Tack Welding Definition: What a Tack Weld Is and How to Do It Right

You have two steel plates lined up perfectly on the table, clamps everywhere, and no third hand to hold anything while you reach for the torch. One quick zap at each end and the job suddenly becomes workable. That tiny, unglamorous weld is a tack weld — and in structural fabrication, getting its definition and its quality wrong is one of the most expensive habits a welder can develop.

What Is Tack Welding? The Official Definition and Why It Matters

Picture a fit-up table with two lengths of plate steel, a fitter holding the near corner in place while someone else reaches for the torch. Two quick arcs at the ends, the clamps come off, and the job finally behaves like a single assembly instead of two loose plates. That tiny weld is a tack weld, and almost every welded structure starts with one.

The trouble is that “a small weld that holds things together” is shop talk, not a definition. When a tack cracks or an inspector asks whether the weld in front of you is a tack, a temporary weld, or a structural weld that merely looks short, the dispute comes down to what the term actually means.

The AWS A3.0 Definition of a Tack Weld

The American Welding Society publishes AWS A3.0, Standard Welding Terms and Definitions, the reference used when a word has to mean exactly one thing. A3.0 defines a tack weld as “a weld made to hold the parts of a weldment in proper alignment until the final welds are made” (AWS A3.0 preview, AWS Publishing). TWI puts it the same way: tacks are small welds that hold components in the correct position before final welding (TWI, “What is a tack weld?”).

Read that definition slowly, because three phrases do all the work:

• Hold the parts — a tack is an alignment device, not a structural connection. It is not meant to carry load in service.
• In proper alignment — the tack’s job is geometric: position, gap, angle, squareness.
• Until the final welds are made — a tack is provisional by definition. It exists to be superseded or consumed by the weld that carries the load.

That last phrase is the one new welders miss. A tack is not a finished weld that happens to be short; it is a temporary alignment aid. Quality still matters, because it is usually left in place and becomes part of the finished joint.

Tack Weld vs. Temporary Weld: Two Different Jobs

This is where shop arguments start. In the same standard, AWS A3.0 defines a temporary weld as “a weld made to attach a piece or pieces to a weldment for temporary use in handling, shipping, or working on the weldment” (AWS A3.0). A lifting lug tacked to a base plate or a shipping tab welded to a machined surface is a temporary weld, and so is a bracing bar that keeps a frame square while it is rolled. Closely related is the construction aid weld, which attaches strongbacks or bracing that assist fit-up and control distortion.

The distinction is purpose, and purpose drives everything downstream:

• A tack weld aligns parts that are about to be permanently joined; it sits in or beside the joint that will be welded over.
• A temporary weld attaches something outside the final assembly. It gets removed, and removal is a controlled operation of its own, usually gouged or ground out and checked for base metal damage.
• A construction aid weld attaches hardware that helps the assembly hold its shape during fabrication, then comes off before delivery.

The AWS engineering forum has threads where this exact distinction gets picked apart in inspection disputes (AWS Technical Forum discussion). If your paperwork calls something a tack and the inspector calls it a temporary weld, the acceptance criteria can change with it.

Why the Precision of the Definition Matters

Tacks are everywhere, and they are not exempt from quality requirements. A cracked tack is a rejectable indication. An undersized tack lets the joint creep as weld metal cools, so the final weld goes in out of position. An oversized tack stands proud and forces over-welding to blend it in. A tack in the tension zone of a cyclically loaded member can initiate a fatigue crack, which is why codes restrict where tacks may sit on structures that see repeated loading.

In practice, the tack is the first weld an inspector sees and the last one anyone thinks about. Aerospace fabrication treats tack discipline as seriously as the final weld (Aerospace TIG Welding: Standards, Techniques, and Training), and tacking an aluminum boat hull works the same way: the tacks decide the shape of the finished platework (Aluminum Boat Welding: The Complete Guide for Strong Hull Joints).

What the Rest of This Article Covers

With the definition settled, we can get specific. The next sections cover what happens when a tiny weld pool freezes, the jobs a tack performs in a fabrication sequence, and the numbers — length, spacing, and placement — that codes and shop practice actually use. You’ll see how tacking sequence fights distortion, how to tack with MIG, TIG, stick, and flux core, and where new welders go wrong. We’ll clear up terminology confusion around temporary welds and seal passes, then finish with safety rules, inspection expectations, and a pre-tack checklist.

One thing to keep in mind as you read: these principles apply to every material, but each alloy reacts differently to the quick, concentrated heat of a tack. If aluminum is your work, the tacking and heat-input rules shift noticeably, and our guide to Aluminum MIG Welding: The Complete Practical Guide covers how tack behavior changes with heat transfer and alloy selection.

How Tack Welding Actually Works: The Metallurgy and Mechanics

A tack weld is small, but it is not a shortcut around welding metallurgy. The arc that makes a tack behaves like the arc that makes a finished bead: it concentrates intense heat into a tiny spot, melts the base metal, and lets that pool freeze in seconds. What happens in those few seconds decides whether the tack holds the joint where you set it or pulls the assembly out of square.

What the Arc Does in a Fraction of a Second

When the arc strikes, electrical energy converts to heat at the electrode tip and the workpiece. In a tack, that heat is delivered to a tiny area, so a small molten pool forms immediately, surrounded by cold, massive base metal acting as a heat sink. The pool solidifies fast, and the metal around it heats only briefly before falling back to ambient temperature.

That short thermal cycle creates a small heat-affected zone — the band of base metal that never melts but is transformed by the heat. Tack welds introduce less total heat than final welds, but the local heating rate can still be steep enough to form hard, crack-sensitive microstructures in carbon and alloy steels. TWI notes that tack welds must meet the same quality expectations as production welds because they are real welds, not decorative dots.

As the pool solidifies and cools, it contracts. That contraction is the source of tack shrinkage stress: each tack tries to pull the pieces it bridges slightly closer together. On a tight, square joint the effect is negligible. Across an open gap, a single tack placed carelessly can close the root opening on one end, kick the other end out of alignment, and create a fit-up problem you did not have before you struck the arc.

