MIG Welding Weave Patterns: Techniques, Charts & Mistakes

You watch a welder lay a wide, rippling bead that looks like stacked silver coins, then you try the same motion and end up with undercut, a wandering arc and a puddle that drops through the plate. The difference usually is not talent. It is knowing which weave pattern suits the joint, the position and the thickness — and exactly how long to hold at each edge.

MIG Welding Weave Patterns: What They Are and Why They Matter

Picture a small fabrication shop at the end of a long day. A welder finishes a bracket with a bead that looks textbook: even ripples, nice width, no obvious holes. The part goes to inspection, and the report comes back with a defect at one toe — lack of fusion. On the next bench, a hobbyist tries the same side-to-side motion on 16-gauge sheet and watches the puddle drop through the plate. Two different outcomes, one shared cause: the weave looked right, but the motion behind it was not matched to the joint, the position, or the thickness.

MIG welding — gas metal arc welding, or GMAW — uses a continuously fed wire electrode and a shielding gas to fuse metal. A weave pattern is a deliberate side-to-side motion of the gun across the joint as the puddle advances. A stringer bead travels essentially straight, with little or no lateral movement. Both are legitimate techniques. The mistake is treating a weave as a default style rather than a decision with consequences.

So what are weave patterns, and when should you use them? In plain terms, a weave is a controlled path that spreads the arc across a wider area. Welders use it to bridge a gap, fill a wide joint, cap a multipass weld, control a vertical-up puddle, or improve appearance on work where looks matter. The pattern name describes the path — zigzag, crescent, triangle, figure-8, loop, box, ladder — while amplitude, dwell, and travel speed determine whether that path helps or hurts.

What the Side-to-Side Motion Actually Does

Weaving is not just a visual trick. Moving the gun laterally widens the puddle, slows forward travel, and changes how heat is distributed. Because travel speed drops, heat input usually rises. A practical estimate is:

Heat input (kJ/in) = (volts × amps × 60) ÷ (1000 × travel speed in inches per minute)

That relationship is why a wide, slow weave can produce a broader heat-affected zone, more distortion, and a shallower penetration profile than a stringer at the same voltage and wire feed speed. The arc spends time at the edges, which can improve toe wetting — or cause undercut if the dwell is too long. It can help bridge a poor fit-up, but it can also hide a gap instead of sealing it properly. Good weave work depends on balancing amplitude, pause time at each toe, and forward progression. The shape in the middle matters far less than what happens at the fusion boundaries.

Procedure limits reflect that reality. Many welding procedure specifications restrict weave width relative to the wire or electrode diameter, and some prohibit weaving altogether on certain joints. That makes pattern choice a compliance question, not only a shop-floor preference. The relevant requirements sit in the applicable code and the qualified welding procedure specification, not in a generic rule of thumb.

The Patterns You Will Meet in This Guide

The main patterns covered here are the zigzag, the crescent or C-weave, the triangle, the figure-8, the circular or loop weave, the T-weave or box weave, and the ladder or step weave. We will also separate true weaves from the whip-and-pause motion, which advances and retreats along the joint rather than simply oscillating across it. Each has a place, and each has failure modes.

This guide is written for home hobbyists learning on a small machine, fabrication shop welders chasing consistent production beads, welding students preparing for position tests, and inspectors who need to read a WPS and recognize whether the motion on the floor matches the procedure. Later sections compare stringer and weave beads, decode each pattern, match patterns to flat, horizontal, vertical, and overhead work, explain settings and motion mechanics, examine heat input and metallurgy, troubleshoot defects, lay out a practice plan, and close with code, safety, and a checklist. Weaving also behaves differently across materials; on aluminum, for example, high thermal conductivity changes how quickly the puddle freezes, which is one reason aluminum MIG welding often favors stringers or very tight manipulation.

Finally, weaving means more arc-on time, which means more fume exposure. The International Agency for Research on Cancer classifies welding fumes as Group 1 carcinogens, so even bench practice deserves ventilation and sensible precautions. None of that makes weaving wrong. It simply makes it a technique worth understanding before you make it a habit.

Stringer vs Weave Bead: The Rule You Should Learn Before Any Pattern

Here is the rule that saves more welds than any pattern: the stringer is the default, and the weave has to earn its place. Most of the beautiful beads you admire online are weaving a joint that genuinely needs width, not a joint that was fitted badly and covered up.

So before you spend hours copying a zigzag, get clear on what each bead style is for. That single distinction decides your heat input, your distortion, your inspection results, and whether the weld passes a bend test or fails one quietly.

What a straight stringer bead gives you

A stringer is a straight run with no side-to-side motion. The arc stays planted, travel speed stays high, and the bead lands roughly one wire diameter wider than the puddle it creates. That focus buys you things a weave cannot:

• A tight, narrow heat-affected zone. Less surrounding metal gets cooked, so hardness and grain structure stay closer to the base metal.
• Deeper, more predictable penetration for a given amperage, because arc energy is not spread across a moving path.
• Lower distortion on restrained and thin parts, since you are not soaking a wide band of metal.
• Better behaviour on heat-sensitive grades. Thin sheet, stainless and aluminium all punish extra arc time. If you weld aluminium, the process rewards speed and a tight arc more than motion, which is why the complete practical guide to aluminium MIG welding leans hard on stringers for thin sections.

Stringers are also the honest choice for root passes, for anything under about 3 mm, and for repair work where you cannot afford to soften a wide strip of an alloy steel component. Filler selection matters as much as motion there, and matching consumables to the base metal is covered in this guide to alloy steel welding electrodes.

