Ultrasonic Rail Inspection Checklist: 10 Best Practices

An ultrasonic rail inspection checklist is a field control list that proves each rail test met 49 CFR Part 213. It confirms the segment was tested inside its 213.237 limit (370 days or 30 MGT on Class 4–5 track), the operator was qualified under 213.238, every suspect was verified within 4 hours, and the 213.241 record is complete.

The ten practices below follow one test from planning to post-test review. Each one names the rule or source behind it, so the checklist can go straight into a procedure, a contract or an RFP. A worked test-day timeline and a printable table follow the ten practices.

This checklist covers internal rail testing on FRA-regulated track (Class 1–5). Visual track inspection is a separate duty under 49 CFR 213.233, and Class 6 and higher track is tested under 213.339.

What does an ultrasonic rail inspection checklist need to prove?

It needs to prove four things: the search was valid, the operator was qualified, every suspect was resolved in time, and the record shows it. A rail test that finds defects but can't show those four things doesn't count as a compliant internal inspection.

Key takeaway: a checklist that only lists "check probes" and "scan the rail" misses most of what an FRA inspector or an auditor will ask for.

1. How should you plan the test: segment, interval and last valid search?

Start from the rail inspection segment and the date and tonnage of its last valid search. Section 213.237(c) sets the maximum interval from that point, and track that wasn't validly searched last time doesn't reset the clock.

Under 49 CFR 213.237(c), the limits are:

Segments are the owner's designated stretches with milepost limits under 213.237(b). The interval and the service failure rate are both tracked per segment. The full rule is explained in our guide to 49 CFR 213.237 rail flaw detection requirements, and the calendar-vs-tonnage arithmetic in how often rail must be tested.

The planning checklist should record:

• Segment ID, milepost limits, track number and class
• Date and tonnage of the last valid search, and any stretches the last test listed as untested
• Annual tonnage, and the resulting due date by days and by MGT
• Test type. An owner using continuous rail testing must designate it before testing, and say whether it counts as the 213.237 inspection (213.240(c))
• Track access, the on-track safety method and the job briefing under 49 CFR 214.315. Rail transit not connected to the general system follows FTA Part 671 and the state safety oversight agency instead

Key takeaway: the due date comes from the last valid search, so carry forward every untested stretch from the previous record.

2. Is the operator qualified under 213.238?

Only an operator qualified under 49 CFR 213.238 can make a valid search. The rule requires a documented training program, at least 160 hours of supervised experience and written authorization from the employer for each equipment type.

Section 213.238(b) lists four duties the operator must be trained and authorized for:

1. Conduct a valid search with the specific equipment types they're qualified on
2. Determine that the equipment is performing as intended
3. Interpret equipment responses and take action under the employer's procedures
4. Determine that each valid search is continuous and "has not been compromised due to environmental contamination, rail conditions, or equipment malfunction"

The 160 hours of rail flaw detection experience must be under direct supervision of a qualified operator or the equipment manufacturer's representative, or a combination. Reevaluation follows the employer's own program and needs a recorded examination; the rule sets no fixed interval. Operators proficient before January 24, 2014 are "deemed" qualified.

For the checklist, confirm before the test that the employer's qualification record names the operator, the equipment type in use that day, the qualification date and the latest reevaluation (213.238(e)). FRA doesn't certify rail flaw operators, and an ASNT or SNT-TC-1A level doesn't replace the 213.238 record.

Key takeaway: check the qualification against the equipment type, not just the person.

3. How should the system be calibrated and verified?

Calibrate and verify the system on a known reference before testing, and typically again after. Part 213 doesn't prescribe a calibration method; it requires the operator to determine that the equipment "is performing as intended" (213.238(b)(2)).

Typical practice, which the owner's or contractor's procedure should define:

• Test vehicles and carts are commonly run over a test rail or reference rail section with known artificial reflectors (drilled holes, saw cuts) at the start of the shift. Each channel is checked for a response at the set gate and sensitivity.
• Handheld flaw detectors are set for zero, velocity, angle and gain, often to 80% full screen height on a known reflector. FRA research found this is usually done on an IIW Type 2 block.
• An end-of-shift check on the same reference is common. If it fails, the procedure should say which data since the last good check must be re-tested.
• Record the instrument and probe serial numbers, the reference used, the settings and the time.