Restraint, Movement, and the Joint Gap

Mechanically, a tack weld is a restraint: it converts two freely moving pieces into one assembly with a fixed relationship, which is why tacks are alignment devices rather than structural connections. Once ends are tacked, the joint gap set with spacers or a root opening tool is locked in place, and the pieces cannot creep apart as clamps move.

The balance matters. Too few tacks and the joint moves freely as the final weld shrinks, so distortion wins. Too many, or tacks far too large, and the assembly is over-restrained: shrinkage strain has nowhere to go and concentrates at the tack toes, a classic recipe for cracking. Fabricators leave some tacks small enough to permit controlled movement, then remove or grind them before the final pass.

Why Tack Strength Follows Throat Size and Length

A tack resists load the same way a fillet weld does, so its strength is proportional to its throat size and its length. Increase the throat, by making a larger effective throat on the same arc length, and the tack carries more load. Make the tack longer without changing the throat, and capacity rises in direct proportion. Both together give the strongest tack, but neither is automatically better:

• Under-sized tacks (too short or too thin) slip and open as parts are hoisted, turned, or welded, ruining alignment.
• Over-sized tacks become hard obstacles the final pass must either consume or weld around, which usually leads to an overwelded, distorted joint.

Designers size tacks for the loads of handling and fit-up, not for service. If a tack will be left in a completed, cyclically loaded member, the engineer must approve its size and location.

Spot Welding Is Not Tack Welding

New welders confuse the two because both are small and both hold sheets together. Spot welding is a resistance process: pressure plus current fuses the interface between stacked sheets without filler metal, producing a permanent, integral joint meant to stay. Tack welding is the opposite by intent — an arc-deposited or fusion tack, usually with filler, that is temporary and expected to be consumed by, or removed before, the final weld. One is a finished joint; the other is a fixture made of metal.

In practice, a Aluminum MIG Welding: The Complete Practical Guide-style short burst on aluminum sheet gives a removable hold point on a body panel, while resistance spot welds on the same panel are permanent. Knowing which process you are running tells you whether you are building an assembly aid or a finished connection, and whether the joint needs the cleaning, preheat, and consumable control in Alloy Steel Welding Electrodes: Selection, Use & Safety for hardenable grades.

The Real Functions of a Tack Weld in Fabrication

Ask ten welders what a tack weld does and most will answer the same way: it holds parts together. That is true, but it is only one job out of six. TWI defines a tack weld as a weld made to hold parts in correct alignment before final welding, and that wording is the clue: alignment is the headline duty, and every other function flows from it. Knowing which job a specific tack is doing tells you how big to make it, where to place it, and what happens if it fails.

1. Holding Assembled Components Together

This is the function everyone sees. Cut parts land on the table as loose pieces; after tacking they behave as one unit you can handle. A typical example is a handrail bay: four posts and two rails get a tack at each corner, the clamps come off, and the frame can be rolled over to weld the far side. Without those tacks you need a helper, a pile of clamps, or both. Tacks also free up clamps that would otherwise block the final weld path, which is exactly why welders reach for them before they reach for anything else.

2. Establishing Positioning and Alignment

A tack is a positioning decision made permanent enough to survive handling. The classic shop example is a bracket welded to a machine base: set it square, tack one corner, check with a square or straightedge, adjust, then finish the tacking. Because a tack can be cut off and re-placed without scrapping the part, it works as a reversible commitment. That reversibility is why precision work, such as aerospace TIG welding, leans on small, deliberate tacks checked at every step rather than one heavy bead placed on faith.

Alignment duty continues right up to the arc. Assemblies are often tacked, then preheated before final welding, and the tacks are what hold the geometry while the metal expands. A joint that was perfectly set at room temperature can walk out of position during a 200°C preheat soak if the tacks are too few or too small to resist that thermal expansion strain.

3. Setting and Maintaining the Joint Gap

Root opening controls penetration, and gaps move as welding progresses because shrinkage pulls the joint shut. Tacks are what keep the gap you set. On a heavy plate butt joint, the fitter spaces the root with a wedge or spacer, tacks across it, and removes the spacer; from that point the tacks fight the shrinkage. When a gap still wants to close, fabricators add a bridge tack or a strongback. If the gap closes you risk lack of fusion at the root; if it opens too far you get undercut or a bead that will not tie in. The tack did not just hold the parts, it held the geometry the weld depends on.

4. Complementing or Replacing Fixtures

A steel jig is fast in production but expensive to design and build, and it only fits one part number. Tacks are free, take seconds, and adjust to whatever shape is on the table. For one-off frames, repair work, and prototypes, tacks substitute for fixtures entirely, and they permit assembly removal or repositioning that a bolted jig would never allow. In a production run you use both: the fixture sets the overall pattern, the tacks lock each sub-assembly so the unit can be lifted off the jig and moved to the weld cell. The practical test is repeatability. If you build the same part every week, build the fixture. If you build it once, tack it.

5. Resisting Component Movement and Distortion During Welding

Every weld shrinks as it cools, and that shrinkage pulls whatever is not fully restrained. Tacks are the restraint: they distribute the pulling force across the whole joint instead of letting one hot bead drag the assembly out of shape. Consider a box beam made from four plates. Tack the two end frames first, check square, then add a center tack and work inward, and the shrinkage from each pass is absorbed symmetrically. Skip the tacks and the last corner weld turns the box into a parallelogram. This is the same principle behind the fabrication techniques TWI describes for distortion control, such as backstep and stitch welding, which simply spread heat rather than dumping it in one run.

6. Temporary Strength for Hoisting, Turning, and Shipping

Sometimes a tack is the only thing standing between a heavy assembly and the floor. Massive weldment components routinely transfer weight through their tacks during handling, turning, and shipping: a rolled hopper shell tacked into a cylinder, then lifted by a crane and flipped; a cab shell tacked shut, painted, and trucked; a beam-column assembly tacked for fit-up and later rehandled twice before final weld. In these cases tack size is a safety calculation, not a habit. Small tacks that were fine for holding alignment on the table may crack under load, and a failed tack during a lift drops a load on someone. Heavy plate here is often tacked with stick electrode for its forgiving arc; if you want the reasoning behind that choice, the advantages of stick welding cover it. The definition itself stays deliberately narrow, and AWS A3.0 ties the term to alignment rather than load, which is exactly why anything expected to carry weight must be treated as a temporary weld, sized deliberately, and approved by the engineer before the lift.

