Level III Audit Support for Aerospace NDT

What the auditor actually samples

The mental model that gets suppliers into trouble is that an audit is a document review. It is not. A Nadcap NDT audit is built around job audits: the auditor picks parts that are moving through the facility against real customer orders and reconstructs everything that had to be true for that inspection to be valid. Planning, drawing revision, technique sheet, operator certification, the process control records for that specific shift, material lot status, equipment calibration, lighting readings, accept and reject decisions, and the paperwork that left with the part.

That structure is why preparation aimed at tidying the document library so often fails. A binder can be perfect while the shop floor tells a different story: a technician working a method they are qualified for in general but not for the specific technique on that part, a UV-A reading taken weekly on a line that requires it before each shift, a penetrant tank topped up from a lot that changed three weeks ago with no record of the change.

The productive preparation is therefore an internal job audit run exactly the way the real one will be run, on live work, with the same backwards traversal. That is the core of what ASNT Level III consulting contributes before an audit: not a checklist walk, but an adversarial reconstruction of a handful of real jobs to find where the chain actually breaks.

The personnel evidence chain that has to hold

NAS 410 is the aerospace personnel qualification standard in North America, maintained in technical alignment with EN 4179 in Europe. Both put the obligation on the employer rather than on a central certifying body: the employer writes a practice, names a responsible Level 3, and certifies its own people against that practice. That structure gives flexibility and it gives an auditor exactly one thing to test, which is whether the practice was followed in every particular.

The chain runs like this. The written practice exists and matches the current standard revision. The responsible Level 3 is named and is qualified in the method. The candidate's training is documented with content and hours. Experience is documented with a source that can be checked rather than asserted. General, specific and practical examinations were administered and graded, with the material retained. Vision was examined within the required period. And the certificate was issued and signed by the person the practice authorizes. Any weak link makes every downstream inspection questionable.

Experience is where most files are thin. Hours claimed from a previous employer need documented verification, not a statement on the application form, and hours accrued in-house need a source that survives an auditor asking to see it - job records, timesheets, a supervised experience log. Where organizations are building this base for the first time, structured NDT training and certification support delivered against the written practice, rather than as generic method courses, produces a file that holds up because the training content and the practice were written to match.

Process control: the records that fail most often

The fluorescent penetrant line is the single richest source of findings in aerospace NDT, because it has more daily control parameters than any other method. ASTM E1417 and the customer specification between them require verification of UV-A irradiance, ambient white light in the inspection booth, system performance against a reference panel, penetrant and emulsifier concentration, water pressure and temperature, and dwell times. Common thresholds include a minimum of 1,000 microwatts per square centimeter of UV-A at the inspection surface, a maximum of 2 foot-candles of ambient white light, and a minimum penetrant dwell of 10 minutes unless the specification says otherwise.

The finding is almost never that a value was out of range. It is that a required check has no record on a day parts were processed, or that the recorded value sits impossibly constant across sixty days, or that the meter used to take the reading was itself out of calibration. Auditors read control charts the way an integrity engineer reads thickness data: the pattern of the record tells them whether the measurement was really taken.

Magnetic particle carries the same logic with different parameters - field direction and adequacy verified with artificial flaw shims or a field indicator, ammeter and timer calibration, bath concentration and contamination checks, light readings again. Radiography adds film or detector performance, density and image quality indicator sensitivity. In each case the discipline that survives an audit is a record created at the time of the check by the person who performed it, which is why organizations with several lines increasingly hold these logs in an inspection data management system rather than on clipboards beside the tank.

Procedures, technique sheets and revision control

Aerospace NDT runs on part-specific technique sheets, and technique sheets are documents with parents. Each one descends from a customer specification, a drawing at a stated revision, and an internal procedure at a stated revision. When any parent moves, the technique sheet has to be reviewed, and the audit question is not whether you have a change process but whether it was applied to this part on this order.

The recurring failure is quiet and systemic: an engineering change updates the drawing, planning picks up the new revision, and the technique sheet continues to reference the old one because nobody linked them. The inspection may be entirely adequate in physical terms. The evidence is not, because the document that authorized it points at something that no longer exists. Auditors find this in minutes by comparing the revision on the router with the revision on the technique sheet.

The second failure is the procedure that is compliant with the wrong document. An in-house procedure written to a general industry standard is not automatically acceptable when the purchase order names a customer specification with additional requirements. Every procedure should state, in its own text, which customer specifications it satisfies and at which revision, and that mapping should be maintained as a controlled list rather than reconstructed from memory when the auditor asks.

What the Level III signs, and what sits outside that authority

The boundaries of the role are worth stating plainly, because both over-claiming and under-using a Level 3 cause problems. Within the employer's written practice the Level 3 is responsible for technical content: approving the written practice itself, writing and approving NDT procedures and technique sheets, establishing and grading qualification examinations, recommending or approving personnel certification, providing technical interpretation on difficult indications, and reviewing examination data independently of the person who produced it.

