Rail vehicle drivers need information telling them when to move and where to stop. For the majority of railways this is achieved using lineside signals. Lineside signals consist of visual devices (coloured lights, mechanical devices, signs) positioned at pre-determined locations at the lineside for the driver to see and react to. In the early days of railways, lineside signals were mainly mechanically-operated semaphore signals and, although many examples can still be found on the network, all modern systems use Colour Light Signals (CLS). While these may look like road traffic lights, CLS actually predated them.
With road vehicles, drivers are ‘driving on sight’ and can change direction or stop in the distance they can see. Railways are very different, and a rail vehicle will obviously go wherever the rails and points take it. A rail vehicle cannot stop quickly (one travelling at 200km/h will have a typical braking distance of over 2.5km), so except for very low speed railways, driving a rail vehicle is not a ‘drive on sight’ operation.
Railway tracks are divided into ‘block’ sections, with normally only one rail vehicle allowed in any one section at a time. The driver of a rail vehicle is likely not to be able to see all sections ahead, so a ‘block’ signal is positioned at the start of each section to indicate if the section is clear of another rail vehicle. An approaching rail vehicle must stop at the signal unless the signal displays an indication to proceed.
Signal visibility
Because a rail vehicle takes a considerable distance to stop, a signal displaying ‘stop’ may not be visible to the driver at the location where they need to start slowing down. To ensure adequate advance information is provided to the driver so they can brake appropriately, one or more ‘distant’ signals are provided on the approach to each stop indication.
Good driver signal visibility is therefore essential, and the driver must know the meaning of the information displayed by lineside signals and how to respond. Drivers need to be able to read the signal, observe the information displayed, understand what it means, and then drive accordingly.
Signals must be clearly applicable to the driver of an approaching rail vehicle, be readily distinguishable from adjacent signals and signals further ahead on the line. All the illuminated lights that comprise the aspect of the CLS must be displayed for a sufficient length of time for the driver to observe the aspect, taking into account the highest approach speed. The signal must also be visible under all conditions and at all times of day and night, including where sunlight or nearby artificial lighting might make readability difficult. The information conveyed by the signal aspect must also be unambiguous, so that the driver is certain of the action they must take.
LED technology, first introduced to UK rail by Unipart Dorman, has enabled signal heads to be much more compact and lighter than traditional filament lamp (bulb) signals. While the visibility of a signal from a distance is important, a driver also needs to be able to see the aspect when stopped close to the signal. Signals must be highly reliable, for both safety and rail vehicle performance reasons, and the use of LEDs in signals has made the sudden failure of a signal very infrequent compared with traditional filament lamps.
Phantom aspects
Visibility of signals can be problematic in locations where, at some times of the day and year, the sun is either shining directly into the signal, shining directly behind it, or reflected by a nearby building. These are known as ‘phantom aspects’ where the signal aspect appears to be lit but is not. A red (stop) signal aspect may also look like a yellow (proceed) aspect with disastrous consequences. Signals facing East (at sunrise) or West (at sunset) are naturally the most vulnerable to phantom aspects, especially those on falling gradients. Modern rail vehicle windows can be more angled, which also increases the risk of phantom aspects from reflections below the horizon. Phantom aspects are not a new problem, and the issue applies to all types of signals irrespective of their manufacturer.
A national Network Rail audit of the alignment of Unipart Dorman signals in use at the time identified several issues that could affect the ability of a driver to reliably read, interpret, and act upon the information presented at the signal. An investigation concluded that there had been issues when installing, inspecting, testing, and commissioning the signals. Many of the signals were not aligned optimally, some not even pointing in the direction of the driver.
This was partly due to the variety of the traditional signal structures in use, and that it was impractical to verify the alignment from the signal head on non-accessible structures. It was also identified that the training and information for people involved in the process had not been sufficient and had not considered the manufacturer’s requirements. Traditional ‘fixes’ for the alignment of signals were also found to actually make things worse not better!
The problems resulted in the issue of Network Rail Special Inspection Notice ‘Unipart Dorman Phantom Aspect Mitigation NR/SIN/192’ in December 2020 and Notice Board ‘Alignment of Unipart Dorman Signal Heads NB 179 issue 3’ in May 2022. NR/SIN/192 applied to signals with Unipart Dorman LED signal heads that performed the function of a mid-platform or platform starting signal (on a platform or less than 200 metres beyond a platform).
