9.2 Evaluating Indication Size, Shape, and Persistence

Once a liquid-penetrant pattern is relevant, official PT topic 3 moves from class to evaluation. The written exam asks whether you can describe shape, watch bleed-out through the allowed time, tell an indication from a discontinuity, use persistence after a wipe, and refuse to invent a length from fluorescent bloom. None of that work replaces the acceptance table — it feeds the table.

ASTM E1417 and ASME Section V, Article 6 control how long developer may sit before you interpret. The construction code or specification named in the procedure — ASME Section VIII, AWS D1.1, a customer aerospace table — decides whether the number you recorded is acceptable. A Level II who quotes "it looks like a quarter-inch crack" without saying how length was taken, and against which sentence, has not evaluated the indication.

Linear versus rounded

After the class is relevant, describe the indication in the words the acceptance standard uses.

The 3:1 length-to-width idea is a common code convention, not an ASNT-published secret. Some specifications say "crack-like" or "elongated" instead of "linear." Use the definition in the governing standard. If a stem does not give a definition, treat a sharp, crack-like bleed as linear and a blob as rounded.

Linear is not automatically rejectable. Rounded is not automatically acceptable. A 25 mm linear lap on a forging lug may fail the table; a 1.5 mm isolated rounded pore may pass. The table decides. Do not promote a rounded indication to linear because you "know porosity is never that sharp," and do not demote a linear crack to rounded because the first developer blush was a dot.

Record length along the indication, width if the procedure asks for it, location (weld metal, toe, HAZ, base metal, distance from a datum), and orientation (parallel to the weld, transverse, circumferential, radial). Some procedures also want a sketch or a photograph. Measure with the scale, comparator, or optical aid the procedure allows.

Bleed-out grows with time

Developer does not freeze the picture at the moment it dries. Capillary liquid keeps leaving the opening. That continuing exudation is bleed-out. As more penetrant reaches the developer, the stain usually grows — wider, longer, or both.

Typical industry practice, and the usual ASTM E1417 / ASME V Article 6 pattern, is to begin interpretation after a minimum developer dwell (often on the order of 10 minutes) and to finish before a maximum the procedure states (often 60 minutes, or the document's own window). Those minutes are industry practice numbers the procedure may tighten. They are not ASNT exam secrets, and they do not outrank a printed procedure paragraph.

What growth means in evaluation:

• A tight crack may be invisible or only a pinpoint at 2 minutes, a small rounded spot at 10 minutes, and a sharp linear indication at 30 minutes. If you evaluate too early you under-call shape and length. If the procedure's window is 10–60 minutes, the 30-minute linear call is the one that counts, not the first glow.
• Open porosity usually stays rounded as it grows. The stain may get larger in diameter, but it does not turn into a crack-like line unless the opening itself is elongated or several pores are aligned.
• Excessive dwell lets adjacent indications bleed together, so two acceptable rounded pores can look like one rejectable linear indication. That is why there is a maximum, not only a minimum.
• Fluorescent Type I indications can look larger sooner than visible Type II stains because the eye (and the bloom) see a small amount of dye very well. Growth still happens in both families.

Watch the same site through the window when the class or shape is in doubt. Evaluation is not a single snapshot at "whenever I walked by the booth."

Measuring the indication versus measuring the discontinuity

These are different lengths. Mixing them is a standard specific-exam trap.

• The indication is the developer stain — the bleed-out you see under the specified lighting at the specified time.
• The discontinuity is the actual opening in the metal — the crack mouth, the pore orifice, the lap interface.
• The defect decision compares whatever the procedure says to measure with the acceptance limit.

Bleed-out is almost always larger than the opening that fed it. A hairline crack can paint a 4 mm stain. If the procedure says "measure the indication," you record 4 mm. If it says to wipe the developer and measure the visible opening with a magnifier, you record the opening, which may be much shorter. If it says to measure at 10 minutes and again at 30 minutes and use the larger indication length, you do that. You do not split the difference because you "know the crack is smaller than the red."

Worked logic the specific exam likes (the stem will print the limit):

• Relevant, linear, indication length 5 mm, linear limit 3 mm → reject (defect).
• Relevant, rounded, indication diameter 2 mm, isolated rounded limit 3 mm, spacing rules met → accept for that indication.
• Relevant, linear, indication length 8 mm, but the procedure says measure the opening after wipe and the opening is 2 mm against a 3 mm opening limit → apply the 2 mm number, not the 8 mm stain.
• Two rounded indications whose bleed-out merged after 90 minutes, outside the procedure's 60-minute maximum → the merged length is not a valid evaluation. Redevelop and read inside the window.

