Test Coupon Welding: Preparation, Testing and How to Pass

A welder walks into a certification shop on a Monday morning, confident after five years of production work, and fails because the cap on his 3G plate was 3/32 inch too tall — a defect he never learned to spot with his own eyes. Test coupon welding is where skill meets paperwork: one small plate decides whether you are qualified for the job.

What Is a Test Coupon and Why Does It Matter?

A test coupon is a sample weldment: a small plate or pipe joint made with the same base metal, joint design, filler metal, shielding gas and welding process that will be used in production, later cut apart and tested to prove the weld is sound. That is the short answer. The longer answer is why a single 6-inch plate can decide whether a welder gets a job, whether a procedure gets approved, and whether a bridge or pressure vessel is allowed to exist at all.

Think of it as a scientific experiment with a deadline. You prepare the joint, weld it under controlled conditions, and then destroy it to find out what actually happened inside the metal. Nothing about a coupon is decorative. Every dimension, pass and cleaning step is chosen so that the test result can be trusted as a preview of thousands of identical welds made later on a real job.

The Coupon, the Specimen, and the Paper

New welders often mix up three things that codes keep strictly separate. Getting this straight early makes the rest of the process much easier to follow.

Term	What it is	What it proves
Coupon	The welded test joint itself, built to a set thickness, bevel and size	That the weld was made under qualifying conditions
Specimen	A strip, tensile bar or macro section cut out of the coupon	Strength, ductility, fusion and soundness of the actual weld metal
Record	The WPS, PQR or WPQ that documents everything	That the test happened, how it was done, and who is covered by it

The paperwork trail follows the same logic. A Welding Procedure Specification (WPS) is the recipe the shop follows in production. A Procedure Qualification Record (PQR) is the laboratory evidence that the recipe works. A Welder Performance Qualification (WPQ) is the individual welder’s record proving they can execute it. AWS D1.1 keeps standard WPS and PQR forms in Annex N, certifying that the test weld was prepared, welded and tested in conformance with Clause 4 of the Structural Welding Code — Steel. ASME Section IX mirrors this with its own suggested formats: Form QW-483 for the Procedure Qualification Record and Form QW-484A for Welder Performance Qualification.

Two Very Different Jobs From One Small Plate

A coupon serves one of two purposes, and confusing them is a classic preparation mistake.

• Procedure qualification (PQR). A procedure or test engineer welds the coupon to demonstrate that a specific set of parameters — heat input range, filler classification, preheat, position, thickness — produces acceptable results. The coupon is cut, bent, pulled in tension and inspected. Pass, and the WPS is approved for production.
• Welder performance qualification (WPQ). The welder themselves runs the test, usually on their own equipment and consumables, under the eyes of an inspector. The coupon is tested to show that this particular person can deposit sound weld metal in the position they will actually work in.

One rule ties them together and trips up even experienced welders: welders qualify to a welding process, not to a procedure, and the qualification record must match the WPS actually used in production. Qualify on GTAW and then run production flux-cored work without requalifying, and the paperwork no longer covers you. The same logic applies across materials, so a builder who learns alloy steel welding electrodes still has to think in terms of the procedure and position listed on the coupon, not just the rod in their hand.

Why Fabricators, Bridges and Boilers All Depend on Coupons

The reason this small plate carries so much weight is that nearly every load-bearing weld in modern construction traces back to one. A structural fabricator needs proof that the shop’s procedure works on the plate thickness it just bought. A bridge owner needs independent evidence that the welder climbing the girder can hold a vertical groove. A pressure vessel maker needs documented strength data because the failure mode is not a bent plate but a rupture.

In the United States, ASME Section IX is not merely industry practice; it is adopted by federal regulation in 46 CFR Part 57 for welding and brazing of power boilers, heating boilers, pressure vessels and piping. Maritime regulators in the Code of Federal Regulations essentially require the same qualification trail. Pipelines follow API 1104, bridges follow AWS D1.5, and state departments of transportation add their own variations on top. No coupon, no qualified welder, no stamp.

That is also why certification training, whether it is a full aluminium welding course or a short structural test-prep class, spends so much time on coupon work. Production skill gets you in the booth; coupon skill gets your name on the qualified list.

So when people ask what a test coupon is for, the honest answer is: it converts trust into evidence. The welder trusts their hands, the shop trusts its procedure, and the inspector trusts the record. The coupon is where all three have to agree before the arc is ever struck on the real structure.

Two Coupon Types: Procedure Qualification vs. Welder Qualification

Every test coupon answers a question, but there are only two questions worth asking. Either “Is this welding procedure sound?” or “Is this welder capable?” Confusing the two is one of the fastest ways to waste a test day, because the coupon you need depends entirely on which answer your employer or your code requires. Once you can tell them apart, the whole certification puzzle gets noticeably simpler.

The Procedure Qualification Coupon (PQR)

A procedure qualification coupon exists to prove that a welding procedure specification, or WPS, actually works. Someone welds a test plate or pipe using a specific process, filler metal, heat input, joint design and preheat, and the finished coupon is then destroyed in testing. The results, along with every variable actually used during the weld, are recorded on a Procedure Qualification Record. ASME Section IX formally captures this on Form QW-483, which lists the essential variables, the supporting data and the test results that back the procedure.