Put the six functions together and the pattern is clear: a tack weld is a multitasking device that holds, positions, spaces, restrains, and sometimes carries, and each role demands a different size and placement. Read the job before you strike the arc, decide which function that tack is performing, and the size and spacing rules in the next sections stop feeling like arbitrary numbers and start feeling like common sense. Get that read right and the rest of the joint follows, including the cleanup: any tack cut off or burned out leaves disturbed metal that must be ground and cleaned before the final pass, and on aluminum that means cleaning the surface properly all over again.

Tack Weld Size, Length, and Spacing: Practical Rules and Code Guidance

Ask ten welders how long a tack should be and you will get twelve confident answers. Some of that confusion is legitimate: the AWS A3.0 welding terminology standard defines exactly what a tack weld is, but it hands you no dimension. Sizing tacks is a blend of code requirements, engineering approval, and shop rules of thumb that have survived decades of real fabrication. Here are the numbers people actually use, and the code language that governs them.

Typical Tack Weld Lengths by Joint Type

A tack only needs enough fused throat to resist the handling forces acting on the joint: clamping spring, gravity, thermal movement, and the pull of the cooling final weld. Beyond that, extra length buys nothing except more heat, more shrinkage, and more grinding later. As TWI summarizes it, tack size is normally determined by material thickness and restraint rather than by a fixed formula.

• Butt joints in light plate (up to about 1/4 in. / 6 mm): 1/2 in. to 3/4 in. tacks at each end, with additional tacks along the seam as thickness increases.
• Heavy plate (1/2 in. / 12 mm and heavier): 3/4 in. to 1-1/2 in. tacks, because fit-up forces and restraint are much higher.
• Pipe and round joints: tacks roughly as long as the thinner wall thickness, with two opposing tacks at 12 and 6 o’clock to keep the root opening even, adding more as pipe diameter grows.
• Corners, flanges, and edges: short tacks, 1/2 in. or less, placed at the ends and wherever the fit is fighting you.

The practical rule: a tack should be long enough that the joint cannot slip, and no longer. An overlong tack shrinks like a miniature weld bead and can pull the very alignment you were trying to lock in.

Spacing: How Far Apart to Place Tacks

Spacing tracks material thickness and how hard the joint wants to move. Thinner material flexes, so it needs more frequent tacks to hold a consistent root gap, while stiffer plate can carry a wider span. These ranges are common shop practice, echoed in shipbuilding and structural training material such as the NSRP/Ingalls weld curriculum guidance:

Material Thickness	Typical Tack Length	Typical Tack Spacing
Under 1/8 in. (3 mm)	1/2 in. (12 mm)	1 to 2 in. (25 to 50 mm)
1/8 to 1/4 in. (3 to 6 mm)	1/2 to 3/4 in. (12 to 19 mm)	2 to 4 in. (50 to 100 mm)
1/4 to 1/2 in. (6 to 12 mm)	3/4 in. (19 mm)	4 to 6 in. (100 to 150 mm)
Over 1/2 in. (12 mm)	3/4 to 1-1/2 in. (19 to 38 mm)	4 to 6 in. (100 to 150 mm), closer at ends
Pipe (all diameters)	About the thinner wall thickness	Two opposing tacks minimum; more on large pipe

Always tack the ends and corners first, regardless of what the table says. Ends are where gaps open and where alignment is judged. On thin aluminum, spacing tightens quickly because the material moves with very little heat: many welders tack every 1 to 2 in. and let the joint cool between tacks. Fit-up cleanliness matters just as much as tack pattern, and aluminum MIG welding setups generally reward closer spacing than steel does, since the joint is less rigid and the surface oxide layer makes poor fusion easier to create.

What AWS D1.1 Actually Says About Tacks

Two clauses in AWS D1.1 shape almost every structural tack decision:

• D1.1 Section 5.18 states that tack welds and temporary welds are subject to the same quality and WPS requirements as final welds: qualified procedures, matching filler metal, the same preheat and interpass temperature, the same visual acceptance criteria, and the same welder qualification. A tack is not a loophole.
• D1.1 Section 5.18.5 limits fillet tack welds that will be incorporated into submerged arc welds to 3/8 in. [10 mm]. Because SAW deposits a heavy bead that remelts the tacks underneath, a larger tack will not be fully consumed, leaving an unremelted, discontinuity-prone zone inside the finished weld.

Notice what D1.1 does not do: it sets no maximum or minimum tack length for general structural work. Length is left to the fabricator, guided by the WPS and engineering design, which is why shop practices vary so widely. Welders discussing this in the AWS technical forum tend to reach the same conclusion: follow your qualified procedure first, then fall back on thickness-based rules of thumb.

When Placement or Removal Needs Engineer Approval

Some tack situations are not the welder’s call. Under D1.1 practice, tacks left in place on members subject to cyclic loading or in tension zones normally require evaluation by the engineer, because a tack is a full-strength discontinuity sitting in the final load path. Engineers also sign off when tacks become part of the completed weld rather than being covered by a final pass, when tacks attach temporary items, and when tacks in pipe or pressure-containing welds are permitted to remain. The AWS D1.1 design examples from the James F. Lincoln Foundation reinforce the underlying principle: tacks are welds, and anything treated as a weld is treated as a weld for both design and inspection.

If your procedure does not explicitly cover a tack, ask before you strike the arc. Removing a tack later is cheap. Explaining an unapproved, cracked tack in a weld map during inspection is not.

One last shop-level note: when tacks do need to come out, grind them flush with a clean grinding wheel, not a contaminated flap disc loaded with stainless or aluminum residue, and inspect the base metal underneath before the final pass. A tack that was cracked or cold at installation usually leaves a scar in the root that the final bead will not hide, and proper surface cleaning before re-tacking gives the replacement tack a far better chance of genuine fusion.

