Weaving Welding Technique: Patterns, Uses, and Best Practices

A welder on a structural job once laid a perfect-looking 6-inch-wide cap on a vertical groove joint, only to have the inspector reject it for undercut on both toes and a cold-look center that never fused. The problem wasn’t his amperage or his filler—it was an uncontrolled weave: too wide, too fast across the middle, no pause at the edges. That gap between a weave that fills a joint and a weave that ruins it is exactly what this guide is built to close.

What Is the Weaving Welding Technique?

Picture this: a welder on a structural job lays down a cap that looks gorgeous in the welder’s mirror, six inches wide, uniform ripples, a textbook-looking crown. The inspector rejects it anyway. Undercut chewed into both toes, and a cold, glassy center that never tied into the root. Nothing was wrong with the amperage or the filler metal. The weave was the problem: too wide, too fast across the middle, no pause at the edges. That is the tightrope every welder walks once they start oscillating an electrode, and it is exactly what this guide is built to close.

So let’s answer the principal question first, plainly.

A Plain Definition of the Weaving Welding Technique

That distinction matters because AWS Welding Procedure Specification (WPS) forms list “Stringer or Weave Bead” as a formal technique variable, right alongside amps, volts, and travel speed. It is not a stylistic preference. It is a documented, procedure-controlled decision, and the qualified WPS governs it.

The single most useful sentence to memorize: weaving is a controlled manipulation, not a wiggle. Its value is coverage and fusion control, not looks.

Three Physics Reasons Weaving Exists

Weaving isn’t a habit welders picked up because it looks impressive. It solves three real physical problems:

• Spreading filler over a wider cross-section. A single arc pass deposits a bead only so wide. When a groove or a fillet leg demands more metal than one pass can carry, oscillation spreads that deposit across the full width instead of stacking beads edge to edge.
• Tying in both sidewalls. A groove joint has two edges that must fuse. Sweeping the arc to each side momentarily directs heat at each sidewall, wetting the joint faces so the deposit fuses at both toes instead of bridging across them cold.
• Modulating heat distribution and puddle size. The arc is a moving heat source. By changing dwell time and travel speed across the width, the welder controls where the heat goes and how large the puddle gets, which in turn controls profile, dilution, and grain structure.

Those three jobs are why weave width, dwell time at each toe, and travel speed are the variables that most directly determine bead profile and fusion quality. Hold that thought; the rest of this article is essentially a deep dive into those three levers.

When Weaving Is the Right Tool — and When It Isn’t

Weaving earns its place in these situations:

• Wide groove joints that demand fill or cover passes beyond a stringer’s reach
• Multi-pass cover and cap passes where a smooth, uniform finish is wanted
• Fillet weld caps on tee and lap joints, especially where leg length must be even
• Vertical-up fill and cap passes, where the shelf of solidified metal needs to be spread and controlled
• Pipe work: cap passes, and some root and hot-pass techniques
• Hardfacing overlay, where the goal is deliberately spreading a hard alloy across a surface

Just as important is knowing when not to weave:

• Thin sheet. The puddle is small and the heat has nowhere to go. Oscillation almost always means burn-through.
• Root passes that need maximum penetration. Spreading the arc thins the heat input per millimeter at depth. If the spec calls for full penetration, a tight stringer is usually the honest choice.
• Out-of-position passes where the puddle will sag. Overhead and vertical-down work punish large, fluid puddles. A wide, slow weave in those positions is a drip waiting to happen.

For processes outside the usual steel playbook, the same physics still apply. Aluminum work, for instance, is far less forgiving of heat buildup, which is why Aluminum MIG Welding: The Complete Practical Guide stresses tight, disciplined travel over wide oscillation, and why Aluminum Flux Core Welding: What Actually Works in 2026 treats heat control as the deciding factor.

The Variables the Rest of This Guide Explores

Six things control every weave you will ever run:

1. Pattern shape — zigzag, crescent, figure-eight, triangle, U, and the rest
2. Weave width — how far the arc travels side to side
3. Dwell time at the toes — the brief pause at each edge that wets the sidewall
4. Travel speed — how fast the arc advances along the joint
5. Work and travel angles — how the electrode or torch is pointed
6. Process — SMAW, GMAW/FCAW, or GTAW each behave differently under the same motion

Before diving into patterns, one safety note worth stating early: longer, hotter weave passes produce more fume and more UV exposure than short stringers, so ventilation, respiratory protection, PPE, and fire watch per OSHA’s welding, cutting, and brazing standard 1910.252 are not optional, no matter how tidy the weave looks.

