Types of Sound Waves Used in Rail Inspection: Wave Modes and the Beam-to-Defect Map

The sound waves used in rail inspection are mainly two bulk ultrasonic modes: a 0° longitudinal beam and shear beams at typically 37–45° and 70°, usually around 2–2.25 MHz. The 70° shear beams find transverse fissures and detail fractures. The 0° beam finds horizontal split heads and split webs, and it watches for loss of the base echo.

Surface (Rayleigh) waves and guided waves also exist in rail steel, but neither is the basis of routine internal rail testing under 49 CFR 213.237. This guide explains why each defect needs a particular beam, maps every 49 CFR 213.113 defect to the beam and channel that normally finds it, and works the beam geometry through a 136RE rail head.

The general physics (velocity, impedance, Snell's law, near field and beam spread) is covered in EMA's pillar article, the science behind ultrasonic testing. This article applies that physics to the one question rail testing turns on: which beam meets which crack face-on.

Which sound waves used in rail inspection do the routine work?

Rail flaw testing relies on two wave modes: longitudinal waves in a 0° beam and shear waves in angled beams. Surface waves and guided waves are real wave modes in rail steel, but in North American practice they are research tools or special monitoring systems, not the internal test that 213.237 requires.

Velocities are general values for carbon steel from Evident's velocity table (longitudinal) and Evident's shear value; the Rayleigh figure is calculated from them. Actual rail steel varies by a percent or two, so procedures calibrate rather than type in a textbook number.

Federal rules don't name any of these modes. Section 213.237 requires "internal rail inspections" with equipment that can detect defects between joint bars, and the word "ultrasonic" does not appear in Part 213. Ultrasonic testing is how the industry meets that rule, and the wave modes above are how ultrasonic testing does it.

Key takeaway: one longitudinal beam and a set of shear beams do the routine work; the other two modes are specialist tools.

Why does defect orientation decide which beam finds it?

A planar crack sends a strong echo back to the probe only when the beam strikes its face close to perpendicular. A crack face is a near-perfect mirror, so a beam that hits it at a steep slant reflects away from the probe, and a beam running parallel to it barely reflects at all. Rail defects grow in three main orientations, so rail testing needs beams in three main directions.

This is the single most important idea in rail ultrasonic testing, and it is where many published explanations go wrong. A longitudinal wave is not "better" at transverse cracks because of its particle motion. What matters is the angle between the beam and the crack face.

Two effects make the picture less tidy than the table. Fatigue crack faces are not perfect mirrors, so some energy scatters back even at an oblique hit. And a crack can block the sound it doesn't reflect, which is how a 0° beam reveals defects it can't echo from, through loss of the base echo.

A crack that is pressed shut by compressive stress can also let part of the sound through and reflect less than its size suggests. That is covered in EMA's physics pillar and is one reason a single channel is never trusted alone.

Key takeaway: match the beam to the crack plane, not the wave type to the defect name.

What does the 0° longitudinal beam find in rail?

The 0° longitudinal beam travels straight down from the running surface through the head, the web and the base, and it finds flaws that lie roughly parallel to the running surface. It also measures the round-trip time to the base, and a drop or loss of that base echo is itself an indication. It does not see vertical transverse planes, so it cannot find transverse fissures or detail fractures by echo.

A 0° channel usually works with two gates. A flaw gate watches the time window between the running surface and the base for early echoes. A loss-of-bottom gate (often called LOB or loss of back wall) watches the base echo and alarms when it drops below a set level.

Defects the 0° beam finds by echo

• Horizontal split head. 213.113(d)(11) describes it as "a horizontal progressive defect originating inside of the rail head, usually 1/4 inch or more below the running surface." Its plane faces the 0° beam directly, and the base echo falls as the split shadows it.
• Head and web separation. This is a progressive fracture "longitudinally separating the head from the web of the rail at the head fillet area." It lies roughly horizontal at the fillet, right in the 0° path.
• Split web. The rule defines it as "a lengthwise crack along the side of the web and extending into or through it." Where the crack runs along the web at a shallow angle, the 0° beam traveling down the web meets it close to face-on.
• The horizontal start of a compound fissure. A compound fissure begins as a horizontal split head, so the 0° channel can see it before it turns.
• Horizontal legs of bolt-hole cracks. Fully developed bolt-hole cracks can "continue horizontally along the head/web or base/web fillet," and those legs face the 0° beam.

