TFR: New, local, ultrasonic broken-rail detection

[Steve Appleton]

From Business Day, 9 Novemeber 2011:

Keeping Transnet on the straight and narrow
Ultrasonic device saves parastatal millions of rand by preventing derailments, writes Sarah Wild

Published: 2011/11/09 10:53:08 AM

LOCALLY developed technology is saving TFR millions of rand by helping the parastatal minimise derailments. Driven by Armscor’s Institute for Maritime Technology (IMT) and the Council for Scientific and Industrial Research (CSIR), the ultrasonic transducer technology runs on Transnet’s (TFR's) coal line to Richards Bay and the Orex line (Saldanha-Sishen). It uses sound waves to detect faults that can cause derailments.

The technology "detects breaks in the line", says Wiehan le Roux of Transnet’s technology management division. "It’s on our coal export line and our iron line. It’s not on the entire line, only part of it." One detector "has already picked up three faults in the line. It has saved Transnet millions of rand," he says. Aside from the costs of derailments, an inoperative rail line weighs heavily on the country’s economy because the two lines transport coal and iron ore, two major exports.

The CSIR has been responsible for the development of the technology, while the Armscor institute drives the process.

According to Francois Burger, programme manager for the ultrasonic broken-rail detector, "IMT’s capability in this regard stems from its expertise in the area of underwater acoustics and signal processing, as applied for the South African Navy." The institute is a division of Armscor Defence Institutes. The CSIR was subcontracted to develop the transducer system.

"It’s a real-time and remote system," says Jeremy Wallis, manager of sensor science and technology at the CSIR’s materials science and manufacturing unit. "That’s really the value proposition to Transnet — someone can sit at headquarters and receive a message saying there appears to be a fault. Then you can stop a derailment."

The transducer emits a pulse, which resonates and creates a vibration on the rail. "It’s a bit like taking a hammer and whacking the rail. The (sound wave) dissipates as the sound propagates to the right and left." Two receivers are on either side of the transmitter, usually about 1km apart. If a receiver does not receive the signal, there may be a fault on the line. "Should the signal not arrive," Mr Burger says, "the receiver raises the alarm, which is an indication of a potential discontinuity in the rail. This is then physically inspected." All is solar powered, Mr Wallis says. "It sits out in the middle of nowhere and powers itself."

While this does seem like an ideal solution for maintaining railway lines, it does not come cheap. "A system consists of transmitter and receiver units, nominally spaced at a distance of 1km," Mr Burger says. "Hence, to cover, say, 100km, 101 units are required. "Equipment per site, including lightning and surge protection, communication link, and solar power supply, ranges between R30000 and R50000, and is dependent on the final specification and installation costs," he says.

However, there is demand for this technology, at home and abroad. Dr le Roux says Transnet is considering the business case for putting more transducer systems on the Orex line, and that it is going to tender.

Mr Burger says some overseas operators are testing this technology on their railway lines. However, it does not work on all railway systems. First, it has to be a continuous line. Heavy freight lines are welded together, so the sound wave can travel along it. But if it is a conventional railway line, with expansion joints, the pulses cannot jump the gap.

There are other problems as well, Mr Wallis says. "It comes down to the physics of how the sound propagates on the rail. A lot of frequencies get generated, and different intensities of frequencies. A lot of those die off quickly. After 100m, more of the frequencies are gone. "There are only certain frequencies that will actually go a few kilometres."

However, by March next year, the CSIR will unveil modelling technology that will map which transducer should be used on which type of railway line.

"It depends on the material and the geometry," he says. "The CSIR is developing technology so that this system can be used on any railway in the world…. We can work out which signals will propagate … and then we can decide which transducer design we’d use to generate that. Then you can pretty much pick up the system and go to any part of the world with a custom solution."