PJP Weld Explained: Partial Joint Penetration vs CJP, Symbols, Strength & Requirements

What Is a PJP Weld? Partial Joint Penetration Explained

A drawing hits the shop floor with a bevel-groove symbol and a small note beside it reading “PJP 5/16.” The welder looks up and asks the question that trips up apprentices and engineers alike: “Isn’t that just a groove weld? Why can’t we just burn it in all the way?” That single acronym changes how the joint is designed, welded, inspected, and paid for.

A PJP weld — partial joint penetration — is a groove weld that does not extend completely through the thickness of the components being joined. Instead of fusing the full depth of the joint, the weld penetrates only partway into the material and leaves an unwelded root section behind. This definition is standard across both bridge and building codes: Caltrans bridge design guidance and AASHTO/AWS D1.5 both describe PJP as a groove weld that does not fuse through the full thickness of the joined parts.

Under AWS D1.1 Clause 5.4.2, PJP groove welds include two common situations: groove welds made from one side without steel backing, and groove welds made from both sides but without backgouging. In plain terms, if any unwelded material is left at the root, you are looking at partial penetration.

PJP vs. CJP in One Sentence

The contrast is simple. A complete joint penetration (CJP) weld fuses through the entire thickness of the joint, so the weld is as strong as the base metal and is treated as a full-strength connection. A PJP weld deliberately stops short. That means it is not designed to carry the full tensile capacity of the members — its strength depends on how much depth actually fused.

Why does the term matter so much? Because the acronym quietly controls four things at once:

• Design strength — the effective throat, not the full plate thickness, governs capacity.
• Weld volume — less deposited metal means less time and filler consumed.
• Cost — PJP welds generally cost less than CJP welds because they use less weld metal and may not require backgouging.
• Inspection strategy — partial penetration welds are usually examined with a different technique and acceptance criteria than full-penetration welds.

It is worth saying clearly: PJP is a legitimate, code-recognized, engineered weld. It is not a shortcut and it is not a defective weld. The Steel Tube Institute notes that fillet and PJP welds are perfectly valid connection types when the loading allows it, while AISC welding guidance treats PJP as a normal design tool with its own rules. You will see PJP welds everywhere in structural steel: shear connections, stiffener welds, and compression splices, where loads never demand full tensile capacity.

Good weld quality still comes down to solid fundamentals. Filler, polarity, and technique all show up in a PJP groove just like any other joint, so building muscle memory through deliberate welding practice pays off long before the inspector arrives. If you are running flux-cored or wire processes on these joints, understanding setup details like contact tip selection and wear will keep your penetration consistent from the first pass to the last.

Over the next sections, this guide walks through the whole picture: the groove joint details you will actually find in the field, how effective throat and weld size are measured, how engineers calculate design strength, the symbols that call a PJP out on a drawing, the WPS and welder qualification rules behind it, and the inspection and defect control that keep it sound. By the end, you will know exactly how to specify, weld, and inspect a PJP weld correctly — and when choosing partial penetration is the smart, economical call.

PJP vs CJP: Definitions, Symbols, and the Core Difference

Strip away the acronyms and the argument comes down to one question: does the weld metal travel all the way through the joint, or does it deliberately stop short? That single distinction separates complete joint penetration (CJP) from partial joint penetration (PJP), and it drives everything from the welding symbol on the drawing to the inspection plan in the shop.

What CJP Actually Means

A CJP groove weld extends completely through the thickness of the components being joined. When the filler metal matches the base metal, a sound CJP weld develops the full strength of the base metal, which is why it is the default for cyclically loaded connections and tension members. Because the metal fuses across the entire thickness, there is no unfused root waiting to act as a crack starter. That is why CJP welds dominate bridge tension flanges and moment connections.

What PJP Actually Means

A PJP weld does not extend completely through the thickness. The groove is filled only to a specified depth, leaving a deliberate root face below it. That is the trade-off: less weld metal, lower heat input, less distortion, and a much cheaper joint, but a reduced effective throat and a connection that develops less than the full capacity of the base metal. That is why a PJP is often paired with a reinforcing fillet when extra capacity is needed.

How AWS D1.1 Draws the Line

The American Welding Society keeps the two separate. AWS D1.1 Clause 5.4.1 governs CJP groove welds, while Clause 5.4.2 defines PJP groove welds. The prequalified joint details live in the figures: Figure 5.1 collects prequalified CJP details and Figure 5.2 collects prequalified PJP details. If you are writing procedures from those figures, a solid grounding in welding practice and WPS writing pays off quickly.

