On 20 November 2016 Londoners awoke to find their city battered by torrential rain and winds of up to 100mph. Storm Angus, which had spent the previous day causing chaos to shipping in the Atlantic and across the western coast had finally reached the city.
Angus fled the Capital almost as fast as it arrived, leaving a trail of minor flooding and train delays in its wake. This was seen as largely unavoidable by London’s councils, train operators and the Underground and was soon dealt with. Angus, however, would have the last laugh – for its brief passage through the capital would trigger a series of unexpected issues on the Piccadilly line that are still causing problems today.
To understand why the Piccadilly line is suffering at the moment one must first be prepared to accept two basic facts:
- ‘Leaves on the line’ is a serious problem
- Trains can get ‘flat tyres’
Both of these are simple truths, yet for those outside the railway industry they can be surprisingly difficult to accept as fact. This is partly because, for the last thirty years, ‘leaves on the line’ has been the kind of lazy headline that the mainstream press have always delighted in running. It sits alongside ‘the wrong kind of snow’ and ‘trains as hot as cattle cars’ in the grand journalistic tradition of stories that can be quickly thrown out on a slow news day to generate traffic and public reaction.
Those stories provoke a reaction for the same reason that these facts are hard to accept – because they seem to run counter to ‘common sense.’ Both fall foul of what is sometimes known within the tech industry as the ‘Stands to reason test’ – that any phrase you can put the words ‘It stands to reason’ in front of is almost invariably wrong:
It stands to reason that trains are large and heavy, whilst leaves are small and light, so leaves can’t stop a train.
It stands to reason that trains have solid metal wheels, so trains can’t get ‘flat tyres.’
Why leaves matter
As anyone who has slipped on a garden path in the autumn knows, fallen leaves are surprisingly good at reducing friction – particularly when wet. They also have a tendency to render down into mulch. Both of these leaf states are particularly problematic to railway operations.
Trains rely on steel wheels running on steel rails to move. Normally such a combination of smooth-on-smooth surfaces would be highly unwise and lead to slippage, but for trains the benefits outweigh most of the disadvantages – which can be offset by a variety of methods (such as light sanding and surface treatment). The overall weight of the average Underground train also helps ensure wheels adhere to the track.
Leaves are particularly effective at disrupting this relationship between wheel and track, finding their way between the two and causing individual wheels to slip. This is why, on railway embankments and land, coniferous plants and trees are prevalent (when foliage is present at all). For the likes of Network Rail and TfL, deciduous trees are almost as much of an enemy as Japanese Knotweed.
Leaves on the Underground
One could be forgiven for assuming that the ‘underground’ part in ‘London Underground’ would help ensure that leaves aren’t a problem on the Tube. Yet in fact a considerable percentage of the London Underground is actually above ground. Indeed as any veteran London pub quizzer will tell you, only two lines actually stay beneath the surface entirely – the Victoria and Waterloo & City lines.
Those same pub quizzers would likely also be able to tell you that the Central line is the one which spends the most time on the surface – and as a result it suffers its own leaf-related issues every year. What they may not know, however, is that the Piccadilly line is also particularly vulnerable to leaves on the line. This is largely thanks to the presence of the Uxbridge branch, where both public and private landowners outside of TfL’s control seem particularly fond of a deciduous tree line.
Leaf drop on the Uxbridge branch (and beyond) is thus something that London Underground have to manage every year during autumn, as the surrounding trees gradually shed their foliage and this falls or is blown onto the line. To help mitigate this problem the Underground, like most railways, periodically runs ‘Rail Adhesion Trains’ (RATs) which are specially augmented both to help clear leaves and mulch from the track and to carry out other actions, such as light sanding, designed to help reduce wheel slip.
This year, regular RAT runs have taken place on the Piccadilly line as planned using the RAT that is compatible with the line’s layout and signalling.
Unfortunately, it is here that Storm Angus re-enters our story. For this year’s autumn was particularly dry and windless until it arrived. As a result, it appears that overall leaf drop had been relatively low until Angus arrived. The storm, plus the sudden change in weather it partly brought about, changed this situation effectively overnight. The result was an unprecedented level of leaf drop in a short period that seems to have quickly exceeded a level that the Piccadilly line’s mitigation could cope with.
No doubt other factors than Angus will surface during TfL’s own investigation as to what has caused such an unprecedented level of rolling stock disruption on the line. That Angus and leaves have been such a factor, however, seems relatively clear. Certainly sources suggest that shortly after the problems began, the testing of S-Stock on various sections of track shared by the Piccadilly and District lines was swiftly pushed back so that the District line’s RAT could be tested (and cleared) for use on the line instead. This in itself suggests that excess leaf fall was very quickly assessed to be a significant issue.
The form that issue took relates to the second of our ‘stands to reason’ statements – because once track adhesion began to drop, the wheels on the Piccadilly line trains began to slip.
Worn slippy
Wheel slip caused in braking is a major problem for trains because of the effect it has on the wheel itself. When wheel slip happens, the train’s overall weight – so important normally for helping it grip the track – stops being a positive factor and becomes a negative one.
This is because wheel slip rarely occurs at the exact same time on every wheel, on every bogie of the train. Instead there is nearly always more than enough force generated to move the train forward it’s just that one (or more) wheels then either fail to rotate at all or begin rotating slightly later than the rest once traction is regained. In braking, a wheel might stop rotating before the train stops causing the wheel to rub along the track, thus creating a flat section.
When this happens, a single part of the wheel thus stays in contact with the track longer than the rest of it. The weight of the train, combined with the forward motion naturally then causes this section of the wheel to wear quicker than the rest. It doesn’t take much for a wheel to be worn to a state where it doesn’t rotate evenly, keeping its worn section in contact with the track even longer and making the problem even more serious.
And soon, in essence, the train has a ‘flat tyre’.
Why ‘flat tyres’ are a particular problem on the Piccadilly line
Despite the lack of tyres, ‘flats’ are as serious a problem for trains as they are for cars. They seriously affect movement and also cause damage to the track itself. Once a train has picked up a number of flats it thus needs to be removed from service until the problem can be corrected.
On a modern Tube – and indeed rail – line flats are not an unheard of occurrence. Whilst uncommon, they are to a certain extent inevitable. This is thanks to the seasonal problems described above and to the fact that the way trains are driven and braked can also cause flats. Indeed long time readers will remember that the first batch of 378s delivered to London Overground were discovered to be particularly susceptible to flats due to certain issues with their design and had to be packed off back to Bombardier in Derby for fixing.
As a result of the above, most modern rolling stock is both fitted with Wheel Slide Protection (WSP) and designed so that wheels themselves are relatively easy to detach and replace in the depot.
Unfortunately, the Piccadilly line rolling stock is not new rolling stock. Not only this, but (to make things worse) the design and construction of the 1973 Stock (as it is known) came after train design had moved on from traditional block braking (which at least had the benefit of helping to clear leaves and mulch from the wheel) but before modern WSP had been invented.
As a result, the Piccadilly line trains are particularly susceptible to wheel slip when track conditions are bad or if not handled well.
Driving and braking
The difficulty of braking Piccadilly line trains well is likely one of the reasons why some media sources have suggested that bad driving may be a cause for the current issues. Certainly some have attempted to draw a line of cause and effect between the training of additional drivers necessary to man the upcoming launch of the Night Tube on the Piccadilly with the current issues.
Correlation, however, is emphatically not causation. New drivers are trained on all London Underground lines every year, and – when it comes to the Piccadilly – that training also includes tuition (and practice) at cadence braking and defensive driving. According to sources both of these have remained a feature of Night Tube driver training as well.
As on the road, defensive driving is learning to read (and behave) on the road differently in adverse weather to minimise issues. Cadence braking, meanwhile, is a braking technique to be used when surface conditions are poor. On the road this means pumping the brake pedal in order to slow down whilst still maintaining control and reducing skid risk. Mechanically, the technique is different on a train but the end result is the same and – as one might imagine – it is particularly useful on the Piccadilly line given the lack of automated WSP.
TfL’s own report may ultimately thus show that driver behaviour played a factor, but it seems unlikely. Instead the two biggest causes for why the problem occurred will likely be much more prosaic – the wrong type of weather hit the wrong line, with the wrong type of rolling stock at the wrong time.
Wear and lathing
Why the problem has persisted, and why it is taking so long to recover, is also largely due to simple logistics.
When a train has a flat there are two ways it can be fixed. Firstly, by replacing the wheel with a brand new one. Secondly, by turning it on a wheel lathe until it is round again. The latter is possible in situations where the flat isn’t too severe, the rules (generally) being that a wheel can be reduced by up to 10% of its diameter before it is no longer fit for service.
Whether new or turned, it is not simply a case of replacing the wheel, however. Whilst wheels can vary across a whole train by up to 10% overall, for obvious reasons of stability and balance they can’t vary that much in close proximity to each other.
