Groove Welding Definition: Types, Symbols, and Process

A fabricator clamps two half-inch steel plates together, runs a bead across the joint, and walks away thinking the job is done. Three weeks later the inspector taps the weld with a hammer and sends it back: there was no groove, so the weld never had a chance to reach full penetration. That callback is exactly why “groove welding” shows up on nearly every structural drawing, pipe spool diagram, and pressure vessel plan you will ever read.

Groove Welding Definition: What It Is and Why It Exists

Start with the definition, because it is cleaner than the way most people say it out loud. A groove weld is a weld made in a prepared channel or opening between joint members so that filler metal can be deposited into the joint. The American Welding Society states it precisely in AWS A3.0: a groove weld is “A weld in a weld groove on a workpiece surface, between workpiece edges, between workpiece surfaces, or between workpiece edges and surfaces.”

The matching definition for the space itself matters just as much. A3.0 defines a weld groove as “A channel in the surface of a workpiece or an opening between two joint members providing space to contain weld metal.” Read those two sentences together and the whole technique becomes obvious: you prepare a channel, you fill it. That is groove welding.

The same standard breaks the family down into nine types of single groove welds and seven types of double groove welds. “Single” means the joint is prepared on one side only, “double” means both sides are prepared. The names behind those counts (square, V, bevel, J, U, and their double variants) are the vocabulary you need to read any groove weld drawing, and they are the catalog this article walks through later.

Why nearly every groove weld starts with a bevel

A butt joint made from two plates butted edge to edge gives the arc almost nowhere to go. To reach the root of the joint and melt through the thickness, the edges are usually cut at an angle through the full thickness of the workpiece. That cut is the edge preparation, or bevel. It removes material, opens the joint, gives the electrode a path to the root, and lets the arc force its way into the joint rather than washing across the face.

The exception is the square groove, where the edges meet square with no bevel at all. It is the simplest preparation to machine and the one with the least weld volume, which makes it a favorite for thinner material and for high-speed processes like laser, electron beam, and friction stir welding. On thick plate it becomes hard to achieve full penetration without extra heat and extra passes.

Two halves of the same job

Worth separating, because shops often blur the terms. Keyence draws the line clearly: grooving (or groove machining) is making the beveled opening, typically on a beveler, a milling machine, or a thermal cutter. Groove welding is depositing weld metal into that prepared opening. One is metal removal, the other is metal addition, and the quality of the finished weld starts with how accurately the first half was done.

Groove welds also work across the full range of joint configurations. They are used on butt, T, corner, lap, and edge joints, which covers everything from pressure vessel shells to structural tube connections to the flanges and hubs of machined parts.

Scale matters too. The Open Washington “Introduction to Welding” text calls groove welds “probably the most widely produced welds” in manual and semi-automatic arc welding. Manual stick welding still handles a large share of that work, and its strengths are well documented in our guide to why SMAW still wins for field repairs and dirty-shop conditions.

Why the technique exists at all

Because a groove weld does things a simple surface weld cannot:

• Full penetration. The weld metal can melt all the way through the joint thickness and fuse with the base metal across the section, rather than just sitting on the surface.
• Lower deformation. Less filler metal is deposited, and it is placed where the load actually is.
• High load capacity. The joint can develop the full strength of the base metal instead of being limited by a small throat of weld metal.
• Better heat distribution. Keyence puts it plainly: groove welding is essential where deformation of the base metal must be limited and high-strength welding is required.

Groove Weld vs. Fillet Weld: Which One You Actually Need

Mix these two up and the bill arrives later. One weld is a triangle on the outside of a joint; the other fills a channel cut into the metal. Choose wrong and you eat a rejected weld, a failed inspection, or a connection that quietly underperforms.

What a fillet weld actually is

A fillet weld is a triangular weld joining two surfaces that meet at roughly a 90-degree angle, like a bracket plate on a column flange. Its cross-section is a right triangle, one leg on each member, hypotenuse facing out. It is the most produced weld in the world and the default detail on nearly every steel drawing. For a refresher, this intro to weld types and joint configurations is a good start.

The critical dimension is the effective throat: the shortest distance from the root of the joint to the face of the weld, and the plane the weld is assumed to fail along. A 3/8 in fillet weld has an effective throat of about 0.265 in, not 0.375 in. Engineers design around it because the load path is shear along the throat.

What a groove weld actually does

A groove weld fuses base metal through the thickness of the joint. Edges are prepared, usually by beveling, to form a channel, filler metal goes into it, and the finished weld becomes continuous with the base metal. The load path runs straight through the material instead of around it.

Feature	Fillet Weld	Groove Weld
Joins	Surfaces meeting near 90 degrees	Butt, T, corner, lap, edge joints
Strength basis	Shear rupture along the throat	Base metal, CJP with matching filler
Edge prep	None	Square, bevel, J or U groove
Production	Fast	Slower, multi-pass
Inspection	Visual only	Visual plus NDT often
Uses	Shear tabs, cover plates, bracing, column bases	Moment connections, splices, HSS

The jobsite reality

Per Miller Electric’s guidance on structural field welding, fillet welds are the most common weld on structural projects and the easiest to produce. They are quick, any qualified field welder can make them, and they are judged by visual inspection alone. Most are done with self-shielded flux-cored or stick welding, which is why SMAW remains such a dependable field process.