Next up: the stringer-versus-weave debate, and why codes and inspectors care more about it than most welders expect.

Stringer vs. Weave Bead: The Choice That Codes Care About

Before you pick a pattern, you have to settle a more basic question: should you oscillate at all? A stringer bead is a straight travel pass with no side-to-side motion. The electrode (or gun) advances along the joint line and the puddle is dragged forward, one consistent width, no weaving. A weave bead adds deliberate oscillation across the joint as the arc moves ahead. Neither one is “the correct answer.” They behave differently in heat, penetration, deposition, and risk, and the difference is what procedures and inspectors are actually measuring.

The Heat Input Math Behind the Choice

The standard formula for heat input in kilojoules per millimetre is:

Heat input (kJ/mm) = (Voltage × Amps × 60) ÷ (Travel speed in mm/min × 1000)

Notice what is missing from that formula: oscillation. The arc does not know whether you are weaving. What it does respond to is time. When you widen a weave while holding travel speed constant, the arc travels a longer path for every millimetre of forward progress, so it spends more time over each unit length of joint. Effective heat input per millimetre rises. Slow the weave down further, as most welders do when a puddle gets wide and fluid, and the number climbs again. That is why a slow, wide weave consistently runs hotter than a stringer deposited at the same amperage and voltage.

A stringer, by contrast, generally delivers deeper, narrower penetration and lower heat input than an equivalent-width weave bead run at the same settings. The arc stays concentrated in one place, cuts in, and moves on.

Stringer vs. Weave: Side-by-Side

Criterion	Stringer bead	Weave bead
Heat input per mm	Lower and consistent	Higher, rises with width and slower travel
Penetration profile	Deeper, narrower, more focused	Shallower, wider, spread across sidewalls
Deposition rate	Moderate; more passes to fill	High; fills a wide cross-section faster
Dilution	Higher at the root, more uniform	Lower and uneven across the width
Trapped slag risk	Low, continuous tie-in	Higher at overlap points and re-entrant corners
Grain structure / HAZ	Finer grain, smaller heat-affected zone	Coarser grain, wider HAZ from extra heat
Skill sensitivity	Forgiving of modest skill	High; width, dwell and speed must all be controlled
Typical procedure stance	Preferred or required for many passes	Allowed but often width-restricted

Why the Industry Leans Toward Stringers

The bias toward stringers is not tradition for its own sake. Straight passes produce finer grain structure, lower peak temperatures, and a smaller heat-affected zone, which generally means better toughness and less distortion. They also reduce the chance of centerline issues, since a wide, slow puddle that freezes down its middle can trap segregation and, in susceptible materials, invite centerline cracking. Stringers are also far less punishing: if your hand wanders slightly, the effect on the finished bead is small. A wandering weave shows up immediately as uneven width, one-sided undercut, or a humped centerline.

There is also a productivity counter-argument, and it is a fair one. In flat position on thick material, a well-controlled weave fills a groove in fewer passes. On out-of-position work and root passes, though, the arithmetic flips: stringers keep the puddle small and controllable. Field stick welding is a good example of where weaving remains completely normal practice, which is why a guide like advantages of stick welding still matters to anyone choosing a process for structural work.

What AWS D1.1 Actually Says

This is where welders get conflicting advice, so it is worth being precise. The AWS D1.1 welding procedure specification forms list “Stringer or Weave” as a required technique field, and the forms also reference maximum heat input, which means the procedure drafter, not the welder, decides which technique is permitted and how hot the pass may run. D1.1 does not set a general blanket bead-width limit for manual SMAW weave beads; AWS discussion records confirming this point are publicly available in the AWS forum thread on weave width limits.

That said, restrictions do exist elsewhere. Some code clauses and most fabricator procedures cap weave width, and for GMAW/FCAW the code limits weld layer width, requiring the welder to split the layer when a single layer exceeds the specified width. Equivalent technique fields appear in ASME Section IX, where QW-483 addresses weave requirements, and sample procedure forms such as the AWS sample WPS form show exactly where stringer-or-weave technique is declared.

In practice, run stringers where the procedure asks for them, keep your weave within the stated width whenever a limit applies, and measure rather than guess: bead width with a machinist’s rule, heat input from your actual voltage, amperage, and travel speed. The comparison only matters if you can prove which one you laid down.

The Main Weave Patterns and When to Use Each One

Ask five welders to name the weave patterns and you will get five different lists, because shop names drift: one calls a motion a crescent, the next calls it a half-moon. The names matter less than three things: the shape of the side-to-side motion, where the dwell happens, and how fast the arc advances along the joint. Get those three right and the pattern gives you predictable bead geometry, flat across the middle and tied in at both toes. Get them wrong and the same motion produces undercut, a humped center, or a cold sidewall. Here is the standard catalog, roughly in the order most welders learn it.