Tack Welding Sequence and Distortion Control

Distortion is a heat problem before it’s a weld problem. Metal expands as it’s heated and contracts as it cools, and a shrinking bead of weld metal will always pull against an unrestrained plate. Tack welds are the first form of restraint, as they decide how much movement is allowed while that contraction takes place. The right tack sequence puts that movement to work. The wrong one will have a joint shifting before the root pass is finished.

Not Enough Tacks Creates Runaway Movement

Most fit-up failures come down to a lack of restraint. As TWI’s guidance on tack welds notes, too few tacks can allow a joint to progressively close as welding proceeds. On long seams run with MMA or MIG, this can be severe enough for the edges to overlap, closing the root and preventing penetration.

The opposite can occur with submerged arc welding. With its high current, fluid weld pool, and heavy flux restraint, an inadequately tacked joint can be driven open as the bead solidifies behind it. Instead of a controlled root opening, the gap becomes larger than intended with no room to correct it.

Proven Tacking Sequences

To counter these effects, TWI’s distortion control job knowledge recommends three sequences that cover most fabrication work:

1. Tack straight through. Starting at one end and placing tacks to the opposite end. This is quick, but any gap error will be pushed ahead of the tacks and concentrated at the far end of the seam.
2. Tack one end, then backstep. Anchor a fixed end first, then place each subsequent tack behind the previous one towards the middle. Each tack’s shrinkage pulls against an already restrained point, slowing the build-up of accumulated shrinkage.
3. Tack the center, then backstep outward. Anchor the middle first to split the seam into two halves. From there, backstep towards each end independently, keeping shrinkage balanced along the full length. This is the most reliable option for long seams.

Joint situation	Preferred sequence	Why
Short butt joint, tight fit-up	Ends, then center	Locks overall length whilst closing any small central gap last.
Long seam (plate or pipe)	Center first, backstep outward	Splits accumulated shrinkage into two manageable halves.
Gap that must stay open (SAW root)	Even spacing, more tacks	Provides enough restraint to prevent the joint from drawing closed.
Symmetric part (box, frame)	Balanced, alternating sides	Allows shrinkage to cancel across the neutral axis.

Controlling the Gap With Direction

Tack shrinkage is directional. If a gap is too wide at one end, placing the first tack at the largest opening and working towards the tighter end will pull the plates together as each tack cools. If the gap has closed too far, reversing that direction can help push it back open. For this reason, experienced welders check the gap after every two or three tacks, not just at the initial fit-up.

Balanced Restraint and Back-to-Back Assembly

On symmetric fabrications, restraint must be balanced. One of the most effective methods is back-to-back assembly: clamp two matching parts face to face, tack and weld them as a pair, then separate them. The shrinkage from each side counteracts the other, leaving both parts closer to flat. The same idea applies without a full back-to-back fixture. For parts such as box sections, brackets, or aluminum boat hull panels, alternate tacks from side to side as work progresses. With long aluminum seams, where heat builds rapidly, this balance is especially important, a practice that is also covered within our aluminum MIG welding guide.

However, restraint is a trade-off. Too much of it, especially from heavy fixturing combined with large tacks, leaves nowhere for shrinkage to go and can buckle the plate between tacks or crack the tacks themselves. Too little allows drift. The goal is controlled movement: enough short, evenly spaced tacks to hold alignment, without locking the assembly rigid. Any cracked tack should be removed and replaced before continuing.

Stitch and Backstep Welding as Support

Tacking sets the stage, but travel technique keeps distortion in check after it. Stitch welding lays short segments with deliberate gaps between them, allowing each to cool before the next begins. Backstep welding runs short beads against the direction of travel, so each increment partially cancels the contraction of the last. These methods don’t replace a correct tack sequence, they rely on one. With well placed tacks and controlled heat input, both become far more effective at holding the joint true through to completion.

Techniques by Process: Tack Welding with MIG, TIG, Stick, and Beyond

The AWS definition of a tack weld never mentions a process, and that is deliberate. You can tack with almost any arc process. What changes from one process to the next is trigger discipline, heat control, cleanup, and how much allowance you leave yourself before the final pass. In practice, a tack is usually made with the same process as the final weld, just at lower power, so the tack behaves like a miniature version of the weld that follows it.

MIG (GMAW): Fast Tacks with Careful Trigger Control

MIG is the fastest tacking process in most shops, which is exactly why it produces the sloppiest tacks. A trigger held too long dumps filler into a joint that only needed a dot of metal. Typical tacking trigger times run roughly 0.5 to 1.5 seconds depending on material thickness: thin sheet sits at the short end, heavy plate toward the long end. Set voltage and wire feed speed to your final weld settings and let duration do all the work.

Stickout matters more than most welders realize. Keeping wire stickout around 1/2 inch with a tight contact-tip-to-work distance gives you far better arc control on thin material, because the arc stays pointed where you aimed instead of wandering as the wire feeds in. For a corner or lap tack, place the wire right at the joint intersection, pull the trigger, release, and hold position for a beat so the crater fills before you move.

Then let the part cool. Tacking a whole assembly in quick succession stacks heat and shrinks the joint incrementally, so a short pause between tacks keeps the gap you set from closing up. Clean, dry base metal is non-negotiable, especially on aluminum, where oxide and moisture cause porosity in the tack and then in the weld over it. Our aluminum cleaning prep guide covers the process, and aluminum MIG welding technique has more on short-burst control on thin sections.

TIG (GTAW): Fusion Tacks Versus Filled Tacks

TIG gives you two tacking options, and choosing between them is a fit-up decision. A fusion tack is filler-free: you strike the arc, let the edges melt and pool together, taper the current, and stop. Fusion tacks are ideal before final fit-up because they are small, low-profile, easy to break or re-melt, and they leave no extra metal to grind flush. A filled tack uses a dab of filler and is the right call on edges that would otherwise melt back, on thicker sections needing a bit of throat, or whenever the tack will stay in the finished weld.

Current tapering is the skill that separates a clean TIG tack from a crater crack. Add filler if you are filling, then back off the pedal while the puddle is still fluid so the crater solidifies under shrinking pressure. For tight tolerance work such as aerospace assemblies, where tacks are held to the same standard as production welds, see our notes on aerospace TIG welding standards and training.

