Types of Welds and Welded Joints: The Complete Guide for welders and Designers

A fabricator once told me about a handrail that tore away from a wall plate three days after installation — the welds looked fine, but the shop had used fillet welds on a butt-style connection that was loaded in the wrong direction, with almost no penetration. That single joint-design decision put a worker in the hospital and cost the company a full rework. This is exactly why understanding types of welds and welded joints is not academic trivia — it is the difference between a structure that lasts 50 years and one that fails in 5 days.

Introduction: Why the Type of Weld You Choose Matters More Than Your Technique

A fabricator once told me about a handrail that tore away from a wall plate three days after installation. The welds looked perfect: clean ripples, nice color, and no visible porosity. However, the shop had used fillet welds on a butt-style connection that was loaded in the wrong direction with almost no penetration. That one decision put a worker in the hospital and forced a full rework. The problem wasn’t the operator’s skill. It was the drawing.

This is exactly why the type of weld you choose, and the joint it sits in, will always carry more weight than technique alone. A steady hand can never correct a poor joint selection.

Joint vs. Weld: Two Terms You Need To Know

Before moving forward, it’s important to separate these two concepts:

• A joint is the geometry of how the base metals are arranged. It exists before the first arc is struck.
• A weld is the deposited metal that fuses those members together, including its size, type, and penetration.

These terms are not interchangeable. The same joint can accept multiple weld types, and the same weld type will perform differently based off of the joint it’s in. Industry practice recognizes five basic welded joint configurations: butt, tee, lap, corner, and edge. Nearly every fabrication joint stems from one of these five.

Who This Guide Is For

If you searched for “types of welds and welded joints,” you’re looking for a practical breakdown to use in real time. This guide is built with three groups in mind:

• Students and new welders who need the vocabulary before their first qualification test.
• Fabricators and shop leads who need to pass inspection the first time without over-welding.
• Designers and engineers who need to specify joints and accurately interpret symbolic callouts.

All of these roles fall under code requirements. In the United States, OSHA’s 1910.252 requires that welding, cutting, and brazing follow accepted industry practices, which heavily reference consensus standards. The most prevalent of these for structural steel is AWS D1.1 (Structural Welding Code—Steel), with the current edition being AWS D1.1-2025. Additionally, when drawings utilize symbols over written instructions, they almost always reference AWS A2.4, the standard for symbolic representation of welding, brazing, and nondestructive examination requirements.

Design Dictates Success

The harsh truth is simple: a highly experienced welder can still produce a defective weld if the joint design is wrong. Improper edge preparation, a weld type mismatched to the load direction, or a lack of access for the required position will lead to issues such as undercut at the toe, lack of fusion at the root, or a fillet with an insufficient throat. These aren’t mistakes in technique, they’re mistakes in selection.

On the flip side, a well-designed joint is forgiving. The correct groove angle allows the arc to reach the root, the correct weld type matches stress flow, and the correct size provides true throat area. With those factors in place, a mid-level operator can consistently produce sound welds. For the machine side of that equation, check out our How to Set Up a MIG Welder: Complete Step-by-Step Guide, but know that no setting can rescue an improper joint.

What We’ll Cover

With the foundation set, this guide will build from the ground up in the following sections:

1. The five basic joints — geometry, applications, load behavior, and limitations for butt, tee, lap, corner, and edge joints.
2. The major weld types — groove variations, fillets, plug, slot, spot, seam, surfacing, and backing welds, including the CJP and PJP distinction.
3. Weld symbols under AWS A2.4 — reference line, arrow side, other side, dimensions, tails, flags, and a worked example.
4. Selection factors — the variables that drive the decision, explained in plain language.
5. Process-to-joint matching — SMAW, GMAW, FCAW, GTAW, SAW, and resistance welding.
6. Defects by joint type — what goes wrong, why, and how to prevent it.
7. Inspection, testing, and qualification — NDT methods, acceptance criteria, WPS/PQR, and welder qualification.
8. Safety and a practical checklist — including a printable 10-point selection checklist.

While this is a steel-focused guide, this taxonomy applies to a multitude of materials. Sheet metal, pipe, and pressure components use these same five joints with added requirements. For thin materials specifically, our Complete Guide to TIG Welding Sheet Metal builds off of joint selection. If your project consists of thin stamped parts, it’s worth reading How Does a Spot Welder Work? Physics, Process, and Practical Setup Explained, as resistance welding changes the entire conversation.

By the end, you’ll be able to correctly name any of the five joints, choose an appropriate weld type based off of loading and access, read the designer’s symbol, and predict the most common defects for that configuration. That’s the true difference between welding to a print, and understanding the print. Let’s start with the five joints that everything else is built on.

The Five Basic Types of Welded Joints (Butt, Tee, Lap, Corner, Edge)

Before you pick a process, a filler metal, or an amperage setting, you have to answer one question: what shape are the parts in? Every welded connection you will ever run falls into one of five basic joint geometries. The distinction matters because a joint is the arrangement of the members you are joining, while a weld is the actual deposited metal that fuses them together. The same weld can perform beautifully in one joint and fail in another. As UTI’s welding joint guide puts it, these five types — butt, tee, lap, corner, and edge — are the foundation everything else in drawing interpretation and procedure writing is built on.

