§3.1 · Judging the weld
Visual examination and weld discontinuities on the CWI
Welding inspection on the CWI starts with your eyes: visual examination (VT) is a Part A domain of its own and the backbone of the Part B practical. You look at the joint as it's prepared, welded and finished, measure what you see, and call each discontinuity by its AWS A3.0 name.
- Exam
- CWI
- Clause
- §3.1
- Tags here
- 16
- Part A minimum
- at least 10%
3.1.1Weld inspection checklist: what to have in hand
- The governing code and its edition. Part C lists D1.1:2020 and D1.1:2025, and the 2025 edition rewrote some visual criteria.
- The loading type. Statically and cyclically loaded members sit under different acceptance columns in D1.1 Clause 8.
- Tubular or not. Tubular connections pick up extra rules in D1.1 Clause 10.
- The drawing and its welding symbols: size, length, pitch, contour and finish (see welding and NDE symbols).
- A gauge that reads the dimension in question. A fillet gauge and an undercut gauge are different tools, covered under weld inspection gauges.
- Which welds the drawing also sends to NDE, so your visual report and the NDE report describe the same joint.
3.1.2Where the eye goes, stage by stage
On Part B this is the biggest block: B5.1:2025 sets Welding Inspection and Discontinuities at a minimum of 34%, the largest minimum on that part, and Visual Examination at a minimum of 10% of Part A (AWS B5.1:2025, Table 8.1, checked October 2026). Visual weld inspection is the cheapest method on the job and the first one applied, so the exam treats it as a skill with its own vocabulary.
Before the arc
Read the joint the way the welder will meet it: groove angle, root opening, backing, alignment. Check the cut edges of plate for laminations, a mill defect in the base metal that a welder can't fix by welding over it. On a back-gouged joint, the gouge has to reach sound metal before the second side goes in.
While the welder works
Watch the numbers on the WPS being met in practice: preheat before the first pass, interpass temperature between passes (the difference is laid out under preheat vs interpass temperature). Bead sequence and placement matter on thick joints, where a buried flaw is expensive to dig out later.
The finished joint
Size and profile come first. A concave fillet carries less throat than its legs suggest; a convex one adds metal and a sharper notch, and D1.1 limits convexity. Then the toe and the surrounding base metal: undercut and overlap sit right at the toe and they're easy to swap. Pipe weld inspection adds the root side when you can reach it, and the same names apply.
3.1.3Names the exam expects, in A3.0 terms
- Undercut
- A groove melted into the base metal next to the weld toe or root and left unfilled.
- Underfill
- The weld face or root surface sitting below the surface of the adjoining base metal. Undercut is in the base metal; underfill is in the weld.
- Overlap
- Weld metal that rolls out over the toe without fusing to the base metal.
- Incomplete fusion
- Weld metal that never fused to the groove face or to the bead beside it. It isn't the same as incomplete joint penetration, where the weld stops short of the full joint thickness.
- Porosity
- Gas pockets trapped as the metal froze: scattered, clustered, linear, elongated or piping (wormhole). The forms and causes are on the porosity page.
- Arc strike
- A spot of remelted or heat-affected base metal where an arc was struck outside the joint.
- Lamellar tearing
- A terrace-shaped crack in the base metal, caused by weld shrinkage pulling through the plate thickness. A lamination is the mill defect; lamellar tearing is weld-induced.
3.1.4Call it from the description
Every tag here describes a joint or a situation in words. Picture it on the bench before you reach for a name.
0 of 16 tagged · 0 accepted
Tag 01
What is the correct term for a depression at the end of a weld bead?
Remarks on every option
- A Not the AWS term. The dip at the end of a bead has its own name.
- B Too general. The term for the depression at the end of a weld bead is crater.
- C Void is a general word for a cavity, not the term for the end of a bead.
- D Correct: A crater is the depression left at the end of a weld bead. Crater cracks start there if it isn't filled.
Pick an option. The remarks on all 4 open here.
Tag 02
How should stop-start points in a continuous weld be evaluated?
Remarks on every option
- A Height is part of it, but a smooth-looking hump can still hide a cold start. Fusion matters more.
- B Stops and starts are where cold laps, porosity and crater cracks show up. You don't skip them.