The Core Pattern Catalog

The list below is the same set of motions you will find in most training programs and pattern overviews, including the common weave pattern guide from Arc Captain:

Pattern	Hand motion	Bead geometry	Best fit
Straight / stepped	Tiny side steps with a small forward advance each cycle	Narrow, flat, tightly stacked ripples	Narrow grooves, fill passes, layered work where width is limited
Zigzag	Sharp back-and-forth with a steady forward creep	Wide, flat cover with visible side breaks	Flat and horizontal caps, multi-pass cover passes
Crescent	Arc from toe to toe along a curved path, pausing at each end	Smooth, gently crowned ripple, good toe wetting	Horizontal fillets, tee caps where leg symmetry matters
Circular / oscillation	Continuous small loops advancing along the joint	Fluid, wide, soft-shouldered bead	Flat-position wide fills and cover passes on thick plate
Figure-eight	Loops that cross over the centerline	Flattens the middle, distributes filler evenly	Wide caps where centerline buildup is a risk
Triangle (“Christmas tree”)	Pause at the two outer corners, draw the apex forward along the center	Layered, stacked shelves with a pointed leading edge	Vertical-up groove fill and cap
U-shape / J-shape	Rounded sweep across the bottom, pull straight back up the side	Deep, directed puddle that pushes slag along	Vertical and overhead fills, fast-fill electrode work
Trapezoid	Flat bottom sweep with angled returns	Wide bead with near-equal time at both toes	Horizontal fillets and wide cover passes needing even toes
Whip and pause	Whip the arc forward to wash the puddle ahead, pause back in the crater	Deep penetration, thin light ripples	E6010 root passes, open-root pipe, poor-fit joints

The triangle and the tight zigzag with edge pauses are the two standard vertical-up SMAW manipulations, and they are what most welders reach for on 3G groove tests. The whip-and-pause is a different animal: it is a penetration tool rather than a coverage tool, and it belongs to advantages of stick welding territory, since flux-cellulose rods like E6010 are designed to be whipped.

Dwell: Where You Pause Matters More Than the Shape

Two welders can run the identical zigzag and get completely different results because one paused in the wrong place. Pausing roughly half a second at each edge lets the puddle wet into the sidewall and fill the toe, which is the standard fix for undercut in vertical-up weave passes. Dragging quickly across the center keeps the middle from piling up into a humped, convex bead. In practice you are trying to spend equal time at both toes and minimal time in the middle, which is why instructors tell students to count a slow “one” at each side and hurry across.

Uneven dwell is just as damaging as wrong dwell. If you linger on the left toe twice as long as the right, you get asymmetrical fusion and, on a horizontal fillet, a lopsided leg. Equal pauses are the whole point of the triangle and crescent patterns.

How Wide Should the Weave Be?

Weave width should stay controlled and roughly proportional to electrode or wire size rather than spanning an entire groove in one pass. A 1/8 in. electrode earns a wider pattern than a 3/32 in. one, but “earns” is the key word: the arc has to stay hot enough at the edges to fuse the sidewall, and a wider sweep spreads the same amperage over more metal. Many fabricators write an explicit weave or layer width limit into the qualified welding procedure specification, and that number overrides any rule of thumb you learned in the field.

Pattern Selection Matrix

Condition	Recommended pattern
Narrow joint, thin section, root or fill	Straight/stepped or tight zigzag with edge pauses
Wide flat cover, thick plate, productivity matters	Circular or figure-eight
Horizontal fillet, leg symmetry required	Crescent or trapezoid
Vertical-up groove, structural test	Triangle or tight zigzag with equal half-second edge pauses
Overhead, gravity working against you	Small triangle or tight crescent, short cycle, fast travel
Appearance-critical cap, controlled pace acceptable	Crescent or figure-eight with deliberate dwell
Root with E6010, open root pipe	Whip and pause

For rod selection that supports these motions, the all position welding rod guide is a useful companion when you are moving between flat, vertical, and overhead passes.

