenous roots suit thin wall and close-fitting joints; filler gives you control over reinforcement and is standard on heavier sections. Keep the torch angle shallow, match amperage to travel so the keyhole stays a small, stable oval, and dip filler at the leading edge of the puddle rather than dragging it through the arc.

In 6G, walk-the-cup is what separates consistent roots from patchy ones

Heat Input, Amperage, and Keyhole Control: The Skill Nobody Can Fake

Walk up behind an experienced pipe welder on a root and you will notice something before you see the bead: the sound is steady. No sputtering, no long hiss, no pause. That steadiness is not style. It is the audible signature of a keyhole that is being held at a constant size, and it is the single best indicator that the arc is doing what the procedure demands.

The keyhole is a small opening at the leading edge of the puddle, roughly the width of the root gap, that proves the arc has broken through the land and is burning into the root opening. If you cannot see or hear a keyhole, you are not welding an open root — you are laying metal on top of a bevel and hoping. Everything else in this section is really about keeping that opening alive without letting it run away from you.

Reading the Keyhole Instead of Guessing

Keyhole size is not a mystery; it responds to four inputs, and once you learn the direction each one pushes, diagnosis gets much faster:

• Amperage: more current widens and lengthens the keyhole. Too much and the walls collapse into a drop-out.
• Travel speed: slower travel lets heat soak and the keyhole grow. Speed up and it closes.
• Root gap: a wider gap needs less current to keep the same keyhole, and it punishes any extra heat instantly.
• Land thickness: a thick land needs more punch to break through; a thin land breaks through before you are ready.

A healthy keyhole looks round, stable, and roughly consistent in size as you travel. A keyhole that flares and drops out means the puddle lost its support — usually too much current, too slow a travel speed, or a gap that opened up mid-joint. A keyhole that closes and dims means you have lost penetration, and the fix is the opposite: more current, slower travel, or a check that the gap has not tightened. The mistake welders make is reacting to the symptom without checking the fit-up, which is the actual cause more often than the machine setting. Getting this right starts long before the arc, in the all-position welding rod selection and joint prep stage.

Heat Input: The Number Behind the Symptom

Heat input is proportional to voltage multiplied by amperage, divided by travel speed. That relationship explains almost every root failure you will see. Slower travel raises heat input even when the machine has not changed, which is why a welder who hesitates through a hard-to-reach section can blow a hole in a joint that was behaving perfectly ten minutes earlier. Excessive heat input on the root produces burn-through, suckback, a wide heat-affected zone, and distortion that pulls the pipe out of alignment. Too little does the quieter damage: TWI identifies low power or low arc energy input as a direct cause of incomplete root fusion and incomplete root penetration — defects that will not show up until radiography or, worse, service.

Arc Force: Cellulosic Versus Low-Hydrogen

Electrode type changes the feel of the root completely. Cellulosic electrodes such as E6010 run a fierce, digging arc with strong arc force — the gas blast from the coating physically pushes into the gap and helps hold the keyhole open. That digging action is why 6010 punches through a land so well, and also why it is unforgiving: a gap that grows by a millimetre turns into a blow-through fast. Low-hydrogen electrodes like E7018 run a softer, less aggressive arc with a smoother puddle. They are kinder on fit-up variation but weaker at breaking through a heavy land, so they suit backed roots or tighter, more controlled joints. Match the electrode’s arc character to the joint, not to what is in the rod oven.

Body Position, Angle, and the 6G Problem

In 6G the pipe is fixed at 45 degrees, so the same joint asks you to weld flat, vertical, and overhead within a single pass. Welders habitually speed up at the transitions — coming out of the overhead into vertical-up, or rushing the top of the joint — and that speed starves the root, thinning penetration exactly where inspection is most likely to look. Keep travel consistent through the transitions and let your body rotate with the arc rather than reaching across it.

