Groove Weld Explained: Types, Symbols, Design, Defects & Inspection Guide

A fabricator once stared at a cracked bridge splice at 2 a.m. — the plate edges had been cut square instead of beveled, and the root never fused. The whole repair could have been avoided with one correct decision: the right groove weld preparation.

What Is a Groove Weld? Definition, Search Intent & Why It Matters

At 2 a.m. on a bridge retrofit job, a fabricator stood over a cracked splice and tried to figure out what went wrong. The plate edges had been cut square instead of beveled, the root never fused, and the joint cracked under load before the crew had even left the yard. The fix wasn’t exotic metallurgy or a fancy procedure. It was one correct decision made hours earlier: the right groove weld preparation.

That story is more common than anyone in the industry likes to admit. Groove welds carry the structures we drive on, the pipes that move pressure, and the frames that hold heavy equipment together. When they’re right, nobody notices. When they’re wrong, the consequences show up at the worst possible time.

The AWS Definition, in Plain Language

The American Welding Society defines a groove weld as “a weld made in a weld groove on a workpiece surface, between workpiece edges, between workpiece surfaces, or between workpiece edges and surfaces” (OpenWA Welding, WA 1.5.2, citing AWS, 2010, p. 21).

Let’s unpack that. A weld groove is simply the gap, channel, or prepared space you create between (or on) the pieces you’re joining. The welder then fills that space with deposited metal. Unlike a fillet weld, which sits on the surface like a bead of caulk along a corner, a groove weld lives inside the joint, fused into the edges themselves.

Why Groove Welds Are Everywhere

According to OpenWA’s welding curriculum, groove welds are “probably the most widely produced welds,” whether they’re made manually or semi-automatically. Think about what that means on a practical level: structural steel splices, pressure vessels, pipe runs, tank shells, heavy equipment booms. Anywhere full-thickness strength or a sealed joint is required, a groove weld is usually doing the work.

They’re common for a simple reason: they let you fuse the entire thickness of the material, or at least a controlled portion of it. That gives engineers a predictable, calculable connection. In structural design codes such as AWS D1.1, groove welds are the backbone of moment connections, butt joints, and pressure-boundary welds.

A Groove Weld Is Not a Joint Type

This is where beginners (and plenty of shop-floor conversations) get tripped up. Groove weld describes how the weld is deposited, not which joint you’re welding. The five basic joint types are butt, corner, tee, lap, and edge. Groove welds apply to all five:

• Butt joint: two edges beveled and joined face to face, the classic pipeline and plate splice.
• Corner joint: edges meeting at a corner, common in box and tank fabrication.
• Tee joint: a groove weld can be used when code calls for deeper fusion than a fillet alone provides.
• Lap joint: overlapping plates where a bevel or square groove may be specified.
• Edge joint: the groove is formed by the edges themselves, typical on thin sheet stacks.

So a fillet weld can appear in a tee joint, and a groove weld can appear in a tee joint. The joint is the configuration; the groove weld is the joining method. Keeping those two ideas separate makes reading WPS documents and AWS symbols dramatically easier.

Who This Guide Is For

This guide is written for everyone who touches a groove weld decision:

• Welders and fitters who need to translate a drawing into the right edge prep, root opening, and pass sequence.
• Inspectors and QC personnel who need to know what “acceptable” actually means.
• Engineering students and apprentices building a mental model of joint design before they ever strike an arc.
• DIYers and hobbyists working up from basic practice, perhaps starting with our guides on is stick welding hard or how to set up a MIG welder.

Whatever brought you here, the goal is the same: make groove weld decisions from knowledge rather than habit. The next section breaks down every groove preparation type — square, V, bevel, J, U, flare, and their double variants — so you can match the shape to the thickness, the code, and the load.

Types of Groove Welds: Square, V, Bevel, J, U, Flare and Double Variants

A groove weld is only as good as the groove cut into it. Under the AWS definition, a groove weld is simply a weld made in the groove between two members, and the shape of that groove varies with plate thickness, access, distortion control and cost. The preparations you will meet in the shop or field are square, V, bevel, J, U, flare-V, flare-bevel and scarf, and each can appear in single or double form.

The Naming Rule: What “Double” Actually Means

Square Groove

A square groove uses straight, square-cut edges with no bevel. It is the fastest to fit up and the cheapest to prepare, since a shear or clean saw cut is usually enough. The trade-off is penetration: arc energy alone must melt the full thickness, so square grooves are practically limited to thin material, low-strength service, or joints where the code permits reduced penetration. As a general guide, a square groove with no root opening works up to roughly 1/2 inch with processes that deliver deep penetration, while a root opening of about 1/16 to 3/16 inch helps ensure fusion at the bottom. If you are learning, our guide on is stick welding hard is a good companion, because the root pass is where beginners struggle most.

V-Groove

The V-groove is the workhorse of structural shop work. Both members are beveled, and the bevel angles add up to the total groove angle, commonly 60 degrees for structural steel under AWS D1.1. It gives reliable full penetration on light to medium plate, is easy to cut and easy to weld, and the wide opening gives the welder a clear view of the root. The cost is filler metal and heat: a thick V consumes a lot of consumables and pulls hard, making distortion more likely than a narrower prep.