Groove welds are a smaller share of the work but carry the load where it matters most: moment connections where a beam framing into a column transfers bending, column splices on tall buildings, and hollow structural section connections where load travels through the tube wall. Those are the joints inspectors examine hardest and welders are tested hardest on.

How each weld is actually sized

Fillet welds are designed for shear rupture along the effective throat. The AISC equation in plain language: weld strength = weld metal strength × effective weld area × a modification factor. Weld metal strength is 0.60 × electrode tensile strength, so E70 gives 42 ksi. Effective area is effective throat × weld length. The modification factor kds is 1 + 0.50 sin(1.5 × theta), which equals 1.0 for a 90-degree fillet and rises as the weld runs nearly parallel to the load.

Strength comes from adding metal, so a designer can over-weld a connection and waste shop hours.

A complete joint penetration (CJP) groove weld with matching filler works completely differently. As the AWS structural welding code and AISC make clear, a CJP joint with matching filler metal is controlled by the base metal, per AWS D1.1 Table 4.3 and AISC Table J2.5. No weld-specific calculation is required; the joint reaches the full capacity of the material it connects.

The trap in specifying CJP by default

Engineers like CJP because the answer is simple: full capacity, no math. But the Steel Tube Institute argues designers reach for CJP when simpler details do the same job. Fillet welds and partial joint penetration groove welds with fillet reinforcement can also deliver full-capacity joints. CJP should be the exception, because fit-up demands, welder skill, and qualification requirements make it harder and costlier to produce.

Tuning strength, and a word on laser welding

According to Keyence, joint strength can be adjusted by changing the bevel shape, its depth, its width, and the resulting joint area. Fillet welds cannot be tuned that way: the gap between base metals limits how load transfers, so a deeper or wider groove is the fix when a joint comes up short.

The same source notes groove welding pairs well with laser welding: narrower weld lines and lower heat input suppress deformation and residual stress, but accuracy demands more precise groove machining and tracer positioning. For a shop moving toward laser work, this primer on laser welding systems and safety covers the basics.

Bottom line: if load runs across the joint and the connection must act as one piece of steel, you need a groove weld. If it carries shear at a corner and you can add metal to reach the required throat, a fillet weld is faster, cheaper, and easier to qualify. Decide before the steel is cut.

The Anatomy of a Groove: Bevels, Root Faces, Angles and Gaps

Every groove weld starts as a shape. Before a torch lights up, someone has cut, ground, milled or gouged a precise opening between two pieces of steel, and every dimension of that opening has a name on the drawing. The symbol, the procedure specification and the inspection report all speak that language.

The five metal terms you need first

• Base metal is the parent material being joined, plate, pipe or tube.
• Filler metal is the consumable you add, electrode, wire or a preplaced strip.
• Weld metal is the solidified result of filler metal melting and combining with base metal.
• Weld pass is one progression of the welding process along the joint, one layer of deposit.
• Weld bead is the visible result of a pass or series of passes, with its own profile, width and reinforcement.

The same definitions run through AWS terminology and classroom references like the Washington State open welding textbook on groove weld preparation.

Bevel or chamfer: which one is on your drawing?

A bevel is an angled cut made entirely through the thickness, so the plate ends in a knife edge at the root. A chamfer is the same cut made only partway through, leaving a square corner behind. Most drawings just say “bevel” for any angled face, but the distinction matters when a detail calls for a partial-thickness edge prep, such as a square groove with a chamfered corner.

Each bevel carries four measurements:

• Groove face is the surface exposed by the preparation cut, where arc, filler and slag interact.
• Bevel angle is the angular dimension of that cut, measured from the perpendicular of the plate surface.
• Root face (also called a landing) is the unmelted land of base metal left at the joint root, a linear dimension such as 1/8 in.
• Groove depth (bevel depth) is how far the preparation reaches into the thickness.

Groove depth matters most. Assembled groove depth has the strongest influence on finished weld size and strength, so it drives filler and pass count.

How the groove angle is derived

In a single bevel groove, only one member is angled, so the groove angle equals the bevel angle. Bevel one side at 45 degrees and the groove angle is 45 degrees.

In a V-groove, both members are beveled, so the groove angle is the included angle between them, equal to the sum of the two bevel angles: two 30-degree bevels give a 60-degree groove.

The same logic runs through every groove family. A drawing that says “double V” shows two bevels facing each other, and each half can be specified independently.

Single versus double, and why the naming matters

The single or double prefix applies across the board: single V, double V, single J, double J, single bevel, double bevel. Double grooves cut the volume of weld metal you must deposit, and less deposit means less heat, less distortion and less time. The trade-off is harder machining, harder access and more complex inspection. Double V machining is fiddly, and a double bevel leaves the joint asymmetric, complicating welding and back gouging.

Radiused grooves: the J and U

J and U grooves replace the flat angled face with a curve. The radiused face spreads welding heat more evenly across the joint and reduces the stress concentration that drives cracking in hard-to-weld materials like high carbon steel, alloy steel and many stainless grades. They are harder to prepare by hand or machine, and many come about through back gouging, where a carbon arc gouging machine shapes the groove after the plate is set up. Our guide to air arc welding and carbon arc gouging covers that equipment and how it preps joints.

In a J or U groove the groove radius is the measured dimension, and when bevel depth exceeds the radius a groove angle is specified too, since the curve alone no longer defines the opening.

Flare bevel and flare V: usually an accident

A flare bevel or flare V groove has an outside radius equal to the bevel depth. The flare usually comes from the shape of the parts themselves, a cone meeting a flat plate for example, rather than deliberate preparation. The problem is that the groove narrows sharply at the root, which limits how far penetration can reach. Flare grooves are rarely intentional, so confirm the actual root opening before welding.