Walking the Cup vs. Freehand (GTAW)

TIG adds a weave motion that belongs in its own category. Walking the cup is a technique where the ceramic cup rests on the joint or bevel and the welder rolls the torch in a figure-eight, walking the cup forward so each step sets the weave width and the pitch between ripples. Cup size effectively controls the pattern: a bigger cup walks a wider, coarser weave, and a smaller cup tightens both step and width. It is repeatable, tiring less than freehand, and dominates pipe root, fill, and cap work. Freehand weave is the alternative, with the tungsten held off the surface and the hand doing all the oscillation, and it is what you use when the joint geometry will not support a cup, such as tight corners, fillets, or repair work on thin material. Filler timing stays the same either way: dab at the leading edge of the puddle, not into the arc. For the wider context of this motion in high-specification work, see our aerospace TIG welding guide.

The practical takeaway: pick the smallest pattern that covers the required width, put your dwell at the toes, and let the procedure set the ceiling.

Heat Input, Puddle Control, and Metallurgy of a Good Weave

Two welders can lay down beads that look almost identical under white light, while the arcs that produced them leave completely different metal behind. Appearance tells you about the hand. Microstructure tells you about the heat. Once you understand what a weave does to the zone sitting just behind the puddle, weave discipline stops being a style preference and becomes a quality requirement.

What excess heat does to the joint

Every arc puts energy into the base metal, and the region that never melts but gets hot enough to change is the heat-affected zone, or HAZ. When heat input climbs, that zone simply stays at elevated temperature longer, and grains in it grow. Coarse grains mean reduced toughness, so a wide, slow weave can leave you with a cap that passes visual inspection and then fails under load. The same excess energy produces greater distortion, because a larger volume of heated material shrinks and pulls on the assembly. When a plate goes out of square mid-job, the cause is almost always travel that was too slow for the amperage setting.

The effect gets sharper when you are not welding plain carbon steel. On stainless and nickel alloys, prolonged exposure in the carbide precipitation temperature range (roughly 800–1500°F / 425–815°C for common austenitic grades) can sensitize the material: carbon combines with chromium and drops out of solution at the grain boundaries, leaving adjacent zones depleted and measurably weaker in corrosion resistance. Toughness-critical alloy work carries the same concern, which is why Alloy Steel Welding Electrodes are matched to the service conditions rather than picked for deposition speed. Procedures for corrosion-resistant alloys, including Alloy 20 Welding, are written the same way: the technique exists to protect the service properties of the alloy, not just to fill the joint.

Reading the puddle instead of the arc

The puddle is the fastest feedback loop you have, and it answers before the finished weld does. A healthy puddle sits roughly semi-circular behind the arc and stacks up as overlapping coins or scales as you advance. When that shape changes, the cause is predictable:

• Teardrop puddle: excessive heat or travel that is too slow. The tail drags out, the metal stays fluid too long, and out of position it sags. Centerline buildup follows.
• Narrow, ropey bead: insufficient heat or travel that is too fast. Filler never wets the sidewalls, so you get cold lap and lack of fusion even though the bead looks dense.
• Humped, convex bead: wrong dwell distribution. Time sits in the middle, the center crowns, and the toes stay thin.

The correction is always one of three levers: narrow the weave, move the dwell, or change travel speed. Raising amperage to “help” a cold wide weave usually just gives you a hotter puddle with the same unfused edges.

Interpass temperature and pass sequencing

Interpass temperature matters as much as the dial setting. In multi-pass work, every pass re-heats the one below it, so thermal history accumulates. A dependable sequence is to fill with stringer beads and reserve the weave for the final cosmetic cap. Stringers keep each pass narrow and let layers cool in between, while a controlled weave on top delivers the width and appearance you need without asking one pass to melt the entire groove. Filling the whole joint with wide weaves stacks heat on heat and pushes both the HAZ and the interpass temperature upward. ASME Section IX treats heat input as a qualified variable for exactly this reason: the procedure, not the welder’s preference, decides how much energy the joint may see.

Dilution: why a weave cap over stringers wins

Dilution is how much base metal ends up mixed into your deposited weld metal. A single wide weave draws in more of the groove faces and root across its length, shifting the deposited chemistry away from what the filler was designed to deliver. That matters whenever the cap is meant to carry toughness or corrosion resistance, and it matters more when the filler is a matching or overmatching alloy. Running stringer fills first and finishing with a weave cap keeps dilution concentrated in the lower layers, where it belongs, so the surface weld metal stays closest to the specified composition. The result is better mechanical properties at the surface than a full-groove weave provides, plus a cleaner profile. It is the same logic behind weaving in hardfacing: spread the expensive alloy evenly while keeping the base-metal contribution predictable.

Weaving by Process: SMAW, GMAW/FCAW, and GTAW

The side-to-side idea doesn’t change from process to process, but almost everything else does. Arc force, puddle fluidity, filler delivery, and how flux or shielding gas behaves all shift the moment you swap a stick electrode for a wire or a tungsten. A weave that looks textbook under a 7018 rod can sag badly in spray-transfer MIG. Here is the same core logic, rig by rig.

