Acceptance criteria: ASME Section V and Section VIII

One of the most common mix-ups in the field sounds like this: "We accepted this weld to ASME Section V." Technically, that sentence is wrong. ASME Boiler and Pressure Vessel Code (BPVC) Section V tells you how to perform an examination; it does not tell you whether a discontinuity is acceptable or rejectable. That decision belongs to the referencing (construction) code — usually Section VIII Division 1 or Division 2. If you get the division of labor between these two documents wrong, no examination report will survive an audit. Before anything else, an inspector has to settle which document supplies the method and which one supplies the verdict. This article walks through that distinction, indication classification, RT and UT acceptance logic, and the philosophical gap between construction codes and in-service (API) codes — from a field point of view.

1. What Section V does and does not do

Section V is the technical recipe for nondestructive examination methods. Each method lives under its own Article: Article 2 radiographic examination (RT), Article 4 ultrasonic examination (UT), Article 6 liquid penetrant examination (PT), Article 7 magnetic particle examination (MT), Article 9 visual examination (VT), Article 10 leak testing. Article 1 gives the general requirements and carries one critical statement: acceptance criteria are outside the scope of Section V; they come from the referencing code. So Section V tells you how to select the IQI/penetrameter, what scanning angle to use, how to construct the DAC/TCG curve, how to prepare the surface, how to qualify the procedure — but it will never answer "is this 3 mm slag inclusion acceptable?"

That split is deliberate, not arbitrary. The same RT technique can serve a pressure vessel, a pipeline, or a bridge; each has a different acceptance threshold because the design stress, the service condition, and the safety philosophy differ. Section V keeps the method common and leaves the accept/reject decision to the code.

2. Where acceptance criteria come from: the referencing code

For pressure vessels, the referencing code is usually Section VIII Division 1 (workmanship- and rule-based, very widely used) or Division 2 (stress-analysis based, a lower design margin — therefore a higher allowable design stress and tighter NDE requirements). For process and power piping, ASME B31.3/B31.1 apply (B31.4 for liquid and B31.8 for gas in true transmission pipelines), and for structural steel, AWS D1.1. The referencing code fixes two things: (a) the extent of examination (full versus spot radiography, 10% versus 100% UT), and (b) the acceptance criteria. In Section VIII Div 1, RT acceptance lives in UW-51 (full radiography) and UW-52 (spot radiography), UT acceptance in the relevant Mandatory Appendices, and surface examination acceptance under Appendix 6 (MT) and Appendix 8 (PT). You look at Section V while applying the method, and at Section VIII while making the call. The report must reference both — for example, "RT per ASME Sec V Art. 2, acceptance per ASME Sec VIII Div 1 UW-51."

3. The language of indications: linear, rounded, relevant, non-relevant

You cannot read acceptance criteria without a firm grip on the shared vocabulary.

Rounded indication: an indication whose length is equal to or less than three times its width. Porosity is the classic example. It may be circular or slightly elliptical.
Linear indication: an indication whose length is more than three times its width. Cracks, incomplete fusion (also called lack of fusion, LOF), and elongated slag fall into this class.
Relevant indication: an indication whose size exceeds the minimum threshold for acceptance evaluation. For surface methods, that threshold is typically around 1.6 mm (1/16 in.); most codes do not evaluate anything smaller. The governing value here is the code value in inches — the metric equivalent is approximate.
Non-relevant indication: a trace caused by geometry, surface condition, or magnetic flux leakage, not by a real discontinuity.

This classification matters because codes evaluate linear and rounded indications in completely different ways. In most cases a linear indication is an outright reject, while rounded indications are judged by size, frequency, and alignment rules.

4. RT acceptance logic — Section VIII Div 1

Full radiography acceptance in Section VIII Div 1 (UW-51) works on this logic:

Unconditional rejection: cracks, incomplete fusion, and incomplete joint penetration (IJP) are rejected regardless of size. These are planar, sharp-tipped discontinuities that carry a crack propagation risk.
Slag and elongated (linear) indications: length is judged against a stepped limit tied to thickness. Roughly: up to 19 mm (3/4 in.) thickness the limit is 6 mm (1/4 in.); between 19 mm (3/4 in.) and 57 mm (2-1/4 in.) the limit is t/3; above 57 mm (2-1/4 in.) the limit is 19 mm (3/4 in.). A cumulative length and spacing rule also applies to aligned indications.
Rounded indications (porosity): these are evaluated with thickness-dependent size and distribution charts, formally under Mandatory Appendix 4 (Rounded Indications Charts); UW-51 refers rounded indication acceptance directly to that appendix.