Bevel Groove

In a bevel groove, only one member is beveled; the other stays square. Bevel grooves are typical where access is limited to one side or where the joint is a corner configuration, with a common total groove angle of 45 degrees for a single-bevel. Bevel and J grooves also introduce the broken-arrow rule in AWS A2.4: the symbol’s arrow points to the member that must be beveled or J-grooved. Fitters who ignore the broken arrow weld the bevel on the wrong plate, and the fit-up fails inspection before the first pass.

J-Groove and U-Groove

J and U grooves are curved preparations that need more machining time but dramatically reduce weld metal volume. Because the groove walls open gradually, the welder deposits far less filler to fill the joint than in a V of the same thickness, which means less heat input, less residual stress and less distortion. AWS forum expert G.S. Crisi has long recommended a J-groove or bevel groove over a double-V on thick plate for exactly that reason: cutting filler volume also cuts welding time and distortion. U-grooves are common on thick pressure-vessel and pipe walls; J-grooves appear where one-sided access or specific vessel procedures demand them.

Flare-V and Flare-Bevel Grooves

Flare grooves are made from the natural shape of a bent member rather than a cut edge. A flare-V groove forms between two rounded bends, and a flare-bevel groove forms between a rounded bend and a square edge, typical on cold-formed steel shapes. No cutting is required, which saves prep time, but the weld throat must be calculated from the specified effective throat, not the full groove depth. Effective-throat calculations for flare grooves are handled in AWS D1.1.

A related but distinct case is the scarf groove, a shallow angled lap preparation used mostly in special applications such as rail and thermite work. For a look at how a groove-like prep is used outside structural steel, see our feature on railroad thermite welding.

Double Variants: Double-V and Double-Bevel

Double grooves work the material from both sides. A double-V-groove spreads the deposited metal above and below the neutral axis, largely balancing shrinkage forces and reducing angular distortion compared with a single-V of the same thickness. That is why DOT procedures, such as the approved NCDOT double V-groove with back gouge detail used across Caltrans bridge procedures, is specified for unlimited thickness: the first side is welded, the root is back-gouged by air arc or grinding to sound metal, and the second side is welded to guarantee a defect-free root. You can see the same principle in pipe welding procedures, where back gouging or purge-side welding secures full root fusion.

Quick Thickness Guidance

• Up to about 1/4 inch: square groove, often with a small root opening.
• 1/4 to 1 inch: single-V or single-bevel; double-V starts becoming economical around 3/4 inch.
• Above 1 inch: double-V, double-bevel, J or U grooves to control filler volume and distortion.
• Thick pressure-retaining work: U or J grooves per the vessel WPS.

Match your consumable to the prep as well. Deeper grooves and vertical-up passes pair well with a controlled rod like the 3/32 E7018, while heavy root and fill work often favors 5/32 E7018 settings, and process setup details for shop groove work live in our MIG welder setup guide. Pick the prep first, and everything else follows from it.

How to Read Groove Weld Symbols (AWS A2.4) the Right Way

A groove weld symbol is shop language. The drawing says “bevel this edge, open it 1/16 inch, and give me 5/8 inch of groove depth,” and the fitter sets the bevel machine accordingly. But symbols are also the most misread language in a fabrication shop. One misplaced letter or a missed broken arrow and you get square-cut edges, a root that never fuses, and a repair that costs a night’s sleep. This section is about reading them correctly the first time.

The Standard Behind the Symbols

Every welding symbol used in the United States traces back to AWS A2.4, Standard Symbols for Welding, Brazing, and Nondestructive Examination. A2.4 defines how symbols are drawn and interpreted, and it is the document your welder, fitter, and inspector are all expected to speak. One important recent change: the 2012 revision standardized the term “Depth of Groove” to replace the older “Depth of Bevel,” since not every groove is technically a bevel. If you work from older charts or shop standards, be aware of that terminology shift.

A2.4 also gives us the full family of groove symbols: square groove, V-groove, bevel groove, U-groove, J-groove, flare-V, flare-bevel, and scarf. The shape of the symbol mirrors the shape of the edge preparation on the plate, which makes the system intuitive once you see it a few times.

The Anatomy of the Symbol

Every welding symbol starts with the same skeleton: an arrow, a reference line, and the weld symbol itself.

• Arrow side: the side of the joint the arrow points to. A symbol placed below the reference line applies to the arrow side.
• Other side: the opposite member. A symbol placed above the reference line applies to the other side.
• Both sides: symbols above and below the line mean groove welds on both sides, as in a double-V groove.

This arrow side / other side convention is where most shop-floor misreads happen. A fitter glances at the drawing, sees a V symbol, and preps the wrong plate. Slow down here; the side of the line is not decoration, it is an instruction.