When weaving is genuinely the right call

Weaving is justified when a single narrow bead cannot do the job:

• Wide joints and gap bridging where a stringer would leave a ropey bead sitting in the middle of a groove.
• Filling thicker sections at a reasonable deposition rate without running an unreasonable number of passes.
• Vertical-up progression, where the side-to-side motion lets each pass tie into the shelf below and keeps the puddle from sagging.
• Capping passes where bead width and profile are specified, and where both sidewalls must be fused and dressed evenly.
• Appearance work on visible fabrication, where a uniform ripple and correct cap width are part of the deliverable.

The trade-offs you pay for width

Wider weaving slows travel speed, and slower travel speed raises heat input and widens the heat-affected zone. That extra heat shows up as more distortion, coarser grain growth in the HAZ, and reduced toughness in the region beside the weld. On quenched and tempered steels and on stainless, that is not a cosmetic issue.

There is a second cost that visual inspection cannot see. The arc spends most of its energy in the middle of the weave while the toes get a brief visit. Weave too fast across the middle or too wide for your amperage, and you get cold lap at the sidewalls even though the bead looks full and flat. With gas-shielded flux-cored and stainless wires, the edges are also where slag and oxide can get rolled into the bead instead of floating out.

How weave limits appear in a WPS

This is why procedure documents get specific. Excessive weave width is commonly limited in welding procedure specifications to a small multiple of the electrode or wire diameter, precisely to protect toughness and sidewall fusion. Under AWS structural welding practice and ASME Section IX qualification work, pass width, travel speed, current range, position and the number of passes are treated as variables that must be qualified, not invented at the bench.

When you read a WPS, look for four numbers before you strike an arc:

1. Maximum pass or weave width, often expressed as a multiple of wire diameter.
2. Travel speed range, which caps how slowly you are allowed to weave.
3. Dwell guidance, or the note that no dwell is permitted at the toes.
4. Number of passes, which tells you whether the engineer expects stringers or a weave fill.

Inspectors care because a weld made outside those limits may look correct and still not represent the qualified procedure. That is a documentation failure as much as a metallurgical one.

Get this hierarchy right and the patterns in the next section become tools instead of tricks.

The Main MIG We

The Main MIG Weave Patterns Decoded: Zigzag, C-Weave, Triangle and More

Every weave you will ever run is a variation on one idea: moving the arc across the joint, then coming back. What separates the patterns is where the arc spends its time. The shape gets the attention, but the dwell does the welding. Here are the common patterns, how each one behaves, and where it does its best work.

Zigzag: The Default Worker

The hand sweeps side to side in straight lines, with a slight pause at each toe before reversing. In zigzag weaving, welders generally spend less time at the centre and more at the edges to secure tie-in and avoid undercut. Best for flat and horizontal fillets, thin-to-medium plate and general fabrication. Typical defect when rushed: undercut at the toes and a bead that looks wide but sits on top of the plate. Keep the amplitude modest on anything under 6 mm (1/4 in).

Crescent or C-Weave: Edges First

The arc travels in shallow arcs, curving up to the top toe, across, down to the bottom toe, then back through the middle without stopping. The curved path wraps the sidewalls better than a straight zigzag and gives a smoother, more uniform toe line. It shines on horizontal fillets and capping passes where appearance matters. Watch the lower toe for sag and overlap when the puddle gets heavy.

Triangle: Wide Grooves, Controlled Centre

The arc peaks at the centre of the joint, then drops to the base corner on one side, pauses, crosses back to the other corner, pauses, and returns up the middle. The pause at the base corners drives fusion into the root of a wide groove; the quick pass through the centre keeps the puddle from piling up. Good for fill passes in V-grooves and butt joints. Too much centre time produces a convex, ropey bead with cold sidewalls.

Figure-8 and Circular Weaves: Wide Caps and Overhead

The figure-8 crosses the centre of the joint on every loop, which spreads heat oddly well and keeps a wide cap flat. It is a favourite for overhead work and for capping wide joints, because the crossing motion keeps the puddle from running away from you. The circular or loop weave is a smaller, rounder version used on light sheet and for cosmetic caps. Both demand a small, cool puddle; if the puddle is fluid, the loops turn into a sagging mess.

T-Weave, Box and Ladder/Step Patterns

The box or T-weave moves in short right-angle steps: across the joint, hold, up or down, across again. The ladder pattern does the same thing with pauses at each corner. These are useful for wide caps and for filling large gaps where you need metal deposited in predictable places rather than smeared across the joint. They are slower and put more heat into the plate, so treat them as problem-solvers, not everyday motions.

Whip-and-Pause: Not Really a Weave

Whip-and-pause moves forward and back along the joint rather than side to side, so it behaves differently from a true weave in terms of heat input and penetration. The forward whip lets the puddle cool for a fraction of a second before you step back into it. It is popular in flux-cored work and on thin material, but it does not widen a bead and can leave a stack of humps if the rhythm slips.

Quick Note: The pause at the toes matters far more than the shape in the middle. A perfect serpentine with no dwell leaves lack of fusion at the sidewalls and undercut at the edges. Two crisp pauses per pass will beat a beautiful shape every time.

