Tack Weld 101: Definition, Rules, Sizing, Sequence & Common Mistakes

A fabricator once spent two full days welding a 12-foot steel frame, only to watch the final bead pull the entire assembly 6 mm out of square. The welds were perfect — but the tack welds that were supposed to hold everything in alignment were placed in a rush, from one end straight to the other. The problem was never the welder’s hand; it was the tacks nobody thought twice about.

What Is a Tack Weld? The Definition Everyone Quotes (and Few Read Closely)

A fabricator once spent two full days welding a 12-foot steel frame, only to watch the final bead pull the entire assembly 6 mm out of square. The welds were perfect — but the tack welds that were supposed to hold everything in alignment were placed in a rush, from one end straight to the other. The problem was never the welder’s hand; it was the tacks nobody thought twice about. This is exactly why understanding the true definition of a tack weld is the most important first step to quality fabrication.

The Official Definition: What AWS A3.0 Actually Says

When most welders are asked, “what is a tack weld?”, they will respond with something along the lines of, “a small temporary weld.” While that may be true in casual conversation, the technical answer comes straight from the industry standard. According to the AWS A3.0 Standard Terms and Definitions, a tack weld is defined as, “a weld made to hold the parts of a weldment in proper alignment until the final welds are made.”

This definition is incredibly intentional. It does not mention the weld’s length, throat size, bead profile, or even the process used to create it. Instead, it solely focuses on its function. Whether you are utilizing MIG welding to tack sheet metal, or utilizing SMAW electrodes for structural steel, the purpose remains identical: to lock your fit-up into place until completion. This functional definition is the cornerstone that the rest of this article will build upon.

Tack Weld vs. Temporary Weld: Not One and the Same

One of the most common misconceptions in the shop is lumping tack welds in with any “temporary” weld. While similar, these two terms are not interchangeable under welding codes and terminology. AWS A3.0 classifies a temporary weld as a weld made to attach pieces for temporary use in handling, shipping, or working on the weldment. These welds are commonly used as erection aids, lifting lugs, or shipping braces, and are oftentimes intended to be removed entirely after serving their purpose.

Tack welds, on the other hand, are not necessarily destined for removal. In many code-compliant fabrications, tack welds are left in place and become an integrated part of the final joint. Depending on the process, they may be fully remelted into the root or fill passes, or they may be incorporated into the completed weldment as-is. This key distinction directly impacts quality requirements, welder qualification, and engineering approval. For projects with strict code compliance, such as pipe welding procedures, these differences become crucial to avoid non-conformances.

Purpose Over Size: Why the Colloquial View Falls Short

Walk into any shop and you’ll hear the age old rule, “keep your tacks small.” While this is sound advice for many applications, it is not a universal requirement. The colloquial belief that tacks must be small and intermittent can be undesirable in certain circumstances. For example, when working with crack sensitive materials, larger or more frequent tacks may be needed to prevent movement while also adhering to preheat requirements. On large diameter pipe or heavy structural members, undersized tacks can crack under the stress of fit-up before the root pass even begins.

This is why fabrication standards focus more towards procedural control rather than an arbitrary size. Factors such as material thickness, joint configuration, and distortion control will dictate tack sizing and spacing far more accurately than the phrase, “just a quick tack”. Regardless of their size, every tack must be made with the same level of care that you would put forth for a final pass. This includes proper joint prep, suitable consumables, and adequate PPE to ensure the utmost safety for the welder.

Setting Up the Scope: What This Article Will Cover

With the technical definition now established, the true search intent becomes clear: not only “what a tack weld is”, but also “how it’s made, and why it matters.” Throughout the remainder of this guide, we will dive into the seven functions of tack welds, the various types that codes recognize, and the specific requirements set forth by organizations such as AISC and AWS D1.1. Additionally, we will cover proper sequencing to combat distortion, step by step techniques for processes such as stick welding thin metal, the most common mistakes to avoid, and the OSHA welding safety regulations that apply to tacking operations.

From the smallest gauge tab to the heaviest structural column, tacks are the unsung heroes of every sound weldment. By treating them as a critical part of the welding process, rather than an afterthought, we can prevent the exact disaster that our opening fabricator experienced. With that being said, it’s time to take a deeper look into the real job that these tiny welds are responsible for.

The Real Job of Tack Welds: Seven Functions Behind Every Good Fit-Up

Before a single structural bead is laid, two separate things have to happen. First comes fit-up: the parts are cut, cleaned, positioned, squared and spaced exactly where the drawing says they belong. Then comes tack-up: small welds are placed to lock that geometry in time. Shops use the words almost interchangeably, but the distinction matters. Fit-up is the geometry you want. Tack-up is what keeps the geometry from escaping while you weld it. A perfect fit-up with sloppy tacks is a temporary achievement, and a rough fit-up with careful tacks still won’t save you.

That is why the real job of a tack weld is never “just holding parts together.” It sets and maintains the prescribed joint gap and establishes the mutual positioning of the components, as the practical guidance on tack weld practice puts it. Everything else flows from those two duties.