The Broken Arrow Rule

For bevel and J grooves, only one member is shaped while the other stays square. The symbol alone cannot tell you which member to bevel, so A2.4 requires a broken (kinked) arrow pointing directly to the member that receives the preparation (A2.4, Clause 5.4.1). If the arrow is straight, do not guess — go back to the engineer. Beveling the wrong plate on a 40-foot girder is not a mistake you fix with a grinder.

Reading the Dimensions: S, (E), A and R

Groove welds are dimensioned with a small, consistent set of letters placed around the weld symbol:

• S — Depth of Groove (formerly Depth of Bevel). How far the edge preparation extends into the member.
• (E) — Effective Throat. The depth of weld metal that actually carries load, shown in parentheses because it is a calculated value.
• A — Groove angle (included angle of the V or U).
• R — Root opening, the gap between the two members.
• C — Weld size (width) for square grooves, and P — Penetration, when depth of penetration must be specified separately.

Notice how S and (E) can differ. The groove may be deeper than the load-bearing throat. This distinction matters because, per AWS D1.1, shop drawings must show the groove depth “S” required to attain the specified weld size “E”. The welder needs both numbers: one for fit-up, one for acceptance.

Common Misreads That Cost Money

• Ignoring the arrow side. Prepping the opposite member because “it looked easier.”
• Missing the broken arrow on a bevel or J groove, then beveling the wrong plate.
• Confusing S and (E). Reading 5/8 as the throat, then under-welding or over-welding the joint.
• Treating a square groove like a V. Adding a bevel that was never specified wastes filler and changes the effective throat calculation.
• Assuming A and R are optional. If the drawing omits them, they come from the WPS, not from improvisation.

For how these symbols play out in real piping work, our guide to pipe welding procedures walks through symbols, positions, and passes together. And when a beveled root shows up with gouged-out undercut along its edge, see our breakdown of welding defects undercut for causes and acceptance limits.

Bottom line: the symbol is a contract between the designer and the shop. Read the reference line, check which side the symbol sits on, look for the broken arrow, and decode S, (E), A, and R before the first arc is struck. Two minutes of careful reading at the fit-up table saves two hours of gouging and re-welding later.

Groove Weld Joint Design: Groove Angle, Root Opening, Root Face & Land

Here is what separates a welder who just follows a rod size from one who understands the joint: edge preparation geometry. Groove angle, root opening, root face, and land are not arbitrary numbers a supervisor hands you. Each one controls how much heat reaches the joint center, how easily the arc reaches the bottom, and whether you finish with a fused, full-strength connection or a bead that never touched the root.

The Four Parameters You Must Know

Every groove joint, plate or pipe, is defined by four geometry terms:

• Groove angle: the total included angle between the two prepared edges. A single-V cut at 30° per side gives a 60° included angle.
• Root opening: the gap between the member edges at the root, the channel the arc reaches through to fuse the very bottom.
• Root face (land): the small flat vertical portion of the prepared edge, typically 1/16 to 1/8 inch (1.6–3.2 mm), which keeps the arc from blowing straight through before the joint fills.
• Bevel depth: how far the edge is cut back, measured from the root face to the member surface.

As open textbook welding curriculum from Washington’s community and technical colleges explains, all four serve one goal: give the arc enough access to fuse both sides and the bottom completely, using the least filler metal possible.

How Each Parameter Affects Penetration and Fusion

Groove angle trades arc access against filler volume. Wider angles let a longer electrode reach deeper, but demand more filler, more passes, more heat, and more distortion. Narrower angles save filler but risk sidewall lack of fusion, especially with short-circuit MIG or small-diameter wire. The code accounts for this: under AWS D1.1, a groove angle of 60° or greater requires a weld size equal to the full depth of preparation. Below 60°, the required weld size drops by 1/8 inch per side, because the tighter opening gives less guaranteed access to the joint walls.

Root opening is the number most often behind incomplete penetration. Too small and the arc never reaches the bottom; too wide and the root cannot be bridged without burn-through or excess filler. Values run from flush (0 on square groove welds in thin plate, where process and backing do the work) up to 3/16 inch or more on heavy structural joints. Root face works the other way, against burn-through: a thin or missing land on heavy plate lets the arc blow out the root, while a land that is too thick guarantees lack of penetration no matter how many passes you stack.

Back Gouging: Your Insurance Policy

On critical joints, the root pass is gouged from the reverse side down to sound metal, then welded from that side. This removes a suspect root and guarantees full fusion. Back gouging is central to heavy structural work, which is why approved procedures like the NCDOT double-V-groove with back gouge procedure spell it out explicitly. That procedure pairs a 60° total groove angle with E7018 low-hydrogen electrodes classified under AWS A5.1, a common, proven combination. For pipe, end preparation geometry follows standards such as ASME B16.25, which defines bevel angles, lands, and root openings for buttweld fittings.

Choosing Your Dimensions From the WPS

Practical reality: you do not invent these numbers on the shop floor. The Welding Procedure Specification, qualified to AWS D1.1 or ASME IX, dictates the exact angle, opening, land, electrode size, and heat input. The fitter’s job is to cut and fit to those numbers within tolerance. One useful access note from AISC’s research on weld clearances: allow roughly 18 inches of clear space around a joint so the welder can position the electrode and hold a consistent travel angle. No access, no quality weld, however good the geometry.

