Butt Weld Symbol Explained: How to Read AWS Groove Weld Symbols on Drawings
Learn to read the butt weld symbol: AWS groove symbols, arrow-side rules, dimensions, CJP vs PJP, common mistakes,…
READ ARTICLE →A pipefitter once told me he lost a whole Saturday grinding a beautiful V into two 3/4-inch plates, only to have the inspector reject the job because the drawing called for a CJP groove with back gouging, not the bevel he cut. The weld looked perfect, the metal was sound, and it still failed the paperwork because nobody read the groove detail before the torch came out.
A pipefitter once told me he lost a whole Saturday grinding a beautiful V into two 3/4-inch plates, only to have the inspector reject the job because the drawing called for a CJP groove with back gouging, not the bevel he cut. The weld looked perfect, the metal was sound, and it still failed the paperwork because nobody read the groove detail before the torch came out.
That story is really a search story. People type “groove welding definition” for three different reasons, and most articles answer none of them properly. So here is the plain-English answer first: a groove weld is a weld made inside a groove — a channel cut into a surface, or the opening left between two members — where the weld metal sits down in that channel and fuses both sides. If you have seen two plates butted edge to edge with a V ground into them and then filled with weld metal, you have seen a groove weld. It is the joint that carries the load in most metal structures you drive past every day.
Quick Note: AWS A3.0, Standard Welding Terms and Definitions, 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.” A groove weld is simply a weld made in that groove. Two separate things: the groove is the prepared gap, the groove weld is the finished fusion of the members inside it.
Once you notice the pattern, you can see why one short definition never satisfies everyone:
Good groove welding definition coverage has to serve all three, because on a real job those three people are looking at the same joint from different angles.
Groove welds dominate structural steel, pipeline, shipbuilding and pressure vessel work for one blunt reason: a properly made groove weld can transfer the full thickness load of the members it joins. A fillet weld sitting on the surface of a lap or T-joint cannot do that reliably, because it only fuses the edge and the face. The Steel Tube Institute’s guidance on full-capacity welds makes the same point — a CJP groove weld develops the strength of the base metal itself, which is exactly what a bridge girder, a pressure part or a transmission line seam depends on.
That is why you will meet groove welds in:
By the end, you should be able to look at a drawing, name the groove, state the required penetration, and only then strike an arc — which is precisely what would have saved that pipefitter his Saturday.
At its core, the term has one clear meaning. A groove weld is a weld made in the groove between two members that have been joined edge to edge. The edges are beveled, squared, or otherwise shaped so that the weld metal can be deposited inside the joint, rather than resting on top of the surfaces. While that sentence is the foundation of any groove welding definition, it does not explain enough to read a drawing, interpret a Welding Procedure Specification (WPS), or pass an inspection. To do that, you need to know the anatomy of the joint itself.
Before dissecting a groove, it is important to separate two terms that are often confused. The joint is the geometric arrangement of the two pieces of metal prior to welding, while the weld is the deposited metal that fuses them together. In other words, the joint is the setup, and the weld is the result. There are five basic joint types in welding: butt, lap, corner, edge, and T. A groove weld can be applied to several of these, but it is not limited to any single one.
It also helps to distinguish the joint type from the weld type. A butt joint can receive a groove weld, while a T-joint may receive a fillet weld, or even a partial penetration groove weld with additional fillets. The same joint can utilize different welds depending on the design requirements. Keeping these two ideas separate is key to interpreting welding symbols correctly.
To speak the language of welding, you must be able to name each part of a groove joint. The following terms are the most common for a single-V groove butt joint:
Short Note: The root opening serves a vital purpose. With no gap at all, the arc will struggle to reach the bottom of the groove, often leading to a surface bridge instead of true penetration. A deliberate gap allows the arc and molten metal to reach the root, while also providing an escape path for gas and slag during the first pass.