Reading the Symbol

The drawing tells you which one you are looking at. A groove weld symbol drawn with no dimension designates a CJP weld, so the fabricator is expected to achieve full penetration. A PJP weld symbol always carries an effective weld size dimension.

On double-sided details that dimension appears as E1 for the other side and E2 for the arrow side, a naming convention documented in Caltrans bridge design guidance. On a CJP, the size is implied by the material thickness, so no number is needed. On a PJP, the engineer must state exactly how deep the weld penetrates, because that depth sets the effective throat and therefore the load the joint can carry.

CJP Is a Result, Not a Shape

Here is the part that trips people up: CJP is defined by the result, not the joint geometry. The same single-bevel or V-groove can be either one. Weld it from one side only and you have a PJP. Backgouge to sound metal and backweld, and that identical groove becomes a CJP. The geometry did not change; the procedure did. That is why a groove weld is never automatically a CJP weld, and why a symbol alone never tells the full story.

It is also worth knowing that a CJP is not always the only full-capacity option. Guidance on full-capacity welds shows that fillet and fillet-reinforced PJP details can sometimes meet the demand at lower cost. Judging whether the required throat was actually achieved is its own skill, much like telling good and bad MIG welds apart.

PJP Groove Joint Designs and Detail Types You Will Actually See

Open a set of structural drawings and you will meet PJP welds in only a handful of familiar shapes. Prequalified PJP joint details are illustrated in AWS D1.1 Figures 5.2 and 5.5, covering V-, bevel-, J-, U-, and flare-groove configurations, while AISC’s Steel Construction Manual Table 8-2 lists the matching prequalified details and sizes. The geometry you choose is a trade between how much filler metal you need, how hard the joint is to fit, and whether a welder can actually reach the root.

The Groove Shapes You Will Actually See

• Single-V and double-V grooves: the workhorse. A single-V opens from one side with a root face and root opening; a double-V mirrors it, splitting the weld volume and balancing heat to fight distortion.
• Single-bevel and double-bevel grooves: one plate is beveled, the other square. Handy for T-joints and corners where only one member can be prepped.
• J-groove and U-groove: a curved, machined root face lets the arc reach deeper at a narrower overall angle, saving filler on thick sections at the cost of machining.
• Flare V and flare bevel grooves: formed where a curved member meets a flat or another curve, with no machined root face at all.

Every groove joint is defined by three numbers: the groove angle, the root opening (the gap), and the root face (the land). Widen the angle or the opening and the arc reaches deeper; narrow them and you save filler metal but risk lack of fusion at the root. Backing, when used, lets you weld from one side and still control the root, but it does not turn a PJP into a CJP.

Z-Loss: Why Narrow Grooves Lose Throat

Here is the rule that surprises people. A Z-loss factor is applied to groove angles below 60 degrees because the weld cannot reliably penetrate to the root, effectively turning the joint into a PJP groove weld even when drawn with a full-penetration symbol (AWS D1.1 Clauses 4.4.3.3 and 4.4.3.4). The narrower the included angle, the more effective throat you lose. Below a 30-degree groove angle, the T-joint detail is no longer prequalified under AWS D1.1 and is not used in strength calculations. Go that tight and the detail needs qualification by testing, not a table lookup.

Flare Grooves and Tubular Connections

Flare V- and flare bevel-groove joints are inherently partial joint penetration. They form where a round or curved member meets a flat one, so there is no machined root face to define penetration. For a flare V-groove the effective weld size is commonly expressed as E = 3/4 r, where r is the radius of the curved member.

Tubular T-, Y-, and K-connections bring their own wrinkles. When a branch meets a chord at a shallow angle, the heel of the intersection becomes an acute, hard-to-reach corner. Welders run a backing pass there, and that pass is not expected to meet volumetric inspection criteria; it supports the main passes and controls burn-through. On these joints, access often decides the design.

Fit-Up Tolerances Are Part of the Design

As-detailed versus as-fit-up is where clean drawings meet a dirty bench. Prequalification assumes the joint is built to the tolerance ranges in AWS D1.1. Let the root opening grow, let the bevel wander, or let a plate sit out of square, and you no longer have the prequalified detail. Exceeding those tolerances voids prequalification and usually forces a qualified WPS plus rework. Treat tolerances as a design input, not a shop-floor courtesy.

Joint selection usually comes down to three pressures:

• Fit-up difficulty: wide angles forgive sloppy fit; narrow grooves do not.
• Distortion: double-sided grooves balance heat and pull, single-sided ones do not.
• Accessibility: an overhead single-V may be impossible where a bevel from below is easy.