A cross-section map of a groove

Keyence’s groove geometry reference labels the dimensions of a prepared joint in a way that transfers to a shop drawing:

Letter	Dimension	What it tells you
A	Groove angle	Included angle between faces
B	Bevel angle	Angle of one cut edge
C	Plate thickness	Total member thickness
D	Groove depth	Depth of the preparation
E	Root face	Unmelted land at the root
F	Root gap	Gap between the members

Types of Groove Welds: Square, V, Bevel, J, U and More

Once you know a groove weld is simply a weld made inside a prepared channel, the next question is which channel to prepare. AWS A3.0 names a groove after the shape of that preparation, and those shapes fall into a handful of families that every shop, inspector and structural engineer recognizes instantly. Learn the five basics and most groove weld drawings on a shop floor stop being mysterious.

Square Groove

A square groove has edges that meet square, with no bevel at all. It is the cheapest preparation to make, needs the least filler metal, puts the least heat into the joint and therefore produces the least distortion. Square grooves are typically welded with a zero root gap (0 mm), which is exactly why they pair so well with laser welding, electron beam welding and friction stir welding, where a tight fit beats a wide open channel. The drawback shows up on thick plate: with no bevel to guide the arc, the filler has to bridge the entire thickness, so full penetration is genuinely hard to achieve.

V-Groove

The V-groove is the most common groove in structural, pressure and ship work. Both edges are beveled, forming a V that opens toward the welder, and machining is relatively simple, usually two straight cuts. The catch is distortion. The bead is asymmetric through the plate thickness, with most of the weld metal sitting on one side of the mid-plane, so angular deformation can be significant. Both the amount of welding and the resulting distortion grow with plate thickness, which is exactly why double-V preparation exists at all.

Bevel Groove

A bevel groove is prepared on one side only, leaving the other side square. It is the answer when the back of the joint is unavailable, for example a member butting into a wall or a plate welded too close to a flange to get a beveller behind it. You fill the prepared side and then weld from the back over the remaining land, so full penetration is still achievable. Machining is simple, but the bevel angle and root gap directly affect how the weld behaves, so both need to be controlled rather than eyeballed.

J-Groove and U-Groove

A J-groove has one radiused side; a U-groove has both sides radiused. Because the groove face is a curve rather than a sharp angle, heat spreads more evenly and the joint is far kinder to crack-sensitive materials. Both cut into less material and need less filler metal, and a U-groove delivers deep penetration with less filler to deposit, making it a mainstay in pipe welding, boilers, pressure tanks and heavy structural components. The trade-off is forming: a true radius needs a radius tool or a specialised bevelling head, and you cannot fake it with a hand grinder. It is a tooling decision, not a labour decision.

The Double Groove Variants

Every groove shape has a double version, and the naming follows the same logic across all of them. A double V-groove prepares both sides of the joint, cutting the weld metal needed and therefore reducing distortion, but the machining is difficult because both faces have to be bevelled and held in alignment. A double bevel groove is simple to machine, yet the asymmetric groove makes welding and back-chipping awkward, since one side has to be gouged back before the second side can be welded. A double J-groove cuts a radius on both sides and is the classic way to reduce weld volume on very thick plate.

Groove	What Gets Prepared	Best Use	Machining Difficulty	Main Drawback
Square	Edges meet square, no bevel	Thin material, laser/EBW/friction stir, zero gap	Lowest	Full penetration hard on thick plate
V	Both edges beveled	General fabrication, most structural and pressure work	Relatively simple	Angular deformation from asymmetric bead
Bevel	One side only	Back side inaccessible, full penetration still required	Simple	Bevel angle and root gap affect workability
J	One side radiused	Less material removal, crack-sensitive base metal	Difficult, needs radius tooling	Special equipment for the radius
U	Both sides radiused	Pipe, boilers, pressure tanks, heavy structures	Difficult, needs radius tooling	Two radiused cuts to form
Double V	Both sides V-beveled	Thick plate where distortion is the enemy	Difficult	Alignment and machining effort
Double Bevel	One side beveled each face	Thick plate, two-sided access	Simple	Asymmetric weld makes welding and back-chipping difficult
Double J	Radius both sides	Very thick plate, minimum weld volume	Difficult	Radius tooling and fixturing

What Thickness Can a Groove Actually Handle?

Penetration capability is where the choice stops being cosmetic. According to Keyence’s guide to welding grooves, V-grooves and U-grooves can achieve full penetration even with thick plate, with no theoretical upper thickness limit, whereas square grooves have a thickness limit for full penetration. Their worked example makes the point vividly: in a TIG operation on 7 mm (0.28 in) plate, a square groove gave no penetration to the opposite side, while a V-groove produced a back bead on the reverse face. That back bead also depressed the top surface, so a second layer had to be welded to restore it. Two extra figures from the same source place the numbers in context: thick plate means 6 mm (0.24 in) or more, and ultra-thick plate means 150 mm (5.91 in) or more. Thin-plate welding problems are a different discipline, which is where resources like aerospace TIG welding techniques become worth studying.

The Five Types Beginners Should Recognize

Lincoln Technology frames it neatly in its comprehensive guide to groove welding: the five basic groove weld types a new welder needs to recognize are the square groove, V-groove, U-groove, J-groove and bevel groove. Master those five, understand that double versions cut heat and distortion at the cost of harder machining, and most groove weld symbols on a shop drawing stop being mysterious.