FRA's 2026 study of handheld rail UT, DOT/FRA/ORD-26/07, found that North American railroads have no standardized calibration procedure for handheld rail UT. It also found that, with a small exception, no two operators produced statistically similar calibration results, and one operator was inconsistent across six occasions. The report warns that calibration errors "can cause excessive false positives and false negatives." Our post on ultrasonic testing calibration for rail inspection RFPs covers how to write calibration into a contract.

Key takeaway: a written calibration step, with a reference standard and a recorded result, is what turns "performing as intended" into evidence.

4. Which probe angles should the test use?

A rail test typically combines a 70° beam for transverse defects in the head, a 0° beam for horizontal defects and loss of back-wall echo, and a 37–45° beam for the web and bolt holes. These are common industry configurations, not a federal requirement.

One published system uses 0°, 38° and 70°. Wheel-probe transducers are "typically centered at 2.25 Mhz" in one patent's description (US 5,777,891). Section 213.237(f) also requires that the equipment detect defects "between joint bars, in the area enclosed by joint bars", so the setup must cover rail ends inside the bars.

A common draft error is saying a straight beam finds transverse defects. It doesn't; the 70° beam does. The checklist item is simple: confirm every channel the procedure requires is active and responding before the test leaves the reference rail.

5. How do coupling and test speed affect the search?

Lost coupling or excessive speed turns a test into an untested stretch. If sound doesn't enter the rail, there is no search, and 213.241(c)(6) requires the record to show that track as not tested.

Wheel probes are fluid-filled and are commonly coupled with a water spray; hand probes use a gel or liquid couplant. Coupling suffers on heavily lubricated gage faces, wet or frozen rail, scale and grinding marks, and badly worn heads. ORD-26/07 notes that flat handheld probes couple poorly on heavy gage face wear. A TTCI digest (TD-16-030) adds that on worn rail fixed-angle probes can inspect "the wrong volume of material."

Part 213 sets no test speed. The speed limit is whatever still gives a valid search on that equipment and rail condition. A 1999 university report (MPC 99-106) described test cars capable of about 30 mph non-stop but averaging 6–8 mph with stop-and-verify, which shows how much verification stops drive production.

Checklist items for coupling and speed:

• Couplant supply and spray working on every probe wheel
• Loss-of-coupling (0° back-wall) alarm active and watched
• Speed reduced where rail condition degrades the signal
• Every coupling loss logged with milepost limits, not dismissed as noise

See our guide to ultrasonic couplant for rail inspection for couplant choice.

6. How do you prove a valid search and continuous coverage?

A valid search must be continuous over the whole segment, so every gap has to be found, logged and closed before the deadline. The definition in 213.237(j)(4) is short:

"Valid search means a continuous inspection for internal rail defects where the equipment performs as intended and equipment responses are interpreted by a qualified operator as defined in § 213.238." (49 CFR 213.237(j)(4))

Under 213.237(g), if the qualified operator decides a valid search couldn't be made over a length of track, that length "may not be considered as internally inspected." Under 213.237(h), before the time or tonnage limit expires, the owner must do one of three things:

1. Conduct a valid search
2. Reduce speed to a maximum of 25 mph until a valid search can be made
3. Replace the rail that wasn't inspected

Coverage checks for the checklist: start and end mileposts match the plan, there are no gaps between runs, rail inside joint bars was covered, and each untested stretch has an owner and a date for one of the three options.

Lost coupling isn't a "false call". It's untested rail. The difference matters: a false call is a suspect that proves benign on verification, while an invalid search is rail that was never tested at all.

7. How fast must suspect indications be verified?

Under 213.113(b), the track owner must verify an indication within four hours, or immediately if it suggests a defect needing remedial action A, A2 or B. Owners using continuous rail testing under 213.240 work to 84-hour and 36-hour windows instead.

"Except as provided in § 213.240, the track owner must verify the indication within four hours, unless the track owner has an indication of the existence of a defect that requires remedial action A, A2, or B … in which case the track owner must immediately verify the indication." (49 CFR 213.113(b))

Under 213.240(e)(4), "A suspect location is not considered a defect under § 213.113(c) until it has been field-verified by a person qualified under § 213.238." A broken rail with separation needs protection immediately in either case.

Verification usually means a hand test at the marked location, at several angles, to confirm and size the reflector. Many suspects turn out to be bolt holes, weld geometry or surface damage. Our article on false calls in rail UT covers how to tell them apart and why a low false-call rate can hide missed defects.