Getting these four parameters right is half the battle. The other half is reading them correctly from the drawing, which is exactly what welding symbols tell you. And before you strike an arc on any prep, make sure your welding PPE and joint access setup are in order. Sound geometry plus a properly followed WPS is what turns a groove weld from a hopeful bead into a load-bearing connection.

CJP vs PJP Groove Welds: Which One Does Your Design Need?

On a shop drawing, two groove welds can look identical in cross-section but carry completely different responsibilities. One transmits the full strength of the plate; the other only joins the edges far enough to handle a specific, smaller load. That distinction lives in two acronyms every welder, fitter, and inspector eventually has to master: CJP (complete joint penetration) and PJP (partial joint penetration).

Getting this wrong in either direction costs money. Specifying CJP where PJP would suffice means paying for beveling, back gouging, and NDT you didn’t need. Specifying PJP where the structure needs CJP is far worse: the joint simply cannot carry the load it was designed for.

What Complete Joint Penetration Really Means

A complete joint penetration groove weld extends completely through the thickness of the joint, producing complete fusion of the entire joint members. In plain language: the weld metal runs from one face of the plate all the way through to the other face, with fusion along the entire prepared surface.

As the Caltrans Memo 12-5 puts it, the purpose of a CJP groove weld is to transmit the full load-carrying capacity of the connected members. When a designer wants a welded butt splice to be as strong as the plate itself, CJP is the specification.

Reaching CJP is not automatic. Depending on thickness and position, it may require a backing strip, a backing weld, or a back gouge from the opposite side to confirm fusion at the root before the second side is welded. Every extra step exists for one reason: the root is the hardest place to fuse, and CJP means fusing it anyway.

What Partial Joint Penetration Means

A partial joint penetration groove weld does not extend completely through the joint thickness. The groove is prepared only partway into the plate, or full-depth prep is intentionally fused only partway, and the unfused bottom portion of the joint acts almost like a built-in crack. That sounds alarming, but it’s acceptable wherever the load demands are lower and the designer has calculated the weld’s actual load-carrying geometry.

That geometry is the effective throat. For a PJP groove weld, effective size is generally the depth of the bevel, with a possible 1/8 in deduction in certain cases, per the Washington Open Courseware welding text. A PJP weld on a 1-inch plate might only be credited with roughly 5/8 inch of throat, and that’s the number structural engineers use in their calcs.

Why Designers Call CJP or PJP Instead of a Groove Shape

Notice what a proper fabrication drawing specifies: “CJP groove weld, double-V, per WPS” or “PJP, 5/16 in effective size.” The penetration requirement comes first, the groove shape second. That’s deliberate. AWS D1.1 and AWS A2.4 symbols are written this way because the same CJP performance can be achieved with a single-V, double-V, or U-groove depending on thickness and process. The shape is a means; penetration is the requirement.

This also explains why design capacity is tied to effective throat, not deposited metal. A PJP weld with a large cap still has the same effective throat as a PJP weld with a small cap. Adding extra weld on the surface changes appearance and cost, not strength. If you’ve ever watched a welder build a beautiful oversized cap on a PJP joint thinking it added strength, you’ve seen this principle ignored in real time.

When PJP Is Prohibited

Codes draw hard lines around where PJP may be used. Under Caltrans and AASHTO/D1.5 practice:

• PJP groove welds are prohibited in butt joints, except for milled compression splices.
• PJP is prohibited where tensile stress is normal to the effective throat — that unfused root becomes a crack-like discontinuity under direct tension.
• Minimum effective PJP sizes are set: 1/4 in for plate thickness T < 3/4 in, and 5/16 in for T > 3/4 in.

The logic is straightforward: the unfused root of a PJP weld is a built-in stress raiser. In tension pulling directly on that throat, it will propagate. In shear or compression, the surrounding geometry can often carry the load safely around it. That’s why PJP shows up constantly in shear-tab beam connections and similar details, and almost never in tension-critical butt splices.

Making the Call in the Shop

As a welder or fitter, you don’t choose CJP versus PJP — the engineer does. Your job is to read it correctly, confirm it against the WPS, and execute it. Check the WPS in Pipe Welding Procedures: The Complete Guide to WPS, Positions, Passes & Inspection, and when a CJP is called for, remember that achieving it often starts well before the arc strikes — with fit-up, root opening, and back gouging covered in our defect prevention guides.

CJP and PJP aren’t jargon. They’re the load path, written in weld metal. Read them right, and every other groove weld decision falls into place.

Welding Processes, Positions and Techniques for Groove Welds

A groove weld is only as good as the process behind it. You can cut a perfect bevel, hit the root opening the drawing called for, and still get lack of fusion if the process, parameters, and position don’t match the joint. Here’s how the five common groove processes compare, followed by positions, multi-pass technique, and the heat controls that keep a groove weld code-compliant.