As a result, even changing one wheel requires work and perhaps replacement of the adjacent and opposite ones as well. On the 1973 stock, and indeed most Underground stock, both wheels on an axle must have almost exactly the same diameter. Beyond that all the wheels on a single bogie must have a very similar diameter and finally all the wheels on the train car or carriage must be broadly similar.
Changing a single wheel, therefore, may trigger an extensive rebalancing exercise across the entire train – and also require the adjustment of everything from tripcock position to suspension height before it can be completed.
It is work like the above that the Piccadilly line depot staff have been working hard to complete. Logistically, it is a situation made worse by the fact that the only wheel lathe on the line is at Northfields’ Depot and using it renders a whole maintenance pit unavailable for wheel replacement work. In previous years London Underground have been able to temporarily hire a mobile lathe, but for reasons not yet clear (but likely related simply to availability) this hasn’t happened this year. TfL have indicated, however, that where possible Piccadilly line trains have now been making use of lathes they can get to elsewhere on the network.
Even with this, however, wheel replacement is a slow process and one that simply can’t be accelerated as it’s a problem of capacity, not effort.
An unusual confluence of events
All of the above hopefully goes some way to explaining why the Piccadilly has struggled so much in recent weeks. As human beings we like to try and ascribe a single cause to events – one single problem or person who is to blame. In part this is because we live in an era of simplified narrative, but it is also because a single issue implies a simple – and hopefully speedy – solution.
Sadly none of that is true on the Piccadilly. TfL’s own investigation will no doubt reveal more about what happened and why, and may even highlight some extra causes and effects. At the most basic level though the issues on the Piccadilly, like many issues that affect the railway, are simply the result of a number of minor factors combining to form a greater problem.
Unusual weather has resulted in a greater-than-usual risk of flats, something the Piccadilly line trains are unfortunately more susceptible to than modern rolling stock. The more flats created, the more trains are withdrawn from service.
Unfortunately there are hard, physical limits to how many trains can be serviced on the Piccadilly line at a time. Once these were hit, the number of trains in service began to drop, forcing cancellations and increasing pressure on the trains still in service a number of which then developed problems themselves.
All of this, from a commuter perspective, has been further aggravated by the fact that one branch of the Piccadilly line was not only directly aggravating the problem but is arguably lower priority than the other (which links London to Heathrow). For both of these reasons the service pattern that has remained in force has generally favoured Heathrow – a major inconvenience to regular travellers on the other branch.
Things can only get better?
Whilst it will come as no comfort to those currently affected, the truth is that the only thing which will restore full service to the Piccadilly line is time. There are no magic means by which the backlog of repairs can be speeded up and whilst TfL’s report will help highlight ways this situation can be avoided and further mitigated (if possible) in future that will not help travellers now.
What remains to be seen is whether this month’s planned launch of the Night Tube on the line will also go ahead. No doubt TfL are under pressure from the Mayor to see that deadline met, but at the same time neither party will want to do so if there’s a risk that it will damage day time services even further.
Perhaps this, then, is the barometer which affected passengers should look to for now. If the Night Tube launches then this would suggest that the worst of the damage has been overcome. Should it be delayed, then that may suggest that there is still some way to go yet.
Cover image by Romazur
Buy a gift subscription for Christmas
Have a friend or relative who loves transport? Then give them a subscription to LR Magazine for Christmas. Buy it now
Whilst train wheels can get “flats” the expression “flat tyre” dates from the days when trains had tyres – a metal rim rather similar to the rim on a cartwheel. Nowadays train wheels don’t have tyres – they have one piece wheels. This is why a wheel lathe is so vital on a modern railway.
One of the reasons a flat tyre sounds so silly is because the average person thinks only of a pneumatic tyre but there were tyres of other types before Dunlop and Michelin et al introduced their version.
And this doesn’t fully explain why there aren’t great problems with the above ground section of the Bakerloo line. As I understand it the 1972 Bakerloo line stock is the last of an old generation with good old-fashioned brake blocks and the Piccadilly 1973 stock was the first of a new generation in the days before wheel slip protection was available.
More worryingly, we are probably at least five years away from the introduction into passenger service of the first new Piccadilly line train and six to eight years away from getting rid of the 1973 stock. Meanwhile, there seems to be no solution – only mitigating measures that would make it possible to get the trains back into service quicker (e.g. a second lathe).
A comment I had heard elsewhere is that the current set of wheels in use were bought “on the cheap” and use substandard Chinese steel – hence why they are more susceptible to flats.
Any truth in this, or simply an unsubstantiated rumour?
Good article making good points. There are three issues here. One is the fact that the Piccadilly line west of Acton Town is a tree huggers paradise. Lots of the stations have trees overhanging the line exactly at the point where the driver needs to start braking. How do I know? I drove trains on that line for 5 years.
The second issue is the braking system on the train is horrid from a driver’s point of view. It’s a bit technical but, trust me, the braking on the 73 Stock is very unforgiving and it is not nice for a driver to apply the most gentle brake he (or she) can and see the speedo drop from 45mi/h to zero in a second. After that you’re in the lap of the gods and their leaves.
The third issue is that the speed on some sections of the Piccadilly line west of Acton is quite high for a tube train and there are some steep downhill bits (approaching Sudbury Hill both roads, Alperton EB, Hillingdon, North Ealing EB) where stopping at the station becomes a lottery in the autumn.
All that said, the skill ranges of drivers varies widely and there will always be some who can cope with the very real issues of leaf fall better than others. The “others” might need some extra training but there isn’t much money on LU for anything nowadays while everyone is running round trying to find ways of maintianing a working railway while the mayor says he won’t raise fares and, now, the wheel turning costs are going to bite hard into an already cut budget. They cancelled 33 Piccadilly trains yesterday.
Any truth in this, or simply an unsubstantiated rumour?
Very much unsubstantiated from my perspective. Not least because wheels are a critical point of failure, so are material tested to hell and back.
Basically I deliberately left it out of the article as there’s no evidence and it smacks to me of trying to find a single magical cause rather than just accepting the obvious evidence. No doubt if there’s any element of truth to it at all it’ll come out in the TfL report.
JB… Thanks for the article. It is a good narrative. I have refrained from commenting on a number of issues….need to wait for the LU report. However some factual engineering comments follow:
Under “Worn Slippy”. you imply that flat wheels can be caused by wheels spinning in motoring. This is not true. Flat wheels only occur in braking. More complex issues occur on more modern trains not on 1973 tube stock. Also “Similarly, during braking, that same wheel will lock in place and stop rotating before the others do.” might be better if it said something like “In braking, a wheel might stop rotating before the train stops causing the wheel to rub along the track, thus creating a flat section.”
In “Why ‘flat tyres’ are a particular problem on the Piccadilly line”……Flatted wheels don’t increase braking distances, it’s the lack of adhesion that does that. Unless the flats are gross – way beyond anything that would be allowed to run in service – they can’t be “grabbed” by the brake shoes or the track, like the sides of a hexagonal nut. The vibration caused by the flats can and does damage the track and doesn’t do the vehicle any good either!
In “Wear and Lathing”… The use of the term “rebalancing” doesn’t make sense. If one wheel is replaces it is probably the new one that needs to be turned on a lathe to make it the same size as the other, good, wheel. Also, you imply that 1973 stock (uniquely?) requires both wheels on the same axle to be the same size. This is pretty much always the case (within about 1mm). As importantly, generally both wheelsets on the same bogie need to be about the same size (within 5mm or so) and often these days, all wheelsets on the same car need to be the same size.
There is also the possibility that the wheels themselves were of inferior quality and the grade of steel being a factor in this debacle. It remains to be seen if this is disclosed in the report. The weather, although a contributing factor cannot be blamed entirely for this; although there was a storm prior to this occurrence, we have had plenty of adverse weather in the past and never have there been so many trains getting flats. It is right that the drivers are not to blame. Driving and braking techniques are the same as they have always been and cadence braking is taught during training as a defensive measure to mitigate against the risk of locking the wheels when braking. One issue not mentioned is the rheostatic brake found on the 73 stock. This brake comes on in addition to the regular brake to assist the train to slow down, however sometimes this applies quite harshly and causes the wheels to lock.
Another factor to bear in mind is that wheel lathes are not just used to remove individual flat spots on wheels, but also to even out wear in general.
This is usually caused by a line having curves in one direction more often than another (very likely unless you have a loop at a terminus), and also when curves occur at points of stress, for example braking zones.
The Victoria Line has a wheel lathe facility at Northumberland Park, for example, despite none of the stations and track in between being above ground and susceptible to leaves, wet track or slippage.
Wheel lathe facilities greatly increase the life of wheels in general and so are used across all lines. Perhaps a single facility at Northfields was just not sufficient for this rolling stock and this line to cope with the pronounced peak this autumn.
PoP…. As you rightly say, all LU passenger stock uses solid wheels, but the world over, tyred wheels are still in use, not least on light rail and tramways where resilient wheels (with rubber elements between tyre and centre) are in use (DLR included)
At bogie level, the brakes on Bakerloo and Piccadilly are similar. One brake block per wheel. Both trains have rheostatic braking. The brake control is different, but I can’t explain why there isn’t a problem on the Bakerloo except to speculate that there’s fewer leaves.