Stick (SMAW): Electrode Angle, Arc Strikes, and Slag

Stick tacking is clumsy by comparison but perfectly serviceable. Use a slightly steeper work angle than you would for a full bead so the tack stays short, and keep the arc short to avoid blowing a hole in thin material. Avoid striking arcs on the visible face of the base metal; arc strikes create hard spots and stress raisers that inspectors flag. Strike on the joint edge, on a run-on plate, or on scrap and carry the arc in.

The one rule that causes rework more than any other: chip the slag out completely before final welding. A tack with flux trapped against it is a cold-lap inclusion waiting to happen. Chip, wire brush, and check the tack edges before you strike again. If you are still deciding whether stick belongs in your workflow at all, our advantages of stick welding piece lays out where SMAW still outperforms wire processes.

Flux Core and Submerged Arc Considerations

Flux-cored tacking behaves like stick with a machine: same slag discipline, same removal requirement before the final pass, plus wire-feed settings dialed back so the short burst does not pile up metal. On aluminum work the shielding and cleaning rules tighten considerably, which our aluminum flux core welding guide explains in detail.

Submerged arc is the special case. SAW cannot tack in open air the way other processes do, so tacks on assemblies destined for submerged arc are usually placed by another process, or by SAW itself under conditions that will not compromise the cover. Codes address this directly: fillet tack welds incorporated into SAW welds must not produce objectionable changes in the appearance of the weld surface, a requirement tied to AWS D1.1 practice discussed in AWS technical forum guidance on tack welds. In plain terms, your tack must melt into the finished bead cleanly, with no visible lump, dip, or pocket of porosity showing through the cap.

Specialized Processes: Electron Beam, Laser, and Resistance

Beyond the shop floor staples, precision fabrication uses laser welding systems, electron beam welding, and resistance spot processes for tacking. These deliver extremely concentrated, low-distortion tacks in small heat-affected zones, which is why thin-wall, aerospace, and medical assemblies favor them. They also require fixturing and shielding that most general fabricators do not have, so they stay out of everyday work. Note that resistance spot welds are not tack welds at all, as the next section clarifies.

Process	Tack method	Watch out for
MIG (GMAW)	0.5–1.5 s trigger burst	Long stickout, stacked heat
TIG (GTAW)	Fusion or filled tack	Crater cracks, no taper
Stick (SMAW)	Short arc, steep angle	Slag left in joint, arc strikes
FCAW	Reduced wire speed burst	Slag entrapment
SAW	By secondary process	Surface appearance of finished bead

The pattern across every process is the same: short duration, controlled heat, clean metal, and a tack small enough that the final weld can absorb it without a trace. Get that habit right and the rest of the joint takes care of itself.

Common Tack Welding Mistakes and How to Fix Them

Every shop has a version of the same story: a tacked column splice is lifted by the crane, and halfway up a loud pop echoes across the yard. The investigation always ends the same place. The tack looked fine, felt solid under the hammer, and never actually fused. Nearly every tack failure traces back to a simple, well-understood error with an equally simple fix.

1. Cold Tacks That Never Fused

The cold tack is the most common defect in the category. It has bead shape and crown, and it passes a quick glance, but there is no fusion at the root. Under real stress it pops off clean, leaving two shiny surfaces. Causes in MIG work are travel speed too fast for the wire setting, excessive stickout, dirty base metal, or triggering the gun before the arc stabilizes. In stick welding it happens when the arc is struck and broken too quickly or the electrode angle throws energy sideways.

• Increase wire feed speed slightly so the arc wets into the base metal before you break off.
• Shorten your stickout. For short-circuit MIG on solid wire, keep it around 3/8 inch to 1/2 inch; long stickout drops current where you need it most.
• Clean the joint. Mill scale, oil, paint, and moisture rob heat. For aluminum, prep is non-negotiable — see aluminum cleaner for welding.
• Hold the trigger a beat longer. Let the puddle form and wet the toes before releasing.

2. Cracked Tacks

Cracks hide under slag, spatter, or the crown. They form when a tack on restrained thick material cools too fast, when a crater tears on shrinkage, or when hydrogen from moisture migrates in during cooling. The code position is unambiguous: cracked or otherwise improper tacks must be removed before welding. You cannot weld over a crack and hope the final pass buries it, because a crack at the root of a tack becomes a crack at the root of the finished weld.

• Preheat thick or highly restrained sections to slow cooling and let hydrogen escape.
• Use low-hydrogen electrodes and keep them dry — damp 7018 is a crack factory.
• Grind any cracked tack completely to sound metal, then re-tack. Never feather and cover.

3. Undersized and Oversized Tacks

An undersized tack is a restraint failure. The tack itself is sound, but too small in throat or length to resist the forces on the joint, so the assembly shifts and the part that was aligned on the table is out of tolerance by final welding. Size tacks to the job, not to habit: thicker plate, longer spans, and heavier assemblies need longer, thicker tacks sized to carry handling and thermal loads until the final weld takes over.

The opposite error is just as costly. A tack that is too long or heavy becomes part of the final weld whether you want it or not, producing a lumpy bead with hard spots, inconsistent penetration, and inconsistent appearance. It also acts as a heat sink at the exact point where you need fusion. Match tack size to joint size, and reduce or remove excessive tacks before final welding rather than trying to burn them in. This matters most in remelting processes: the AWS D1.1 requirement that multipass tacks have cascaded ends or otherwise be prepared for incorporation exists because a blunt, cold tack end can create a defect when the final pass remelts it.

4. Arc Strikes Left on Base Metal

An arc strike is any place the electrode touches the work outside the weld zone, usually from striking an arc a few inches away and walking into position. On carbon steel it may be cosmetic; on pressure vessels, alloy steels, and cyclically loaded structures it creates a small, rapidly quenched hard zone whose localized hardening and residual stress can affect future structural integrity. Removing the tack does not remove the strike.

• Strike the arc inside the joint, on the tack location itself.
• Use run-on and run-off tabs when the starting position is awkward.
• If a strike occurs, grind it out and have it inspected — do not paint over it.
• Train the habit early; welders who start arcs on scrap carry that behavior into structural work.

Aluminum is especially sensitive because of its oxide layer and thermal conductivity, and the principles in this aluminum MIG welding guide apply directly to how you strike and finish a tack.

