What an ultrasonic flaw detector is
An ultrasonic flaw detector sends a pulse of sound into a part and reads what comes back. A discontinuity inside the material reflects part of the pulse earlier than the back wall would, and the time it takes and the amplitude it returns at together give the position and a measure of the size of the defect. Because the sound travels through the material rather than around it, the method finds defects that are entirely internal, which is what makes it the usual companion to a surface method on welded and forged work.
It is a two-part measurement. The instrument generates, times and displays; the probe decides what can be found. A straight probe sends sound along the beam axis and finds laminations and defects parallel to the surface. An angle probe refracts the beam into the material at an angle and is what weld inspection is built around, because the defects of interest in a weld usually lie at an angle to the surface and need a beam that crosses them.
The difficulty is that the instrument shows echoes rather than defects. Deciding which echo is a flaw, which is a geometric reflection from the weld cap or the root, and which is electrical noise is the operator judgement the method depends on, and it is why calibration with a reference block and a written procedure sit at the centre of any inspection rather than at its edges.
How an ultrasonic flaw detector measures
A piezoelectric probe converts an electrical pulse into sound and the returning sound back into a voltage. The instrument times the interval between the transmitted pulse and each echo, converts it to a distance using the sound velocity of the material - which for steel is around 5920 m/s in longitudinal mode - and displays the result as a trace in which echoes appear as vertical deflections along a distance axis. The velocity has to be set for the material, because the same time delay becomes a different distance in aluminium or in austenitic stainless steel.
Sizing rests on a relationship between echo amplitude and reflector size, and that relationship is established with reference reflectors of known size in a block of the same material. The instrument draws an amplitude-distance curve from those references, and subsequent echoes are compared against it, so a defect is reported in terms of equivalent flat-bottom hole diameter or a length at a stated amplitude threshold. Because the curve depends on the probe, the material and the geometry, it is built for each job rather than carried over from the last one.
The practical limits come from the beam. Sound has to get in, so the surface has to be smooth enough and a couplant has to bridge the gap; a rough or scaled surface loses most of the energy before it enters the part. The beam also has a width and a near field, so a defect smaller than the beam width is sized by the beam rather than by itself, and a defect close to the surface sits in a zone where the beam has not yet formed. Angle probes add geometry to the arithmetic, which is why the probe angle and the weld preparation are part of what the inspection procedure states.
Configuration choices
Most of the specification is fixed by these decisions rather than by the accuracy class. Each one constrains the next, which is why they are settled in order.
| Configuration | Options to fix | What it actually decides |
| Probe type | A straight beam probe, or an angle probe at a stated refraction angle | What the instrument can find. A straight probe cannot cross a defect lying parallel to its own beam. |
| Probe frequency and crystal size | The nominal frequency of the probe and the size of its active element | The trade between penetration and resolution. A higher frequency resolves smaller defects and penetrates less; a larger crystal sends a wider beam with more energy. |
| Instrument bandwidth and gain range | The frequency range the receiver accepts, and the maximum gain available | Whether the probe can be driven at the frequency it is designed for, and whether weak echoes can be lifted far enough to read. |
| Distance and amplitude curves | DAC, AVG or TCG, and whether the instrument can build and store more than one | How a defect is sized and recorded. Without a reference curve the echo amplitude on its own means very little. |
| Sampling rate and displayed resolution | The rate at which the returning signal is digitised, and the resolution of the display | Whether closely spaced echoes stay separate, and whether thin sections can be read at all. |
| Display and mode | A-scan alone, or A-scan with B-scan or a stored reference waveform | Whether the operator can see the geometry of a defect rather than a single set of echoes, and whether a trace can be kept as evidence. |
| Storage and reporting | Internal memory, removable storage, and the report format the instrument can produce | Whether an inspection record can be issued and archived, which is what most procedures require after the fact. |
| Reference blocks and calibration | The blocks and probes kept for calibration, and the interval at which the set is checked | Whether the inspection can be shown to have been carried out on a calibrated instrument, which is what makes the record admissible. |