Mark…..Wheel, axle and rail steels are controlled by international standards and would be akin to “original sin” if they are not complied with. They are all safety critical items and it is inconceivable that substandard steel (Chinese or otherwise) would be used.
Might withdrawn D stock have provided a supply of wheel sets that could have been pressed into use on the Piccadilly Line? D stock was unusual in having wheelsets that are similar, if not identical, to those on 1973 tube stock. Have subsequent modifications rendered them incompatible? Lack of wheels would complicate disposal of withdrawn trains, but that is a problem less urgent than getting the Piccadilly trains back into service.
thanks 100andthirty – this was tricky one to find the balance between simplicity and engineering on. I’ll tweak in line with your suggestions.
@100andthirty
I’d agree with your speculation regarding the Bakerloo. The above ground areas of track are significantly shorter than those of the Piccadilly, but also they are within a much wider track corridor, with significantly greater vegetation clearance either side.
A few comments. I don’t dispute any of the basic remarks about flats and how they’re caused. They happen, they’re a fact of railway life.
1. I am not sure I buy the line of argument that somehow things were exceptional this year. We have had severe storms before and not had anything like the fall out we’ve seen this time.
2. I am not sure I accept the premise put by some that new drivers are somehow responsible for the increased instance of flats this year. As the article states drivers are recruited across the network year in, year out and we do not see endemic issues.
3. The nature of the infrastructure and the presence of trees has not materially changed. Yes parts of the Picc Line have specific issues and “interesting” climatic conditions but they are not unique. Other lines like the Jubilee to Stanmore have similar issues.
I been through enough autumns and winters to be reasonably familiar with what can happen to both station and train assets on the Underground. I’ve seen the preparation plans, I’ve seen hundreds of incident reports and been involved in the fault attribution and associated arguments. I’ve read more legal opinion about what an “exceptional storm” or “exceptional weather” is than some people have hot dinners. I’m pretty convinced that something else has gone on in the last year or so that has contributed to the problems this year. I am aware of the allegations about wheelsets and I understand why JB has steered clear of embracing that issue. I am keen to see just how transparent LU is with its report on the problems and where responsibility finally rests. It will be an interesting read assuming it is published. I note questions to the Mayor are already in place for next week’s MQT.
TfL have said (reported in the press) that the Picc Line Night Tube will launch as planned but with the current scale of cancellations I don’t see that as tenable. To offer a better service on weekend nights relative to what peak time commuters are getting feels like an own goal in the making. Restoring fleet resilience and reliability for the majority of your passengers has to have a higher priority than launching the Night Tube. I’d also argue that getting that resilience is essential if you wish to run a viable overnight service on New Year’s Eve when the Picc Line’s capacity really is required.
@Ex Piccadilly Driver: are the “more highly skilled at braking” drivers the ones runniong the Acton shuttle at the moment, then?
Apropos a couple of statements about less leaves on the Bakerloo Line – this may be true as I have noticed a lot of recent co-ordinated vegetation/tree clearance on the route north of Wembley towards Harrow and beyond.
If my garden is anything to go by, it is certainly the case that leaf fall this year is much later than usual. It is now only 2 weeks to Christmas yet I have spent the morning raking up leaves and there are still more to come down. This means that the leaves are much wetter than usual and consequently harder to clear. In most years they have fallen much earlier when the weather was dry(ish).
There has been some interesting research in recent years into leaf fall, which suggests it is not just about mulch. A common problem that manifests is black spots burned into the railhead. This is thought to be caused by a chemical reaction between the pectin in the lead mulch and iron oxide on the rail, forming iron pectinate which has a particularly poor adhesion factor.
As for portable wheel lathes, these are limited in number and not surprisingly in high demand at this time of year. It may be coincidence, but I understand one such unit is on long term lease to Greater Anglia’s Crown Point depot to try and avoid the wheel problems that almost wrecked their regional services in autumn 2015.
Pretty much fully agree with 130s comments but would add:
London Plane leaf mulch paste* on polished steel has teflon-esque coefficients of friction so is the rail equivalent of “black ice” vastly increasing stopping distances (up to 10x) leading to far more defensive driving.
*the oily content from leaves of certain species making the paste far more slippy than most leaves
Turning a new wheel to match existing reduced diameters is incredible inefficient overall as you reduce the future number of times the wheel can be turned.
New Picc wheels are 790mm and the minimum diameter is 710mm.
Any bets on Hegenscheidt getting a cheque in the post shortly???
An excellent article as always! This should be required reading for any disgruntled commuter who says that ‘leaves on the line’ is a silly excuse.
Not sure if I approve of the use of Arsenal in the heading picture, though…@WW
Yes, I agree with Littlejohn above. It’s not just the storm, it’s that the storm came after a very mild autumn that had delayed the leaf-fall. In my own area of NE London I didn’t see any significant leaf-fall from trees until November.
@John Bull: “They seriously affect affect movement” doesn’t look right. That’s as far as I’ve got; maybe more later.
Lots of comments crossing each other – was referring to the original 130 comments but agree on the subsequent ones too.
Re PoP,
Differences between Bakerloo and Piccadilly:
As Steven Taylor has pointed out there NR have been having a massive and well publicised de-vegetation exercise on the in London section of the WCML for at least 2 months earlier this year.
Mobile wheel Lathes:
Agree with Chris J that Anglia (new franchise awarded since last autumn) have got a mobile one in at Crown Point as they had a disastrous autumn 2015 performance which they can’t afford to repeat with new higher PPM targets. (After a string of poor performances when a former franchisee (not Abellio) didn’t replace a knackered wheel lathe when they thought they might lose the franchise…)
Ngh 17.39…..couldn’t agree more about the inefficiency of turning down a new wheel to match an existing wheel. I made the point mainly to correct JB’s drafting rather than advocating the process (which I have seen done occasionally when the brown stuff hits the mechanical air mover).
If it is due to lack of training among new drivers surely the duty on which flats was first detected on a damaged train could be identified & related to a particular driver. If there is no pattern, the inadequately trained driver theory could be laid to rest. In the 1990’s sudden heavy wear on all Victoria line train wheels provoked all sorts of research. It included analysis of who had been driving the affected trains. Although automatic, they could be manually driven. The wear stopped. I believe the cause was never discovered.
Jim elson
In the early 1990’s there was a problem with excessive wheel wear on the Victoria line as you say. The cause was lack of lubrication. It seemed that there had been a loss of tacit knowledge about setting the track lubricators. In recovering the situation, some of the lubricators were found to deliver too much lubricant – it was not possible to set them precisely, nor was the process intuitive. This caused flats! All was eventually fixed by using train mounted dry lubricant which stays where it’s put!
Idle thought; Wasn’t there a massive track replacement programme on the Uckbridge Branch in summer 2014 (closure for several weeks). Any bets on there the new rails are a
lotbit harder and less forgiving on the wheels (in micro/macroplasticity terms) than the old ones? [Same issue seen in heavy rail in many times and places over the years…]The article states ‘Not only this, but (to make things worse) the design and construction of the 1973 Stock (as it is known) came after train design had moved on from traditional block braking (which at least had the benefit of helping to clear leaves and mulch from the wheel) but before modern WSP had been invented.’
This is not correct the 73TS has conventional brake blocks and 4 of the 6 cars have Rheo braking. What has changed is that the modern brake blocks no longer have Asbestos in them but this has been the case for at least a decade.
73TS was originally fitted with WSP but this was not very effective due to the speed of response and was not operative in the Westcode Step 7 (emergency) brake rate. This was removed. Braking is via the Westcode 7 step valve – only the top 3 or 4 brake rates are available to the operator.
Mraking arrangement is similar to that on D78.
@ngh – a couple of comments from a non-wheel-turning aficionado:
– in BR, we found that the switch from hand tamping to mechanical tamping led to a much harsher ride, especially over switches and crossings. I don’t know what, if any, changes in practice resulted from the tube PPP – just a thought
– if the problem of matching wheels/wheel sets is so severe (and as I will explain next, that seems to be the case), it would be worthwhile keeping complete spare sets of matched wheels so that the delays in returning the trains to traffic was limited only by the time taken to lift them. Perhaps they do this already? [The seriousness of the issue can be seen at the Franco-Spanish frontier, where there are fields full of carefully identified wheel sets from wagons that have had their gauge changed when crossing the frontier, awaiting the return of their “own” wagon – the insurers insist on nothing less].
Building on JohnM’s comments:
There has been no asbestos in LU brake blocks since 1984
With reference to the formation of ‘Iron Pectinate’ above there is a reference to the process at http://proceedings.asmedigitalcollection.asme.org/proceeding.aspx?articleid=1595791 (abstract available but the rest is behind a paywall). In my experiance water with a touch of detergent (Tepol) was sufficient to prevent a 73TS stopping even with the WSP working. The other method that provides a good simulation is to stick brown sticky paper tape to the railhead and dampen it. With air operated brakes one has to be careful the WSP does not discharge so much air that no braking effort is left. Not so much a problem on trains with large compressors for air doors but trains with electric doors have much smaller compressors.