SMAW, GMAW, FCAW, GTAW and SAW: Picking the Right Process

GMAW is the shop workhorse for groove welds. A typical prequalified D1.1 procedure for carbon steel uses ER70S-6 wire on DCEP spray transfer at 190–280 A, 22–30 V, and 350–600 ipm of wire feed — wide enough that you still have to dial in to your actual thickness and position. Setting up from scratch? Our How to Set Up a MIG Welder guide walks through gas, polarity, and parameter selection, and AWS D1.1 is where the prequalified procedures live.

SMAW remains the field default because nothing beats a stick holder for reach and portability. The NCDOT double-V groove WPS calls for 1/8 to 5/32 in E7018 low-hydrogen electrodes at roughly 90–150 A for a 1/8 in rod. Electrode diameter drives amperage more than anything else, so check our 5/32 E7018 amperage guide before stepping up rod size on a fill pass, and What Are Electrodes in Welding? if the E7018 numbering still looks like alphabet soup.

FCAW splits the difference: it penetrates like stick, runs like MIG, and tolerates the mill scale and wind that would wreck a gas-shielded arc — see Flux Core Wire Welding Tips. GTAW is the choice for the root pass on pipe and critical alloy joints where you need a flawless, slag-free tie-in, as covered in our TIG sheet metal guide. SAW dominates flat butt grooves in shipyards and tank work because its deposition rate is unmatched, but it is essentially flat-position only and needs a clean flux bed to do its job.

Welding Positions and Vertical-Up Progression

Positions are designated 1 through 5: flat (1), horizontal (2), vertical (3), overhead (4), and fixed-position (5) for pipe. Plate groove welds qualify as 1G through 4G, pipe as 5G (fixed horizontal) and 6G (fixed at 45°). Flat groove welds are the cheapest welds you can buy — gravity helps fill the joint, spatter stays put, and travel speeds climb. Overhead is the opposite: the puddle wants to fall, so you shorten the arc, tighten your weave, and keep heat down. Our Overhead Stick Welding (4F & 4G) guide covers the technique, and Pipe Welding Procedures breaks down positional pass sequencing for root, hot, fill, and cap.

Vertical-up progression is standard for groove welds thicker than about 5/16 in. Running uphill gives deeper penetration and a heavier, more controllable bead than downhill, at the cost of speed. The trade-off is worth it on structural work where penetration is the whole point; the downside is more heat input, which makes distortion and interpass temperature your next problem.

Multi-Pass Strategy: Clean Between Every Pass

Anything past roughly 1/4 in is a multi-pass job: root, optional hot pass, fill passes, then a cap. Before the arc strikes, remove coatings, rust, dirt, and mill scale within one inch of the weld area — contamination there migrates into the puddle as porosity or inclusions. Between passes, chipping and wire-brushing slag is non-negotiable, especially with FCAW and SMAW; trapped slag is one of the most common reasons a groove weld fails RT. Overlap the previous bead slightly and stagger stringer beads rather than running wide weaves, which pile heat into the center of the joint. Laying tacks to hold fit-up before the root? Our MIG tack weld guide covers spacing and sizing so the tacks don’t become part of the problem.

Preheat and Interpass Control

Heat control is where groove welds quietly pass or fail. The NCDOT procedure specifies preheat of 100 °F up to 3/4 in thickness, rising to 225 °F for material over 2½ in, with a maximum interpass temperature of 450 °F. Preheat slows the cooling rate, drives off moisture that would otherwise crack the weld, and helps the root fuse; interpass limits keep you from overheating the HAZ. Measure with a contact thermometer or thermal stick — never by eyeball or by how warm the plate feels. Cleanliness ties in directly: E7018 and other low-hydrogen electrodes must stay in a holding oven, and a damp or rusted rod defeats every preheat calculation you made. For the wider heat-and-hydrogen picture, Caltrans’ preheat and hydrogen control memo is a solid companion read, and this open welding textbook chapter walks through process and position fundamentals.

Pro tip: treat your WPS as the floor, not a suggestion. If the plate is cold, the ground clamp is on painted metal, or the interpass temp is already at 450 °F, stop and fix the condition. Heat management is the one groove weld variable you can’t grind out afterward.

Common Groove Weld Defects, Causes and How to Prevent Them

A groove weld can look perfect on the surface and still hide a crack starter underneath. The bead is shiny, the cap is even, the inspector’s tape measure passes the profile check. Then the assembly goes into service, cycles a few thousand loads, and a crack grows from a root that never fused. That is why defect knowledge matters more in groove welds than in almost any other joint: the geometry hides the evidence.

Under the ISO 6520 classification system, welders and inspectors work with roughly ten common groups of imperfections: cracks, incomplete penetration, lack of fusion, undercut, porosity and blowholes, spatter, overlap, inclusions, and geometric imperfections such as warpage and underfill. Each one has a cause, a signature, and a prevention rule. Learn them here and you will stop guessing when something goes wrong.

Incomplete Penetration: The Groove Weld’s Worst Enemy

Incomplete penetration (also called lack of penetration) is the most detrimental defect in groove welds. It acts as a stress raiser, a built-in crack tip sitting exactly where the load path is highest, right at the root. Once a structure starts cycling, cracks can originate there or propagate straight through the joint.