Quick Note: While every job will vary, a common shop starting point for mild steel plate is a 60-degree included angle, a 1/16 inch (1.6 mm) root face, and a root opening between 1/16 and 3/32 inch (1.6–2.4 mm), increasing to 1/8 inch (3 mm) for heavier or out-of-position work. These are guidelines only. The drawing, the WPS, and procedure qualification always take precedence. If you’re looking to dial in your parameters, 3/32 E7018 amperage settings and technique can help get you into a workable range before testing on scrap.
Two sets of terms tend to trip up even experienced fitters:
The following sequence illustrates how all of these parts come together in a typical single-V groove:
While the cap is what is visible, the true quality lies beneath it. A joint can appear to have full penetration, yet still lack sidewall fusion. That is exactly why penetration and fusion are listed as two separate elements, both equally vital to the integrity of a groove weld.
Not all groove welds are created equal. Once you know the definition, the next question is which groove the joint actually needs. Standard welding terminology recognizes six families of groove geometry: square, V, bevel, J, U and flare. The V and bevel families each come in single and double versions, giving you eight shapes to reason about. Pick the wrong one and you either burn filler you didn’t need to buy or leave an arc that can’t reach the bottom of the joint.
A square groove is two straight edges butted together with no bevel at all. The sides stay parallel, and a root opening, if one is specified, is the only gap between them. It’s the cheapest preparation possible because there’s nothing to cut, but it only works where the arc can physically reach the bottom. That limits it to thin plate, roughly a few millimeters thick, or to thicker material welded from both sides where the second side is gouged or otherwise prepared to guarantee root fusion. On heavy plate in a single-sided square groove, you get a beautiful cap over a cold, unfused root. Distortion is modest simply because the deposited volume is small.
The single-V groove is what most people picture when they hear “groove weld”: two bevels meeting at an angle to form a V. It’s practical from about 6 mm up to around 20–25 mm of plate thickness. Beyond that the included angle makes the top of the groove very wide, and filler volume climbs fast because weld volume increases roughly with the square of the groove width. More filler means more heat input and more angular distortion, which is why thick plate work often skips the single-V entirely.
The double-V groove, often called an X groove, is the fix. Instead of one V through the full thickness, you cut shallow bevels on both faces and weld from each side. Each bevel only spans half the thickness, and since the width of a V grows linearly with depth, deposited metal drops to roughly half of the single-V equivalent. You also get less total heat input, roughly half the distortion, and shorter passes that are easier to control. The trade-off is a second setup: someone has to flip or roll the assembly, and the second side usually needs back gouging to confirm a sound root before capping passes go in. For plate above about 25 mm, the double-V almost always wins on cost once you count filler, grinding and straightening.
A bevel groove is a single-V with one straight edge, used where only one member can or should be beveled. You’ll see it constantly on T-joints and corner joints, where the flat plate stays square and the second member carries the angle. A single-bevel works in the same thickness band as a single-V, and a double-bevel mirrors the double-V logic for heavier sections. Volume, heat and distortion follow the same pattern: the wider the top of the groove, the more metal you’re buying.
On thick material, V profiles get punishing. A J-groove uses a narrow included angle with a radiused bottom; a U-groove uses a radiused bottom with slightly wider, near-parallel sides. Both cost more to prepare, requiring milling or a dedicated beveling machine, but they consume far less filler on heavy sections while still giving the arc reasonable sidewall access. That’s the whole trade: pay once for preparation, save repeatedly in deposited metal, passes, heat input and distortion. You’ll meet J and U grooves in heavy pressure vessel plates, thick structural nodes and large-diameter pipe where a V would mean dozens of passes and a stack of distortion problems. As the NSARC groove guide lays out, the profile is chosen to balance penetration, access and deposited volume at a given thickness.
