How to Pass a Welding Test: A Step-by-Step Welder’s Guide

The weld looks fine. You grind the crown, sign the travel log, and go home thinking it’s done. Then the inspector cuts a coupon, bends it 180 degrees, and the face of the weld opens up like a zipper. That crack isn’t a failure of character — it’s a failure of preparation, and it is almost always preventable. This guide walks through exactly how to pass a welding test, from choosing the right qualification code to the last crater fill, so the coupon you deliver the first time is the coupon that passes.

What a Welding Test Actually Tests (and What It Doesn’t)

Here is the direct answer. To pass a welding test, you do two things in order: weld a coupon that conforms to a written Welding Procedure Specification (WPS), then survive the examinations the governing code requires. For most structural and pressure work that means visual inspection followed by guided bend tests. Where the code calls for it, volumetric non-destructive examination such as radiography or ultrasonic testing replaces the destructive test.

That is all it is. What trips people up is assuming the test measures “good welding” in the abstract. It does not. It is a documentation-driven check that one named person can produce one defined weld type, within one defined set of parameters, to one published standard.

The three documents welders constantly mix up

Document	What it actually is	Whose responsibility
WPS (Welding Procedure Specification)	The blueprint. Process, filler metal, parameters, preheat, joint preparation, inspection requirements.	The company or engineering office that owns it.
PQR / WPQR (Procedure or Welding Performance Qualification Record)	Proof the procedure is sound, made by welding test coupons and testing them.	Engineering or the procedure qualifier.
WPQ (Welder Performance Qualification)	Proof that you can execute that procedure within a stated range.	You, and whoever administers the test.

ASME Section IX calls the middle document the PQR. AWS D1.1 calls it the WPQR, and there that record doubles as written verification of the welder’s own qualification, so one sheet in a structural shop may be doing two jobs at once.

You will also see WPSs described as prequalified or qualified by test. As AISC explains in its welding procedure FAQ, prequalified procedures are those AWS considers sound by default, so you may use them without qualifying them yourself, which includes SMAW, GMAW except for short-circuit transfer mode, SAW and FCAW. GTAW, GMAW-s, and the electroslag and exothermic processes cannot be prequalified, so they must be qualified by test, meaning extra paperwork and extra coupons before a welder ever gets to the test itself.

What AWS means by “certification” and “qualification”

Per the AWS Welding Digest explanation of welder performance qualification and welder certification, welder certification is written certification that a welder has produced welds meeting a prescribed standard of welder performance, while welder performance qualification is the demonstration of a welder’s ability to produce welds meeting prescribed standards. Read that again and notice what it does not say. A certification is a written record of a demonstration, not a licence and not a judgement of craftsmanship.

The current structural code is AWS D1.1/D1.1M:2025, the 25th Edition, approved by the American National Standards Institute on March 19, 2025, revising the 2020 edition. Clause 6, Part C holds welder performance qualification, and the resulting WPQR must list all applicable essential variables from Table 6.12. That table is why your test coupon must match the WPS so precisely: if the amperage range, filler metal classification, position, or base material changes, you are looking at a different qualification.

What the test does not measure

• No metallurgy exam. Nobody asks about grain structure or hydrogen cracking on test day.
• No speed test. A slow, sound weld beats a fast, ugly one.
• No blueprint reading test. Reading a shop drawing is a separate skill.
• No verdict on job fitness beyond your qualified range. A 3G SMAW test says nothing about stainless, aluminium, pressure vessel fillets, hot work permits, or purge setups.

What it does measure is consistency: can you repeat, under examination conditions with an inspector watching, a weld that meets the written procedure? If you have welded stick in all positions before, you already know how. The test is about documenting it.

The three pieces of paper you walk away with

1. The WPQ record. Your symbol or number, the WPS used, essential variables, test description, actual values achieved, date, and results.
2. The continuity log. Evidence you have used that qualified process within the last six months. This keeps the qualification alive and is the piece most often neglected.
3. The range qualified. Process, position, base material, thickness range, backing, and diameter limits. This is what tells a fabricator whether they can put you on their work.

Why fabricators and owners care so much

AISC has reported that welding qualification tests in 2021 cost as much as $10,000, which is why fabricators treat documentation seriously rather than as bureaucracy. That figure covers coupon material, the testing lab, mounting and machining, and the failed attempts, not just the happy path. Multiply it by every welder on a job and it becomes obvious why an owner or general contractor may reject a technically brilliant welder who cannot produce documentation on request. It is not about trust in your hands, it is about trust in your paperwork. Specialized sectors lean on codes for the same reason: in aerospace welding and aerospace TIG work, an unqualified weld is not a small mistake, it is a weld that gets cut out.

Keep that framing in mind for the rest of this guide. Passing a welding test is a consistency and documentation problem far more often than it is a talent problem.

Which Welding Test Do You Need? Codes, Processes and Positions

“A welding test” is not one exam. It is a specific combination of code, process, joint type, position, base metal and thickness, and every variable changes what you will be asked to do and how your coupon is judged. Sort this out before you book anything or buy a single rod.

The code decides the rules of the game

The governing code sets coupon geometry, essential variables, bend test and NDT acceptance criteria, and the paperwork. Most readers meet one of these:

• AWS D1.1 — structural steel: buildings, bridges and general structural work. This is the code most shop welders qualify under. The 2025 edition is available from AWS as a public review copy, with position rules in Table 6.11 for plate and Table 10.8 for tubular work.
• AWS D1.6 — stainless steel, with its own filler metal and cleaning requirements.
• AWS D1.8 — aluminum, a different animal from cleaning to preheat to filler selection. Our guides to aluminum MIG welding and preparing aluminum properly cover the practical side.
• AWS D1.3 — bridge welding, with its own structural details and acceptance rules.
• ASME Section IX — boilers, pressure vessels and piping. The ASME BPVC Section IX preview is worth a read even if you never touch a vessel, because the logic is similar and the paperwork is more formal.
• ISO 9606-1 — the international standard used across Europe, the UK, Australia and many other markets.