Does anyone have any information if the problem on the Picc is with the Rheo braked Motor car axles or the air braked trailer cars ?
It’s a shame they didn’t think about utilising the remaining D Stock to up the Rayners Lane shuttle to about 6 per hour instead of 3. There are several still in use on the District, and always idle S stock in Ealing Common depot that could be used on that line instead of the Ds. I was waiting at Alperton this week when the Adhesion train made up of A stock went through, and when the track was replaced there was a class 47 diesel on the line, so there is no gauging issue with D stock.
“When a train has a flat there are two ways it can be fixed. Firstly, by replacing the wheel with a brand new one. Secondly, by turning it on a wheel lathe until it is round again.”
There is a third way: by welding new steel onto the surface to restore the lost material, then turning it down to the specified diameter. I once worked in a factory in France which did just this.
This isn’t especially relevant to the case in hand, because it’s a specialised task that I don’t expect to be handled in house, but I thought it might be interesting to readers.
‘@Geoff in Wembley
The remaining D Stocks are supposed to be there to cover for S Stocks being returned to Bombardier for modifications for fitting the new ATC system. The problem with using them on the Rayners Lane Branch is that the Piccadilly Line Operators are not trained to drive or fix D Stock and the District Line Operators are not trained to drive to Rayners Lane. I suppose the trains could be dual manned if there were sufficient District Drivers surplus.
The other problem would be the Customer Information Systems would need updating to include the stations to Rayners Lane. Due to the Disibility Discrimination Requirements to have CIS I guess they could not be operated without a functioning CIS but perhaps others might comment on this.
I can’t remember if D Stock is cleared to run on that route in customer service or not so might require some gauging and route approval before this could happen.
How much does a wheel lathe cost? Can TfL buy another 10 for next year? Is it more a question of space in the depot?
They had the same issue last year to a lesser extent. LUL hired a wheel lathe and put it into Cockfosters Depot. During the summer this lathe was transported to Stratford Market Depot. Did LUL think that the leaves were not going to fall this year?
Apparently there are issues with how the drivers are operating the brakes and the Depot staff suspect it’s linked to ongoing industrial action.
@Paul and other comments about ground wheel lathes, the ground wheel lathes have problems with turning wheels with flats as the drive rollers bear on the wheel tread that has the flats and thus the wheel will drop as the flat goes over the drive roller. Difficult to draw a diagram here but the arrangement at the bottom of the wheel being turned is :
Drive Roller 1 —– Turning Tool ——- Drive Roller 2 with the mass of the train supported on the drive rollers. The wheelset is turned via friction with the drive rollers.
The last thing you want is a wheel that is not concentric to the axle.
Conventional wheel lathes turn between centres so this is not a problem.
Re John M,
Don’t the newer ground wheel lathes support the mass of the train on the axle boxes which are clamped and used as a fixed centre.
Re GPOC,
Isn’t portable one at Stratford a permanent fixture there due to the archiology found underneath the depot during construction so pit based one couldn’t be fitted. Hence possibly loaned to Cockfosters?
Re Ratty,
10 is about the annual global production rate from all manufacturers.
The Tramlink wheel lathe cost north of £5m when eventually purchased from Hegenscheidt and installed. It’s not normally a piece of kit that can be simply brought in on a whim.
@ngh – I was always taught that the steel rail was harder wearing than the steel wheels that ran on them, if only for the reason that it was more expensive and time-consuming to change worn rails than to deal with worn or flatted wheels. That is why many railways and light railways prefer tyred wheels rather than unitary wheels. With tyred wheels, one can use a wheel lathe, or change the tyre when it is worn down without having to replace the whole thing.
Mention was made above of resilient wheels. What may not be so well known is that London Transport (as I was told at the time), concerned as they were about the noise of the Victoria Line trains, decided to experiment with resilient wheels, since they were proven abroad to demonstrate a positive reduction in rail-generated noise through the wheels. So far, so good.
However, they only had a limited budget and so only equipped one car (maybe two) on a full-length train and then wondered why there was no significant reduction in volume! Another aspect surfaced, in that the workforce at Acton Works, so it was said, was resistant to handling the resilient wheels in the first place, whilst those more closely involved made up (to my mind) the spurious excuse that the wheel rims would be distorted out of alignment with the axles because of the rubber inserts, despite obvious lack of evidence of such from e.g. Germany at the time, indeed right up to the present.
I know what conclusion I drew from all this but I leave you to draw your own.
@Graham F: Deutsche Bahn’s decision to replace the monobloc wheels on their high speed trains with resilient wheels, (to reduce vibration), caused the Eschede disaster. The tyres of the resilient wheels turned out to be liable to distortion that accelerated fatigue cracking.
Why should the Night Tube be in doubt? It requires few trains and low additional mileage. Not even serving Uxbridge branch. The daytime service suffers since there are insufficient serviceable trains to fill timetabled paths, but overnight there should be no such problem.
Just a couple of comments: @Bollo Mouse: The rheostatic brake is selected automatically as the preferential brake with electro-pneumatic added as necessary.
The brake rate selection is very coarse and the e.p. feed up is quite rapid so wheel slide is almost inevitable in bad conditions. If the rheo cuts out (due to wheel slide) the e.p. applies automatically, exacerbating the slide. The only way out is to release and re-apply. That takes nerve.
@ 130: Don’t the 73s still have slip slide control? They did originally but was it taken off because is was ineffective? I remember some trials were done in the early days.
Ec-Picc Driver. As commented earlier, WSP was originally fitted but removedcbecause, basically, it didn’t work.
Graham Feakins. The only train that was fittedcwith resilient wheels in the period 1969 to date was the BREL prototype train of circa 1986. This didn’t have a conventional primary suspension on its steerable bogies so the resilient wheels reduced vibrations into the bogie frames. The bogies were not sufficiently successful to adopt.
Is there any simple way of tweaking the rheo-brake setting, to give more sensitive handling?
Either rheo or regenerative brakes are the equivalent of a car-driver using his/her gearbox to slow down in slippery/dangerous conditions.
[ It’s assumed, f’rinstance, that Land_Rovers are “immune” to snow & ice, which is simply untrue. You still need a different driving-technique & selection of the correct braking, gears & drive-combinations to proceed without ending up in unwanted places. ]
From an engineering p.o.v. trains have the same problems.
@130 The only one on the Underground, perhaps, but resilient wheels were certainly one of the measures used by British Rail to try to reduce the damage to the permanent way caused by heavy unsuspended weight of the axle hung traction motors on the Class 86 locos.
Wheel flats can occur on rubber tyres. Pneumatic ones usually have hard enough compounds not to be a problem, but solid tyres had to use softer rubber and could get flats. I also recall the problem on some of my Dinky Toy collection. Formula 1 racing cars use pneumatic tyres, but have very soft rubber in order to trade durability for grip. Pirelli produce five versions, from “hard” to “ultrasoft”, but even the hard formulation rarely lasts more than half the standard race distance of 200 miles. Locking up the brakes and flat spotting the tyres is not as serious as a puncture, but the resulting vibration is often serious enough that a new set of tyres is needed.
“Resilient” railway wheels.
Has anyone else here come across the Mansell Wheel” ?
Which had carefully-cut wooden sections between the wheel-centre & the steel rims.
Used a lot on railway carriages in the period (IIRC) 1860-1920.
With a return for subsurface operations to having just one type of train – S stock (who remember R and O stock?) I just wonder if we might also see the Tube returning to a standard – most obvious in the form of 1938 stock but also in the later 59 and 62 versions. That way problems on one line could be mitigated by transfer of some reserve trains used on other lines.
If there is a need to hold the braking in rheo to avoid the severe step in braking, then perhaps a driver override ‘logged’ when used, should be available.
Overground is another case to hold up as having some commonsense (but it could still have been improved). All trains use the same driving and trailer vehicles, and trains have ‘grown’ by adding extra trailer vehicles. Unlike many a/c trains these ones have lockable opening windows catering for the days when the condenser might ‘overheat’ and cause the a/c to malfunction (a possible cause of Boris bus issues). The only flaw in the ordering was to buy Class 170s for the GOBLink, when a Stadler-style diesel-electric module inserted between 2 cars would have allowed the use of Class 378 carriages, ready for electrification. The modules would have a re-use value with other operators saddled by having isolated diesel routes and a large standard fleet of electrostars (can’t imagine who ..?)
@NGH ‘Don’t the newer ground wheel lathes support the mass of the train on the axle boxes which are clamped and used as a fixed centre.’ I must admit I am a bit out of touch but the proposal for the Vic line wheel lathe had clamp downs on the axle boxes to increase the force on the rollers so more cutting force could be used. It was not commisioned when I left the project as the 67TS axle boxes had not been desidned to have external force applied on the part outside the primary suspension rubbers.