If the governing specification is silent on how to measure, the Level II does not invent a shop custom. The report states how the length was taken and that the document did not define the method — and the Level III or the customer resolves it. On the written exam, use the method printed in the stem.

Persistence after wipe

Persistence is the evaluation partner of the confirmation ladder in 9.1. After a gentle wipe and redevelop (when the procedure allows that check):

• A relevant opening returns in the same place, often with the same orientation. A crack returns as a line (or as a spot that grows back into a line). An open pore returns as a rounded bleed at the same mouth.
• A false lint line or developer clump does not rebuild from a metal path. It is gone, or it reappears only if you reintroduce the fiber or the clump.
• Incomplete-removal background may return as a blush or streak that does not lock to a single opening.
• A nonrelevant geometric collector persists in the geometry — the thread root bleeds again — which confirms nonrelevant, not relevant.

Persistence answers "is there a reservoir of penetrant in an opening?" It does not by itself decide accept or reject. A persistent 1 mm rounded pore can still be acceptable. A persistent 10 mm linear crack is still measured against the table.

Do not confuse persistence with over-wipe. If you solvent-flood the site, you can empty a tight crack and then claim it "did not come back." That is a process error, not proof of a false indication. The wipe is light, the redevelop is controlled, and a full reclean-and-reprocess is the tie-breaker.

When a rounded indication is porosity — and when it is a tight crack

A small rounded fluorescent or red spot is the most misread relevant indication in the booth.

Surface-open porosity usually:

• Appears rounded and stays rounded as bleed-out continues
• Sits in weld metal or a casting skin, not preferentially at a toe or a fillet radius
• Shows a pore mouth under the optical look
• Returns as a rounded spot after wipe-and-redevelop
• May have neighbors that are also rounded (a porosity field)

A tight crack that only shows a dot usually:

• Appears first as a pinpoint because little penetrant has escaped yet
• Grows in length during the developer window and becomes linear or crack-like
• Prefers a stress raiser — toe, radius, thread root, pit, grind mark
• Looks like a fine opening, not a hole, under magnification
• Returns after wipe and then lengthens again

The exam stem that shows "a rounded indication at a fillet radius that became 8 mm long at 25 minutes" is a linear crack evaluated at 25 minutes, not "porosity that got bigger." The stem that shows "an isolated 2 mm rounded spot in the center of a weld face that stayed round from 10 to 40 minutes, with a visible pore mouth" is rounded porosity.

Do not wait past the procedure's maximum hoping a pore will "declare itself." Do not freeze the first 3-minute snapshot and call every later line a false bleed. Shape is taken inside the specified viewing time.

Do not size from fluorescent bloom alone

Bloom (or halo) is the glowing cloud around a Type I indication. High UV-A, a sensitive penetrant, a thick developer layer, and a long dwell all make the halo larger than the stain you would record under a controlled viewing condition. Bloom is useful for finding small indications. It is a poor ruler.

Rules the written exam expects:

• Do not put a scale on the outermost fluorescent haze and call that the official length unless the procedure's measurement method says length is taken that way.
• Use the lighting the procedure states — commonly UV-A intensity at the surface in the ASTM E1417 / ASME V Article 6 band (industry practice often ≥ 1000 μW/cm² with ambient visible light held down, frequently ≤ 20 lux / 2 fc) — and view at the working distance those checks used.
• If the procedure wants the indication measured as the distinct stain, not the haze, record the stain. If it wants a wipe-back and an optical measurement of the opening, do that instead.
• Visible Type II red or color-contrast indications can also spread in developer; the same rule applies: measure by the documented method, not by the palest pink halo.

A candidate who answers "always use the bloom; bigger glow means a worse crack" has invented a severity scale the standards do not define. Brightness is not a code dimension.

Realistic exam scenarios

A relevant indication is 9 mm long and 2 mm wide. The standard's printed definition is linear when length is three times width or more. Description: linear. Compare it with the linear table, not the rounded table.

At 8 minutes a fillet-radius site is a faint rounded spec. At 20 minutes, still inside a 10–60 minute window, it is a sharp 6 mm line. Evaluate the 6 mm linear indication. Calling it porosity from the first glance is wrong.

A Type I indication has a 12 mm fluorescent halo and a 4 mm distinct stain. The procedure says measure the indication, not the bloom. Record 4 mm, then apply the table. If the procedure had said to measure the opening after wipe and the opening is 1 mm, record 1 mm.

A rounded spot wipes away and never returns after a controlled redevelop and a full reprocess. That is not a persistent relevant pore. Reclassify toward false (clump, lint, leftover dye) and stop treating the first stain as a diameter for the porosity table.

Topic 3 evaluation language is short: shape, time, what you measured, persistence, then the table. Anything else is storytelling.