Decoding the position codes

The shorthand is easier than it looks. The number is the position and the letter is the joint type: G for groove, F for fillet.

• 1G flat groove, 2G horizontal groove, 3G vertical groove, 4G overhead groove.
• 1F to 4F — the same four orientations applied to fillet welds.
• 5G — pipe in the horizontal position, rolled 45 degrees.
• 6G — pipe fixed at a 45 degree incline with the pipe rotating. The classic pipeline ticket, and the hardest position test most welders will ever attempt.

One point of confusion: pipe positions are not just harder plate positions. They are a separate qualification track with their own table.

Test performed	What you become qualified to weld
3G + 4G (vertical + overhead) groove on plate	All groove positions (1G, 2G, 3G, 4G) and all fillet positions (1F, 2F, 3F, 4F)
2G groove on plate	Flat and horizontal groove positions only, plus the matching fillet positions
6G groove on pipe	All pipe and tube positions including 5G
5G groove on pipe	5G and the flat and horizontal pipe positions

That first row is the most useful fact in this section. Pass vertical and overhead plate tests and one set of coupons unlocks every position you will ever need on structural plate. That is why serious training programmes push candidates toward 3G plus 4G rather than 1G alone.

Thickness range rules, and why they matter commercially

Qualification is not tied to one thickness. Under AWS D1.1 Table 6.2, a CJP groove test on plate thicker than 1 inch (25 mm) qualifies you for 1/8 inch (3 mm) minimum material with no upper limit, so one 1-inch coupon covers unlimited plate thickness. Below 1 inch the rule flips: a test generally qualifies you up to twice the tested thickness, so a 1/4 inch coupon caps you at 1/2 inch material.

That ceiling is why fabricators push welders to test on thicker plate, and why someone who qualified on 1/8 inch plate finds out the hard way on a 3/8 inch tank job. For heavy plate, ask the test centre to use coupon stock at or above 1 inch. There is also a bridge between joint types: AWS D1.1 Clause 6.12 lists the conditions under which a CJP groove qualification can also cover PJP groove welds, provided the WPS meets the stated limits on size and reinforcement. Otherwise PJP is its own test.

Why vertical and overhead are harder

In flat and horizontal welding, gravity helps. The puddle sits where you put it, heat rises away from the joint and the bead lands where your hand wants it. In 3G vertical, gravity pulls the molten pool downward, so you fight it with shorter beads, controlled heat input and a pause to let the puddle freeze. It is a rhythm skill more than a strength skill, but the rhythm is unforgiving.

Overhead is hard in a different way: you are welding above your head in an awkward stance and molten metal now falls toward you, so reinforcement and crater control become the whole game. A clean overhead bead needs a stable body position, a well-placed joint and a torch angle that pushes arc force into the puddle instead of blowing it out.

Add pipe. In 5G and 6G the pipe rolls or sits at an angle, so you must hold the torch and joint in a fixed relationship to your body while the workpiece moves. A 6G welder manages filler hand, torch hand, bead tracking on a rotating cylinder and heat control in three axes at once.

Matching the test to the work you actually want

Do not chase a ticket you will never use. Ask three questions: what material do you weld, what joint type do you build most, and what position do you end up in most? A shop building stainless tank shells gets more value from a D1.6 groove qualification in its real positions than from a structural 3G ticket. A pipeline contractor lives and dies by 6G. A fab shop that mostly lays out plate wants 3G plus 4G and is done.

Then check the code’s own position tables instead of assuming. Plate positions come from Table 6.11, pipe and tube from Table 10.8. Read the row for the test you plan and the column for the work you plan, and confirm the credit you expect to receive. Five minutes of reading saves a wasted test day.

Make those decisions and the test stops being a mystery. It becomes a work order: a specific WPS, a specific coupon size and shape, a specific position, a specific set of acceptance criteria. The next section turns that work order into a practice plan you can follow for four weeks.

Preparing to Take the Test: A Realistic Practice Plan

Most welding test failures are decided before the coupon is lit. The welder arrives having read the WPS that morning, spends the opening pass trying to remember the amperage range, and produces a beginner-looking coupon. None of that had anything to do with skill.

The fix is unglamorous: prepare the way the job will actually be performed. AWS D1.1 requires that a welder performance qualification follow a Welding Procedure Specification applicable for qualification, including all of its essential variables and limitations. A practice plan that copies the WPS exactly is not extra credit. It is the qualification.

Read the WPS before you walk in

Get the WPS from the shop or the testing agency a week ahead and read it until the essential variables are boring. A typical structural WPS pins down roughly this:

Essential variable	What to memorise
Welding process	SMAW, GMAW, FCAW or GTAW, plus the variant (FCAW-S vs FCAW-G)
Amperage range	Both ends of the range, not your favourite setting
Voltage	Where the WPS allows it, mostly GTAW and GMAW
Travel speed or arc time	Inches per minute permitted and the resulting heat input
Preheat and interpass	Minimum preheat and the interpass ceiling
Filler metal	Exact classification: E7018, ER70S-6, E71T-1C
Shielding gas	Type and flow rate, or none for SMAW
Polarity	DCEP, DCEN or AC

Amperage outside the range is not a small technicality. AISC’s guidance on the WPS describes qualification as welding a test assembly followed by visual inspection and bend testing or RT, with the WPS as the document everything hangs on.

Reproduce the test, not a version of it

Same process. Same filler metal. Same positions. Same base material. Same thickness if you can manage it. AISC reports that in 2021 many structural steel welders tested on 1-inch plate, which qualifies them to weld any material over 1/8 inch, so practice coupon thickness matters as much as technique.

Match the process and drill it honestly. If the ticket is SMAW, spend your hours on stick welding technique, not the gun. For GMAW, study MIG welding practice, keep GTAW drills on TIG control, and for flux core, work through flux core welding until you control the wider, hotter puddle.

