Key Points of Ultrasonic Testing for Tungsten and Molybdenum Materials

Ultrasonic testing of molybdenum billets often runs into one of two situations: a clear back-wall echo accompanied by dense noise, or a low back-wall echo with no obvious defect to be found. The former is related to the coarse-grained structure; the latter may be poor coupling or excessive attenuation. Rule out these two kinds of interference before interpreting the result.

Before testing, calibrate the instrument and probe with a standard reference block, covering sensitivity, time-base linearity and dead zone. Keep the calibration record with the report, and reconfirm it after changing the probe or the shift.

How to Select the Probe

Raising the frequency improves resolution but also increases attenuation. Molybdenum has a high sound velocity and coarse grains, so at too high a frequency the grass-like noise becomes pronounced and masks small defects instead. Standard practice is to balance penetration against resolution: first scan the whole part at a lower frequency, then re-check anomalous areas locally at a higher frequency. The crystal size of the probe affects near-field length and coverage; small crystals suit edges and thin walls, large crystals favor deep scanning.

The timing of the test also affects the result. Forgings or rolled stock carry high internal stress after deformation, and testing too early may miss delayed cracks; scheduling the test after heat treatment and before finish machining balances defect exposure against subsequent removal.

Coupling and Scanning

If the molybdenum surface has been ground or turned, roughness has a marked effect on coupling. The direct-contact method uses machine oil or a dedicated couplant, and care must be taken to exclude air bubbles; the immersion method suits regular shapes and automated scanning, and the water path and water temperature must be controlled. The scan path must cover the end face, the side faces and critical areas; move the probe at a constant speed and leave sufficient overlap between adjacent paths. Steps, holes and threaded sections readily produce structural echoes, so familiarize yourself with the part geometry before scanning. Sensitivity on the end face and on the side faces may differ, and the position of the same defect measured from two directions should correspond.

Interference from Coarse-Grained Structure

• Grass-like noise: scattering at grain boundaries produces dense noise and lowers the signal-to-noise ratio.
• Lower back-wall echo: increased attenuation or poor coupling, which must be distinguished from a large internal defect.
• Phantom echoes: caused by geometric reflection and mode conversion; re-checking from another angle can rule them out.
• Near-surface dead zone: affected by the probe near field and the initial pulse, and especially pronounced on thin parts.

When the noise is dense, try lowering the frequency, increasing the averaging count, or switching to a focused probe. The echo characteristics of the same defect scanned from different directions should remain consistent; signals that appear in only one direction need further verification.

Common defects include inclusions, porosity, cracks and laminations, and the echo characteristics differ by type: inclusions are mostly single strong reflections, porosity appears as dense low-amplitude signals, and laminations give clear indications during parallel scanning. Scan speed and instrument pulse repetition frequency must be matched; too fast a speed misses small defects. Attach a scan diagram to the report to make re-inspection and locating easier.

Evaluation and Records

The evaluation basis should be agreed in advance: defect equivalent size, indication length, density and the area concerned. Acceptance criteria may cite documents recognized by both parties or be set by drawing and process requirements, and should not be adjusted after testing. Records include probe parameters, instrument settings, scan coverage, abnormal positions and waveform screenshots, to support re-inspection and traceability.