Amplitude, Frequency and Pitch in Plain Language

Amplitude is how far the arc travels side to side, measured as total weave width. Frequency is how many complete cycles you make per second or per minute. Pitch is the distance the arc advances along the joint in one full cycle, so pitch equals travel speed divided by frequency. Robotic and cobot welding software typically exposes weave pattern plus amplitude, frequency and dwell time as separate parameters, which is why a robot can hold a 6 mm weave at 2 Hz all shift while your hand drifts wider as you get tired. Human welders control the same three variables by eye and rhythm. On aluminium, that rhythm has to be tighter still; the puddle goes from stiff to collapsing in a heartbeat, as covered in our aluminum MIG welding guide.

Which Pattern Where

Pattern	Best position	Thickness / joint	Main risk
Zigzag	Flat, horizontal	3–10 mm fillet, butt	Toe undercut
Crescent / C	Horizontal, vertical-up	6 mm+ fillet, cap	Lower-toe overlap
Triangle	Flat, horizontal	Wide V-groove fill	Cold sidewalls
Figure-8	Overhead, flat cap	Wide caps, 8 mm+	Excess heat input
Circular	Flat, light sheet	1.5–4 mm cosmetic	Burn-through
Box / ladder	Flat, horizontal	Wide gaps, fill	Slow travel, sag
Whip-and-pause	Flat, vertical-down	Thin sheet, FCAW	Humping, cold lap

Pattern choice always sits underneath the procedure. If the WPS caps weave width, that limit wins over anything in this table.

Matching the Pattern to the Position: Flat, Horizontal, Vertical and Overhead

Gravity does not care how clean your hand motion looks. The zigzag that lays a wide, even cap flat can drop molten metal on the floor overhead, and a crescent that fills a horizontal fillet nicely can undercut the top edge if you stretch it too far. Position changes how fast the puddle flows, how long it stays fluid, and how much arc force you are fighting. Pattern selection starts with where the joint sits.

Flat Position: Where Weaves Are Most Forgiving

Flat is the one position where the puddle works with you. Heat builds quickly, fluidity is high, and you have room to move. That makes the zigzag and the circular or loop weave the go-to choices for wide caps and gap bridging. A steady side-to-side zigzag with a short hold at each toe lays down a broad, uniform layer. Circular motion keeps the puddle bunched and rolling, which helps when you bridge an open root or a sloppy fit-up gap.

The risk in flat is too much heat, not too little. A wide, slow weave on thin plate will punch through, and repeated wide passes pile up heat input. Keep amplitude to roughly what the puddle can support, keep moving, and remember that two or three stringers usually beat one heroic weave on anything but the widest cap.

Horizontal: Two Toes, Two Different Problems

In a horizontal fillet or groove, the puddle sits on a shelf of base metal and wants to sag. That creates two failure points at once. The upper toe is prone to undercut because arc force eats into the top edge and filler metal drains away before it can wet in. The lower toe is prone to overlap, rollover and sag, where metal piles up instead of tying in.

The answer is a tight, symmetric weave: short zigzag or crescent strokes with a deliberate pause at the top edge to let it fill, then a quick trip across the bottom. Amplitude that works fine flat will cause trouble here, so narrow it. If the upper toe keeps gouging, reduce amplitude and slow your travel slightly rather than adding voltage.

Vertical-Up: Build a Shelf, Then Stand on It

Vertical-up is where weave discipline really shows. The puddle is held up by the solidified metal below it, so your job is to keep the pool small and build a fresh ledge every cycle. Use a narrow zigzag or crescent, pause distinctly at each side, and let the leading edge freeze before you swing back. Each pause deposits a shelf that supports the next increment of molten metal.

Stretch the weave and the puddle grows past the point where surface tension can hold it, and it sags or falls. Widen the amplitude on a vertical-up cap and you will usually get a heavy, convex bead with lack of fusion at the sidewalls.

Vertical-down is the exception. Because travel speed is fast and the puddle is tiny, weaving rarely helps. A stringer bead, or a very tight step-and-pause technique, is standard on thin material where vertical-down is used for speed and to keep heat from blowing through.

Overhead: Small Puddle, Tight Motion

Overhead punishes a big puddle immediately. Everything you deposit is trying to drip, so the recipe is a smaller puddle, a tighter weave and a lower wire feed speed than you would run flat for the same material. A small figure-8 or tight circular motion gives you control because the arc stays centered and the metal is constantly worked back into the joint rather than hanging below it.

The trade-off is reduced travel speed and deposition, which means more heat cycles over one area and a real risk of lack of fusion if you get lazy with the pause at each toe. Keep stitches short, keep amplitude narrow, and resist the urge to widen the pattern to finish faster.

Joint Type and Pass Sequence: Root, Fill and Cap

Position is only half the decision; the pass you are running is the other half.

• Root pass: almost always a stringer or a very tight weave. The root needs penetration and consistent fusion, not width. On an open-root butt joint, a small side-to-side motion just enough to tie both lands together is plenty.
• Fill passes: moderate weaving is justified to cover area and bridge any remaining gap, but keep each layer within the width your settings can actually fuse.
• Cap passes: wider zigzag, figure-8 or circular patterns earn their place here, because appearance and even tie-in at both toes matter most.

Joint geometry also constrains you. In a lap joint, the puddle wants to run along the lower plate, so aim the arc slightly at the top sheet and keep the weave tight to the edge. An outside corner has almost no backing and burns through easily, favoring short, narrow strokes. A groove weld gives you sidewalls to pause against, while a fillet asks you to manage a triangular puddle that sags from the top toe. For filler and technique choices across these joints, our All Position Welding Rod Guide is a useful companion reference.