Defects the 0° beam finds mainly through loss of bottom

Piped rail is "a vertical split in a rail, usually in the web, due to failure of the shrinkage cavity in the ingot to unite in rolling." Its main plane is vertical and runs parallel to the 0° beam, so the beam rarely gets a clean echo from it. The irregular cavity scatters and blocks the beam instead, and the base echo drops or breaks up.

The same loss-of-bottom response also flags things that are not defects. Lost coupling, a dry spot, heavy surface damage, a weld collar or a bolt hole in the path all reduce the base echo. Telling those apart is part of the operator's interpretation, and EMA's false calls in rail UT article covers the common causes.

What the 0° beam cannot do

A transverse fissure stands vertically across the head, at a right angle to the rail's length. The 0° beam travels straight down, parallel to that plane, so it meets the fissure edge-on and returns essentially no echo. A small transverse defect is also too thin to block much of the beam, so even loss of bottom is unreliable for it. This is why rail systems always run angled shear beams for the head.

Key takeaway: the 0° beam owns horizontal planes and the loss-of-bottom check; it is blind to the transverse defects that cause many service failures.

How do 70° shear waves find transverse fissures and detail fractures?

A 70° shear beam enters the head from the running surface and travels steeply along the rail, so it meets a vertical transverse crack only about 20° from face-on. That is the closest any practical top-surface beam gets to a transverse plane, and it is why 70° channels are the standard tool for transverse defects. Systems usually run one 70° beam looking forward and one looking backward, so each defect is seen from both sides.

The angle is measured from the vertical, so a 70° beam is only 20° below horizontal. A transverse crack's face points along the rail. The beam therefore strikes it at 20° from the face's normal, while a 0° beam strikes it at 90°.

Transverse fissures

213.113(d)(15) defines a transverse fissure as "a progressive crosswise fracture starting from a crystalline center or nucleus inside the head," spreading as a round or oval surface "substantially at a right angle to the length of the rail." It grows from inside the head, so the 70° beam has to reach the nucleus depth within the head. The worked example below shows the beam reaching the full depth of a 136RE head within about 5.3 in. of travel along the rail.

EMA's transverse fissure in rail article covers the defect itself: causes, growth and remedial action.

Detail fractures

A detail fracture is "a progressive fracture originating at or near the surface of the rail head," and 213.113(d)(7) says it may arise "from shelled spots, head checks, or flaking." It then turns and grows transversely, which is the plane the 70° beam is designed to meet. The difficulty is not the beam angle but the entry: the same shells and head checks that start the fracture sit where the sound has to enter, and they scatter or block it.

Detail fractures often start near the gauge corner, where shelling forms. Whether a 70° beam placed over the running surface covers that corner depends on the probe position, its aim and its beam spread, so systems differ in how they cover the gauge side. EMA's detail fracture development article covers how the defect grows.

Engine burn fractures

An engine burn fracture starts "in spots where driving wheels have slipped on top of the rail head." In developing downward, the rule says, these fractures "frequently resemble the compound or even transverse fissures." Their transverse growth is a 70° target, but the burned running surface above them is a poor entry surface, so the same masking problem applies.

How a shear beam is made at 70°

No probe launches shear waves into a liquid, because liquids have no shear stiffness. A wheel probe or wedge sends a longitudinal wave through fluid or plastic at a set incident angle, and at the steel surface it refracts and converts to shear. The refraction geometry, including why a worn running surface can swing a nominal 70° beam by several degrees, is worked in EMA's physics pillar and in shear wave ultrasonic testing.

Key takeaway: 70° shear beams, forward and backward, are the transverse-defect channels; shelling at the entry surface is their main enemy.

What do 37–45° shear beams look for?