The Seven Things a Tack Weld Actually Does

1. Holds components in place. The obvious one. Parts stay where a fitter put them instead of creeping, twisting or sliding as the assembly is rolled, flipped or simply left alone overnight.
2. Establishes alignment. Squareness, parallelism, concentricity and rotation are all frozen by tacks. Remove a tack that was doing alignment work and the part will move, usually in the direction you least want.
3. Sets and maintains the joint gap. Root opening and land are held by the tacks, not by gravity. On a long seam, too few tacks let the joint progressively close up: with MMA or MIG the edges can actually overlap, and with submerged arc the joint can open up instead. Both outcomes are fit-up failures caused entirely by tack quantity, not by the welding process itself.
4. Complements fixtures. Jigs and strongbacks do the heavy positioning, but tacks finish the job and, just as importantly, let you remove the assembly from the fixture so you can reposition, rotate or weld in a better spot. A fixture alone can trap you; a fixture plus tacks gives you options.
5. Controls distortion. Tacks are the first and cheapest line of defense against weld shrinkage. Balanced, well-sequenced tacks resist the pull of each passing bead far better than clamps alone. TWI’s distortion control guidance treats restraint planning as a fabrication decision, not an afterthought.
6. Provides temporary strength. Tacks carry component weight during hoisting, turning and manipulation. Their strength scales directly with throat size and length, so a 3/8-inch assembly lift is not a job for three 1/2-inch tacks. Think of tacks as temporary structural welds, because on the crane hook that is exactly what they are.
7. Allows removal for adjustment. A good tack is designed to be cut out without gouging the base metal. That reversibility is what makes trial assembly possible: fit it, check it, correct it, then commit.

Three Situations Where Tacks Do the Real Work

Pipe fit-up. A root gap that shifts by a millimetre mid-weld changes penetration and can ruin a root pass that has to pass radiography. Tacks bridge the gap, hold the land, and keep the two spools aligned while the fitter checks the tack-up against the WPS. Our pipe welding procedures guide walks through how tacking fits into the wider pass and inspection sequence, and pipe tacking reference material covers the clock-position logic behind it.

Frame building. Here tacks act as a checkable draft. You tack a rectangular frame on the diagonals, measure, and if the numbers are off you cut two tacks instead of grinding out six feet of bead. Do this before you commit and the frame stays square; skip it and every subsequent bead inherits the error.

Heavy lifts. A welded assembly that must be hoisted, turned and set back down is relying entirely on tacks for structural continuity. Check throat thickness, count them, and don’t be shy. The Fabricator’s tack welding walkthrough makes the same point: tacks are engineered, not improvised.

Once you see tacks as a multi-function tool rather than glue, most fit-up problems get diagnosed faster. Whether you are running short tacks from a MIG welder set up for tacking or a stick rod from an open root, ask what each tack is doing for you. If you can’t answer, it probably isn’t doing much.

Tack Weld Types: Incorporated, Remelted, Non-Incorporated and Temporary Welds

Not every tack weld is created equal. Some are meant to disappear into the finished joint, some are meant to be eaten alive by the final pass, and some are strictly forbidden from staying in the structure at all. Knowing which category you’re making decides your electrode choice, your heat input, your removal obligations, and ultimately whether your weldment passes inspection.

The AWS and AISC framework sorts tack welds into four practical buckets. Let’s walk through each one.

1. Incorporated Tack Welds (Inside the Joint)

An incorporated tack weld sits inside the joint groove and becomes part of the final weld. When the production passes cover it completely, the tack effectively ceases to exist as a separate feature. This is the most common scenario in groove welding: you tack the pieces, check fit-up, then run root, fill and cap passes right over the top.

Because the tack is literally absorbed into the weldment, it must meet the same quality bar as the final weld. That means the same WPS, the same preheat, the same low-hydrogen consumables, and a qualified tack welder. If you’re running 7018 tacks over preheated plate, your 3/32 E7018 amperage settings for the tack should match what the procedure calls for, not whatever the machine happened to be set at.

2. Tack Welds Designed to Be Remelted

A special subset of incorporated tacks is built to be completely remelted by a high-energy final process, most often submerged arc welding (SAW). Because SAW deposits heavy, high-penetration passes, a properly shaped tack simply vanishes into the melt pool.

The counterintuitive sizing rule here: these tacks should be small in throat thickness but longer in length. A long, thin tack gives the SAW pass enough volume to remelt the entire tack without leaving an un-fused island. A tall, chunky tack may survive inside the bead as a discontinuity.

It’s worth noting that remelting logic shows up elsewhere in welding. Aluminothermic rail joining relies on a similar principle, as covered in our guide to railroad thermite welding.

3. Non-Incorporated Tack Welds (Outside the Joint)

A non-incorporated tack weld lands outside the final weld region, on the base metal surface, to hold fit-up. Nothing melts it later. It stays exactly where the arc left it, which is why codes treat it far more strictly.