5. Crater Cracks and Trapped Contamination

Every weld ends in a crater, and a crater is a shrinkage cavity waiting to crack. If you simply cut the arc, the crater tip becomes a stress riser and cracks within seconds in a restrained joint. Fix it by backstepping: reverse direction briefly and add filler as you withdraw, or circle the puddle back over the crater before breaking off. TIG tacks are the most exposed because they often end without filler; on restrained joints, add a small dab at the end of every tack.

A tack left dirty is a tack welded over dirty. Trapped slag becomes lack of fusion and slag inclusion; spatter prevents the next pass from wetting in. Chip and brush every tack before the next operation, remove slag completely at the tack roots, and apply anti-spatter before welding, never on the joint face. On aluminum, removal technique changes by process — this aluminum flux core welding breakdown explains what actually works.

6. Tacks Placed in Tension Zones of Cyclically Loaded Members

This mistake has the most serious long-term consequences. In fatigue or cyclic service, a tack in a tension zone is a stress concentration that can initiate a crack even after final welding. AWS D1.1 prohibits tack welds not incorporated into the final weld from being placed in tension zones of cyclically loaded structures, except where other provisions specifically permit it.

• If tacks will be removed from a cyclically loaded structure, keep them out of tension regions.
• If placement is unavoidable, get engineering direction in writing first.
• Incorporated tacks must be clean, sound, and compatible so the final weld absorbs them without defect.
• Document everything — inspectors look for evidence of compliance, not just visual quality.

For teams working to ISO frameworks, ISO/TR 17671-1 complements the AWS approach with heat input and quality expectations affecting tack placement and removal.

A Quick Reference Table of Defects and Fixes

Defect	Root Cause	Fix
Cold tack (no fusion)	Low heat, long stickout, dirty metal	Raise wire speed, shorten stickout, clean, hold arc longer
Cracked tack	Fast cooling, hydrogen, restraint	Preheat, low-hydrogen consumables, remove and re-tack
Undersized tack	Sized for convenience, not load	Match tack size to joint and handling loads
Oversized tack	Habit or speed	Trim or remove before final weld
Arc strike	Striking outside joint	Start in the joint, grind out strikes, inspect
Crater crack	Arc broken without filling crater	Backstep and fill before breaking arc
Slag/spatter	No cleaning between tacks	Chip, brush, and clean before next pass
Tack in tension zone	Ignoring cyclic loading	Follow AWS D1.1, get approval, document

Building Good Habits Into Every Tack

Most of these mistakes disappear with two changes: better heat control and better cleaning. Everything else is discipline. A welder who confirms the arc is actually fusing, fills every crater, cleans between tacks, and knows where a tack is allowed to go will produce tacks that pass inspection and survive real service loads. Treat every tack as if it matters, because the quality of the tack is what keeps a structure standing. When a tack is sound, you will not hear the pop.

Tack Welding vs Temporary Welds, Spot Welds, and Backing

Few things start an argument on the shop floor faster than two people using the same word for different things. “It’s just a tack.” “That’s a temporary weld.” “It’s basically backing.” Usually both welders are describing the same bead, and the disagreement only surfaces when an inspector writes it up. The fix is not better welding, it is better vocabulary.

Tack Weld vs Temporary Weld: The Distinction That Matters

Both are small, both are quick, and both may eventually come off. The difference is what they attach. A tack weld joins parts that belong to the weldment itself — the two plates, the pipe segments, the stiffener and the shell. A temporary weld attaches something foreign to the weldment: a lifting lug, a handle, a clamp tab, an alignment bracket, a grounding tab. That foreign piece is an accessory, not a component.

The same bead can therefore be described correctly two ways. Weld a temporary lifting lug onto a beam and it is a temporary weld; weld the identical lug onto a beam you intend to keep forever, and it stops being temporary — it becomes a permanent attachment, subject to the full welding procedure and inspection requirements for that joint. The geometry never changed, the intent did. As TWI’s technical FAQ on tack welding puts it, tack welds are temporary in purpose but are still real welds, with real fusion and real consequences.

Question	Tack Weld	Temporary Weld
What does it join?	Parts of the weldment	Something foreign to the weldment
Typical example	Holding a corner joint in position	Welding on a lifting lug or handle
Removed later?	Usually absorbed into the final weld	Yes, the attached item is cut off
Left in place?	Often, if it fuses into the final bead	Only if the attachment is permanent

Spot welding belongs to a different family entirely. Resistance spot welding joins overlapping sheets at discrete points using pressure and heat generated by electrical resistance, with no arc and usually no filler. A “tack” made with a MIG gun is an arc weld, not a spot weld, and calling it one in a procedure document will get your paperwork bounced. If you are tacking thin aluminum sheet, the fusion-tack technique in Aluminum MIG Welding: The Complete Practical Guide is the practical reference.

Backing Is Not a Tack

Backing answers a different problem. AWS A3.0 defines backing as material or a device placed against the back side of the joint to support the weld and shield molten weld metal from the atmosphere. A backing strip, a ceramic backing, a consumable insert — none of them hold the parts in alignment. They shape and protect the root pass. Remove the backing and the joint is still positioned; remove the tacks and it falls apart.

The “Continual Tack” and the Seal Pass

Welders sometimes describe a short, intermittent run laid along a long joint as a “continual tack.” Strictly, the moment a weld is intended to remain and carry load it is no longer a tack — it is production welding performed in short segments, which is what stitch welding actually is. A seal pass is similar: a light continuous bead that closes a joint against leakage or corrosion. Neither is a tack by definition, and inspectors will treat them accordingly.

When a Tack Stops Being a Tack: The SAW Exception

Codes make one careful allowance. In submerged arc welding, tack welds left in the joint are fully remelted by the deep-penetrating SAW pass, so the finished weld metal absorbs them and their chemistry and soundness become part of the joint. AWS discussion on tack weld acceptance reflects the same logic: if the tack is completely consumed, it does not need separate acceptance criteria. AWS D1.5 extends this remelting exception beyond SAW to electroslag and electrogas welding, where the pool likewise reworks the entire joint volume.