Butt Joint

A butt joint is two members lying in the same plane with their edges abutting. It is the workhorse of structural steel, pressure vessel, and pipeline work, and it is usually welded with a groove weld that can be either complete joint penetration (CJP), where the weld consumes the full thickness, or partial joint penetration (PJP), where it does not. Heavy plate gets edge preparation — square, V, bevel, U, or J — so the arc can reach the root; groove geometry selection is essentially a trade-off between deposition volume and access.

• Typical uses: ship hull panels, pressure vessel shells, pipe spools, tank floors, plate girders.
• Load behavior: loads travel straight through the weld throat, so a CJP butt weld can be designed as strong as the base metal.
• Strengths: smooth profile, no added thickness, easy to inspect with radiography or ultrasonics.
• Limitations: needs accurate fit-up and often back-gouging or dual-side access; edge prep and multi-pass work add cost and distortion risk.

Lap Joint

In a lap joint the members overlap and are joined along the edges of the overlap, most often with fillet welds, though plug, slot, and seam welds also qualify. It is forgiving: no precise edge prep, and gap variation is easily absorbed. That tolerance makes it the default for sheet metal and thin-plate assemblies, including resistance spot-welded car bodies, a process we unpack in How Does a Spot Welder Work?. The downside is eccentricity — the load path is offset, which introduces bending and stress concentration at the weld toes, and the overlap creates a crevice where moisture and corrosion can hide.

Tee Joint

One member sits perpendicular to the other, forming a T. Tee joints are everywhere in frame fabrication: machine bases, column-to-base-plate connections, stiffeners, brackets, and handrails. They are normally fillet-welded on one or both sides, though a bevel-groove component is added when the designer needs real penetration. Load is applied across the joint, so the weld root and toes see high stress. Access drives everything here — a tee in a tight corner may only allow flat or horizontal fillets, while overhead position work demands real skill and careful technique, as covered in our overhead stick welding guide.

Corner Joint

A corner joint brings two members together at an angle to form an L. Depending on how the edges meet, it can be flush, lap, or edge corner. Corner joints dominate sheet metal box, cabinet, enclosure, and tank fabrication, where the appearance of the outside corner matters as much as the strength. They are efficient on thin material but sensitive to heat input: too hot and you burn through, too cold and you get lack of fusion on one side. Controlling that balance is exactly what TIG welding sheet metal is all about.

Edge Joint

The most underestimated of the five. An edge joint lines two members up side by side and welds along the shared edge, normally as a groove weld. It is practical only on thin stock — generally about 3 mm (1/8 in) or less — and appears on radiator tanks, fuel cells, ductwork, trimmed edges, and hinge leaves. Because the weld never has to carry a heavy structural load across the full thickness, it is a sealing and light-duty joint, not a primary strength connection. Stick welding thin metal exposes how unforgiving this geometry is on an overpowered machine.

Quick Comparison Table

In practice, you will see these combined. A pressure vessel is a long series of CJP butt joints in rolled shell courses with corner-joint end caps and tee-joint nozzles — the National Board treats longitudinal and girth butt welds as the primary pressure-boundary seams. A ship hull is mostly butt-welded panel lines. A machine frame is dozens of tee joints with fillet welds. A sheet metal toolbox is corner joints with a few edge joints along the seams. Recognizing the geometry first tells you which weld type, which inspection method, and which welding procedure to reach for — and that is the habit that keeps the wrong weld out of the wrong joint.

Types of Welds Explained: Groove, Fillet, Plug, Slot, Spot, Seam and More

The distinction that trips up most beginners: a joint is the geometry — how two pieces sit relative to each other — while a weld is the actual fused metal you deposit into that joint. The five basic joints can each be joined with several different weld types. A tee joint can take a fillet weld, a groove weld, or both. The drawing tells you which, and the load tells you whether the answer is right.

Groove Welds: Seven Preparations, One Job

Groove welds are fusion welds made in the gap between two members edge-to-edge, and their name comes from the shape cut or formed into the edge. The American Welding Society’s groove guide recognizes seven basic profiles:

• Square-groove — edges butted flat with a small root gap. Cheap, no beveling, but limited to thinner material unless a backing strip is used.
• V-groove — both edges beveled to form a V. The workhorse of plate and pipe welding; the included angle (usually 60°) trades deposition volume for easy arc access.
• Bevel-groove — only one edge beveled. Common on tee joints and where the back side is inaccessible.
• U-groove — a curved root profile that reduces filler volume on heavy plate compared with a V, at the cost of more expensive edge preparation.
• J-groove — one edge machined into a J. Same deposition savings as the U, used where only one side can be prepared.
• Flare-V-groove — formed by the natural radius of bent or rolled sections such as square tubing and pipe, requiring no cutting at all.
• Flare-bevel-groove — a flanged flat edge meeting a straight edge, again relying on the material’s own geometry.

CJP vs PJP: How Far Does the Weld Go?