Under AWS D1.1 Section 5.18.2.3, non-incorporated tack welds shall be removed. The code allows one notable exception: for statically loaded structures, they need not be removed unless the Engineer so requires. For cyclically loaded structures, the language tightens further. In the tension zones of cyclically loaded structures, non-incorporated tacks must be removed per AWS D1.1/D1.1M Section 7.17, consistent with AISC’s welding guidance. A surface tack in a tension zone is a ready-made crack initiator under fatigue loading.

Removal isn’t a casual grinding job either. The tack must be cut or ground out and the base metal blended smooth, with any resulting gouges repaired and re-inspected. Rough removal that leaves a notch trades one defect for a worse one.

4. Temporary (Erection-Aid) Tack Welds

Temporary welds exist purely as erection or assembly aids: lifting lugs, alignment tabs, brace attachments, and fit-up aids that never belong in the finished structure. They fall outside the structural welding code’s acceptance pathway entirely and must be removed by cutting, then the area ground flush and the surface examined for cracks. On pressure vessels and piping, temporary attachment removal carries similar obligations.

How the Four Types Compare at a Glance

The practical takeaway is simple: decide the tack’s category before you strike the arc. If it will be incorporated, weld it like production metal. If it will be removed, plan the removal, the grinding, and the re-inspection while the fit-up is still being planned, not after the crane is already on the load. For process-specific tack parameters, our MIG tack weld guide walks through spacing, settings and sizing in detail.

What the Codes Say: AWS D1.1, AISC and Welder Qualification Rules

Every shop has an unwritten rule about tack welds: make them small, make them quick, and nobody will ask questions. The codes disagree. A tack weld is a weld in every sense that matters to a structural inspector, which means it carries the same paperwork, the same preheat, the same consumable controls and the same qualified welder behind it as the final bead.

AWS D1.1 Section 5.18: The Two Sentences That Settle It

In the AWS D1.1 Structural Welding Code – Steel, Clause 5.18 draws a useful distinction between two terms that get mixed up constantly. Clause 5.18.1 states plainly: “Temporary welds shall be subject to the same WPS requirements as the final welds.” Clause 5.18.2 follows with the line every welder should memorize: “Tack welds shall be subject to the same quality requirements as the final welds.”

Read those two clauses together and the meaning is unambiguous. If your Welding Procedure Specification calls for a specific preheat temperature, shielding gas, travel speed or electrode for the root pass, it calls for exactly the same things when you are putting three small tacks on a beam copé. The WPS does not pause while you fit things up. You can read the standard’s own framing in AWS’s free standards resources before you buy the full document.

The Two Narrow Exceptions (SAW Only)

D1.1 does allow relief, but only in two very specific cases involving tack welds made with the submerged arc welding process that will be completely remelted and consumed by the final weld. For those tacks, the code permits:

• Preheat flexibility — the preheat requirements that would otherwise apply need not be enforced for the tack itself.
• Discontinuity removal — small discontinuities in the tack do not have to be individually removed, because the subsequent weld deposit consumes them.

Outside of that narrow SAW scenario, no exceptions exist. An arc strike outside the weld zone, an undercut, a crater crack or a cold lap on a tack is evaluated exactly as it would be on the finished joint. The Fabricator’s guide to performing a successful tack weld shows what “consumed by the final weld” looks like in practice.

Preheat, WPS Compliance and Low-Hydrogen Consumables

Two supporting requirements cause the most shop violations. First, preheat: if the base metal or thickness triggers preheat under Table 5.8 of D1.1, tack welds get preheated too, and the interpass temperature still applies. Second, consumable control: structural tacks are normally made with low-hydrogen electrodes such as E7018 or a low-hydrogen FCAW/GMAW wire, stored and handled under the same drying rules as the rest of the job. Hydrogen does not care whether the deposit is 3 inches long or 3/8 inch long. For a refresher on classification numbers, see What Are Electrodes in Welding? Types, Numbers, Polarity & How to Choose, and for a common low-hydrogen rod, 3/32 E7018 Amperage: Settings, Polarity & How to Run It.

Process-specific WPS control matters as much as rod choice. A GMAW tack set to the wrong voltage or wire-feed speed is a procedure violation whether or not it looks fine, which is why How to Set Up a MIG Welder: Complete Step-by-Step Guide and our MIG Tack Weld Guide both start with settings before technique.

Who Is Allowed to Tack: Clause 6 Qualification

This is the rule most often quietly ignored. Under AWS D1.1 Clause 6, welders, welding operators and tack welders must all be qualified for the process they use. Qualification does not last forever: a tack welder remains qualified only while welding with that process at least once every six months. Let the interval lapse and requalification is required before that person lays another tack. Welding is one of the few trades where a four-week holiday can technically invalidate your skill record.

