Handover and the geometry record

New track is not finished when the last fastening has been tightened. Before trains can run on it, someone must certify that it meets the standard, which means inspecting and measuring it against the design and the tolerances of Chapter 9. Then it must settle: the ballast consolidates under the first traffic, and the track’s geometry changes as it does, so it is measured again after a set tonnage and corrected. From then on, its geometry deteriorates slowly under traffic, and the record of measurements, starting with the as-built survey, shows how fast and predicts when it will need maintenance. A track whose record is kept well is maintained when it needs it; one without a record is maintained when something goes wrong.

This chapter covers the handover inspection, using the requirements of ARTC’s 2005 NSW Track Standards Construction, TCS 01, as an example; the remeasurement after initial traffic; the deterioration of geometry and how it is tracked; and the geometry record. It closes by bringing the book together.

The handover inspection

TCS 01 required a detailed inspection of new main line work, against its standards, before a line could be certified for general traffic.

Figure 1Handover inspection of new main line track, as described in ARTC TCS 01

• Avoid: Survey pointsall survey points checked against the design
• Avoid: Cant, cross level and gaugechecked throughout with a manual track gauge
• Avoid: Alignment on curveschecked over a fifth or more of every curve, and wherever it looks visually irregular
• Avoid: Surfacevisually checked; any apparent variation measured
• Do: ? Track recording runwhere possible, a recording vehicle run before handover, meeting the recording standard
• Avoid: Remeasure after settlementafter at least 200,000 tonnes of trailing load, measured again and corrected to the original standard

The inspection combines measurement at every point where it is cheap, such as gauge and cross level throughout, with sampling where measurement is slower, such as alignment on a fifth of each curve, and visual checks guided by experience, with any doubtful point measured. A recording vehicle’s run, where one can be arranged, gives a continuous record of the whole length, which catches faults between manual measurement points and becomes the first entry in the track’s recorded history.

Sidings

TCS 01 required the same detailed inspection before certifying a siding fit for use, and set a maintenance period of three months from the date of certification, during which any defects that developed were to be corrected to the original standard by the builder. During the maintenance period, for both main lines and sidings, the track was to be maintained above the network’s maintenance standard. The running case’s balloon loop, being a siding, had a three-month maintenance period after handover; its junction turnout, part of the main line, was subject to the main line requirements.

Settlement and the second measurement

New ballast is always loose. Under the first traffic, it consolidates, and the track settles, unevenly, by several millimetres or more. Geometry that was within tolerance on the day of handover can drift out of it in the first weeks. TCS 01 required new main line track to be measured again after at least 200,000 tonnes of trailing load and any defects corrected to the original standard, which catches the settlement while the builder is still responsible.

Figure 2How long the junction takes to see 200,000 tonnes of grain traffic

weeks = 200,000 ÷ weekly trailing tonnage over the track

Loaded wagons per week

4 × 40 × 92 = 14,720 t

Empty wagons per week

4 × 40 × 22 = 3,520 t

1. weekly trailing tonnage ≈ 14,720 + 3,520 = 18,240 t
2. weeks to 200,000 t ≈ 200,000 ÷ 18,240 ≈ 11 weeks on the diverging route
3. the main route, carrying all main line traffic, reaches 200,000 t much sooner

The second measurement for the junction’s diverging route therefore falls about eleven weeks after opening on the grain traffic alone; the main route, which carries every train on the main line, reaches the threshold within days or weeks depending on the line’s traffic. The cooperative’s contract with its track builder set out the remeasurement and correction for each part of the work, with the infrastructure manager’s inspector to witness it.

How geometry deteriorates

After the initial settlement, track geometry deteriorates more slowly, but steadily and without stopping. Each passing wheel compresses the ballast a little more, and unevenly; faults that start small, at a weld, a soft spot or a turnout, grow faster than the track around them, because the dynamic loads there are larger. Railways track deterioration by measuring geometry at intervals, often by recording vehicle, and expressing it as the growth of faults, or of a combined track quality index, per million tonnes of traffic.

Figure 3Illustrative growth of a surface fault on the loading straight with traffic

Surface departure

Surface departure on a 20 m chord (mm)Million gross tonnes since handover

Illustrative growth of a surface fault on the loading straight with traffic data

Million gross tonnes since handover	Surface departure
0	3.0
1	4.5
2	6.0
3	7.5
4	9.0
5	10.5
6	12.0

The record turns into a forecast. If a fault starts at 3 mm and grows at about 1.5 mm per million gross tonnes, it reaches a 12 mm planning limit after about 6 million tonnes, and at the cooperative’s traffic of about a million gross tonnes a year over the loop, that is about six years. Track built to tight tolerances starts lower on the line and takes longer to reach the limit; track on a poor formation grows faster. Maintenance can then be planned in advance, combining work on several faults in one possession, rather than reacting to each as it appears.

The geometry record

The geometry record for a length of track holds its design geometry, the as-built measurements and corrections, the handover inspection and recording run, the remeasurement after settlement, and every later measurement, with the maintenance done between them. It is referenced to the track’s kilometrage and to fixed survey marks, so that measurements taken years apart can be compared point by point. Railway Clearances by the Numbers explains why the same record matters for clearances: a structure’s margin depends on where the track is, and only the record shows whether it has moved.

Maintenance and the design geometry

Tamping machines restore track to a target geometry, and the target must be the design, not whatever the track has drifted to. Machines that measure and correct relative to the existing track, without a design target, smooth out short faults but can let curves, transitions and grades drift over many cycles. Providing the design geometry, curve radii, transition points, cant and levels, to the maintenance machines keeps the track true to its design for its whole life, and is one of the most valuable things a project can hand over.