Note: Position-specific adjustments are normal. Many welders drop amperage or wire feed speed by roughly 10–15% when moving from flat to vertical or overhead, and typical MIG shielding gas flow sits in the 15–25 cubic feet per hour range, with drafty or out-of-position work often needing the higher end or added draft protection. These are starting points, not rules: the procedure specification always overrides preference. If your WPS limits weave width to a stated multiple of electrode diameter or prohibits weaving entirely, that limit governs the weld. Qualification and inspection standards from the American Welding Society are the usual basis for those requirements.

Position	Best patterns	Main risk	Typical adjustment
Flat	Zigzag, circular	Excessive heat, burn-through	Wider weave allowed; watch travel speed
Horizontal	Tight zigzag, crescent	Upper-toe undercut, lower-toe sag	Narrow amplitude, symmetric pauses
Vertical-up	Narrow zigzag, crescent	Sagging, sidewall lack of fusion	Smaller puddle, pause at each shelf
Vertical-down	Stringer, tight step	Burn-through on thin stock	Lower heat, fast travel, no wide weave
Overhead	Small figure-8, tight circle	Dripping, lack of fusion	Reduce wire feed speed, short stitches

One last caveat on aluminum: it forgives even less than steel out of position. High thermal conductivity pulls heat away fast, and the puddle gives you almost no warning before it sags. If you are taking these patterns to aluminum sheet or structural work, read Aluminum MIG Welding: The Complete Practical Guide first, because preheat, travel speed and weave width all shift compared with carbon steel.

Settings and Motion Mechanics: Making a Weave Pattern Actually Work

A weave is only a motion. Settings decide whether that motion produces a sound weld or a pretty failure. Shift voltage two volts, add 5 mm of stick-out, or swap CO2 for a 90/10 blend, and the same hand movement starts to undercut, spatter, or refuse to wet in at the toes.

Wire, Gas and Transfer Mode First

For carbon steel, MIG runs on direct current electrode positive (DCEP) with CO2 or an argon/CO2 blend. Wire diameter sets your current window. 0.030 in (0.8 mm) suits thin sheet and out-of-position work where a small, fast-freezing puddle is an advantage. 0.035 in (0.9 mm) is the general-purpose shop size. 0.045 in (1.2 mm) and up push deposition on thick plate, where the puddle is already broad and a wide weave adds little.

Gas choice changes puddle fluidity and penetration shape. Straight CO2 penetrates well and costs less, but the arc is harsher and spatter heavier. Argon-rich blends such as 75/25 or 90/10 smooth the arc and improve wetting, which makes toe lines noticeably easier to control during a weave. Aluminum is different again: pure argon, fast heat conduction, and weaves kept narrow. Our aluminum MIG welding guide covers that reasoning in detail.

Transfer mode decides what weaving does to the puddle. Short-circuit transfer occurs at lower current and gives a colder, more controllable puddle, which is why it dominates thin material and out-of-position work. Spray transfer, with argon-rich gas and high current, is smooth and high-deposition but dumps heat into a wide area fast. Globular sits awkwardly between them: rough, spattery, and a poor partner for toe control. The American Welding Society classifies these modes by how molten metal crosses the arc, and the mode you are in changes everything a weave does to penetration and bead profile.

Voltage, Wire Feed Speed and Inductance

Wire feed speed sets current; voltage sets arc length. Raise WFS without voltage and you get a stubbing, digging arc that burns through thin sheet and buries the wire mid-weave. Raise voltage without WFS and the bead flattens and spreads, inviting undercut at the toes, especially where you pause.

Inductance, sometimes labelled arc force, matters in short-circuit transfer. More inductance slows the current rise, softens the arc, cuts spatter and gives a wetter puddle. Less gives a crisper, colder arc. A slightly wetter puddle helps toes tie in, but overdo it on vertical work and the puddle sags before you can step up.

Angles, Stick-Out and the Arc You Actually Get

Travel angle and work angle change penetration and bead shape independently of the pattern. A push angle, torch tilted forward roughly 10 to 15 degrees, flattens the bead, improves visibility and reduces penetration. Drag digs deeper, narrows the bead and throws more spatter. Work angle splits heat between the plates; on a fillet, 45 degrees is a starting point, shifted toward the heavier member when thicknesses differ.

Contact-tip-to-work distance is the quiet variable that ruins more weaves than any pattern choice. A longer CTWD reduces current density and destabilizes the arc, so the puddle surges and stalls while your hand keeps a steady rhythm. Hold stick-out short and constant, roughly 10 to 15 mm (3/8 to 5/8 in) in short-circuit. Worn tips and clogged nozzles wreck consistency too, so keep spares and check gas flow; the Airgas welding equipment guide is handy when replacing consumables, and a supplier such as AllGas Welding Supply can match gas and tip sizes to your machine.

Rhythm, Dwell and Weave Pitch

Three numbers describe a weave: amplitude, frequency and dwell. Amplitude is how far you swing, frequency how fast you complete a cycle, dwell how long you hold at each toe. Dwell does the most work. Hold at the toe just long enough for the puddle edge to wet and wash in, then move with purpose. Count out loud while learning; uneven counting shows up immediately as uneven ripple spacing.

Weave pitch is the distance the torch travels in one full cycle, and it is the easiest thing to measure: each ripple marks one cycle, so ripple spacing is your pitch. Draw a 150 mm (6 in) line, weld it with your chosen weave, and time the pass. Distance divided by time gives a travel speed you can log and repeat. Consistent pitch means consistent heat input per millimetre, which is what inspectors and procedure limits are really about.

Quick Note: Judge stitch spacing against the puddle width, not a fixed millimetre figure. Each pass should overlap the previous one by roughly a third to a half of the puddle width. Narrower overlap leaves valleys that trap slag and cause lack of fusion.