In terms of fixing the geometry the ‘hook holding the axle box would have to be rigid and the rollers forced up against it by hydraulics that would be capable of responding as the wheel flat passed over. The images I can find on the internet do not seem to show the detail and I did not look that closely at the last wheel lathe I saw at Ashford.
There is an RSSB report into ovality introduced by Ground Wheel Lathes that concludes that the problem has largely been resoved by holddowns so I am not sure what is happening.
Some of the latest ‘wheel lathes’ seem to be a combination of a lathe with a milling cutter. This presumably reduces the drive requirements to turn the wheel – and incidently gets rid of the problems of long spirals of sharp red hot metal coming off the lathe tool. The chip breaker is vital to operator safety.
@Ex Picc Driver one of the problems with the Picc Line WSP was that the volume of air in the pipework between the dump valve and in the wheel cylinders was quite large and you have to remove a significant proportion of this to get a stopped wheel rotating. The system was also on a per-bogie basis which meant if either axle stopped you lost the braking on both as the WSP dumped the air. Also the electronics had no Safety Integrity Level ( was it even about in the 1970’s ?) so it was disabled in emergency so when things became critical either running up to a red signal or overshooting a platform or getting tripped the Emergency Brake rate would be applied and the wheels lock.
Would this not also affect the D78s on the District Line (or were their bogies etc replaced in the 2008 refurbishment)? Or it may be that the D Stock is now to small (7 or 8 trains) make an operational difference to the District
@DH
There has never been a single type of Tube stock on all lines. Even the “Standard” stock, (which had many detail differences, albeit they could work together) was not used on all lines at the same time, as the Central Line was not converted to 4rail electrification until well after 1938 stock had appeared on the Northern.
@Nick Biskinis.
The main problems seem to have been caused by a particularly bosky branch of the Picc. There are, apparently, not so many trees overhanging the District Line.
@DH -whilst there are some advantages in terms of holdings of spares in having a standard tube stock,the differing maximum lengths (and in some cases, tunnel geometry, preclude a single fleet. There are, in any case, good financial and technical reasons to avoid a peaky build and a single generation technology. Any large scale owner of assets such as trains (or power stations, or aircraft, or…) will try and even out the replacement of the portfolio so as to avoid a spike,with all the polticial and fiscal difficulties that go with that. And if you are successful in that evening out process, then it’s pretty certain that in the case of long life assets such as trains, the technology will have moved on considerably between the assets’ introduction and their removal from service. Even NTfL will probably go through several versions if construction is spread over two or three decades, as seems likely.
Very interesting discussions about wheel lathe principles. Just a bit of nit-picking – it is the weight of the train which has to be supported (on rollers on or something else) not the mass.
timbeau: my Dinky Toys also had solid rubber tyres, but they didn’t get flats. I put this down to the lack of brakes, but perhaps yours were better equipped?
Taz asks why the Night Tube on the Picc should be in doubt. Some suggestions:
(1) Management efforts may be directed towards “fixing the roof”, and suggestions of adding on a nice-to-have “conservatory” at the same time might be unwelcome.
(2) Trains used on the night service need their daily checks done during the inter-peak, and the depots may be too cluttered up with unserviceable trains
(3) Disgruntled passengers may be slightly regruntled by evidence that tube management are taking their problems seriously enough to cancel something else (even if it happens to be something rather irrelevant).
Would not the simple expedient of a weighted scraper or wire brush ahead of several bogies not clear the rails ?
NM
No
They wear out – quite quickly
How do I know? Because such an obvious & simple solution has already been tried, I’m afraid.
No, my Dinky toys didn’t have brakes, but they did get seized bearings through being left out in the rain.
Greg
“Has anyone else here come across the Mansell Wheel” ?
Which had carefully-cut wooden sections between the wheel-centre & the steel rims.
Used a lot on railway carriages in the period (IIRC) 1860-1920.”
LSWR 1520 Brake third and SR3363 (SE&CR 1084) Brake Lavatory Third / Composite had Mansell wheel sets and are in service on the Bluebell Railway. I presume they still are running on the Mansell wheels… but there is some debate about how they can have new tires fitted (no one has done this for quite a few years!), but I believe there was no problem turning them.
A few comments:
It’s not true that railways in general in the UK have ‘coniferous’ trees and plants on their cutting sides and embankments. Out here in Hampshire, it must be about 95% deciduous at least, and I’m sure that’s true for all lines through rural areas. Anyway, I doubt that coniferous trees turn up much even on urban railways. As the railside vegetation usually reflects the natural* local non-railway vegetation, it will often reflect what has been planted in urban gardens. That means you will see more evergreen but non-coniferous trees on railways in London. Usually something like Laurel. The other characteristic of land beside railways is that ‘weed’ trees absolutely thrive on it. The London ‘weed’ is almost certainly Sycamore. Just like the ubiquitous street tree, the London Plane, Sycamore has massive leaves, perfect for plastering themselves over rail lines especially when wet.
Things will always be different in different years, due to two variables, the state of the trees (secondarily dependent on the weather), and the weather itself. Wind speed, wind direction, rainfall – all combine to make a bit of mystery of why different years and different places are affected by leaf-fall.
* ie not crops
Well, Hampshire is not the UK, but you’re probably right about the limited amount of conifers. This may also be connected with the fact that conifers are slow-growing, so they do not take well to the sort of random every-ten-year blitz which railside land gets.
@Malcolm
conifers are slow growing
To be honest, no more so than deciduous trees. The Norway Spruce and variants are grown as cash crops and bought this time of year. And as for the Leyland Cypress, it grows at a phenomenal rate.
End-state woodland – generally oak and beech in England – grows slowly. What you tend to see on the sides of railways – sycamore, ash, plane, birch, willow – grows much quicker and is quite happy to be coppiced. You are right to suggest that most conifers would object to being coppiced, but there are counter-examples (Leylandii again for one).
If we restrict to native and naturalised trees, what you are saying is sort of true, but there are only three native conifers, and yes, they are slow growing. However, one of the three (yew) is quite happy being at least pollarded.
Also don’t be fooled that conifers in particular and evergreens in general don’t drop their leaves: they do, just not all at the same time, and the leaves don’t create the sticky mess we know from sycamores and planes.
Ian, my Leylandii hated it when I coppiced them. But then I was hoping they would! As a daily commuter is S.E. Land I know there are sod all conifers along my route. There are some around Chislehurst, but that’s about it… Mostly it’s deciduous…
@Malcolm
“Well, Hampshire is not the UK…”
At first, I misread that as “Hampshire is not in the UK” (d’oh!) when of course you actually mean that Hampshire is not a representative sample of the UK.
Historically, leaves on the line were not a problem as in steam days, embankments were kept clear of foliage to minimise the chance of trackside fires. Since steam locomotives have been withdrawn, however, this is no longer done and hence the issue arises.
There is also the problem of not having supporting trees as it can make an embankment more vulnerable to landslides in heavy rain.
A problem with no easy solution.
@100andthirty 10 Dec at 07:09 – LT certainly ran trials with resilient wheels and this report (which I haven’t read) coincide exactly with when I recall being told about the half-hearted experiment on the Vic. Line:
Rutty, F. G. Resilient Wheels—Experience of London Transport,
London Transport Report 8/74-c239 (revised September
13, 1974).
There is also this, which I assume is related:
London Transport, Office of the Scientific Advisor, “Report on
the Effectiveness of Noise Damping Rings on Train Wheels,”
Investigations into Noise on Railways, Laboratory Investigation
128/X215 (September 30, 1974).
A related quote is here:
“London Transport reports a total noise reduction of 3 and 5 dBA at
curved track with the use of the Penn Bochum and SAB [resilient]
wheels, respectively [on curves in tunnel]. Taken from this document:
http://www.tcrponline.org/PDFDocuments/TCRP%20RPT%2023-D.pdf
@ Taz – reading the experience of the P Line night running trial on another forum doesn’t paint a very encouraging picture. I also don’t call 21 trains [1] in service overnight a small number when the Picc Line fleet team are struggling with nearly 40 trains cancelled a day. I’m sorry but the situation with the daytime service is appalling. I appreciate it’s a tough task to rectify so many trains quickly but LU really should not doing anything that may increase the stress on the fleet or run the risk of more flatted wheels. In the great scheme of things restoring the best possible daytime service has to be the priority given how heavily used the Picc is at this time of year with shopping, leisure and airport traffic all piled on top of the usual crush.
[1] and yes I realise this may be higher than what is planned to run in normal night service but my view remains the same.
Graham Feakins…..’you could knock me down with a feather’. I must have missed the Rutty work having been merely a young ‘erk’ at the time. The Scientific Advisor report however relates to the steel noise deadening ring which is fitted to all LU’s passenger wheels to reduce ‘ringing’ noise.
WW: point of clarification please. Are both your references to numbers of trains counts of physical trains (rather than trips)? I assume they are, but it would be a trip (or a whole diagram) which would be cancelled, A train is either in service or out of service (for the whole or some part of the day or night in question).