AISC’s Engineering FAQ 8.5.1 reinforces the same discipline from the design side, directing readers to AWS D1.1/D1.1M Section 7.17 for the rules on removing tack and erection-aid welds rather than leaving them to field judgment. On pipe work, where tacks often become part of the root pass, the qualification and procedure expectations are even stricter, as covered in Pipe Welding Procedures: The Complete Guide to WPS, Positions, Passes & Inspection.

Engineering Approval of Tack Procedures

Finally, the engineer of record or fabricator must approve the tack procedure itself — including tack size, location, removal method and whether the tack will be incorporated into the final weld. Any tack that will be gouged out later needs a defined removal process that does not cut into base metal. For work governed by OSHA’s cutting and welding standard, the general employer requirements in 29 CFR 1910.252 apply alongside the code clauses.

How Long, How Many, How Far Apart? Tack Weld Sizing, Spacing and Placement

Ask ten welders how big a tack should be and you’ll get ten different answers. Ask them how far apart to space the tacks and the room gets quieter still. In reality, sizing and spacing aren’t matters of opinion — they’re governed by joint geometry, material, process, and a handful of numbers experienced fitters carry in their heads. Learn those numbers and your fit-ups stop springing apart.

Tack Length: The Half-Inch to Three-Quarter-Inch Rule

For structural steel work, a typical tack weld runs about one-half to three-fourths of an inch, and in most shop situations it should not exceed one inch. That sounds small, but remember what a tack actually is: a small weld whose job is to hold parts in place, not carry the load of the finished structure. A one-inch tack on 3/8-inch plate is already more metal than the fit-up needs, and oversized tacks cause real problems — distortion, hard spots, and tacks the final pass can’t remelt.

Length scales with thickness. On light sheet, a tack as short as 1/4 inch may be plenty, especially with TIG welding sheet metal, where heat control is precise and a huge tack would warp the panel instantly. On heavy plate or structural members, 3/4 to 1 inch is the working range. What matters more than length is penetration: a tack sitting on the surface like a caterpillar is worse than no tack at all, because it looks secure while sliding out of alignment under the first real heat cycle.

Pipe Tacking: Clock Positions Everyone Should Know

Pipe fit-up has its own well-established convention, taught early in any pipe welding procedure. The standard approach is four evenly spaced tacks at the 12, 3, 6 and 9 o’clock positions. Many crews instead use 11, 2, 5 and 8 o’clock, offsetting the tacks from the top and bottom — a smart move when the root pass will start or stop at 12 or 6, because the tacks won’t sit in the arc-start and arc-stop zones.

Each pipe tack should be roughly 3/4 inch long and penetrate through to the root. On large-diameter pipe, tack length should be about 5 to 10 percent of the pipe diameter, so a 24-inch line gets tacks around 1-1/4 to 2-1/2 inches while a 6-inch pipe stays close to the 3/4-inch standard. Two tacks are enough for small, light fittings held in a clamp; four is the default for anything free-standing.

Root penetration is non-negotiable. A tack that bridges the bevel without tying into the root leaves an unfused island the root pass must burn through or work around — and in a radiographed joint, that island becomes a reject. This is why pipe tacking uses the same low-hydrogen consumables and, where required, the same E7018 amperage settings as the weld itself, not whatever rod is closest to hand.

Spacing: Why Material Matters More Than Distance

No code stamps a universal spacing number, which frustrates beginners. What exists instead is a principle: the more the assembly wants to move, the closer the tacks need to be. On mild steel, tacks every 12 to 24 inches are often adequate for a rectangular frame. On austenitic stainless steels, spacing must be much shorter, because stainless expands and contracts far more under heat. A gap set at room temperature on stainless will wander if the tacks are spread too far, and the joint walks out of alignment the moment the first bead goes in.

Thin material tightens spacing too, since a long unsupported span between tacks can buckle before the weld starts. The practical habit: place enough tacks that no section of joint can move independently. If you can flex the gap with light finger pressure between two tacks, you need another tack — or a better one.

Placement Rules That Prevent Trouble

Placement is where sizing and spacing meet distortion control. Start tacks at the middle of the joint and work outward in opposing directions, rather than running from one end to the other — the pattern that ruined the 12-foot frame in our opening story. Corners, ends, and changes in section thickness deserve tacks first, because those are where movement starts. Avoid placing a tack where the final bead will stop and restart, and never leave a tack sitting proud of the joint where it can’t be incorporated or feathered.

Finally, treat every tack as a weld, not a placeholder. It gets the right filler, the right heat, the right preheat where the base metal demands it, and a careful inspection before the first production bead is struck. Get the numbers right — half an inch to three-quarters, four tacks on the clock, closer spacing on anything that grows in the heat — and the rest of the job becomes dramatically easier.

Tack Sequence and Distortion Control: Why Order Matters More Than You Think

Two welders can place the same number of tacks, the same size, on the same joint, and get completely different results. One assembly comes off the table square and stress-free. The other is bowed, twisted, or short by several millimeters. The difference is almost never torch skill. It is sequence.

