Prepare for CSWIP 3.0 by structuring your revision around the inspection workflow: checks before welding, during welding and after welding, then drilling the imperfection catalogue and acceptance reasoning within each stage. The distinguishing skill the subject demands is telling imperfections with similar appearances apart, linking each one to its likely cause, and recording findings as a visual inspector would. Work through sketch-based weld symbol exercises, gauge-based measurement examples, and written scenarios where you state the imperfection, its probable cause, and the correct documentation action. Check your readiness with the rubric in the final section rather than with a single practice score.
What changes between before, during and after welding inspection
Visual inspection is a staged activity, not one final look. Before welding you verify materials, joint preparation and consumables; during welding you monitor process parameters and inter-run condition; after welding you assess the finished joint against the acceptance criteria.
Pre-welding checks focus on inputs. Compare the material certificates and marking against the drawing, examine bevel angle, root face (land) and root gap with a rule or gauge, and confirm cleanliness, freedom from moisture, and correct consumable selection and storage. An error here, such as an excessive root gap, is invisible in the finished weld but produces root defects later, which is why staged inspection catches problems the cheapest way.
During-welding checks shift to process control. You observe preheat and interpass temperature where specified, run sequence, interpass cleaning and slag removal, and the condition of each run as it is deposited. Post-welding inspection is the stage most people picture: surface finish, weld profile, dimensional checks and the surface imperfection catalogue. Treat the three stages as three separate checklists when revising, because questions and site practice both hinge on knowing which check belongs to which stage.
- Before welding: material identity, joint preparation geometry, fit-up, consumables, cleanliness
- During welding: preheat and interpass temperature, run sequence, interpass cleaning, arc behaviour
- After welding: weld profile, dimensions, surface imperfections, final visual assessment against the acceptance criteria
Telling undercut, lack of fusion and other look-alike imperfections apart
Several imperfections present similar surface appearances but differ in cause and significance. Undercut is a groove melted into the parent metal at the weld toe; lack of fusion is an unwelded interface, typically along a fusion face, and the two require different reasoning.
Train yourself to locate the imperfection before naming it. Undercut sits in the parent metal adjacent to the weld toe, has a sharp edge and a continuous or intermittent groove, and is often linked to excessive current, excessive travel speed or a wrong torch angle. A local dip in the weld metal at the toe, by contrast, may be a slight depression in the weld profile rather than metal loss from the parent, and the distinction changes what you write on a report.
Lack of fusion is harder because it is frequently internal, but surface indications can appear where the weld metal has rolled over without bonding, leaving a sharp, overlapping edge rather than a groove. Compare that with overlap, where excess weld metal has flowed onto the parent without fusing, producing a smooth, rounded lip. Practice with photographs or samples: identify the boundary, ask whether parent metal was lost, whether weld metal was added without bonding, and whether the feature is continuous or intermittent. Make this habit of questioning before naming explicit in your practice by forcing yourself to answer those three questions in writing before stating any imperfection name.
Surface evidence of internal imperfections: the honest limits of visual testing
Visual inspection assesses what is visible on the surface. Features such as root concavity, excess penetration, burn-through and surface-breaking porosity are directly observable; lack of side-wall fusion or internal slag generally is not, and must be inferred only from visible evidence.
On a single-sided butt weld viewed from the root, concavity appears as a shallow, saucer-shaped sag of the root bead; excess penetration appears as proud, bulbous metal protruding through the root; burn-through shows a local hole with a rim of collapsed metal. Each has a distinct cause family: concavity often reflects excessive root gap with low current, burn-through reflects excessive heat input or gap, and excess penetration reflects high current at low travel speed. Matching the visible geometry to these cause families is the reasoning chain the scenario material tests.
Equally important is knowing what you cannot conclude. A smooth, sound-looking external weld can still contain internal lack of fusion; you should never certify an internal condition from visual evidence alone. In reports and exam answers, describe what is visible, note any evidence that suggests a buried imperfection, and recognise that volumetric methods exist precisely because VT has this boundary. Reviewers of your reasoning should see cautious, evidence-based language rather than confident claims about hidden conditions.
Reading weld symbols and drawings without guessing
Inspection decisions start from the drawing. Weld symbols on a joint line specify weld type, size, extent, and side of the joint, and misreading them leads to measuring the wrong feature or accepting a weld that does not match the specification.
Work through the symbol system systematically: the reference line, the arrow line pointing at the joint, and the symbols placed above or below the reference line to indicate which side of the joint is welded. A fillet symbol with a size written to its left gives the leg length; supplementary symbols and notes cover peripheral welds, field welds and weld-all-round instructions. Symbols from both the ISO and AWS systems differ in conventions, so identify which system your reference documents use before interpreting dimensions.