Two categories of variables matter here:

• Essential variables — changes that require re-qualification, such as a change in welding process, filler metal classification, thickness beyond the qualified range, or preheat level.
• Non-essential variables — changes you can adjust within the WPS without re-testing, such as travel speed, stringer versus weave bead placement, or minor fit-up details.

The PQR does not care who welded it as much as it cares what was welded and how. The welder used for a procedure test must be qualified and the welding must be supervised and documented, but the record proves the metallurgy, not the person.

The Welder Performance Coupon (WPQ)

A performance coupon proves something entirely different: that a specific welder can produce sound work with their own hands. Nobody destroys the procedure here. Instead, the coupon is visually inspected and tested, and the results are logged on a Welder Performance Qualification record, sometimes called a WPQR or, in ASME language, a performance qualification record.

Performance coupons are position-specific, which is why candidates spend so much time agonizing over which test to take. Under AWS D1.1 practice, a 3G or 4G groove test plate does far more than qualify those two positions: it also covers 1G, 2G, 1F, 2F, 3F and 4F plate positions. Thickness matters just as much. A 3/8-inch test plate qualifies groove welding up to 3/4 inch thick, and qualifies the welder for fillet welds on unlimited thickness.

When One Test Does Double Duty

The two qualification paths sometimes overlap. A welder can be credited with a successful performance qualification bounded by the performance variables of the coupon welded for a PQR, provided the welding was supervised and properly documented. In plain terms: if your company runs a procedure test and a qualified welder lays the bead under supervision with the records to prove it, that welder can often walk away with both a PQR and a WPQ from the same coupon.

One important exception: ASME Section IX treats machine welding as a different qualification from manual welding. A coupon run on an automatic rig does not credit a hand welder, and vice versa.

Coupon Menu: Pipe, Plate and Fillet

Coupon	Typical Use	What It Proves
6G pipe	Pressure piping, structural pipe, rolling positions	Hand skill in a fixed-incline position, all positions on pipe
Plate groove 1G–4G	Structural steel, buildings, frames	Varying position control plus visual, bend and tensile results
Fillet coupon	T-brace and lap joints	Leg size control, fusion, and root soundness on simple joints

Pipe coupons carry the heaviest reputation. NCDOT’s 6G certification, for example, uses ASTM A500 7/16-inch material with E7018 electrodes, welded vertical-up with a 50°F minimum preheat. That single test packs together consumable control, preheat discipline and uphill travel technique. Before you practice on anything else, review an all position welding rod guide so you understand how rod selection influences your position range, and confirm your E7018 handling against a proper alloy steel welding electrodes reference, since hydrogen pickup from damp coating is a leading cause of coupon rejection.

The takeaway is simple. Ask first whether you need to prove a procedure or prove yourself. Then pick the coupon whose thickness, position and diameter ranges match the work you actually intend to run, because a qualification you never mapped to your real work is just an expensive certificate.

Which Code Governs Your Test Coupon

Before you cut a single bevel or strike an arc, one question determines everything about your test coupon: which code governs the work you want to qualify for? The answer dictates the coupon dimensions, the number of specimens, the acceptance criteria, and even how long your qualification stays valid. Walk into a test shop knowing the wrong code, and you will prepare the wrong plate, weld it to the wrong dimensions, and fail a test you never should have taken.

AWS D1.1: The Structural Steel Workhorse

The AWS D1.1/D1.1M Structural Welding Code—Steel governs most structural plate qualification tests in the United States. If you are welding buildings, towers, or general steel structures, D1.1 is almost certainly your code. It defines the 1G through 4G groove plate positions, fillet weld tests, and the essential variables that control what your qualification covers. A typical D1.1 test might use a 3/8-inch A36 plate with a single-V groove, E7018 electrode, tested in the 3G (vertical) position. D1.1 also specifies whether a backing bar is required, the number of test specimens for each thickness range, and bend-test acceptance criteria. Because D1.1 is so widely used, many welders encounter it first during their initial certification. It is the code behind most SMAW structural test coupons you will see in a certification shop.

AWS D1.5: The Bridge Code

Bridge work runs under AASHTO/AWS D1.5M/D1.5 Bridge Welding Code, which is significantly more restrictive than D1.1. D1.5 requires visual inspection plus additional testing, and it mandates specific preheat and interpass temperature controls that D1.1 does not always require for the same thickness. State departments of transportation reference D1.5 in their own field welder test programs. For example, the North Carolina DOT’s approved structural steel welding procedures and Minnesota DOT’s field welder qualification review both build on D1.5 requirements but add agency-specific documentation and testing steps. If you plan to work on bridges, expect tougher acceptance criteria and a longer list of paperwork.

ASME Section IX: Pressure Equipment

Anyone welding pressure vessels, boilers, or high-pressure piping needs to understand ASME Section IX. This code governs procedure qualification (PQR) and welder performance qualification (WPQ) using a variables-based system. Section IX divides variables into essential, non-essential, and supplementary essential categories. If you change an essential variable outside the qualified range, you must requalify. Section IX also uses the QW-484A welding data form to record every detail of the test. One critical distinction: under 46 CFR 57.04-1, US marine welding qualification follows ASME Section IX as modified by Coast Guard requirements for specimen types and qualification ranges. This means a marine welder’s coupon may look different from a refinery welder’s coupon even though both work under Section IX.

