Visual Testing (VT) Standards: What ASME Section V Actually Requires

The Method Everyone Assumes They Already Know

Visual testing is the one NDT method every welder, QC inspector, and shop supervisor believes they already understand — look at the weld, check for obvious defects, sign the report. That casual confidence is exactly why VT generates more code compliance findings during audits than almost any other method on the list. ASME Section V, Article 9 (VT) and the referencing requirements scattered through Section VIII, Section I, and B31.3 specify lighting levels, viewing distances and angles, resolution verification, and procedure qualification requirements that most shop-floor "visual inspections" never actually meet. Visual testing done to code is a measured, documented, calibrated inspection method. Visual testing done by habit is a liability sitting in a job file waiting for an auditor or a failure investigation to find it.

This distinction matters because VT is almost always the first line of defense and frequently the last independent check before a weld gets covered, a vessel gets closed up, or a pipeline gets buried. Understanding what ASME Section V actually requires — not what shops assume it requires — is foundational NDT knowledge that gets surprisingly little rigorous attention compared to UT or RT.

Direct vs. Remote Visual Testing: The Code's Actual Distinction

ASME Section V Article 9 splits VT into two fundamentally different categories with different requirements, and conflating them is a common procedure error.

Direct Visual Examination

Direct VT requires the inspector's eye to have unobstructed access to the surface, at an angle not less than 30 degrees to the surface being examined, and within a maximum viewing distance of 24 inches (610 mm) unless the procedure demonstrates equivalent resolution capability at a greater distance using a documented method (mirrors, borescopes with verified resolving power, or other aids qualified under the article). That 30-degree angle rule is not arbitrary — visual acuity for detecting fine surface discontinuities like tight cracks or shallow undercut drops sharply below about 30 degrees off the surface, because the eye is essentially looking along the surface plane rather than across the discontinuity. A common field shortcut — inspecting a weld from a comfortable standing position at a shallow grazing angle because the geometry makes the proper angle awkward to achieve — technically fails the code's direct VT requirement even if the inspector genuinely looked at the whole weld.

Remote Visual Examination

Remote VT uses mirrors, telescopes, borescopes, fiberscopes, or cameras when direct access isn't achievable — inside a vessel through a limited manway, inside piping too small for entry, or on elevated structural welds. Article 9 requires that remote VT systems have a resolving capability at least equivalent to direct visual examination, verified using a resolution test target (commonly a 1951 USAF resolution target or an equivalent qualified gauge) before use. This verification step is frequently skipped in the field — a technician grabs a borescope, runs the inspection, and never actually confirms the scope's resolution meets the code's equivalency requirement for that specific job. When a client or third-party auditor asks for the resolution verification record on a remote VT report and there isn't one, that inspection's validity is legitimately in question, regardless of whether the inspector's findings were accurate.

Lighting: The Requirement Almost Everyone Underestimates

ASME Section V Article 9 specifies a minimum illumination level of 1,000 lux (100 fc) at the examination surface for VT performed to the article's general requirements, though many referencing sections and client specifications call for higher levels — some critical weld inspections specify 2,150 lux (200 fc) or more. This is not a "make sure there's decent light" guideline; it's a measurable, verifiable number, and code-compliant VT procedures require the illumination to actually be measured with a calibrated light meter and documented in the inspection record, not estimated by eye.

The practical gap here is significant. A weld inspected under typical shop fluorescent lighting often falls well short of 1,000 lux, especially at a real working distance and angle rather than directly under a fixture. A weld inspected in the field — inside a vessel, under a pipe rack, during a night shift turnaround — is even more likely to be under-lit unless the crew specifically brings supplemental lighting and verifies it. Shops that run VT to a documented lighting standard, with a light meter reading logged on the report, are a distinct minority, and that's precisely the gap that shows up when a missed weld defect surfaces later as a leak or a failure and the investigation pulls the original inspection record.

