A sun kink is a heat buckle in continuous welded rail (CWR): when rail temperature climbs far above its neutral temperature, compressive force (about 2,600 lb per °F in a 136 lb rail) can exceed the track's lateral resistance and throw the track sideways. Under 49 CFR 213.119, every CWR owner must have a written plan to control it.
The sun kink railroad problem is not just "hot weather". It is the gap between today's rail temperature and the temperature at which the rail carries no force, multiplied by the stiffness of the steel, set against how well ties, ballast and anchors hold the track in line. This guide covers the physics with worked numbers, what the federal rules require, how railroads use slow orders, and where rail flaw testing and weld inspection fit in.
Contents: What is a sun kink? · Why rail heats · The force · Why RNT drifts · What the rules require · Slow orders · Prevention · Where NDT fits · Limits · FAQ
What is a sun kink on a railroad?
A sun kink is a lateral buckle in rail track caused by thermal compression, not by a broken or defective rail. The FRA's 2012 safety advisory on the subject described derailments caused by "rail buckling under extreme heat conditions (commonly referred to as 'sun kinks' in the rail)".
The definitions in 49 CFR 213.119(l) give the vocabulary every CWR plan uses:
"Rail neutral temperature is the temperature at which the rail is neither in compression nor tension." — 49 CFR 213.119(l)
The same paragraph defines a buckling-prone condition as a track condition that can result in the track being laterally displaced, and tight or kinky rail as CWR with minor alignment irregularities that show it is under substantial compression. A buckle usually starts at one of those small irregularities, where the stored force has somewhere to go.
| Term | Meaning | Where defined |
|---|---|---|
| CWR | Rail welded into lengths of more than 400 ft | 49 CFR 213.119(l) |
| Rail neutral temperature (RNT) | Rail temperature at which the rail carries no longitudinal force | 49 CFR 213.119(l) |
| Desired rail installation temperature range | The band, set for a geographic area, in which CWR should not buckle in extreme heat or pull apart in extreme cold | 49 CFR 213.119(l) |
| Buckling-prone condition | A track condition that can result in lateral displacement | 49 CFR 213.119(l) |
| Tight or kinky rail | Minor alignment irregularities showing substantial compression | 49 CFR 213.119(l) |
| Disturbed track | Roadbed or ballast disturbance that reduces lateral or longitudinal resistance | 49 CFR 213.119(l) |
| Pull-apart prone condition | Rail below RNT at or near a joint, so tension may affect the joint fastenings | 49 CFR 213.119(l) |
Key takeaway: a sun kink is a stability failure of the whole track structure, which is why it is managed through a CWR plan rather than through rail defect rules.
How much hotter than the air does rail get?
Rail in direct sun typically runs 25 to 35 °F above the air temperature, and can slightly exceed 35 °F on a clear summer day. That range comes from FRA research. The 2015 FRA report on rail coatings (DOT/FRA/ORD-15/37) gives the example of a 100 °F day producing a rail temperature of about 135 °F with little cloud and no rain.
Network Rail in Great Britain gives a matching figure in metric. Its buckled-rail page says rails in direct sunlight "can be up to 20°C hotter than the air temperature", and that most of its network can operate with track temperatures up to 46 °C.
| Source | Rail temperature above air | Notes |
|---|---|---|
| FRA DOT/FRA/ORD-13/16 (2013) | 25–35 °F | Track buckling prevention report, Kish and Samavedam |
| FRA DOT/FRA/ORD-15/37 (2015) | Up to about 35 °F, slightly more | 100 °F air ≈ 135 °F rail, clear sky |
| Network Rail | Up to 20 °C (36 °F) | Rails recorded as high as 51 °C |
The practical point is that the air temperature on a weather forecast understates the load. Railroads that set slow orders by measured rail temperature, rather than by forecast air temperature, are working with the number that actually drives the force.
Sun exposure, shade, cloud, wind and track orientation all change how fast a given stretch of rail heats. For the wider weather picture, see how weather affects train safety and operations.
Long-term data show how strongly buckles follow the heat. A 2025 study of the Czech rail network by Nezval, Andrášik and Bíl (European Transport Research Review 17:28) found that 83.1% of buckled-rail incidents from 2002 to 2022 happened in June, July and August, and that daily air temperature had a strong influence on whether a buckle occurred. Broken rails showed the opposite pattern, with 78.3% in November to March, which is the cold-weather side of the same neutral-temperature problem.