API 1104: Pipeline Welding

Pipeline welders test under API 1104, which covers welds on pipelines and related facilities. API 1104 uses pipe coupons, typically in the 5G or 6G positions, with radiographic examination playing a central role in acceptance. The code allows a single test to qualify a welder for a range of pipe diameters and wall thicknesses, making it efficient for contractors who need broad qualification quickly.

CSA W47.1 and W59: Canadian Standards

Canadian welders work under CSA W47.1 (certification of companies for fusion welding of steel) and CSA W59 (welded steel construction). CSA W47.1 requires visual testing plus root and face bend tests for thicknesses up to 10 mm, with side bends required above 10 mm thickness. Macro-etch testing serves as a supplementary test in many Canadian qualification programs. These thresholds differ from AWS D1.1, so a welder qualified under one standard cannot assume equivalence under the other without a formal bridging process.

Government and Agency Variations

Beyond the main codes, government agencies impose their own layers. The Nuclear Regulatory Commission references ASME Section IX with additional requirements for nuclear facility work. The Kentucky Transportation Cabinet KM 64-108-0305 limits qualified status to two years from the completion of testing unless continuity records are maintained. Commercial testing services like PSI charge per plate for radiographic testing of submitted AWS D1.1 test plates, which illustrates that coupon testing is both a technical and commercial process. The American Welding Society’s guidance on reading and interpreting qualification records helps candidates understand how these agency variations interact.

Code	Applies To	Key Testing Requirement
AWS D1.1	Structural steel	Bend tests, visual inspection
AWS D1.5	Bridges	Visual, bend, and impact testing
ASME IX	Pressure vessels, piping	Bend, tensile, macro; per QW-484A
API 1104	Pipelines	Radiographic examination, bend tests
CSA W47.1	Canadian steel	Root/face bends to 10 mm; side bends above

Specialty welding in aluminum, stainless, and exotic alloys often falls under these same codes but with modified filler metal and preheat requirements. For those applications, resources on alloy steel welding procedures and aerospace welding standards provide the additional detail those materials demand. The key takeaway remains the same: identify your governing code before you touch a grinder, and every step that follows becomes clearer.

Preparing a Test Coupon: Fit-Up, Materials and Dimensions

Most welders who fail a coupon test fail it before they strike an arc. The bevel is cut a few degrees shallow, the root gap closes up halfway down the joint, mill scale still clings to the bevel faces, or the backing bar is clamped with a gap under it. None of these mistakes look dramatic, and all of them end the test as surely as a cold lap. As AWS guidance on passing welder qualification tests makes clear, the coupon is a physical record of how you prepared and executed the joint. This section walks through preparation the way an inspector checks it, in the order it actually happens.

Material Selection and Grouping

Start with paperwork, not steel. The base metal on the coupon must fall inside the range the Welding Procedure Specification (WPS) is written for, and the WPS itself is built on material grouping. Under AWS D1.1 and ASME Section IX, common steels are grouped by weldability and composition, so a procedure qualified on A36 plate (P-No. 1) generally covers other P-No. 1 steels within stated thickness and strength limits. You need the mill test report for the exact plate you are about to weld. A coupon welded on unverified metal can be rejected on paperwork alone, no matter how good the bead looks.

Filler metal matters just as much. The electrode or wire must sit within the F-number and A-number the procedure allows, and its strength must be compatible with the base metal. If your coupon involves carbon-moly or chrome-moly steels rather than plain carbon steel, the consumable choice gets tighter still, and alloy steel welding electrode selection becomes a study in its own right. Get the rod lot and container on the record before the test starts.

Bevel, Root Gap and Backing: Typical Dimensions

The most reproduced structural coupon in North America is the AWS D1.1 carbon steel plate groove test. Its standard fit-up is simple enough to memorize:

Fit-Up Item	Typical D1.1 Plate Coupon Value
Test plate	Two pieces, 3/8 in. (9.5 mm) thick A36 steel
Included bevel angle	45° total, meaning 22.5° per side
Root gap	1/4 in. (6 mm) nominal
Root face (land)	Around 1/16 in. (1.5 mm), held uniform along the joint
Backing bar	0.250 or 0.375 in. thick, at least 3 in. wide

The backing bar must sit in close contact with the test plate along its full length. A daylight gap between backing and plate changes root fusion conditions and gives you an unpredictable root pass, which is exactly where bend specimens crack. Plate length is commonly 12 inches, though some programs specify otherwise.

Pipe coupons follow a different script. For a typical 6G groove test, welders see a root gap in the range of 1/8 to 3/16 inch on a 1/16 to 3/32 inch land, with high-low (mismatch between the two pipe ends) held to roughly 1/32 inch. ASME Section IX defines its own coupon geometry and orientation rules in the Section IX qualification documents, so treat those numbers as typical shop practice, not a universal constant. Always measure against your governing code sheet.