Symptom Table: What to Change First

Symptom	Likely cause	Adjustment
Undercut at toes	Long dwell, high voltage, slow travel	Shorten dwell, drop 0.5–1 V, speed up slightly
Ropey, convex bead	Low voltage, short arc, cold puddle	Raise voltage in small steps, match WFS
Puddle too fluid, sagging	Excess inductance or heat, long weave	Reduce inductance, narrow amplitude, lower WFS
Wandering arc, spatter burst	Long stick-out, poor ground, dirty wire	Reset CTWD, clean ground, check liner and tip
Uneven ripple spacing	Inconsistent dwell and frequency	Slow down, count the pause, use rails
Burn-through on thin sheet	Too much current, weave too wide	Smaller wire, short-circuit, tighter weave

Drill: Soapstone Rails on a Practice Plate

Draw two parallel soapstone lines on a clean plate, about 10 mm apart for a first attempt. Weld between them, pausing at each line without crossing it. The lines remove the guesswork about amplitude, so your attention goes to the pause and the arc length. Run six beads, move the rails 3 mm wider, repeat. Then mark the plate in 50 mm segments and place the same number of cycles in every segment. When ripples line up with your marks, your rhythm is repeatable. Log voltage, WFS, gas flow and travel time beside each coupon, because a bead you cannot reproduce is a coincidence, not a technique.

Heat Input, Penetration and Metallurgy: The Hidden Cost of a Wide Weave

A weld can look like jewelry and still be a problem. The ripples are even, the toes are straight, and the crown sits nice and full — then a bend test opens a crack, or a macro etch reveals lack of fusion hiding under that pretty surface. Often the culprit is not the pattern itself but how much heat the pattern put into the plate, and where that heat went.

Heat input is the energy delivered per inch of weld. It is not the same thing as amperage. You can run the same amps and volts as the next welder and still put twice as much heat into the part if your travel speed is half as fast. That is the quiet cost of a wide weave: the side-to-side motion slows you down, and slower travel means more heat per inch.

Heat Input: The Number Behind the Bead

The standard estimate for arc welding is:

Heat input (kJ/in) = (amps × volts × 60) ÷ (travel speed in ipm × 1,000)

Say you are running 200 A at 24 V. A stringer at 12 ipm gives (200 × 24 × 60) ÷ (12 × 1,000) = 24 kJ/in. The same settings with a wide weave at 6 ipm gives 48 kJ/in. You never touched the machine, but you doubled the heat going into the joint. That is why heat input rises as travel speed falls, and why a slow wide weave deposits more heat per inch than a fast stringer at the same amperage.

What changes	Likely effect
Travel speed drops for a wider weave	Higher heat input, wider HAZ, slower cooling
Dwell at the toes gets longer	Better wetting, but more undercut risk and more heat at the edges
Amplitude gets wider	Broader, shallower penetration and more distortion across the joint
Weave outruns the puddle	Sidewall lack of fusion, even with a full-looking crown

HAZ, Cooling Rate and What They Do to the Metal

The heat-affected zone is the band of base metal that gets hot enough to change microstructure without melting. A wider heat-affected zone generally means slower cooling, which can reduce toughness in some steels and alter hardness near the fusion line. In hardenable carbon steels, slow cooling can leave hard, less ductile regions; in other grades it may soften the metal. Neither outcome is visible from the surface.

That is why procedure limits exist. A welding procedure specification may cap heat input, weave width, pass thickness, or interpass temperature for reasons that have nothing to do with how the bead looks. A beautiful cap that exceeds those limits is still a nonconforming weld.

Penetration Shape and the Sidewall Trap

A stringer tends to concentrate the arc, producing a narrower, deeper profile. A weave spreads the arc across a wider area, producing a broader and often shallower penetration profile. Done well, that broader profile helps wet into both joint faces. Done poorly, it creates a hidden defect.

Sidewall lack of fusion is common when the weave moves faster than the puddle can wet into the joint faces. The center fills, the crown looks proud, and the edges never truly bond. Extra wire does not fix this. The motion has to slow at the toes long enough for the puddle to bite, without dwelling so long that you undercut the edge or pile up heat.

Distortion, Thin Gauge and Residual Stress

Heat spread across a wider band causes more shrinkage across that band. The result can be angular distortion, buckling, or residual stress that pulls a assembly out of tolerance after welding. On thin sheet, a slow weave keeps the puddle fluid for too long and makes burn-through more likely. Weaving is not a substitute for good fit-up, and it is rarely the right fix for a gap that should have been tacked or backed.

Stainless and Aluminum: Different Rules

Stainless steel adds another concern: sensitization. When austenitic stainless stays in roughly the 800–1650°F (427–900°C) range, chromium carbides can precipitate at grain boundaries and reduce corrosion resistance. Long dwell and slow, wide weaves keep the metal in that window longer. Low-carbon or stabilized grades help, and so does removing heat tint, which is itself a sign of oxidation.

Aluminum is often welded with stringer beads or very tight weaves because of its high thermal conductivity and tenacious oxide layer. Heat runs away from the puddle quickly, and a wide, slow motion can spread the arc without improving fusion. If you are setting up for aluminum, our aluminum MIG welding guide walks through the machine and travel details.

Note: When a joint is wide enough that one weave would exceed the procedure’s width or heat-input limit, two controlled stringers are often the better engineering choice. They are easier to inspect, put less heat into any one band, and give you a second chance at sidewall fusion.