If your world is curved members, our guide to weld tube fittings walks through fit-up for those joints. And before you chase a tight root, deliberate welding practice pays off more than any symbol on a drawing.

Effective Throat and Weld Size: The Rules That Govern PJP Strength

When a PJP weld carries load, the number that matters is not the width of the cap sitting on the surface. It is the effective throat — the shortest distance from the joint root to the weld face — because that dimension sets how much metal actually resists the force.

Engineers multiply it by the weld length to get the effective area: Awe = effective length × effective throat. Stress in the weld equals the applied load divided by that area, so a throat 20% thinner than assumed means 20% more stress than the design allowed. That is why the drawing specifies penetration depth, and why the welder must actually achieve it instead of simply filling the bevel to the top. Two welds that look identical from the outside can carry very different loads.

How the Effective Weld Size Is Determined

For PJP groove welds, the effective weld size is generally taken as the depth of bevel — the penetration the joint detail promises. But codes subtract a deduction under certain conditions, because the arc cannot reliably fuse to the very bottom of a narrow or awkward groove.

The 1/8-inch (3 mm) deduction applies when:

• The process is SMAW or SAW and welding is done in the vertical or overhead position; or
• The groove angle at the root is less than 60 degrees, regardless of process.

No deduction applies when the groove angle is 60 degrees or greater for any process, or 45 degrees or greater in the flat or horizontal position using GMAW or FCAW. Those combinations give the arc enough access and heat to consistently reach the root.

Why the distinction? A wide, open groove gives the electrode a clear line to the root. A tight groove, or a position where gravity pulls molten metal into the root, makes fusion at the very bottom harder to guarantee. The deduction is the code’s way of admitting that the last fraction of the bevel is not dependable.

Minimum Effective Throat and Weld Size Tables

Codes also set floors so the weld receives enough heat for fusion. The AISC Specification covers minimum effective throat thicknesses for PJP groove welds in Table J2.3, with Table J2.1 addressing groove-weld throat geometry. For bridge work, AASHTO/AWS D1.5 Table 2.2 sets minimum effective weld sizes for PJP groove welds: 1/4 inch (6 mm) for material under 3/4 inch (20 mm) thick and 5/16 inch (8 mm) for thicker material. Designers cannot call out a hair-thin throat even when the stress calculation would tolerate it.

Why the Throat, Not the Face, Runs the Numbers

A cap can look deep and full while the root sits unfused, leaving no effective throat where it counts. The visible face size tells you how much weld was deposited; the effective throat tells you how much actually ties the two members together. Inspection can verify the face, but confirming the throat often takes root-focused or volumetric NDE. That gap is why process control matters. A worn contact tip in GMAW can wander the arc and rob you of root fusion, and consistent root penetration is a skill built through deliberate welding practice. High-deposition setups like robotic submerged arc welding help hold those depths across long PJP runs.

The takeaway: the effective throat is the design number, the deduction rules define it, and the welder’s real penetration depth is what makes it true.

How Engineers Calculate PJP Weld Design Strength

A PJP weld is not sized by eye or by feel. It is a designed joint with its own allowable stress, its own resistance factors, and a paper trail that has to survive review. Here is how the numbers actually work.

The Governing Equation

AISC Specification Section J2.4(a) sets the nominal stress for PJP groove welds at 60% of the filler metal tensile strength: Fnw = 0.60 FEXX. Multiply that stress by the effective weld area, Awe, and you get the nominal strength:

Rn = Fnw × Awe

The effective area is the effective throat multiplied by the weld length. From there, AISC applies resistance factors that depend on how the load hits the joint:

• Shear loading: φ = 0.75, with a safety factor Ω = 2.00
• Tension normal to the weld axis: φ = 0.80, with Ω = 1.88

So the design strength in shear becomes 0.75 × 0.60 FEXX × Awe, while tension normal to the axis allows the slightly higher 0.80 factor. Those factors are codified in the AISC welding provisions and are the backbone of PJP capacity checks.

Where the 0.6 Factor Comes From

That 0.6 multiplier is not arbitrary. It has been used since the early 1960s to compensate for two things: the notch effect of the unfused area at the root, and uncertainty about root quality where nondestructive evaluation is difficult. The unpenetrated land, the very feature that defines a PJP, acts as a stress riser, and inspectors cannot always see or size it reliably. The AISC’s research on fillet and PJP welds traces this history directly. The factor is a deliberate discount for a joint you cannot fully verify.

