THE US Transportation Research Board (TRB) awarded Protran Technology and QinetiQ North America a grant in November 2013 to further develop and test QinetiQ's Intelligence Rail Integrity System (Iris) which monitors changes in rail neutral temperature (RNT) to determine track deterioration. Protran Technology partnered with Maryland Transit Administration (MTA) to provide the first real-time network of remote sensors installed on continuous welded rail (CWR).

"MTA has been researching technology that would continuously monitor track changes in rail temperature, stress and rail neutral temperature for some time," explains Mr Jason Lurz, MTA's director of rail operations. "Protran Technology and QinetiQ have been selected as part of this demonstration project to expand our track monitoring capabilities."

Thermal loads develop in rail due to daily heat/cooling cycles. When a rail heats it expands and is in compression, but when it cools the rail contracts and is in tension. Somewhere in between these two points, the rail is neutral and there is no compression or tension. Rail temperature with no force due to thermal loading is the rail neutral temperature (RNT). Extreme fluctuations in rail temperature from RNT can cause buckling and breaks and lead to derailments if undetected.

Iris continuously monitors changes in rail temperature, longitudinal rail force and RNT. If rail buckling or breaks are detected, Iris alerts the location to designated personnel by text or e-mail. In these times of dwindling resources, Iris can be used to prioritise maintenance and establish the performance trends of different track structures.

ProtranDuring an initial meeting at MTA's Heavy Rail Facility in January, the committee recommended that locations currently requiring heat watches when rail or ambient temperatures reach critical levels, should continue but use Iris to target known problem areas.

The committee agreed to develop recommended installation and operations instructions. For example, whenever possible the rail should be set to the RNT at the time Iris is installed. Iris should be installed with all new rail installation especially in curves and at abutments. Warning notification ranges must be identified as they will differ between geographical locations. The committee recommended that maintenance should be undertaken in-house by the operator because of limited track time and RWP requirements.

The committee also recommended that the trains should be notified once the change in delta is identified and tested. As a minimum, notification must be transmitted to the control centre.

Comprehensive training is critical. Two main issues discussed were the need to change track inspection procedures and training on railways that install Iris and to pinpoint who exactly will develop and train staff with the new technology.

Iris provides a network of remote sensors installed on CWR and offers the following four primary functions:

• continuous monitoring for broken or buckled rail

• emergency notification to the railway

• web-accessible GIS user interface, and

• assists in rail de-stressing operations.

Iris has three main components: a sensor unit (SU), a collector unit (CU) and a configuration status unit (CSU). The SU continuously monitors rail temperature, RNT and rail forces for hazardous rail conditions. When detected, rail failures trigger an alert system that sends an emergency text and/or e-mail. The information collected populates a database that offers a simple web-accessible user interface which can be used for data analysis and presentation. Each CU can receive data from up to 13 individual SUs and transmit the data to the web-accessible database.

The CSU is a hand-held device that communicates with individual sensors and is used for various functions, including configuration during sensor installation and assistance in field-weld maintenance operations.

Once the equipment is properly installed it immediately starts recording data and looks for changes in temperature. Rail temperature is much greater than RNT prior to buckling and causes excessive compressive loads. Most of the compressive force in the rail is released after a buckling event.

By definition, the RNT is the rail temperature when the force in the rail is zero. After a rail break, the RNT approaches the ambient rail temperature as typical CWR condition no longer exists and the longitudinal force in the rail is reduced. Prior to breaks however, rail temperature is much lower than RNT which causes excessive tensile loads. Force at the rail is released after a break and, again, approaches zero. After the rail break, the RNT approaches the ambient rail temperature, as typical CWR condition no longer exists. By monitoring dramatic changes in RNT, a railway can predict potential rail breaks and lateral buckling to help prevent derailments.

Although Iris can be used for any track application, the optimum location for Iris is in curves and at abutments in non-direct fixation track. The sensors should be installed 183 to 244m apart and the sensor units can be linked to the control centres to cover a 40km area of "dark" track.

Installation of Iris is relatively straightforward. To install the strain gauges for the rail-mounted sensor units, the rail requires light grinding to remove the external layer of corroded steel, followed by fine sanding to remove any depressions in the rail caused by the grinding. After grinding and sanding, the rail needs to be wiped with a clean rag to clear the area of rust grindings.

The strain gauges are covered with petroleum-based M-Coat to make them weather proof. The SUs are magnetically mounted on top of the strain gauges. The battery-powered sensors communicate wirelessly for continuous data collection. An onboard processor calculates RNT and monitors for possible track failure.

The CU has two power configurations, either direct through 120/220V ac (wiring at emergency trip stations is a good option) or solar power. The control units should be installed approximately 3.6m high. The CUs collect all the data from the SUs and send the data to a web-database. They are versatile with power connection and data link access and use available ac or dc, cellular, and RR-LAN.

Cellular modem

The CU transmits data via cellular modem (GSM, CDMA, 3G, LTE) and its enclosure includes a mounting bracket for poles up to 7.6cm in diameter. After installing the CU hardware, three connections must be made: cellular antenna, Iris link antenna and CU power supply. The cellular antenna has a threaded connection which can be secured to the port located on the bottom surface of the CU enclosure. The Iris link antenna is the same and the threaded port is also located on the bottom of the CU enclosure. The signal from the cellular antenna is stronger than a typical mobile phone and has adequate signal strength.

The CSU consists of a small dongle that connects to a standard laptop computer. The CSU hardware and pre-loaded firmware are accompanied by operation software, and the CSU was used for configuration during installation. The CSU can be taken to the site of the SU to download raw data, graph it, and analyse data, and is compatible with multiple operating systems including Windows XP/7/8 (USB dongle) and android.

After the sensors are installed, warning and alert levels must be determined. Each region will differ but typically the summer alarm temperature should be set around 15 to 20oC above the area's high ambient temperature. For example, the likely setting for the warning and alarms in the Philadelphia/Washington DC area are as follows. The summer warning rail temperature is set at 54oC and the summer alarm rail temperature is set at 60oC. The winter warning temperature is set at -7oC and the winter alarm temperature is set at -12oC.

Normally, the sensors are configured to populate the Iris database and to send text alerts to the appropriate individuals, however, alarms can be sent directly to the trains approaching the sensors, but Protran's Protracker train unit must be installed on each train for this to work.

There are two other ways to alert trains prior to them entering the area of concern. One is to install wayside signals in known problem areas at a distance sufficient to allow the trains to stop or slow down. The second, less costly and more efficient application, is to install portable Protracker units in areas which Iris identifies with a high potential for rail buckling or breaks. The units are attached magnetically to the rail web and, following maintenance, they can be removed and installed elsewhere.

Although we are not yet in an age where technology renders human interaction irrelevant we are at a point where technology can help to protect employees better in dangerous situations. Protracker has been accomplishing this since 2008, while Iris is the next generation of technology to avert costly derailments.