I do agree with the point you are making about the priority, as I tried to make clear (??) with my roof/conservatory analogy.
I have also just realised that I think I contradicted (through forgetfulness) what someone with evident inside knowledge told us a while back. If what that person said is correct, trains used on the night service are not (as I suggested) given their daily look-over during the daytime inter-peak. They get it at some point in the night, hence the need to have all-night access to depots. No train (I think) is allowed to run during evening peak, all evening, all night, and the following morning peak.
@Silent Hunter: leaves on the line as a post-steam phenomenon is true in general, but in this case I don’t think that the Uxbridge branch of the District/Piccadilly ever had steam trains. So I suppose the trees on the line have had 100+ years to grow. As PoP mentions above, the pre-1973 electric trains’ brake blocks acting on the wheels would have helped dislodge the leaf mulch.
I’m surprised that no one has mentioned the track circuit issue? Or is all of LT axle counters? I gather early BR/RailTrack/Network Rail issues was that once block brakes went out (replaced by discs) on sprinters, that trains ‘disappeared’, as leaf mulch built up on the treads and the wheels became insulated from the rails. I’m told it was a while before the occasional disappearance of trains in Thetford Forest was understood!
Ian J
The opening day of the Uxbridge branch was steam-hauled, by a Met 0-4-4 Tank.
Quote from old (1972) booklet by LT on the “Met”, showing the inaugural train on 4 July 1902 …
“Although the branch was intended for multiple-unit electric trains, it was worked by steam for six months”
You can see the conductor-rails laid in, in said picture
Two points come to mind when I read this thread. First, steam locomotives were all fitted with sanding gear which I believe was regularly used. Whether the larger diameter drivers helped or hindered the wheel slip problem others may be able to comment on. Second, during my forty plus years of driving vintage trams I was taught to use EITHER rheostatic brake OR wheel brakes but NEVER both at the same time. If the rheostatic brake caused a lock up then it had to be shut off and the wheels got rotating before applying the wheel brake. This might involve applying power briefly. Sand was also an essential aid. Assuming the rheostatic/wheel brake changeover on LU stock is automatic it sounds as if the system needs a bit of fettling.
I would have thought Rheostatic braking would be less likely to lock up the wheels than wheel braking?
Or is it that wheel braking can be applied more gradually?
@JimJordan – large driving wheels were always said to be more prone to slipping, but the as the largest diameters were usually on the single driver designs,which had other reasons to slip, the evidence may not be 100% conclusive.
@Greg T -yes, that was my understanding, too. (Off topic, but presumably the cause of the delay was a shortage of stock?)
Graham: I presume the same presumption. Which just goes to show that mis-synchronising track electrification and electric-train acquisition is not a new problem, even if it’s not quite solved yet…
SHLR: Yes, good point. Any kind of electric braking should (in theory at least) have the characteristic that the braking torque acting on the wheels is very low if the wheels are turning very slowly. From the point of view of not bumping into things, this difference is probably unimportant (1), but from the point of view of not producing flats, it could be very important.
(1) The retarding force slowing the train is probably no different as between slowly turning wheels and locked wheels.
A third effect of wheel lock (additional to “bumping into things” and “making wheel flats”) would be a sudden reliance on those pesky flanges to keep the vehicle on the track, as the conical-wheel effect suddenly disappears.
@Greg, Malcolm, Graham H. Yes that is right for the Met but the District extension to Uxbridge didn’t open until 28 June 1903 and so far as I know was always electric. Ian J specifically referred to ‘the Uxbridge branch of the District/Piccadilly’, which I took to be Rayners Lane – Ealing Common.
“Wheel slip caused in braking”
“Slip” occurs when motoring in poor adhesion conditions. “Slide” occurs when braking.
Littlejohn.
Given your caveats – you are also correct.
The Rayners La – Ealing Common section has always (IIRC) been electric traction.
“The Uxbridge branch of the Picadilly” refers nowadays to the whole section from near Ealing Common to Uxbridge, notwithstanding the fact that Met and District trains have formerly used, and/or still use, some of the relevant track. I don’t think the article spelled out whether the leaf problems were restricted to one side or the other of Rayners Lane, but I would imagine that the whole branch is affected. (Anyway the very limited use of steam traction on part of the branch, many years ago, is of course historically interesting, but is also, as someone said, off topic).
@Greg, Malcolm: The Met and then BR also operated steam-hauled freight on the Uxbridge branch into the 1960s, I believe, so would have had a reason to keep the section north of Rayner’s Lane clear of vegetation. As Littlejohn surmised I was thinking of south of Rayner’s Lane. At least north of there the Metropolitan provides an alternative service (and is the main route to central London).
There is no officially accepted method to cut out the rheostatic brakes, but it would solve a lot of problems, especially when you know the rheo rates apply and feel differently from unit to unit. I get the feeling that those up higher don’t want to effect it because it would mean adding another mod that may/may not work to an already heavily botched train that’s on its last legs, never mind square wheels.
We’ve always had leaf fall. We’ve always complained about it. We’ve never had any remedial action besides the occasional cutting back of the greenery.
I’m sure there are plenty of current and ex-Picc drivers (including myself in the latter) that will attest to the rheo coming on so harshly on some units (at somewhere completely inappropriate, say the approach to Southgate on the westbound) that it sets the runback off, causing an emergency brake application to come on that you have absolutely no control over that the time. Now imagine that on the approach to Sudbury Hill on the westbound, on a particularly mulchy day… and I bet you any money the speedometer reads at a confident 0-15mph fluctuation the whole time you’re applying the brakes. It’s not a great experience – it’s actually quite stressful. Especially when you change ends and find the train then behaves like a completely different beast altogether.
@IanJ = and indeed just a little bit further south than Rayners Lane,too, to serve the S Harrow gasworks siding, but I suspect little lineside vegetation was encountered. [I’m not sure that BR cared much about the freights to Uxbridge by the end – certainly as far as motive power was concerned – I have a vivid memory of an extremely rusty 76042 taking 3 attempts to take a single wagon up the bank into the yard at Eastcote]. As Anne says, LT didn’t seem to go out and hack back the greenery much, even then.
The leafiest stretch of LUL track is probably Leytonstone – Epping.
Are there any problems with leaf-fall/slippage there?
And if not, why is the Picc suffering so much?
Yes, I know, the traction/control/braking systems on the Central are different & newer, but, even so ……
All the coal yards along the Uxbridge branch that I remember from my youth are now car parks, but the last time I travelled the line I am sure that the unelectrified siding was still in situ (and rusting nicely) at Rayners Lane, between the running lines and the car park. Is it still there?
TfL have just confirmed that the “Night Tube” will start on the Picc this next weekend ( 16-18 December ), notwithstanding the problems mentioned above ….
@Greg Tingey
You sort of answered your own question there. As you say, the motoring/braking systems are completely different. The main difference however is technology; how leaf fall effects are monitored, and the mitigation plans that are in place.
The Central line operates a Rail Adhesion Train (in the form of a converted 1962 tube stock), and whether or not it is required to run is assessed by engineers who are able to monitor the braking rates/wheel slippage of 1992 tube stock remotely. If the numbers say that the rail adhesion is poor, then the RAT will go out to deposit sandite on the railhead to help with adhesion issues. This has been going on for years – the Piccadilly has never had the luxury of sandite treatment until now, and even so, it came too late.
@Littlejohn
The trackbed is still there at Rayners Lane, but the rails have long since been lifted. Only a few sleepers remain, nearest to the station.
Anne
I have a photo of said RAT at Ealing Broadway, taken about 4 years back …..
With Paddington closed for Crossrail works at Christmas and almost till New Year and hence no Heathrow Express or Connect it will be interesting to see if the Piccadilly line is able to cope with a reduced number of trains and the extra passenger load.
(No NR services east of Ealing Broadway so expect lot of extra Piccadilly use from Acton Town as a faster alternative to the District).
@ngh: Do you think it will take that long to get most trains back into service? I notice that the TfL page on the problems is sadly devoid of any hints or expectation about when a full service will resume. Is there any expected date for a full resumption of service?
@ Graham H: an extremely rusty 76042
Really? It can’t have been more than 10 years old as it transferred up to Birmingham in 1964. Admittedly it was scrapped two years later…
At least according to this.
Did BR really look after its engines that badly?
76042: Here it is at Cricklewood in September 1962, a couple of months after its move from Neasden
https://www.flickr.com/photos/gricerman/16047967049
@SHLR – a lot of transfers, especially at the end,were on paper only. The answer to your question is in timbeau’s photo. After about 1961*, nothing seemed to be cleaned or maintained, on the assumption that they could be – and were – scrapped once they became unuseable. A friend who started work as a traction apprentice in the early sixties recalls travelling on the footplate of an absolutely clapped out Black 5, which expired at Woodford Halse,where the inspector took one look at it and condemned it on the spot, thus causing the train to terminate also – a not uncommon experience.
+some sheds were notoriously worse than others, Neasden and Cricklewood seemed especially bad.