Flare-V and flare-bevel grooves are the odd ones. Nobody cuts them. They come from the naturally rolled edges of structural shapes and cold-formed tubing: the curved side of a round HSS, the radius on a cold-formed angle, the rolled edge of a wide-flange flange. A flare-V appears where two rolled edges face each other, a flare-bevel where a rolled edge meets a straight member.
Because the “groove” is a curve rather than a machined bevel, you can’t treat it like a square, V or bevel joint. The effective throat is a calculated fraction of the member thickness, not the full groove depth, so the joint is sized by formula rather than by the visible gap. The Steel Tube Institute’s guidance on full-capacity HSS welds walks through exactly this, and it’s why a flare groove that “looks” generous often tests out weaker than expected.
Quick Note: Which groove fits which thickness?
One habit worth building: whenever you’re unsure what a groove symbol on a drawing means for fit-up and pass sequence, walk through the drawing properly before striking an arc. Our companion piece on pipe welding procedures shows how groove choice, pass sequence and inspection expectations lock together on real work. The geometry on paper is a load-bearing decision, not a suggestion.
Ask ten welders what the difference is and you’ll get ten versions of the same idea. Here’s the version that holds up in the field: a groove weld sits inside the prepared edge or gap between two members, while a fillet weld sits in the corner of an overlapping or perpendicular joint and is sized by its throat. One fills a slot you cut on purpose; the other lays in a corner like a bead of caulk. That single distinction drives everything else: how you prep the joint, how you size it, what the inspector will do with it, and what the job costs.
| Factor | Groove Weld | Fillet Weld |
|---|---|---|
| Geometry | Fills the prepared groove or gap between edges | Triangular cross-section in a corner |
| Where placed | Between abutting or butted edges | Inside corners of lap, corner and T-joints |
| How sized | Bevel depth, root penetration, effective throat | Leg length, with throat derived from it |
| Penetration | Full (CJP) or a specified depth (PJP) | Typically designed as fillet-only, no groove |
| Load transfer | Carries load through the full member thickness | Transfers shear across the corner |
| Cost per inch | Higher: edge prep, fit-up, NDT | Lower: fit-up and weld only |
| Inspection | Often volumetric (RT/UT) for CJP | Usually visual plus surface examination |
Every drawing you’ll ever see is built from five joint arrangements: butt, lap, corner, edge and T. Groove welds work in butt joints (the classic plate-to-plate case), corner joints, edge joints and T-joints, where a bevel or J-groove is cut into the stem member so the weld can reach down into the thickness. A lap joint normally takes a fillet weld, because the members overlap instead of butting; it only becomes groove territory when one edge is grooved so the weld bites into the plate beneath it.
That’s why the symbol matters. A groove weld symbol on a drawing normally carries the depth of bevel or root penetration written to the left of the groove letter — information a fillet symbol simply doesn’t carry. The AWS weld symbol chart shows this layout clearly: reference line, arrow, basic symbol on the arrow side or other side, and the size notation hanging off it (AWS A2.4 standard weld symbols). If your symbol has a depth number with a groove letter next to it, someone is asking for penetration into the plate, not a surface bead. For a refresher on reading symbols from scratch, Miller’s weld symbol guide walks through the same rules step by step.
Quick Note: When a drawing shows a bevel symbol and a fillet symbol on the same reference line, you are looking at a combination weld, not a duplicated callout. The groove goes in first to the specified depth, then the fillet caps the corner. The two are welded as one detail.
The classic confusion on the job is a partial-penetration bevel groove plus a fillet weld. The drawing shows both, and the welder runs only one of them — usually the fillet, because it’s faster. In this detail the effective throat is measured as the shortest distance from the root of the groove to the outer surface of the fillet, so the groove has to be there for the throat number to be true. Skip the groove and you’ve undersized the weld on paper and in the metal. Details like this are why combination and penetration callouts show up on lists of commonly misapplied welding symbols (STRUCTURE magazine).