SMAW: Weaving With Flux and Arc Force

Stick welding rewards discipline over flash, and the first habit is arc length: roughly equal to the electrode core diameter, which for common stick sizes works out to about 2 to 3 mm. That short arc concentrates the force that drives the puddle into the joint. On a vertical-up groove it preserves the upward arc force holding metal in the joint; stretch the arc and you lose that push, the puddle goes flat and wide, and spatter replaces penetration.

Work angle for most groove and fillet passes is a drag or pull angle of about 10 to 15 degrees back from perpendicular, with vertical-up fillet work often run closer to straight up. Then comes the weave itself: move across the joint, pause briefly at each toe to wet the sidewall, cross the center at speed. Keep both edge pauses equal, or one side undercuts while the other builds a shoulder. Why out-of-position stick work is so demanding is covered in our all position welding rod guide.

The electrode changes the rhythm more than anything else. With E6010, the classic motion is whip and pause: a short forward push into the puddle, then a brief pull back off it, letting the deep-penetrating arc dig while the cellulosic flux gases find an easy way out. With E7018 the whip disappears completely. You run a tight side-to-side weave with the arc buried in the puddle, pausing at the toes as the heavy coating lays down a clean, high-deposition bead. Rush a 7018 weave and you get trapped slag, lumpy ripples, and a cap the inspector grinds out. Stick welding’s enduring strengths on structural work are explored further in advantages of stick welding.

GMAW and FCAW: Transfer Mode Decides Your Weave Width

Wire processes give you a smoother puddle and faster travel, which makes wide weaves tempting and risky at the same time. Everything hinges on transfer mode. In short-circuit transfer, heat input is low and the puddle is small and comparatively stiff, so a modest weave works well even overhead and vertical. In spray transfer, a continuous stream of droplets plus higher heat makes the puddle large and very fluid. Weave too wide in spray and the metal sags, especially out of position, and the cap you were trying to fill comes out soft and humped instead of flat. That is exactly why many GMAW and FCAW procedures restrict weave width outright and specify stringer passes instead: splitting a wide layer into several narrow ones keeps peak heat input per pass lower, reduces distortion, and lowers the risk of centerline cracking in the final cover pass.

Gun angle matters more than most welders expect. At identical settings, a push angle, with the gun pointed in the direction of travel, produces a wider and flatter bead because the arc force spreads the puddle forward, while a drag angle concentrates it into a narrower, deeper profile. On thin aluminum in particular, a push technique combined with a small, quick weave is usually the safe combination, and the torch handling and settings behind that are covered in our aluminum MIG welding guide. Flux-cored wire adds a slag layer on top of all of this, so treat it more like stick: consistent overlap, clean interpass, and no patterns that leave sharp re-entrant corners where slag can hide.

GTAW: Walking the Cup vs. Freehand Weave

TIG is where weaving turns into a metronome. Walking the cup rests the ceramic nozzle on the joint and rolls it side to side, so the cup diameter effectively becomes your step pitch: a larger cup takes bigger steps and covers more width per cycle, a smaller cup gives finer and more controllable increments. Because the nozzle physically indexes each step, walking the cup is the most repeatable weave motion in welding, which is why it dominates pipe cap work. The trade-off is geometry. The cup has to sit and roll against the work, so it works beautifully on pipe and open joints and poorly in tight corners or fillet joints where it cannot contact the surface. Holding that kind of repeatability is the baseline expectation in aerospace TIG welding.

Freehand weaving is the alternative: the torch hand floats, usually propped on a heel rest or a planted finger, and the oscillation comes from the wrist. It is less consistent but far more versatile, and it is the normal choice on positional sheet and thin-wall tubing. For either method, start with the tungsten roughly 15 to 20 degrees off perpendicular in the direction of travel and adjust from there based on puddle shape. The filler rod enters at the leading edge of the puddle, and the dab timing should lock to the weave rhythm: dip at one toe, feed as you cross, brief pause at the other.