Cleaning, Tack Welding and Fit-Up Discipline

Before any fit-up work, remove rust, paint, oil and mill scale from both test faces. Structural programs commonly require bare, bright metal within at least one inch of the joint area on both sides. Contaminants baked into a root pass produce porosity and slag inclusions that only surface when the coupon is bent. If your test coupon is aluminum instead of steel, the cleaning chemistry and timeline change completely, and the sequence in aluminum welding surface prep is worth knowing before you walk in.

Clamp the backing bar tight, check the root gap with a gauge at several points along the joint, and correct any drift before tacking. Tack welds are part of the finished weld: they go in at the specified procedure, they must be sound, and they are inspected like every other inch of the joint. On plate coupons, tack the ends, then check squareness and gap one final time. On pipe, tack three or four spots evenly and re-measure mismatch at each.

Preheat, Position and Records Before the Arc

If the WPS calls for preheat, apply it and verify it with a temperature stick or infrared thermometer before proceeding. A representative example appears on state DOT procedures, where NCDOT pipe coupon documents specify a 50°F minimum preheat and a 450°F maximum interpass temperature (NCDOT structural steel welding procedures). Skipping preheat will not fail you visually, but it can crack the root days later, after everyone has gone home.

Then set the coupon in the required position and leave it there. Test plate welds must be made in the position being qualified, and the coupon cannot be removed from that position until the weld is complete. Finally, stamp identification on the coupon edge or backing bar before assembly, record base metal heat numbers, plate thickness, joint dimensions, consumable lot and ambient conditions on the fit-up sheet, and have the inspector initial it. Provincial practical exam rules such as Manitoba’s treat that record as part of the test, and once final visual inspection is done, removing imperfections by grinding or re-welding is prohibited.

Once the coupon is tacked, marked, gauged and signed off, the only thing left is the arc. Preparation will not win the test on its own, but skipping any step here will lose it for you before the first bead goes in.

Welding the Coupon: Hold Points, Technique and Self-Inspection

Fit-up is accepted, the root gap measured one last time, and the inspector steps back. You are alone with the arc. Everything from here turns on three habits: run the procedure exactly as written, respect the hold points, and be honest with yourself between passes.

Run the WPS as Written, Not as You Feel Like Running It

A welding procedure specification is not a suggestion, and a test coupon is the one place nobody gives you room to improvise. Amperage stays inside the range published for that filler classification and diameter. Preheat and interpass temperature get checked with a temperature stick or infrared thermometer, not judged by the color of the plate. Heat input, travel speed, bead width and pass sequence fall under the same rule.

AWS guidance identifies deviating from WPS parameters as one of the most common reasons welders fail qualification tests, and the example it gives is a classic: weaving when the procedure requires stringer beads. The weld can look beautiful and still be technically non-compliant, so the coupon fails even though the metal is sound. Read the WPS line by line before test day and ask what each variable actually forbids. If you want a refresher on why stick still dominates structural work, the reasoning is laid out in Advantages of Stick Welding: Why SMAW Still Wins.

Hold Points: Where the Inspector Stops the Test

A hold point is a scheduled interruption in the welding sequence where the examiner takes over. On pipe and pressure-vessel style tests, syllabi commonly require the root, fill and cap passes to be visually inspected while the coupon remains in the test position, so reinforcement, penetration and workmanship are judged exactly as they were deposited. On plate groove coupons the inspector may stop you after the root, before any back gouging or grinding, and the coupon is re-welded only once that hold point is released.

A second rule catches experienced welders off guard. Under state DOT test protocols such as NCDOT’s, coaching on machine set-up, fit-up, welding or inspection during the test is prohibited. The welder must demonstrate competency unaided: your instructor cannot whisper amp settings, a buddy cannot suggest a rod change, and you cannot ask the examiner for technique advice. Talking yourself through the process out loud is fine. Accepting a hint is not, and the examiner is under no obligation to treat it as anything less than a disqualification.

Stringer Beads, Weave Width and Pass Geometry

The simplest way to stay inside most structural procedures is to run stringer beads. They are easier to repeat, give more consistent penetration, and carry a lower risk of trapped slag or excessive dilution. When a procedure does allow weaving, it usually caps the width, commonly at roughly two to two-and-a-half times the electrode or filler diameter, because a wide weave lets the trailing edge of the puddle cool ahead of the arc and invites lack of fusion at the toes. Out-of-position coupons, especially 3G and 4G, tighten that limit further.

Match technique to the position you are actually welding. What works flat is not what works overhead, and rod selection matters as much as hand speed. Our All Position Welding Rod Guide covers which filler classifications behave predictably across 1F through 4F and 1G through 4G, useful reading before you commit to a coupon position you have not practiced lately.

Consumable Control: The Low-Hydrogen Discipline

Filler metal handling is part of the test, whether anyone says it out loud or not. Under AWS D1.1, low-hydrogen electrodes such as E7018 require an approved storage and usage plan, typically hermetic containers, holding ovens or rod quivers with defined exposure limits before rods must be redried or discarded, while E60XX electrodes carry no comparable storage requirement. That contrast is not trivia: damp E7018 is one of the quietest ways to put hydrogen cracking into a coupon hours after you have gone home. Fundamentals are covered in Alloy Steel Welding Electrodes: Selection, Use & Safety.