Rod angle matters as much as current. The work angle should bisect the joint so both bevel faces get equal heat; the travel angle should push slightly into the keyhole on a drag technique for stick, while TIG generally runs a slight push. Keep the arc length short and constant — long arcs lose focus, widen the puddle, and invite lack of fusion. And when you stop, stop on the bevel, not in the gap. Ending in the open gap leaves a thin, unsupported crater that cracks as it cools.

Restarts and Crater Cracks

Crater cracks at root-pass stop points are one of the most common radiographic rejects, and they are almost entirely a technique problem. Feather the crater back to sound metal, then restart slightly ahead of the stop and back-feed into the crater so the new puddle remelts the old one. Welding forward from a cold crater edge without back-feeding leaves a notch and a shrinkage crack waiting to be found. On TIG, a remote amperage control or foot pedal lets you taper out instead of snapping the arc, which does more for crater quality than any downstream grinding.

Why Nobody Else’s Numbers Work in Your Shop

A welder who tells you “run it at 92” is giving you half a sentence. Amperage readings shift with cable length and condition, ground quality, work lead placement, and whether the machine has been calibrated recently — a point regulars on the AWS forum raise often, and one that AWS guidance on root welding treats as basic practice. A long, undersized cable run can leave you 10 to 15 amps below the dial reading at the arc. Treat every borrowed number as a starting point, then tune by the keyhole, the sound, and the root you can actually see.

Purging, Shielding, and Root Oxidation on Stainless and Reactive Metals

Everything so far has focused on the side of the joint you can see. On stainless steel, duplex, titanium, and nickel alloys, the side you cannot see decides whether the weld survives five years of service. The root pass is the only pass with a face inside the pipe, and while the arc is still moving, that face sits at incandescent temperature in whatever atmosphere is trapped in the spool.

Why the Backside Needs Its Own Atmosphere

Air is roughly 21 percent oxygen, and hot metal is greedy for it. On austenitic stainless, oxygen and nitrogen combine with the hot root to form a visible oxide layer and, beneath it, a chromium-depleted zone where pitting corrosion starts. On duplex, the wrong atmosphere can upset the ferrite-austenite balance. Titanium is the extreme case: above a few hundred degrees Celsius it absorbs oxygen, nitrogen, and hydrogen readily, and a straw-colored root is a rejection, not a cosmetic issue.

The fix is displacement, not dilution by hope. Argon, or an argon-hydrogen forming gas for some austenitic work, is fed into the sealed joint volume until oxygen drops below what the procedure allows. AWS D10.11 covers gas purging as part of open-root pipe welding practice, and it is the reference most procedures lean on for the how rather than the marketing.

Shielding vs. Pressure Control: Two Different Jobs

Welders tend to blur these together, but they are separate settings on separate devices. Shielding is chemistry: getting oxygen out of the joint so the root solidifies clean. Pressure control is mechanics: a slight positive pressure pushes back against the puddle and holds a flat-to-slightly-convex root profile instead of letting it sag or suck back.

The second one is where people get hurt. Over-pressurize and the puddle pushes outward, giving a convex root, an unstable keyhole, and sometimes a blow-through at six o’clock. Too little and the root concaves. The sweet spot is a gentle flow you can barely feel at the vent, not a jet.

Measuring the Purge Instead of Guessing

A purge monitor reads oxygen in parts per million at the vent, turning purging from an article of faith into a number. Targets are procedure-dependent, not universal. Common practice puts many stainless and duplex roots below roughly 100 ppm, tighter on critical corrosion service. Titanium and some nickel alloys go far lower, and those numbers come off the WPS, not a forum post. Titanium rewards the discipline in this aerospace TIG welding guide.