- C Counting starts tells you nothing about whether they're sound.
- D Correct: At each restart, check that the new bead fused into the old one and tied in smoothly, with no crater or cold start left.
Pick an option. The remarks on all 4 open here.
Tag 03
In a pipeline project, a welding inspector notices that each pass in the weld joint ends at the same location. What potential issue could arise from this observation?
Remarks on every option
- A Stacked stops don't help corrosion. Craters and porosity at the stops can make it worse.
- B Correct: Stacking every stop in one spot piles start/stop flaws (craters, porosity, cold laps) into one zone. Stagger them.
- C Bunched stops aren't a convenience, they're the problem. Each one is a likely flaw site, and here they all line up.
- D Flexibility isn't the concern. The concern is a weak, flaw-prone spot through the weld thickness.
Pick an option. The remarks on all 4 open here.
Tag 04
Which condition indicates improper cleaning between weld passes?
Remarks on every option
- A Surface roughness comes from technique or the process. It doesn't prove the passes weren't cleaned.
- B Some spatter is normal and doesn't point to poor interpass cleaning.
- C Color mostly reflects heat and shielding, not cleaning between passes.
- D Correct: Slag left on a pass gets trapped under the next one. Slag inclusions are the classic sign of poor interpass cleaning.
Pick an option. The remarks on all 4 open here.
Tag 05
When examining a multi-pass weld, which layer requires the most scrutiny?
Remarks on every option
- A Correct: The root sets penetration and fusion at the bottom of the joint, and it's the hardest pass to fix once it's buried. Look hardest there.
- B The hot pass cleans up and backs up the root, but the root itself carries the higher risk.
- C Fill passes matter, but they're not where penetration and fusion at the bottom of the joint get decided.
- D The cover pass is what you see last, but a nice cap can sit on a bad root.
Pick an option. The remarks on all 4 open here.
Tag 06
Which of the following is a critical factor in determining the acceptability of a weld profile?
Remarks on every option
- A Length gets checked, but you can't judge a profile from length alone.
- B Color tells you about heat and shielding, not the profile.
- C Correct: The toe is where weld meets base metal. A sharp notch, undercut or overlap there is what makes a profile reject.
- D Heat tint is oxidation color, not a profile feature.
Pick an option. The remarks on all 4 open here.
Tag 07
When examining a series of tack welds, what is the primary factor to evaluate?
Remarks on every option
- A The count only matters if the WPS or drawing calls for one. Sound tacks matter more.
- B Length gets checked, but a cracked tack is the bigger problem because it gets buried in the weld.
- C Tack color says nothing about soundness.
- D Correct: A cracked tack gets welded over and carries the crack into the final weld. Remove or repair it first.
Pick an option. The remarks on all 4 open here.
Tag 08
What is the primary focus when examining temporary attachment removal areas?
Remarks on every option
- A Coating color says nothing about damage left under it.
- B Size alone misses the point. A small spot with a crack is worse than a big clean one.
- C Location gets recorded, but what you're judging is the surface itself.
- D Correct: After a lug or clip comes off, check the spot is ground smooth with no cracks, gouges or leftover weld metal.
Pick an option. The remarks on all 4 open here.
Tag 09
How should the extent of arc strikes be documented?
Remarks on every option
- A A count without where they are and how big doesn't let anyone find or repair them.
- B The electrode doesn't help the repair. The record has to say where each strike is and how big.
- C Correct: Record where each strike is and how big, so it can be found, ground out and checked for cracks per the spec.
- D Timing doesn't matter for disposition. Location and size do.
Pick an option. The remarks on all 4 open here.
Tag 10
What is the correct method to measure weld reinforcement?
Remarks on every option
- A The throat is a fillet strength dimension. Reinforcement is height above the plate surface.
- B Root to face is close to the whole weld depth. Reinforcement starts at the base metal surface.
- C Correct: Groove weld reinforcement is the weld metal standing above the adjacent base metal surface, measured at the highest point of the face.
- D Toe to toe is the face width, not reinforcement height.
Pick an option. The remarks on all 4 open here.
Tag 11
How should surface oxidation on stainless steel welds be evaluated?
Remarks on every option
- A Wiping isn't the test. Heavy oxidation can still cost corrosion resistance; judge it against the agreed criteria.
