Ultrasonic testing at Level II is less a memory test than a decision test: every indication forces you to choose between interpreting, re-scanning, or classifying, and every setup forces a choice of angle, frequency, and reference sensitivity. Build your study around those decision points. For each topic, ask what choice the knowledge supports and rehearse the choice aloud. Compute a skip distance, then decide which inspection angle it justifies; convert a dB difference, then decide what it means for sizing. Treat physics formulas as tools for decisions and your written procedure as the referee, and your review becomes directly usable under exam-style conditions.
Split UT Physics Into What You Compute and What You Reason
Level II UT knowledge divides cleanly: computations (refraction angles, skip distance, near-zone length, dB ratios) have exact answers, while reasoning tasks (probe selection, interpretation) have justified answers. Drill the first group for speed and the second for defensible language.
Start with Snell's law at the wedge-to-part interface, because it determines which wave modes actually enter the material. For any wedge angle, compute the refracted longitudinal and shear angles using the velocity ratio, and identify the first critical angle, the wedge angle above which the refracted longitudinal wave no longer propagates and only the shear wave remains. Then link each angle to a skip distance on the part. These linked computations decide where a 60-degree shear-wave beam lands and how many skip distances separate the probe from a given weld depth.
The reasoning tasks have no single formula. Choosing a frequency for coarse-grained material means trading attenuation and penetration against resolution; choosing an inspection angle means matching the beam to the expected discontinuity orientation, and that choice is bounded by your written procedure. Rehearse these as spoken justifications: state the constraint, state the trade-off, state the choice. A justification you can say in two sentences while studying is the same one you will need to construct when a scenario asks you to defend a probe selection.
Calibration Setups: What Each Reference Configuration Actually Proves
Each calibration proves a specific property: range calibration proves depth-display accuracy, angle and exit-point checks prove beam geometry, and a DAC curve proves amplitude reference across sound paths. Know what each setup would fail to prove if done wrong.
On an IIW (V1) block, placing backwall multiples at known graduations validates the horizontal range, while the block's radii verify exit point and refracted angle for an angle-beam probe. Practice narrating each adjustment out loud, because scenario questions probe the purpose of each step, not just the sequence. If you can explain why range is established before sensitivity, and what a mislocated exit point does to every subsequent depth reading, you have converted a memorized routine into understood physics you can repair when something looks wrong.
A distance-amplitude curve (DAC) plotted from side-drilled holes at several depths defines reference amplitude as a function of sound path; time-corrected gain (TCG) is the instrument feature that electronically flattens that curve so equal reflectors read equal across the screen. They are not interchangeable terms: DAC is a plotted reference, TCG is compensation applied to the display, and DGS replaces plotted curves with theoretically derived equivalent-size curves. Your written procedure determines which of these is acceptable, so study them as distinct named tools rather than synonyms.
dB Arithmetic: Amplitude Ratios You Must Convert Instantly
Decibel comparisons are UT's core arithmetic: every 6 dB change halves or doubles amplitude, and 20 dB is a tenfold ratio. Convert between dB differences and amplitude ratios in both directions before layering attenuation or transfer corrections.
Worked example: a side-drilled-hole reference is set to 80% full screen height. An indication peaks at 40%, a 6 dB difference, meaning half the reference amplitude. An indication 12 dB above reference would read four times higher, and at an 80% reference it must be re-referenced down to stay on screen. Drill the standard pairs, 6 dB is a factor of 2, 12 dB is 4x, 18 dB is 8x, 20 dB is 10x, until the conversion happens without pencil work and you can reverse it just as fast.
Two traps deserve deliberate drills. First, dB differences add while amplitude ratios multiply: a 6 dB gain increase followed by another 6 dB increase is a fourfold amplitude change, not a doubled one. Second, attenuation corrections scale with sound path, and a deeper reflector has a longer round trip, so its amplitude deficit includes path-length loss, not just reflector size. An indication sitting 6 dB under reference at twice the depth is not automatically equivalent to a smaller reflector at reference depth, and treating it that way corrupts every sizing decision downstream.
Geometry Echo or Discontinuity? A Weld-Root Case to Rehearse
Scenario: a 60-degree shear-wave scan of a butt weld shows an indication near the far-side root that repeats consistently and changes with probe angle. Rehearse the full classification routine before deciding, not after.
The plausible mistake is recording the echo as a flaw on first contact. The better decision is to plot the sound path first: compute the skip distance for the part thickness, confirm the indication appears at the sound-path position corresponding to root or cap geometry, and observe its behavior. Corner-trap geometry echoes typically appear at a stable, geometrically predictable depth, may weaken or vanish when the probe angle changes to 70 degrees, and show little amplitude variation as the probe orbits along the weld.