On test day, take rods only from the approved container, keep them in the quiver between passes, and never bake your own supply in an unapproved oven. If the facility provides consumables, use theirs without exception. Changing classification, diameter or brand mid-test is a deviation from the WPS, and deviations are exactly what visual inspectors are trained to spot.

Self-Inspection Between Passes

Treat every layer as if it will be judged on its own, because eventually it is. Between passes, wire brush the surface, chip the slag completely, look for porosity, undercut, overlap and poor tie-in at the toes, and confirm the previous bead is clean and correctly shaped before you cover it. Verify amperage after every change and confirm interpass temperature. Catching a defect while you are still welding is a minor problem. Catching it after the coupon is finished is a career event.

Honesty matters here too. Cracks on a coupon are unacceptable under AWS D1.1 visual criteria, and all craters must be filled to the full cross section of the weld. If you see either, deal with it while the coupon is still yours to work on.

The Point of No Return

Once the welder confirms the weld test is complete, a final visual inspection is performed, and no further welding is permitted on the coupon. You cannot add a cap pass. You cannot fix that crater. Grinding, gouging or touching up a declared-complete coupon is treated as though the test never happened or, worse, as a failure outright.

So declare complete only when you would be comfortable with that weld being radiographed. Walk around the coupon, check the stop-start points, the tie-ins, the crater at the end of the last bead, and the crown height against the limit you memorized during preparation. If something is questionable and the arc is still legal to strike, fix it now. If you are certain, say so, set the electrode down, and step away. The self-discipline to stop when it is good and to keep going when it is not is the actual skill being tested here, and it is the same discipline that keeps production welds out of trouble years later.

Testing the Coupon: Destructive and Non-Destructive Methods

The arc is out, the welder has called it done, and the coupon looks clean. Now the shop takes that weld apart, piece by piece, until nothing is left but specimens and a report. This is where a coupon stops being a weld and becomes evidence, and knowing how that evidence is built turns guesswork into deliberate preparation.

From One Coupon to Several Specimens

A finished coupon is marked first, so every cut traces back to a specific location and orientation in the joint. It is then sectioned with a bandsaw or an abrasive cut-off wheel, never with an arc torch that could alter the metal beside the cut. Depending on the test, the backing bar is machined off flush with the root surface so the root can be inspected and bent, and the face reinforcement is ground down to the specified profile. Each strip that comes out of the coupon is a specimen: a face bend, a root bend, a side bend, a tension bar, or a macro-etch cross-section.

Preparation matters more than people expect. A specimen cut too wide, cut from the wrong location, or chewed up by a careless saw cut can fail for reasons that had nothing to do with the welder’s hands. It is the one stage the candidate cannot control, which is exactly why understanding how it works removes the mystery.

Bend Tests: The Industry’s Favorite Interrogation

Bend testing stretches the weld until anything trapped, unfused or cracked opens up where an inspector can see it. The specimen is clamped and bent 180 degrees around a mandrel of specified diameter, with the mandrel size set by the code according to coupon thickness and the ductility of the metal. Under ASME Section IX the acceptance criterion is blunt: after bending, no open flaw exceeding 1/8 inch in dimension is permitted on the tension surface.

Each orientation asks a different question:

– Face bends pull the cap surface apart, exposing cracks, undercut, porosity and lack of fusion near the top of the joint.
– Root bends put the root surface in tension, which is where incomplete penetration and poor tie-in show up first.
– Side bends cut across the full weld and expose the fusion zone layer by layer, making them the best single indicator of overall penetration and interpass fusion.

Specimen Counts by Thickness

ASME Section IX Table QW-451.1 sets the required specimens for a typical groove coupon:

Coupon thickness	Typical specimens required
1/16 in. to 3/8 in.	Two tension tests plus two face bends and two root bends
3/4 in. to 6 in.	Four side bends

Tension tests complement the bends. Two reduced-section specimens are pulled until they break, and the result is compared against the minimum specified tensile strength of the base metal. Beyond the number, the fracture location tells a story: a break in the base metal away from the weld is a far better sign than one running down the weld centerline or the heat-affected zone. An all position welding rod with the wrong ductility for the base metal is one common reason a bend or tension result looks worse than the visual inspection suggested.

Macro-Etch and Fillet Fracture

A macro-etch specimen is a polished cross-section of the joint, etched with acid so the weld metal, fusion lines and heat-affected zone become visible to the naked eye. It reveals root penetration depth, sidewall fusion, internal porosity, slag inclusions and weld profile in a single glance. For fillet weld coupons, the fracture test breaks the specimen so the fusion face can be examined; a rough, torn surface with clear fusion is acceptable, while a smooth, glassy face means the weld never bonded. These same checks are how shops confirm quality on everyday aluminum work, as covered in Aluminum MIG Welding: The Complete Practical Guide.

Free-Bend and Guided Side Bend: The 46 CFR Variation

Marine work under 46 CFR Part 57 runs its own arrangement. Test plates 3/4 inch or less require one reduced-section tensile specimen and two free-bend specimens; thicker plates add a guided side bend. Free-bend specimens are judged by ductility rather than flaw size: they must show at least 30 percent elongation for most classes, and 20 percent for Class II and II-L pressure vessels. Guided side bend specimens are machined to 0.350–0.380 inch in thickness so they fit the standard test jigs shown in ASME Section IX Figures QW-466.1 through QW-466.3.