Two abbreviations govern groove weld intent. Complete joint penetration (CJP) means the weld metal runs through the entire thickness, so the joint is effectively as strong as the parent metal. Partial joint penetration (PJP) fuses only a specified depth; the unfused root acts like a built-in crack-like discontinuity, which matters enormously under fatigue loading.

CJP from one side usually requires a backing strip, a backing ring, or a back-gouge-and-reweld sequence to guarantee fusion at the root. That is why pressure vessels and structural moment connections specify CJP, while simple statically loaded splices often accept PJP. The National Board’s guidance on CJP welds makes the point plainly: the designation is a through-thickness requirement, not a visual quality comment, and it drives both the examination method and the acceptance criteria. In practice, PJP joints are usually supplemented with fillet welds sized to carry the balance of the load.

Fillet Welds: The Triangle That Carries Everything Else

A fillet weld is roughly triangular in cross-section and joins two surfaces at an angle — lap joints, corner joints, and tee joints are its home turf. Strength is governed by the throat (the shortest distance from root to face), not the leg length, which is why code tables size fillets by effective throat. Equal-leg fillets on 90° joints put the throat at about 0.707 times the leg.

Two variations show up constantly on drawings:

• Intermittent fillet welds — specified with three numbers: leg length, individual weld length, and pitch. A callout of 5/16 – 2 – 6 means a 5/16 in. leg, 2 in. long, repeating every 6 in. center to center.
• Staggered intermittent fillets — alternating welds on opposite sides of a lap or tee joint, shown with a zigzag symbol. Staggering spreads heat input and cuts distortion while still meeting the combined length requirement.

Fillet welds are inexpensive and forgiving of fit-up, but weaker per unit of deposited metal than a groove weld, and they create a sharp root that concentrates stress. On cyclically loaded structures that root is a fatigue crack initiation site — a fact that drives much of AWS D1.1’s fatigue category table.

Plug, Slot, Spot, Projection and Seam Welds

Plug welds fill a round hole in one member to fuse it to an underlying member; slot welds do the same through an elongated slot. Both are common in built-up beams, heavy lap joints, and situations where a groove weld would be impractical or would seal off a drainage point.

Spot welds and projection welds are resistance welds: clamped between electrodes and fused by heat from electrical resistance rather than an arc. Spot welds are the backbone of sheet-metal assembly, while projection welds use a formed bump to localize current at a predictable point, allowing several welds in one stroke. For the underlying physics, our guide on how a spot welder works covers heat input, nugget formation, and electrode force; spot welding copper handles the tighter parameter control that copper alloys demand. Seam welds are essentially a continuous series of overlapping spots along a joint, producing a leak-tight line used on fuel tanks and ductwork.

Surfacing, Overlay, Backing and Insert Welds

Surfacing (hardfacing or overlay) welds join nothing — they deposit a wear- or corrosion-resistant alloy onto a base surface, as on bucket teeth, crusher rolls, or stainless-lined pressure vessel interiors. Backing welds sit at the root to support the groove weld, either as a permanent strip or as a consumable insert: a shaped wire ring that melts into the root and eliminates the back-gouge step. Consumable inserts are beloved in orbital pipe welding for producing a smooth, uniform root bead from one side, a technique covered in our pipe welding procedures guide.

Bottom line: the joint sets the geometry, the load sets the requirement, and the weld type is how you satisfy it. Get that sequence right and the rest — symbols, procedure, inspection — follows logically.

How to Read Weld Symbols on Engineering Drawings (AWS A2.4)

Two welders can look at the same drawing and walk away with completely different parts. Usually it is not a skill problem. It is a symbol problem. Weld symbols are the shorthand that lets an engineer say “full-penetration groove weld, 6 mm leg size, weld all around, field erection only” without writing a paragraph on every joint. The standard that governs that shorthand is AWS A2.4, Symbols for Welding, Brazing, and Nondestructive Examination, and almost every structural and pressure-vessel drawing you meet will follow it. Once the grammar clicks, a busy fabrication print starts to read like plain English.

The Anatomy of a Weld Symbol

Every weld symbol starts with the same skeleton: a horizontal reference line, a leader line (the arrow) angling off one end to point at the joint, and an optional tail at the back of the reference line. Everything else hangs off that line.

• Below the reference line = the arrow side. This is the side of the joint the arrow physically points to.
• Above the reference line = the other side. Same joint, opposite face.
• In the tail = supplemental information: a process identifier like “GMAW,” a reference to a specific Welding Procedure Specification, or an unusual specification the drawing cannot fit elsewhere.

The critical rule is placement, not the orientation of the drawing on the bench. If the arrow points at the near face of a tee joint and the symbol sits below the line, you weld the near face. Flip the symbol above the line and you weld the far face. If symbols appear on both sides, you weld both. The classic beginner mistake is assuming “below” always means the bottom of the plate. It does not. It means the side the arrow touches.

Groove Symbols: The Letter Shapes

Groove weld symbols are drawn as a stylized cross-section of the joint preparation, which is why they become intuitive the first time you cut a bevel on a bandsaw. Groove geometry drives penetration and fill volume, and the symbol tells the operator exactly which preparation the engineer called for.