Process Comparison at a Glance

Process	Typical angle	Preferred weave	Main risk
SMAW, E6010	Near perpendicular, slight drag	Whip and pause	Over-long arc, spatter
SMAW, E7018	10–15° drag	Tight zigzag with equal toe pauses	Trapped slag, undercut
GMAW short-circuit	Push or drag 5–15°	Small crescent or zigzag	Cold sidewalls
GMAW spray	10–15° drag	Wide weaves restricted; stringers preferred	Sagging, excess heat
GTAW	15–20° off perpendicular	Walking the cup or freehand crescent	Inconsistent step pitch

What the Paperwork Controls

One detail worth knowing before you head back to the booth: your procedure paperwork has opinions about weaving. AWS D1.1 procedure forms include specific fields for oscillation traverse length, traverse speed, and dwell time on mechanized and automatic applications, and ASME Section IX, QW-483, defines equivalent weaving variables for qualification. Manual freehand weaving is harder to quantify, which is precisely why inspectors fall back on procedure limits such as maximum weave width or layer thickness. If the qualified WPS calls for stringers, run stringers, no matter how good a weave would look. Reference material is available from the AWS sample WPS form and the ASME Section IX QW-483 reference.

Choosing a Weave by Welding Position and Joint Type

A crescent weave that lays a glassy, flat cap in the flat position will dump a molten blob into your helmet overhead. Position decides which weave patterns survive: gravity may support the puddle, pull it sideways, or try to tear it off the joint, and joint geometry narrows the choice further. Here is how to match pattern to position and joint.

Flat Position: The Widest Freedom

Flat is the only position where the puddle sits fully supported by the backing or the metal beneath it, so it gives the widest pattern latitude and the highest productivity. Circular and wide zigzag weaves work well here; travel speed is usually the limiting factor, not puddle control. You can stretch weave width to the top of what your procedure allows because sagging is not a threat, but heat still accumulates: wide plus slow equals a coarse heat-affected zone, exactly what the metallurgy section warned about. Stringer fills under a woven cover pass keep total heat lower than filling the whole groove with wide weaves.

Horizontal Fillets: Keeping the Throat Balanced

In a horizontal fillet, gravity pulls the puddle downward, so the lower leg grows heavy while the upper toe wanders or undercuts if you let it. The fix is a tight crescent or short stepped motion with an even pause at the top toe, which wets the upper sidewall before the filler can sag away from it. Keep the weave short: about one and a half to two electrode or wire diameters, because wider than that and the top half of the puddle outruns its own surface tension. Done correctly, the face reads as shallow, overlapping crescents with equal legs. Training texts treat horizontal fillet work as puddle shaping first and deposition second; General welding curriculum material is a useful cross-check.

Vertical-Up: The Most Weave-Dependent Position

Vertical-up is where weaving earns its reputation: the puddle is held up by the frozen metal behind and beneath it, so the pattern does structural work, not cosmetic work. Two motions dominate: the triangle, often called the Christmas tree, and a tight zigzag with pauses. The triangle drives heat into both sidewalls on the up-stroke and lets the center freeze before it piles up; a zigzag with brief, equal dwell at each toe ties in with an easier hand rhythm. Advance about 1 to 2 mm per cycle: larger steps give a rough, scalloped bead with inconsistent fusion at the tie-ins, while tiny steps simply overheat the joint. Watch the trailing puddle, not the arc, and keep both edge pauses identical, because unequal dwell is the classic cause of one-sided undercut on vertical groove welds. For stick work, a common pairing is a whipping E6010 root pass followed by a tightly woven low-hydrogen fill and cap; our all position welding rod guide breaks down which electrodes behave best where.

Overhead: Compact and Quick

Overhead punishes wide, slow puddles because anything large and fluid drips. Keep the weave minimal and fast: a small triangle or a tight circular motion, narrow width, a quick sweep across the center with brief toe pauses, and a tight arc length. Slightly lower amperage and a smaller electrode (3/32 in. or 1/8 in. stick sizes) gives a puddle you can actually hold up. The finished bead should stay slightly convex and narrow; a flat, wide overhead cap means you were too hot or moving too slowly.

Matching Pattern to Joint Type

• Butt groove joints: multi-pass and layered. Use stringers or narrow weaves for fills, then a weave cap if the WPS permits it. Watch layer-width limits, which some procedures impose on GMAW and FCAW passes.
• Tee and lap fillets: a slight crescent or stepped cap balances throat and face and ties in both toes, especially on larger legs built in multiple passes.
• Pipe: root work favors a freeze or keyhole technique rather than a wide weave; the hot pass and fills may take tight weaves; the cap is where walking the cup shines, with cup size setting your step pitch and weave width. Freehand weaving remains the fallback for tie-ins, misaligned joints, and restricted access.
• Overlay and hardfacing: weaving spreads wear-resistant alloy across a surface while managing dilution with the substrate and limiting heat into the base metal. Cross-weave and stripe-pass approaches both exist, and the procedure, not habit, decides between them. Filler selection matters as much as motion, so see our alloy steel welding electrodes guide.