Key takeaway: log the time of each indication and its verification; the clock is the checklist item.

8. What remedial action does 213.113 require?

Once a defect is confirmed, a person designated under 213.7 decides whether the track may stay in use, and operation over the defective rail isn't allowed until the rail is replaced or repaired or the 213.113(c) remedial action is started. The action depends on the defect type and size.

Selected rows of the 49 CFR 213.113 remedial action table, for defects sized by percent of rail head cross-section:

Compound fissures use different bands: B from 5% to under 70% of head area, A2 from 70% to under 100%, and A at 100%.

What the letters mean, in short: A is visual supervision of each operation over the rail; A2 is a visual inspection, then no more than 10 mph for up to 24 hours; B is a speed of 30 mph or less set by a designated person with at least one year of supervisory experience; C is joint bars within 10 days; D is joint bars within 7 days; E is joint bars bolted under 213.121; H is 50 mph. Notes C and D also set speed limits until the bars are applied, so read them in full. Compound fissures have no joint-bar option.

The checklist items for this step:

• Defect marked with a highly visible marking on both sides of the web and base, or on or next to the head where those are inaccessible (213.237(e))
• Designated person notified, with the time
• Remedial action letter, speed and any joint-bar deadline recorded
• Size recorded if the rail isn't removed before the next train

Defect names and remedial letters are explained in FRA defect codes explained.

9. What must the inspection record contain under 213.241?

A 213.237 inspection record must give the date, the track and its limits, each defect's location, type and size, the initial remedial action and date, and any track not tested. Records are kept for at least two years after the inspection and one year after the initial remedial action.

For electronic records, 213.241 requires that each author is uniquely identified, completed records can't be modified, amendments are stored separately, and the person making the inspection initiates storage within 72 hours of completing it. Continuous rail test records under 213.240 add the dates and times of data collection and of each suspect's identification and field verification.

Good practice goes beyond the minimum: add the operator, equipment, calibration results and each suspect found to be benign. Those extra fields are what make the post-test review possible.

10. What belongs in the post-test review?

The post-test review closes every open item and checks the test's quality before the record is filed. Part 213 doesn't name a "post-test review", but several of its duties can only be met by one.

A post-test review should confirm:

• The end-of-shift calibration check passed, or the affected rail is scheduled for re-test
• Every suspect has a verification time inside its window and a result
• Every confirmed defect has a remedial action, a marking and a designated person's decision
• Every untested stretch has a plan under 213.237(h) and a date before the segment's deadline
• Electronic storage of the record was initiated within 72 hours
• Service failures in the segment are counted toward the rate calculated within 45 days after each 12-month period (213.237(a))

The review is also where operator performance is checked. Section 213.238(d) says reevaluation can draw on this data: "Proficiency may be determined by a periodic review of test data submitted by the operator." Given the ORD-26/07 finding on operator-to-operator calibration differences, trend the calibration results and false-call counts per operator over time.

Worked example: one test day, hour by hour

This example is illustrative. It shows how the ten practices fit together on a stop-and-verify test, where 213.113(b) sets the verification windows.

Inputs: Class 4 freight segment, MP 10.0 to MP 34.0 (24 miles), single track. Traffic is 36 MGT a year. The last valid search was January 15, 2026, with no untested stretches.

Due date: 30 MGT at 36 MGT a year is reached after 30 ÷ 36 × 365 = 304 days, on November 15, 2026. The 370-day limit falls on January 20, 2027. The shorter limit controls, so the test must be done by November 15. It is scheduled for Tuesday, November 3, 2026, which leaves 12 days to close any gaps.

Checks on the numbers:

• The first indication was verified 35 minutes after it appeared, well inside the 4-hour window that ended at 13:05. A 15% detail fracture is a remedial action C defect, not A, A2 or B, so "immediately" didn't apply.
• The second indication was verified in 15 minutes.
• Test time from 07:20 to 14:45 is 7 hours 25 minutes. Over 24 miles, that's about 3.2 mph overall, including two verification stops and the re-run.
• Electronic storage must be initiated by 14:45 on November 6, 72 hours after the inspection ended.
• The segment is compliant only once the 0.25 mile gap is validly searched, slowed to 25 mph or replaced before November 15.

Printable ultrasonic rail inspection checklist

Use this table as a field sheet. The owner's or contractor's procedure should add equipment-specific settings.