Heat makes steel expand, and the cooling metal that follows pulls whatever it is locked to. A tack weld locks two pieces together while that pull is still active, so the order in which you commit those locks decides which direction the assembly moves and how far. This is why TWI lists tack welding first among the principal fabrication techniques for minimizing distortion — before balanced welding, before heat sinking, before peening. Sequence is the cheapest distortion control you have, and it costs nothing but attention.

Why Tacking From One End Fails

The most common bad habit is tacking progressively from one end to the other. Each tack shrinks as it cools, and because everything behind it is already rigid, that shrinkage has nowhere to go except forward. The gap at the far end closes. Keep going and the sheets do not just close up — one sheet end can actually overlap the other. By the time you notice, the fit-up you carefully measured at the start no longer exists, and the final weld simply locks the error in permanently.

The Sequences That Work

Practical fit-up guides recommend a short menu of reliable patterns:

• Center-out alternating. Tack the middle of the joint first, then alternate sides as you work toward each end. Shrinkage pulls equally in both directions and mostly cancels itself.
• Center, then back-step. Place a tack at the center, then work back toward one end in short steps, return to the center, and work toward the other end.
• One end, then back-step. Anchor one end deliberately, then back-step the remainder so accumulated shrinkage moves toward the anchored point instead of piling up at the open end.
• Balanced or back-to-back. Place matching tacks on opposite sides or opposite corners of the joint at the same time. Equal pull on both faces keeps the assembly flat.

Back-stepping deserves special attention because it works at the bead level too, not just the tacking level. Each short segment welds against the direction of travel, so each new segment’s shrinkage fights the previous one instead of adding to a single long pull. The same idea scales down to tacks.

Directional Tacking: Closing a Gap That Is Too Wide

Sometimes the problem runs the other way — the fit-up is loose and the gap is growing. In that case you tack deliberately in one direction, starting at the end where the gap is already acceptable and working toward the wide end, letting each tack draw the material slightly. This is a controlled, intentional version of the same shrinkage force that wrecks careless sequences. Used on purpose it closes an oversized gap; used accidentally it creates an overlap. On pipe welding procedures where the root gap is held to tight tolerances, this directional habit is close to mandatory — a drifting root gap on pipe shows up as a fit-up rejection long before the arc is struck for the root pass.

What the Research Says About Tack Placement

Sequence is only half the picture; placement, size and spacing are the other half. Research summarized in AWS Welding Journal found that properly designing the location, size and spacing of tack welds is a highly effective way to reduce distortion, because the tacks change the stiffness of the assembly before any production welding begins. A well-placed tack does not just hold a part — it stiffens the joint so the later beads have less room to pull it out of shape.

That is a useful mental shift. Instead of thinking of tacks as temporary staples, think of them as the framework that decides how your weldment will move. On thin sheet, tack placement matters even more because there is so little material to resist pull; our guide to TIG welding sheet metal makes the same point about heat input and restraint working together. Even a MIG tack weld follows these rules — the process changes, the sequence logic does not.

A Practical Rule of Thumb

Before you strike an arc, decide where shrinkage is allowed to go, then place tacks so it goes there. Anchor what must stay put, balance what can move, and leave one direction open for accumulated pull to escape. Mark the planned tack order on the part if you have to. Thirty seconds of planning prevents the kind of silent, cumulative drift that only shows up when the final bead goes down and the assembly is already out of square.

How to Make a Good Tack Weld: Step-by-Step Technique for SMAW, GMAW and GTAW

Knowing what a tack weld is and where to place it only gets you to the starting line. The quality of the tack itself decides whether your final bead ties in cleanly or fights you the whole way. A good tack is small, sound, clean, and shaped so the finished weld can absorb it without a trace. A bad tack becomes a crater, a crack, or a hard spot you have to gouge out later.

Here is the practical technique, process by process, for turning a fit-up into tacks you can trust.

Step 1: Joint Prep and Fit-Up Checks Before You Strike an Arc

Every tack starts before the rod or wire comes out. Clean the joint faces back at least 1 in. on either side of the gap, removing mill scale, rust, oil, paint, and cutting burrs. Confirm the root face, root opening, and land against the drawing or the pipe welding procedures you are working to, then re-check squareness and diagonal measurements after the first tack. Steel moves as you add heat, so fit-up is a repeating check, not a one-time event.

Step 2: SMAW Tacks, Especially on Pipe Roots

Stick tacking on pipe is where technique matters most. According to the pipe tacking walkthrough in Open Washington’s welding pipe course, strike the arc about 1/2 in. ahead of where you want the tack to begin. Hold a long arc for a moment to preheat the area and let the flux shielding build, then ease in and hold position until a keyhole forms. That keyhole is your proof of full penetration.

Run the tack roughly from the 12 o’clock position toward 3 or 9 o’clock, keeping the keyhole small and consistent. To extinguish it, push the electrode through the keyhole with a quick thrust. This technique closes the arc without leaving a crater and preserves keyhole size, which matters when you restart or tie the root in later. Immediately fill any crater dimple while the metal is still hot.