Material	What the purge protects	Typical spec style
304/316 stainless	Chromium at the root, corrosion resistance	Oxygen ppm limit plus heat tint acceptance
Duplex	Phase balance, corrosion resistance	Oxygen limit, hydrogen-free gas
Titanium	Ductility, contamination-free root	Very low ppm plus trailing shield
Nickel alloys	Clean root, crack resistance	Oxygen limit, often color comparator

Heat Tint and What the Code Actually Says

Heat tint is the rainbow that appears on the root as it oxidizes. Color roughly indicates oxide thickness, and thicker oxide means more chromium pulled from the metal below. Here is the part that surprises people: ASME B31.3 does not itself address weld root oxidation or heat tint. If a project wants it controlled, the owner or specification writes it in, and a reference weld or color chart becomes the comparator during inspection. Without that, you are arguing about taste.

When the Root Comes With Slag

Flux-cored and stick roots on stainless or nickel alloys leave a glassy slag on the inside face, and it must come off completely before the hot pass. A wire brush used on carbon steel will contaminate the next stainless joint, so keep dedicated stainless brushes and never share grinding wheels between carbon and stainless.

The Safety Side Nobody Enjoys Talking About

Argon is an asphyxiant: invisible, odorless, heavier than air, and it will quietly fill a confined space and displace the oxygen a welder needs. Purging a vessel or spool while someone is inside it is a documented killer. Atmospheric monitoring, forced ventilation, a competent-person entry plan, and a standby attendant are not bureaucracy. OSHA 1910.252 sets ventilation and health protection requirements for welding, and the confined space rules sit alongside it. NIOSH has published guidance on welding fumes and gas hazards worth reading once a year.

Purging a Big Spool Without Burning Through Gas

Nobody purges a forty-foot spool end to end. You dam it. Inflatable purge dams, foam plugs, and water-soluble plugs isolate a short section around the joint, sometimes only a few feet each side. Small volume means fast displacement and low gas cost.

1. Measure the dammed volume in cubic feet.
2. Multiply by the number of volume changes your procedure calls for, commonly around five or six for air displacement.
3. Divide by your flow rate in cubic feet per minute for a minimum purge time.
4. Verify with the oxygen monitor at the vent, not with the clock.

Then hold a low flow through the weld until the root drops below the oxidation temperature. Aluminum is the exception, since shielding only happens at the torch, as covered in this aluminum MIG welding guide. On stainless and reactive metals, the inside of the pipe is where the weld is really being made.

Root Pass Defects: What They Look Like, Why They Happen, and What Codes Accept

A bad root rarely announces itself. The cap can look perfect while incomplete penetration, a cold lap, or a crater crack sits underneath, waiting for pressure or fatigue. That is why experienced inspectors look at the root before the hot pass covers it, and why radiography and ultrasonic testing exist in the first place.

The defect library

Most root pass defects fall into a handful of recognizable families. The trick is connecting the shape you see to the condition that caused it.

Defect	Main cause	Appearance	Detection	Prevention
Incomplete penetration	Low current, thick land, tight gap, fast travel	Straight dark line on the root side of a radiograph; visible ledge inside pipe	RT, UT, internal visual	Correct gap and land, enough amperage, stable keyhole
Lack of fusion at root or sidewall	Low heat input, fast travel, arc aimed off the bevel, contamination	Tight planar flaw, sometimes a shadowy line; easy to miss visually	UT or phased array; RT if favorably oriented	Wash both bevel faces, correct work angle, adequate heat
Suckback or concave root	Excess current, wide gap, insufficient filler, weak purge support	Smooth concave groove on the inside surface; thin root	Internal visual, RT	Balance heat and gap, add filler, control purge pressure
Burn-through or excess reinforcement	Too much current, slow travel, thin land, wide gap	Hole, icicles, or heavy bulge inside the pipe	Visual, RT	Verify fit-up, reduce heat, maintain travel speed
Internal undercut	Long arc, excessive current, poor electrode angle	Groove at the root toe on the inside surface	UT, RT, internal visual	Short arc, correct angle, controlled heat
Root concavity	Insufficient fill, excess heat, poor shielding	Smooth dip below the desired root profile	Visual, RT	Feed enough filler, control heat input
Porosity	Moisture, oil, rust, wet electrodes, lost gas coverage	Rounded dark spots on RT; surface pores may be visible	Visual, RT, UT	Clean surfaces, dry consumables, maintain shielding
Slag inclusions	Poor cleaning between passes, slag trapped at the root	Irregular elongated indications on RT	RT, sometimes UT	Clean thoroughly, use correct angle and travel
Crater cracks	Stopping abruptly, unfilled crater, high restraint	Star-shaped or transverse cracks at stop points	Visual, PT, MT, RT	Taper the crater, restart ahead of the stop, feather tacks
Root oxidation	No purge, oxygen or moisture ingress, excessive heat	Blue, straw, gray, or black oxide on the inside surface	Visual, comparator or reference weld	Purge and monitor oxygen, control heat input