What this checklist can't do

A checklist controls the process; it can't make ultrasonic testing see what the physics hides. These limits should be written into the procedure and the report.

• Masking by surface damage. Shelling and head checks can block the beam. Detail fractures, which start at or near the surface from shelling or head checks, are the defects most often masked. See rail shelling and surface defects.
• Thermite weld bases. Test cars miss most of the rail base, so field welds may need a separate hand test of the base (TTCI TD-00-013).
• Joint bars themselves. Section 213.237(f) covers the rail between the bars, not the bars. Joint bar condition falls under 213.121, and CWR joint inspections under 213.119(h).
• Rail hidden at crossings. In TSB Canada investigation R19M0018, web corrosion under a crossing "had not been identified by the visual or ultrasonic track inspections, because the web of the rail was hidden by the crossing surface."
• Detection is never 100%. A 1999 TTCI test of six hi-rail systems on 49 known defects found 72% overall, and only 33% of bolt hole cracks. Part 213 sets service failure rates, not a detection rate.
• Owner procedures govern. This checklist doesn't replace the owner's procedure, the contractor's procedure or a Level III's judgment.

How EMA Industries applies this

EMA Industries performs ultrasonic rail flaw testing for transit agencies, short lines and industrial track owners, from Southern California and mobilizing nationwide. Whoever does the testing, the deliverable should map onto the table above: qualification record, calibration results, every suspect with its verification time, every untested stretch, and a 213.241-ready record. To scope a test program or an RFP, see our ultrasonic rail testing services or contact EMA Industries.

Related reading: FRA track safety standards (Part 213) · 49 CFR 213.237 explained · False calls in rail UT · What is ultrasonic testing? · Ultrasonic vs visual inspection

Frequently asked questions

How often must rail be ultrasonically tested?

On Class 4 and 5 track, and on Class 3 track with scheduled passenger trains or a hazmat route, at least every 370 days or 30 MGT, whichever comes first. Other Class 3 track is tested annually or every 30 MGT, whichever is longer, with a 5-year maximum. Part 213 sets no internal testing interval for Class 1 and 2 track.

How quickly must a rail flaw indication be verified?

Within four hours under 213.113(b), or immediately if the indication suggests a defect needing remedial action A, A2 or B. Owners using continuous rail testing under 213.240 have 84 hours, or 36 hours for suspected A, A2 or B defects.

What probe angles are used in ultrasonic rail testing?

Typically a 70° beam for transverse defects in the head, a 0° beam for horizontal defects and loss of back-wall echo, and a beam of about 37–45° for the web and bolt holes. Many systems add side-looking beams for the head corner. These are industry practice; Part 213 doesn't specify angles.

What happens if the test loses coupling over part of the track?

That stretch isn't counted as inspected under 213.237(g) and must be listed as untested in the record. Before the time or tonnage limit runs out, the owner must re-test it, limit speed to 25 mph, or replace the rail.

Who can operate rail flaw detection equipment?

An operator qualified under 213.238: trained under a documented program, with at least 160 hours of supervised experience, demonstrated proficiency and written employer authorization for each equipment type. The employer qualifies the operator; FRA doesn't certify them.

How long must rail inspection records be kept?

Internal rail inspection records are kept for at least two years after the inspection and one year after the initial remedial action, under 213.241. Electronic storage must be initiated within 72 hours of completing the inspection.

Sources

• 49 CFR 213.237, Inspection of rail, eCFR (text read on Cornell LII, October 2026)
• 49 CFR 213.238, Qualified operator, eCFR
• 49 CFR 213.240, Continuous rail testing, eCFR
• 49 CFR 213.241, Inspection records, eCFR
• 49 CFR 213.113, Defective rails, eCFR
• 49 CFR 213.233, Track inspections, eCFR
• 49 CFR 214.315, Supervision and communication, eCFR
• DOT/FRA/ORD-26/07, model-assisted POD for handheld rail UT, FRA (ENSCO), July 2026
• TTCI Technology Digest TD-16-030, July 2016
• MPC 99-106, rail inspection technology report, Mountain-Plains Consortium, August 1999
• Track Safety Standards; Improving Rail Integrity (79 FR 4234, January 24, 2014), Federal Register
• TTCI Technology Digests TD-99-006 (1999) and TD-00-013 (2000); TSB Canada report R19M0018 (2019)