Radiography and Ultrasonics: Tests That Leave the Coupon Whole

Where destructive testing destroys the evidence, radiographic and ultrasonic examination preserves it. RT puts film or a digital detector behind the weld and produces a permanent image of the interior. Interpretation separates linear indications such as cracks, incomplete fusion and incomplete penetration, which are treated much more harshly, from rounded indications such as porosity and slag, where acceptance limits are tied to coupon thickness. UT uses sound waves instead and works especially well on thick sections where film would be impractical.

Neither method replaces visual inspection; they extend it. An inspector reads the surface, the machine reads the interior, and only together do they qualify the weld. Preparing for that combination is exactly what AWS guidance on passing a weld test emphasizes: know the required specimens, know the acceptance criteria, and inspect your own weld between passes as if the laboratory were already cutting it up.

Acceptance Criteria and the Paperwork That Proves It

Once the arc is out and the coupon has cooled, your weld stops being work and becomes evidence. Acceptance is never a matter of opinion. The code gives the inspector a fixed sequence: visual first, then non-destructive examination where required, then destructive testing, and finally the records that turn good results into a qualification you can legally work under.

Visual Acceptance: The First Gate

Nothing advances past a visual failure. For structural steel, AWS D1.1 Clause 4.9.1.1 sets the visual acceptance rules for groove welds, and they are blunt:

• Any crack is an automatic rejection, whatever its size or location.
• Craters must be filled to the full cross section of the weld, not just dabbed flat at the stop.
• Reinforcement is limited to 0.125 inch above the base metal surface.
• Undercut is limited to 0.031 inch deep.

A weld gauge measures those numbers in seconds, which is why experienced candidates check cap height and toe undercut between passes instead of waiting for the inspector. Visual also catches arc strikes, ragged tie-ins and surface contamination. Oxide and residue can make sound metal look dirty, so prep guidance in Aluminum Cleaner for Welding: How to Prep Aluminum Right belongs before test day, not after a rejection.

Bend Tests: What Counts as a Failure

After visual, the coupon is sectioned into strips and each strip is bent around a mandrel. The failure definition is precise: any crack or open discontinuity larger than 1/8 inch, measured in any direction, on the convex surface of the specimen. A shorter witness line does not fail the test.

Corner cracks are the trap. A crack at the corner of a bent specimen is generally not counted as a failure unless it results from a visible slag inclusion or a fusion-type discontinuity, so examiners study that corner carefully before ruling. The orientations also interrogate different zones: face bends stress the cap and heat-affected zone, root bends expose lack of fusion at the joint root, and side bends reveal fusion through the full thickness.

Tensile, Macro-Etch and Radiographic Results

Tensile specimens measure strength instead of ductility. Under ASME Section IX QW-150, the required tensile strength must exceed the minimum specified strength of the base metal, such as 58 ksi for A36 plate. A bar that breaks in the weld but reaches that number still passes; the code cares about achieved strength, not where the break happened.

Macro-etch sections answer what the eye cannot: penetration depth, root fusion and any lack of side wall fusion once acid reveals it. Radiographic film, where the code calls for it, exposes internal porosity, slag and incomplete penetration without destroying the coupon. Inspectors read every result against fixed tables rather than by judgment, as AWS’s Inspection Trends column makes clear.

From One Coupon to a Range of Qualification

This is where a single pass turns into a working limit. Ranges come from what was actually tested:

• Thickness: testing on 1-inch material qualifies a welder for all material thicknesses, while testing on 3/8-inch material qualifies only up to 3/4 inch.
• Diameter: a pipe coupon carries the diameter and schedule limits of the size welded, capped at the maximum outside diameter tested.
• Position: the positions tested are listed individually on the record, and what each demands of your consumables is covered in the All Position Welding Rod Guide.
• F-number: filler metals qualify by classification group, so consumable choice must stay inside your qualified group, the logic behind guides like Alloy Steel Welding Electrodes: Selection, Use & Safety.

The Paperwork: WPS, PQR, WPQ and Continuity

Results only mean something once they are documented. A procedure qualification record supports the WPS, and the WPS plus a performance test produces the WPQ or WPOR for an individual. Under ASME Section IX the procedure record goes on form QW-484A and the welder’s performance record on QW-483; AWS D1.1 uses the qualification forms in Annex N for the same purpose.

State and federal programs add their own records on top: state DOT field welder qualification reviews and marine rules under 46 CFR Part 57 both expect documentation beyond the base code, and continuity records proving you stayed qualified tie the whole file together.

Common Test Coupon Mistakes and How to Fix Them

Plenty of welders who can run a beautiful bead in production still walk out of a certification booth with a failed coupon. The reason is rarely a lack of skill. It is a small, avoidable slip that happened either before the arc was struck or after the welder declared the test complete. The AWS Welding Digest guidance on passing weld tests puts cracks caused by improper preheat, rapid cooling and hydrogen contamination among the most frequent qualification failures, and those are only the beginning. Here is how coupons fail, organized by phase, with a practical fix for each.