Incomplete penetration is more than a cosmetic flaw. ESAB’s welding defects guide notes that it reduces the effective throat, lowering static strength and fatigue life, and in process piping or pressure components it can be a critical rejectable condition. Lack of fusion is equally deceptive because it can sit entirely below the surface and escape visual examination, which is why UT is often needed to find it.

How the root gets inspected

The first and best inspection happens immediately after the root pass, before the hot pass buries it. On a single-V groove in plate, the root is visible from the back side. In pipe, you need to look inside with a mirror, flashlight, or borescope. Check for a consistent bead, complete fusion at both bevels, adequate reinforcement, and no cracks or deep concavity.

Once the joint is complete, surface inspection cannot tell you much about the root. Radiography reveals volumetric flaws such as porosity, slag, and heavy reinforcement, and it can show incomplete penetration as a straight dark line. Ultrasonic testing and phased array are better at finding tight planar flaws like lack of fusion and cracks, especially when their orientation makes them nearly invisible on a radiograph.

What codes actually reject

Code acceptance is not a single yes-or-no answer. Under ASME B31.3 Table 341.3.2, incomplete penetration and lack of fusion are generally rejectable in normal and severe service, with limited exceptions. TWI’s guidance on incomplete root penetration goes further for its quality levels: incomplete root penetration is not permitted for Quality Level B (stringent) or Level C (intermediate), apart from a partial-penetration exception at Level C. Pressure vessel rules are similarly strict.

Reinforcement limits are where confusion often starts. A commonly cited acceptance figure for internal root protrusion is about 2.4 mm (3/32 in.), provided the profile is smooth and does not create a stress raiser. Treat that as a starting reference, not a universal rule. The project specification, service category, and applicable code govern the final acceptance criteria.

The real cost of burying a bad root

A root defect found before the hot pass might cost ten minutes of grinding. Found after radiography, it means excavation, cut-out, re-fit, re-weld, and a second inspection. On a large spool, that can pull a crew off schedule for a full shift. Carbon arc gouging is often the fastest way to remove the old root, but the repair still has to be feathered, cleaned, and welded to procedure.

That economics is why the root deserves more attention than any other pass. The cap is what people see. The root is what determines whether the joint is actually joined.

Root Pass Troubleshooting: Symptom, Cause, Fix

Most root pass problems are not mysterious. They repeat, joint after joint, and they usually announce themselves early — in the sound of the arc, the shape of the keyhole, or the color of the bead on the inside. The trick is diagnosing them in the right order. Fit-up comes first, always. If the land is 3 mm when the procedure calls for 1.6 mm, no amount of amperage fiddling will rescue the joint.

Diagnose Before You Touch the Machine

Change one variable at a time. If you raise amperage and widen the gap and speed up all at once, you learn nothing and you ruin the joint. Check the gap with a gauge at four points around the circumference, look at the bevel faces for scale or oil, confirm the land, then and only then start adjusting heat and travel speed.