When the WPS and an Engineer Take Over

In code work, follow the procedure and the applicable standard. AWS publishes structural welding codes and qualification requirements that may restrict weaving, heat input, or pass sequence. For critical joints, pressure equipment, or unfamiliar alloys, bring in a welding engineer before you experiment. A good-looking bead is not evidence of soundness; only the procedure, the tests, and the inspection behind it can tell you that.

Troubleshooting Weave-Related Defects: Undercut, Lack of Fusion and More

Most weave defects announce themselves at the toes, where the arc dwells longest and the puddle freezes first. Nearly every one leaves a motion signature you can read off the bead. Ask three questions each time: what was the pattern doing, what was the machine doing, and what was the joint doing?

Short Note: Rule out the basics before you change your rhythm: gas flow, leaks or a wrong mix; wire condition and drive-roll tension; oil, mill scale or moisture on the metal; a loose ground or a failing liner. Chasing a weave fix on a gas-starved torch wastes plate.

Undercut at the Toes

Undercut forms when the arc lingers at the toe long enough to gouge the base metal without filler refilling it, or when edge travel is so fast the puddle never catches up. In short, dwell and travel are out of balance. Reduce amplitude slightly, shorten the pause at each toe, and move decisively across the middle instead of drifting. Then check the machine: excess voltage, too much wire feed speed, a stick-out that has crept long, or a work angle aimed at the vertical leg rather than into the joint.

Overlap and Rollover

Rollover is the mirror image: the puddle climbs the upper toe and freezes there without fusing, leaving a cold lap. Typical causes are weaving too wide for the available current, a slow edge pause, or a drag angle that shoves metal ahead of the arc. Tighten the weave, quicken the edges, and switch to a slight push. If the bead still sits on the plate like a rope, the puddle is too cold, which is a current problem, not a rhythm one.

Lack of Fusion at Sidewalls and Root

Lack of fusion at weave sidewalls happens when the puddle is too cold or the weave is too wide for the current. The arc never wets the sidewall; it just lays metal alongside it. Narrow the amplitude, hold at the toe until the puddle flattens and bites, and raise wire feed speed or voltage so the arc can melt the joint face. At the root, suspect the prep as much as the motion: a tight gap, a thick land, or an unclean bevel defeats a perfect weave. This defect is often invisible from the surface, which is why AWS qualification work relies on bend tests, macro sections and volumetric inspection rather than a good-looking cap. On aluminium, where oxide fights every pass, pair the fix with proper cleaning as covered in our aluminium MIG welding guide.

Porosity and Arc Instability

Porosity is usually a gas coverage or contamination problem rather than a pattern problem, but erratic weaving makes it worse: a whipping, jabbing motion drags the nozzle around, changes contact-tip-to-work distance and disturbs the shielding envelope. Smooth the motion, shorten stick-out, hold a steady push angle and confirm flow at roughly 15–20 cfh. Rising spatter mid-bead usually means arc length is fluctuating with your hand, a rhythm problem dressed as a machine problem. Prep matters too; our guide to aluminium cleaner for welding covers why a wiped surface is not the same as a clean one.

Burn-Through on Thin Sheet

Thin gauge punishes weaving because every sweep returns heat to a spot already near melting. Drop the weave for stringers, or use a very tight step technique with fast travel. Lower the current, speed up, and back the work with a chill bar or copper backing where possible. If you must widen the bead, use two overlapping stringers rather than one slow weave.

Humping, Convexity and Uneven Ripples

Humping at high travel speed happens when the puddle outruns the arc and piles up ahead of it; slow down, or raise current so the puddle stays flatter. Excess convexity, which concentrates stress at the weld toe, comes from slow weaving with insufficient current: the metal piles instead of wetting out. Raise current, cut the edge pause and add a slight push angle. Uneven ripple spacing is almost always a counting failure. Mark rails with soapstone, hold the same beat at each toe, and drive from the forearm, not the wrist.

Slag Entrapment with Gas-Shielded Flux-Cored Wires

A wide weave gives molten slag more chances to roll into the puddle, especially at the reversal points where the arc stalls. Keep the weave tight, the arc on the leading edge of the puddle, a slight drag angle, and clean between passes. Any slag left behind becomes a planar discontinuity.

Defect	Likely Cause	Weave Fix	Setting Fix	Prevention
Undercut	Long toe dwell	Shorter pause, tighter amplitude	Less voltage, fix stick-out	Watch the toes
Rollover	Weave too wide for current	Narrow, decisive edges	Raise voltage or WFS	Match width to thickness
Lack of fusion	Cold puddle, wide weave, poor prep	Hold until sidewall wets	More current, check inductance	Bend or macro test coupons
Porosity	Gas loss or contamination	Smooth, steady nozzle path	Flow rate and mix check	Clean metal, sealed torch
Burn-through	Weaving on thin sheet	Stringers instead	Lower current, faster travel	Backing or chill bar
Humping	Travel outruns the puddle	Steady speed, lower frequency	Raise current against travel	Rehearse on scrap
Slag entrapment	Wide reversal points	Tight weave, arc on leading edge	Check polarity and drag angle	Clean between every pass

Fix one variable at a time and log it: change the dwell, run a coupon, cut it open, then decide. A welder who can name the defect from the toe line and reach for the right correction is worth more than one who simply weaves harder and hopes.

Practice Plan: Drills, Self-Assessment and Robotic Weave Parameters

A weave pattern only becomes useful when your hand can repeat it without thinking. The practice plan below builds that repetition in layers: first the motion, then the joint, then the position, and finally the feedback loop that tells you whether the bead is actually sound.