@timbeau -and yet some optimist has replenished the tender…
Train Braking:
The trouble with many trains is that much of what happens under the train is automatic and beyond the driver’s control. When the driver calls for brake by moving the controller handle into a braking position, all sorts of stuff is going on under the train. For example, the rate of braking on each car is adjusted automatically, depending on its weight. Of course, the weight varies from station to station as passengers get on and off so the way the train handles at each station is different, despite what the engineering would have you believe.
Engineers will tell you that the brake rate should be the same because it is adjusted for the changing weight but it doesn’t work like that. Actually, the weight affects the speed achieved, although the rate of acceleration is also adjusted for weight, the top speed will usually be lower on a fully loaded train. Then, the speed is affected by time of day because of voltage fluctuations on the power supply, particularly a DC supply like the Underground’s. There is less voltage at peak times (more trains running) and more at off peak times (less trains). So, train speed varies at each station approach.
Now, drivers don’t normally brake according to speed, they brake at a fixed point on the approach to each station, assuming they haven’t been checked by adverse signals. So, if the approach speed is out by just 5mi/h, the train braking performance will appear to be different. You tend to stop short at peak times and long at off peak times and have to adjust accordingly.
Then there’s the preference regime. This is where the brake system prefers to use the motors for braking (dynamic braking – rheostatic or regenerative, depending on the design or response) but it needs friction braking as a back up and to add braking effort if the dynamic brake is insufficient.
When the driver calls for brake, the dynamic brake is given preference (friction brakes wear so they come second) and, if the rails are slippery and the wheels start to slide, the dynamic brake will give up, the wheels will start to rotate again and the friction brake will take over. This will simply lock the wheels again. Either way, you (on the front) are stuffed and in the lap of the leaves.
A feature of the swap between dynamic and friction braking is the lag when the brake control unit realises that the dynamic brake has gone and it needs friction to make up the loss. The driver feels the brake release, even though he (she) didn’t change the controller position, and then reapply and then lock again.
Already mentioned (@Anne) is the rate of brake feed up. Safety considerations would tell you that the quicker the brake goes on, the better. Fine if the adhesion is OK but not good if the rails are greasy. Feed up rates on many brake systems are too high for poor adhesion and effectively design in wheel slide. Incidentally, release rates are too high too, making stopping with a jerk inevitable. Feed up is critical. These needs to be more design effort in getting this right.
So, all in all, you have an unpredictable mix that means the driver has little idea how the train will behave at each station. Flats are inevitable in bad conditions, no matter how hard you try to avoid them. With brake systems designed to standards rather than real life, the sort of difficulties we see today will continue. That said, I wonder how the S Stock is performing in these conditions. Do they have sanders? Do they work?
@graham H
Had things turned out slightly differently the 760xx locos seen at Neasden might have been these………
http://railphotoprints.uk/img/s5/v127/p487430659-3.jpg
Ex Picc Driver……nicely put.
I think, but am open to contradiction having only driven them on test tracks, that more modern trains with electronic control of both the dynamic and friction brakes are more consistent and predicable.
Re SHLR,
Standard are reporting that the RMT are currently estimating February and TfL early Jan at best for resolution.
64 of 86 currently operational (79 required for peak service).
Assuming the easiest to sort units have been fixed first, the TfL estimate looks sensible assuming there are mostly the worst units left to fix and they have enough new wheelsets available where needed.
@Ex Picc Driver – Absolutely well put, sir. I can’t think of anything more to add to your excellent response!
As for S Stock performance: When they were first introduced to the line, the braking took a bit of getting used to, so earlier seasons meant we did have a lot of S8s with flats to begin with. You are getting drivers who may have driven with the Westinghouse brake for many years to now adjust to a braking system where handling-wise, you keep the brakes on and if slippage should occur, you let the WSP do the work. This unnerved a lot of people to start, but as the years have gone by and people come to accept that electronic control of both dynamic and friction brakes are more consistent and predictable with something like the S stock, this year on the Metropolitan we only had two reported trains with flats that have required withdrawal from service.
Emphasis every year has been to start braking much earlier than is required, with a lighter application, so should the WSP be activated, there is plenty of braking distance to account for the low adhesion without having to throw all the brakes on. This has held us in good stead so far, as I say, and in conjunction with:
1) vegetation works since last winter, especially north of Harrow
2) our Rail Adhesion Train running an extensive path to all branches, with a more rigorous underframe cleaning regime (to mitigate against signals/points failures where blobs of sandite may fall on points)
3) a better timetabled Network Rail Rail Head Treatment Train (RHTT) (that no longer blasts our freshly-laid sandite away as it did with previous timetables) and
4) the continued of implementation of a temporary speed restriction (operating at 40mph instead of 60mph in areas where the downhill gradient is pretty steep)
we have been able to mitigate a lot of the leaf fall effects.
As for sanders: Sanding equipment on S Stock will *fingers crossed* be going live for next year, not that we’ve been in that much of a hurry for it, because the Metropolitan line has fared fairly well for so many years without it. Part of the problem with sanders is working out the algorithms for the train to sand at what would be considered the best time. We’ll get there.
Timbeau
Actually, not an EM1 ( Class 76 ) as they were freight locos, but an EM2 ( 77 ) would have been nice.
Electra; Ariadne; Aurora; Diana; Juno; Minerva; Pandora
I’ve been hauled by all of them (!)
Retro pictures HERE
Anne:
Rail Adhesion Train @ Ealing Bdy
@Greg
Had BR adopted 1500V as standard and extended the electrified former GCR network, it would surely have carried freight too!
But as the picture of “Jason” I linked to illustrates, fourteen EM1s (not EF-, be it noted!) also had steam heat and, in all but one case, names as well. So I have travelled behind Pluto!
Drifting back in the general direction of on topic, they also had regenerative braking, which proved rather more reliable on the Woodhead line than the contemporary Metadyne system did on the Underground.
@timbeau: Be glad BR didn’t adopt 1500V DC… As they have found out in the Netherlands, these days, the frequencies and train lengths required (8 coach double deckers). The power supply really can’t cope! This is already after doubling the contact wire decades ago to stop them from melting.
So they are now engaged in a long term project to convert to 25kV AC, so all rolling stock being procured these days, must be dual voltage or can be modified to be so.
Anyway back to the topic at hand…
Assuming the Uxbridge Branch is particularly susceptible then improved vegetation management and efficient rail treatment both have a role to play. I would prefer that the District line takes over the Uxbridge Branch, swapping with Piccadilly which would run into Ealing Broadway. Campaign is driven by the need for level access (a good start to end stepping up/ stepping down on shared platforms). There are other benefits. An extra batch of S stock would be needed. However this would free up 1973 stock for strengthening elsewhere. In relation to the article this would significantly reduce the Piccadilly surface track, particularly for the low frequency section.
I am getting to the point where I feel like banning and deleting suggestions about swapping the Ealing Broadway and Uxbridge branch between the Piccadilly and the District. It is not that it is a stupid idea as such but that it comes up with monotonous regularity.
Short answer (repeated ad nauseum). It doesn’t address the need to reduce the number of branches on the District line whereas in future the upgraded Piccadilly should be able to cope with having an extra branch. This is just one reason why a swap is not on London Underground’s radar.
Almost back to topic … There is no fundamental problem with the Piccadilly line serving the Uxbridge branch – a well-established traffic flow. The fact that there are problems with 1973 stock trains but much, much less of a problem with S stock trains suggests the problem is not a fundamental one. By the time you finished getting distracted swopping the branches around (non-trivial) the first of the new trains Piccadilly line trains would probably be available and these could be prioritised to serve the Uxbridge branch.
I hope this isn’t off-topic, but is TfL pushing for a unified deep tube system so that trains will be interchangeable should incidences put a significant amount of trains on a line out of service, then extra rolling stock from other lines can cover the line affected?
@tyteen4a03
We used to have that, more or less, from 1940 (when the Central Line was converted to the LER standard electrification system) until 1967 (when the Victoria Line, with its special signalling, opened). Since then, each new project seems to have had bespoke signalling so very few types of stock, if any, can now work away from their “home” line. NTfL may be a step back towards the ideal, but it will be a long time, if ever, before full interchangeability is restored over all seven deep tube lines. The Victoria Line, which currently has the newest rolling stock, may well be the last to succumb, but we are probably looking at the second half of the century.
The surface lines, because of the amount of shared track, have always had to have more commonality. Indeed, electrification and signalling were about the only things the District and the Metropolitan were ever forced to settle their differences over, although even the method of electrification had to go to arbitration!
timbeau
And the completion of the “circle” which, IIRC required if not an Act of Parliament, a summons before the Bar of the House, to get Watkin & Forbes to explain themselves (!)
An informed source told me yesterday that a directive has been issued from on high in LU that a 3-car set of 1973 Stock is to be converted into a Rail Adhesion Train (RAT) for the Picc. It is hoped to be ready for next year’s leaf fall season.