Choosing the right consumable and amperage is half the battle. If you are running 7018 tacks, this 3/32 E7018 amperage guide gives you a sane starting window, and our broader electrode selection guide explains what those digits on the flux actually mean.

Step 3: GMAW and GTAW Tacks

For MIG tacking, stable setup comes first: liner, contact tip recess, drive roll tension, and gas flow all affect arc start quality, as covered in our MIG welder setup guide. Short-circuit settings that are too cold produce cold lap; too hot and you burn through thin fit-up edges. If you want process-specific spacing and settings, the MIG tack weld guide covers sizing and spacing in detail.

TIG tacks are the controlled option on thin and stainless work. Build the puddle slowly, add filler deliberately rather than dabbing blindly, and taper the current at the end to avoid a crater. The same taper-and-fill discipline applies on sheet, as explained in the complete guide to TIG welding sheet metal.

Step 4: Clean, Feather, and Inspect Before the Final Weld

This is the step most welders skip, and it is where tack problems turn into weld defects.

• Remove all slag. Chip and wire-brush every tack completely before the hot pass. Trapped flux becomes a slag inclusion, one of the most common reasons a root or hot pass gets rejected.
• Feather the ends. Use a hand grinder to thin each tack end to a minimum feathered-edge thickness of 1/16 in. so the final bead can wash into it instead of stacking a cold stop-start on top.
• Grind arc strikes. Any strike outside the joint needs to be blended smooth; arc strikes outside the weld zone are treated as surface defects by most codes.
• Inspect. Look for craters, undercuts, cracks, and lack of fusion. Cracks are never acceptable in a tack and must be removed, not covered.

Note: Tack removal deserves the same care as tack placement. Gouging or grinding that notches the base metal leaves a stress raiser you will never get back. Cut tacks cleanly, then blend flush.

Done right, tacks disappear into the finished weldment. Nobody notices them, which is exactly the point.

The 10 Most Common Tack Weld Mistakes (and How to Fix Each One)

Experienced inspectors will tell you that tack welds rarely fail in strange or mysterious ways. Almost every problem traces back to the same handful of habits: a rushed arc, a skipped step, or a tack that was never meant to be there in the first place. The good news is that every one of them is preventable once you know what to look for. Here are the ten errors that cause the most rework, along with the fix for each.

1. Craters at the End of the Tack

Pulling the arc straight off the puddle leaves a dish-shaped crater, and craters are classic crack starters because the shrinking metal has nowhere to go. Back off the current slightly or step back into the puddle for a fraction of a second before breaking the arc, then use the post-flow on a MIG setup to keep shielding gas over the solidifying metal. If a crater does form, grind it out and re-tack rather than welding over it.

2. Cracks in or Around the Tack

Surface cracks usually come from a crater, a contaminated joint, or too little heat for the thickness being joined. Cracks are never acceptable in a tack, even a temporary one, because the final weld will consume the cracked metal and inherit the defect. Cut or grind the tack out completely, clean the area, and re-tack with correct settings and a low-hydrogen consumable, as covered in 3/32 E7018 amperage selection.

3. Arc Strikes Outside the Joint

Flicking the arc on adjacent base metal leaves small, hard spot marks that act as stress raisers and are grounds for rejection under most structural codes. Strike the arc on a start tab, on the tack location itself, or inside the joint. Any accidental arc strike should be ground smooth and examined, since the spot can be hard enough to crack later.

4. Slag and Weld Oxide Left in Place

Chip and wire-brush every stick or flux-cored tack before the next pass goes over it. Slag trapped under a bead causes lack of fusion, and a buildup of weld oxides on a tack that will be remelted contaminates the final weld pool. The same discipline applies to flux core wire welding, where slag removal between passes matters even more.

5. Hard Spots and Underbead Cracking from Low Heat Input

This is the mistake fabricators discover late. A tack made with too little heat on crack-sensitive steel can leave an excessively hard heat-affected zone or a hidden underbead crack that grinding will not remove. Bump the amperage into the proper range for the electrode diameter, keep travel speed moderate, and treat every tack on a sensitive alloy as if it were a production bead. The same metallurgy logic applies when you are welding hardened steel.

6. Skipped Preheat

Cold tacks on thick, restrained, or high-carbon joints chill too fast and crack. Preheat the tack zone exactly as the WPS requires, and verify with a temperature indicator rather than by feel. Codes exempt almost nothing here, which The Fabricator’s tack welding guidance reiterates: tacks follow the same preheat rules as production welds.

7. Oversized Tacks That Will Not Remelt

A tack larger than the root pass can bridge over instead of fusing, and welding over large intermittent tacks can disrupt the arc and degrade the appearance of the final weld. Keep tacks modest, typically 1/2 to 3/4 inch, so the root pass can fully consume them.