The symbol you will meet most often is the fillet weld, drawn as a right triangle with the perpendicular side on the left. Miller’s welding-basics resources point out that the fillet triangle is one of the most frequently encountered symbols on fabrication drawings, which makes sense: lap, tee, corner, and edge joints all default to it. A single triangle below the line means a fillet on the arrow side; triangles on both sides mean a double fillet.

Numbers, Modifiers, and Special Flags

Dimensions follow a strict order. For a fillet weld, size comes first (leg length, in millimetres or inches), then length, then the pitch when the weld is intermittent. For groove welds you will see preparation angle, root opening, and penetration depth spelled out, often with a melt-through or backing symbol when full joint penetration is required.

Three modifiers change the meaning of an otherwise ordinary symbol:

1. Flag at the break in the leader line — a field weld. The joint is welded on site rather than in the shop.
2. Circle at the break in the leader line — weld all around, the entire perimeter of a member, common on pipe-to-flange and bracket work.
3. Staggered intermittent notation — two rows of short welds offset from each other, drawn as triangles offset above and below the reference line. The numbers read as “weld length, then pitch.” A 60-240 callout means a 60 mm weld every 240 mm centre to centre.

A Worked Example

Picture a handrail bracket bolted to a column. The arrow points to the top joint of a tee. Below the reference line sits a fillet triangle with “8” to its left and “60-240” to its right. Above the line sits a second fillet triangle with no dimensions. A flag sits at the leader break.

In practice that means: weld a fillet on the arrow side with an 8 mm leg, intermittently at 60 mm welds spaced 240 mm apart; weld a continuous fillet on the other side using the default size rules printed in the drawing notes; and perform the work in the field, not the shop. The engineer specified asymmetric stiffness on purpose — the top chord carries the load, the bottom only needs restraint.

Now change one element: swap the fillet triangle for a bevel-groove symbol with a dimension of 6 below it, and the job shifts from a quick fillet pass to joint prep, grinding, and likely a backing or root-pass strategy before you ever fill the cap. That is the power of a symbol. It does not just describe a weld; it defines the labour, the cost, and the failure risk of the connection.

The habit worth building is simple: before you fit up, decode the symbol twice — once for weld type and size, once for side and distribution — then check the tail and the flags. It takes thirty seconds. The rework it prevents rarely does.

Choosing the Right Weld and Joint: 8 Factors That Drive the Decision

Knowing the five basic joints and the common weld types is only half the job. On the shop floor the harder question never stops: which weld goes on which joint, for this plate, under that load? Eight factors decide it, and they rarely all point the same direction. A joint that wins on cost often loses on fatigue life, and the joint built for the heaviest load may be impossible to reach with a torch. Weigh them in this order and the right choice usually becomes obvious.

Thickness, Load and Penetration: Where Strength Actually Comes From

Thickness sets the stage. Thin sheet usually means a fillet or a spot weld with carefully controlled heat, while thick plate demands a prepared groove simply to get fusion down to the root. Load direction then decides whether that weld works in shear, tension, or bending. Fillet welds are efficient in shear and easy to position, but they are the wrong instinct when a lap or T-joint sees repeated pull-out tension.

Penetration is the tiebreaker. CJP (complete joint penetration) groove welds are typically required for full-strength splices where the weld must develop the full strength of the member. A PJP (partial joint penetration) weld is legitimate when the calculated effective throat is enough, but it is a design decision, never a shortcut. The strength numbers live in AWS D1.1 Table 4.3, which summarizes available strengths for various welded joints. One trap worth flagging: design strengths must match the load-resistance method your shop uses. LRFD and ASD apply different factors, so a value copied from the wrong column quietly under-designs the joint. Always pull from the column that matches your method, as explained in the ANSI overview of AWS D1.1-2025.

Fatigue and Cyclic Loading: Reading the D1.1 Categories

Static strength is the easy part. If the connection sees repeated cycles, fatigue governs, and fatigue barely cares how much metal is in the weld. The fatigue provisions of AWS D1.1 for non-tubular connections sort details into categories A, B, B′, C, D, E, E′ and F, based on joint geometry and stress range. Category A is essentially unwelded base metal or a ground-flush CJP groove weld. Every step further down the alphabet is a steeper penalty for the same stress range.

In plain language, the category is a score for stress concentration. Two joints with identical throat area can differ enormously in allowable stress range because one has a smooth, ground toe and the other an abrupt notch. If your detail lands in E or F, adding weld metal will not rescue it. Change the detail instead: move the weld out of the high-stress zone, grind the toe flush, or replace a sharp-cornered attachment with a tapered one. Undercut at the toe does exactly this damage, which is why welding defects: undercut — causes, acceptance limits and how to fix it matters far beyond cosmetics.

Access, Position and the Operator’s Reality

A perfect joint on paper is worthless if the torch will not fit. Overhead fillet and groove work (4F and 4G) runs at lower current, smaller beads and slower travel than flat position, so productivity drops while defect rates climb. The usual answer is to rotate the part in a positioner rather than pay an operator to fight gravity for eight hours. On pipe, position and access are baked into the procedure itself; our guide to pipe welding procedures, positions, passes and inspection walks through how that plays out. When overhead cannot be avoided, technique and rod selection carry the day, as covered in overhead stick welding: the complete 4F and 4G guide.

