Return to Archive -by date - by topic - 2001 Archive.
Production engineering done and tested; HMRI confirms it complies with the Regulations – now the train operators have to decide how many units to cup and cone
| 401 days of cup and cone roulette to go |
Let's get one thing straight, in terms of cost per equivalent fatality averted, retrofitting fitting Mk 1 stock with anti-override protection so that it can stay in service until the end of 2004 is a waste of money. But the Railway Safety Regulations says that it will be illegal to run Mk 1 stock in passenger service without such protection after 31 December 2002 – or the end of next year or in 624 days time.
Now, the three southern train operating companies are rather like the low life in the Clint Eastwood detective film ‘Dirty Harry'. If they still harbour the belief that by ordering new trains and doing nothing about modifying the existing stock, they can claim a derogation from the Act from January 2003, ‘Dirty' Vic Coleman is likely to growl, ‘This here's an Act of Parliament, the most powerful legal document you can have. You're wondering whether I'm willing to use it, Are you feeling lucky, Punk'?
Well are you? Would the safety lobby let a transport Minister ease up on railway safety – particularly when the papers have the photos of crumpled wreckage at Clapham on file?
So, cup and cone here we come and I spent a morning at Selhurst Depot at the end of March being taken through the first production standard installation – the result of much work by the HSBC Rail Team, which has acted as project manager for the Rolling Stock Companies, plus a number of consultants and suppliers (see box).
Why is override protection needed? Simply because the Mk 1 stock has a structurally weak body mounted on a rigid underframe. In a collision, this doesn't matter if buffers collide with buffers and then maintain contact until everything comes to rest.
But usually contact is not maintained. In which case one underframe rides up over the other and slices into the superstructure which is not designed to take such end loads.
This happened in the Clapham Junction Accident. Sir Anthony Hidden's resulting public inquiry required British Rail, among other things, to investigate the crashworthiness of Mk 1 stock and to agree appropriate modifications to improve crashworthiness for vehicles with a service life of eight years and over.
Well, the structural modifications were tested. But Chris Green was planning to build 400 Networker vehicles a year forever, so spending a fortune on Mk 1 stock that was going to be out of service within the eight year timescale was not cost effective
Then came the recession and privatisation and nobody ordered trains for three years. All this passed Her Majesty's Railway Inspectorate by. And when the three Southern franchises were let on the basis that Mk 1 stock was fit for purpose for another seven years no one said ‘err, what about Hidden'.
When the HMRI woke up, industry declined to do anything about Mk 1 crashworthiness since, the Value Per Prevented Fatality (VPF) was outside even the £2.85 million applied to railways where there is a risk of a catastrophic accident. Miffed, HMRI asked consultants WS Atkins to produce an anti-over-ride protection modification for Mk 1 stock and the result was cup and cone.
This name refers to the cone on the headstock of one vehicle which, in a collision, engages with the cup on the headstock of the other vehicle, preventing over-riding.
Following live tests, which were only partially successful at 40mile/h, the Railway Safety Regulations (1999) were published. These make it illegal to operate Mk 1 stock after 31 December 2002 , unless fitted with an anti-override device, in which case they can remain in service up to 31 December 2004 .
With some cup and coning inevitable, HSBC, acting on behalf of all three Rolling Stock Companies, set up a team to manage the ‘productionising' of the HMRI's concept. Technical support was provided by Halcrow Rail.
While the cup and cones do the work, the clever bit is getting the couplers out of the way so that the cups and cones can engage. To achieve this, a new coupler is fitted with the shank cut at an angle of 45 degrees and a sheer pin inserted in a hole drilled though both halves at right angles to the shear plane. This pin takes the end loads on the coupler.
Two pairs of bolts, one top, one bottom, clamp the two parts of the shank together, but the bolts run in slots, rather than through holes. If the pin shears in a collision the two parts of the shank slide together, the bolts lose tension and the head of the coupler falls down out of the way.
A reality check by the new team soon junked the proposed cup and coning of intermediate vehicles. There were a number of practical problems.
First, would you be happy walking through an inter-car gangway knowing that the coupler below was designed to shear away under high load? No? Quite.
Apart from which, by the time the leading end of the unit had absorbed the collision energy there was not enough force to make cup and cone work at the intermediate vehicles, even before the inter-car gangways got in the way.
So, cup and cone is now limited to the outer ends of the end vehicles. Of course protection is provided only if Mk 1 smites Mk 1. Should a modern EMU with an integral body hit a Mk 1 unit, over-riding is still possible. And, of course, as Mk 1 stock is withdrawn, the chance of hitting another Mk 1 will reduce, and with it the cost benefit.
Each head stock has a cone on the left hand side looking towards the cab and the cup on the right mounted where the buffers would be. The oval shape of the cup accommodates misalignment between the two vehicles.
Cups and cones will engage down to a minimum radius of 500m which equates to 90% of the network. Misalignments of plus/minus 65mm vertically and plus/minus 160mm can be accepted.
It was assumed that the shear-away couplers would clear the decks for cup and cone to engage, but full scale tests by DERA at Shoeburyness showed that for some reason, the new couplers weren't shearing away. The force required to shear a coupler is around 170tonnes – equivalent to a 7mile/h closing speed. The measured force at that speed was around 75 tonnes.