One important warning: the relevance threshold for rounded indications in RT is not the ~1.6 mm (1/16 in.) threshold used in surface methods, and it varies with thickness. Appendix 4 sets that threshold in roughly the 0.5–1.5 mm band, depending on shell thickness. Carrying the surface-method threshold over to RT rounded indications is a common habit — and a wrong one.

For UT acceptance in Div 1 where UT replaces RT, the reference is Mandatory Appendix 12 (together with the UW-53/UW-11 provisions), while the RT rounded indication chart is Mandatory Appendix 4. State clearly in the report which appendix supports which decision; it improves traceability.

Spot radiography (UW-52) is somewhat more lenient: the slag length limit rises to about 2t/3, although at any thickness anything below 6 mm (1/4 in.) is always acceptable and anything above 19 mm (3/4 in.) is always rejected. The reason is that spot RT is a partial-coverage, sampling-based approach, so the design stress is held lower to match (joint efficiency E). The critical point here: acceptance criteria only make sense together with the extent of examination applied and the joint efficiency. You cannot consider one without the other.

5. Surface indications: PT/MT acceptance

Div 1 gives surface examination acceptance under Appendix 6 (MT) and Appendix 8 (PT), and the logic is very clear:

All cracks and linear indications are rejected (relevant indications whose length exceeds three times their width).
Rounded indications carry a typical limit: any relevant rounded indication larger than ~4.8 mm (3/16 in.) is rejected; so are four or more rounded indications in a line separated by less than ~1.6 mm (1/16 in.) edge to edge.
Indications below ~1.6 mm (1/16 in.) are treated as non-relevant.

Here too the code value in inches governs; the metric figures are rounded, and in a borderline dispute you argue from the inch value. Surface methods are more conservative than volumetric ones because surface and near-surface cracks are the most critical initiators under service loading — which is why tolerance for linear indications is essentially zero.

6. UT acceptance logic — amplitude-based versus flaw-based

Ultrasonic examination has two distinct acceptance philosophies, and confusing them causes serious errors:

Amplitude-based: the accept/reject call comes from the echo amplitude of the discontinuity relative to a reference level (DAC/TCG). The general UT method under Section VIII Div 1 Mandatory Appendix 12 uses this logic; reflectors that exceed a given percentage of the reference level, or a given length, are rejected. It is fast to apply, but it measures how much sound came back, not the real size of the flaw.
Flaw-based: you measure the actual height and length of the discontinuity and compare them against a maximum allowable flaw size derived from fracture mechanics/ECA. Note this carefully: the provision that permits UT in lieu of RT (originating in Code Case 2235, today applied through UW-11/UW-53 in Div 1 and through Part 7 in Div 2) uses flaw-based acceptance, not amplitude-based. Encoded PAUT and TOFD fit that philosophy far better, because they give flaw height directly.

The distinction between the Divisions is not absolute; it is more a historical tendency. Div 2, Part 7 offers both an amplitude-based route (7.5.5) and a fracture-mechanics, flaw-based option — so it is not "inherently flaw-based only," but it is more open to flaw-based acceptance. Likewise, current Div 1 permits flaw-based UT (with encoded PAUT/TOFD) as an option under UW-53/Article 4. Because Div 2 allows a higher allowable design stress (a lower design margin), it also tightens examination. That is the code philosophy staying consistent: if you reduce the design margin (thinner shell), you must raise the examination assurance.

7. Workmanship or fitness for service?

The real dividing line here is philosophical. Most construction codes (Div 1, the workmanship acceptance in B31.3) are workmanship- and rule-based: the acceptance limits represent "the quality expected of good fabrication practice," not the true load-carrying limit of the structure. Those limits are conservative on purpose; they usually reject flaws far smaller than what the component could actually tolerate. The aim is a repeatable, defensible quality barrier.