Four drills that transfer straight to the coupon

1. Stringer-bead consistency. One pass, one bead, no weaving. Stop the pass and the puddle freezes. Practise until profile, reinforcement and width look identical every time.
2. Restart and crater control. Start and stop are the hardest points on a coupon. Drill ramp-up, crater fill and step-back-and-blend until you can start without a blob and stop cleanly.
3. Fillet leg consistency and sidewall fusion. Equal legs, a straight line, no cold lap along the vertical wall. Most production welders lose points here.
4. Heat management in vertical and overhead. Shorter beads, deliberate pauses, and a puddle you control so gravity works for you, not against you.

Practice is not the same as qualifying

After two weeks you can weld a pretty bead. Qualifying is different: one assembly, one continuous fill, no stopping to fix a mistake, no changing your mind halfway through a pass. Rehearse it timed and uninterrupted, because a coupon is judged on one run, not your best run of the day.

Keep your technique boring. The inspector is looking for consistent, defect-free work, and hesitation is visible. Overthought torch angles, sudden travel-speed changes, and second-guessing a bead you already laid down show up as undercut, porosity, or a wandering seam. Pick your technique in practice and repeat it on test day.

A four-week plan

Week	Focus	Checkpoint
1	WPS study plus flat coupons matching process, filler metal and thickness	Ten consecutive beads in range, no stops
2	Horizontal groove and fillet work, arc length and puddle control	Consistent leg size and sidewall fusion every pass
3	Vertical and overhead drills, starts and crater fills	Full coupon per position without a stop
4	Timed rehearsal of the actual test assembly	One uninterrupted run, plus sectioning rehearsal

One detail to check in week one: if your test assembly is a partial joint penetration groove weld not qualified by a supporting CJP groove WPS, AWS D1.1 Clause 6.12 requires three additional macroetch specimens. Rehearse your sectioning technique too, so the coupons read cleanly.

Coupon Preparation: Bevels, Root Gaps, Tacks and Fit-Up

Most welding test failures are born before the arc is struck. A textbook bead still fails if the coupon underneath had a wandering root gap, a cold tack or a bevel angle the procedure never called for. The bead is half the evidence; the joint you built is the other half.

Coupon Geometry: How Big It Has to Be

For a CJP groove weld on plate, AWS D1.1 requires the test coupon to be at least 15 in (380 mm) long in the welding direction, with a minimum width of 7 in (180 mm) per member. Figure 6.6 in D1.1 shows the expected layout.

Coupon element	D1.1 requirement	Why it exists
Length in welding direction	Minimum 15 in (380 mm)	Enough sound weld for specimens
Width per member	Minimum 7 in (180 mm)	Specimen extraction and edge distance
End discard	Minimum 1 in (25 mm) each end	Removes arc start and crater stop

At least 1 in (25 mm) is discarded from both ends of the test weldment, because start and stop are the hardest places on any weld to make sound and the code declines to grade them. What gets judged is the middle of your bead, the part you control with a steady travel speed.

Bevel Prep, Land and Root Opening

Groove geometry must match the WPS. A single-V groove is standard for CJP plate work, and the bevel angle, the land (the flat face left at the bottom of the bevel), the root face and root opening are all essential variables. Grind a different angle than the procedure calls for and you have welded to a procedure nobody qualified.

Root opening moves your qualified thickness range up or down. A 1/8 in root opening that opens to 3/16 in halfway down the coupon is a different joint from the one you rehearsed. Hold the gap with spacer bars or consistent tacking, and check it with a feeler gauge before welding.

Square or beveled groove is a real decision, not a style choice.

Groove	Code rule	Effect on your range
Square, CJP, no back gouging	Table 6.2: maximum qualified thickness is the test thickness	Test 1/4 in plate and you qualify 1/4 in and nothing thicker
Beveled, CJP	Qualifies the thickness you test	Test on the thickness your job needs, or back gouge the root before the final pass
PJP groove	Table 6.3: groove depth S of 1/8 in (3 mm) up to 3/8 in (10 mm) or greater	Qualified range changes with S; a shallow S on 1/2 in plate gives less range than S of 3/8 in

Fit-Up and Clamping

Fit-up decides where the heat goes. On flat coupons, plates sit on a flat table with mill scale ground back at the joint, held by clamps or hold-downs close to the tacks so the root gap stays uniform. Keep clamps clear of the root line and the weld path.

Horizontal coupons get mounted in a vertical plane. Clamp the ends near the tacks, not across the root, and check for bowing with a straightedge first. A coupon clamped out of plane fights you for the whole length, and you lose the test to heat input rather than skill.

Tack Welding Without Regret

A tack is a mini weld and it gets judged like one. Use the same process and consumable as the WPS, place tacks where they will be ground out, then grind and clean them before the main pass. Never bridge the root opening with a tack, and never leave a tack inside the joint you are filling.

Using SMAW for tacking is still common practice for good reasons, and the tradeoffs are worth understanding in why SMAW remains a shop favourite. A dirty rod or a different filler makes a tack a porosity site waiting for the main pass to bury, and buried defects still get found.

Marking the Coupon

Shops identify welders with soapstone, low-temperature paint or steel stamps. Keep that identification clear of the weld zone on a formal coupon: mark on the discarded end, or on the plate edge outside the joint. AWS D1.1 Clause 6.10.1 requires the whole weldment to be visually inspected before any mechanical test specimens are cut, so a coupon that fails visual inspection never reaches the bend test.

Fillet Test Coupons

Fillet coupons follow their own rule. AWS D1.1 Table 6.4 requires the fillet weld qualification test to be based on the maximum single-pass and minimum multi-pass fillet size intended for construction. If the job is a 3/8 in fillet, do not qualify on a 1/4 in bead and hope it carries over.

Filler Metal, Electrodes and Consumables: Getting the Chemistry Right

New welders often treat the electrode or the wire as a personal preference. On a qualification test it is a qualification variable. The WPS names the process and the filler metal, and your qualification extends only to the process and consumable you actually used on the coupon. Reach for the wrong rod on test day and you can produce a technically excellent weld that counts for nothing.