The causes are almost always setup or technique issues you can fix before the arc is even struck:

• Root opening too small for the electrode to reach
• V-angle too narrow, so the arc wanders and freezes before the root melts
• Electrode or wire oversized for the joint
• Travel speed too fast
• Current too low or heat input insufficient

The fix is equally practical: verify the fit-up against the WPS before welding, keep the arc pointed at the root, and if the joint is thick enough, run a root pass and then back-gouge to sound metal before capping. Our pipe welding procedures guide walks through back-gouging in detail, because on pipe it is standard practice for CJP work.

Lack of Fusion: Where the Weld Stops Sticking

Lack of fusion means the filler metal froze against base metal or against a previous pass without actually fusing into it. Common causes include heat input too low, wrong electrode angle, weaving too wide for the amperage, and a travel speed that lets the puddle run ahead of the joint edges. For stick work, running the rod too cold is a classic trigger, so check settings against our 3/32 E7018 amperage chart or the 5/32 E7018 settings guide rather than dialing by feel.

Porosity, Slag Inclusions, Undercut, Underfill and Overlap

Porosity and blowholes are trapped gas pockets. The usual suspects are excessive current, arc length too long, poor or disrupted shielding gas, moisture in the electrode or flux, and contaminated base metal (oil, rust, paint, mill scale). Prevention is 80 percent cleaning and gas discipline: grind the joint clean, store low-hydrogen electrodes in an oven, check your cable connectors and gun liners for leaks that let air in, and set the machine correctly using a proper MIG welder setup routine.

Slag inclusions happen when flux or oxide gets trapped between passes. Typical causes: inadequate interpass cleaning, too narrow a groove for the bead to wet the sides, wrong travel angle, and a first pass that overbuilds one wall. Wire brushing and chipping between passes, plus proper flux core technique, clear most of them.

Undercut is that groove melted into the base metal beside the weld toe. It reduces the section and creates a notch. Excessive current, too long an arc, too fast travel, and an incorrect electrode angle all produce it. For a deeper dive into acceptance limits and repair, see our undercut defects guide.

Underfill means the weld face sits below the required groove depth, leaving the joint thinner than designed. It is a coverage problem, usually from insufficient passes or travel speed too fast to fill. Overlap is the opposite issue: filler spilled over the base metal without fusing, caused by too little heat, wrong angles, or travel speed too slow. Both are profile defects that a careful visual inspection catches before the piece leaves the table.

Cracks: The One Defect You Never Talk Yourself Out Of

Cracks can be longitudinal, transverse, crater, or under-bead hydrogen cracks. Causes include high restraint, too-fast cooling, hydrogen from wet consumables, and improper cooling of the previous pass. Preheat and interpass temperature control are the main defenses, especially on higher-carbon steels.

The pattern across every defect on this list is the same: most groove weld failures are decided at fit-up and settings, not at the final pass. Keep the right electrode at the right amperage, verify joint geometry, clean between passes, and control heat input. Defects rarely appear when the fundamentals are handled in order.

Inspecting Groove Welds: Visual, RT, UT, MT and PT Explained

A groove weld can look textbook-perfect on the surface and still hide a cold root that never fused. The cap is the only part anyone sees, but the load travels through the root. That gap between appearance and reality is exactly why inspection exists: it is how you prove the weld is sound instead of just tidy.

Visual Inspection: The First Gate, Every Time

Every groove weld gets looked at before any machine does. Visual inspection (VT) catches underfill, overlap, undercut, excessive reinforcement, misalignment, and visible porosity in seconds, and it costs almost nothing. The AWS B1.11 Guide for the Visual Inspection of Welds is the standard reference here: it walks inspectors through detecting porosity, lack of penetration, undercut, underfill and incomplete fusion using proper lighting, magnification, and inspection technique.

Two things trip people up. First, B1.11 is a guide — it tells you how to look, not what to accept. Acceptance criteria come from the governing code, usually AWS D1.1 Clause 8. Second, visual work isn’t unskilled work: D1.1 requires inspectors to meet visual acuity standards and to be qualified for the level of inspection the project demands, and most specs call for an AWS Certified Welding Inspector. In pipe work, visual acceptance is woven directly into the Pipe Welding Procedures: The Complete Guide to WPS, Positions, Passes & Inspection workflow, where each pass is checked before the next one covers it.

Surface Methods: MT and PT

Visual stops at the surface — magnetic particle (MT) and liquid penetrant (PT) go a step further by revealing discontinuities the eye can’t see.

• MT magnetizes the part and applies iron particles that cluster where magnetic leakage occurs. It finds surface and near-surface cracks, but only works on ferromagnetic materials — steel yes, stainless, aluminum and copper no.
• PT uses a penetrant that seeps into surface-breaking openings, then bleeds back out after developer is applied. It works on any nonporous material, but it ignores anything buried below the surface and struggles on rough or weld-spattered faces.