Symptom	Likely cause	Fix on this joint	Permanent fix
Arc won’t break through the land	Land too thick, gap too tight, amps low	Add a little current, shorten the arc, slow down	Gauge every land; recut bevels that drift from the WPS
Keyhole opens and drops out	Too much heat, wide gap, thin land	Back off amperage, speed up, feed filler faster	Clamp or tack to hold gap; verify land thickness
Keyhole closes up mid-bead	Insufficient heat or gap tightening as the pipe shrinks	Raise current slightly, reduce travel speed, widen gap	Strongbacks or bridge tacks to resist closure
Looks fine outside, fails RT	Incomplete penetration or internal undercut	Stop welding and inspect from inside	Inspect root with mirror and light before the hot pass
Bead sits on top, no penetration	Low amps, fast travel, long arc	Slow down, tighten arc length, reduce land	Match machine setting to cable length and ground quality
Concave, “sucked back” root	Overheating, thin root deposit, high purge pressure	Cut heat, add filler, reduce purge flow	Set purge by pressure gauge, not by feel
Icicles or grapes inside the pipe	Excess heat, slow travel, wide gap	Grind out and reweld; do not bury it	Reduce heat input; log the settings that worked
Porosity only at restarts	Crater not feathered, dirty restart point	Grind back to sound metal and restart ahead of the stop	Feather every crater as a habit, not a repair
Root cracks at tacks	Cold tacks, unfused tack ends	Remove completely — no blending	Feather tacks and fuse them into the root
Uneven root at 6 and 12 o’clock	Position transition, gravity, travel speed drift	Reposition and re-time the travel	Practice 6G transitions deliberately
Oxidation colors on stainless root	Purge insufficient or lost mid-weld	Stop, re-purge, measure oxygen	Purge monitor and recorded readings
Pipe pulled out of alignment	Uneven heat, too many passes one side	Re-align before the fill; cut out if cracked	Balance pass sequence; check alignment after root

Repairing a Bad Root Without Making It Worse

Where internal access exists, grind the defect out from the inside, feather the edges, and reweld to the procedure. Where it does not, carbon arc gouging from the outside followed by grinding to bright metal is the standard route — but gouging heat can drive a crack deeper, so verify with PT or RT before you weld. Welding over a bad root is a false economy; the pass you save today becomes a cut-out on Monday.

Cracks and lack of fusion must be removed entirely. Blending them leaves a stress concentrator under a cosmetic cap, and TWI notes that incomplete root penetration and root fusion are treated as rejectable in quality levels where full penetration is required. Radiography exposes these as density changes — incomplete penetration reads as a dark line on film because less metal means less absorption.

Field Habits That Save Joints

• Keep a mirror and a bright flashlight in your hood bag. The root is only inspectable before the hot pass goes on.
• Hold a light behind the joint before welding; uneven glow around the circumference is uneven gap.
• Mark the pipe at your start and stop point so restarts land in a known, feathered spot.
• Log the settings that produced a passing root — machine, cable length, amperage, travel feel.

Common Root Pass Mistakes, Practice Drills, and Certification Reality

Walk into any pipe-test booth and you will hear a version of the same story. A welder with years of structural experience lays a neat cap, then watches the inspector fail the root. The root pass is where most experienced welders fail their first pipe certification attempt on open-root joints. The problem is rarely nerve. It is a set of habits that structural work never punished.

The Mistakes That Repeat in Every Shop

Most bad roots come from a short list of shortcuts:

• Fitting by eye. Gap and land are measurements; use a gauge before every tack.
• Welding over mill scale, oil, or rust. Contamination turns into porosity and lack of fusion.
• Too much current because it feels faster. It opens the keyhole too wide and invites suckback.
• Running a long arc. You lose directional control and the bevel faces stop wetting in.
• Traveling too fast to escape the puddle. The root goes thin, convex, or unfused.
• Weaving the root. A controlled keyhole beats a wide oscillation every time.
• Skipping the hot pass or letting the joint cool too long. Both can leave the root vulnerable to shrinkage stress or cracking.
• Not feathering tacks and restarts. Cold laps and crater cracks start there.
• Welding an out-of-alignment joint. Mismatch makes one side dig and the other side starve.
• Ignoring interpass temperature. Too hot or too cold changes the weld you thought you made.
• Never inspecting the root from inside. You cannot fix what you refuse to look at.

