Return to Archive -by date - by topic.
System Authorities were intended to manage the interfaces orphaned by the vertically separated railway. The Train Protection and Warning System (TPWS), with its combination of track and train mounted equipment linked by radio frequency transmissions across an air gap, was a classic justification of the need for system authorities.
When retrofitting equipment to traction and rolling stock you are not only introducing more things to go wrong. With TPWS piggy-backing on the existing Automatic Warning System (AWS), the reliability of AWS was crucial to the success of TPWS. As a result the TPWS System Authority set up a sub-group to manage the AWS aspects of the project and, in particular, improve its reliability.
Now the operating principle of AWS is very simple. Mounted on the track in advance of the signal is a permanent magnet with its South pole uppermost, followed by an electro magnet which, when energised, creates a North pole.
Beneath the train the receiver contains a pivoted permanent magnet incorporating electrical contacts or a reed relay with a magnetically biased armature. These control the AWS equipment which the TPWS uses as its interface with the braking system.
When the signal is at caution the electromagnet is not energised. As the AWS receiver on the train passes over the permanent magnet it detects the separate South Pole. Since opposite poles attract, the magnet or relay armature moves, opening the contacts . If they stay open for more than a second the AWS horn sounds and the warning is either acknowledged by the driver, which resets the receiver, or the brakes come on.
When the signal shows a clear aspect, the electromagnet is energised. The receiver now detects the permanent magnet's separate South followed by a separate North pole.
As before, the permanent magnet opens the contacts, but the opposite polarity of the electro-magnet closes them. The opening and closing of the contacts is detected by the control unit and the driver gets an indication and a bell.
This simple system works with the receiver flashing past the magnets at up to 180 feet per second. And the receiver has a particularly brutal operating environment. A solid state receiver has been developed, but most units in service still rely on moving parts.
Now supposing a train comes into a depot with an AWS problem reported by the driver. How do you check the operation of the receiver? Up to now, the answer has been 'with difficulty'.
For a start, I have oversimplified the combination of poles the receiver can encounter.
There is also Separate North, where new AWS has been installed but is not in use or there are traction current cables which can generate spurious magnetic fields. On lines with bi-directional signalling there is a Separate North-Separate South sequence which should produce a horn plus brake application if not acknowledges. Finally there is a Separate North-South-North sequence to indicate a track fault on bi-directional sections.
At some locations there are also extra strength magnets.
Historically, the receiver has been tested by passing a permanent magnet beneath its housing while your mate in the cab monitors the reaction of the AWS equipment. This test provided a 'Go/No Go' which had some benefit but was not repeatable.
Applying the South pole should cause the horn to sound and some deft manipulation might get the North pole into play quickly enough. But applying this technique underneath a £12 million train is not very 21 st Century.
In many cases when a fault is reported, the magnet test cannot replicate it in the depot and if a fault cannot be replicated it is almost impossible to trace the cause. Around 80% of AWS failures are attributed to receivers due to testing limitations.
In such cases, and under pressure to get a train back into service, there is a strong tendency for a Depot Shift Manager to replace the receiver and its connecting cable 'just in case'. And, as you might expect, when these ‘failed' receivers are returned for repair it often turns out be a cases of NFF - No Fault Found. Yet, remarkably, in all the years AWS has be in use, no one had thought of producing a dedicated test unit for the receiver - until two old chums of mine got together.
Despite his youthful appearance John Hampshire claims to have seen me lying on top of hot Napier engines retrofitting steel mesh guards to the radiator fan carden shafts of Deltic locomotives when he was an apprentice at Finsbury Park . He is coming up to retirement at Bounds Green Depot
Dr Mike Rees, I first met at Bounds Green in the late 1980s when the Class 91's traction equipment was interfering with its own electronics. He was sent from GEC Laboratories to sort out the problem, rapidly identified the cause and devised a simple solution which he tested on the spot. Though well past retirement he seems to be as busy as ever and is still keeping an eye on Class 91s at Bounds Green for Alstom.
What was needed was obvious to these two; a means of persuading a stationary receiver that it was seeing the magnetic flux produced by the AWS track magnets. This should include the various combinations of polarities lasting for the time taken for the receiver to pass the magnets across the train's speed range.