8. Too Few Tacks, or Poorly Placed Ones

Four tacks on a twelve-foot frame will not hold. Space tacks by the joint length, material stiffness, and gap requirement, adding more on thin or flexible work. See MIG welder setup for wire speed and voltage ranges that keep small tacks controlled.

9. Improper Removal That Gouges Base Metal

Thermal cutting too close to the final surface leaves a gouge that must be filled and ground, adding a defect where none existed. Cut slightly proud, then grind flush, and inspect the area for undercut before welding over it. Acceptance limits are covered in undercut causes and acceptance limits.

10. Unqualified Tackers

AWS D1.1 requires tack welders to be qualified under the same procedures as production welders, with requalification every six months. Using an untrained helper to “just throw a few tacks on” puts unqualified weld metal into a qualified structure, a point AISC’s welding FAQ makes plainly. Check qualifications, and follow pipe welding procedure requirements on any pressure work.

Run through these ten before every fit-up and the majority of tack-related repairs disappear. The weld that follows a clean, correctly sized, correctly placed tack almost always comes out better too.

Tack Weld Safety: OSHA Fire Watch, Ventilation and Fume Rules You Can’t Skip

For most welders, tacking is the quickest part of the job. A few seconds to strike an arc and secure a joint can feel far less hazardous than laying down a full pass. However, the short duration of a tack weld does not make it any safer. In fact, the sparks, slag, fumes and heat from even the smallest tack can create a dangerous work environment if the proper precautions are not taken. With that in mind, the following safety measures should be non-negotiable for any fabricator or shop.

Fire Prevention: The 35-Foot Rule and Fire Watch

The greatest threat from a tack weld is often not the weld itself, but the fire it can start. Grinding, spatter and stray sparks can ignite flammable materials in seconds, which is exactly why fire prevention is the first line of defense.

Under OSHA 29 CFR 1910.252(a)(2)(vii), where practicable, combustibles must be relocated at least 35 feet (10.7 m) from the welding or cutting work site. If moving these materials is not possible, they must be properly shielded with fire-resistant barriers. This rule applies to tack welds just as heavily as it does to multi-pass structural welds.

Even after the last tack is complete, hidden embers can reignite hours after the arc is extinguished. For this reason, OSHA also requires that a qualified fire watch be maintained for at least a half hour after completion of welding or cutting operations. This is especially crucial for jobs in congested shops, around wood scaffolds or when working on overhead stick welding tasks where spatter can fall a considerable distance.

Ventilation, Fumes and Hazard Communication

Tacking may only last a moment, but it still releases welding fumes and gases that can be hazardous to your health. Proper ventilation is not a luxury, but a requirement for compliance and long-term wellbeing. In general, OSHA calls for general mechanical ventilation of at least 2,000 cubic feet (57 m³) per minute per welder unless local exhaust ventilation (LEV) or approved airline respirators are used.

These fumes can become exponentially more dangerous in confined spaces, where gases can displace oxygen or reach toxic concentrations with little airflow. For high-fume processes such as flux core wire welding, and even when performing a precise MIG tack weld, exhaust capture should be prioritized.

Fume exposure prevention starts before the arc strikes. All filler metal containers must carry the notice: “Welding may produce fumes and gases hazardous to health. Avoid breathing these fumes and gases. Use adequate ventilation.” This falls under OSHA’s Hazard Communication requirements, and it is your responsibility to read and heed it. For more on consumables, take a look at our guide on what are electrodes in welding.

PPE, Inspection and Non-Destructive Testing of Tacks

With fire and fumes accounted for, personal protection is the final barrier between you and injury. Proper welding PPE, including a shaded helmet, flame-resistant clothing, gloves, hearing protection and respirators when needed, must be worn for tacking. Never use cracked lenses, exposed skin protection or damaged leads, as these faults can be just as dangerous for a single tack as they are for a full bead.

Lastly, do not assume a tack is ready to be covered. Tacks can contain craters, cracks or porosity that will propagate if left unchecked. Critical fabrications, such as pipe welding procedures, may require visual inspection or specific Non-Destructive Examination (NDE) of tack welds before full welding. Even for general work, every tack should be visually inspected, with any questionable discontinuities removed or repaired by a qualified welder. For guidance on correcting such flaws, our resource on welding defects undercut is a great starting point.

Tack Weld vs. Clamps, Fixtures and Temporary Welds — Plus a Final Checklist

Tacks are only one way to hold a fit-up. Before you commit to a welding sequence, it is worth asking whether a clamp, a fixture, or a temporary erection weld would do the job better. Each option solves a different problem, and each carries a different distortion price tag.

Tacks vs. Clamps, Fixtures and Backing

Clamps are fast, infinitely repositionable, and leave no weld metal in the structure. Their weakness is physics: clamping eliminates plastic distortion but cannot prevent the elastic portion of the shrinkage strain, which springs back the moment you release the clamps. A frame that looked dead-square under pressure can walk out of square as soon as the last clamp comes off.