Cost is the honest tiebreaker. Groove welds cost more per inch because of edge preparation, fit-up and inspection, but they are the only option when the joint must carry the full thickness load — which is exactly why designers reserve them for the connections that matter most (Steel Tube Institute, full-capacity welds). Before you decide which weld you’re actually running, confirm the call on the print and run it against our full groove weld vs fillet weld comparison, then check the joint details in pipe welding procedures — pipe is where groove and fillet decisions get written into the WPS rather than guessed at.
When working with a groove weld, the most important design decision isn’t the angle or the land. It’s penetration. A groove weld doesn’t always need to fuse the entire thickness of the joint, and whether it does or not changes everything from prep to inspection. The two levels you’ll see on a drawing are complete joint penetration (CJP) and partial joint penetration (PJP).
A CJP groove weld extends completely through the thickness of the components being joined, fusing the full cross section from one face to the other. When done correctly, this allows the joint to develop the full strength of the member itself. For structural applications, this is what qualifies it as a full-capacity connection, a concept outlined by the Steel Tube Institute for HSS and tube work.
This level of penetration is typically specified for tension-critical butt joints, main member splices, pipeline girth welds, pressure vessel seams, and anywhere cyclic or dynamic loading is present. Many agencies spell this out directly, with the Caltrans memo on welded connections being one example of how CJP requirements are enforced for structural steel.
Quick Note: CJP means the joint is no longer the weak link, not that the weld will outlast the base metal. If the member yields, a properly executed CJP weld is meant to hold.
Full penetration isn’t something to assume. It must be proven by one of three accepted methods:
Visual appearance alone isn’t enough. Proof of one of these methods is required to claim CJP.
A PJP groove weld is fused only to a specified depth of joint preparation, measured from the face of the member. The material below that depth is intentionally left unfused and is not a defect. This makes it a practical choice when full penetration isn’t needed for the load, saving weld metal, heat input, time and distortion.
You’ll often see PJP in lighter structural connections, built-up members or T-joints where only a portion of the thickness needs to transfer force. In many of these cases, it’s paired with fillet welds to build the required strength. When reading these joints, the AWS weld symbol chart is your first stop.
One of the most misread points under AWS D1.1 is the effective throat for a combination PJP bevel groove plus fillet weld. It’s the shortest distance from the root of the groove to the face of the fillet (minus any reinforcement), not the sum of the two throats. This directly affects your weld size calculation, so it’s worth double checking.
The penetration level drives how the weld is verified. CJP welds can’t be confirmed by sight alone, so volumetric nondestructive testing, such as radiography or ultrasonic examination, is almost always required. PJP welds, on the other hand, are frequently accepted by visual inspection with surface methods like magnetic particle or dye penetrant, since the unfused root is intentional.
This inspection gap is a major cost driver. For long seams, full volumetric NDT can outweigh the welding cost itself, which is why many pipeline and pressure applications fall under their own specifications, as outlined by the National Board and the North Shore Arc groove guide.
Cost isn’t the only factor, though. The unfused root of a PJP weld creates a built-in stress raiser. For static, non-reversing loads, it can be perfectly adequate. Under cyclic or fluctuating loads, however, cracks can initiate there and propagate. That’s why PJP groove welds are treated as fatigue-sensitive details in bridges, cranes and machinery. Combined with surface flaws like those covered in our guide to undercut causes and acceptance limits, that notch effect only compounds the risk. In many of these cases, CJP is the safer requirement.
Note: Before you strike an arc, confirm three things: whether the drawing calls for CJP or PJP, the specified depth, and the required method to prove it (backing, back gouging, or a qualified single-side procedure). It’s the simplest way to avoid a rejection.
Half the arguments on a job site never reach the welder. They happen over a drawing. Someone sees a V on the print, cuts a bevel, fills it, then the inspector points at the small number left of that V and asks why the root was never gouged. Welding symbols are a compact language: ten minutes of study pays back, and guessing costs you.