The practical test is simple: will the final pass completely remelt this tack? If yes, it can stay and be judged as part of the finished weld. If any part of the tack survives unmelted — as it typically does with MIG, TIG, or stick — it is a separate weld, it must meet tack quality requirements, and in tension zones of cyclically loaded members it may need to be removed and repaired.

Tack Weld Safety, Codes, and Inspection Requirements

Tack welding is usually treated as prep work, the unglamorous stuff you knock out before the real welding starts. Regulators and inspectors do not see it that way. A tack uses the same arc, the same fumes, and the same molten pool as any production pass, so every rule that applies to your root pass applies to your tacks. The catch is that tacks get placed in a hurry, in awkward positions, with less attention than a full bead. That is when people get burned, fire watch gets skipped, and a cracked tack gets buried under a final pass nobody re-examines.

OSHA’s Rules Apply to the Tack, Not Just the Final Weld

Under OSHA 29 CFR 1910.252, the general industry standard for welding, cutting, and brazing, the fire prevention requirements are blunt. Combustible material within 35 feet of the point of operation must be removed or protected with flameproof covers, and where appreciable combustible material cannot be moved, a fire watch must be posted. That fire watch stays on duty for at least 30 minutes after welding or cutting stops, watching for smoldering in insulation, flooring, or wall cavities. Two quick tacks on an assembly near cardboard packaging and scrap bins trigger exactly the same obligation as a full day of production welding.

Ventilation rules apply on the same footing. For spaces holding less than 10,000 cubic feet per welder, or rooms with ceilings under 16 feet, OSHA requires mechanical ventilation of at least 2,000 cubic feet of air per minute per welder, or an equally effective method such as local exhaust at the arc. The full text sits in OSHA’s Subpart Q regulations.

Eye Protection and Fume Exposure

OSHA 1910.133 requires eye and face protection selected for the specific process and current. In OSHA’s shade recommendations, shielded metal arc welding with 1/16 to 5/32 inch electrodes lists shade 10, while larger electrodes and higher currents call for darker filters. Wire processes such as stick welding and GMAW generally land in the shade 10 to 14 range depending on amperage. The temptation on tacks is real: a welder flips the hood for a one-second zap, or looks away and tacks on instinct. That habit causes arc flash burns to the cornea. Auto-darkening helmets rated for the process, with the shade matched to the current, remove most of the temptation.

The AWS Z49.1 safety standard, Safety in Welding, Cutting, and Allied Processes, is freely available as a free download from AWS and expands on OSHA with guidance on PPE, fume control, confined spaces, and cylinder handling. Fume exposure deserves particular attention because tacks are often placed in poor positions where the welder’s head sits directly in the plume. Alloyed consumables, including stainless and high-alloy stick electrodes, produce fumes with different toxicity profiles than plain mild steel, so check the electrode manufacturer’s safety data before selecting consumables for alloy steel applications.

What AWS D1.1 and ISO/TR 17671-1 Expect

On the quality side, tack welds carry code obligations that catch many fabricators off guard. Under AWS D1.1 Section 5.18, tack welds are treated as real welds: they must be deposited by welders qualified for the applicable process using a qualified procedure, and tacks that will not be incorporated into the final weld must still satisfy visual inspection before the member is accepted for welding. A rejected tack does not go away because it is small. It gets repaired or removed and replaced before the final pass covers it.

ISO/TR 17671-1 states that tack welds should be applied in a balanced sequence to minimize distortion, and that any tacks not incorporated into the final weld should be removed. Both expectations matter to inspection. An unbalanced tacking pattern leaves the joint locked in a distorted position, and a leftover tack sitting outside the finished weld becomes an unexamined piece of weld metal in a load-carrying member.

Terminology matters just as much in the paperwork. AWS A3.0, the standard welding terms and definitions document, gives inspectors and welders a shared vocabulary so that a note reading “tack” on a traveler does not get argued into three different meanings on the shop floor.

What an Inspector Looks For in a Tack

When an inspector walks a fit-up, the tacks are not a formality. Expect these checks:

• Fusion and profile. Tacks should be fused at both toes, not sitting on the surface as a cold lump.
• Cracks and craters. Crater cracks are the most common tack defect; inspectors reject them visually and may call for magnetic particle or penetrant testing in critical applications.
• Size and location. Undersized tacks that let the joint move, and oversized tacks that force an overbuilt final pass, both draw comments. Tacks in tension zones of cyclically loaded members draw the sharpest objections.
• Removal quality. Tacks that must come out should be removed without gouging the base metal, then ground smooth. A removal gouge is a stress raiser just as bad as the tack it replaced.
• Arc strikes. Stray strikes outside the joint are base metal damage and are rejected under D1.1 visual criteria.

Why so strict? Because a tack left in a member becomes part of that member. A cracked or unfused tack in a load path is a crack initiator, and the final bead laid over it does not reliably heal the defect beneath. High-stakes industries such as aerospace welding treat tack discipline as a core inspection checkpoint for exactly this reason.

Turning Compliance Into Shop Habits

Before striking an arc for tacking, confirm the fire watch is assigned, ventilation is running, your filter shade matches the current, and the fit-up surfaces are clean. After tacking, walk the joint: check each tack for cracks and fusion, confirm nothing landed outside the joint, and remove or flag any tack that will not be incorporated. Done that way, safety and inspection stop being separate events bolted onto the end of the job. They become part of the tacking motion itself, which is exactly where they belong.

Practical Tips, Quality Checks, and Final Takeaways

After going through every detail of tacking, it’s time to bring it all back down to where it matters most: the shop floor. Going back to the opening scenario, those two plates that seemed impossible to hold by hand were made workable by nothing more than a few, properly placed tacks. Those miniscule welds bought you time, control, and accuracy. With that being said, the difference between a clean fit-up and a distorted, rejected joint almost always comes down to the discipline behind those tacks.