Surface Condition Requirements

Article 9 also requires the examination surface be free of conditions — scale, weld spatter, grinding dust, oil film, paint overspray — that could mask a discontinuity or produce a false indication, and this requirement is procedurally tied to the acceptance criteria being applied, most commonly AWS D1.1 Table 6.1 for structural steel or the applicable ASME Section VIII/Section I acceptance criteria for pressure-retaining welds. A weld that's been wire-brushed but not fully cleaned of slag and spatter can hide undercut or crater cracking under the debris, and a VT report signed off on that surface condition is not actually verifying what the code requires it to verify.

What Skilled VT Actually Finds

Done properly, VT is remarkably effective at catching a specific category of defects before they ever need a volumetric or surface NDT method to confirm them:

• Undercut, overlap, and excessive reinforcement on weld caps
• Surface porosity and crater cracks visible without magnification
• Misalignment (hi-lo) at pipe and vessel joints exceeding code tolerance
• Arc strikes outside the weld joint — a defect category that's purely visual and that PT/MT often miss if not specifically looking for them
• Weld profile discontinuities — excessive convexity, concavity, or abrupt transitions that create stress risers
• Base metal surface defects — laminations, gouges, mechanical damage — before welding even begins
• Dimensional and fit-up issues caught during in-process VT, before a weld is even made, which is where VT delivers its highest value per hour spent

That last point deserves emphasis: the highest-value VT in a fabrication sequence often happens before welding — checking joint fit-up, root gap, bevel angle, and cleanliness — because a fit-up defect caught before the arc starts costs nothing to fix and a fit-up defect caught after the weld is complete can mean a full repair cycle. Shops that only think of VT as a post-weld inspection step are missing where the method earns its keep the most.

VT Personnel Qualification

ASNT SNT-TC-1A and the referencing code sections require documented training, experience, and a near-vision acuity test — typically Jaeger J1 or J2 equivalent at the specified reading distance, verified annually — for anyone performing or signing off VT. This near-vision test is a genuinely important and genuinely under-enforced requirement; a technician whose corrected vision has drifted since their last annual check, running VT on a critical weld, is a real and preventable gap in an inspection program. Far vision (20/20 or corrected) is also typically required and tested. Building and tracking this certification and vision-test cadence for a technician roster is exactly the kind of administrative load that Atlantis NDT ERP is built to manage — flagging when a technician's annual vision recheck or SNT-TC-1A recertification is coming due before it lapses mid-project, rather than discovering the gap during a client audit.

VT Under In-Service Inspection Codes

New construction isn't the only place ASME Section V Article 9 requirements apply. In-service inspection codes governing existing equipment — API 510 for pressure vessels, API 570 for piping, and API 653 for aboveground storage tanks — all specify their own VT scope, frequency, and personnel qualification requirements layered on top of the base ASME Section V approach, and these are worth understanding even though Atlantis NDT does not train or certify inspectors under these API programs. API 653, for instance, calls for external VT of tank shells, roofs, and foundations on a defined interval driven by the tank's corrosion rate and prior inspection history, plus internal VT (supplemented by UT floor scanning) at intervals set by the same corrosion-rate calculation — and the external VT specifically requires attention to settlement patterns, shell distortion, and coating condition that a generic "look at the tank" inspection would miss entirely.

API 570 similarly specifies VT scope for piping circuits based on risk classification, with particular attention to VT at supports, insulation penetrations, and any location where external corrosion under insulation (CUI) is credible — a mechanism that's purely a visual finding at the insulation jacket level (staining, jacket damage, low points where water collects) long before it becomes a UT thickness finding underneath. A technician who understands why the VT is being performed at that specific location — not just that a checklist says "inspect insulation jacket" — catches the early-stage indicator instead of waiting for it to become a wall-loss problem discovered by chance during a later strip-and-inspect. Recognizing where a code-driven VT scope is pointing the inspector, and why, is exactly the kind of code literacy that separates competent field VT from checklist VT, regardless of which specific in-service code governs the asset.