Both are specified heavily for root passes and for crack-suspect areas. Surface-breaking undercut is a classic MT/PT finding that visual inspection may flag but cannot fully characterize.

Volumetric Methods: RT and UT

When the concern is what is happening inside the groove, you need volumetric testing.

• Radiographic testing (RT) passes X-rays or gamma rays through the weld onto film or a digital detector. It produces a permanent record and is excellent at porosity, slag inclusions, and — critically for groove welds — lack of root penetration and fusion. D1.1 Clause 8 addresses RT specifically, including RT of groove welds in butt joints. The introduction to welding and NDT from CED Engineering is a good plain-language primer on how these methods complement each other.
• Ultrasonic testing (UT) bounces high-frequency sound waves through the weld and reads the echoes. No radiation, fast setup, deep reach — but it is technique-sensitive, harder to document, and needs a skilled operator to interpret indications.

Acceptance Criteria: What Actually Counts as a Reject

Codes, not inspectors, decide pass or fail. D1.1 Clause 8 lays out discontinuity limits, and those limits shift with the loading condition: cyclically loaded connections get separate acceptance criteria for members in tension versus compression, with tension zones held to tighter tolerances because cracks grow faster under reversed loading.

DOT projects add their own radiography rules. Under Caltrans DOT RT requirements, hole-type image quality indicators (IQIs) must be placed on the source side of the weld, each exposure is limited to a maximum 24-inch weld width, and if a rejectable defect is found, the entire weld must be re-examined — not just the flagged spot.

The practical rule: start with visual, then specify MT/PT for surface integrity and RT or UT when the design demands verified full penetration. Choose the method the drawing and code require before welding begins, not after the metal has cooled.

Safety, Common Mistakes and Pro Tips for Better Groove Welds

Groove welding concentrates heat, fumes, and molten metal inside an open joint, which is exactly why it punishes sloppy habits faster than a fillet weld. Sound geometry earns you a strong joint. Disciplined safety and consistent technique keep you in the trade long enough to build the next one.

PPE: The Non-Negotiables

OSHA requires a filter shade matched to your arc and current under 1910.133(a)(5), and AWS F2.2 gives the shade-selection detail. Thin metal at low amps wants a lighter lens; heavy multi-pass groove work at high amps wants a darker one. Under the hood, wear safety glasses at all times, plus a flame-resistant jacket, leather sleeves and bib, gauntlets, safety boots, and hearing protection for the din of a shop running SAW.

Burns are not a rare accident. The Bureau of Labor Statistics counted 370 lost-time welding and heating-tool injuries in 2018, and more than a third of them — 140 — were thermal or heat burns. Most were avoidable with a buttoned collar and a proper cuff. Our welding PPE guide covers every layer in depth, and if spatter has already got through, weld burns on arms covers first aid and prevention.

Ventilation and Confined Spaces

Groove joints are deep and narrow, and deep, narrow welds trap fume. Use local exhaust or fume extraction right at the arc; an open door is not ventilation. OSHA’s general welding, cutting, and brazing standard puts the duty squarely on the employer to remove fumes and keep oxygen at safe levels. Confined spaces raise the stakes further. Shielding gases such as argon and CO2 are heavier than air, they displace oxygen, and a leaking cylinder can suffocate a welder in minutes. Test the atmosphere before entry, keep a spotter at the opening, and never purge a confined space with shielding gas while anyone is inside. Our MIG welding safety precautions guide walks through gas handling in detail.

Hot Work and Fire Prevention (NFPA 51B)

NFPA 51B is the reference standard for fire prevention during hot work, and OSHA points to it from Subpart Q. In practice, that means clearing combustibles within 35 feet, protecting what cannot move, assigning a fire watch during and after welding, and re-inspecting the area once you stop. Cutting a groove into a painted or oil-soaked plate adds fuel on top of fuel, so clean the joint before you strike an arc.

The Mistakes That Ruin Groove Welds

• Wrong groove angle for the process. A wide V that suits SAW starves a GTAW root.
• Skipping the back gouge. You leave incomplete fusion on the far side and a built-in crack starter.
• No interpass cleaning. Slag trapped between passes becomes an inclusion, not a layer.
• Ignoring preheat. Thick, restrained steel quenches too fast and cracks. Follow the WPS number.
• Welding without a WPS. If you do not have the procedure, you do not know the parameters. Our pipe welding procedures guide shows how a real WPS reads on the shop floor.

Pro Tips Checklist

1. Inspect and fit up before welding: gap, land, and angle first.
2. Tack in the planned sequence to keep distortion under control.
3. Hold travel speed steady, especially on vertical-up roots.
4. Clean every pass and watch the toes against the acceptance limits in AWS D1.1, undercut included.
5. Log heat input and verify it with a thermometer, not a glove.

Do these five things and most groove-weld defects never reach the inspector. The next section explains how that inspection actually works.

Conclusion: Put Groove Weld Knowledge Into Practice

You have now walked the full path of the groove weld, from the definition on the drawing to the final acceptance stamp. The good news is that nearly every groove weld failure you will ever meet traces back to one of five decisions made early. Get those five right and the rest of the process follows almost mechanically.