Position	Preferred patterns	Why	Main caution
Flat (1G/1F)	Circular, wide zigzag	Puddle supported; productivity driven	Wide, slow passes accumulate heat
Horizontal (2F/2G)	Tight crescent, stepped	Counters downward puddle pull	Heavy lower leg, upper-toe undercut
Vertical-up (3G/3F)	Triangle, tight zigzag with pauses	Frozen metal supports the puddle	Steps over 2 mm; unequal toe dwell
Overhead (4F/4G)	Small triangle, tight circular	Minimizes fluid puddle volume	Sagging from wide, hot, slow weaves

Pick the pattern the position allows, keep the width the procedure allows, and let the puddle, not the arc, tell you whether the choice was right.

Common Weave Mistakes and How to Fix Them

Almost every weave defect an inspector flags comes back to one thing: where the welder spends time. Heat follows the hand, and the hand follows the eyes. Get the timing wrong on a single pass and the bead tells on you, no matter how steady the arc looked while you were running it. Here are the seven faults welders actually get blamed for, each tied to a specific weave error and a concrete fix.

The Seven Faults and Their Fixes

• Undercut at both toes. The classic weave failure. It happens when you cross the edge too quickly, or worse, pause in the middle of the pattern instead of at the sides. The fix is repeatable: a brief, equal dwell at each edge to let the puddle wet the sidewall, then a quick move across the center. Equal pauses matter, because a longer hold on one side produces undercut on the opposite toe.
• Humped, convex centerline. The mirror-image error. Dwelling in the middle stacks filler where it isn’t needed, giving you a crowned bead that grinds down badly and reads as a cold-look center. Speed up across the middle; the center of a good weave should be the thinnest part of the travel cycle, not the hottest.
• Cold lap or lack of fusion on the sidewalls. Usually a weave too wide for the amperage, or travel that outran the puddle’s ability to bite into the joint face. Narrow the weave first. If the joint geometry demands width, raise heat slightly within the limits of your qualified electrode selection and procedure rather than stretching a cold arc across the groove.
• Sagging or dripping in vertical and overhead work. A teardrop puddle is a clear signal: too much heat or too large a weave for the position. Drop the amperage, tighten the pattern, and shorten each weave cycle so the puddle solidifies before gravity wins.
• Trapped slag between weave segments. Risk climbs when a weave lays irregular valleys, sharp re-entrant corners, or inconsistent overlap that the next pass cannot fully clean or re-melt. This shows up most with low-hydrogen stick electrodes and flux-cored wires. Chip and brush interpass thoroughly, keep the overlap rhythm consistent, and avoid the jagged shoulders a wandering pattern creates.
• Uneven ripples and inconsistent bead width. The cause is almost never the electrode. It’s an unstable arc length, a wandering hand, and no cadence. Brace your body, pick a rhythm you can repeat, and drop to a smaller pattern until the width holds steady.
• Watching the arc instead of the trailing puddle. The single most repeated piece of weave coaching in any shop. The arc is brightness; the information is at the trailing edge of the puddle, where tie-in, width, and profile are actually decided. Experienced instructors emphasize the same point: keep your eyes behind the arc.

Troubleshooting at a Glance

Symptom	Likely weave cause	Correction
Undercut on both toes	Too fast at the edges, or dwelling in the center	Equal brief dwell each side, quick across the middle
Humped centerline	Too much time in the middle of the pattern	Accelerate across center; thin the middle of the cycle
Cold lap, sidewall non-fusion	Weave wider than amperage can support, or travel too fast	Narrow the weave; raise heat slightly within the WPS
Sagging or dripping	Puddle too hot and too large for the position	Reduce amps, narrow pattern, shorten the weave cycle
Trapped slag	Irregular valleys, sharp corners, inconsistent overlap	Clean interpass, keep overlap even, smooth re-entrant edges
Uneven ripples, drifting width	Unstable arc length, no rhythm, unbraced posture	Brace, set a cadence, step down to a smaller pattern
Porosity or erratic tie-ins at stops	Poor restart technique, crater left at the tie-in	Grind the stop, blend into the previous bead, restart ahead and walk back in

Work through the list in order on your next practice plate and the same pattern emerges: most weave faults are timing faults, not equipment faults. Fix where the puddle spends its time and the bead profile usually fixes itself.

Practice Drills, Code Compliance, and Final Takeaways

You don’t learn a weave by reading about it. You learn it in deliberate, measured reps, and the fastest way to progress is a ladder that starts with no weave at all.