Complete vs. Partial Joint Penetration (CJP vs. PJP)

Penetration separates a joint that carries the full load of the members it connects from one that is only as strong as the metal someone deposited in the groove. Get this right and groove weld design becomes simple. Get it wrong and the connection is under-designed from day one.

Two Penetration Terms to Keep Separate

Root penetration is how far weld metal reaches into the joint root, the tightest part of the preparation where the two members meet. Joint penetration is the depth of weld metal past the base metal surface: the groove or bevel depth plus any root penetration beyond it.

In a groove weld, size and joint penetration are the same dimension. A one-inch groove weld must have one inch of joint penetration. That is not a nominal label, it is the measured depth of weld metal, and the groove weld symbol calls that dimension out directly.

CJP: Where the Base Metal Governs

A complete joint penetration weld melts all the way through the thickness of the joint, leaving one continuous piece of weld metal fused to the base metal. With matching filler metal, the joint is then controlled by the base metal rather than the weld.

AWS D1.1 Table 4.3 sets the matching-strength rules and AISC Table J2.5 sets the limit states. E70 filler metal, with FEXX = 70 ksi, counts as matching for ASTM A500 Grade C HSS at 50 ksi yield and for ASTM A1085 HSS at 50 ksi yield. A matching CJP detail between those members is automatically full capacity: no weld-specific calculation, no reduced allowable stress.

For hollow sections, AWS D1.1 Clause 10.10 covers the CJP groove weld requirements for HSS-to-HSS T-, Y- and K-connections. The question there is not whether the weld is strong enough, it is whether the detail you drew is one the standard permits. These connections also mean welding in multiple positions, so filler selection and welder qualification become part of the design instead of an afterthought. Our all-position welding rod guide covers the consumable side of that call.

PJP and the PJP-Plus-Fillet Detail

Partial joint penetration stops partway through the thickness. On its own, a PJP weld is designed around its own effective throat, far less than the plate thickness. You can make a PJP detail behave like a full-strength connection by adding a fillet weld, and AWS D1.1 Clause 4.4.2.7 states: “The effective throat of a combination PJP bevel groove weld and a fillet weld shall be the shortest distance from the joint root to the weld face of the diagrammatic weld minus 1/8 in. [3 mm] for any groove detail requiring such deduction.”

That 1/8 in. subtraction is what engineers call Z-loss: the shortest path from the root to the weld face does not always pass through sound weld metal, so the standard trims it back before you calculate with it.

The Geometry Numbers That Must Be Right

Both CJP and PJP groove welds can be prequalified when the joint geometry falls inside the AWS D1.1 limits. For a T-joint welded by FCAW or GMAW, with the AWS tolerances in parentheses:

Detail	Root opening R	Groove angle	Root face landing
CJP with backing	1/4 in. (3/16 to 1/2 in.)	45° (40° to 55°)	Root closed by backing
PJP or PJP plus fillet	0 (-1/16 to 1/8 in.)	45° (40° to 55°)	f = 1/8 in. min. (1/16 to 3/16 in.)

Read that carefully: the PJP detail wants a closed root, zero opening, a real land, and a 45 degree groove angle. Drift outside those numbers and the detail is no longer prequalified, which means a qualified test coupon and a slower approval process.

A Worked Example: HSS Connection

The Steel Tube Institute worked example shows why designers should not default to CJP. The section is an HSS 7.000 x 0.375 ASTM A500 Grade C with Fy = 50 ksi, Fu = 62 ksi and t = 0.349 in. A 7/16 in. fillet leg on the plate and the HSS gives an effective throat S = 0.468 in., with phi = 0.9 for HSS yield, 0.75 for fillet weld shear rupture and 0.80 for the fillet-reinforced PJP weld.

The headline is that a fillet-reinforced PJP detail can often match the capacity of a fillet-only design, so you are not forced into an expensive CJP detail. But note the fine print: the kds = 1.5 strength enhancement factor available to a round HSS fillet weld does not apply to the PJP-plus-fillet case, so the two options must be compared on the same basis.

How Groove Welds Are Made: Processes, Passes, Positions and Backing

What decides whether a groove joint works is how it gets made: the process, the passes, the root, and the support under the puddle.

Groove welds come out of manual arc welding (SMAW) and the semi-automatic processes, GMAW and FCAW, and from GTAW (TIG), submerged arc welding (SAW), laser and electron beam welding. What changes job to job is heat input control, root depth, and how well a molten pool holds its place when the position gets awkward.

Passes and Beads: How Many Welds Are in a Groove Weld?

A pass is a single progression of the welding operation along the joint. A bead is the deposited weld metal left behind by that pass. One pass makes one bead. A thin bracket groove may need only that, while a thick column splice might take a root pass, three or four fill passes, and a cap pass.

Multi-pass welding is not sloppy workmanship. It keeps heat input per unit length low enough that the plate does not twist, and places metal where it is needed.

The Root Pass Does the Heavy Lifting

The root pass decides whether the joint works. It is the first bead in the groove, setting three things at once: penetration depth, root reinforcement behind it, and the geometry later passes follow.

Get it right and the fill passes are forgiving. Get it wrong and a grinding wheel cannot save you. Poor penetration, excessive root gap or a cold start create defects that must be gouged out and rewelded, slow and expensive.