Mid-angle shear beams, typically somewhere between about 37° and 45°, reach down into the web and are aimed mainly at bolt-hole cracks and other inclined web flaws. A 213.113 bolt-hole crack runs across the web from the hole, "inclined upward toward the rail head or inclined downward toward the base." A crack on an incline is best met by a beam traveling on a similar incline.

The direction matters as much as the angle. A crack inclined at 45° can be hit face-on by a 45° beam traveling toward it from one side, while a 45° beam from the other side runs parallel to the crack and misses it. That is why mid-angle channels, like 70° channels, are usually paired forward and backward.

Bolt holes themselves are strong reflectors. A sound bolt hole gives a characteristic, repeatable echo pattern at a known position, and the analyst looks for an extra echo, a shifted echo or an echo that persists past the hole's normal signature. Multiple cracks at one rail end count as one defect, but cracks in adjacent rail ends within the same joint are reported separately, per 213.113(d)(1).

Section 213.237(f) requires equipment that can detect defects "between joint bars," and the mid-angle and 0° channels are what reach the rail ends inside a joint with the bars in place. The bars themselves are a separate matter under 213.121. EMA's bolt hole cracks rail inspection article covers the defect and its detection.

Published practice varies on the exact angle. One historical account of North American test cars lists wheel probes carrying 0°, 45°, 70° and side-looking transducers, while one published system uses 0°, 38° and 70°. Treat "37–45°" as a typical range, and check the angles named in the system's own procedure.

Key takeaway: mid-angle beams are the web channels, paired in both directions because a bolt-hole crack faces only one way.

How are vertical split heads found?

A vertical split head is a vertical split along the rail, through or near the middle of the head, so its plane is parallel to both the 0° beam and to beams aimed along the rail. None of the core channels meets it face-on. Systems therefore use a side-looking or gauge-corner shear beam aimed across the head, and they also watch the 0° channel for indirect signs.

213.113(d)(16) defines a vertical split head as "a vertical split through or near the middle of the head," adding that "a crack or rust streak may show under the head close to the web or pieces may be split off the side of the head." A beam aimed across the width of the head, from one side toward the other, meets that vertical longitudinal plane close to face-on.

One published patent design (US 5,419,196, 1995) aims a shear beam at roughly 45–75° toward the lower corner of the head, where a split shows as either an early echo or a lost corner echo. The 0° channel can add a second clue: when the split lies under the 0° beam path, it can deflect or block the beam and reduce the base echo. Treat that as supporting evidence, not as the primary detection.

EMA's vertical split head in rail article covers the defect's causes and its remedial action.

Key takeaway: vertical splits need a beam aimed across the head; the 0° and along-rail beams see them only indirectly.

Which beam and channel finds each rail defect?

Each 49 CFR 213.113 defect has a beam that normally finds it, a backup response, and a typical channel signature. The table below is the core of rail ultrasonic testing. Channel names and counts vary by system, so read the channel column as a typical arrangement, not a specification.

Two rows deserve more explanation.

Compound fissure: horizontal turning transverse

The compound fissure is the defect most often described wrongly. It is not a crack that combines random directions. The regulation defines it precisely:

"Compound fissure means a progressive fracture originating from a horizontal split head that turns up or down, or in both directions, in the head of the rail. Transverse development normally progresses substantially at a right angle to the length of the rail." — 49 CFR 213.113(d)(3)

That definition tells the inspector which beams to use. The horizontal origin is a 0° target, and the transverse development is a 70° target, so a compound fissure often shows on both channel types at the same location. EMA's compound fissure in rail and horizontal split head articles cover each defect in detail.

Defective welds

213.113(d)(6) defines a defective weld as one with discontinuities "oriented in or near the transverse plane" exceeding 5% of the head area individually or 10% in aggregate. In the head, those planes are 70° targets; in the web, mid-angle beams help. TTCI's thermite weld study (TD-00-013) found that "the ultrasonic rail flaw detection process only scans the head and web of the rail, and not the majority of the base," so field weld bases need hand UT. EMA's aluminothermic weld inspection article covers that work.

Key takeaway: read every indication against the defect's plane; a defect on the wrong channel for its type is a reason to verify, not to dismiss.