The Fusion-Face Check

Design strength is only half the story. Engineers also check the base metal at the fusion face using:

FnBM = 0.6 Fu

evaluated over the same effective area. If the base metal is weaker than the weld deposit, the fusion face governs and controls the design. Whichever of the two checks is lower wins. This is why specifying an overmatched filler metal does not automatically fix a weak parent material; overmatching raises Fnw but does nothing for the fusion face.

Weld Groups, ICR, and Other Codes

Real joints rarely load a single weld line in pure shear. For eccentrically loaded weld groups, AWS D1.1 Section 2.6.4.3 permits the instantaneous center of rotation (ICR) method, which models the weld group rotating about a point and sums the resulting resistances. Europe takes a different path: Eurocode 3’s directional method resolves stresses into components and checks them against the weld strength. Different math, same goal, prove the effective throat carries the load.

The Conservatism Question

Modern testing has pushed back on the 0.6 factor. Luo et al. (2020), Reynolds et al. (2020), and an AISC 2021 report by Dowswell all suggest the PJP design equations may be overly conservative, with theoretical rupture stress at the effective throat approaching 100% of weld metal strength. The code has not relaxed yet. Note too that the 50% directional strength increase is permitted for fillet welds loaded transverse to their axis. It is not a free pass on PJP welds.

A Worked Example

Take a 5/16 in. PJP bevel groove welded with E70 filler, where FEXX = 70 ksi. The nominal weld stress is Fnw = 0.60 × 70 = 42 ksi. In shear, design stress drops to 0.75 × 42 = 31.5 ksi. If the effective throat and length give an effective area of, say, 2.0 in², the design strength is 31.5 × 2.0 = 63 kips. Run the same joint at the fusion face with 0.6 Fu and use whichever result is smaller.

Comfort with these equations is part of building real welding skills, not just shop-floor technique.

WPS, Prequalification, and Welder Qualification for PJP Welds

A PJP weld on a drawing is not a suggestion. Once it reaches production, it is a controlled procedure. Every partial joint penetration groove weld must be made to a Welding Procedure Specification (WPS) that is either prequalified under the code or qualified by testing. Skip that step and you are welding without a defensible procedure, which is exactly where inspection failures and liability begin.

What Can Be Prequalified, and What Cannot

Under AWS D1.1, a handful of processes earn prequalified status because their behavior is well documented. AWS D1.1 permits prequalified WPSs for:

• Shielded metal arc welding (SMAW)
• Gas metal arc welding (GMAW), except short-circuit transfer
• Submerged arc welding (SAW)
• Flux-cored arc welding (FCAW)

Four processes fall outside that circle and must be qualified by testing: GTAW, GMAW-S (short-circuit), electrogas welding (EGW), and electroslag welding (ESW). If a drawing calls for any of those, budget for testing time.

The Limits of Prequalification in Clause 5

Prequalified does not mean anything goes. AWS D1.1 Clause 5 and Table 5.1 cap the variables that keep a procedure inside proven territory: maximum electrode diameter, maximum current, maximum root and fill pass thickness, and maximum single-pass fillet weld size. Push past any of these and your WPS is no longer prequalified. It becomes a procedure that must be tested.

Two details shops miss constantly: the footnotes in Table 5.1 and the as-detailed and as-fit-up tolerances. Those are not decoration. They are mandatory, and a groove fit up beyond tolerance voids the prequalification.

Welder Qualification: The CJP Carryover Rule

Here is the rule that saves a lot of testing money. Per AWS D1.1 Subclauses 4.25 and 4.28, a complete joint penetration (CJP) groove weld qualification also qualifies a welder for PJP groove welds. Put simply, if a welder has already passed a CJP test, they do not need a separate PJP qualification for the same process, position, and thickness range. A dedicated PJP test is usually just a supplement, a way to confirm the welder can deposit the required weld size and reach the specified penetration depth, as outlined in this welder qualification breakdown.

PQR and the Variables That Matter

When a procedure is qualified by testing, the results are recorded in a Procedure Qualification Record (PQR). The PQR captures the actual test coupon, the results, and the essential variables you cannot change without requalifying: process, position, base metal, and filler metal. Add supplementary essential variables when notch toughness is required, including CVN (Charpy V-notch) testing.

Qualification testing is not cheap. WPS qualification tests were reported to cost as much as $10,000 in 2021, and all of that documentation lives in the PQR. That price tag is why thoughtful shops lean on prequalified procedures wherever they honestly can.