The symbol language used across North American fabrication is defined in AWS A2.4, Standard Symbols for Welding, Brazing, and Nondestructive Examination, and most shops tape a copy of the AWS weld symbol chart inside the trailer. Everything below follows that standard.
Strip a symbol down and it has four pieces:
Quick Note: The groove symbol sits on the reference line, not beside the arrow. On the arrow side means the symbol goes below the line; for the other member it goes above. If both sides are welded you get a symbol above and below, and the two are read independently.
Dimensions attach to specific places, and those places carry meaning. For groove welds, the depth of bevel, root penetration, or weld size is written to the left of the groove symbol. That single placement is the most misread detail on structural drawings.
Read it this way: a number left of the V describes how far the preparation or penetration goes. A number to the right of a fillet symbol describes leg size. Mix them up and a 5/8-inch bevel gets cut short on plate ordered for full penetration.
Other dimension rules worth memorizing:
Miller’s guide to reading weld symbols walks the same anatomy with illustrations if you want the picture version.
Example 1: single-V butt joint, welded from both sides. The arrow touches the joint, with a V below the reference line and another above it. Left of the lower V sits a depth note; to the right sits a size note. Read it as: bevel the arrow side to the stated depth, weld that side, then weld the reverse. If a backing-bar square also appears, a backing strip is required before the root pass.
Example 2: bevel groove on one side of a T-joint. A single-bevel symbol appears below the reference line only. The depth of bevel sits to the left of the symbol, and a fillet symbol may also be present below the line. That combination means groove first, then a fillet over it for reinforcement, which is a classic partial-penetration detail.
Example 3: double-V with different depths. Symbols appear above and below the line, each with its own number to the left. The arrow-side depth might be 3/4 inch while the other side is 1/4 inch. These are not a matched pair, and cutting both sides to the same angle is exactly how a Saturday disappears into rework. Pipe welding procedures documents run the same reading logic with the WPS alongside it.
Short Note: Common misreadings, in order of frequency: ignoring the arrow-side versus other-side rule, confusing the number left of the groove symbol with weld size, assuming a tail implies a special process, and missing the field-weld flag entirely. STRUCTURE magazine’s list of misapplied symbols covers most of what goes wrong in practice.
Before you cut, grind or strike an arc, resolve every symbol and write the interpreted dimensions on your fit-up checklist. A drawing is a contract in miniature, and the groove detail always wins the argument.
A groove weld is won or lost before the arc is struck. The bevel, the land, the root opening and the cleanliness of the sidewalls are where the strength comes from, and every one of them carries a tolerance you either hit or pay for at inspection.
Start with the drawing. When the groove geometry is not specified there, the choice comes down to thickness, access from one or both sides, and how much deposited metal you can afford. A square butt groove handles light plate; past roughly 3/8 inch a V, J or U bevel is needed so the electrode can actually reach the root.
Whatever the method, cut quality drives sidewall fusion. Oxides, slag and the hardened heat-affected zone left by thermal cutting must be ground back to bright metal along the full depth of the groove, and the land itself left clean, flat and within the specified tolerance.
Short Note: Fit-up tolerance discipline separates professionals from improvisers. Check the root opening with a gauge at every tack, not once at the start. A root that drifts closed at one end and open at the other will give you burn-through on one side and lack of penetration on the other, in the same weld, and no amount of cap-pass skill fixes it.
Clean both groove faces, remove mill scale for at least an inch either side of the joint, square the members, set the specified root opening and root face, and check alignment for mismatch. Then tack, using the same filler and procedure as the production weld, since tacks are part of the finished joint and are inspected like everything else. Space them evenly, keep them small enough to fuse fully into the root pass, and feather the ends so the root ties in without a stop-start defect. Our guide on how to tack weld with a MIG welder covers spacing and sizing in detail.