Distortion Control: Fighting Shrinkage

Every weld cools, shrinks, and pulls. Distortion is managed with sequencing, back-step and skip techniques, fixturing, and balanced heat input. Alternate sides of a double-V groove instead of filling one completely, stitch short segments moving backward along the joint, and tack firmly before the main pass. Strong tacking is cheap insurance, with spacing and settings laid out in our MIG tack weld guide. Thin material punishes sloppy heat control fastest, so stick welding thin metal is where burn-through and warpage show up first.

Cost, Deposition Rate and Material

Cost comes last on purpose. Labor, not filler metal, is the largest slice of most weld bills, so a cheaper consumable that takes twice the passes is rarely cheaper overall. Deposition rate and edge prep drive the quote: a square-groove joint that needs no beveling saves setup before the arc is even struck. Material then overrides everything above it. Aluminum pushes you toward GTAW or a capable pulse MIG setup, stainless demands heat input discipline to avoid sensitization, and galvanized or cast steel brings fume and cracking concerns that no clever joint design will fix.

The habit worth building is a two-minute mental pass: How thick? Which way does the load pull? Does it need full penetration? Does it cycle? Can I reach it? Which way will it warp? What does it cost? What is it made of? Run that list on every joint and you will catch the failures that technique alone never could.

Welding Processes and the Joints They Suit Best (SMAW, GMAW, FCAW, GTAW, SAW, RSW)

Choosing the right joint geometry is only half of a solid design. The welding process you pair with it will determine your deposition rate, penetration, working positions, and whether the finished connection holds up to its intended load. A tee joint that runs smoothly on a mechanized GMAW fixture in a shop can be incredibly difficult to weld overhead on a jobsite. Likewise, a lap joint that is routine for sheet steel would be a poor fit for 40 mm plate.

With that in mind, here is how the most common welding processes line up with the five basic joints and the weld types that suit them.

SMAW (Stick): The Field and Structural All-Rounder

Shielded metal arc welding is one of the most portable options on the market. All you need is a power source, an electrode holder, and a box of electrodes, making it ideal for sites where wind, tight access, or a lack of shielding gas would halt other processes. Its true all-position capability allows for overhead fillet and groove work (4F and 4G), which is why it remains a staple for structural steel erection, pipe root passes, repairs, and maintenance.

• Best joints: Butt, tee, corner, and lap on plate from roughly 3 mm upward, as well as grooved pipe and structural members.
• Typical welds: Fillet, V-groove, bevel-groove, and multi-pass root, fill, and cap welds.
• Trade-off: Slag removal, a lower deposition rate than wire-fed processes, and high operator dependence. Electrode selection is critical, so it’s worth reviewing electrode classifications and polarity before striking an arc.

GMAW and FCAW: The Shop Workhorses

Gas metal arc welding (MIG) and flux-cored arc welding are built for speed and efficiency when producing fillet and groove welds on plate and sheet. GMAW offers a continuous wire feed, minimal slag, and high deposition rates. Short-circuit transfer keeps heat low for thin sheet, while spray transfer builds speed on thicker plate. These gains rely on fine-tuning, and a consistent MIG welder setup routine is one of the easiest ways to avoid defects.

FCAW trades external shielding gas for a flux core, providing deeper penetration and greater tolerance for dirt, mill scale, and wind. Self-shielded FCAW is well-suited for outdoor structural and shipyard work, while gas-shielded variants thrive in heavy fabrication shops. Variables like stick-out, voltage, and drag versus push technique all shift with the wire, so studying flux-core wire welding technique is crucial before switching between joints.

GTAW (TIG): Precision, Root Passes, and Thin Material

Gas tungsten arc welding prioritizes control over speed. It produces clean, high-quality welds with precise penetration, making it the top choice for pipe root passes, aerospace and food-grade work, nonferrous metals, and jobs where appearance is just as important as strength. Its slower travel speed and higher skill requirement are most worthwhile on thinner materials. For those applications, the settings, filler technique, and common pitfalls are covered in the TIG sheet metal guide.

SAW: High-Deposition Work on Thick Plate

Submerged arc welding delivers some of the highest deposition rates for long butt and fillet welds on thick plate. With the arc hidden beneath a blanket of granular flux, it is nearly unmatched for long straight seams, large-diameter pipe, pressure vessels, ship panels, and heavy plate tees. However, it is limited to the flat position and requires proper fixturing, so it is rarely used for on-site erection work.

RSW, Oxy-Fuel, and Other Sheet Processes

Resistance spot welding is the industry standard for lap joints in thin sheet, especially in automotive body assembly. Electrodes clamp the sheets together and pass a high current through the overlap, creating a weld nugget via resistance with no filler, flame, or shielding gas. The underlying physics and setup are broken down in how a spot welder works. Projection and seam welding extend this to staged or continuous joints, and even conductive materials like copper can be joined with careful parameter control, as outlined in spot welding copper.

Oxy-fuel welding and brazing still earn their place for thin sheet, repairs, and sites without power. They offer a portable flame for lap and corner joints, though heat input and distortion are harder to control than with arc processes.