From the Drawing to the Shop Floor

Translating the code into action looks like this:

1. Confirm the process is prequalified, or that a tested WPS exists for it.
2. Check that the WPS variables stay inside Table 5.1 limits and its footnotes.
3. Verify the welder holds a qualification covering that process, position, and thickness.
4. Confirm the as-detailed and as-fit-up tolerances on the drawing.

One more practical point: prequalification covers welding variables, not fit-up reality. When a fabricator changes a groove detail to suit production, that change can push the joint outside the prequalified envelope, forcing a tested WPS and a fresh PQR. Keep the drawing and the procedure aligned. A procedure on paper only works if the welder can execute it, so build those fundamentals through focused welding practice, sharpen machine setup with a guide to using a wire welder, and fine-tune heat input with the right amperage settings for your electrode.

Welding Techniques and Process Selection for PJP Grooves

A PJP weld is a controlled partial burn-in. The aim is to fill the groove down to the depth of bevel with sound metal, leave the unfused root where the drawing says it belongs, and stop before you gouge through the backing. Nail the depth and the joint is strong; overshoot and you leave a defect for the inspector.

Matching the Process to the Joint

Each arc process reaches the depth of bevel differently, so the choice comes down to position, thickness, and deposition rate.

• SMAW is the field default for structural PJP work. Low-hydrogen E7018 gives good toughness, and a tight arc with a slight drag angle drives penetration at the root. Amperage and polarity matter, so check settings in our guide to 3/16 E7018 amperage and technique.
• FCAW splits into gas-shielded and self-shielded. Gas-shielded flux core runs hot and deposits fast in flat and horizontal, ideal for deep bevels. Self-shielded wire tolerates wind, and it is a staple of gasless MIG welding for outdoor repairs.
• GMAW in spray and pulsed-spray transfer lays clean, high-deposition beads, and pulsed spray is the safest all-position choice for groove work. Globular transfer is a poor fit for PJP grooves because it spatters and gives erratic penetration. Keep consumables sharp; worn tips ruin arc stability, as covered in our contact tip maintenance guide.
• SAW is the flat-position deposition king, laying several times more metal per pass than SMAW. High-volume runs suit robotic submerged arc welding.

Not sure which wire you have? Our explainer on flux core vs MIG clears up the confusion.

Whichever process you run, polarity governs how deep the arc digs. DCEP concentrates heat at the plate and penetrates deeper at the root; DCEN runs cooler and flatter. The right pick for tubular wire is covered in our guide to polarity for flux core welding.

Travel Angle, Bead Sequence, and the Unfused Root

For a bevel groove, hold the electrode or gun close to perpendicular to the groove axis, adding a slight push or drag as the WPS requires, and keep the arc focused on the root for the first pass. Build the groove with stringer beads rather than wide weaves, stepping side to side or back-stepping to limit heat buildup. The unfused root is intentional in a PJP weld, so never “chase” it with extra amperage or dwell to burn deeper. That is how you blow through the root face and turn a sound PJP into a reject. Consistency here takes deliberate reps, and our welding practice guide lays out how to build it.

Fit-Up, Preheat, and Tolerances

Prequalification lives and dies on fit-up, a point the AISC guidance on PJP and fillet welds drives home. Hold the as-detailed and as-fit-up tolerances or the WPS no longer applies. One prequalified BTC-P4-GF detail, for example, allows a root opening from −1/16 in. to 1/8 in. and a groove angle from 40 to 55 degrees. For that same detail, a root face of at least 1/8 in. with a 45-degree groove angle needs no effective throat deduction. These limits come straight from Caltrans bridge design details.

Control heat input: too low and you risk lack of fusion, too high and you lose toughness and add distortion. Clean every pass, because slag, spatter, and silicates left behind become inclusions. Between passes, hold interpass temperature inside the WPS range. AWS D1.1 also treats PJP grooves differently in welder qualification, as this TheFabricator breakdown explains.

Position, Distortion, and Shop Setup

Position changes the math. Flat and horizontal work with GMAW or FCAW generally fills the groove more completely. SMAW and SAW run vertical or overhead require a 1/8-inch deduction from the depth of bevel when calculating effective weld size, so budget extra passes. To fight distortion, use balanced welding: alternate sides, back-step, and sequence passes around the neutral axis. A tidy welding facility setup with good fixturing and lighting makes all of this repeatable.