The Five Decisions, in Order

1. Define the load path first. Before anyone picks up a torch, the design has to answer one question: does this joint need complete joint penetration or will partial joint penetration carry the load? Designers specify CJP or PJP for a reason, and as we covered in CJP vs PJP, the shape of the groove is only the means, never the requirement.
2. Choose the preparation to match. Only after the penetration requirement is set do you select square, V, bevel, J, U, or a double variant, along with groove angle, root opening, and root face.
3. Pick the process and position. SMAW, GMAW, FCAW, GTAW, or SAW, and the welding position, drive how much heat you can put into the joint and how fast you can fill it.
4. Work to a qualified WPS. Preheat, interpass temperature, travel speed, pass sequence, and filler classification all live in the procedure, not in the welder’s memory.
5. Plan inspection before the first pass. Visual, RT, UT, MT, or PT, and who signs off, should be written into the plan on day one.

Who Should Choose the Groove Details?

This is worth stating plainly: the fabricator or detailer is usually best placed to select the specific groove geometry, because they are the ones who have to manage heat input and distortion on the shop floor. Caltrans makes exactly this point in Memo 12-5, reminding designers to leave preparation details to the fabricator unless a specific joint is structurally essential. A detailer who understands distortion will happily trade a wider groove angle for a symmetric double-V that keeps the plate flat.

Your Action Checklist

• Confirm whether the drawing calls for CJP or PJP, and why.
• Verify the groove type, angle, root opening, and root face against the WPS.
• Check fit-up tolerance and cleanliness before tacking.
• Confirm preheat and interpass temperature with a calibrated thermometer.
• Run the specified process, position, and pass sequence exactly as qualified.
• Back-gouge and re-weld where the procedure requires it.
• Schedule visual and NDT acceptance criteria before production starts.
• Document everything: heat numbers, preheat logs, operator IDs, NDT reports.

Keep Learning From Trusted Sources

Three documents deserve a permanent place on your shelf or bookmark bar. AWS D1.1 covers groove weld qualification, inspection, and acceptance. AWS A2.4 is the authority on symbols, and AWS B1.11 guides visual examination of welds. OSHA’s 1910.252 sets the baseline for cutting, welding, and brazing safety in general industry.

On weldgallery.com, keep going with our guides to pipe welding procedures, weld undercut causes and acceptance limits, and welding PPE. If electrode selection still feels intimidating, our electrode guide breaks down the classification system in plain language.

Final thought: a groove weld looks simple from the outside, two plates, one bead. Everything that makes it strong is hidden inside the joint, decided long before the arc is struck. Make those decisions deliberately, and the inspection report takes care of itself.

Frequently Asked Questions

What is a groove weld in simple terms?

A groove weld is weld metal deposited into an opening or ‘groove’ cut between two workpiece edges or surfaces. It is the weld you make in a butt joint after beveling the plate edges, and it applies to butt, tee, corner, lap and edge joints.

What is the difference between a groove weld and a fillet weld?

A groove weld is deposited IN the joint between the members, while a fillet weld is deposited ON the joint, joining overlapping edges at an angle. Groove welds carry load through the plate thickness; fillet welds carry load along the edge surface.

What does CJP mean on a groove weld symbol?

CJP stands for complete joint penetration — the weld runs fully through the member thickness so the joint can develop the full strength of the connected parts. PJP (partial joint penetration) welds only reach a specified depth, defined by the effective throat E.

Why do welders cut a V or bevel into plate edges before welding?

The bevel lets the electrode reach the root so the weld fuses through the full thickness. A square butt edge would require far more current and travel control to achieve penetration, which is why prepared grooves are used on anything beyond thin material.

What groove angle should I use?

A common included groove angle is 60° (30° per member for a V-groove), which is the standard for many prequalified AWS D1.1 details. Note that if the groove angle is under 60°, AWS D1.1 reduces the effective weld size by 1/8 inch per side because full root fusion cannot be assumed.

What causes lack of penetration in a groove weld?

The usual culprits are a root opening that is too small, a groove angle that is too tight, an electrode that is too large, travel speed that is too fast, or heat input that is too low. It is one of the most damaging groove weld defects because it acts as a stress raiser.

How are groove welds inspected?

Inspection starts with visual examination against AWS B1.11 and D1.1 Clause 8 acceptance criteria, then moves to NDT — radiographic testing (RT) for butt groove welds, ultrasonic testing (UT), plus magnetic particle (MT) or penetrant (PT) testing for surface-breaking defects.

Which welding process is best for groove welds?

It depends on thickness, position and access. SMAW with E7018 is the workhorse for field and heavy groove welds; GMAW/FCAW suit production and out-of-position work; GTAW is used for root passes on pipe; SAW is efficient for long, flat, heavy groove welds.

Do groove welds need a WPS and a qualified welder?

Yes — code work under AWS D1.1 requires either a prequalified welding procedure specification or one qualified by testing, and welders must be qualified for the process, position and joint type they are welding. Keep records; inspectors will ask for them.