A Four-Step Practice Plan

1. Straight stringers first. Run flat stringer beads on plate until you can hold a consistent width and a flat profile for the full length. This teaches puddle support, the skill every weave later depends on.
2. Tight zigzags. Add a small, rhythmic side-to-side motion roughly one electrode diameter wide. Keep the cadence even and the toe pauses equal so neither side builds a cold corner.
3. Widen on purpose. Scribe a target width on the plate and expand the weave to reach that line deliberately, rather than drifting wider by accident. Stop when you hit the mark, not when the joint merely looks full.
4. Repeat out of position. Move the same pattern to horizontal, vertical-up, and overhead, shrinking the weave each time gravity gets a vote. A good all position welding rod makes this stage far less frustrating.

Objective Self-Checks

Don’t grade your welds by vibes. Use chalk guide lines and scribed width marks as hard boundaries, then measure the finished bead with a machinist’s rule against your target. Brush or lightly grind a cross-section on practice plate and look for sidewall tie-in and a flat-to-slightly-convex profile. Photographing each pass gives you a record that makes improvement measurable instead of memorable.

The Paperwork Behind the Technique

Here’s the part many welders learn the hard way: your qualified procedure, not your hand, decides whether weaving is legal.

AWS WPS forms explicitly list “Stringer or Weave Bead” under technique, and the mechanized versions go further, adding fields for oscillation traverse length, traverse speed, and dwell time (AWS sample WPS form). On the pressure vessel side, ASME Boiler and Pressure Vessel Code Section IX QW-410 carries the same technique fields: stringer or weave bead, oscillation, and multipass or single pass per side. If the WPS says stringer, a flawless-looking weave is still a reject. If it allows a weave, it may also cap your layer width and heat input range, and you’re bound by those numbers too.

Safety Scales With Pass Length

Wide weaves mean longer, hotter passes, which means more fume and more UV exposure for everyone nearby. OSHA 1910.252 requires mechanical ventilation for welding most metals, with local exhaust used to keep toxic fumes below permitted exposure limits. That same standard incorporates fire-prevention precautions referencing NFPA 51B, so fire watch and combustible clearance aren’t optional after a long hot cap pass.

Final Word

A weave is a tool chosen for joint geometry and controlled by three numbers you actually own: width, dwell, and speed. Master the small, tight patterns before you ever reach for the wide ones. The welders who look effortless aren’t wiggling the rod; they’re running a disciplined cadence they built one measured bead at a time.

Frequently Asked Questions

What is the weaving welding technique?

Weaving is a controlled side-to-side or looping motion of the electrode, torch, or gun across the joint while the arc moves forward. It produces a wider bead than a straight stringer pass, helps tie in both sidewalls, and lets the welder spread filler metal over a broader area.

When should you weave instead of running a stringer bead?

Weave when the groove or fillet is wider than one stringer pass can cover, when you are filling or capping a multi-pass weld, or when you need to tie in both sidewalls of a wide joint. Use stringers when penetration, lower heat input, or procedure requirements matter more than coverage.

How wide should a weave bead be?

Keep the weave as narrow as the joint allows and always within the width your WPS permits. A practical field rule is to keep weave width proportional to the electrode or wire diameter rather than spanning the entire groove in a single pass, since wider, slower weaves raise heat input.

Why does my weave bead have undercut on the edges?

Undercut happens when the arc passes the toe too quickly or dwells in the center too long. Pause briefly and evenly at each edge so the puddle wets into the sidewall, then move quickly across the middle to avoid a humped centerline.

Which weave pattern is best for vertical-up welding?

The triangle (“Christmas tree”) pattern and a tight zigzag with short pauses at each edge are the two most reliable choices. Advance only about 1–2 mm per cycle and watch the trailing puddle — if it turns teardrop-shaped and starts to sag, you are too hot or moving too slowly.

Does weaving increase heat input?

Yes, indirectly. A wider weave forces you to move slower along the joint, and heat input rises as travel speed drops. That can coarsen the heat-affected zone, increase distortion, and, on stainless steel, risk sensitization — so control weave width and travel speed deliberately.

Is weaving allowed by welding codes?

Usually yes, but the qualified WPS decides. AWS and ASME procedure forms both require the technique to be declared as stringer or weave bead, and some procedures limit weave width or layer width, particularly for GMAW and FCAW, where a split layer may be required.

What is walking the cup in TIG welding?

Walking the cup is a GTAW weave method where the ceramic nozzle rests on the workpiece and is rocked in a figure-eight motion to advance the torch in consistent steps. It produces very uniform weave ripples, while freehand weaving gives more control over puddle shape at the cost of consistency.