The list is unglamorous, but that is the point. Root quality is built by boring discipline: checking, cleaning, and watching the keyhole instead of racing the clock.

Drills That Build Root Skill

Root skill comes from repetition with feedback, not from running longer beads.

1. Coupon plates first. Fix the gap and land, then run a root until the keyhole stays round and the back bead is even.
2. Move to pipe in 2G, then 5G, then 6G. Each position exposes a different travel and body-control weakness. An all-position welding rod guide helps you match electrode and technique.
3. Weld deliberately bad fit-ups. Open a gap too wide, tighten one too far, and practise correcting with angle, current, and travel rather than the grinder.
4. Try a no-grind drill. If every restart needs grinding, your stop and start technique is doing the hiding.
5. Time your travel. Mark a length, weld it, and compare seconds per inch with the root profile you get.
6. Take blind tests scored by radiography. Self-judged visual practice is far less informative than a film or digital image that shows what you cannot see.

What a Pipe Test Is Really Checking

A pipe test is not a cap contest. It evaluates root quality first, because the root is the pass that proves full penetration. Most tests combine a visual check with bend testing or radiography, and a coupon with a beautiful cover and incomplete penetration is still a failed coupon.

This is why capable structural welders struggle. Structural shop welding rarely requires open-root, full-penetration technique; much of it is fillet work, partial penetration, or grooves with backing. Pipe work removes those comforts and asks for a consistent keyhole in every position.

Qualification is also tied to a specific WPS covering process, position, thickness, and backing conditions. The WPS is backed by a procedure qualification record, and essential variables matter. Switching from open root to backing, or from 2G to 6G, can put you outside the qualified envelope. AWS’s recommended practices for root pass welding of pipe without backing are a useful reference, but the project WPS still governs.

Before you test, ask five things: which process, which position, what wall thickness, whether the root is open or backed, and how the test will be inspected. Then confirm the acceptance code. Under a code such as ASME B31.3, incomplete penetration and lack of fusion are generally rejectable in normal and severe service; TWI’s defect guidance explains why these are treated so seriously.

Career Leverage Without a Big Budget

Root skill is the fastest route from structural work into pipe and pressure work. It is also affordable to build. You do not need full spools: offcuts, short pipe sections, and coupon plates teach the same lesson, and stick welding’s advantages make it a cheap way to practise the basics. Consumable cost is small next to a failed test, a lost day, and a repeat trip.

Treat every practice root as a test coupon. Log the settings, the fit-up, and the result. The cap is what people see; the root is what keeps the line from leaking.

Conclusion: Build the Weld From the Inside Out

The cap is the part everyone photographs. It also proves the least. The root pass is the only pass in the sequence that demonstrates the joint is actually joined: the arc broke through the land, fused both bevel faces, and carried full penetration through the wall. Every fill pass and every cap ripple after it only covers that work up. Get the root wrong and the weld is wrong, no matter how good the surface looks on Friday afternoon.

What separates good welders from lucky ones is that the root is mostly decided before the arc starts. Fit-up geometry, clean bevel faces, and a procedure matched to the base metal shape root quality more than any wrist movement. Hand skill executes a decision already made at the fit-up bench.