AISC’s guidance on the WPS puts it plainly: the procedure specification must state the welding process and the filler metal, because everything downstream depends on those two choices. AWS D1.1 enforces it through Table 6.5, the list of essential variables that force a new qualification test. Process, position, base metal outside its group and certain filler metal changes all sit there.

The AWS A5 Families You Will Actually Meet

The AWS A5 series classifies welding consumables, and most test coupons use a handful of them.

Process	Classification	Typical use
SMAW	E60xx / E70xx	E7018 low-hydrogen workhorse for structural steel; E7024 fast-fill on flat and horizontal plate; E6011 deep penetration, easy slag cleanup in all positions
GMAW	ER70S-6	Carbon steel solid wire, short-circuit and spray transfer
GMAW / GTAW	ER308L, ER309L	308L for 304 and 316 stainless, 309L for dissimilar joints
GMAW	ER4043, ER5356	Aluminium: ER4043 general purpose, ER5356 for 5000-series alloys
FCAW	E71T-1C, E71T-1M	Flux-cored wire, 1C for CO2, 1M for argon/CO2 mix

The all-position welding rod guide breaks the E numbers down in more detail, and if your test is stick, the argument for SMAW in a test bay still holds. On aluminium, silicon content changes how the puddle behaves, so aluminium MIG practice deserves its own prep routine.

Low-Hydrogen Electrodes and Moisture Control

The E70xx family is low hydrogen and dominates structural work for that reason. The coating keeps moisture out of the deposit, and moisture is the enemy: it breaks down in the arc, releases hydrogen that drives centreline cracking, disrupts the arc itself and traps porosity. Electrodes that have been sitting in a damp shed are a bend-test failure waiting to happen.

Store them the way the manufacturer says: in a holding oven at roughly 250 to 300 °F, in original packaging, with the manufacturer’s baking instructions followed if the electrodes have been exposed. If the label gives a different temperature or time, the label wins.

Electrode Conditioning, Shielding Gas and Polarity

Preparation happens before the arc. Tap the electrode against a hard surface to knock the flux coating loose, and never drag it across the plate, because dragging chips the coating and leaves the stub unshielded. Keep holder jaws clean and free of spatter, since a dirty holder causes erratic arc and repeated stubbing.

For GMAW and FCAW the shielding gas is doing as much work as the wire. A 75/25 argon/CO2 blend handles short-circuit and spray transfer on carbon steel, while 100% CO2 suits FCAW and short-circuit GMAW and burns hotter. Set flow at roughly 20 to 25 CFH and check the gun, hoses and O-ring seals for leaks. Wind is the most common field cause of porosity, because a gust strips away the protective column. Cylinders, regulators and delivery details are covered in this welding gases and supply guide.

Polarity is simple enough to memorise. DCEP, electrode positive, is the norm for SMAW and FCAW, while DCEN, electrode negative, is the norm for GTAW. Angle matters too: too little and you lose penetration, too much and you drag the puddle ahead of the arc.

Filler Metal and Mechanical Properties

The deposited metal inherits its strength from the filler, not the base plate. A 70 ksi electrode laid on 50 ksi plate gives you a 70 ksi weld. The chemistry also governs toughness, ductility and crack resistance, and those are precisely the properties the guided bend test interrogates. A coupon welded with the wrong consumable does not count toward the range you need, and it does not transfer to a job specifying a different classification.

Fillet welds add another layer. AWS D1.1 Clause 6.13 governs fillet weld qualification, and Table 6.4 sets your qualified fillet sizes from the maximum size tested in a single pass and the minimum size tested in multiple passes. A shop that tested half-inch fillets in multiple passes has not automatically qualified half-inch single-pass fillets. Finally, respect shelf life: manufacturers date their products and specify storage conditions, so record exactly what you used on the coupon.

Equipment Worth Investing In

A reliable helmet with a good auto-darkening filter, a fume extraction gun that stays close to the plume, and a solid electrode holder are not luxuries. An auto-darkening filter still needs a welder’s grade check, fume extraction only works with real airflow, and cheap jackets and gloves still get uncomfortably hot. If you are assembling a kit, this welding equipment guide is a sensible place to understand what each category involves.

Welding the Coupon: Technique for 1G, 2G, 3G and 4G

Coupons are won or lost on technique, not talent. A test coupon is judged on consistency: the same arc length, travel speed and bead shape from first inch to last.

Fundamentals for every position

• Arc length of roughly 1/8 to 1/4 inch (3 to 6 mm). Too short and you stub the arc, throw spatter and, on GTAW, contaminate the tungsten. Too long and penetration drops, the bead widens and air gets into the pool.
• Steady travel speed inside the WPS range. Drifting fast starves the puddle and undercuts the toe; drifting slow piles reinforcement and invites burn-through.
• Constant work angle and travel angle. Work angle is the electrode-to-joint angle; travel angle is how far the torch is pushed off the line of travel. Change either mid-pass and penetration and bead shape change.

If amperage control is shaky, work through an all-position welding rod guide first. An electrode that behaves in your test position is half of puddle control.

Stringer beads versus weave beads

A stringer is one pass; a weave moves the electrode side to side. Most codes and WPSs prefer stringers on qualification coupons because a weave changes the heat pattern, dwells the arc at the weave edges where it can undercut, and traps slag against the sidewall. Weaves are allowed where the WPS permits, and on wide, heavy multipass welds. If you weave, keep it narrow, slow down and clean the sidewall between passes.

1G flat

Flat is the easiest position, so most people get it wrong through carelessness. The risks are a wandering arc, inconsistent bead placement and drifting travel speed. Keep the bead centred in the groove, count your travel rhythm aloud for the first few inches, and check the profile with a straightedge. Undercut at the toe in flat position is almost always a speed problem, not an angle problem.

2G horizontal

Horizontal is where puddle control starts to matter. Let the puddle run ahead of the arc and it undercuts at the leading edge or burns through; hold it too far back and the trailing edge lacks fusion. Aim for the puddle in the leading third of the pool. Pushing gives deeper penetration and a narrower bead but more spatter; dragging gives a flatter, wider bead, easier on thin material. For a 60-degree fillet, hold a work angle near 45 degrees and a travel angle of 10 to 15 degrees.