- B Grinding hides the evidence and doesn't fix poor shielding inside the joint. Judge it first.
- C Oxide film thickness isn't a field measurement an inspector makes. Heat tint is judged by color.
- D Correct: Heat-tint color tracks how much oxidation the stainless took. Compare it to the color chart or samples the spec or customer accepted.
Pick an option. The remarks on all 4 open here.
Tag 12
How should surface irregularities in the heat-affected zone be evaluated?
Remarks on every option
- A Color hints at heat but won't show cracks or gouges in the HAZ.
- B You can't measure HAZ width with a rule on the surface, and it wouldn't tell you if the area is sound.
- C The HAZ is part of what you inspect. Cracks and arc strikes there count even if the weld is fine.
- D Correct: The HAZ is where hydrogen cracks and arc strikes like to show. Look for cracks and other discontinuities and judge them against the code.
Pick an option. The remarks on all 4 open here.
Tag 13
What surface condition must be achieved before performing a visual inspection of a weld?
Remarks on every option
- A Correct: VT only judges what you can see. Slag, spatter, scale and dirt hide cracks, porosity and toe flaws.
- B Penetrant belongs to PT, which comes after visual and needs its own prep. It isn't part of VT.
- C Paint hides surface flaws. Visual is done before coating.
- D Slag left on can hide cracks, porosity and undercut. It has to come off before you inspect.
Pick an option. The remarks on all 4 open here.
Tag 14
What should be verified first before beginning a visual weld examination?
Remarks on every option
- A Scheduling is the supervisor's business. It doesn't affect whether you can see the weld.
- B Weather matters for welding, but before VT you confirm you can actually see and reach the weld.
- C Correct: Visual inspection depends on enough light and a clear line of sight. Check both before you start.
- D The coating comes after inspection and has no bearing on whether you can do the VT.
Pick an option. The remarks on all 4 open here.
Tag 15
What is the correct term for the visual examination of a weld without the use of special equipment?
Remarks on every option
- A Not a standard term. Looking at the weld with your own eyes is called direct visual.
- B Not a recognized term. The split in visual testing is direct vs. remote.
- C Correct: Direct visual means looking at the weld with your own eyes, maybe helped by a mirror or low-power glass, with no borescope or camera.
- D Not a standard term. Visual testing is classed as direct or remote.
Pick an option. The remarks on all 4 open here.
Tag 16
In a heavy machinery plant, which NDE method would be the most efficient for quickly assessing a large number of surface welds with minimal tools?
Remarks on every option
- A PT is a good surface method, but each weld needs cleaning, dwell and developer time. It's slower than VT.
- B UT needs a trained tech, couplant and calibration, and it looks inside the weld. Not quick or tool-light.
- C RT needs a radiation source, film or detector and an exclusion zone. The opposite of quick and tool-light.
- D Correct: VT needs little more than light, gauges and a trained eye. It's the fastest way through a lot of welds.
Pick an option. The remarks on all 4 open here.
3.1.5Visual examination: what candidates ask
What tools do you get for the Part B practical?
AWS supplies them: an inspection tool kit and a set of plastic weld replicas, plus the Part B Book of Specifications as the open book (AWS CWI Examination User Guide, September 2026, checked October 2026). You don't bring your own gauges. More on the format under Part B practical.
Does the CWI cover aluminum or pipe weld inspection?
Through the code you pick for Part C. AWS lists D1.2:2014 for aluminum, API 1104 (22nd edition) for pipelines and ASME B31.1/B31.3 (2024) for piping, next to D1.1 for structural steel (AWS Exam References and Editions Guide, August 2025, checked October 2026). Part A metallurgy stays on ferrous metals.
Is PAUT weld inspection part of visual examination?
No. Phased array is an ultrasonic technique and belongs with the other methods on the NDE methods page. Visual is its own domain.
What eyesight do you need?
A near-vision test reading Jaeger J2 at 12 in (30.5 cm) or more, with or without glasses, plus a color perception test, done no more than a year before the exam on the AWS form (AWS Visual Acuity Form, August 2026, checked October 2026).
Visual calls between shifts
Our CWI prep app puts practice questions on your iPhone or iPad for the gaps between shifts.