A practical exercise: take any welded assembly you have access to, sketch its joint, write an intended weld, and then convert it into a symbol, then reverse the exercise by drawing a symbol and sketching the joint it describes. Expected observations include noticing how easily a fillet size placed on the wrong side of the reference line reverses the meaning, and how a double-sided symbol requires you to specify each side independently. Repeated sketching cements symbol interpretation far better than reading a symbol chart once.
Measuring welds with gauges: fillet legs, throat and profile decisions
Dimensional assessment uses simple tools correctly rather than sophisticated tools loosely. Fillet gauges, rules and calipers establish leg lengths, throat thickness, convexity, concavity and alignment, and the measurement technique determines whether the reading means anything.
For a fillet weld, the required leg length is usually the specified dimension, and the actual throat thickness depends on the weld profile: a concave fillet can have adequate legs but a throat below the required value, while a convex fillet can show large legs with excessive reinforcement. Learn how a fillet weld gauge seats against the leg and toe, and how misalignment of the gauge or measurement at a crater or a stop-start position distorts the reading. Measure at representative points and note where readings vary along the joint.
Consider two worked examples. Example one: specified fillet leg 8 mm, measured legs 8.5 mm and 8.5 mm with a distinctly convex profile; the legs pass but convexity must be assessed against the profile criteria, so the better decision is to record the profile observation, not just the legs. Example two: measured legs 8.0 mm and 7.0 mm with a concave toe; relying on the larger leg alone would miss the reduced throat, so the better decision is to evaluate the throat thickness against the requirement. The point both examples share is that a single dimension never describes a weld profile; geometry and dimensions are assessed together.
| Feature | Tool typically used | Common measurement pitfall | What to record |
|---|---|---|---|
| Fillet leg length | Fillet weld gauge | Gauge not seated square to the toe | Legs at representative points |
| Fillet throat thickness | Gauge / calculation from profile | Assuming throat from one leg only | Throat with profile description |
| Undercut depth | Gauge or depth comparator | Measuring the groove length, not depth | Depth, length, location |
| Alignment / misalignment | Straightedge, rule | Measuring across a crowned surface | Offset relative to reference surfaces |
| Excess weld metal | Rule or gauge | Measuring on an uneven plate surface | Height relative to parent metal |
Linking imperfections to causes and to the inspector's correct response
Every named imperfection has a cause family rooted in process variables: heat input, technique, fit-up, consumable condition and material condition. The inspector's role is to identify, describe, locate and record, and to refer findings through the reporting route rather than decide repairs unilaterally.
Build a cause map rather than a flat list. Porosity clusters point toward gas coverage problems, contaminated or damp surfaces, or consumable condition; solid inclusions point toward inadequate interpass cleaning and slag trapping in a narrow groove; hydrogen cracks in the heat-affected zone of a hardenable material link to preheat and moisture. A useful practice exercise is to take each imperfection name and write a short, plausible weld history that could produce it, then trace the visible imperfection back through these variables rather than jumping to a repair instruction.
Worked scenario: a butt weld shows intermittent surface cracking running along the weld centreline shortly after welding. A plausible mistake is to name it as slag entrapment because the marks look linear and dark, and to recommend immediate re-welding. The better decision is to recognise that crack-like indications are the most significant category, to record the exact location, length, orientation and appearance, and to refer the finding for appropriate follow-up examination and a qualified repair decision. Why it matters: mislabeling a crack masks its significance, and an unauthorised repair decision exceeds the visual inspector's role. The same reporting discipline applies to arc strikes, which are small but reportable because of their potential to initiate cracking in some materials.
Documentation, ethics and the boundaries of the visual inspector's role
A visual inspector's value lies in accurate, dated, traceable records and in knowing the limits of the role. Reports describe observed facts against the specified criteria; they do not fabricate readings, sign off unseen work, or authorize deviations from the specification.
A usable visual inspection record links each finding to its location, the acceptance criteria applied, and the stage of inspection. Compare two styles: 'weld looked acceptable' tells a reviewer nothing, whereas 'root run visual check at joint 14, no visible surface defects, interpass temperature within the specified range, reference drawing X' allows anyone to reconstruct what was verified. Practice writing one-line findings in this style for every imperfection name you revise, and you will internalise both vocabulary and reporting discipline together.
Ethical boundaries appear in scenario form: a colleague asks you to record a fillet as the specified size when your gauge shows a shortfall; a weld is covered by paint before your final visual check. The professional response in both cases is to record actual observations, decline to certify what was not inspected, and escalate through the reporting line. When revising, pair each imperfection topic with one reporting or ethics question so the knowledge and the judgement are rehearsed together. For administrative details of the certification itself, such as entry requirements and renewal, refer to TWI Certification rather than to secondary sources.
- Record facts: what was seen, where, when, against which criteria
- Never record an inspection that did not take place, including work covered by paint or assembly
- Refer cracks and other significant findings for specialist follow-up rather than deciding repairs yourself
- Keep measurement records traceable to the drawing revision in force
References and further reading
Use these references to explore the concepts and check the latest information from the relevant organizations.