How to choose an ultrasonic flaw detector
| Factor | Why it decides the answer | How to settle it before enquiring |
|---|
| The material and its sound velocity | The distance scale is computed from the velocity, and coarse or austenitic material also scatters the beam and raises the noise floor | State the material and grade. For cast or austenitic material, say so, because it changes the probe and the frequency. |
| The defect being looked for and where it lies | An internal lamination, a weld toe crack and a forging defect sit at different depths and orientations and need different probes | Describe how the defect is expected to have formed and where in the section it is expected to sit. |
| Section thickness and geometry | Thin sections rule out low frequencies and put defects inside the near field, and complex geometry sends echoes back from the shape itself | Give the thickness range and a drawing of the joint or section, including the weld preparation. |
| The acceptance standard | It names the reference blocks, the probe angles, the scan pattern and the reporting thresholds, and it often fixes the instrument class | Quote the standard and the acceptance level, rather than describing the requirement in general terms. |
| The surface condition | Sound has to enter the material, and rough, scaled or coated surfaces lose the beam before it is useful | Describe the surface as it will be at the time of inspection, not as it is drawn. |
| Whether the record has to stand up | A full report with stored traces and calibration data changes the instrument class and the time the inspection takes | State whether traces must be archived, whether reports are issued, and who will read them. |
If any of these are still open, send what is known together with the joint drawing. A quotation can be issued against the geometry and the standard and refined once the remaining figures are available.
The nearest alternatives
These are the types an enquiry for an ultrasonic flaw detector usually turns out to be, and the point at which the choice is genuinely decided.
| Nearest alternative | What separates it from an ultrasonic flaw detector |
| Magnetic particle testing equipment | Finds surface and near-surface defects in ferromagnetic material by making the part a magnet, and shows the indication directly on the surface. It cannot see a defect that lies deep in a section, which is exactly what an ultrasonic instrument is used for, and it does not work at all on non-ferromagnetic material. |
| Ultrasonic thickness gauges | Uses the same pulse-echo principle but reads only the interval to the back wall, so it reports remaining material rather than looking for a discontinuity. It is a corrosion-mapping instrument; a flaw detector is a defect-finding instrument, and the two are distinguished by how much of the return signal they interpret. |
| Coating thickness gauges | Measures a layer on the surface by magnetic or eddy current means and says nothing about what is inside the material. On a coated part the coating often has to be accounted for in the ultrasonic reading rather than measured by the same instrument. |
| Radiographic inspection | Transmits radiation through the part and images the internal volume, which gives a permanent two-dimensional picture of a defect. It needs access to both sides and carries a radiation safety regime, and it is less sensitive than ultrasound to tight planar defects that do not reduce the wall thickness in the beam direction. |
| Eddy current testing | Finds surface and near-surface defects in conductive material without a couplant and scans tubes quickly. The penetration is shallow, so it complements ultrasonic testing on tube and bar rather than replacing it on heavy sections. |
| Acoustic emission testing | Listens for the sound released by a defect that is actively growing, under load, rather than actively looking for it. It surveys a whole structure at once and localises, but it needs a load cycle and it says little about the size of a dormant defect. |
A flaw detector looks for a discontinuity anywhere in the section and reads the echoes that come back from it; a thickness gauge reads only the back wall. That boundary - how much of the signal is interpreted - is what separates the two product types.
What to include in the enquiry
The first reply can be a quotation rather than a set of questions if the enquiry carries these. Where a figure is not known, send the duty description and we will confirm what it implies.
- The material and grade, and whether it is coarse-grained or austenitic
- The defect being looked for, how it is expected to have formed, and where in the section it lies
- The section thickness range and a drawing of the joint or geometry
- The acceptance standard and the acceptance level to be worked to
- The surface condition at the time of inspection
- The probe types expected to be used, including angles and frequencies if the standard fixes them
- Whether stored traces and a written report are required
- Whether the instrument is used in a workshop or on site, and the power available
- Reference blocks or calibration arrangements available locally
- Quantity, destination port and required delivery date