Quick Process-Selection Matrix

The most practical takeaway is simple: match the process to joint geometry, thickness range, access, and position before the first tack goes in. Never assume a shop-preferred process will translate to the field. Above all, protect yourself from arc radiation, spatter, and fume with the right welding PPE from the moment you set up.

Common Weld Defects by Joint Type — and How to Prevent Them

Most weld defects don’t happen by chance. They’re the result of joint design, heat input, preparation, consumable condition and technique. By knowing which flaws are most likely for each joint, you can prevent them before they reach inspection.

Undercut: A Fatigue Risk for Fillet and Lap Joints

Undercut is a groove melted into the base metal at the toe of a weld and left unfilled. It’s most common on fillet welds and lap joints, especially when travel speed is too high, amperage is too high, the electrode angle is off or the arc lingers at the edges. On a tee joint under tension, that small notch creates a stress concentrator. For this reason, AWS D1.1 sets strict limits on undercut depth.

Prevention: Slow your travel slightly, angle your electrode to push filler into the toe and drop amperage if a groove begins to form. If undercut appears, grind it out and re-weld. Do not cap over it. For more details, read our guide on welding defects: undercut — causes, acceptance limits, and how to fix it.

Lack of Fusion and Incomplete Penetration: Hidden Risks for Groove Joints

These two flaws are the biggest concerns for partial-penetration groove welds and undersized fillet welds. Lack of fusion means the weld metal never bonded to the base metal or a previous pass. Incomplete penetration means it never reached the specified root depth. Both are planar flaws, so visual inspection won’t catch them. Radiography or ultrasonics are often required.

Prevention: Bevel and clean joint faces, verify the root opening and land during fit-up, use the correct amperage for your rod size and let the puddle wet into the sidewall before moving forward. For complete joint penetration, use backing or a proper root pass, and don’t assume one pass will bridge a wide root. For repeatable results on piping, follow a proven pipe welding procedure.

Porosity, Slag Inclusions, Burn-Through and Overlap

These defects are often tied to joint preparation, shielding and heat control. Porosity comes from trapped gas, caused by oil, rust, damp consumables, wind, lost shielding gas or a too-short arc in stick welding. Slag inclusions form when you skip cleaning between passes or use poor weaving. Burn-through, excessive penetration and overlap are especially common on thin sheet. Overlap occurs when weld metal spills onto the base metal without fusing. This is a frequent issue when stick welding on thin metal.

Prevention: Clean and dry all materials, verify gas flow and connections (a faulty welding cable connector can cause erratic current and penetration), chip and wire-brush between passes, and use lower amperage or stitch techniques on thin material.

Cracks: Cold Cracking and Crater Cracks

Hydrogen-induced cracking, or cold cracking, needs three things: a susceptible steel microstructure, high hydrogen content and high restraint stress, often triggered by rapid cooling. It can appear hours or days after welding, so it demands foresight. Thick sections, high-strength steels and stiffly restrained corner or tee joints are most at risk. Low-hydrogen rods, like E7018, are your first defense. Learn more about rod classification in our guide on what electrodes are and how to choose them. For a specific setting, see our breakdown of proper 3/32 E7018 amperage, as overcooked rods lose their low-hydrogen benefit.

Crater cracks form when you snap off the arc without filling the crater at the end of a bead. They’re common in sheet work, especially when TIG sheet work requires a backpedal or crater-filler technique.

Prevention: Preheat where the procedure requires it, control heat input, use correctly stored low-hydrogen consumables and always fill the crater.

Lamellar Tearing and Distortion

Lamellar tearing occurs in rolled plate under high through-thickness restraint, most often at corner and tee joints in heavy, stiffened structures. Flattened sulfide inclusions can split in layers as shrinkage pulls them apart. Prevention calls for specifying through-thickness (Z-quality) steel, using lower-hydrogen processes, redesigning to relieve restraint, or applying buttering layers where needed.

Distortion is the defect that rarely triggers NDT, but ruins assemblies. It’s caused by unbalanced shrinkage from excessive heat, uneven weld placement or poor sequencing. The fix starts before striking an arc. Use proper fit-up, clamping, back-step or skip techniques, balanced weld placement and realistic amperage. A clean, well-organized rig, like the one outlined in our MIG welder setup guide, will prevent most distortion issues from the start.

Weld Inspection, Testing, Codes and Qualification

A weld can look perfect and still be hiding a crack. That is the uncomfortable truth that makes inspection, testing, and qualification the backbone of any welding operation. No matter how good your joint design or your operator technique, quality is only proven when someone checks the work against a recognized standard. This section breaks down how that checking actually works in the field and the shop.

Visual Welding Inspection: The First and Cheapest Line of Defense

Before any sophisticated equipment comes off the truck, visual welding inspection (VT) does the heavy lifting. AWS D1.1 and most other codes treat visual inspection as the mandatory first step, and for good reason: it is fast, inexpensive, and catches a surprising share of problems before you spend money on more advanced testing. A trained inspector checks weld size, profile, undercut depth, overlap, porosity visibility, spatter, and cracks right on the surface.