Back Gouging and the Back Bead

On a V-groove, the bead on the opposite side of the plate is the back bead. Keyence’s guide to groove welding problems explains why: because the top surface becomes depressed when the back side is welded, a second layer must be welded to restore sufficient strength.

So many V-groove joints get gouged out on the back side first. The welder removes the underside bead with a carbon arc gouging set, grinds it flush, and welds a fresh back bead into the depression. The result is full penetration with reinforcement on both faces.

Backing: The Support You Do or Do Not Leave Behind

Backing is the plate, bar or strip under the groove that supports molten root metal. AWS D1.1 Clause 7.9.1.2 says steel backing should be used whenever possible and be continuous for the length of the joint. Interrupted backing leaves gaps where root metal drops through or the arc burns through.

Whether it stays depends on loading. Backing may be left in place on a statically loaded weld. Under cyclic loading it cannot, and the weld must be made from one side without backing, a harder procedure because the root pass must hold its own shape under the arc.

Position Changes Everything

Groove welds are made flat, horizontal, vertical and overhead. Root pass position drives process choice, since that is where you most need metal to stay put. Vertical and overhead roots favour higher deposition rates and good pool control: FCAW run slow-travel and high-fill, or GTAW for precise control. The Lincoln Electric guide to groove welding is worth reading before committing a crew to an overhead splice.

Procedures, Qualification and Hold Points

Every production groove weld runs to a Welding Procedure Specification (WPS) covering process, filler metal, joint details, preheat, heat input limits and inspection. AWS D1.1 Clause 5 provides prequalified welding procedure specifications and joint details, which is how the Steel Tube Institute compares a CJP detail to a fillet-reinforced PJP detail.

The welder qualifies separately. AWS offers the Welder Performance Qualifier (WPQ1) and Welding Procedure Qualifier (WPQ2) certifications alongside CWI and SCWI credentials, and they map to who may run a joint in a position. Inspection holds sit between the steps: groove geometry before tacking, root pass before filling, finished profile before release.

Why Groove Welds Cause Scheduling Headaches

As Miller Electric notes on field welding structural steel point out, groove welds demand highly skilled welders, making resource allocation harder, and they become bottlenecks on a structural jobsite. You cannot simply add labour: the qualified welder, the beveling upstream and the inspection time downstream must all line up.

Book the welder and beveling capacity early, confirm the WPS before the joint is prepped, and treat the root pass as the milestone that gates everything after it.

Groove Weld Symbols: How to Read the Drawing

A groove weld symbol is a stack of engineering shorthand. Misread one and a shop will cut the wrong root opening, weld the wrong depth, and fail inspection.

Every symbol shares the same anatomy, per the OpenStax welding curriculum: the reference line, arrow line, tail, optional WPS information, and the groove weld symbol itself.

The Three Lines That Anchor Everything

The reference line is the long horizontal backbone. The arrow line points to the joint, telling you which member it applies to. The tail is the short leg at the opposite end, and it is optional; it carries the WPS number, the process, the electrode designation, or a shop note.

Which side of the reference line you read tells you where the weld goes: below describes the arrow side of the joint, above the other side.

Location	Meaning
Below the reference line	Arrow side of the joint
Above the reference line	Other side of the joint
Both sides	Groove weld called out on both sides

Two marks at the junction of the reference line and tail carry big meanings. A circle there means weld all around the joint, such as a ring around a pipe. A darkened flag there means the weld is performed in the field, not in the shop.

Two Kinds of Information on One Symbol

A groove symbol combines two things: the preparation, the groove shape itself, and the weld, meaning its type, depth, and the size of the deposited metal. The preparation shows as the house-shaped groove profile with its depth callout; the weld portion carries the size figure, effective throat, contour, and finish symbols.

Groove symbols also stack with fillet weld symbols on the same reference line, which is how a drawing calls out a partial joint penetration groove weld reinforced by a fillet. Read them as one combined detail, not two separate welds, as the Steel Tube Institute explains.

Reading the Callouts

Letters in the preparation portion specify the joint geometry before the arc is struck.

• G is the root opening, the gap between the workpieces. It gives the electrode access to the root and improves penetration.
• A is the groove angle, the included angle between the two prepared members. When only one member is angled, the symbol shows its bevel angle instead.
• D is the groove depth, measured from the base metal surface down to the root of the preparation.
• R is the root face, the square landing at the bottom of the bevel, and the radius callout appears on radiused J and U preparations.
• Effective throat T, in parentheses, appears only when the weld is not full penetration. It gives the welder and inspector the required penetration depth.

Face Contour and Finish Symbols

The face of a groove weld can be specified as flat, convex, or concave, each with its own symbol. Beside it there may be a finish symbol for dressing the surface: chipping (C), grinding (G), hammering (H), machining (M), rolling (R), or peening (P). Contour defines the face shape; finish defines the dressing allowed, so specified reinforcement is not ground away.

Two Worked Examples

A full penetration groove weld symbol with a welding depth of 5 mm (0.20 in.), a root gap of 0, and a groove angle of 70 degrees tells you the arc must fuse through the full 5 mm thickness, the edges meet with no gap, and the included angle is 70 degrees. Set up anything else and you weld a detail the engineer never approved, as the Keyence example shows.

Now take a PJP groove with a 6 mm effective throat in parentheses plus a fillet symbol. The groove does not have to go through, but the combined throat of groove plus fillet must reach 6 mm. That lets a shop get a full-capacity connection without the cost of CJP, the case the Steel Tube Institute makes before defaulting to CJP.