3G vertical

Vertical travel direction is an essential variable under AWS D1.1, so a welder qualified only in vertical up is limited to vertical up, and one qualified only downhill is limited to downhill. Uphill is harder because you fight gravity and heat at once, but it is the traditional qualification for structural work and a key ticket for pipeline jobs. Control heat with short beads, brief pauses and letting each bead freeze before the next. Downhill is faster but runs away on thin material, so burn-through is the usual failure. Advanced TIG references help with vertical puddle control in GTAW.

4G overhead

Gravity helps overhead with a properly set-up joint. Put the coupon on a frame high enough to work with your arm overhead without locking your shoulder. Angle the electrode so the arc sits on the leading edge of the puddle, pushing molten metal ahead of itself. Reinforcement is the enemy: hold a flatter puddle, never linger over the crown, and resist a second pass because the first looked high. A bead that droops into a lumpy roof fails the 3/16 inch reinforcement limit.

Multi-pass welding

• Cap with a stringer over the previous layer, or a slight controlled weave where the WPS permits.
• Grind, chip or gouge slag and spatter between passes; carbon arc gouging is covered in this air arc and gouging guide.
• Keep interpass temperature inside WPS limits, using a preheat thermometer rather than a guess.
• Cap fillet welds with a stringer for consistent leg lengths. AWS D1.1 Figure 6.16 defines the fillet weld soundness coupon as a T-joint or lap-joint assembly verified by macroetch cross-sections.

Arc starts and crater fills

The start and stop are the hardest places to make sound, which is why one inch is discarded from each end of the coupon before specimens are cut. Start at the very edge of the coupon or on a runoff tab if the code allows, so the crater lands in the discarded zone. Ramp the arc up gradually rather than stabbing the puddle. Never snap the arc off at the end: break it and hold the electrode still in the pool so the metal freezes, or step back half an inch and blend into the bead. An abrupt stop leaves a crater crack.

Work position and fixture setup

The coupon must be welded in the position being tested, and it must be rigid. Any movement, from a slipping clamp or a flexing plate, shows up as cracking in the bend specimen. Flat coupons sit on a flat table, horizontal coupons need clamps or strong magnets, vertical coupons go in a frame, and overhead coupons need a purpose-built jig at a workable height.

After the Arc: Visual Inspection, Bend Tests and NDT

The arc is off, the slag is chipped, and the inspector is crouching at the toe of your bead. That is the moment your coupon stops being yours. Acceptance criteria change how you weld. The standards are the recipe, and you are the cook.

The Sequence: Visual, Then NDT, Then Destruction

Each step can end the process before the next one begins.

1. Visual inspection of the completed weld against workmanship limits.
2. Volumetric examination (radiography or ultrasonic testing) where the code requires it.
3. Mechanical testing, where specimens are cut, machined, and bent or pulled.

A coupon that fails visual inspection is never sectioned. There is no “see how the bend test goes anyway” option. AWS D1.1 sets the visual acceptance criteria in Clause 4.9, and for structural qualification this filter decides whether further testing happens at all.

What AWS D1.1 Clause 4.9 Actually Rejects

Defect	AWS D1.1 visual acceptance limit
Cracks	Not permitted, no matter how small.
Incomplete fusion	Not permitted.
Undercut (groove welds)	Maximum 1/32 in (0.8 mm).
Weld reinforcement	Maximum 3/16 in (4.8 mm).
Overlap	Not permitted.
Arc strikes outside the weld zone	Automatic rejection of the weldment.

A stray strike, even on mill scale, even ground flat, rejects the coupon. That is why test plates are cleaned of paint, oil and rust before you strike an arc, and why you never “touch up” a coupon on a plate you did not prepare.

Reinforcement deserves a second look. Many welders aim for a tall, proud bead, but 3/16 in (4.8 mm) is the ceiling. That is less metal than most people picture.

Guided Bend Testing: The Real Test

After visual and NDT, specimens are bent in a jig around a plunger to roughly 180 degrees, putting weld metal in tension across the root. Defects show as cracks on the tension surface, measured with a scale.

Acceptance is governed by ASME QW-160.1.7:

• No single discontinuity may exceed 1/8 in (3.2 mm) measured in any direction.
• The sum of the greatest dimensions of all discontinuities over 1/32 in (0.8 mm) but not exceeding 1/8 in (3.2 mm) may not total more than 3/8 in (9.5 mm).
• Corner cracks may not exceed 1/4 in (6.4 mm), unless the crack results from a visible slag inclusion or other fusion-type discontinuity, in which case the stricter 1/8 in (3.2 mm) maximum applies.

Corner cracks get a longer leash than face cracks because a corner crack is a tip problem where bend stress concentrates. But one tracing back to a visible slag inclusion is a fusion-type defect, judged against the tighter 1/8 in limit.

If a specimen has corner cracks exceeding 1/4 in (6.4 mm) and no evidence of a slag inclusion or other fusion-type discontinuity, the specimen is disregarded, and a replacement is cut from the original weldment and tested.

ASME lists the qualification mechanical tests, including tension, guided bend, fracture and macro examination, in QW-302.1, with purposes defined in QW-140. Tension pulls weld metal until it fails. Macro examination means cutting the weld open and polishing it to inspect the penetration profile and fusion lines.

Radiography and Ultrasonic Testing

Not every coupon gets bent. For full joint penetration groove welds, AWS D1.1 Clause 6.10.2 requires 100% radiographic or ultrasonic examination of the CJP groove weldment before mechanical test specimens are cut. Cutting the coupon open for a bend test would destroy the evidence RT or UT was meant to find.

Radiography images internal defects with X-rays or gamma rays. It is excellent for porosity and slag but sensitive to surface condition, and it wants simple plate geometry. Ultrasonic testing sends sound through the metal, catching planar defects like cracks and lack of fusion, but it needs skilled technicians and a clean surface. RT is the usual choice because it leaves a permanent record.