Fixtures and jigs go further by repeating the same job accurately on production work, but they cost real money to design and build, they must be periodically checked for wear, and a fixture that has drifted will produce a consistent run of wrong parts. Backing bars and chill blocks handle heat and root geometry, but they restrain rather than align.

The interesting case is the mirror-symmetrical back-to-back assembly. When two identical halves are welded face to face, shrinkage forces cancel each other, so such assemblies can be tack welded instead of clamped and left to balance their own distortion. On that kind of work, tacks beat clamps because they let the joint move in a controlled, symmetric way instead of storing strain to release later.

A sensible rule from the fabrication literature is the three-step order: tack symmetrically, recheck diagonals, then weld in a balanced sequence from central or stiff regions toward free ends. Note the recheck: tack shrinkage itself moves diagonals, so measure after tacking, not only before.

Temporary Welds: A Different Animal

Temporary or erection-aid welds are made outside the joint to lift, brace, or align a member, then removed before final inspection. They are convenient, but removal is a real cost: every gouge risks carving into base metal, and every removed weld leaves a surface that must be ground and often re-inspected. If the member will be machined after welding, stress relief can reduce the residual stresses behind elastic distortion by roughly 90 percent before machining, which is usually cheaper than chasing a distorted part around the shop.

When you do need temporary strength for a heavy lift, size it deliberately and remember the removal path. The same care applies to the process you tack with: MIG tacks are fast and forgiving, while a root tack on pipe follows the keyhole technique described earlier in our pipe welding procedures guide.

The Pre-Weld Tack Checklist

Run this list before you strike the final bead:

1. Joint is clean — mill scale, oil, paint and moisture removed; consumables dry.
2. Fit-up verified — gap, land, and diagonals measured and within WPS tolerance.
3. WPS and preheat confirmed — tack procedure approved, interpass temperature ready.
4. Tacker qualified — qualifications current, including the six-month requalification rule.
5. Tack size and spacing correct — typically ½ to ¾ in., closer on stainless and thin material.
6. Sequence balanced — start mid-joint and alternate outward, never one end to the other.
7. Cranes and PPE ready — see what a welder wears before any hoisting.
8. Tack ends feathered — ground to at least 1/16 in. and inspected for cracks or craters.
9. Slag and spatter removed — no inclusions waiting to become a repair.
10. Fire watch and ventilation set — OSHA 1910.252 requirements in place.

Conclusion: The Frame That Held Square

Go back to the 12-foot frame from the opening story. Nothing was wrong with the beads; the failure lived in tacks placed in a rush, in a straight line from one end to the other, pulling progressively as they cooled. A symmetric start at mid-frame, alternating outward, with diagonals rechecked after every pair, would have held the assembly square for a fraction of the effort.

Treat every tack as a small, permanent decision about geometry. Size it, sequence it, qualify it, inspect it, and remove only what you planned to remove. Do that consistently, and your weldments come off the table square the first time — no second chances, no two-day rework, no surprises when the clamps finally come off.

Frequently Asked Questions

What exactly is a tack weld?

A tack weld is a small weld made to hold parts of a weldment in proper alignment until the final welds are made, according to AWS A3.0. It is a positioning device first and a structural weld second — but it must still meet quality requirements.

Do tack welds have to meet the same code requirements as final welds?

Yes. AWS D1.1 Section 5.18.2 states tack welds are subject to the same quality requirements as final welds, and temporary welds must follow the same WPS as the final welds. The main exceptions are preheat and discontinuity removal for single-pass tacks later remelted by submerged arc welding.

How long and how far apart should tack welds be?

Shop tacks typically run about 1/2 to 3/4 inch and rarely exceed 1 inch, while pipe tacks are about 3/4 inch or 5-10% of the pipe diameter on large pipe. Spacing depends on thickness, joint length, and material — stainless steel needs closer spacing than mild steel because of its higher thermal expansion.

Who is allowed to place tack welds?

Tack welders must be qualified under AWS D1.1 Clause 6 just like welders making final welds, and they stay qualified only while using that process at least every six months. Many shops require the same qualified welder who will run the final bead to place the tacks.

Do tack welds have to be removed?

Non-incorporated tack welds outside the joint must be removed for most applications, though AWS D1.1 allows them to remain in statically loaded structures unless the Engineer requires removal. In tension zones of cyclically loaded structures they must always be removed.

Why does my tack weld crack?

Cracks usually come from too little heat input, missing preheat on thicker or alloy steel, a hard heat-affected zone, or an unfilled crater. Fix it by following the WPS preheat, using the correct filler, and blending crater ends.

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

A tack weld holds parts in proper alignment until final welding, while a temporary weld attaches hardware for handling, shipping, or fabrication support and is normally removed afterward. Both are held to the quality standards of final welds.

How do I stop my joint from closing up while tacking?

Tack from the center outward using a back-step or skip sequence instead of running tacks from one end to the other, and add tacks as needed to resist transverse shrinkage. Directional tacking can also deliberately close a gap that has become too wide.