Documenting VT to Survive an Audit

A defensible VT report needs more than "visual — satisfactory." At minimum, a code-compliant record documents: the specific procedure and revision used, illumination level measured (with the meter reading, not an estimate), viewing angle and distance (or remote equipment resolution verification), the acceptance criteria applied, a description or sketch of any recordable indication, and the technician's current certification level and vision test date. Reports that skip straight to a pass/fail checkbox without these supporting details look identical for a clean weld and a weld that was actually never properly lit or angled during inspection — there's no way to distinguish a rigorous inspection from a rushed one after the fact.

This is where structured field data capture changes outcomes, not just paperwork quality. Atlantis NDT's reporting software prompts for illumination readings, angle/distance confirmation, and acceptance criteria selection as required fields rather than optional notes, and timestamps photo documentation directly against the weld or joint ID — which is exactly the traceability an ASME Section V audit, or a post-failure investigation years later, is going to ask for. For shops running distributed field crews across multiple job sites, having that data captured consistently in the field rather than reconstructed from memory back at the office is a real difference in defensibility.

Aids and Enhancement Tools Within the Code's Boundaries

Article 9 permits certain visual aids without reclassifying the examination as "remote" — magnifying lenses, for instance, are commonly used to resolve a questionable indication once it's already been located by unaided or minimally aided direct viewing, and this is a meaningful distinction: the code's direct-VT distance and angle rules govern how the surface is scanned to find an indication in the first place, while magnification is typically an evaluation aid applied after something catches the inspector's eye, not a substitute for proper scanning technique across the full weld length. A technician who tries to scan an entire 20-foot seam through a magnifying lens at close range, rather than performing a proper full-length scan at the correct angle and distance first, is misapplying the aid and likely missing indications outside the narrow field of view magnification creates.

Weld gauges deserve equal procedural weight, since much of what VT reports as a rejectable profile defect — undercut depth, reinforcement height, fillet weld leg length, misalignment — is a dimensional measurement, not a subjective call. A technician who eyeballs undercut as "looks shallow, probably fine" instead of actually running a gauge against the AWS D1.1 Table 6.1 threshold (typically 1/32 inch for cyclically loaded members, more permissive for statically loaded ones, though the exact figure depends on plate thickness and loading category) is substituting judgment for a measurement the code specifically intends to be objective. This is a small discipline gap that shows up constantly in the field — not because inspectors don't own gauges, but because reaching for one on every borderline call slows down a production-paced shop floor, and that time pressure is exactly where a rejectable defect gets waved through as "close enough."

Where VT Fits in a Multi-Method Program

VT is never meant to stand alone on critical welds — it's the first gate, not the only gate. A typical pressure vessel weld sequence runs VT during fit-up, VT after root pass, VT after each fill pass on multi-pass critical welds, VT on the completed cap, and then the volumetric or surface method specified by code (RT, UT, PT, or MT) layered on top. Skipping intermediate VT passes to save time is a common shortcut on tight schedules, and it's also how root-pass defects end up buried under six more weld passes before anyone with a trained eye looks at them again. A weld sequence with disciplined in-process VT at every stage catches problems when they're still a five-minute grinder fix, not a full cutout-and-reweld after final NDT flags a rejectable indication.

For QC managers and procedure writers, the practical takeaway is to stop treating VT as the "free" inspection that happens by default and start treating it with the same procedural rigor as UT or RT: a written procedure, verified equipment (light meter, resolution target for remote systems), qualified and vision-tested personnel, and documentation that would hold up if pulled six months later during a client audit or a failure investigation. Getting a shop's VT program from "we obviously look at the welds" to genuinely code-compliant is one of the highest-return, lowest-cost improvements an inspection program can make — and it's a common starting point in ASNT Level III consulting engagements precisely because the gap between assumed and actual compliance is so often larger than shop management expects.

Atlantis NDT Products & Services

Atlantis NDT pairs field expertise with software: NDT inspection management software — Atlantis ERP, a digital twin platform for asset integrity, and NDT reporting software. Build your team with NDT training & certification (ASNT SNT-TC-1A) and ASNT certification pathways, or bring in ASNT Level III consulting. Affordable, accessible, fully customizable — book a free consultation.