For the wider view of how the qualification package fits together, read our guide to welder performance qualification and welder certification.

Fillet Weld Macro and Fracture Testing

Fillet welds are qualified differently, which matters for structural tube or plate connections. Two approaches are common: the fillet macro examination, and the fillet fracture test, where a fillet weld is loaded in a fixture that stresses the weld throat until it fractures.

Per ASME QW-183, the acceptance criteria for a fillet weld macro examination are:

• Leg length difference: no more than 3 mm between the two legs.
• Concavity or convexity: no more than 1.5 mm.
• Complete fusion through the throat, with no cracks.

The 3 mm leg difference is stricter than most shop practice allows, so expect plenty of coupons that “look fine” to fail it. A fillet measuring 6 mm on one side and 9 mm on the other is not a weld you would hang a load on, yet no fatigued eye catches it. Cut open and measured, the difference becomes obvious.

Complete fusion with no cracks is where many fillet qualification coupons fail. A fillet that looks like a solid triangular blob can still be cold-lapped along the sidewall if you pushed the puddle ahead of the arc without dwelling at the toes. On the macro section it shows as an unfused boundary, and the specimen is rejected.

Why Knowing This Changes Your Welding

Notice the pattern: undercut capped at 1/32 in means a consistent arc length, reinforcement capped at 3/16 in means no fat beads. Every criterion maps back to a decision you make seconds before you strike an arc. For how a WPS is built and why essential variables matter, see the AISC welding procedure specification FAQ.

Why Welders Fail: The Most Common Reasons and How to Fix Each

Most failed coupons trace back to a short list of preventable defects. The root cause is rarely a lack of skill. It’s a lack of discipline on the day. Here is what inspectors actually find, and how to fix each one before it reaches your bend test.

1) Slag Inclusions

Cause: Weaving too wide, slag left behind after multi-pass grinding, or skipped interpass cleaning.

Code Consequence: Slag is an internal discontinuity, and ASME guidance holds corner cracks from a visible slag inclusion or other fusion-type discontinuity to much stricter limits, which can change the acceptance judgement on a bend specimen entirely.

Fix: Run stringer beads instead of wide weaves, chip and brush between every pass, and grind any questionable spot clean before laying the next bead.

2) Incomplete Fusion and Lack of Sidewall Fusion

Cause: Too little arc length, an incorrect work angle, or insufficient preheat.

Code Consequence: Fusion is non-negotiable. Under AWS D1.1 Clause 6.10.2, any internal defect that violates the acceptance criteria of Clause 8 mandates rejection of the coupon, meaning destructive sectioning cannot proceed on a defective weldment.

Fix: Direct your arc into the sidewall, not just down the middle of the joint. Keep arc length tight, verify preheat against your WPS, and take time to wash into the toes of the bevel.

3) Porosity

Cause: Damp electrodes, contaminated shielding gas, oil, paint or mill scale on the plate, or a gas leak in the torch.

Code Consequence: Pervasive porosity will not be accepted and will cause a failed bend test.

Fix: Store low-hydrogen electrodes in a holding oven per the manufacturer’s instructions, clean plate down to bare metal, check gas fittings for leaks, and account for wind. Outdoors, block the breeze or reposition, because even a slight draft pulls your shielding away.

4) Undercut

Cause: Too much travel speed, excessive current, or a wandering torch.

Code Consequence: Under AWS D1.1 Clause 4.9, undercut on groove welds is limited, and excessive undercut results in rejection.

Fix: Slow your travel down, hold the puddle slightly in front of your arc, and keep arc length consistent. A slightly slower, controlled pass beats a rushed one every time.

5) Excessive Reinforcement and Convexity

Cause: Too much filler metal, an improper electrode angle, or too slow of a travel speed.

Code Consequence: Excessive reinforcement fails visual inspection and creates stress risers, putting your bend specimens at much higher risk of cracking.

Fix: Deposit only what the joint needs. A slightly convex, controlled bead is acceptable on many coupons, but a tall, ropey cap almost never is. Build the cap gradually with stringer passes.

6) Burn-Through

Cause: Thin plate, an excessively wide root gap, or too much heat input.

Code Consequence: Burn-through creates an open discontinuity, which is not permitted and is an automatic fail for a CJP groove coupon.

Fix: Dial amperage down, hold a tight and consistent root gap, and use smaller beads with proper pauses. If fit-up is off before you strike an arc, fix the fit-up rather than welding over a gap that’s too large.

7) Cracks (Including Centreline Cracking)

Cause: Excessive heat input, rapid cooling, hydrogen from damp consumables, or a highly restrained joint.

Code Consequence: Cracks of any kind are not permitted.

Fix: Follow your WPS for interpass temperature, use properly stored low-hydrogen consumables, and allow controlled cooling. Restraint is your enemy on test fixtures, so avoid over-clamping the coupon.

8) Arc Strikes Outside The Weld Zone

Cause: Carelessness while striking, positioning, or moving the electrode holder.

Code Consequence: Under AWS D1.1 Clause 4.9, arc strikes outside the weld zone are grounds for rejection, so stray strikes must be ground out and re-welded into sound metal.

Fix: Be deliberate with every strike and keep your electrode over the joint area. If you do land a stray strike, remove it down to sound metal by grinding and repair it properly, or prevent it entirely.

9) Start and Stop Crater Defects

Cause: Abruptly breaking the arc without filling the crater, or poor restart technique.

Code Consequence: Starts and stops are the most inconsistent areas of any weld, which is why AWS D1.1 requires at least 1 in (25 mm) to be discarded from each end of the weld before specimen extraction.

Fix: Use proper crater fill techniques such as ramping down or a step-back and blend. Give the puddle a moment to fill before breaking the arc.