How much force builds up in hot CWR?
Thermal force is P = E × A × α × (T − RNT), which for 136 lb rail works out to about 2,600 lb per °F above neutral, per rail. Steel expands about 0.0000065 in/in/°F. In a jointed rail the expansion closes the gaps, but in CWR the anchors and ties stop the rail moving, so the expansion turns into force.
The constants below are the values in FRA training Module 2, Track Buckling Prevention and Improved Rail Stress Management (2023), recomputed.
| Rail section | Cross-section area (in²) | E × A × α (lb per °F, per rail) |
|---|---|---|
| 136 lb | 13.35 | ≈ 2,600 |
| 115 lb | ≈ 11.2 | ≈ 2,185 |
E = 30,000,000 psi for rail steel; α = 0.0000065 in/in/°F.
Worked example: the same hot afternoon, two different rails
A 136 lb CWR string is laid at an RNT of 100 °F, inside a typical preferred laying range of 90–110 °F. On a 100 °F afternoon in clear sun the rail reaches about 135 °F.
| Case | RNT | Rail temp | T − RNT | Force per rail | Force in both rails |
|---|---|---|---|---|---|
| As laid | 100 °F | 135 °F | 35 °F | 35 × 2,600 ≈ 91,000 lb | ≈ 182,000 lb |
| After RNT has drifted down | 75 °F | 135 °F | 60 °F | 60 × 2,600 ≈ 156,000 lb | ≈ 312,000 lb |
The rail and the weather are identical in both rows. The difference is 25 °F of lost neutral temperature, which adds about 65,000 lb of compression to each rail. In stress terms that is E × α ≈ 195 psi per °F, so 60 °F above neutral puts about 11,700 psi of longitudinal compression in the steel.
Whether that force produces a sun kink depends on the track's resistance: tie and ballast friction, shoulder ballast, fastener and anchor restraint, curvature and any existing misalignment. FRA's CWR-SAFE model, described in the 2013 report, is built on exactly those inputs. A track with a full ballast shoulder and fully anchored ties can carry a force that would buckle freshly surfaced track on a curve.
Key takeaway: the RNT, not the thermometer alone, decides how much force a hot rail is carrying.
Why does rail neutral temperature drift down?
RNT is set when the rail is laid or destressed, then moves with every repair, creep and maintenance cycle, usually downward. FRA research found that RNT is typically 90–115 °F at installation but can fall to 60 °F or lower in service, and that there are no convenient ways to measure RNT in track.
Common causes, all named in FRA research and TSB investigations:
- Rail creep. Traffic and temperature cycles push rail along the ties. Rail that moves into a section raises its compression and lowers its effective RNT there.
- Ineffective anchors or fasteners that let the rail slide.
- Repairs at low temperature. When a broken rail or defect is cut out in cold weather and a plug is welded in without adjusting for temperature, extra steel stays in the track.
- Curve movement and track work such as tie renewal, surfacing and undercutting, which also cut lateral resistance until the ballast reconsolidates.
The Transportation Safety Board of Canada's investigation R25W0047 shows the pattern.
On June 28, 2025, 13 loaded grain cars derailed near Devlin, Ontario. Investigators found rail anchors that had moved up to 4 in. on the main track and up to 6 in. on the adjacent siding, all in the direction of heavy traffic.
The TSB treated that creep as a sign of compressive stress and said a thermal misalignment (track buckle) may have occurred before the derailment. It was a limited-scope investigation with no formal findings, but the TSB's safety communication to Transport Canada asked for a review of how rail on that subdivision is destressed and secured.
Worked example: how much rail to remove when destressing
Destressing restores RNT by cutting the rail, unanchoring it so it can move freely, removing the excess length and welding it back at the target condition. The length change follows ΔL = α × L × ΔT.
Suppose a 400 ft section has an estimated RNT of 70 °F and the plan's target is 100 °F.
- ΔT = 100 − 70 = 30 °F
- L = 400 ft × 12 = 4,800 in.
- ΔL = 0.0000065 × 4,800 × 30 ≈ 0.94 in.