@tyteen
I am reminded that tube stock could also work on surface lines (although obviously not vice versa) as evidenced by the use of such stock at various times on the (ex-Metropolitan) Northern City, East London and Stanmore branches, and on the (ex-District) Hounslow and Rayners Lane branches, in most cases alongside surface stock.
Tyteen etc….”back in the day”, as they say, you could move a train from underground line to underground line and, provided it would fit, all you needed to change were the maps and destination blinds. The variety of stock on the ELL was testiment to this as was the 1972mk2 stock which worked Northern, Jubilee and Bakerloo. Today it’s not just signalling systems that make trains hard to move, but other electronic sytems such as the passenger information systems, especially if these are triggered by a wayside system. There are further complications if the trains are owned or maintained by a third party. It can all be done but not at a moment’s notice.
NTfL, now Deep Tube Upgrade, is hoping for standard trains but there will be at least two different train lengths (some of the car bodies themselves might have to be different lengths to maximise use of the available space, not just different numbers of cars), and the most complex electronic interfaces ever seen on a tube train. Also it can’t be assumed that the signalling will be the same on the three lines as has already been mentioned. The main aim of common trains from the same supplier was for economies in procurement, design, production, training and spares.
On moving trains between lines, it might also be worth suggesting that, given the inevitable complications of doing this, it might be better to spend any available money and effort on fixing the original problem (wheel flats, say) rather than spreading the problem more thinly over the whole underground system.
One of the reasons the London Underground is supposed to be “easier” to operate than other rail systems is the extent to which problems on a given line can sometimes be confined to that line, rather than rippling out to Horsted Keynes, Milton Keynes and Milford Haven.
Ex Picc Driver,
Well let us hope that the “new” Rail Adhesion Train doesn’t develop flat wheels. If it is made up from ex-1973 stock this, to my naïve mind, has the same mentality that puts flood monitoring centres in disused tunnels in central London close to the Thames.
Re PoP,
May be scrapping the last of the ill fated ’83 stock recently was a bad idea.
Was the 7/7 damaged Picc unit (e.g. half a train) ever repaired (the SSR stock was)? if not then there will be a spare ’73 unit gathering dust most of the time so that makes sense.
I have not contributed for a while but it is good to see that LR still maintains the tradition of explaining in simple terms what is a complex interaction of circumstances. I propose to use your material as a case study that I shall be presenting next week of causal and contributory factors in systems failures.
per PoP: “Well let us hope that the “new” Rail Adhesion Train doesn’t develop flat wheels. If it is made up from ex-1973 stock this, to my naïve mind, has the same mentality that puts flood monitoring centres in disused tunnels in central London close to the Thames.”
It was actually a basement rather than a disused tunnel but it was only two streets from the Thames and, in late 2000/early 2001, Defra’s then new Departmental Emergency Control Centre could not be used for its first emergency (extensive flooding – !) because – you guessed it – the DECC was itself flooded out.
The DECC was relocated to the eighth floor after that, where it was “wittily” (sic) referred to as “The Bunker”!
@ Ronnie MB – you couldn’t make it up could you? (flooded flood control centre) 😉
@ Ngh – the remaining 83 stock at South Harrow was in appalling condition and way beyond saving. I believe that the damaged 73 stock is still around (others will be able to confirm) but whether LU will use that remains to be seen.
Ronnie MB,
I was thinking of when the flood control centre was located in a portakabin within Holborn tramway tunnel. It was still there a few years ago when I visited and even had a sign on it stating it was the flood control centre. I suspect it is still there now.
Walthamstow Writer, ngh
I tend not to follow the finer details of rolling stock but I am pretty sure the 73 stock order included half a train for the Aldwych shuttle so there should be half a train spare unless 7/7 put paid to that.
Thinking about it further, if the main problem is the Uxbridge branch then you don’t need 1973 stock and old D stock would do fine. It should even be able to run to Acton Town which I believe is permitted – but not in passenger service and there are severe speed restrictions under bridges.
@PoP
“I tend not to follow the finer details of rolling stock but I am pretty sure the 73 stock order included half a train for the Aldwych shuttle”
It did
” so there should be half a train spare unless 7/7 put paid to that.”
It did.
The Piccadilly Line has also acquired some extra mileage since the 1973 stock was introduced, namely the Terminal 4 loop and the Terminal 5 extension.
@PoP: To be fair to LUL, it was the GLC who put their flood control centre in the Kingsway Tram Tunnel – the Underground located theirs much more sensibly under Hampstead Heath.
As an a side There are still two sets disused tracks next to Rayners Lane Station
They are not connected to the running Lines however
There was a suggestion of using D Stock on the Rayners Lane branch. All perfectly feasible in the past but not in today’s risk-averse society. Problems include – No Correct Side Door Enable or CCTV for the trains concerned and incompatible platform/floor heights, plus speed restrictions under certain bridges.
@ PoP – yes it did include 1/2 a train. IIRC the max train fleet on the Picc was 86.5 trains from which the contractual train declaration had to be made. I suspect it may well have changed since I was last involved in the detail. I see the new Picc timetable needs 78 trains which is around the 90% mark of 86 trains so looks like the old parameters still apply. And to clarify a question from Malcolm much further up the thread the night service requires 21 trains in service – confirmed in the WTT and also on BBC News today.
As built the 1973 stock also included one six car train (two double ended units) which was for many years a test bed for solid-state traction equipment- known as the ETT (Experimental Tube Train). One unit had Westinghouse equipment and the other GEC (we had our own indigenous electronics industry back in the 1970s!) . Both were converted to standard spec in the late 1980s to provide the additional train needed to run the extra mileage when the Terminal 4 loop opened.
ah – two different flood control centres in equally sensible places – fools seldom differ
@timbeau et al, the other thing about 73TS is that it is not reversible – there was no need for it to be reversible before the Heathrow loop was constructed. As a result you can end up with two workable 3 car units in a depot that you cannot form into a train.
I think D stock is the same. From memory C Stock was reversible. The limitation was on the number of coupler studs from memory – there were not enough to make the connections symmetrical and thus make the units reversible.
With regard to the comments about the flood control centres being located in ‘vulnerable’ locations the original reason for the control centres was to close the gates if bombs were about to be or had been dropped in the Thames thus protection against bomb blasts was more important than protection against flooding.
Not sure what the function of the control centre in the Kingsway Tunnel was but I suspect the work Crossrail have done has removed that.
With reference to the resilient wheels there was also a plan (from a Chief Mechanical Engineer) who shall not be mentioned to use wheels with polyurethane treads. There were a pair of modified wheel at Acton for years with red treads. Needless to say the Chief Signal Engineer said NO ! Also I don’t think there was any consideration on how the train body would be earthed if the wheels had insulated tyres.
Still polyurethane wheels did a great job on escalators repalcing the cast steel wheels.
@john m
Indeed, C stock was reversible, and was reversed by certain workings on the third side of the Kensington or Aldgate triangle in order to even up wear on the wheels. A stock was also reversible to allow operation on the Watford triangle. D stock is not reversible, nor was R stock or most Tube stock (1972 and 1992 being, I believe, the only exceptions)
@timbeau – for the avoidance of doubt, you mean C69 and D70 stock presumably; prior to that, the original pre-WW1 C and D stock was handed as was most, if not all, District stock, with designated A and D end cars.
Indeed, C69, C77 and D78 stock is what I meant.
Sincere thanks to the author for an informative and comprehensive survey of the factors which converged to banjax the Piccadilly Line services for the best part of a month.
Acknowledging the incongruity of TfL pointing frustrated ‘customers’ towards this article, rather than its own non-commital batch of weekly prepared ‘stakeholder management 101’ declarations for a meaningful resumption of ‘good’ service, TfL might of dissipated some of the (rightful) customer aggro by flagging this article, along with the usual caveats…
The very informed comments also provided much low-level insight; many thanks to all for these.
@JohnM: Not sure what the function of the control centre in the Kingsway Tunnel was but I suspect the work Crossrail have done has removed that.
I think that both flood control centres were effectively abandoned once the Thames Barrier was completed. The barrier’s own control centre is in a tower overlooking the gates.
And I wasn’t being ironic when I said Hampstead Heath was a sensible location – as deep underground as you can get on the tube network (for bomb protection), yet also 50 metres-ish above sea level.
Daft question perhaps but at one time there was a bit of a panic when it was discovered that the Picc stock was riding on the bogie vertical bump stops under heavy load. This possibly gave a false reading from the load/weigh transducer to the brakes, making the car appear lighter than it actually was.
The effect on the car(s) concerned is unlikely to cause wheel slide, but the extra momentum of an under-braked car obviously was transferred to the rest of the train. I’m guessing that the problem was fixed years ago, though.
Any comments from the brake wizards there?
Doubled up word:-
They seriously affect affect movement
[I have corrected the affected text. PoP]
TfL’s internal investigation was surprisingly published, given the warts’n’all content. Addressed most or all of the issues raised above. Happily all the short-term recommendations were acted upon and there were no such problems in 2017.
http://content.tfl.gov.uk/piccadilly-line-leaf-fall-report.pdf