As a proof of concept, all that was needed was an electro-magnet which could be put under a receiver in the depot, connected to a battery power supply and energised to see what happened.
With two hands on-chaps like John and Dr Mike this was soon done. As Group Standards specified the magnetic flux required it was a simple matter to calculate the number of coils in the electro magnet for a given current
As my photograph of them with their prototype shows, the coil was wound round a handy plastic tub of the correct diameter. When the current was applied with the plastic tub up against the receiver, the AWS reacted.
This was in 2002. John Hampshire was a member of the TPWS SA AWS sub-group which recommended development of the concept. The System Authority endorsed the recommendation and asked the sub-group to come up with a specification for a production test unit which could be put out to tender.
Functionally, the unit would have to create a magnetic field equal to the minima specified for the permanent and electro-magnetic track magnets, switchable to 90% of minimum field strength. It would also have to be able to simulate the various combinations of north and south poles at speeds between 20mile/h and 125mile/h for standard strength magnets and 20-100mile/h for extra strength magnets.
Rugged construction to take the knocks of depot life had to be combined with the sophistication to work with all types of AWS receivers, including the latest solid state units. Other requirements were use by one person, ease to operating and battery power.
Meanwhile, STS Signalling which, among other things, makes AWS equipment and had been getting back failed receivers which turned out to be fault free, had also been considering the need for a depot test unit. Their approach had been based on clipping a test device, which simulated the magnetic flux seen by a moving train, to the receiver.
So when TPWS SA went out to tender for the development of the new test unit early in 2003, STS was already well into the learning curve. After some refinement of specification the Company won the contract in September.
Under Stage 1 of the contract STS developed a prototype for review by the AWS sub-group. Endorsement of the design initiated the manufacture of two production units for user trials
And in July I visited Bounds Green where one of the units was being assessed. Obviously you need to evaluate the equipment under the widest range of traction and depots at Southampton and Bletchley have used the units, with Central Rivers to come.
As you can see from the photographs it really is a very simple piece of kit which can be carried and put in place by one man. Power is provided by an inbuilt rechargeable lead acid battery. And, being software controlled it is both clever and versatile,
In addition to the nominal magnetic flux it can be set to levels of 90%, 80% and 70% to test the sensitive of the receiver as well as 110% and, following requests during the trials, 120%.
In addition to the carrying case there are two ladder frames. At the centre of one is mounted the cylindrical flux generator with a brass ‘sighting boss' on top to aid positioning under the receiver, although location is not critical. The other acts as a handle.
To locate the equipment the handle is hooked onto the equipment frame while it is slid under the receiver resting on the rails. With the flux generator is in place its cable is plugged into the case together with the long wander lead for the handset.
This T-shaped handset has a small screen and four buttons. The button on the left selects the parameters displayed and that on the right initiates a test. The two buttons on the vertical axis are used to scroll through the parameters.
It is all very intuitive and easy to use. Select the parameter – say sequence of poles, with the left button. Scroll up and down to get the combination required, press the test button and that's it.
So a test goes like this. Put the flux generator under the receiver, connect the cables, climb up into the cab with the handset – hence the long wander lead – and start testing. And I found it uncanny the first time I selected separate South North, pressed the button and heard the AWS chime.
And it's all very rugged. The handset is an off the shelf unit. ‘Is it tough enough for depot use'? I asked John Hampshire. ‘STS threw it on the floor as a demonstration' he replied.
And that is it. As far as the receiver is concerned it is seeing what it sees in real life and the test is not only controlled it is repeatable – as a long suffering fitter in the Class 91 cab found out once I got hold of the handset.
One person can use it and you don't need a pitted road to get access to the receiver. Indeed you could use it in a depot yard.
Given that AWS currently has a failure rate of around 50,000 miles/casualty, the new system should have a considerable effect on overall reliability. And at around £5000 for a complete test set it could soon pay for itself – particularly where trains are on maintenance contracts with heavy penalties for non availability.
TPWS SA owned the Intellectual Property Rights which were transferred to Rail Safety & Standards Board when the Authority was wound up. STS is the licensed manufacturer.
Overall, I was very impressed, but I'll leave the last words to the originator ‘It's better then I thought it would be' said John Hampshire. Dr Mike Rees added that if a University had done the work from first principles they would have got 38 PhDs from the project. There's a lot to be said for experience and pragmatism.