CNC-machined parts require internal defect inspection when hidden cracks, pores, inclusions, lack of fusion or heat-treatment discontinuities could affect strength, sealing, fatigue life or safety after machining. Visual inspection and dimensional inspection can confirm shape, surface and tolerance, but they cannot prove that the internal material is free from rejectable indications. CNC-machined parts made from forged, cast, welded, additively manufactured or heat-treated stock may carry internal risks that remain invisible after milling, turning or grinding. This is most important for load paths, pressure boundaries, rotating parts, sealing lands, critical holes and features exposed to cyclic stress. The need also increases when a part cannot be destructively sectioned after machining. Buyers should decide which internal defects matter for the application, which NDT method can detect them, and what acceptance criteria should appear in the RFQ.
Ultrasonic testing is often selected because sound waves can travel through the part and return echoes from internal boundaries. The method is useful for thick metallic sections, selected forgings, billet-derived components, welded or additive features, and parts where radiography, magnetic particle inspection or dye penetrant inspection cannot answer the main risk. UT is not a universal health check. It has limits near surfaces, around complex geometry, in coarse-grained materials, and when the defect is poorly oriented to the sound beam. A reliable UT plan therefore starts with material, stock form, defect type, section thickness, probe access, surface condition, applicable standard and buyer reporting requirements. The buyer should also state whether the defect risk comes from raw material, welding, additive build, heat treatment, rough machining or service damage. Each source changes the expected location and orientation of the indication.
Ultrasonic testing sends high-frequency sound into a component through a probe and couplant. When the sound reaches a boundary, a pore, a crack face, an inclusion or the back wall, part of the energy reflects to the probe. The instrument displays the return signal as amplitude, depth and time-of-flight information. The inspector compares those signals with reference sensitivity, scan coverage and the acceptance rule. A small echo is not automatically a rejectable defect, and a quiet screen is not proof that every smaller indication is absent. The inspection result depends on calibration, material noise, beam angle, defect orientation, surface condition and reporting threshold. Reference blocks or known reflectors are needed because the screen response must be tied to a repeatable sensitivity, not only to operator judgment.
For buyer decisions, UT should be described as a controlled measurement method, not as a black-box quality promise. The report should state probe frequency, inspection mode, reference block, scan path, material thickness, rejected zones, detectable volume and whether the result is pulse-echo, angle-beam, phased-array or immersion C-scan. UT is most useful when the inspection stage is tied to raw material verification, intermediate machining, heat treatment or final acceptance. The RFQ should also ask whether the UT result is for process screening, final release, or customer compliance documentation. This distinction matters because a screening scan may guide process development, while a release scan must satisfy the purchase order and acceptance standard. It also helps purchasing compare quotes, because a supplier offering only a simple shop-floor scan is not quoting the same risk control as a supplier offering documented final acceptance.
UT can be stronger than X-ray or magnetic particle inspection when the main risk is a planar crack, a thick-section discontinuity or a defect that lies below the surface. Radiography can show volumetric density differences well, but it may miss tight cracks in an unfavorable orientation. Magnetic particle inspection is useful for surface and near-surface defects in ferromagnetic materials, but it cannot inspect nonmagnetic alloys or deep internal zones. Dye penetrant is limited to surface-breaking defects. UT can inspect deeper material volumes, estimate reflector depth and guide further investigation when sectioning is not acceptable. The method is also practical for in-process decisions because a scan can be repeated after rough machining, heat treatment or repair without destroying the part.
The advantage is conditional. A planar crack reflects best when the sound beam strikes the crack face at a useful angle. A rounded pore, an inclusion and a lack-of-fusion indication can return very different signals. UT may also struggle near edges, holes, thin walls, coarse grains or curved surfaces. Buyers should not choose UT only because it is portable or non-destructive. For precision machining services, the better decision compares failure mode, material, geometry, allowable access and acceptance standard. If the defect is surface-breaking, magnetic particle or penetrant testing may be better. If the defect is volumetric porosity in a complex casting, radiography or CT may be needed. When no single method covers the full risk, a supplier should explain the combined inspection route instead of forcing every defect type into UT. That explanation should name what UT will detect, what UT may miss, and what secondary method closes the gap.