For welders looking to sharpen their eye, macro films are invaluable. After a coupon is cut and etched, a macroetch cross-section shows the true profile of your fusion, penetration, and any hidden discontinuities. Be aware that for certain joints, AWS D1.1 Clause 6.12.6 and Figure 6.15 require these macroetch cross-sections to verify soundness when not supported by a CJP groove WPS. Even outside those requirements, a good welder reads their own coupon after it is cut. With fillet welds, remember that AWS D1.1 Table 6.4 requires the fillet qualification test to be based on the maximum single-pass and minimum multiple-pass fillet size used in construction, so an undersized fillet is a failed coupon.

At the end of the day, most test failures aren’t from a lack of natural talent. They’re from rushing the things that are boring to practice. Clean metal, clean consumables, controlled heat, and disciplined technique will take you much further than anything flashy. Hone those fundamentals, and you’ll be cutting bend specimens out of a passed coupon.

Retakes, Continuity and the Paperwork That Keeps You Qualified

This part of passing a welding test catches experienced people by surprise. You can nail the coupon, pass the bend test, and walk out with a Welder Performance Qualification (WPQ) that is worthless six months later because you never used that process again. Qualification is not a trophy. It is a live status you either maintain or lose.

The Six-Month Continuity Rule

Both major codes tie qualification to recent use. AWS D1.1 Clause 6.2.3.1 and ASME Section IX QW-322.1 state the same principle: a qualification remains valid provided no more than six months have passed since the welder last used the qualified process. Both treat welding as a perishable skill on paper. You are not disqualified because you got rusty. You are disqualified because the code cannot show you are still current.

Here is what costs people money: a lapsed qualification cannot be reinstated with paperwork. If six months pass without use of the qualified process, the only path back is a new coupon and a fresh test. A supervisor’s letter, a training record or a course certificate does not substitute for a retest.

What the WPQ Record Must Contain

Your qualification lives or dies on one document. ASME Section IX requires that each qualified welder and welding operator be assigned an identifying number, letter or symbol by the organization (QW-301.3), and that this identifier be used to identify that person’s work (QW-301.4). A reviewer expects to see:

• Welder identification — the number, letter or symbol assigned by your organization
• The WPS used, identified by number or specification
• Essential variables — the ranges you were tested within
• Type of test — bend, macro, radiographic or ultrasonic
• Test results — measured values against the acceptance criteria
• Ranges qualified, per QW-452 and Table QW-461.9
• The date

ASME supplies suggested forms QW-484A and QW-484B for documenting welder and welding operator performance qualification. They are formats, not mandates, and many shops use their own, but starting from the ASME layout saves arguments later with an inspector. If your employer’s form omits an item above, you are the one sent back to the test booth.

Retesting Versus Starting Fresh

When a coupon fails, know what happened. A retest re-examines the same coupon or weldment where the code allows it. A new coupon means fabricating a fresh test assembly and welding it from scratch. Codes permit limited retesting of failed specimens, and ASME Section IX QW-322.2 adds a caveat: when there is specific reason to question a welder’s ability to make sound welds, the qualifications supporting the welding in question are revoked while other qualifications remain in effect. One bad coupon can put your ticket under review without wiping out unrelated tickets. Treat a failure as a signal to fix technique, not an excuse to try again unchanged.

On the employer side, AISC states that properly documented evidence of previous qualification of welders, welding operators and tackers should be accepted without re-qualification provided the period of effectiveness has been maintained. A well-documented WPQ from a previous employer, still inside its window, is worth real money. Arrive with that documentation already organized.

Moving Employers and Keeping Your Records

When you change jobs, get a copy of your WPQ before you leave, not after. Keep the pages showing your identity, the process, the positions and the qualified range, plus the continuity log of recent use. Your new employer will typically re-verify continuity and may want to witness a small continuity weld or re-test if your records are thin. The WPQ is a personal credential: employers change, but the stamp on your weld does not.

Passing the coupon is half the job. Keeping the qualification valid, documented and defensible decides whether you can sign for production welds six, twelve or thirty-six months from now. Treat the WPQ as a living record rather than paper in a drawer, and the administrative side becomes a career asset instead of lost pay.

Safety on Test Day and Your Next Step Plan

Nothing ruins a test coupon faster than a distraction. The fume hood is 15 feet away instead of 6 inches, your eyes are watering, and you keep welding because stopping feels like failure. Then the undercut, the porosity and the sloppy crater show up in the bend test. Safety on test day protects the quality of your work.

The big one for welders is hexavalent chromium. Welding stainless steel and chrome-bearing alloys releases Cr(VI), a known carcinogen. OSHA’s welding and hexavalent chromium fact sheet spells out the rules: the permissible exposure limit is 5 micrograms per cubic metre of air as an 8-hour time-weighted average, enforced under 29 CFR 1910.1026 for general industry, 29 CFR 1926.1126 for construction, and 29 CFR 1915.1026 for shipyards.

NIOSH recommends something far more protective: limiting airborne exposure to all Cr(VI) compounds to 0.2 µg Cr(VI)/m³ as an 8-hour TWA over a 40-hour workweek, according to the NIOSH current intelligence bulletin 68. NIOSH also estimates roughly 558,000 U.S. workers are exposed to airborne Cr(VI) in the workplace, with welding among the industries with the most exposed workers. If you touch stainless, chrome-moly or hardfacing, that is why fume control gear is not optional.

The hierarchy of controls, in plain language

The order matters:

1. Eliminate or substitute. Can the job be done without a Cr(VI)-generating material? Is there a lower-hazard process or consumable?
2. Engineering controls. Local exhaust ventilation, extraction guns with the nozzle kept close to the arc, designed welding booths. OSHA’s interpretation on ductless welding booths notes that eliminating recirculation back into the workplace meaningfully reduces Cr(VI) exposure.
3. Administrative controls. Rotation, schedules that limit time near the plume, training, restricted access while welding.
4. Respiratory protection. The last line, used when the controls above cannot keep you below the PEL.