Outside that boundary are decisions that belong to other authorities. The Level 3 does not release the part as airworthy or conforming; that is a quality and, where applicable, a delegated airworthiness function. The Level 3 does not disposition a nonconformance in the sense of accepting a deviation from design intent; that is engineering and material review authority, with customer concurrence where required. And a consultant Level 3 does not become the certifying body: certification remains the employer's act, issued under the employer's practice.

The distinction matters during the audit itself. A Level 3 representing the supplier on technical questions is answering for procedures, qualification, technique adequacy and data interpretation. When an auditor's question moves to product acceptance, contract flowdown or corrective action commitments, the correct answer comes from the organization, and a technical representative who blurs that line creates commitments the business has not agreed to. Independent review of inspection reports and data before an audit gives that representative something solid to stand on.

Gap review before the audit versus remediation after a finding

These are different exercises with different economics. A gap review is voluntary and therefore unbounded. It samples live jobs and the trailing twelve months of records, tests them against the current checklist and the customer specifications actually flowed down, and produces internal corrective actions that never become external findings. Because it is voluntary, its scope can follow the evidence: if the first three jobs show a technique sheet revision problem, the review can pivot and sample twenty more.

Remediation after a finding is bounded on every side. The scope is the wording of the nonconformance. The clock is short - Nadcap responses are due within 21 calendar days of the audit report being issued, submitted through eAuditNet - and a weak response consumes the remaining time in a rebuttal cycle. The standard of proof is higher, too: root cause must be a mechanism, not a person, and corrective action must include objective evidence of implementation and a means of verifying effectiveness.

The hardest part is rarely the corrective action. It is the containment question that follows a personnel or process control finding: if the process was out of control, or the examiner was not validly certified, what parts were inspected during that window, where are they now, and what does the customer need to be told? Suppliers who have thought about that question in advance answer it in days. Those who have not spend weeks reconstructing traceability while the response clock runs.

Findings that recur across aerospace NDT audits

A short list accounts for a large share of nonconformances. Process control checks missing on days production ran, usually around holidays, shift changes or equipment moves. Vision examinations that lapsed while the technician kept working. Technique sheets pointing at superseded drawing or specification revisions. Experience hours in a certification file with no auditable source. Approval signatures from a Level 3 dated before that person held qualification in the method. Materials in use after their lot or qualification status changed. Subcontracted NDT placed with an agency not approved for the method performed.

What these have in common is that none of them is a technical failure of inspection. They are failures of the system that makes an inspection provable. That is also why they are so persistent: the shop knows the parts were inspected properly, so the finding feels like paperwork, and the corrective action gets written as a reminder rather than as a control. The same finding then reappears at the next audit, which changes its character entirely, because a repeat finding raises questions about the corrective action system itself.

The durable fixes are structural. Interlock the process control record with the ability to run the line. Drive vision and certification expiry from the same system that schedules work, so an expired examiner cannot be assigned. Link technique sheets to drawing revisions so a revision change raises a task. None of these require a large investment; they require someone to treat the audit findings as a design specification for the quality system.

Flowdown, outside agencies and the prime's own approval

Almost every aerospace NDT operation subcontracts something: a method it does not hold, an overflow of radiography, a specialist technique. The moment work leaves the building, two obligations attach. The customer's requirements must flow down to the subcontractor in full, including the specification revision, and the subcontractor must be approved for that method by whoever the contract says approves it - which is frequently the prime, not the supplier.

The audit tests both directions. It looks at the approved supplier list and asks how the agency got on it and what evidence supports the approval. It looks at the purchase orders placed on that agency and asks whether the customer specification and revision were named. And it looks at receiving inspection and asks what was verified when the reports came back, because accepting a subcontractor's report without review is itself a control failure.

Prime approval sits alongside accreditation rather than beneath it. A supplier can be Nadcap accredited for a method and still not be approved by a particular prime for a particular part family. Both statuses have scope and both expire. Keeping a live matrix of method, customer, specification revision, approval status and expiry is unglamorous work, and it is the difference between a smooth audit and a scramble.

How Atlantis engages on aerospace audit support

The scope is technical authority and evidence, delivered independently of whoever performs the inspections. That includes writing or reviewing the written practice against the current standard revision, developing and qualifying NDT procedures and technique sheets, building qualification examinations and supporting personnel certification within the employer's practice, running pre-audit job audits and gap reviews, representing the supplier on technical questions during the audit, and providing independent review of examination data and reports.

It does not include acting as a certifying body, dispositioning nonconformances on the customer's behalf, or making product acceptance decisions - those stay where the contract puts them. Engagements are usually structured either as a defined pre-audit review with a written gap report and a prioritized action list, or as an ongoing technical authority arrangement across a certification cycle. Where a team is deployed to support a campaign, the same personnel stay with it, so knowledge of the written practice and the shop's real habits does not have to be rebuilt each visit.

The most useful first conversation is usually about timing. A gap review nine months before a scheduled audit can change the outcome; the same review three weeks before it can only reduce the surprise. A consultation to scope the review against your current accreditation dates and customer approvals is the normal starting point.

Request a consultation