Shrinkage porosity and looseness usually come from casting, solidification, powder metallurgy or additive processes before final machining. CNC cutting can expose the part shape, but machining does not remove every internal void. During superalloy machining, internal porosity can reduce effective load area and create stress concentration in hot or cyclic service. UT can help locate larger clusters or reflectors when the material has suitable acoustic response. The buyer should identify whether the suspected defect is isolated porosity, distributed shrinkage, lack of bond or a density-related indication, because each one affects UT sensitivity differently. A porosity cluster may require different acceptance logic from a single planar crack. If density variation is the main concern, radiography or CT may provide a clearer map.
Internal cracks may come from raw material, forging, quenching, welding, additive processing, rough machining release or overload before final inspection. They may also become visible only after multi-axis machining changes the stress state or removes support material. Non-metallic inclusions can reduce fatigue strength or corrosion performance in titanium alloy machining and stainless steel machining. UT is most valuable when the crack orientation can be reached by the beam angle. If the likely crack lies parallel to the beam, a different angle, PAUT sweep or another NDT method may be required. The RFQ should state whether crack detection, inclusion screening or thickness verification is the primary purpose, because the scan plan changes with that purpose.
Lack of fusion appears in welded or additively manufactured material when adjacent layers or weld passes do not bond as required. Porosity may also form from trapped gas, powder contamination, poor shielding or process instability. In prototype manufacturing and low-volume production, the inspection plan should separate development learning from final release. A prototype scan may identify where process parameters need correction. A production acceptance scan must define reportable indication size, scan coverage, calibration block, acceptance criteria and reinspection rules after repair or rework. If the part will later be machined from an additive blank, the buyer should decide whether UT happens before machining, after rough machining, after finishing, or at more than one stage.
A practical UT workflow starts by reviewing the drawing, material certificate, manufacturing route, service risk and customer acceptance requirement. The supplier should identify the likely defect types, defect orientation, inspection volume, critical surfaces, access limits and relevant standards before quoting the method. This step also prevents a common RFQ mistake: asking for ultrasonic testing without defining what defect must be detected. Buyers should request a proposed scan plan that names probe type, frequency range, coupling method, reference block, report format and any areas with reduced sensitivity. If the part has multiple stock forms or welded zones, the plan should separate each zone instead of using one sensitivity statement for the whole component. The plan should also say who reviews indications and when the buyer must approve disposition.
Surface preparation controls how much sound enters the part. Rough, dirty, oxidized or sharply curved surfaces can create inconsistent coupling and noisy signals. A surface roughness around Ra 6.3 micrometers may be usable in some contact UT situations, but this value is not universal. The required surface condition depends on probe type, frequency, couplant, part geometry and acceptance sensitivity. The inspection plan should define whether machining, grinding, cleaning, coating removal or local access pads are needed before scanning. For finished parts, the buyer should also confirm whether couplant residue or water exposure is acceptable. If inspection must occur after coating, the supplier should confirm that the coating does not block coupling or hide the required surface condition.
Scanning should follow a documented path that covers the required volume from accessible surfaces. Conventional pulse-echo UT may be enough for simple sections. Angle-beam UT or phased-array UT may be needed when expected defects are directional or geometry is complex. Data acquisition should record gain, time base, probe angle, focal law, scan index, calibration check and rejected zones. Buyers should be careful with statements such as full coverage. Full coverage means coverage of the specified inspection volume under the declared sensitivity, not every microscopic flaw in every possible orientation. If a hole, fillet, rib or curved wall blocks the beam, the report should name that limitation rather than silently accepting incomplete coverage.
UT reporting should turn waveform or scan data into a decision that the buyer can review. A useful report identifies the part number, revision, material, thickness, inspection standard, procedure, equipment, probe, calibration reference, operator qualification, scan coverage, indications found, acceptance criteria and final disposition. When an indication is found, the report should separate location, amplitude, estimated size, type, confidence and recommended action. Personnel qualification should be documented by the applicable scheme and scope. It should not be assumed from a marketing claim or from the word certified alone. If the acceptance standard allows engineering review, the report should show whether the lot was accepted, rejected, sorted, repaired or sent for additional testing. The report should also state if any surface, depth zone or feature was excluded.