The flaw-based / fitness-for-service approach asks a different question: "Does this flaw threaten the integrity of the structure under this service condition, over the design life?" It answers with fracture mechanics, stress analysis, and material toughness. A flaw can fail the workmanship criterion and still pass a fitness-for-service assessment. That distinction drives both cost and safety, especially in repair decisions.

8. Construction codes versus in-service (API) codes

Construction codes (ASME Sec VIII, B31.3, AWS D1.1) accept newly built equipment on workmanship quality. Once equipment goes into service and runs for years, the rules of the game change: now you are dealing with service-induced damage such as corrosion, erosion, creep, fatigue, and environmental cracking. This is where the API in-service inspection codes take over:

API 510 — in-service inspection, rating, and re-rating of pressure vessels.
API 570 — in-service inspection of process piping.
API 653 — inspection, repair, and reconstruction of aboveground (atmospheric/low pressure) storage tanks; it covers fixed-roof, floating-roof, and open-top types alike.
API 579-1 / ASME FFS-1 — Fitness-For-Service: flaw-based assessment of remaining life and acceptability for corrosion, pitting corrosion, cracks, laminations, and distortions.

The key difference: a construction code says "this should not have been built with that flaw," while an in-service code asks "can it keep operating with that flaw?" An indication that would be rejected in fabrication may well be found serviceable until the next inspection interval under an API 579 assessment. That is why writing clearly in the report which code's acceptance criteria you applied is both a technical and a legal obligation.

9. Related standards

ASME BPVC Section V: examination methods — Article 2 (RT), Article 4 (UT), Article 6 (PT), Article 7 (MT), Article 9 (VT), Article 10 (leak testing). Article 1 states that acceptance criteria come from the referencing code.
ASME BPVC Section VIII Division 1: RT linear acceptance in UW-51/UW-52; RT rounded indication acceptance in Mandatory Appendix 4 (Rounded Indications Charts); surface examination acceptance in Mandatory Appendix 6 (MT) and Appendix 8 (PT); UT acceptance in Mandatory Appendix 12 and the UW-53/UW-11 provisions.
ASME BPVC Section VIII Division 2: Part 7 examination and acceptance (amplitude-based and flaw-based options); a lower design margin than Div 1 (against tensile strength, Div 1 ≈ 3.5; Div 2 Class 1 = 3.0, Class 2 = 2.4), a higher allowable stress, and tighter NDE.
ASME B31.3 / B31.1: process and power piping — extent and acceptance criteria (B31.4/B31.8 for true transmission pipelines).
API 510 / API 570 / API 653: in-service inspection codes.
API 579-1 / ASME FFS-1: Fitness-For-Service assessment.
EN ISO 5817 / EN ISO 10675-1: weld quality levels and RT acceptance levels (the reference equivalents in European practice).

Always state in the procedure and in the report which code you applied and which edition of that code you used. Design margins and UT acceptance options are revised from edition to edition, so an acceptance statement written without the edition is incomplete.

From the field

The trap people fall into most often during audits is squeezing the examination method and the acceptance criteria onto the same line. "Accepted per Section V" is a technically empty statement; Section V does not accept anything. Always write the two separately and in full: method "per ASME Sec V, Art. X," acceptance "per ASME Sec VIII Div 1/2, applicable paragraph." The second common mistake is applying construction acceptance criteria to in-service equipment. Force a fabrication workmanship limit onto a vessel that has run for years and you usually end up with unnecessary repair and downtime — when the right tool is an API 579 fitness-for-service assessment. In short: first which code, then which criterion, and only then which method. Any report that scrambles that order is open to challenge and leaves the inspector exposed.

This article is for educational purposes and does not replace the official standard, the code or the written test procedure — the final decision always rests with the code in force and the responsible inspector.

Take it to the field: To keep these NDT methods, standard references and field steps in your pocket — completely offline and free — take a look at the Doawise NDT Guide app.

At DoaWise we carry out nondestructive testing services to international standards. By keeping construction and in-service code acceptance criteria properly separated, we produce recordable, auditable results — from RT and UT evaluation all the way to fitness-for-service assessment.