From there the sequence is root, hot pass, fill, cap. The root pass establishes penetration and is the highest-skill, lowest-deposition part of the job. The hot pass, common in pipe and thick structural work, follows quickly to float out trapped gas and slag from the root. Fill passes build the section; cap passes bring it flush or slightly reinforced and set the final profile.
Backing bars, consumable inserts and back gouging exist to solve the same problem: a fully fused, sound root. A backing bar supports the molten root pool so it cannot blow through, and it is common in structural CJP joints. Consumable inserts melt into the root and are a favorite of pipe GTAW work. Back gouging means cutting a groove into the reverse side of a completed root to expose sound metal, then welding it out — the most reliable route to a true CJP joint when only one side is otherwise reachable.
| Process | Strengths in groove work | Trade-offs |
|---|---|---|
| SMAW | Any position, minimal equipment, great for root passes and field fit-up | Low deposition, slag to chip, stop-start stops |
| GMAW | High deposition, fast fill, easy restarts | Position-limited, gas coverage issues, needs clean metal |
| FCAW | Deep penetration, works outdoors, strong fill rates | Slag removal, fume generation |
| GTAW | Best root control, cleanest weld, no slag | Slow, expensive, high skill |
| SAW | Highest deposition, flat and long seams, thick plate | Flat/horizontal only, flux handling |
The groove weld symbol on the drawing carries this geometry into the shop, and the AWS weld symbol chart is the standard reference for reading it. The economics are simpler: submerged arc welding is the favorite for long, straight, flat groove welds on thick plate because of its very high deposition rate, while GTAW is chosen for root passes where control matters more than speed. Many shops combine them: GTAW or SMAW root, FCAW or GMAW fill, SAW on long downhand seams. If you are still setting up, our MIG welder setup guide covers machine configuration for the fill passes.
Preheat is not optional guesswork. In structural welding, requirements depend on steel grade, member thickness, ambient temperature and joint restraint, and the values are tabulated by category in AWS D1.1. Measure with a temperature stick, IR thermometer or contact pyrometer on the back side of the joint, not by holding your hand over the plate.
Interpass temperature matters just as much. Every pass reheats the previous ones, and exceeding the specified ceiling can degrade toughness and drive hydrogen into the joint. The standard defense against hydrogen-induced cracking, especially on restrained, thick or high-carbon steels, is the combination of low-hydrogen consumables properly stored and re-baked, controlled heat input and adequate preheat.
Shrinkage forces in a groove weld are enormous, and thin or long members will pull out of alignment if you let them. The main tools are sequence and restraint: back-step welding, running short segments in the opposite direction of travel; skip welding, jumping between distant points to spread heat; balanced welding, placing equal weld metal on both sides of a neutral axis; and temporary restraints, strongbacks and clamps that hold geometry until the joint cools. On double-V joints, alternating passes from each side is the single most effective control, because the shrinkage of the second side cancels the pull of the first.
Quick Note: None of the above replaces paperwork. A welding procedure specification (WPS) supported by a procedure qualification record (PQR) is what turns a chosen groove design into a repeatable, code-acceptable result. Weld to the procedure you were given, and if conditions drift outside its range, stop and get it addressed rather than adjusting on the fly. Our piece on pipe welding procedures walks through the WPS, PQR and qualification chain end to end.
Every rejected groove weld tells a story. The bead looked decent, the plates held, and then the inspector circles a spot on the radiograph and the grind-out begins. Trouble is, most groove defects are predictable minutes before they happen, if you know what caused the last one.
Quick Note: If your settings are drifting toward too cold, revisit the chart. Our 3/32 E7018 amperage guide covers the low end, and the 5/32 E7018 amperage guide covers heavy root and fill passes.
Short Note: Half of “bad welder” problems are actually setup problems. Loose cable connections cause erratic arc start, and poor gun prep leads straight to porosity.
None of this is theory. Every defect above has a cause you can control with the next pass: clean, fit, set your heat, read the symbol, and inspect your own work before anyone else does.