Phase 1: Fit-Up and Preparation Mistakes

• Misreading the governing code. This is the most expensive mistake because it happens before anyone picks up an electrode. A bevel angle, root gap or backing requirement that is correct under one standard is wrong under another. Fix: pull the coupon sketch from your actual governing standard and verify every dimension against it before cutting metal. State DOT pipe and plate certification packets, such as the NCDOT approved structural steel welding procedures, publish exactly these details for their tests.
• Dirty or misaligned fit-up. Mill scale, oil, rust and an uneven root opening show up later as lack of fusion, and the welder gets blamed for what the prep did. Fix: grind the land and both bevel faces bright, check the root gap at three points, and tack symmetrically so the plate does not pull out of alignment. If you are working on aluminum, surface prep deserves the same discipline covered in Aluminum Cleaner for Welding: How to Prep Aluminum Right.
• Wrong material or thickness for the range claimed. Using a plate outside the qualified grouping proves nothing. Fix: confirm the P-number grouping and thickness before the test, not after.

Phase 2: Welding-Time Mistakes

• Wrong progression. Running passes in the wrong sequence, skipping ahead of interpass temperature, or drifting into wide weaves where stringers are required leaves inconsistent penetration. Fix: follow the WPS pass sequence exactly and keep each bead within the width limits the procedure calls for. Reviewers who train for multiple positions cover this in All Position Welding Rod Guide: Types, Uses & Tips.
• Hydrogen cracking. Cold electrodes, skipped preheat and excessive travel speed are the usual culprits, and cracking usually appears hours after the coupon is “done.” Fix: bake or dry low-hydrogen consumables as the manufacturer specifies, hold preheat and interpass temperature with a thermometer rather than by eye, and let the coupon cool slowly. Electrode handling and storage guidance is covered in Alloy Steel Welding Electrodes: Selection, Use & Safety.
• Capping errors. Cap reinforcement that is too tall or too concave causes visual rejection before destructive testing even begins, and the extra metal hides a weak root. Fix: measure the finished cap against the visual acceptance limits while the coupon is still clamped, and stop before adding another pass.
• Post-inspection touch-ups. Once the welder declares the test complete, no further welding is allowed, and disqualification occurs if imperfections are removed by grinding or welding after final visual inspection. Fix: treat your verbal declaration as the point of no return and self-inspect hard before you say it.

Phase 3: Post-Weld Handling Mistakes

• Quenching the coupon. Community college programs warn not to quench bend coupons — they must air cool. Dipping hot steel in water traps hydrogen and changes the microstructure the test is meant to evaluate. Fix: set the coupon on a dry bench and let it cool in still air, exactly as the Portland Community College qualification guidance describes.
• Arc strikes and grinding marks after final visual. A stray arc strike on base metal is a crack-initiation site and a common rejection reason. Fix: keep the electrode holder away from finished surfaces, and never attempt to “clean up” a strike with a grinder after inspection has started.

Phase 4: Specimen Preparation Mistakes

• Cutting specimens too wide. Community college programs warn not to cut coupons wider than 3/8 inch because they will not fit the bend jig. Fix: measure each strip before cutting and trim to the width the jig and the code require.
• Wrong grinding direction and sharp edges. Bend coupons should be ground smooth with grinding marks running lengthwise and all edges rounded to avoid false failures. Crosswise scratches and square corners crack in the jig even when the weld itself is sound. Fix: finish each strip lengthwise, then break every edge with a light pass.
• Removing backing incorrectly. Cutting into the root rather than the backing bar destroys the joint you are about to test. Fix: follow the code’s specified removal method and check root thickness before slicing strips.

When a specimen does fail, the road back is predictable: re-test rules commonly require two new test specimens for each specimen that failed, and some agencies impose a 90-day wait after a second failure. Fix the root cause in writing before you schedule the re-test, because the same defect rarely forgives twice.

Retests, Continuity, Staying Qualified and Final Tips

Passing the coupon test is a milestone, not a finish line. Welder qualification is a living status: it can lapse, expire, or be challenged years after the day you welded the plate. An inspector can ask for continuity records on an ordinary Tuesday five years later, and if the paperwork does not hold up, the card in your wallet means nothing. This closing section covers how qualifications are maintained, what happens when a coupon fails, and how to walk into the test facility with everything in your favor.

Continuity: Evidence, Not Memory

Continuity means documented proof that you have actually welded with the qualified process recently enough for the qualification to remain credible. AWS D1.1 treats it exactly that way: the welder must provide documentation showing the welding process has been performed within the qualification period, verified by inspector stamps, weld stamps, or employer production records. A certificate hanging in the shop is not evidence. A signed, dated record is.

Every authority sets its own clock:

• AWS D1.1 requires continuity documentation for structural welding qualifications, verified through stamps or records.
• ASME Section IX requires work continuity records, and a welder who has not welded with a qualified process for six months must requalify in that process.
• Kentucky’s DOT specification limits qualified status to two years, unless the welder has not welded with the process for more than six months or their competence is questioned (Kentucky Transportation Cabinet K108).
• Shipyard programs under 46 CFR Part 57 require a visual acuity examination every year, a workmanship examination every three years, and program audits every two years.