A fusion-face discontinuity generally behaves differently: its amplitude responds to angle in a way consistent with the expected discontinuity orientation, it may not sit at a clean geometric position, and it often confirms from the opposite side at a consistent depth. Why it matters: classification drives disposition. Recording geometry as a discontinuity triggers unnecessary repair and reinspection; recording a real discontinuity as geometry leaves it uninspected. Rehearse both outcomes from the same starting observation and write a two-line justification for each classification.
Sizing Decisions: 6 dB Drop, DAC Endpoint, and Equivalent Size Compared
Each sizing technique carries its own assumptions. The 6 dB drop locates a reflector's edge as scanned; DAC endpoint sizing reads where an indication crosses the reference curve; DGS reports an equivalent reflector size from dB above a curve.
The most instructive exercise is choosing between techniques for the same echo. A long, roughly uniform discontinuity running along a fusion face suits the 6 dB drop, because the beam passes across its ends and amplitude falls to half as the beam leaves the edge. A small rounded reflector does not: scanning across it produces a bell-shaped amplitude whose 6 dB width reflects the beam, not the reflector. The procedure decides which technique is authorized, and the reflector's shape decides whether the result is physically meaningful.
Rehearse the failure mode of each technique, not just its steps. DAC endpoint sizing assumes the reference curve represents the part, which is exactly the assumption that transfer correction tests. DGS reporting gives an equivalent size under a theoretical model, so a stated equivalent indication is a comparison label rather than a measurement of the actual reflector. As you study, write one sentence per technique naming what would make its result invalid; that sentence is the reasoning scenario practice is designed to test.
| Technique | Best suited to | Key assumption | Watch for |
|---|---|---|---|
| 6 dB drop | Edges of long, roughly uniform reflectors, such as slag lines or planar indications along a fusion face | Amplitude falls to half as the beam steps across the reflector boundary | Small or rounded reflectors, where the drop measures beam width rather than length |
| 20 dB drop | Cases where the written procedure specifies a stricter edge criterion | The reflector's amplitude profile is steep at its boundary | Only valid where the procedure explicitly allows it |
| DAC endpoint | Indications compared directly against a plotted reference curve | The curve was built from reflectors representative of the part | Transfer differences between block and part invalidate the comparison |
| DGS | Reporting an equivalent flat-bottom-hole size for classification | Part geometry and attenuation match the theoretical model | Equivalent size is a classification label, not a physical dimension |
Transfer Correction: When Your Reference Block Doesn't Match the Part
Scenario: sensitivity is set on a calibration block, then the test part shows a weaker backwall. Rehearse measuring the difference and applying transfer correction, because uncorrected comparisons quietly shift every amplitude result.
The plausible mistake is transferring the reference gain straight from block to part and comparing indications anyway. The better decision is a deliberate transfer check: measure backwall amplitude on the block and on the part at comparable sound paths, express the difference in dB, and add that correction to the reference sensitivity. Surface finish and coupling differences are physical facts, not operator errors, and the correction quantifies them. Skip the check, and every 'dB above reference' figure you report carries a hidden offset in an unknown direction.
Why it matters concretely: in a simplified example, a 4 dB uncorrected coupling loss makes a reflector truly 2 dB above reference read 2 dB below it, which can move an indication across a recording threshold. On thick, attenuative material the difference also grows with sound path, so the offset is depth-dependent rather than constant. Rehearse the arithmetic both ways: given a transfer difference, restate the indication's true level; given a required true level, back-calculate the gain setting you must actually use on the part.
Readiness Checks, a Self-Scoring Exercise, and a Study Sequence
Close preparation by testing decisions, not recall: compute, classify, and justify. The exercise and rubric below turn the sections above into a repeatable weekly loop you can adapt to whatever study time remains.
Exercise: draw a plate cross-section to scale, choose a thickness, and compute skip distances for 45-, 60-, and 70-degree shear waves. Mark three echo positions supplied by a practice scenario, a colleague, or a training text, then classify each as geometry or discontinuity and write a two-line justification citing movement, angle response, and opposite-side behavior. Finally, state the transfer correction you would apply if the part were 3 dB weaker on backwall than your calibration block, and restate one indication's true level.
Score the exercise against a rubric: skip-distance values match your computed geometry; every classification cites at least two independent observations rather than amplitude alone; dB conversions are accurate to within about half a dB; and each justification names the procedure constraint being satisfied. Repeat sessions should show the justifications arriving faster while staying specific. As an adaptable sequence: spend early sessions on physics computations, middle sessions on calibration setups and dB drills, later sessions on classification and sizing scenarios, and a final session repeating the full exercise from scratch without notes.
- Compute the first critical angle and skip distance for a given wedge and thickness without notes.
- Convert any pair between dB difference and amplitude ratio within about a minute, in both directions.
- Narrate a full calibration (range, angle, exit point, and DAC) naming what each step proves.
- Classify three practice indications with written, observation-based justifications.
- State in one sentence each how DAC, TCG, and DGS differ and when a procedure would choose each.
References and further reading
Use these references to explore the concepts and check the latest information from the relevant organizations.