A groove weld can look perfect from the hood and still hide a lack of fusion below the cap. Inspection is an ordered chain, and knowing that order lets you call for it confidently instead of getting red-tagged on something you could have caught yourself.
Every groove weld, in every code, starts with visual examination, and in most shops the welder is his own first inspector. Most welds pass or fail right here, before anyone picks up a meter or a camera. Visual checks are what you can see and measure: profile, reinforcement height, undercut depth, misalignment between plates, crater finish, surface cleanliness, and whether the deposited size matches the symbol.
The limits are not opinions. They come from the governing code or contract documents, and they are dimensional. Excessive reinforcement is not a cosmetic complaint; it creates a stress raiser exactly where the load concentrates. Undercut works the same way, and the undercut acceptance limits article breaks those numbers down.
When the spec calls for surface examination, two workhorses cover it:
Both find cracks and surface-breaking lack of fusion. Neither looks inside the weld. That belongs to the volumetric methods.
Radiographic testing (RT) produces a film or digital image of the interior, read like an X-ray. It reliably shows porosity, slag inclusions, and lack of fusion on the plane perpendicular to the beam, and it leaves a permanent record. It is also slow, needs access to both sides, involves radiation controls, and can miss defects oriented parallel to the beam.
Ultrasonic testing (UT) bounces high-frequency sound through the weld and reads the echoes. It is portable, has no radiation hazard, handles thick sections well, and is highly sensitive to planar defects like lack of fusion and cracks. Its weakness is operator dependence: a poorly qualified technician misses what a skilled one catches instantly.
Quick Note: CJP groove welds in structural steel and pressure service are the details most often assigned volumetric NDT by drawing or code. PJP welds are frequently accepted on visual plus surface examination alone unless the engineer specifies otherwise. Read the NDT notes on the drawing; they override your assumptions.
Inspection examines the welder too. Qualification starts with a welding procedure specification (WPS), backed by a procedure qualification record (PQR) proving that procedure was tested successfully on a coupon. The welder then runs his own coupon using the actual process, position, and technique he will use in production. That coupon is evaluated, and the result goes on a welder performance qualification (WPQ) card.
Qualification is not permanent. Continuity rules require demonstrated production use at least every six months under AWS D1.1, or the qualification lapses. Pipe work gets considerably more involved; the pipe welding procedures and inspection guide walks the full chain.
Different work follows different rulebooks. Structural steel generally follows AWS D1.1. Pressure vessels fall under the ASME Boiler and Pressure Vessel Code, with the National Board documenting how those requirements play out in its published guidance. Pressure piping follows the ASME B31 series, and pipeline work follows specifications like the California DOT Special Provisions or operator standards. The NDT method, extent, and acceptance criteria are written into the code, the specification, and the WPS; know which one governs.
Run this before you call the inspector and you save a second call, a schedule slip, and your credibility:
Do that and you walk into inspection knowing the answer before the question is asked. And inspection gear and safe work habits go together: the complete welding PPE guide covers the protection you want long before the examiner arrives.
When tackling a groove weld, the joint itself is only half of the job. The space you’re working in can be just as critical as your technique. From open fabrication shops to elevated scaffolding or the inside of a pressure vessel, these environments can quickly go from manageable to dangerous. Because groove welds often require multiple passes and extended arc time, they concentrate heat, fumes and risk. For this reason, safety should always be built into the procedure, not added at the end.
Welding fumes are a mixture of metal oxides, flux decomposition products and gases. In the United States, OSHA 29 CFR 1926.353 requires proper ventilation and protection for welding, cutting and brazing. For general industry, 29 CFR 1910.252 establishes the same baseline, along with fire prevention and mandatory protective equipment. A simple rule to follow is this: if you can smell the weld, you are breathing it. Utilize a fume extractor at the joint, cross-ventilation, or an approved respirator program. For a broader breakdown of protective equipment, our welding PPE guide is a great starting point.