Worked example: beam paths in a 136RE rail head

This example traces the 0°, 70° and mid-angle beam paths through a 136RE rail to show where each beam goes, when its echo arrives, and how squarely it meets each defect plane. It uses the published section dimensions and textbook velocities, and it ignores the wedge or wheel-fluid delay, which each system adds as a fixed offset.

Inputs

Depth below the running surface of the bolt-hole center is 7.3125 − 3.0938 = 4.2188 in., and the top of the hole is 4.2188 − 0.625 = 3.5938 in. down.

Step 1: 0° arrival times (depth = v × t / 2, so t = 2 × depth / v)

A 0° echo at 4.31 µs with a falling base echo fits a horizontal split head about 1/2 in. down, which satisfies the "usually 1/4 inch or more" description. An echo at about 16.7 µs points to the head fillet, the location of a head and web separation. Each is a suspect until verified.

Step 2: 70° shear paths in the head

For an angled beam, sound path SP = depth / cos θ, surface distance SD = depth × tan θ, and the round trip t = 2 × SP / v_S. At 70°, 1 / cos 70° = 2.9238 and tan 70° = 2.7475.

Check the 0.75 in. row: 0.75 × 2.9238 = 2.193 in.; 2 × 2.193 / 0.1280 = 34.26 µs; 0.75 × 2.7475 = 2.061 in. All target depths sit inside the 1.9375 in. head, so the geometry is possible.

Read the last row first. A 70° beam needs about 5.3 in. of travel along the rail to descend through the full head depth, so it sweeps the whole head cross-section in its plane over a short length of rail. A transverse fissure nucleus 3/4 in. down returns its echo when the entry point is about 2.06 in. from the defect, and the backward-looking channel sees the same defect from the other side at the same distance.

Step 3: how squarely each beam meets each plane

The angle off normal is the angle between the beam and the line perpendicular to the crack face. Zero means face-on, and 90° means edge-on.

For a vertical transverse plane, angle off normal = 90° − beam angle; for a horizontal plane, it equals the beam angle; for the 45° inclined crack, it is the difference between the beam angle and 45°. The table makes the beam-to-defect map visible: the 70° beam is the best top-surface choice for transverse planes, the 0° beam for horizontal planes, and a mid-angle beam for inclined web cracks.

Step 4: a mid-angle beam to the bolt-hole center

A 37° beam reaches the bolt-hole center about 1 in. sooner along the rail than a 45° beam and with a shorter path, while a 45° beam meets a 45° crack face-on. That trade-off is one reason published systems choose different mid angles.

What the example does not show

The ray model treats the beam as a line. Real beams spread, the near field of a wheel or wedge probe sits partly in the delay path, and worn running surfaces tilt the entry point, which shifts the refracted angle. EMA's near field vs far field in UT and angle beam testing in UT articles cover those effects. A Level III should confirm the head-depth figure against the rail's actual section and wear before using these numbers in a procedure.

Key takeaway: in a 136RE head, a 70° beam crosses the full head depth in about 5.3 in. of rail and meets a transverse crack 20° from face-on; a 0° beam meets it edge-on.

Can surface waves find rail defects?

Surface (Rayleigh) waves can find surface-breaking cracks on smooth steel, but they penetrate only about one wavelength, which at rail testing frequencies is far too shallow to reach where detail fractures start. On rail they remain a laboratory and research method. They do not provide a 213.237 valid search, and they are not how detail fractures are found.

The Federal Highway Administration's NDT glossary puts the depth limit plainly: Rayleigh-wave particle motion decreases "rapidly with depth below the surface," and "the effective depth of penetration is considered to be about one wavelength." At about 3,000 m/s, the wavelength at 2.25 MHz is 3,000 / 2,250,000 = 1.33 mm (0.052 in.).

Worked check: can a surface wave reach a detail fracture's origin?

At normal rail testing frequencies, a surface wave samples the top 1 to 2 mm of steel, while the shells that start detail fractures can lie 6 to 10 mm down in those tests. Dropping the frequency to roughly 0.3–0.5 MHz to reach that depth makes the wave long and the mode behavior complicated. Hesse and Cawley's modeling of surface wave modes in rail found several similar modes whose amplitude depends on position, so "accurate defect sizing will be difficult."