Inspection, NDE, and Common Defects in PJP Welds

When a PJP weld leaves the fixture, the inspector faces a problem: the root — the one thing that matters most — is buried inside the joint. The penetration the engineer designed for can’t be seen from the surface, and that single fact shapes every decision on the shop floor and in the field.

Visual Testing Comes First, and It’s Mandatory

Under AWS D1.1, visual testing (VT) is the one method performed on every weld, with no exceptions. It runs in three stages. Pre-weld: confirm bevel angle, root face, gap, fit-up, and cleanliness before the arc strikes, because a mismatched joint can’t be welded into compliance. In-process: check interpass cleaning, layer placement, preheat, and technique as passes stack up. Post-weld: inspect size, length, profile, undercut, overlap, and crater fill once the joint cools. A trained eye catches a surprising share of problems before any advanced method is called — the same skill behind telling good and bad MIG welds apart.

The NDE Toolbox: PT, MT, RT, and UT

Beyond VT, four methods cover most structural work:

• Dye penetrant (PT) — surface-breaking flaws on virtually any material; simple and inexpensive.
• Magnetic particle (MT) — surface and near-surface flaws in ferromagnetic steel.
• Radiographic (RT) — volumetric; radiation reveals internal porosity, inclusions, and cracks.
• Ultrasonic (UT) — volumetric; high-frequency sound detects planar flaws such as LOF and LOP, especially in thick sections where RT struggles.

UT and RT Are Not Automatic

A common misconception is that every structural weld gets ultrasonic testing. It doesn’t. Per AWS D1.1 Clause 8.15, UT is required only when specified in the contract documents. The AISC Specification Chapter N sets QA/QC inspection frequencies and splits responsibilities between fabricator and erector. Inspection and test plans typically build hold points around these methods so work pauses until results clear. Read the contract before assuming a method is in scope — an inspector ordering UT nobody requested is a cost problem, and a project skipping specified UT is a liability problem.

Why the Unfused Root Complicates NDE

A PJP weld stops short of full penetration by design, leaving a land of unfused base metal at the root. That land is intentional, so it always shows up during inspection — it is not a defect. The hard part is separating the intended unfused root from unintended lack of fusion or penetration along the sidewalls. UT signals scatter and mode-convert around that root geometry, so interpretation is genuinely difficult and depends on probe selection and a qualified operator. That difficulty is a major reason PJP designs lean on conservative strength factors such as 0.6.

Defects That Get Rejected

Under AWS D1.1 Annex O, three discontinuity types are never acceptable: cracks, lack of fusion (LOF), and lack of penetration (LOP). Surface defects like overlap are also rejectable — see our breakdown of the overlap welding defect. For UT, AWS D1.1 Table 8.2 sorts discontinuities into Classes A through D, with Class A rejected at any length, Class B above 3/4 in. (20 mm), and Class C above 2 in. (50 mm). Table 8.3 provides the companion acceptance criteria for statically loaded groove welds.

PAUT for Thicker Grooves

When conventional UT isn’t reliable enough, phased array ultrasonic testing (PAUT) steps in. AWS D1.1 Annex H governs it for groove welds and heat-affected zones from 3/16 in. to 8 in. (5 mm to 200 mm) using encoded linear scanning, giving a recorded picture of flaws single-probe UT can miss.

PJP Repair Workflow

Every rejectable discontinuity follows roughly the same repair path:

1. Document and locate the discontinuity; mark it clearly.
2. Remove it completely by grinding or arc gouging, with the correct contour.
3. Re-inspect the excavation before welding resumes.
4. Reweld per the qualified WPS with correct preheat and interpass control.
5. Re-examine the repair and record it against the original report.
6. Track repair rates — repeat defects point to a process problem, not a welder problem.

Common Mistakes, Cost Comparisons, and PJP vs Fillet vs CJP

The single most expensive habit in steel fabrication is overspecifying CJP welds. A designer defaults to “full penetration,” and suddenly every joint demands extra NDE, qualified procedures, and hours of grinding. Yet in most cases a fillet, or a PJP with fillet reinforcement, delivers the same full-capacity joint at a fraction of the cost.

The Steel Tube Institute is direct about it: for HSS connections, CJP welds are difficult and expensive and should be the exception, while fillet welds are generally the simplest and least expensive weld type. Their economy, ease of fabrication, and adaptability make them the preferred way to join components—choosing one isn’t cutting corners, it’s matching the weld to the load.

