How to Audit Forging NDT Records for Compliance

Why Most NDT Audit Failures Start with the Contract

A successful audit begins before you open the test reports. If the contract does not clearly state which defects are unacceptable, the auditor has no baseline. Procurement managers often assume that “meets standards” is enough. It is not. A forged flange and a pressure vessel rotor have different risk profiles. The audit must verify that the vendor tested for the right things.

Start by pulling the purchase specification. Look for the specific NDT method. Does it require ultrasonic testing for internal voids? Or does it require radiography for weld seams? If the spec is silent, the audit fails immediately. You cannot verify compliance against a requirement that does not exist.

The goal is to match the paper trail to the physical part. Every record should be traceable. If you cannot follow the lot number from the furnace to the final report, the part is unqualified.

Prerequisites Before You Begin the Audit

Before you review a single document, gather three items. First, the approved work instructions from the vendor. Second, the as-built drawings for the specific component. Third, the calibration certificates for the equipment used.

You need the operator training logs as well. An untrained operator can produce a valid looking report that is actually meaningless. If the vendor cannot produce the training records for the person who pressed the “go” button, you have a gap.

Keep these documents in a single folder. Do not rely on email threads. The audit needs to be repeatable. If a new buyer takes over in six months, they should be able to run the same check without asking the previous buyer for context.

Step 1: Verify the Test Method Against the Spec

Check the test report header. The method listed must match the contract. If the contract calls for ultrasonic testing for internal porosity, the report must say ultrasonic. It cannot say “visual inspection” as a substitute.

Reason: Different methods detect different flaws. Visual inspection only finds surface defects. Ultrasonic testing finds internal voids. Radiography finds inclusions and porosity in specific planes. Swapping methods silently changes the safety profile of the part.

If the vendor used a less sensitive method, flag it. Do not assume it is equivalent. The auditor must be able to point to the line in the contract that was violated.

Step 2: Confirm Equipment Calibration Dates

Pull the calibration certificate for the probe or the X-ray source. Check the date. It must be current. If the calibration expired two weeks before the test, the data is suspect.

Reason: Sensors drift. A probe that is out of calibration may miss a crack that is just below the detection threshold. The calibration record proves that the tool was accurate at the time of use.

Look for the serial number on the report. It must match the serial number in the calibration certificate. If the numbers do not match, you are looking at a report from a different machine than the one that was calibrated.

Step 3: Check Operator Qualifications and Training

Review the signature on the report. Find the operator’s name. Then find their training records. The training must cover the specific material and thickness being tested.

Reason: Testing a thin sheet and a thick forging block require different techniques. An operator trained on thin sections may not detect a deep flaw in a thick rotor. The training record must match the job.

If the operator is a third party from a subcontractor, verify their credentials too. The primary vendor remains responsible for the final quality of the part. They cannot outsource the liability.

Step 4: Review the Specificity of Defect Reporting

Read the body of the report. Does it say “no defects found”? Or does it say “no defects detected above 2mm depth”? The first answer is vague. The second is measurable.

Reason: Vague language hides gaps. If the report does not state the sensitivity level or the minimum detectable size, you do not know what was actually looked for.

For ultrasonic testing, check for the depth range. For radiography, check the film density or digital resolution. If the report lacks these details, request a re-test. Do not accept a “pass” that was not defined.

Step 5: Cross Reference Lot Numbers and Part IDs

Take the part serial number from the shipping label. Match it to the test report. Then match it to the heat number on the mill certificate.

Reason: Forged parts are often made from multiple heats. If the heat number does not match, the material properties may be different. The NDT data might be valid for one heat but not the one you received.

This is where many audits fail. The vendor ships a mixed lot. The test report covers the batch, but the specific part you bought was not individually tested. If the contract requires piece-by-piece testing, a batch pass is not enough.

Step 6: Validate the Disposition of Rejected Parts

If the report lists a defect, check the disposition. Was the part scrapped? Was it reworked? If it was reworked, was it re-tested after the repair?

Reason: A rework without a follow-up test is a safety risk. The repair might have introduced a new flaw. The auditor must see the “after” picture to confirm the part is safe.

If the part was accepted despite a defect, request the engineering waiver. Check if the waiver was signed by a qualified engineer, not just a salesperson. Without that signature, the acceptance is invalid.

Step 7: Archive the Records for Traceability

Once the audit is complete, file the records. Store the reports, the calibration certs, and the training logs together. Label them with the project number.

Reason: In a failure, the regulator will ask for the records. You want them ready. If they are scattered across three different drives, the investigation stalls.

Set a retention period. For critical safety components, keep the records for the life of the part or longer. Do not delete them after the purchase order is closed. The data has a shelf life, but the liability does not.

Common Mistakes in NDT Record Audits

Most buyers make the same three errors. First, they trust the vendor’s summary. A “passed” stamp is not a record. It is a conclusion. You need the raw data.

Second, they ignore the date. A report from a year ago is not valid for a part manufactured last week. The date must be within the production window.

Third, they skip the calibration check. They assume the equipment is fine because it was used yesterday. Yesterday’s calibration is not today’s assurance.

Another common error is accepting “equivalent” methods. If the spec says ultrasonic, eddy current is not equivalent for internal flaws. Do not let the vendor negotiate the method after the fact.

Final Verification Step

Before you sign off, run a final check. Pick one random part from the batch. Trace it through the entire chain. Mill cert to furnace log to forging log to NDT report to shipping label.

If you can do that in under ten minutes, your audit process works. If you get stuck, your records are fragmented. Fix the system before the next batch arrives.

A forging NDT audit is not about catching a single bad part. It is about proving that the process is under control. When the records are clean, the risk is low. When they are messy, you are buying blind.

Table: Key Data Points to Verify in NDT Records

Data Point	Why It Matters	Typical Error
Test Method	Ensures the right defects are found	Using visual instead of ultrasonic
Calibration Date	Proves equipment accuracy	Expired or missing certificate
Operator ID	Links work to trained personnel	Unsigned or unknown operator
Lot Number	Traces material to specific part	Batch number instead of piece ID
Defect Depth	Defines sensitivity level	Vague “no defects” statement
Disposition	Confirms rejected parts are handled	Missing rework test after repair