Pipe spools, box sections and pressure vessels can turn a routine groove weld into confined-space work. These spaces require atmospheric monitoring for oxygen, flammables and toxic gases, a permit system, and a second person stationed outside before striking an arc. Residual vapors can be lethal, so never rely on your nose. You will also need a written rescue plan in place. For specifics on welding inside piping systems, take a look at our gas line welding safety rules.
Welding stainless steel, chromate-coated metals and many superalloys can generate hexavalent chromium, a recognized carcinogen. OSHA regulates exposure under 29 CFR 1910.1026 with a permissible exposure limit of 5 micrograms per cubic meter of air as an 8-hour time-weighted average. Always ask for the alloy you’re working with, and never assume stainless can be welded safely without effective local exhaust.
Quick Note: The arc can injure just as severely as fumes can. OSHA requires welders, helpers and bystanders to use proper filter lenses and protection. For shade selection and gear breakdowns, check out what a welder wears.
Back gouging a groove weld will throw sparks far beyond the joint, with hidden embers able to smolder long after the arc stops. For this reason, a fire watch with an extinguisher is mandatory, and it must remain in place after welding ends. Flash suits, hearing protection, ANSI-rated eye protection, fire-resistant leathers and respirators should be used when grinding or gouging. Many of the same precautions found in our MIG welding safety precautions apply here, especially when it comes to securing gas cylinders, inspecting leads, and checking your welding cable connectors. On top of that, don’t neglect your body. Overhead grinding and awkward positions can lead to chronic strain and burns. Our weld burns guide covers prevention, while regular breaks and stretching can save your joints for the long haul.
A flawless bevel means nothing if the procedure wasn’t followed or the environment wasn’t safe. Groove welding rewards those who respect both the metal, and the risks that come with it.
A groove weld is a weld made in the groove between two members to be joined, where the groove is a channel or opening prepared in the edges or surfaces to hold weld metal. It is the joint that allows the weld to reach into and through the thickness of the material rather than sitting on the surface.
A groove weld is deposited inside a prepared gap or beveled edge between two members, while a fillet weld fills the corner where two surfaces meet at an angle. Groove welds can be sized through the full thickness, whereas a fillet weld is sized by its throat.
A 60 degree included angle with a small root opening and roughly a 1/16 inch root face is the standard starting point for manual welding on moderate plate thickness. The governing answer is always the drawing, the welding procedure specification and the applicable code, because changing the angle changes penetration behavior and weld volume.
A complete joint penetration (CJP) groove weld extends all the way through the thickness of the joined members, so the joint develops full member strength. A partial joint penetration (PJP) groove weld is welded only to a specified depth, and the remaining thickness is intentionally left unwelded.
The basic groove shape sits on the reference line, on the arrow side or the opposite side depending on which member the arrow points to. Any depth of bevel or root penetration value is written to the left of the symbol, and supplementary marks such as the all-around circle or field weld flag add further instruction.
There is no single best process. Submerged arc wins on long flat butt seams in thick plate, GMAW and FCAW give high deposition in many positions, SMAW is the most flexible for field and out-of-position work, and GTAW is preferred for precision root passes.
You need one of them when the design requires complete joint penetration and the weld cannot be proven to penetrate fully from one side. A backing bar supports the molten root, while back gouging removes the imperfect root from the reverse side before the final pass is made.
Visual inspection is required for essentially every groove weld. Depending on the code, the joint type and the contract, the drawing may also call for magnetic particle or penetrant examination for surface discontinuities, or radiographic or ultrasonic examination for internal defects, which is most common on CJP details.
In the United States, OSHA standards 29 CFR 1926.353 and 29 CFR 1910.252 require proper ventilation, protective clothing and eye protection, and fire prevention controls during welding, cutting and brazing. Work in confined spaces adds atmospheric testing and permit requirements, and welding stainless steel adds hexavalent chromium exposure control.