Fillet-Reinforced PJP Welds

Where a PJP groove alone lacks capacity, you can stack a fillet on top. This fillet-reinforced PJP is common in skewed and transverse connections. The effective throat is the shortest distance from the root to the weld face—the groove plus the reinforcing fillet—minus 1/8 in. where the governing code requires that reduction. Geometry rewards a clean fit-up; sloppy gaps and poor overlap control eat straight into throat thickness and strength.

Common Mistakes That Cost Money

• Specifying CJP everywhere out of habit, then paying for gouging, back-gouging, and full-length UT the load never required.
• Sizing a PJP by groove depth alone and forgetting the 1/8 in. effective-throat reduction, which quietly shrinks capacity.
• Using a PJP where the applied tensile stress runs normal to the effective throat—a code violation, not just a weakness.
• Assuming a fillet cap can “fix” an under-penetrated CJP. On a CJP it cannot add strength; on a PJP it can.

Rules That Surprise People

• Undermatching filler works. Undermatching filler metal may be used for all fillet and PJP groove welds when consistent with design requirements; you don’t have to match base-metal strength.
• Some PJP welds are prohibited. PJP groove welds cannot be used where applied tensile stress is normal to the effective throat, per Caltrans bridge design criteria and AASHTO/AWS D1.5. If the load pulls straight across the throat, specify CJP.
• Transverse fillets get a bonus. AISC permits a 50 percent directional strength increase for fillet welds loaded transversely, because the force acts across the throat.
• Fillets can’t rescue a CJP. Adding a fillet to a CJP groove weld cannot increase joint strength—the full thickness is already fused. Add that same fillet to a PJP and it genuinely adds capacity.

PJP also brings less heat input and distortion than a gouge-and-reweld CJP, which matters when you’re joining hardened steel or thin-wall tube. Process setup still counts: dial in your polarity and travel speed before you chase weld size.

CJP also drives rework. One lack-of-fusion indication can mean excavating sound metal and rewelding—all under the same UT that made the joint costly in the first place. That risk premium rarely appears in a fabrication quote until the weld fails inspection.

Cost and Performance at a Glance

Bottom line: specify the least expensive weld that safely carries the load—fillet first, fillet-reinforced PJP second, and CJP only when stress direction or code genuinely demands full fusion.

PJP Weld Checklist and Key Takeaways

A PJP weld is a partially penetrated groove weld, and that partial penetration is engineered, not accidental. Unlike a CJP, where fusion reaches the full joint thickness, a PJP deliberately leaves an unfused root and makes up the difference with a specified effective weld size. On a drawing, the groove symbol carries that size, and the design intent lives in the welding symbol, never in a shop-floor assumption.

Three rules anchor everything else. The symbol and its effective weld size govern the joint. A prequalified detail or a supporting PQR must back the design. And the effective throat, not the visible weld face, controls capacity — it must equal or exceed the size specified on the drawing, which is why code rules like the 1/8-inch deduction exist. Get the throat wrong and you either overbuild the joint or leave it understrength.

Pre-Weld to Post-Weld Checklist

1. Confirm the symbol. Verify the groove type and required effective weld size.
2. Verify backing. Is the detail prequalified under AWS D1.1, or supported by a PQR? Prequalification is a shortcut only when every variable fits.
3. Qualify the procedure. The WPS and the welder must cover the process, position, and thickness.
4. Hold fit-up tolerances. Root opening and bevel-angle drift quietly steals throat.
5. Apply the deduction correctly. Use the 1/8-inch deduction where required, or none where it is not.
6. Control preheat and heat input. Stay inside the WPS.
7. Confirm NDE. Match the required method and acceptance criteria.
8. Inspect for LOP and LOF. Lack of penetration and lack of fusion are always unacceptable.

PJP vs CJP vs Fillet: A Short Decision Guide

Match the weld to the load, not to habit. For a deeper read on where each type earns its keep, see the AISC fillet and PJP welds report.

• Fillet: modest loads, T-joints and laps, code permits it. Cheapest to fit, weld, and inspect.
• PJP groove: the design load exceeds fillet capacity, but full penetration is not justified.
• CJP groove: calculated tension, cyclic loading, or demand-critical joints. Full fusion, and typically ultrasonic testing.

Final Takeaway

PJP is an engineered, code-compliant weld, not a shortcut. The three biggest risks are fit-up drift, unverified procedures, and mistaking a designed root for a defect. Start with the symbol, trust the procedure, verify the throat, and inspect with intent — work the checklist and you will dodge all three.

Keep learning with our guides on building real welding skills, telling good and bad welds apart, and setting up a safe shop.

Frequently Asked Questions

What does PJP mean in welding?