If the freed rail is at 100 °F when it is closed up, about 15/16 in. of excess rail has to be cut out of that 400 ft section for the RNT to return to 100 °F. Work done at a colder rail temperature also needs a rail heater or hydraulic tensor to make up the difference before welding. This is an illustration of the physics. The actual method, adjustment length and records are whatever the owner's CWR plan prescribes under 213.119(a) and (d).
What do 49 CFR 213.119 and 213.7 require for buckling?
Every owner of CWR track must have, file with FRA, and follow a written CWR plan, and only persons qualified under 213.7(c) may inspect CWR or supervise its installation, adjustment and maintenance. Federal law does not set a universal buckling temperature. It requires each owner to set its own limits and procedures.
Under 49 CFR 213.118, the plan is filed with FRA's Associate Administrator for Railroad Safety, and FRA has 30 days to review it. Under 49 CFR 213.119, it must contain:
| 213.119 paragraph | What the CWR plan must cover | Link to sun kinks |
|---|---|---|
| (a) | A desired rail installation temperature range for the geographic area, and destressing procedures that achieve it | Sets the starting RNT |
| (b) | Rail anchoring or fastening that limits rail and tie movement, with attention to bridges and approaches | Resists creep |
| (c) | Joint installation and maintenance | Joints are pull-apart and buckle points |
| (d) | Keeping rail within the range when CWR is cut for repairs, in-track welding, or adjustment of tight track, buckles and pull-aparts | Stops repairs from lowering RNT |
| (e) | Monitoring curved CWR for inward shifts caused by disturbed track | Catches early misalignment |
| (f) | Train speed on CWR when track is disturbed, or when rail temperature versus neutral temperature creates buckling-prone conditions | The legal basis for heat slow orders |
| (g) | When physical track inspections are made to find buckling-prone and pull-apart prone conditions, and the remedial actions | Heat patrols |
| (h) | Joint inspection procedures and frequencies | CWR joint inspections |
| (i) | A comprehensive training program, with annual retraining | Who is competent |
| (j) | Records of installation temperature, location and date | Evidence of RNT |
| (k) | Procedures kept in one manual and available at job sites | Field use |
Paragraph (g) is specific about what the inspections must target: locations where tight or kinky rail is likely, and locations where track work that disturbed the roadbed or ballast has recently been done.
49 CFR 213.7(c) then sets who may do this work. A person who inspects CWR track, or supervises its installation, adjustment and maintenance under the owner's written procedures, must hold a current 213.7(a) or (b) designation, have completed a comprehensive CWR training course, have demonstrated competence in the procedures, be authorized by the owner, and have a recorded examination.
The rules sit beside the rest of Part 213; see the overview of FRA track safety standards. Rail transit agencies may also answer to a state safety oversight agency (in California, the CPUC), so confirm which rules your property follows.
Key takeaway: if a heat restriction, a destressing job or a heat patrol isn't written into the CWR plan, it isn't the owner's procedure under 213.119.
When do railroads slow trains for heat?
Heat slow orders are set by each railroad's own procedure, usually triggered by measured rail temperature; there is no single federal number. 213.119(f) requires the plan to govern train speed when rail temperature relative to neutral creates buckling-prone conditions, but leaves the values to the owner.
Published examples show how much the procedures vary:
| Railroad | Published trigger | Action | Source and caveat |
|---|---|---|---|
| Amtrak (Northeast Corridor and other lines) | Rail temperature about 131 °F | Maximum 100 mph | Amtrak extreme-heat fact sheet (2016), as reported in news coverage; coverage also notes the limits protect overhead catenary, not only track |
| Amtrak | Rail temperature 140 °F | Maximum 80 mph | Same fact sheet |
| Network Rail (Great Britain) | Thresholds by location, from remote sensors | Local speed restrictions | Network Rail, June 2025 |
These are one operator's procedures for its own track, rail sections and neutral temperatures. They are not a buckling threshold for anyone else's railroad, and a freight short line with lighter rail, a lower RNT or recently surfaced track may need restrictions at much lower temperatures.
FRA's Safety Advisory 2012-03 followed four heat-related derailments in June and July 2012, one of which killed two people when cars fell from a bridge in Northbrook, Illinois. Its recommendations included applying "heat-restriction slow orders at necessary locations, with consideration of populated areas", and slow orders at disturbed track until enough time and tonnage have passed to restore stability. The advisory also stressed bridges, grade crossings and turnouts, where FRA said rail conditions are "considerably tighter".