The tail holds the WPS reference and other notes: a specification number, a process code, an electrode or filler designation, and notes on backing, preheat, or inspection. The MTM guide to groove weld symbols covers the variations you will meet.

Read a symbol well and you can plan the beveling, set the root gap, run the right procedure, and know what the inspector will measure. That is the difference between a weld that passes the first time and one that returns three weeks later.

Common Groove Welding Mistakes and How to Troubleshoot Them

Ask any experienced fabricator and they will say the same thing: most bad groove welds start as a geometry problem, not a welding problem. The arc does what arcs do; the groove sent it somewhere else. So diagnose with a tape measure on the joint, not with the machine settings.

Start with the preparation dimensions

Preparation decides most failures before the arc is struck. Keyence’s groove welding troubleshooting reference maps each dimensional error to the defect it produces, and the mapping holds up well in real shops.

Preparation error	What it causes
Root face (landing) too large	The first layer does not achieve sufficient penetration.
Root face too small	The first layer melts too easily and burrs appear on the first pass.
Groove or bevel angle too narrow	Incomplete fusion or slag inclusion, which can lead to high-temperature cracking. The torch nozzle also starts contacting the base metal, making welding physically difficult.
Groove or bevel angle too wide	Melting occurs too easily on the first layer and burrs appear.
Root gap too large	Poor welding efficiency and large weld deformation.
Gap over backing too large	Melting and dripping of the root pass.

Read that table as a symptom list. An unfused line down the center of a V-groove points to a tight angle. A first pass that melts through and leaves a burr points to a narrow root face or an over-wide groove. Neither is a wire problem.

The classic defect list

Defect	What it looks like	What to do
Lack of penetration	Root pass visible as a cold line at the bottom of the groove, or no root bead at all on the back side.	Increase heat input within the WPS, tighten the root face, correct the groove angle, or add a backing bar.
Incomplete fusion	Dark, unfused slivers along the groove walls or at the root.	Open the groove angle, clean the sidewalls, slow down, and use proper travel angle and weaving.
Slag inclusion	Rolled-in dark lines or pockets within the bead.	Grind out and reweld, or add a cleaning pass between layers. Narrow grooves trap slag.
Porosity	Round gas holes, often clustered near the root or at stop points.	Clean the joint, verify shielding gas flow and nozzle distance, dry the filler, and check for drafts.
Undercut	A groove melted into the base metal at the toes of the weld.	Reduce travel speed slightly, adjust work angle, and avoid excessive weaving or grinding out of position.
Burn-through and root melt	Collapse of the root pass, spatter, holes, or a blown-out back bead.	Lower heat input, add backing, reduce root gap, and repair by gouging out the bad metal and depositing a sound pass.
Hot cracking	Fine cracks in the root or toe, usually straight and running along the weld.	Check restraint and fit-up, reduce heat input, use a more crack-sensitive filler, and preheat when the material calls for it.

Reinforcement, convexity and the weld profile

A little reinforcement above the plate is normal. How much is the question. Positive reinforcement (convexity) is not considered an effective part of the weld, because too much convexity can be detrimental, particularly in fatigue-loaded service where the extra metal becomes a stress raiser at the toes.

Grooves must be filled at least flush, and concavity may be acceptable or even desirable in some applications. Convexity or concavity, with joint penetration and weld width, forms the weld profile, which has significant importance in determining the quality and acceptability of the finished weld. Acceptance criteria in the AWS codes and standards are written against exactly these three elements.

Prep-related failures: thermal versus mechanical cutting

Thermal cutting (oxy-fuel, plasma, laser) changes the dimensions you measured ten minutes ago. The cut edge expands while heated and contracts as it cools, so re-measure groove angle and root gap after cutting, and remove slag stuck to the groove surfaces. Skip that and you are welding over inclusions that later show up as slag inclusion.

Mechanical cutting (beveler, milling machine) does not distort the metal, but it can leave lamination on the cut surface and deformation in strained areas. Inspect bevel angle, root width and surface roughness. Lamination is the one to watch: it will not clean up and it will not fuse.

A troubleshooting sequence that works

1. Measure before you weld. Check groove angle, root face, root gap and butt joint alignment with welding gauges, scales and straight edges. In mass production, 2D and 3D instruments measure groove shape instead.
2. Check heat input and travel speed against the WPS. If amperage sits below the range for the thickness in front of you, look at the joint before you touch the machine.
3. Confirm alignment and tacking. Root geometry problems are usually fit-up problems.
4. Re-inspect. Visual for profile, and NDT where the code requires it, before the weld is covered up.

Safety First: Fume Control, Heat, Distortion and Ergonomics

Safety comes first, ahead of joint preparation. Lincoln Electric’s comprehensive guide to groove welding puts it at the top of the agenda, with the practices behind it drawn from ANSI Z49.1 welding safety standards.

Groove welding also changes the safety picture compared with fillet welding. A fillet bead sits on a surface with the arc riding above the metal, while a groove weld drives the arc into a prepared channel, so more heat lands in a smaller volume and the fume plume per foot of joint is denser. That is manageable in an open bay with good air movement, and not manageable inside a vessel, a pipe pen, or a tightly framed structural detail.

Capture the plume at the source

Ventilation is a hierarchy, and source capture sits on top. A flexible exhaust arm or hood close to the arc pulls the plume away before it reaches the welder’s breathing zone. Below that sit general dilution ventilation and, for high-volume work, a booth or tent with filtered extraction. LEV wins because dilution depends on room volume, and a bay full of plate and steel shelving has no free air to spare.