Build the Motion Before You Build the Joint

Pad-of-beads practice on flat plate is the standard way to build consistent weave amplitude and rhythm before moving to joints. Draw two soapstone rails about one bead width apart and run a zigzag or crescent between them, pausing briefly at each line. The rails give your eyes a fixed target and stop the weave from growing wider every pass. Keep each pass short at first, using two-inch weave segments, then overlap the next bead by about one-third.

Use a pause-count drill to fix the timing at the toes: say “one” at the left rail, “two” across the middle, “one” at the right rail. The count should stay boringly even. If the middle gets more time than the edges, the bead crowns; if the edges are skipped, undercut appears.

Once the flat pad looks consistent, move through a simple progression:

1. T-joints. Run short fillet welds and watch both toes instead of the arc.
2. Vertical-up. Use a narrow weave, keep the puddle small, and pause at each shelf.
3. Overhead. Reduce the puddle, tighten the motion, and accept a slower pace.

The position progression mirrors what any all-position welding guide teaches: flat first, then horizontal, vertical, overhead. Do not jump ahead because the flat bead looks good. Gravity changes the puddle in every position, and your rhythm has to adapt.

Read the Bead Like an Inspector

Surface appearance is only the first check. Review each practice bead with the same questions an inspector would ask:

• Toe lines: Are they straight and even, with no undercut or rollover?
• Ripple spacing: Do the ripples repeat at a steady interval, or do they bunch and stretch?
• Weave pitch: Is the side-to-side spacing consistent from start to finish?
• Spatter and surface defects: Are there signs of arc instability, porosity, or trapped slag?

Then cut a coupon and look inside. Cutting and etching a practice coupon reveals penetration and fusion that the bead surface cannot show. A macro test is especially useful for checking sidewall fusion and root penetration; a bend test, if you have the equipment, exposes lack of fusion that a visual inspection would miss.

Keep a Log That Makes Success Repeatable

Keeping a settings and travel-speed log makes it possible to reproduce a good bead instead of guessing. Record voltage, wire feed speed, gas flow, contact-tip-to-work distance, travel angle, pattern, weave width, dwell count, position, joint type, and a short note on the result. Add a photo of every keeper bead. After a few sessions, patterns emerge: you will see which combination produced a clean toe line and which one only looked good under shop lighting.

Translate the Same Motion to a Robot

Robotic weaving parameters typically separate pattern shape, amplitude, frequency and dwell, allowing repeatable bead width control. Amplitude sets the side-to-side width, frequency sets how many cycles occur per unit of travel, and dwell controls the pause at each edge. The robot holds what a human struggles to repeat: identical pitch, identical toe timing, and identical travel speed on every pass.

The trade-off is flexibility. A robot follows the programmed path, while a person can react to poor fit-up, a warped edge, or a repair that needs a different approach mid-bead. Manual weaving still wins on fit-up variation and repair work; robotic weaving wins on production repetition. Even an automated cell should run inside a qualified procedure, and AWS provides the qualification framework many shops follow.

Factor	Manual weaving	Robotic weaving
Consistency	Depends on operator rhythm	Highly repeatable
Fit-up variation	Adapts in real time	Requires a stable joint
Best use	Repair, small runs, awkward joints	Production, long straight welds
Parameter control	Learned by feel	Set numerically

Quick Note: Realistic improvement takes several focused practice sessions, not one afternoon. Short, deliberate drills with a log and a cut coupon will teach you more than hours of uncontrolled bead-running.

Common Mistakes, Code Compliance, Safety and Final Checklist

You know the patterns, positions and settings by now. What separates the welder who passes inspection from the one who grinds and re-runs is judgment: knowing when not to weave, staying inside the procedure, and reading the puddle instead of the arc.

The Mistakes That Cost the Most

These seven show up in rework reports again and again, and almost all of them are technique problems, not machine problems:

• Weaving to compensate for bad fit-up. A wide gap or proud land is a fit-up problem; weaving across it leaves a wide bead floating over an unfused root.
• Using one pattern everywhere. The triangle that works overhead will undercut a horizontal fillet. Match the pattern to the position and joint.
• Weaving too wide on thin metal. Past roughly two to three electrode diameters, extra dwell dumps heat faster than thin sheet can sink it: burn-through with pretty ripples around the hole.
• Ignoring WPS limits. If the procedure calls for stringers and you weave, the weld is out of procedure however it looks.
• Inconsistent dwell. Uneven time at the toes is the top cause of wandering toes, uneven ripple spacing and local undercut.
• Watching the arc instead of the puddle edges. The light hides everything. Track the puddle’s leading edge and both toes, not the wire tip.
• Copying a video without matching settings. A technique filmed on 3/8-inch plate with 0.045-inch wire and 75/25 gas will not translate to 16-gauge sheet with a different blend. See our guide to Aluminum MIG Welding: The Complete Practical Guide for how far settings shift results.

Short Note: Before blaming your hand, rule out the cheap culprits: contaminated base metal, wet or dirty wire, a blocked contact tip, a failed gas line, and a machine that is not putting out the voltage you dialed in.

Codes, WPSs and Welder Qualification

A welding procedure specification is not a suggestion. It lists essential variables, and technique is one of them: amplitude, dwell, travel speed, work angle, pass count, and whether weaving is permitted at all. AWS structural standards impose technique limits on some processes, and flux-cored work often caps weave width relative to electrode diameter. The governing standard plus the WPS overrides your preference every time.