Network Rail's 2025 rollout shows the direction of travel. Its sensors take readings every 15 minutes, send alerts when rail reaches set thresholds, and let the railway restrict speed only where the rail is actually hot instead of across wide areas.
How do railroads prevent sun kinks?
Prevention comes down to three things: keep RNT high and known, keep lateral resistance high, and control speed when either is compromised. Each measure below maps to a part of the CWR plan.
| Measure | What it does | CWR plan link |
|---|---|---|
| Lay and destress at the target range | Sets RNT high enough for local summer rail temperatures | 213.119(a) |
| Adjust rail length during repairs | Stops cold-weather plugs from lowering RNT | 213.119(d) |
| Full anchoring and sound fasteners | Resists creep and keeps RNT from migrating | 213.119(b) |
| Full ballast section and shoulders | Provides lateral resistance; ballast condition matters | 213.119(f) |
| Slow orders on disturbed track | Limits dynamic force until ballast reconsolidates | 213.119(f)(1) |
| Heat patrols at tight-rail and work locations | Finds kinky rail before a train does | 213.119(g) |
| Curve monitoring | Detects inward shifts | 213.119(e) |
| Rail temperature monitoring | Ties restrictions to actual rail temperature | Supports 213.119(f)(2) |
| Reflective (white) rail coating | Cuts heat input on selected stretches | Supplementary |
Reflective coating is the newest visible measure in North America. Union Pacific reported on June 8, 2026 that it sprays white paint on both sides of the rail in selected high-heat areas from a hi-rail truck. Its chief safety officer said crews had seen "about a 20-degree drop in the rail temperature"; the railroad didn't give the unit, though Fahrenheit is the likely reading. Union Pacific described the coating as an addition to anchors, fasteners, maintenance and inspection, not a replacement.
FRA has studied the idea before. Report DOT/FRA/ORD-15/37 found that some low-solar-absorptivity coatings can significantly reduce rail temperature compared with uncoated rail, but said more work was needed on durability in service. Network Rail also paints parts of its rail white.
Where does NDT fit into sun kink prevention?
Rail flaw testing and weld inspection don't measure buckling risk directly, but they cover the welds, joints and repairs that every buckle and pull-apart cycle creates. A CWR program that controls RNT still depends on sound field welds and joint bars, because those are where the rail is cut, heated and rejoined.
Where NDT work meets CWR management:
- Field welds made during destressing and repairs. Every cut-and-weld under 213.119(d) leaves a field weld, usually thermite. A TTCI research digest (TD-00-013, 2000) found that test cars miss most of the rail base, so field welds benefit from hand ultrasonic testing of the base. See aluminothermic weld inspection and flash-butt weld inspection.
- Repair of a buckle. A buckled section is often cut out and replaced. The new rail and its two welds should be tested before the section is left in service, and the plug rail's internal test status checked under 213.237(c).
- Joints in CWR. A joint in CWR is where a pull-apart starts in cold weather and where tension cycles crack bars and bolt holes. 213.119(h) sets on-foot joint inspections; ultrasonic testing of rail joints can find bolt hole cracks the eye can't.
- Bridges and approaches. FRA's 2012 advisory singled out track on or near bridges, where rail is tighter and a derailment is worse. Bridge approach rail inspection covers the testing side.
- Records. A test report that gives mileposts, rail section, weld locations and the rail temperature at the time of testing gives the track owner one more data point next to the 213.119(j) installation records. Rail temperature on the report is good practice, not a 213.237 requirement.
Internal testing itself stays on its own schedule under 49 CFR 213.237, and defects found are handled under 213.113. A sun kink is not one of the 213.113 rail defects.
What rail flaw testing can't tell you about buckling risk
Ultrasonic rail flaw testing finds internal defects; it does not measure rail neutral temperature, longitudinal force or lateral resistance. A rail can be free of flaws and still be 60 °F below its intended neutral temperature.
| Question | Can rail flaw UT answer it? | What does |
|---|---|---|
| Is there a transverse defect or bad weld? | Yes, within the limits of a valid search | 213.237 internal testing |
| What is the RNT of this string? | No | Installation and adjustment records, anchor creep marks, and rail stress measurement where an owner uses it |
| Is this curve losing lateral resistance? | No | Visual inspection, geometry measurement, curve monitoring under 213.119(e) |
| Has this location been disturbed recently? | No | Maintenance records and the 213.119(g) inspection plan |
| Will this track buckle at 135 °F? | No | The owner's CWR plan limits, engineering judgment and models such as CWR-SAFE |
Other limits are worth stating plainly. FRA research says there are no convenient RNT measurements, so most owners rely on installation records and creep monitoring, and those records degrade every time the rail is cut. Surface damage such as shelling can also mask internal defects from the ultrasonic beam, which is a separate limit on what any test proves.