Good visual inspection starts before welding even begins. Fit-up, joint preparation, cleanliness, and root gap all get checked, because a defect introduced at fit-up cannot be fixed with technique later. For a practical look at the details inspectors scrutinize, our guide to undercut causes and acceptance limits walks through one of the most commonly flagged issues, including how much undercut a code will actually tolerate.

Non-Destructive Testing: The Four Main NDT Methods

When visual inspection is not enough, non-destructive testing (NDT) examines the weld’s interior or surface without damaging the part. The four methods you will meet most often are:

• Ultrasonic Testing (UT): High-frequency sound waves are bounced through the weld, and flaws like lack of fusion or cracks reflect the signals back. UT has deep sensitivity to internal defects and does not involve radiation, but it is operator-dependent and less useful on very thin or coarse-grained materials.
• Radiographic Testing (RT): X-rays or gamma rays create an image of the weld’s interior, much like a medical X-ray. RT gives a permanent record and is excellent for porosity and slag inclusions, but it requires strict radiation safety controls and takes longer to process.
• Magnetic Particle Testing (MT): A magnetic field is applied and fine iron particles are sprinkled on the surface. Discontinuities like surface and near-surface cracks disrupt the field and collect the particles into visible lines. MT works only on ferromagnetic materials, so it is perfect for carbon steel but useless on stainless or aluminum.
• Liquid Penetrant Testing (PT): A colored or fluorescent dye seeps into surface-breaking flaws, is cleaned off, and a developer draws it back out, revealing cracks and porosity. PT works on any non-porous material, making it the go-to for stainless and aluminum.

Choosing between them depends on the flaw you expect, the material, the geometry, and the code. Pressure vessel work under ASME rules frequently requires RT or UT on butt joints, while structural fabrication may use MT or PT on critical details.

Acceptance Criteria: What “Passing” Actually Means

Acceptance criteria are the numeric limits that define a good weld: maximum porosity, undercut tolerances, minimum reinforcement, allowed crack types (which are almost always zero), and required weld size. Different codes set different limits. ASME pressure vessel work is generally stricter than structural work. The philosophy is the same, though: criteria tie directly to how the weld will be loaded and how much risk a flaw introduces. Welds carrying cyclic loads get far tighter scrutiny than static, lightly loaded connections.

WPS, PQR and Welder Qualification

Inspection only works if the weld was produced by a controlled process. That is where qualification documents come in:

• WPS (Welding Procedure Specification): The recipe. It specifies process, filler metal, polarity, preheat, interpass temperature, heat input, position, and gas. We covered practical procedure writing in pipe welding procedures and WPS fundamentals.
• PQR / WPQR (Procedure Qualification Record): The proof that the recipe works. A test plate is welded and tested (often with bends and tensile tests), documenting the results.
• Welder Performance Qualification: The proof that the welder can follow the recipe. This is a coupon test in the specific position and process the welder will use.

Two codes dominate this space. ASME Boiler and Pressure Vessel Code Section IX governs welding, brazing, and fusing qualifications, including WPS, PQR, and welder qualification records. For structural steel, AWS D1.1:2025 is the latest revision of the Structural Welding Code—Steel, and it updates design and qualification provisions from earlier editions. If you are working from an old D1.1 binder on the shelf, it is time to check your revision.

One last practical point: qualification is only useful if welders can actually execute the procedure. Settings matter, whether you are running a 7018 rod or a MIG job on thin sheet, and our guides to 3/32 E7018 amperage settings and MIG welder setup translate code requirements into dial settings you can use.

Safety, Best Practices and Final Takeaways for Welders and Designers

Every joint type, weld symbol and penetration decision in this guide eventually ends up in someone’s hands: the operator running the bead, the fitter holding the plate, the inspector signing the report, and the person who leans on the finished handrail years from now. Technique matters, but as the opening story showed, a perfect-looking fillet welded in the wrong place is still a failed joint. This closing section pulls the safety obligations and the selection logic into one place you can actually use.

Ventilation, Fume Control and Hot Work Permits

Welding produces fumes, gases and UV radiation that are invisible until they cause harm. The governing standard in most U.S. shops is OSHA 1910.252, which covers welding, cutting and brazing hazards including fire prevention, hot work permits and fume control. The rule many fabricators forget is the ventilation number: OSHA 1910.252(c)(2) requires mechanical ventilation at a minimum rate of 2,000 cubic feet per minute per welder whenever welding or cutting is done in a confined or poorly ventilated space.

That airflow is not a suggestion and it is not the same as cracking a door. Confined spaces such as tanks, pits, vessel internals and open-bottom ducts also trigger atmospheric monitoring, entry procedures and often a standby attendant. For fume and gas exposure more broadly, OSHA’s welding, cutting and brazing guidance lays out the control hierarchy: engineering controls first (local exhaust and general ventilation), exposure monitoring to confirm they work, and appropriate respiratory protection when controls alone are not enough. The agency’s fact sheet on controlling hazardous fumes and gases during welding is worth reading in full before you next weld coated, galvanized or stainless material.