The Five Controls That Actually Matter

Strip away the process debates and the brand loyalty, and root quality reduces to five controls:

1. Verified fit-up geometry. Bevel angle, root face, root gap, and alignment checked with a gauge instead of by eye, then re-checked at four points around the circumference.
2. Clean bevel faces. Mill scale, oxidation, oil, and moisture removed from the bevel and surrounding parent metal before the first arc.
3. Matched amperage and travel speed. A keyhole that opens, holds a consistent size, and closes behind the puddle — not one that drops out or never forms.
4. Correct shielding or purge on the root side. Argon coverage or oxygen displacement on stainless and reactive alloys, held at a pressure that supports the bead instead of blowing it through.
5. Inspection before the root is buried. The root is the one layer you can still see and still fix. Under ASME B31.3, in-process examination can include the condition of the root pass after cleaning and before the next pass — exactly the window where a grinder is cheap and a cut-out is not.

Your Next Shift: A Short Action List

• Gauge the gap at four points before tacking, and again after.
• Grind bevel faces bright and wipe them with solvent — both sides of the joint.
• Run a test bead on scrap with the exact machine, cable, and ground you will use.
• Watch the keyhole, not the arc, and listen for a steady sound.
• Inspect the root from inside with a mirror and flashlight before the hot pass.
• Log what worked: amps, volts, travel feel, gap, purge flow.

Where the Trade Is Heading

Mechanized and orbital welding keep taking over large-diameter, high-volume work, and for good reason: they deliver root consistency that human hands struggle to match across a twelve-hour shift. But orbital heads still depend on verified fit-up and purge quality — the machine cannot compensate for a gap that varies by three millimeters. Digital radiography and phased array raise the bar too, since defects that once slipped through film now show in sharp detail, and the documentation behind each joint has to match. Position skill still matters; if you are building consistency in flat, vertical, and overhead, an all position welding rod guide is a useful reference for matching electrode to position before you test.

What has not changed is the baseline. Manual root skill remains the foundation of field work, repair, and tie-ins, because no orbital rig fits a cramped rack or a half-cut spool at two in the morning.

So treat every root as a test coupon. Fit it like an inspector is watching, because one probably is. Keep the log. And remember: the cap is what people see, but the root is what keeps the line from leaking.

Frequently Asked Questions

What is a root pass in welding?

The root pass is the first weld bead deposited in a joint, made to produce the root bead and fuse the base of the groove through the full thickness. It establishes complete joint penetration, and every fill and cap pass is laid on top of it.

What is the difference between a root pass and a hot pass?

The root pass creates the initial penetrated bead at the bottom of the joint. The hot pass follows immediately, usually hotter and faster, to burn out slag, temper the root, and prepare a clean surface for fill passes.

Why does my root pass keep lacking penetration?

The usual causes are a root face that is too thick, a root gap that is too tight, too little amperage, or travel that is too fast to let the arc break through the land. Check fit-up with gauges before changing machine settings.

Can you weld over a bad root pass?

You can, but you should not. Covering a defective root hides incomplete penetration, lack of fusion, or slag that inspection will find later, and repairs after fill and cap cost far more than fixing the root while it is accessible.

Do you need backing or purging for a root pass?

Not always. Carbon steel open roots are commonly welded without backing. Stainless, duplex, titanium, and nickel alloys usually need an inert gas purge on the root side to prevent oxidation, and some procedures specify backing rings, tape, or consumable inserts.

Why is the root pass inspected before the next pass?

Once the hot pass and fill passes are applied, the root is buried and can only be evaluated indirectly by radiography or ultrasonic testing. Inspecting early catches problems while they are still cheap to grind out and re-weld.

Is downhill or uphill better for a root pass?

Downhill is faster and popular on large-diameter pipe with cellulosic electrodes, but it offers less fusion control. Uphill gives better fusion and is common on thicker wall and critical service. The qualified procedure decides which is permitted.

How long should a root pass take to learn?

Most welders need weeks of focused practice on coupons and pipe sections to produce consistent open roots, and much longer to be reliable in the 6G position. Progress depends on feedback — ideally radiography or an experienced inspector, not self-judgement.