Visual Acuity and Physical Requirements

Vision screening is part of qualification in most serious programs, and the shipyard rules above show the pattern: repeated on a fixed cycle rather than checked once at hire. Near and distant acuity are tested at distances the program specifies, and if you wear corrective lenses, expect the prescription to be recorded on the same form as your results. Screening is cheap and fast; discovering an uncorrected problem the morning of your test is neither.

What Happens After a Failed Coupon

A failed visual inspection or bend test usually means a waiting period, a retest procedure, and another fee. North Carolina’s DOT requires re-registration and payment of the normal test fee after a failed visual inspection or bend test (NCDOT pipe certification requirements). Other codes permit retests under conditions, and the failed coupon and its specimens are typically retained for review.

Do not book the retest before you know why the first one failed. Diagnose from the specimens: a cold-lapped root, an over-height cap, or low tensile strength caused by lack of fusion rather than base metal. A coupon welded exactly the same way fails exactly the same way.

Field Verification: When the Engineer Checks Your Papers

Qualification does not end at the test shop. Engineers reviewing field qualifications check that the documentation matches the welding procedure actually used in the current welding situation (MnDOT field welder qualification review). That means the process, position, thickness range, diameter range, backing condition and filler classification on your performance record must all cover the joint in front of you: a 3G plate qualification does not authorize a 6G pipe weld, however cleanly it bent. Continuation records such as the ASME QW-484A welding continuation record keep that story straight.

A Practical Action Plan for Test Week

1. Confirm the governing code and get the actual WPS in writing, never a procedure from memory.
2. Practice on the same material type and thickness as the test coupon in the days beforehand; AWS guidance recommends exactly this (AWS Welding Digest). An all-position rod guide helps when the coupon puts you out of position.
3. Refresh material-specific prep. If the coupon is aluminum, a proper aluminum cleaning routine is worth a rehearsal, and a structured aluminium welding course suits candidates testing in an unfamiliar process.
4. Check your vision the week before, not the morning of, and bring your corrective lens prescription.
5. Lay out PPE, consumables and grinder discs the night before, and confirm the fee and any re-registration requirement in writing.
6. Plan the weld sequence, root through cap inside the height limit, and remember that once you declare the test complete, no further welding is permitted.

Conclusion: The Plate and the Paperwork

The welder from our opening story did not fail because he could not weld. He failed because five years of production work had never taught him to judge his own cap height, and nobody had shown him what a 3/32-inch overrun looks like before it cost him a Monday morning. Test coupon welding compresses a career’s judgment into a few passes and a stack of forms: the weld gets you qualified, the records keep you qualified, and the discipline you maintain between tests is what lets you say yes when the next job calls for your stamp.

Frequently Asked Questions

What exactly is a test coupon in welding?

A test coupon is a sample weld joint made to a defined procedure and then cut into specimens for destructive testing, or examined non-destructively, to prove the weld is sound. It is the physical evidence behind a Welding Procedure Specification or a welder’s qualification record.

What is the difference between a coupon and a specimen?

The coupon is the complete welded test assembly — plates or pipe joined together. Specimens are the individual strips or bars cut from that coupon, such as face bends, root bends, side bends and tensile bars, that are actually pulled or bent in the testing machine.

How long does a welding test coupon qualification stay valid?

Validity depends on the governing code and on continuity of welding. Some state specifications limit qualification to two years unless the welder has kept actively welding with that process, while AWS D1.1 continuity allows qualification to be maintained with documented welding activity.

How many bend specimens are cut from a test coupon?

Under ASME Section IX, coupons up to 3/8 inch thick typically require two face bends and two root bends plus two tension tests, while thicker coupons use four side bends. AWS D1.1 programs commonly cut two face and two root bend specimens from a plate test.

What is the most common reason welders fail a coupon test?

Inadequate preparation and deviating from the WPS — weaving when stringer beads are required, skipping preheat, or poor fit-up — cause most failures. Visual problems such as excessive cap height, undercut deeper than 1/32 inch or unfilled craters reject the coupon before bending even starts.

Can I repair a test coupon if I find a defect?

Repairs are generally only allowed before the welder declares the test complete and before final visual inspection. Once final visual inspection is done, removing discontinuities by grinding or welding over them disqualifies the specimen under most code programs.

Do I need a new coupon test for every position?

Not always. Under AWS D1.1, a 3G and 4G plate groove test also qualifies the welder for 1G, 2G and all fillet positions on plate, so passing the harder vertical and overhead tests covers the easier positions.

What is the difference between a bend test and a radiographic test for coupons?

Bend tests are destructive: specimens are bent 180 degrees and rejected for open discontinuities larger than 1/8 inch, exposing lack of fusion or slag. Radiography is non-destructive and evaluates internal indications against linear and rounded indication acceptance limits, allowing the coupon itself to remain intact.

Which code applies to my test coupon?

It depends on the work: AWS D1.1 for structural steel, AWS D1.5 for bridges, ASME Section IX for boilers, pressure vessels and piping, API 1104 for pipelines, and CSA W47.1 in Canada. Identify the governing code before the test day because joint design, specimen count and acceptance criteria all differ.