The SIN identified that 4-aspect Unipart Dorman LED signals could display a phantom top yellow in the unilluminated upper aperture caused by sunlight or headlamp reflection at low angles, and by rail vehicle windscreen reflection of sunlight (or other source of sufficient intensity) from potentially any angle. It also identified that both the 3- and 4-aspect signals could have their light output swamped and the aspect(s) appearing yellow, whatever the aspect colour was being displayed.
It required that the potentially affected signals were accurately aligned vertically on their respective alignment points, and no higher. NB 179 provided guidance on how best to align the affected signals to mitigate the effects of phantom aspects following scientific research, technical investigation centre recommendations, and experience.
Unipart Dorman Mk3 CLS
Resolution of these problems resulted in the much-improved Unipart Dorman Mk3 CLS. This is an all new, CLS design with improved anti-phantom performance that can be retrofitted into the previous Dorman Mk1 Classic and Mk2 iLS/CLS LITE signal heads. Mk3 has new electronics and patented optical assembly, with a universal ‘close up’ display. It is available as 110V current & DC proved versions, and a 24V version with DC proving.
All signals in the range are fully compliant to NR/L2/SIG/19820/E03 Iss 1 and GKRT0057 Iss 2 and, after thorough off-rail trials by independent optical experts and extensive industry consultation, it received full product approval with certificate PA05/07576. Prior to the full approval, extensive testing had been undertaken, with dozens of signals, across the network with no reported problems.
With the exception of the internal optics, Mk3 is identical to its Unipart Dorman Mk1 and Mk2 predecessors, allowing it to be installed on conventional structures and the straight or offset signal posts in the Unipart Dorman folding Assisted Lift Trunnions. The Mk3 also retains all the maintenance free properties of the previous CLS versions, including the self-cleaning features.
Previous versions of the signal had an extra row of LEDs at the top of the displayed aspect that allowed the driver to view the aspect when standing next to the signal. When the signal is mounted below a driver’s eye level this extra array is mounted in the bottom half of the aspect. Mk3 improved on this configuration with a central viewing hotspot, and this also removed the need to stock upper, lower, left, or right close up viewing module configurations.
The improved Mk3 has been designed so it allows less light to enter the signal, thereby reducing the risk of phantom aspects, but maintaining the light output from the signal. The placement of the LEDs and the surrounding mask has also been redesigned to ensure the light output meets the requirements of the latest Network Rail requirements, while virtually eliminating phantom aspects. All Mk3 modules now have a tinted outer lens which reduces stray light transmissivity from sunlight and reflections from rail vehicle windscreens. This applies both into and out of the module when it is not lit.
Unipart Dorman says that there is now also exceptional anti-phantom aspect performance delivered by a redesigned close tolerance mask coated in ‘super black’ paint, and that the ‘future proofed’ electronic design allows the signal to be used with relay, SSI, and CBI interlocking systems. This has included extensive testing at Network Rail training centres with various interlocking systems.
The 24V signal is also being considered as a battery-backed solution of mechanical signalled areas with no 110V supply. Yes, it would be great to re-signal these areas, but the industry needs to find cost effective creative and innovative solutions, as signalling renewals are continually being deferred because the schemes are unaffordable.
When LED signals were first introduced it was thought they would need replacing after 10 years, but it was found that they continued to perform far longer. However, every asset has to be replaced at some point and many LED signals are now approaching the end of their life, requiring replacement as per Network Rail’s Technical Instruction TI180. With the Mk3 CLS being compatible with the Mk1 and Mk2 housings this makes renewals easier, saving time, cost, and carbon, along with the far better phantom aspect performance. The future of signalling may be cab-based, but at best some parts of the network will have to wait for Control Period 9 (2034 – 2039) for ETCS cab-based signalling, and it could take even longer. So, LED lineside signals are likely to be required for many years to come.
The author thanks Peter Williams, the author of NR/SIN/192 and NB 179, for his assistance with this article. Peter carried out the audit referred to and was instrumental in the years of collaborative effort by the Network Rail product approval team, asset engineers, and Unipart Dorman, to resolve the problems and introduce the Mk3 CLS.
Image credit: iStockphoto.com/ColdSnowstorm