Some explanations claim that surface waves can detect early-stage detail fractures. They can at best show surface damage near the origin. The fracture itself grows transversely inside the head, which is a 70° shear target.

Where surface methods do fit

Rolling contact fatigue, head checks and shallow surface cracking are real targets for near-surface methods. In practice, rail operators rely mainly on eddy current, visual inspection and preventive grinding for those conditions, and FRA's rolling contact fatigue review (DOT/FRA/ORD-11/24) notes that surface fatigue "prevents the successful transmission of ultrasonic waves to the base of the rail and back." EMA's surface wave UT for rail article covers the research in more depth.

Key takeaway: surface waves are too shallow at rail frequencies to reach detail fracture origins; detail fractures are a 70° shear job.

Where do guided waves fit in rail inspection?

Guided waves are whole-section vibrations that travel a long way along the rail, and in practice they are used for fixed monitoring that detects complete breaks, not for the internal rail test 213.237 requires. Calling them "Lamb waves" is imprecise for rail. Lamb waves are the guided modes of a flat plate, and a rail's head, web and base form a much more complex waveguide.

Researchers model rail guided waves with semi-analytical finite element (SAFE) methods, which predict many modes, each with its own speed at each frequency. Setshedi, Loveday, Long and Wilke (Ultrasonics, 2019) fitted SAFE models to measured rail signals and identified three propagating modes in pulse-echo measurements from thermite welds. The web behaves a little like a plate at some frequencies, but the rail as a whole does not.

The best-documented service use is broken-rail monitoring. Ramatlo, Loveday, Long and Wilke (MATEC Web of Conferences, 2021) describe a permanently installed system on a heavy-haul line in South Africa that "detects complete rail breaks at long-range using guided wave ultrasound." They note that real damage signals in service are hard to obtain, so they evaluated detection performance by simulation.

Why guided waves don't replace a 213.237 search

Section 213.237 doesn't name a test method, but the internal inspection exists to find 213.113 defects while they are still small; the remedial action table's smallest band for transverse defects starts at 5% of head area. A guided wave that detects a complete break is reporting an event that has already happened. Finding a transverse fissure at a small percentage of head area, inside a joint or under a weld, needs the focused bulk beams described above.

"Guided wave UT" and "long-range UT" are the same method under two names, not competing techniques. EMA's guided wave ultrasonic testing article covers the method where it is established, mainly pipelines.

Key takeaway: guided waves monitor rail for breaks; the bulk 0° and shear beams find defects before they break.

What frequency is used for ultrasonic rail testing?

Rail flaw testing typically runs in the low-megahertz range, about 2 to 2.25 MHz for wheel probes, though exact frequencies vary by system and probe. One published test-car patent (US 5,777,891) describes wheel-probe transducers "typically centered at 2.25 Mhz." One manufacturer's hand-pushed cart lists 2.5 MHz head probes and a 4 MHz web probe, which shows the range is not fixed.

The choice is a trade-off. A 0° beam has to travel from the running surface to the base of a 7-5/16 in. rail and back, which favors lower frequency for penetration. A shear beam has to resolve cracks in the head, which favors shorter wavelength.

At 2.25 MHz, the shear wavelength in steel is about 1.44 mm (0.057 in.) and the longitudinal wavelength about 2.62 mm (0.103 in.). Those numbers are worked in EMA's physics pillar, along with why 10 MHz probes don't suit rail.

A band of 0.5–10 MHz is sometimes quoted for rail, but that is the general flaw-detection range, not a rail testing range. Guided-wave and joint-bar work sits lower: MxV Rail research found about 0.5 MHz worked best for joint bar UT, a different part and a different problem.

Key takeaway: treat about 2–2.25 MHz as typical for rail, and take the actual value from the system's procedure.

How are the beams delivered into the rail?

Test cars and carts deliver the beams through fluid-filled wheels or sleds riding on the running surface, and hand probes use contact wedges with gel or water couplant. The delivery method decides how consistently the beam enters the steel, and therefore how reliable each channel is over a given stretch.