Who certifies

Certifying track fit for traffic is a formal act with legal weight: the person who signs is stating that the track meets the standard and is safe for the trains that will use it. Certification is done by competent people authorised by the infrastructure manager, on the basis of the inspection and measurements, never on the builder’s word alone. The cooperative’s track builder prepared the measurements and a statement that the work met the design and standard; the infrastructure manager’s inspector reviewed them, walked the work, witnessed check measurements, and certified the junction for main line use. The loop, on the cooperative’s land, was certified under the arrangements agreed between the cooperative and the infrastructure manager for the siding’s connection to the network.

Defects in the maintenance period

The maintenance period exists because some defects only appear under traffic: a weld that dips, a sleeper that was not packed, a soft spot in the formation, a fastening that loosens. During the period, the builder inspects the track regularly and corrects any defect to the original standard, not merely to the maintenance limit. The cooperative’s contract required weekly inspections of the loop during its three-month maintenance period, with a recording of gauge, cross level and twist at the loading straight and the bifurcation turnout every fortnight, so that settlement there would be caught early.

Recording vehicles and track quality

Track recording vehicles measure gauge, cross level, twist, alignment and surface continuously at speed, along with the track’s curvature and cant, and plot them against kilometrage with any exceedance of limits flagged. Many railways also combine the measurements into a single track quality index for each length of track, often based on the standard deviation of the surface and alignment over a set length, so that a whole network’s condition can be compared and trended at a glance. A newly built track should start with an excellent index; a sudden rise in a later recording points to a fault developing, and a steady rise shows the normal deterioration that maintenance plans for.

The handover package

The record handed to the infrastructure manager is as much a deliverable as the track. For the cooperative’s project, it comprised the design geometry in the format the infrastructure manager’s systems used, the as-built survey of alignment and levels against the design, the measurements at handover of gauge, cross level, twist, versines and surface, the recording run’s output, the record of defects found and corrected, the remeasurement after initial traffic, the turnout and catchpoint details, the material certificates for rail, sleepers and ballast, and the formation test results. Each was referenced to kilometrage and to survey marks that the infrastructure manager could find again.

When the traffic changes

The geometry record also shows when the track is being asked to do something new. More trains, heavier axle loads or higher speeds change the rate of deterioration, the cant that suits the curves, and the tolerances that the track must be maintained to. When the cooperative later considered longer trains and a second user, the loop’s record showed how its geometry had held under the first years’ traffic, which gave the engineers evidence for how much more it could take before its formation or components needed upgrading.

Maintenance limits and speed restrictions

As geometry deteriorates, it passes through levels that call for different responses. Below the planning limit, the track is simply watched. At the planning limit, maintenance is scheduled. At the maintenance limit, it must be done within a set time. Beyond that, the network’s operating limits for each speed band require speed restrictions until the fault is fixed, and the worst faults close the track. TCS 01 referred to the network’s maintenance and base operating standards for those limits, which are not quoted here. A good record keeps track in the first two levels, where work is planned and cheap, and out of the last two, where it is urgent and disruptive.

Bringing it together

This book began with the track as a structure, and the reminder that geometry lasts only as long as the structure beneath holds it. It described curves by their radius and their versines, chose cant for the speeds of the trains on a curve, and turned cant and the allowed deficiency into the window of speeds a curve suits. It sized transitions to ease trains into curves, grades and vertical curves to carry them over the country, and calculated the resistance a heavy train meets on a loading grade. It placed turnouts and catchpoints, laid out a balloon loop and its loading bin, built the track to tolerance and handed it over, and finished with the record that predicts its maintenance.

The numbers point to a few principles. Every curve is a compromise among the trains that use it, and when the traffic changes the compromise should be reviewed. Geometry changes gradually or not at all: curvature, cant and grade are eased in, never stepped. Twist derails, so it is designed carefully and measured closely. Tolerances are tightest where speed, tonnage or a process like loading demand it. And the record, from the design through every measurement, is what keeps the track true to its design for the decades it must serve.

A failure: the missing design

This is an illustration. A siding built decades earlier had no surviving design drawings; its curves and grades were known only from the track itself. Over many tamping cycles, with machines correcting only relative to the existing track, its main curve had drifted into a series of slightly different radii, and its grade through a loading point had developed a gentle hump. Nobody could say what the track should be, so nobody could correct it properly.

A survey and a new design geometry were prepared, the track was realigned and relevelled to it in one programme, and the design was given to the maintenance contractor as the target for every future tamping. The siding’s geometry has held since. The cooperative made sure its own design, as-built survey and records were handed over in a form the infrastructure manager’s maintenance systems could use.

Exercise: planning maintenance from the record

On a length of track, a twist fault measured 3.5 mm over 3 m at handover and 5.0 mm a year later, after 1.2 million gross tonnes. Assuming it continues to grow at the same rate, how long until it reaches a planning limit of 9 mm? What would you check first?

The fault grew 1.5 mm in 1.2 million tonnes, about 1.25 mm per million gross tonnes. To grow from 5.0 to 9.0 mm, a further 4.0 mm, takes about 4.0 ÷ 1.25 ≈ 3.2 million tonnes, about 2.7 years at 1.2 million tonnes a year.

That forecast lets the correction be planned, but a twist fault growing steadily at one place usually has a cause beneath it, such as a soft formation, a failed sleeper or poor drainage, so the first check is the track’s support at that point. This is an illustration; correcting the cause, not just the geometry, is what stops the fault returning.