• Local exhaust: flexible arm, hood, or downdraft table a few inches from the plume.
• General ventilation: adequate for low fume volumes, weak for enclosed or heavy fabrication.
• Enclosed or filtered booth: best practice for stainless, chrome, and coated or plated parts.

The chromium question

Stainless steel, chrome-plated parts, and many alloy steels put chromium in the plume. OSHA’s hexavalent chromium standards for general industry (1910.1026), shipyards (1915.1026) and construction (1926.1126) require employers to assess potential employee exposure to Cr(VI), then apply feasible engineering and work practice controls to bring exposure at or below the PEL. Respiratory protection supplements rather than replaces those controls, and general industry calls for a regulated area with limited access. The rules also cover protective clothing, hygiene areas, housekeeping, medical surveillance, and training.

The numbers in OSHA’s occupational chemical database are strict and sit far below what a worker’s senses will warn about:

Exposure limit	Value for Cr(VI)
OSHA PEL-TWA	5 micrograms per cubic meter
OSHA action level	2.5 micrograms per cubic meter
NIOSH REL-TWA	0.0002 mg/m³
ACGIH TLV-TWA (2017)	0.0002 mg/m³, inhalable particulate matter
Cal/OSHA PEL-TWA	0.005 mg/m³

What welding fumes do to a body

The NIOSH Pocket Guide to Chemical Hazards entry for welding fumes notes that no OSHA PEL exists for welding fumes as a category, so the hazard is controlled by other means. Exposure routes are inhalation and skin or eye contact. Symptoms include metal fume fever (flu-like onset with fever, chills, muscle pain, cough, and dyspnea) and interstitial pneumonitis. Welding fumes are listed as a potential occupational carcinogen with the lung as the cancer site, and target organs include the eyes, skin, respiratory system, and central nervous system. That is why the Monday cough deserves attention instead of a shrug.

PPE, hot work and fire watch

• Respirator: match type and filter class to the material and fume, and follow NIOSH recommendations for fitting, seal checks, and cartridge change schedules.
• Eyes and face: welding helmet with the correct shade number, safety glasses underneath, and face protection for grinding and chipping.
• Body: flame-resistant clothing, leather gloves, safety footwear, sleeves closed at the wrist. Arc UV burns show up on skin the shirt does not cover.
• Fire watch: hot work permits, a named fire watch, and checks of adjacent compartments for sparks and slag. Slag travels further than people expect.
• Ventilation and gas: check the room, cylinder, and hoses before striking an arc. Our guide to air gas welding supply, cylinders and safety basics covers storage and handling.

Heat input, distortion and residual stress

Heat is the second safety issue, because it hurts the part and the welder. A V-groove bead is asymmetric through the plate thickness, which produces significant angular deformation, and that distortion causes fit-up problems on the next joint in the assembly. Double-V, double-J and U grooves reduce the weld metal deposited and therefore the deformation. For crack-sensitive alloys, J and U profiles are the smart choice twice over: the radiused groove face spreads welding heat more evenly, and the smaller weld volume lowers both distortion and residual stress. Where you can choose the groove, choose the one that heats the part least.

Heat input management is mostly technique: sequence welds to balance shrinkage, back-step to even out heat, keep beads small and passes numerous rather than one heavy pass, control preheat per the WPS, and plan tack welds where they will be remelted and inspected.

Ergonomics

Groove welds in overhead, vertical, and confined positions ask more of the body than a flat fillet. Plan for position changes, keep the torch lead and fume arm from fighting each other, schedule breaks, and hydrate. Fatigue causes arc strikes, poor bead placement, and burns on the opposite arm when the sleeve rides up.

A Step-by-Step Workflow for Planning and Executing a Groove Weld

Groove welds are easy to plan badly and hard to repair later. This sequence keeps callbacks off your schedule: read first, prepare second, measure third, weld fourth, document fifth.

Step 1: Confirm the Requirement

Start at the drawing, not the torch. Find the groove weld symbol and write down what it demands: complete or partial joint penetration, effective throat, groove angle, root opening, root face, backing, and the field-weld flag. If the detail calls for CJP and the symbol reads PJP, stop and ask. Check terminology against the current AWS codes and standards: AWS A3.0M/A3.0:2025 for definitions and AWS D1.1/D1.1M:2025-AMD1 for structural steel work. Two minutes here saves a rejected weld later.

Step 2: Choose the Groove Type

• Square for thin material or laser and electron beam welding.
• V-groove for most general fabrication.
• Double V for thick plate where distortion is a worry.
• J or U for crack-sensitive material and thicker sections.
• Single bevel when the back side is inaccessible but full penetration is still required.

Step 3: Prepare the Edges

Lincoln Electric’s groove welding guide sets out the shop sequence plainly: know the hazards first, clean surfaces to remove oil, grease, rust and paint, pick the groove for the material thickness, then gather the machine, torch, angle grinder, clamps and safety gear. Nothing wrecks a root pass faster than a shop that ran out of clamps.

Thermal cutting (oxy-fuel, plasma, laser) is fast but leaves two problems: the heat-affected zone changes the dimension you just cut, and slag can stick to the face. Mechanical preparation (a beveler or milling machine) cuts clean but can expose lamination. Know which trade-off you accepted.