Two facts from the OSHA fact sheet surprise people. Welding outdoors does not guarantee adequate ventilation: a gentle breeze is not local exhaust, and the plume must be captured before it reaches your breathing zone. And surfaces must be cleaned of paint and solvent residue first, because solvent vapour plus a hot arc is serious, while chromium-bearing paint turns a routine weld into a Cr(VI) exposure.

Stay upwind of the plume in open areas, keep the extraction nozzle within a few inches of the arc, and never assume a big space is a safe space. If the test is scheduled inside a tank, vessel, hopper or pipe, stop. Confined space welding without ventilation kills workers, and permit-required confined space rules cannot be waived because the inspector is waiting.

PPE is the last layer and worth doing properly: a properly fitted respirator, safety glasses under your helmet, fire-resistant clothing, gloves that let you feel the torch. Aluminium fume differs but is not harmless, and preparation matters most. Our guide to preparing aluminium for welding covers degreasing and cleaning, and our aluminium MIG welding guide covers fume extraction setup.

Your 30-day action plan

Passing a welding test is a consistency problem far more than a talent problem. Almost every failed coupon traces back to a decision made weeks earlier.

Window	Action	Outcome
Week 1	Pick your code and position: D1.1, D1.6, ASME IX or ISO 9606-1, matching the work you want.	A test you qualify for, not the hardest on the wall.
Week 1 to 2	Get the WPS. Put the essential variables on a card: process, amperage, voltage, travel speed, preheat, filler metal, gas, polarity.	No surprises on test day.
Week 2 to 3	Two or three practice coupons a week in the test position, welded straight through without stopping.	Repeatable bead shape and heat input.
Week 3	Master fillet and groove control: consistent legs, sidewall fusion, uniform root reinforcement.	Visual inspection passes before the bend test.
Week 3 to 4	Drill starts and craters: ramp-up, dwell, step back, blend. Practise the first and last 30 seconds of every weld.	You stop losing points at the ends of the coupon.
Week 4	Book the test. Bring your own filler metal, PPE, and paperwork that shows your continuity.	The coupon you deliver the first time is the one that passes.

One last piece of advice. Read your coupon after it is cut, find the fracture face, and work out why it went there. A welder who studies their own failures becomes qualified fast. Then take the paperwork as seriously as the arc: set up a continuity tracker, keep copies of every WPQ you earn, and passing a welding test turns into doing the same thing the same way, on paper and off.

Frequently Asked Questions

What is the difference between a welder qualification test and a welding certification?

A qualification test is the hands-on event where you weld a test coupon and it is examined; the result is a Welder Performance Qualification (WPQ) record. A certification, such as the AWS Certified Welder program, is a separate credential issued by AWS after a test administered under their own rules. Many states and employers accept either, but a code WPQ is what the contract documents usually require.

Which welding test is the hardest — 1G, 2G, 3G or 4G?

For most welders, 3G (vertical) and 4G (overhead) are the hardest because gravity changes puddle control, travel speed and filler placement. Testing 3G and 4G together is the standard route to a ticket that qualifies you for every position. Flat and horizontal positions are easier to hold consistent, which is why 1G and 2G alone give you the narrowest qualification range.

How thick a plate can I weld after I pass my test?

It depends on the thickness of the coupon you welded. Under AWS D1.1 Table 6.2, a CJP groove test on a plate 1 inch thick or thicker qualifies 1/8 inch minimum through unlimited thickness. A test on thinner plate generally limits you to twice the test thickness. This is why testing on 1-inch plate is the common choice for welders who want maximum range.

Can I qualify by radiography or ultrasonic testing instead of bend tests?

Yes, most codes allow it. AWS D1.1 Clause 6 lets you qualify a groove weld by RT or UT instead of guided bend testing, and ASME Section IX QW-302.2 likewise allows qualification by volumetric NDE. The catch is that a CJP coupon must pass 100% volumetric examination before any mechanical specimens are cut, so a defect found by RT stops the process before sectioning.

Do I really have to retest every six months?

You do not retest, but you do have to keep using the process. AWS D1.1 Clause 6.2.3.1 and ASME Section IX QW-322 both require you to have welded with the qualified process within the last six months. If you stop, the qualification lapses and you must produce a new coupon and pass the examinations again — there is no paperwork-only reinstatement.

What are the most common reasons welders fail a bend test?

Slag inclusions, lack of fusion, porosity, and cracks that start in the heat-affected zone. Most trace back to four habits: weaving too wide, not cleaning between passes, letting the arc length get long, and welding with damp or contaminated consumables. A 1/8-inch maximum single discontinuity is the limit, and corner cracks over 1/4 inch without visible slag can cause a specimen to be thrown out entirely.

How long does it take to prepare for a welding test?

A focused four-week plan is usually enough for a welder who is already productive. Two weeks of position-specific practice coupons, one week drilling starts, crater fills and fillet leg consistency, and one week of mock test coupons welded with the same WPS, filler metal and position as the real test. Experience with the process matters more than elapsed time.

What should I bring and wear on test day?

Bring your safety glasses, gloves, leather jacket or sleeves, your own helmet if the shop allows it, and a copy of your WPS and any prior WPQ records. Wear cotton or other flame-resistant clothing, closed leather boots and a welding cap or bandana. Expect the shop to supply the coupon, the filler metal and the tacks, and expect the welder to be asked to sign the travel log before the arc is struck.

Is AWS D1.1 the same as ASME Section IX?

No. AWS D1.1 covers structural steel construction, while ASME Section IX governs welding and brazing qualifications for boilers, pressure vessels and similar pressure-boundary work. The test structure is similar — coupon, visual, bend or NDT — but the essential variables, test positions and qualification ranges are different, so you need separate tickets for structural and pressure work.

What safety limits apply to welding fumes on a test day?

OSHA sets a permissible exposure limit of 5 micrograms per cubic metre for hexavalent chromium as an 8-hour time-weighted average, and NIOSH recommends a much stricter 0.2 µg/m³. Practically, that means a fume extractor or local exhaust positioned close to the arc, clean unpainted steel, gas flow set correctly, and never welding in a confined space without ventilation and a permit.