CTQ features on nuclear-sector machined parts need an inspection strategy tied to function, drawing definition, part state, measurement capability, and release authority. No single instrument proves every characteristic. Build a feature-level plan that names the datum simulation, method, environment, restraint condition, frequency, record, and decision rule before machining begins.
The buyer or authorized design organization identifies what is critical and why. The supplier can recommend a practical method, but cannot infer safety significance from a tight tolerance alone. A modest dimension can be critical to assembly identity, while a close local size may have no nuclear classification.
Feature geometry determines what can be observed. A bore diameter can be sampled with an internal gauge, yet diameter alone does not prove straightness, cylindricity, axis location, or relationship to a sealing face. A CMM may evaluate relationships, but probe access, stylus qualification, point strategy, fixture force, and software settings influence the result.
Surface texture has its own measurement definition. Ra cannot stand in for waviness, lay, local damage, or sealing performance. The drawing needs the applicable texture standard, parameter, filter or cutoff, evaluation length, direction, location, and acceptance rule. The instrument then needs adequate range and access for that setup.
CTQ Characteristic | Method Question | State to Define | Failure if Undefined |
|---|---|---|---|
Bore size and form | Does the method resolve local size, straightness, and cylindricity where needed? | Temperature, cleanliness, final process, and free or restrained condition | Passing diameter masks taper, lobing, or fixture distortion |
Datum-related hole pattern | How are datum features contacted and simulated? | Final surface condition and assembly-relevant support | Reported position changes with an undocumented alignment |
Sealing or bearing surface | Which geometry and texture parameters represent function? | Post-grind or post-finish state, measurement direction, and filter | A single roughness value releases a functionally unsuitable surface |
Thin wall or flange | Can contact force or fixturing change the feature? | Free state, support points, stabilization time, and orientation | Inspection fixture forces the part into compliance |
For complex relationships, a precision machining review should connect datums, setup strategy, finish operations, and inspection access. Measurement planning after the route is frozen can reveal that a datum is coated, inaccessible, or altered by the final process.
Consider a hypothetical thin-wall flanged sleeve. A three-jaw fixture can make the bore appear round while clamped. After release, residual stress and wall flexibility may change the form. If the drawing requires free-state acceptance, the final record must reflect that state. If restraint is functionally intended, the drawing needs to define it.
Temperature affects both part and instrument. The significance depends on material, dimension, tolerance, thermal history, and the difference between calibration and inspection conditions. “Room temperature” is not a complete record when the acceptance margin is small. Record the agreed stabilization and environmental controls when they affect the decision.
Measurement uncertainty belongs in acceptance planning. A result close to a specification limit cannot be treated as unquestionably conforming without the contract’s decision rule. The drawing system, quality plan, or purchase document needs to define how uncertainty and guard bands are handled. Do not blend ASME and ISO defaults without agreement.
Calibration status is necessary, but it does not prove method suitability. A calibrated micrometer used across a flexible wall can still deliver a misleading result. The quality-assurance review needs evidence of equipment status, method selection, operator competence, and traceable raw results.
A feature control plan translates drawing requirements into executable checks. It includes characteristic ID, requirement source, manufacturing stage, method, equipment, datum setup, sample or coverage rule, environment, recording format, reaction plan, and release authority. It also distinguishes process monitoring from final acceptance.
Inspect characteristics at the stage where they exist and again after later work if that work can alter them.
Use independent verification where the purchase documents require separation of roles.
Preserve actual readings when variable data are required, rather than recording only pass or fail.
Define the response to an out-of-tolerance result before remeasurement, adjustment, or rework.
Link every report to the item, operation, drawing revision, method revision, and inspector.
Repeated measurements need control as well. Rechecking until one value passes can hide method variation. A documented reaction plan addresses suspected equipment error, part contamination, temperature, setup, and genuine nonconformance without discarding unfavorable data.
Before release, review the drawing revision, CTQ list, completed operation status, equipment validity, actual results, exceptions, and approved dispositions. The record should make clear whether inspection occurred before or after heat treatment, coating, grinding, cleaning, or any other condition-changing operation.
For an RFQ, provide the controlled 3D model and 2D drawing, governing GD&T system, CTQ rationale where available, required part state, sampling or coverage rule, data format, uncertainty decision rule, and witness points. That gives the machining team enough information to propose a route that can actually be inspected and released.