A historical account of North American practice (MPC 99-106, Colorado State, 1999) describes fluid-filled wheel probes carrying three to six transducers per wheel and up to 12 per rail, at 0°, 45°, 70° and side-looking angles. The same report says test cars could run "up to 30 mph in a non-stop testing mode," but in practice stopped often to hand-verify indications and averaged 6 to 8 mph. Equipment has changed since 1999, so read that as a dated description of the method, not a current speed.

Coupling is the weak link. A wheel probe needs a continuous fluid film between tire and rail, and rust, grease, sand, ice and corrugation all break it. When coupling fails over a stretch, that rail was not tested, and 213.241(c)(6) requires the record to show any track not tested. EMA's ultrasonic couplant for rail inspection article covers field coupling.

Hand verification

Every suspect a test car or cart flags has to be checked by a qualified operator, usually with a hand probe at the same angles. Under 49 CFR 213.238, the operator is qualified by the employer, and under 213.113(b) the owner verifies a suspected defect within 4 hours, or immediately for defects requiring remedial action A, A2 or B. The beam-to-defect map above is what the verifying operator uses to classify the defect type.

Phased array

Phased array UT (PAUT) can sweep many angles from one probe, which in principle covers several defect planes at once. In North American rail it remains mainly a verification and research tool; TTCI tested a head-contact PAUT prototype in 2016 (TD-16-050) that found transverse defects under shells but could not inspect the base flanges. EMA's phased array ultrasonic testing article explains the method.

Test-car software usually presents each channel as a B-scan style map along the rail, with A-scans for detail. It does not routinely produce C-scan images of the rail. EMA's UT fundamentals: A-scan, B-scan and C-scan article explains the displays.

Key takeaway: the beam angles matter only as long as the beam actually enters the steel; coupling and hand verification are part of the method.

How well do these beams find defects in practice?

The best public benchmark is TTCI's 1999 test of six hi-rail detection systems on 49 known defects, which found 72% of the flaws overall, with large differences by defect type. That spread follows the beam-to-defect logic in this article: defects that face a core beam squarely were found far more often than those behind shells or in complex geometry.

The flaws had been found in revenue service, verified, and installed at the Rail Defect Test Facility at FRA's Transportation Technology Center near Pueblo, Colorado. The study (Garcia and Reiff, February 1999) is old and covers hi-rail systems of its time, so treat it as evidence of which defects are harder, not as a current detection rate.

The sample sizes are small for every group except transverse defects, so treat the split-head, weld and bolt-hole rates as indicative.

The pattern is the useful part. Horizontal split heads, which face the 0° beam squarely and also cut the base echo, were found most often. Small transverse defects, defects under shells and bolt-hole cracks in joint geometry were found least often.

Unsourced claims of "90%+ detection" circulate online. No single figure describes rail UT performance; the honest answer depends on defect type, size, surface condition and the system. No federal rule requires 100% detection; 213.237 sets valid searches, intervals and service failure rate limits instead.

Key takeaway: detection rates rise with defect size and with how squarely a core beam meets the defect; shells and joints cut them.

What are the limits of sound wave rail inspection?

Sound wave rail inspection reports only reflectors that the beam reaches, that face it closely enough to echo, and that stand out from noise. Each limit below follows from the beam-to-defect map, and an engineer quoting a rail test result should know them.

What this article doesn't cover

This article maps beams to defects. It doesn't set reject levels, size defects, or replace a written procedure; sizing and remedial action follow the 213.113 table, and calibration follows the system's procedure. It also doesn't cover joint bar inspection, which is a separate problem under 213.121, or structural weld UT under AWS D1.1.

Key takeaway: every result means "no reportable reflector found by these beams on this day," never "this rail is flawless."

How EMA Industries applies this

EMA Industries performs ultrasonic rail flaw testing for transit agencies, short lines and industrial track owners, along with rail weld inspection and NDT Level III services. The beam-to-defect map in this article is what a verifying technician uses on site: 0° for horizontal splits and loss of bottom, 70° for transverse defects, mid-angle beams at bolt holes, and hand UT of thermite weld bases that top-surface beams don't reach. EMA's Level III services also cover written procedures and technique review for owners who want their channel set and verification steps documented. EMA Industries is based in Southern California and mobilizes nationwide; to discuss a rail testing or verification scope, contact EMA Industries.