In the aerospace sector, UT may be used for selected forged, wrought, welded or additively manufactured components where internal integrity affects fatigue or fracture risk. Materials such as Inconel 718 can require tight control of grain structure, heat treatment state and scan sensitivity. The buyer should not rely on a generic aerospace label. The RFQ should provide the applicable material specification, customer NDT requirement, acceptance class, part criticality and whether the supplier needs a specific accreditation or customer approval for that UT scope. It should also clarify whether the inspected item is raw stock, a near-net blank, a repaired part or a finished CNC component.
For medical device parts, UT suitability depends on part type, material, risk classification, surface condition and validation requirements. A trial instrument, fixture or surgical tool may have a different inspection requirement from an implant. Materials such as Ti-6Al-4V can transmit sound well in some forms, but geometry, surface finish and regulatory documentation still matter. Buyers should specify whether UT is for raw material screening, manufacturing validation, lot release or failure investigation. That distinction changes the standard, sample size and report evidence. If the surface is polished, coated or passivated after UT, the buyer should also define the final inspection state.
In the power generation and oil and gas sectors, UT may support inspection of pressure-bearing, rotating, welded or high-temperature components. The method can identify internal reflectors before parts enter service, but acceptance normally depends on a code, material specification or project requirement. A valve body, manifold, rotor, shaft or pressure component may need different probe angles and reporting thresholds. Buyers should ask whether ASME, ASTM, ISO, customer specifications or project drawings define the decision rule. The inspection plan should also state whether repair, reinspection or rejection is required after a relevant indication. For pressure-related parts, the report should connect indication disposition to the construction code or purchase specification, not only to an internal supplier note.
The main advantage a buyer should require when choosing Neway for ultrasonic testing is a documented connection between the inspection plan and the manufacturing route, not a list of equipment. For a one-stop manufacturing service, the practical value comes from deciding when to inspect raw stock, when to inspect after rough machining, when to repeat after heat treatment, and when final dimensional inspection should follow UT. The buyer should ask how ultrasonic testing is linked with machining, deburring, heat treatment, surface finishing, dimensional inspection and nonconformance control. The answer should include workflow, not just a certificate name. This is especially important when a defect found after finishing would require expensive rework or when raw material rejection must be separated from machining responsibility.
For heat-treated parts, UT may be used to screen for quench cracks, internal discontinuities or changes that become critical after hardness and stress state change. For parts that have undergone deburring processes, the supplier should confirm that surface condition still supports coupling and that edge rounding has not hidden a critical access point. The RFQ should state the inspection stage, finish state, cleaning requirement and whether UT data must be retained with other quality records. If UT is performed before a later machining cut removes material, the buyer should confirm whether the remaining critical volume is still fully represented.
Material response should drive UT parameters. Conventional stainless steel SUS304, nickel alloy, titanium, aluminum, plastic and ceramic parts can require different probe frequencies, couplants, scan paths and acceptance standards. The best buyer request is specific: provide drawing revision, material grade and condition, manufacturing route, expected defect type, critical section thickness, surface finish, acceptance standard, required report fields and any customer approval requirements. That information lets the supplier return a defensible UT plan instead of a generic promise about internal defect detection. A practical review also separates defect screening from dimensional acceptance. UT can support an integrity decision, but final release still needs the drawing's datums, tolerances and surface requirements checked by the appropriate dimensional or visual method. If the supplier cannot state inspection volume, reportable indication size, calibration method and acceptance criteria, the buyer should treat the UT proposal as incomplete. The next step is to clarify those items before price comparison or production release.
How small can internal defects be for ultrasonic testing to detect them?
Is ultrasonic testing applicable to all materials, such as plastics and ceramics?
What are the main differences between X-ray inspection and ultrasonic testing?
Which international standards do Neway’s ultrasonic testing services comply with?