PPE: The Non-Negotiable Layer

Ventilation protects the air; PPE protects the body. A full setup covers a flame-resistant jacket and sleeves, leggings, gloves matched to the process, and a helmet with the correct lens shade. If you are still guessing at shade numbers, gear fit or head protection, start with Welding PPE: The Complete Guide to Protective Gear Every Welder Needs (2026) and the companion breakdown in What Does a Welder Wear? The Complete Welding PPE Guide. Process-specific hazards deserve process-specific reading too: MIG Welding Safety Precautions covers wire-feed specific risks, and Weld Burns on Arms explains why sleeve gaps cause most of the injuries new welders suffer.

The 10-Point Joint Selection Checklist

Run through this before you strike an arc or release a drawing:

1. Load direction: Is the load shear, tension, bending or cyclic? Match the joint, not the habit.
2. Material thickness: Both members, and whether they are equal or unequal.
3. Required penetration: CJP, PJP or partial fillet? Say so on the drawing.
4. Access side: Can the welder reach both sides, or is a single-sided joint with backing mandatory?
5. Welding position: Will this be 1F/1G flat, or forced overhead? Position drives amperage and rod choice.
6. Distortion risk: Balance welds, sequence passes, and use staggered intermittent welds where appropriate.
7. Material type: Carbon, stainless, aluminum, galvanized or copper each push you toward a different process.
8. Process capability: Does your machine and filler actually achieve the specified penetration?
9. Inspection method: If it will be radiographed or ultrasonically tested, design the joint for it.
10. Cost and rework: Groove prep costs time; a wrong fillet costs a teardown.

For procedures that formalize these choices into a qualified document, see Pipe Welding Procedures: The Complete Guide to WPS, Positions, Passes & Inspection.

The Mistakes That Keep Showing Up

• Using a fillet weld where the load pulls perpendicular to the weld throat, then sizing it as if it were a butt joint.
• Skipping edge preparation and hoping for penetration. Square-groove joints only work within a defined thickness range.
• Tacking poorly. Bad tacks crack and become crack initiation sites, so follow a spacing and sizing routine like the one in MIG Tack Weld Guide.
• Choosing filler by availability instead of matching the base metal, as explained in What Are Electrodes in Welding?
• Ignoring undercut, which quietly becomes a fatigue crack starter. See Welding Defects Undercut: Causes, Acceptance Limits, and How to Fix It.
• Burn-through on thin material from amp settings that belong on plate, covered in Stick Welding Thin Metal.
• Skipping the WPS and welding from memory on code work.

Final Takeaway: Design First, Then Strike the Arc

Go back to that handrail. The operator was not unskilled; the joint design was wrong for the load, the penetration was inadequate, and nobody caught it before installation. The fix was never better hand control. It was asking five questions earlier: what loads this joint will see, how deep the weld must penetrate, whether the geometry lets the weld sit where the force travels, how it will be inspected, and what could go wrong if the answer is “we’ll figure it out on the floor.” Get those right and the weld becomes the easy part. That is the whole point of understanding types of welds and welded joints: the taxonomy is not trivia, it is the checklist that keeps structures standing and people safe.

Frequently Asked Questions

What are the five basic types of welded joints?

The five basic welded joint configurations are butt, tee, lap, corner and edge joints. Every welded connection you will encounter is a variation of one of these five arrangements, and each is joined with specific weld types such as groove or fillet welds.

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

A groove weld is deposited between the two members being joined, inside a prepared groove such as a V, bevel, square, U or J shape. A fillet weld is deposited beside the members, filling the corner or lap between them, and has a roughly triangular cross-section.

Which weld joint is the strongest?

A butt joint made with a complete joint penetration (CJP) groove weld is generally the strongest option because the weld can develop the full strength of the base metal. Fillet-welded lap and tee joints are usually weaker per unit length and are typically designed as partial-strength or shear connections.

What type of weld should I use for thin sheet metal?

For thin sheet around 3 mm (1/8 in) or less, edge joints and lap joints with small fillet welds or resistance spot welds are common choices. Square-groove butt joints with full penetration also work well, and lower heat input processes like short-circuit GMAW help prevent burn-through.

What does CJP versus PJP mean on a weld symbol?

CJP means complete joint penetration: the weld fully fuses through the joint thickness, often requiring backing, gouging or a specific groove design. PJP means partial joint penetration: the weld fuses only part of the thickness, which is economical but limits the load capacity of the connection.

What do the letters V, bevel, U and J mean on weld symbols?

They describe the shape of the groove prepared in the members. A V-groove tapers both edges symmetrically, a bevel tapers only one edge, and U- and J-grooves use curved profiles that reduce weld volume on thick material while maintaining penetration.

How do I stop my welds from cracking?

Control hydrogen (clean and dry consumables, remove coatings), reduce restraint where possible, preheat when the code or procedure requires it, use the correct filler matching the base metal, and avoid crater cracks by filling the crater before breaking the arc.

Do I need certification to weld structural steel?

Most structural codes require welders to be qualified to the applicable procedure. AWS D1.1 requires welder performance qualification against a written procedure (WPS), and ASME work requires qualification under Section IX, with continuity records maintained by the employer.