How EMA Industries applies this
EMA Industries performs ultrasonic rail flaw testing and rail weld inspection for transit agencies, short lines and industrial track owners, from Southern California and mobilizing nationwide. On CWR territory, the useful work is testing the field welds and plug rail that destressing and buckle repairs create, and reporting weld locations and rail conditions in a form the owner can file beside its CWR records. To plan weld or rail flaw testing around a destressing or heat-repair program, see our ultrasonic rail testing services or contact EMA Industries.
Related reading: Continuous welded rail: how CWR works and how it is inspected · Ultrasonic rail flaw detection · Railway track inspections
Frequently asked questions
What causes rail tracks to buckle in hot weather?
Rail tracks buckle when compressive force from thermal expansion exceeds the track's lateral resistance. The force depends on how far rail temperature is above the rail neutral temperature, at roughly 2,600 lb per °F in 136 lb rail. Weak ballast, missing anchors, recent track work, curves and train forces all lower the temperature at which buckling starts.
How do welded rails avoid buckling if they have no expansion gaps?
CWR is laid or destressed at a neutral temperature high enough that summer rail temperatures stay within what the track structure can resist. Anchors, fasteners and a full ballast section hold the rail in place so the expansion becomes stored force instead of movement. Under 49 CFR 213.119, the owner's CWR plan sets the installation range, repair procedures, inspections and heat speed limits.
At what temperature does railroad track buckle?
No single temperature applies. Buckling depends on the difference between rail temperature and the neutral temperature of that string, and on the track's lateral resistance. Track with a well-maintained 100 °F neutral temperature may tolerate rail temperatures that would buckle recently surfaced track with a neutral temperature of 70 °F.
What is a heat slow order?
A heat slow order is a temporary speed restriction a railroad applies when rail temperature, or recent track disturbance, makes buckling more likely. 49 CFR 213.119(f) requires the CWR plan to govern speed in those conditions, but each railroad sets its own triggers. Amtrak's published procedure, for example, capped speed at 100 mph at about 131 °F rail temperature and 80 mph at 140 °F.
Is a sun kink the same as a rail defect?
No. A sun kink is a track stability failure, a lateral misalignment of sound rail. Rail defects such as transverse fissures and defective welds are internal or surface flaws listed in 49 CFR 213.113 and found by internal rail testing under 213.237.
Sources
- 49 CFR 213.119, Continuous welded rail (CWR); plan contents, eCFR (read on Cornell LII, October 2026)
- 49 CFR 213.118, CWR; plan review and approval, eCFR
- 49 CFR 213.7, Designation of qualified persons, eCFR
- 49 CFR 213.237, Inspection of rail, eCFR
- FRA Safety Advisory 2012-03, Buckling-Prone Conditions in Continuous Welded Rail Track, 77 FR 41881, July 16, 2012
- Kish and Samavedam, Track Buckling Prevention: Theory, Safety Concepts, and Applications, DOT/FRA/ORD-13/16, FRA, March 2013
- FRA, Track Buckling Prevention and Improved Rail Stress Management (Module 2), July 2023
- Ritter and Al-Nazer, Quantification of the Effectiveness of Low Solar Absorptivity Coatings for Reducing Rail Temperature, DOT/FRA/ORD-15/37, FRA, October 2015
- TSB rail investigation report R25W0047, Transportation Safety Board of Canada, February 2026
- Network Rail, Buckled rail and Smart sensors, fewer delays: tackling heat on the tracks, June 13, 2025
- Union Pacific, Tracking Rail Heat, June 8, 2026
- Nezval, Andrášik and Bíl, Identification of factors contributing to broken and buckled rails: insights from long-term data, European Transport Research Review 17:28, 2025
- Amtrak, Extreme heat fact sheet, July 22, 2016 (no longer at its original Amtrak media URL)