Step 4: Inspect Before You Weld

Keyence’s groove troubleshooting guidance is direct about it: inspect groove angle, root face and gap, plus butt joint alignment, before welding to prevent weld defects. Re-measure angle and root gap on thermal-cut grooves and check for adhering slag; on mechanically cut grooves check bevel angle, root width and surface roughness. Do not arc a joint whose geometry you have not measured.

Step 5: Clean and Fit Up

Remove oil, grease, rust, paint and moisture from the groove faces. Aluminum has its own rules, so prepping aluminum correctly before welding is worth reading if that is your material. Then clamp, align and tack exactly as the procedure specifies.

Groove width is a live trade-off. Too narrow and tracer control gets difficult, so defects get more likely. Wider and control gets easier, but you add welding metal and the deformation that comes with it.

Step 6: Run the Procedure

Root pass first, with the travel angle and technique the procedure calls for. It sets the penetration and root reinforcement everything else sits on. Then fill passes, gouging out defects where required, and a cap bead with controlled reinforcement. For SMAW root work, why stick welding still wins on structural steel is a useful reference, and air arc and carbon arc gouging covers back gouging technique.

Step 7: Inspect and Document

Check the finished profile visually against the drawing, run NDT where required (RT, UT, MT or PT), and keep the records: welder identification, WPS number, inspection results. That paper trail is what protects you at the final walk-around.

Conclusion

A groove weld is a weld made in a prepared channel between joint members so that filler metal can be deposited through the thickness of the joint. Choose it over a fillet weld when you need full penetration, moment capacity, or a joint that behaves like the base metal; choose a fillet when a shear connection will do. Safety and inspection are not extra steps at the end, they are part of the process.

Your next steps are short:

• Review groove weld terminology in AWS A3.0M/A3.0:2025 so your shop speaks one language.
• Read the WPS before the job, not during it.
• Measure the groove dimensions before the arc is struck.

Do those three things and Monday morning starts a lot smoother than the callback in the story above.

Frequently Asked Questions

What is groove welding?

Groove welding is a joining process in which the edges of two workpieces are prepared (beveled or otherwise shaped) to form a channel or opening, and weld metal is then deposited into that groove to fuse the joint. AWS A3.0 defines a groove weld as ‘a weld in a weld groove on a workpiece surface, between workpiece edges, between workpiece surfaces, or between workpiece edges and surfaces.’ The defining feature is the prepared geometry, not the welding process.

Is a groove weld stronger than a fillet weld?

A properly made complete joint penetration (CJP) groove weld can develop the full strength of the base metal, because the weld metal fuses through the full thickness of the joint. A fillet weld is designed around shear rupture along its effective throat, so it is generally the more economical choice for non-critical connections. Steel Tube Institute guidance notes that fillet welds and PJP-plus-fillet details can also deliver full-capacity joints, so CJP should be an exceptional specification rather than a default.

What is the difference between CJP and PJP groove welds?

CJP means the weld metal melts completely through the thickness of the joint, giving full penetration; PJP means penetration stops partway, leaving unmelted base metal at the root. AWS D1.1 Clause 4.4.2.7 requires a 1/8 in. deduction from the effective throat of a combination PJP bevel groove weld and fillet weld for groove details that need it, because penetration to the root is not guaranteed.

What is the most common type of groove weld?

The V-groove is the most widely used, because beveling both edges to form a V is relatively simple to prepare and works across a wide range of plate thicknesses. Single-bevel and square grooves are next most common, while J and U grooves are chosen for radiused preparation on crack-sensitive or thick material where distortion matters.

Do you always have to bevel both sides?

No. In a single bevel groove only one member is beveled, and the groove angle equals the bevel angle. A V-groove bevels both members so the two angles add up. In a J-groove only one side is radiused and in a U-groove both sides are radiused. Single-preparation details are common where the back side of the joint is inaccessible.

What is a root gap and how much should it be?

The root gap, or root opening, is the space left between the two members at the joint root; it gives the electrode access and helps penetration. For a prequalified AWS D1.1 FCAW or GMAW CJP T-joint with backing, the detail uses a 1/4 in. root opening with a 45 degree groove angle and permits 3/16 in. to 1/2 in. after fit-up. PJP details typically use zero root opening with a root face landing of at least 1/8 in.

Can you make a groove weld without a backing plate?

Yes, but it is harder. Steel backing should be used whenever possible and generally must be continuous for the length of the joint under AWS D1.1 Clause 7.9.1.2. Backing may stay in place on statically loaded welds, but under cyclic loading it cannot be left in place, and the weld must then be made from one side without backing, which is a significantly more difficult procedure.

What causes incomplete fusion and slag inclusion in a groove weld?

The most common cause is a groove or bevel angle that is too narrow, which prevents the arc and filler from reaching the side walls, and it can also make the nozzle contact the base metal. A root face that is too large starves the first pass of penetration, while too small a root face makes the first layer melt too easily and form burrs. Verify angle, root face and root gap with gauges before welding rather than after.

What safety rules matter most when groove welding?

Control the fume at the source with local exhaust positioned close to the arc, because OSHA’s fact sheet notes chromium is converted to hexavalent Cr(VI) during welding. OSHA’s Cr(VI) standards require exposure assessment and feasible engineering controls first, with respiratory protection as a supplement; OSHA’s permissible exposure limit for Cr(VI) is 5 micrograms per cubic meter as an 8-hour TWA. Wear proper eye and face protection, flame-resistant clothing, and use a hot work permit and fire watch on construction sites.