Analysis showed that the buffer supporting the inter unit gangway was absorbing some of the collision energy. When the gangways and their supports was removed, the couplers worked as advertised.
This means that cup-and-coned end cars will have their gangways removed and the end doors locked out of use. This has obvious implications for safety and operations which the TOCs will have to resolve.
Now for the techie bit, with nary a microprocessor in sight. For tools, all you need is a gas axe, sorry, oxyacetylene cutter, a drill with the appropriate bits and a template so that you can drill the holes to attach the components. Plus a skip for the discarded parts and a fork lift truck.
So take one Mk 1 end vehicle, remove the gangway and gangway support brackets and chuck in skip, followed by the buffers, the step irons, the coupler release mechanism and the coupler, taking care to keep the coupler tail pin assembly with bushing and followers and put it somewhere safe.
Now you get your torch and remove rectangular sections of the underframe, in both side and inner longitudinals. This creates a crumple zone between the headstock and the centre casting which supports the bogie pivot to ensure that the headstocks remain parallel as the underframe collapses. This collapse will absorb 1.3MJ – 30% above the Railway Group Standard
Taking your templates you position them over the spaces in the headstock where the buffers were located, get your drill with a magnetic base stand, and drill 10 x 22.2mm holes on each side to take the huck bolts which will secure the cup and cone castings, plus two smaller holes each side for the bolts to secure the new step irons.
On the cone side you drill two 22.2mm holes in the underside of the headstock for additional huck bolts to restrain the cone vertically. Holes drilled, you get the oxy cutter out again to open up the existing hole in the head stock where the buffer went to make room for the cup.
Then you take the castings, which come pre-drilled, put them on the headstock and bolt them into place from the back. Remembering where you put the tail pin you install the new coupler, bolt on the new step irons and coupler release lever and chain and there you are, one end done. Oops, I forgot the composite cover to tidy up the gap left by the scrapped gangway and the new end door lock.
Once the fitting teams get into the swing of things it should be possible to do a vehicle end a night. The time consuming part is drilling the holes: fitting the new hardware took around 30min on the pre-series vehicle I inspected at Selhurst.
An obvious concern is that a hard shunt could weaken the shear pin, allowing the coupler to fail at a lower load at a later date. So on the side of the coupler is a glass phial containing yellow paint, epoxied to the two parts of the shank at right angles to the shear plane – and protected by a wire mesh cage. If the pin starts to shear, the halves of the shank will slide relative to each other breaking the glass and releasing the paint. Initially this was going to be white, but yellow was chosen to prevent a train being withdrawn from service because of an incontinent pigeon perched on a coupler.
Tests showed that a four car unit of Mk 1 stock hitting a stationary four car unit at 4mile/h, generated a compressive force on the coupler of 100tonnes followed by a rebound force of 50tonnes. As this was being written a 10,000 cycle fatigue test, at plus/minus 100 tones was underway on a production shear coupler.
But if the pin is designed to shear under a 7mile/h collision might this not happen during over-robust shunting?
When coupling, drivers are instructed to stop when the trains are six ft apart, then move forward slowly. But if a train keeps rolling, rather than ‘pause and engage', 4mile/h is a very big bang. At 6mile/h a four car Mk1 EMU will push another four car unit with its brakes on back a metre. So experience suggests that even hard shunts are in the 2-3mile/h range and well within the capability of the shear pin.
Who supplies what
Original concept WS Atkins
Cups and cones Couplers plus other hardware William Cook
Templates and tooling Aircraft Unit Engineering
Gangway cover Dartford composites.
End door lock Pickersgill-Kaye
Drawings, risk assessment and VAB Halcrow
Dynamic Testing DERA Shoeburyness
Pre-Series installation Selhurst depot |
A kit of production parts will cost around £9000, probably a bit under. If the work can be done during a C4 exam, you avoid the cost of overhauling the existing coupler. Assuming one end can be fitted in a night, modifying both ends of a unit would add a day to a C4.
Added to which there is the labour charge. I could see one of the maintenance firms setting up a ‘cup and cone while u wait' line which would probably reduce both cost and time. So, say £15,000-20,000 a unit.
Meanwhile, HMRI has confirmed that the proposed solution meets the requirements of the Rail Safety Regulations and has no objection to the engineering modifications – subject to a number of operational issues. Which is where the TOCs come in.
What does the loss of inter-unit gangways mean in terms of over-crowding, short platforms, emergency evacuations, staff access through an 8 or 12 car train?
Of course these issues are not unique. Several EMUs, including the Networker 465/466 fleet, don't have corridor connections.
But cup and cone will reduce the flexibility of the Southern work horses. If the ‘pascom' is pulled the guard could walk back through the train, unlocking the end doors and stepping over the gap – but the doors would have to be locked immediately.
Bearing in mind that as new trains are introduced Mk 1 stock will increasingly run peak only, one option would be to run fixed formations of eight or 12 cars, leaving the intermediate inter-unit gangways in place.
But the real issue for the TOCs is how many units to cup and cone, which only in part depends on how many new EMUs they think Adtranz, Alstom and Siemens can get into service for 1 January 2003. No space this month for an analysis, but, whichever way you cut it, cup and cone will be a nice little earner over the next 401 days.