Related reading: how ultrasonic rail flaw detection works, rail shelling and spalling and 49 CFR 213.237 rail flaw detection inspection requirements.

Frequently asked questions

Which ultrasonic beam finds transverse fissures in rail?

A 70° shear-wave beam, usually run in both forward and backward directions, finds transverse fissures and detail fractures. A transverse fissure stands vertically across the rail head, and a 70° beam meets that plane about 20° from face-on. A 0° straight beam runs parallel to the fissure and returns essentially no echo from it.

Can a 0° straight beam find a detail fracture?

A 0° beam can't reliably find a detail fracture, because the fracture grows in the transverse plane, parallel to the 0° beam. The 0° beam finds horizontal flaws such as horizontal split heads, split webs and head and web separations, and it flags loss of the base echo. Detail fractures need the 70° shear channels, and shells at the surface can mask even those.

What frequency is used for ultrasonic rail testing?

Rail testing typically uses about 2 to 2.25 MHz for wheel probes, though systems differ. One published test-car patent describes transducers "typically centered at 2.25 Mhz," and one hand-pushed cart lists 2.5 and 4 MHz probes. The figure balances penetration to the rail base against resolution in the head; check the system's procedure for the actual value.

Are guided waves used for FRA rail flaw testing?

Guided-wave testing is not how owners meet 49 CFR 213.237. Guided waves are used in fixed systems that detect complete rail breaks over long distances, and in research. Internal rail testing to find 213.113 defects before they break relies on 0° longitudinal and angled shear beams with hand verification by a 213.238-qualified operator.

What is a compound fissure, and which beam finds it?

Under 49 CFR 213.113(d)(3), a compound fissure is a progressive fracture that starts as a horizontal split head and then turns up or down, developing transversely across the head. Its horizontal start is a 0° target, and its transverse development is a 70° target. It often shows on both channel types at the same rail location.

Sources

• 49 CFR 213.113 Defective rails (eCFR); definitions read verbatim on Cornell LII
• 49 CFR 213.237 Inspection of rail (eCFR); text read on Cornell LII
• 49 CFR 213.238 Qualified operator (eCFR)
• 49 CFR 213.241 Inspection records (eCFR)
• TTCI Technology Digest TD-99-006, Benchmarking Results of North American Rail-Flaw Detection Technologies (Garcia and Reiff, February 1999)
• TTCI Technology Digest TD-00-013 (Kristan, July 2000), thermite weld failures and base coverage
• TTCI Technology Digest TD-10-004 (Szablewski and Robles Hernández, March 2010), shelling
• TTCI Technology Digest TD-16-030 (Witte and Poudel, July 2016)
• TTCI Technology Digest TD-16-050 (November 2016), head-contact PAUT prototype
• Jeffrey and Peterson, Assessment of Rail Flaw Inspection Data, MPC 99-106, Colorado State University (August 1999)
• FRA DOT/FRA/ORD-11/24, Rolling Contact Fatigue: A Comprehensive Review (Magel, 2011), ROSA P dot/23669
• FRA RR 22-25, ultrasonic inspection of cast manganese special trackwork, ROSA P dot/64412
• FHWA NDT glossary: Rayleigh waves
• Hesse and Cawley, "Surface wave modes in rails," JASA 120 (2006) 733–740
• Setshedi, Loveday, Long and Wilke, Ultrasonics 96 (2019) 240–252, SAFE modeling of rail guided waves
• Ramatlo, Loveday, Long and Wilke, MATEC Web of Conferences 347:00034 (2021), guided-wave broken rail detection
• US Patent 5,777,891, ultrasonic rail testing system with wheel probes (2.25 MHz)
• US Patent 5,419,196, rail head side-looking probe
• Evident, Material sound velocities
• Harmer Steel, 136 lb. AREMA (136RE) section drawing
• TSB Canada rail investigation report R19M0018