Qualification works the same way. A welder qualified on stringer beads demonstrated that technique; a weave that deviates from the qualified procedure puts the performance qualification at risk. On inspected work, expect visual examination of toe blending and profile, plus radiographic or ultrasonic inspection on groove welds where a wide weave can hide lack of fusion at the sidewalls. High-spec sectors such as aerospace welding go further, with procedure-specific technique controls and documentation. If your WPS bans weaving, the fix is more stringer passes, not a cleverer pattern.

Safety: Longer Arc-On Time Means More Fume

Weaving slows travel and lengthens arc-on time per inch, which raises fume generation at the point of work. The baseline is OSHA’s welding, cutting and brazing standard, 29 CFR 1910.252, covering ventilation, personnel protection and fire prevention. OSHA’s welding program page and its welding fume materials stress that local exhaust ventilation at the arc is the primary engineering control.

Manganese is the exposure worth knowing. OSHA sets a permissible exposure limit of 5 mg/m³ as a ceiling for manganese fume; NIOSH recommends 1 mg/m³ as a time-weighted average and 3 mg/m³ short-term, and NIOSH’s welding program treats welding fume as a priority health concern. Longer, wider weaves push exposure toward those limits, so keep the extractor hood within a few inches of the arc. Add respirator selection, eye and skin protection, a hot work permit and a fire watch with an extinguisher wherever sparks can travel.

On stainless and aluminum, prep matters as much as technique. Residual cleaner, oil or mill scale vaporizes into the puddle; correct aluminum surface preparation before you strike an arc prevents porosity no weave can hide.

Pattern Quick Reference

Pattern	Best positions	Strongest at	Primary risk
Zigzag	Flat, horizontal, vertical-up	Filling, gap bridging, capping	Undercut if dwell too long
Crescent (C-weave)	Horizontal, vertical-up	Wrapping toes, blending	Sag on lower toe
Triangle	Overhead, vertical	Fluid puddle control	Thin sidewalls, edge burn
Figure-8	Flat, overhead	Wide caps, even heat	Centerline buildup
Circular/loop	Flat, overhead	Smooth cap appearance	Excess heat in center
Whip-and-pause	Root passes, all positions	Root fusion, out-of-position control	Humping at high speed

The Final Pre-Weld Checklist

1. Fit-up inspected, gap and land within tolerance, joint cleaned and dry.
2. WPS read; confirm weaving is permitted and note the width limit.
3. Wire, gas and polarity verified; flow rate checked with the meter, not by ear.
4. Settings dialed in on a scrap coupon of the same thickness.
5. Pattern chosen for the position, with a defined amplitude and dwell count.
6. Fume extractor at the arc; PPE on; fire watch briefed.
7. After the pass: toes fused, ripples uniform, no undercut, no rollover; log the settings.

Conclusion

Remember that bead that looked like stacked silver coins and still failed? Looking good was never the hard part. The engineering underneath is: the right pattern for the position, a width inside procedure limits, even dwell at both toes, and enough fume control to stay healthy while you weld. Next time you are in the shop, take 3/8-inch plate, snap two soapstone rails, set a zigzag amplitude of about three wire diameters, and run five passes at a steady dwell count. Photograph each one and log the settings. Two weeks of that beats two years of guessing. For the aluminum side at the same depth, continue with our full aluminum MIG welding guide.

Frequently Asked Questions

Is weaving bad in MIG welding?

Not inherently. Weaving is useful for wide joints, gap bridging, fill passes and vertical-up work, but it raises heat input and can cause lack of fusion if the weave is too wide for the current. Follow the welding procedure and keep the weave within its limits.

What is the best weave pattern for vertical-up MIG welding?

A tight zigzag or crescent weave with a clear pause at each side is the most common choice because it builds a shelf that holds the puddle. Keep the puddle small, use lower wire feed speed than flat welding, and avoid wide, slow motions.

How wide should a MIG weave bead be?

Many procedures cap weave width at roughly two to three times the wire or electrode diameter, and some codes set their own limits. Treat any number you see online as a starting point and check the WPS or code governing your job.

Should I weave or use stringers on thin sheet metal?

Stringer beads are usually safer on thin sheet because they keep heat input low and reduce burn-through and distortion. If you must widen the bead, use a very tight weave with higher travel speed and lower amperage.

Why does my weave bead have undercut at the edges?

Undercut usually means the arc spent too long at the toe without enough filler to refill the gouge, or the travel speed at the edges was too fast. Shorten the dwell slightly, reduce voltage a little, and make sure the wire is aimed at the leading edge of the puddle.

What is weave pitch in welding?

Weave pitch is the distance the arc advances along the joint during one complete side-to-side cycle. A consistent pitch produces even ripples; a changing pitch signals an inconsistent hand speed or rhythm.

Does weaving weaken a weld?

A wide, slow weave can widen the heat-affected zone, slow cooling and reduce toughness in some steels, and it can cause sidewall lack of fusion. A controlled weave within procedure limits generally produces acceptable mechanical properties.

Can I use the same weave pattern for stainless and aluminum?

No. Stainless steel is sensitive to heat input and heat tint, so tight weaves or stringers are preferred. Aluminum is usually welded with stringer beads or very tight weaves because of its high thermal conductivity and tenacious oxide layer.

How long does it take to learn a consistent weave pattern?

Most welders see clear improvement after several focused practice sessions on pad beads and simple T-joints, but consistency under different positions takes weeks of deliberate practice. Cutting and inspecting coupons speeds up the learning curve.