PJP stands for partial joint penetration, a groove weld in which the weld metal does not extend through the full thickness of the components being joined. It is a deliberate, code-recognized design choice rather than a defect. PJP welds are used where applied loads are less than the full capacity of the connected parts.

Is a PJP weld as strong as a CJP weld?

No. A CJP weld extends through the full joint thickness and, with matching filler metal, can develop the full strength of the base metal. A PJP weld has a reduced effective throat, so its capacity must be calculated by the engineer using the code equations rather than assumed.

What is the effective throat of a PJP weld?

The effective throat is the shortest distance from the joint root to the weld face. The PJP effective weld size is generally the depth of bevel, with or without a 1/8-inch (3 mm) deduction depending on the welding process, position, and groove angle. The effective area is the effective throat times the weld length.

Why is 1/8 inch sometimes subtracted from a PJP weld size?

The 1/8-inch deduction accounts for the likelihood that the weld will not reliably reach the joint root. It applies for SMAW or SAW in the vertical or overhead position and for any groove angle under 60 degrees at the root. No deduction is required for groove angles of 60 degrees or more, or for 45 degrees or more in flat or horizontal GMAW/FCAW.

Can a welder qualified on CJP also weld PJP?

Yes. Under AWS D1.1 Subclauses 4.25 and 4.28, a complete joint penetration (CJP) groove weld qualification qualifies the welder for partial joint penetration (PJP) groove welds. A dedicated PJP test is typically used only as a supplement to confirm the welder can achieve the required weld size.

What are the most common defects in PJP welds?

The critical defects are lack of penetration and lack of fusion, which are always unacceptable because they reduce the effective throat. Porosity, slag inclusions, and undercut can also occur. Fit-up drift and unverified procedures are the most common root causes on the job.

When should you use a PJP weld instead of a fillet weld?

Use PJP when the joint geometry or loading calls for groove penetration beyond what a fillet can provide but full joint penetration is not required. Fillet welds are usually more economical and are the preferred method for many connections. A PJP weld with fillet reinforcement can be designed to reach full-capacity in some cases.

Can a PJP weld be ultrasonically tested?

Yes, but it is more challenging than testing a CJP weld. A PJP root presents an inherent unfused area that always produces an indication, and UT of that root is not straightforward. This inspection difficulty is one reason the design equations apply a conservative 0.6 factor to filler metal strength.

Are PJP welds allowed where tensile stress acts on the joint?

Not always. Many bridge and structural criteria, including Caltrans and AASHTO/AWS D1.5, prohibit PJP groove welds where the applied tensile stress is normal to the effective throat. PJP is most commonly permitted for shear connections, stiffener welds, and compression splices.

What welding processes can be used for a prequalified PJP WPS?

SMAW, SAW, FCAW, and GMAW (except short-circuit transfer) can be prequalified under AWS D1.1. GTAW, GMAW-S, EGW, and ESW procedures cannot be prequalified and must be qualified by testing. All prequalified procedures must comply with AWS D1.1 Clause 5 limits and joint detail tolerances.

Authoritative Sources & Further Reading

• https://dot.ca.gov/-/media/dot-media/programs/engineering/documents/memotodesigner/w-12-5m-a11y.pdf
• https://dot.ca.gov/-/media/dot-media/programs/engineering/documents/bridgedesigndetails/chapter-11/202502bdd113completejointpenetrationandpartialjointpenetrationgrooveweldsa11y.pdf
• https://www.aisc.org/aisc/solutions-center/engineering-faqs/8-welding/
• https://www.aisc.org/aisc/solutions-center/engineering-faqs/9-welding-inspection-and-non-destructive-examination-nde/
• https://www.aisc.org/media/rp2pj0bm/fillet-and-pjp-welds-final-report-rev1.pdf
• https://www.aisc.org/globalassets/modern-steel/archives/2015/12/si.pdf
• https://steeltubeinstitute.org/resources/full-capacity-welds-cjp-vs-fillet-vs-fillet-reinforced-pjp/
• https://www.thefabricator.com/thewelder/article/arcwelding/arc-welding-101-evaluating-partial-joint-penetration-grooves-for-welder-qualification
• https://pubs.aws.org/Download_PDFS/D1.1-D1.1M-2025WebPV.pdf
• https://connect.ncdot.gov/resources/Materials/MaterialsResources/Approved%20Welding%20Procedures%20-%20Jack%20and%20Bore.pdf
• https://weldinganswers.com/how-to-write-aws-d1-1-prequalified-welding-procedure-specifications/