Pre-Tack Checklist and Cleaning Requirements

Before striking an arc for a single tack, run a quick check over your joint. The most consistent way to do this is with a pre-tack checklist to prevent avoidable mistakes:

• Verify fit-up: Check for proper joint alignment and your required root gap. Even a slight misalignment now will be exponentially harder to correct later.
• Clean the joint: Remove any mill scale, rust, oil, grease, or moisture from the faying surfaces. For more sensitive metals, such as those that would benefit from using a specialized aluminum cleaner for welding, extra care is needed to remove oxides before tacking.
• Confirm your filler: Tacks that are to be incorporated into the final weld must be made with electrodes meeting final weld requirements and cleaned before incorporation. This is non-negotiable for code compliant work.
• Check your environment: Be aware of fire hazards and ventilation. Any arc operation, including tacking, falls under the fire prevention requirements of OSHA 1910.252 as well as the guidance found within AWS Z49.1.

Size, Cooling, and Restraint

When it comes to tacking, less truly is more. Minimizing tack size to about 1/16 inch below the final weld size helps eliminate overwelding and reduces welding-induced distortion. A good quality tack of proper size generally does not require grinding. Allow your tacks to cool at a natural rate whenever possible, avoiding any forced, rapid cooling methods that could introduce unnecessary stress into the joint.

Now, the question of restraint always arises. Fixtures are your best friend when you have repetitive, high precision work or are dealing with thin sheet metal that is prone to warping. However, don’t lean on clamps alone as a crutch for poor fit-up. The correct approach is to utilize fixtures to compliment your tacks, not replace the need for correct placement and sequence. When working with more demanding filler selections, such as those covered within an alloy steel welding electrodes guide, the strength of your small tacks becomes all the more vital to hold that high strength joint in place.

Removing Tacks Without Damaging The Base Metal

There will be times where a tack cannot, or should not, be incorporated into the final weld. In these cases, removal of tacks needs careful control to avoid causing defects in the component surface. The safest methods are typically a controlled chipping process with a sharp chisel, or carbon arc gouging used with extreme precision. Avoid reckless grinding that thins the parent metal adjacent to the joint. After removal, inspect the spot thoroughly. Any gouges, notches, or undercuts left behind can act as stress risers and will need to be blended or evaluated against your acceptance criteria, such as those referenced in ISO/TR 17671-1.

Training New Welders For Tack Quality

Tacking is one of the first skills to teach, and one of the easiest to overlook. For apprentices, don’t have them practice tacking by making them as big as possible. Have them practice restraint. Utilize coupons, teach them to tack opposite ends first, and critique fusion over length. For processes that are less forgiving, such as SMAW, teaching correct manipulation from resources like the all-position welding rod guide or going over the practicality of the advantages of stick welding will allow them to place precise, easily removable tacks with far more confidence. Above all, teach them to inspect their own tacks for fusion. If you can see it holding under light pressure and it fused into both members, you’re on the right track.

Final Takeaways

At its core, a tack weld is far more than a “quick zap”. As defined by TWI, it is a temporary weld used purely to locate and hold components in place until they can be permanently joined. When done right, it is barely noticeable. When done wrong, it can pull an entire assembly out of square, crack, or fail inspection.

Keep it fused, keep it small, keep it clean, and only incorporate it when it meets final weld requirements. Remember the plates from the beginning of this article. They didn’t need to be fully welded to become workable, they just needed a couple of accurate tacks to give the welder control back. That is the true purpose of tacking. Master those tiny welds, and you’ll master the control that separates a rushed fit-up, from a precision fabricated joint. That is a skill that will never go out of style on any shop floor.

Frequently Asked Questions

What is the definition of a tack weld?

AWS A3.0 defines a tack weld as a weld made to hold the parts of a weldment in proper alignment until the final welds are made. It is a temporary alignment weld, not a structural connection, though it must still meet the same quality requirements as the final weld when code applies.

How long should a tack weld be?

Most codes do not set a universal length. In practice, butt joint tacks commonly run 1/2 to 3/4 inch, tee joint tacks 1/2 to 1 inch, and heavy plate over 1 inch may use 3/4 to 1-1/2 inch. AWS D1.1 caps fillet tacks incorporated into submerged arc welds at 3/8 inch.

How far apart should tack welds be spaced?

Spacing depends on material thickness and joint type. Thin material under 1/8 inch may need tacks every 1 to 2 inches, medium 1/8 to 1/4 inch material every 2 to 3 inches, and plate over 1/2 inch every 4 to 6 inches. Pipe tacks are usually placed closer together than plate tacks.

Is a tack weld as strong as a final weld?

No, a tack weld is deliberately small and is not intended to carry the full service load. However, the quality standard is not lower: AWS D1.1 subjects tack welds to the same quality requirements as final welds, because a defective tack can initiate cracks that propagate into the finished structure.

What is the difference between a tack weld and a temporary weld?

A tack weld joins the parts of the weldment itself to hold alignment. A temporary weld attaches something foreign to the weldment, such as a lifting lug or strongback, for handling, shipping, or working on the assembly. The same physical weld can be classified either way depending on whether the attached piece stays.

Can you leave tack welds in the finished work?

Yes, if the tack is incorporated into the final weld and meets the applicable requirements. Tacks not incorporated into the final weld must be removed when required by the engineer, and AWS D1.1 prohibits unincorporated tacks in the tension zones of cyclically loaded structures except as specifically permitted.

Why do my tack welds keep cracking?

Cracks in tacks usually come from insufficient size for the load being carried, contamination, rapid cooling that produces hard heat-affected zones, or crater cracks at the end of the tack. Preheating thicker sections, increasing tack size modestly, cleaning the joint, and tapering the end of the tack all help.

Do tack welds need to be cleaned before the final weld?

Yes. AWS D1.1 5.18.4 requires tacks incorporated into final welds to be cleaned prior to incorporation, and ISO/TR 17671-1 says the same. Slag, spatter, oxides, and porosity left on a tack can be trapped inside the finished weld.

Is there an official tack weld symbol on a drawing?

No. There is no separate tack weld symbol in standard welding symbology. Tacking requirements are normally described in the weld procedure specification, drawings notes, or a general fabrication note rather than with a unique symbol.

Do you need a qualified procedure and qualified welders to tack?

Yes, for code work. AWS D1.1 5.18.1 requires tack welds and construction aid welds to be made with a qualified or prequalified WPS and by qualified personnel. Treating a tack as exempt from procedure requirements is one of the most common and costly errors in fabrication.