Background
Rail is an ever-expanding and increasingly important method of transporting goods and people. With the ever-expanding urban areas, existing railways often carry hazardous materials very close to urban areas. Train derailments, the added danger of hazardous material contaminants, and the risk of explosions can be very costly to society and fragile ecosystems, and may cause irreversible damage to the environment and aquatic life. Faulty railcar wheel carriages also greatly contribute to noise pollution in densely populated urban areas.
Contents
- Background
- Wheel Impact Phase Detector WIPD
- On Board Mote Sensing Technologies
- P waves and S waves
- References
Existing methods of wayside railcar wheel impact sensing are complicated and expensive to install and requires regular maintenance and calibration.
Seismic rail impact sensing are proving to yield much higher accuracy at substantially lower installation and maintenance cost. It finally provides a low cost inspection solution to scan railcar wheels for compliance in order to mitigate the risk of environmental damage caused by derailments.
Seismic wheel sensing works on the principle of seismic wave phase analysis, the sensors do not have to be installed on the rail track but can be installed within a convenient distance parallel to the track. The system can be installed as a permanent installation or through mobile deployable stations. These sensor stations can then be deployed strategically across major rail arteries.
Wheel Impact Phase Detector (WIPD)
The search for effective methods to detect in motion broken railcar wheels has delivered promising new technological advances with the arrival of WIPD or Wheel Impact phase detection. The phase sensing method uses two or more seismic sensors connected to the side of a rail track to capture seismic noise generated by railcar wheels. A wheel that exceeds a preset seismic noise threshold in amplitude, will trigger a wheel tracking algorithm that calculates seismic phase shift data related to the actively tracked wheel noise level, to determine the precise location, in real time, of the faulty wheel carriage while moving at full speed. With the sensors spaced from each other such that seismic phase time propagation delays between sensors, in relation to the signal source, can be measured to calculate the exact position of the faulty wheel in real time.
Knowing the precise location of the tracked wheel allows the system to isolate the railcar and capture the railcar and wheel carriage identification information.
Subsequently, a railcar log is made on a computer database with the railcar identification information and made available to control centers via ground or satellite links.
On-Board Mote Sensing Technologies
The promise of mote technology vibration sensors embedded on the railcar itself, could also be used to measure vibrations and possible wheel impact failures. This type of sensing however would not be able to distinguished between the source of the vibrations originating from either the rail track or wheel/bogie. There are many environmental factors that would make mote sensing highly subjective.
P-waves and S-waves
Seismic waves consist of primary and secondary waves in nature. We know from seismology that it is possible to measure the distance from an earthquake epicenter by measuring the phase time delay between the arrival of first, the primary body p-wave and then the arrival of the secondary slower surface s-wave. With distance this delay increases and is one of the commonly used methods to locate earthquake epicenters. While this method works well over long distances where the p-wave and s-waves are clearly separated, it becomes very difficult to differentiate between the (p) and (s) waves over distances smaller than a few meters as outlined below.
This p-wave to s-wave delay was put to test to see if it could be used to sense the real time position of a faulty wheel impact on a train on a track while moving at full speed. Test calculations has shown that the relative short distances of less than a few meters complicated the ability to clearly differentiated between the start of the p-wave and the subsequent arrival of the s-wave. Typical wheel impact pulse frequencies are below 2 kHz.The wave propagation delay time between the p-wave and s-wave over a distance of 1 meter is smaller than half the wavelength of the typical average frequency of the wheel impact impulse. This condition makes the method of measuring p-wave vs s-wave propagation delays virtually impossible since the s-wave will be superimposed on the p-wave before the ADC had the chance to even capture one full wavelength. Since seismic sensor calibration to KIPS are very critical it is important to capture at least a number of full wave cycles in order to establish whether the wheel exceeds allowed compression levels.
Thus a different method had to be discovered in order to guaranty consistent detection and locating faulty train wheels.