Gauges and Wheels

by Michael Frewston

  These trains ran on Lignes à Grande Vitesse (LGV), which were built to far higher standards than even the best traditional lines, and involved very low tolerances in track irregularities. They went over hills rather than around them – curves are anathema to high speeds, but relatively steep gradients are no problem with such powerful trains.

  Gradually increasing the speed as experience was gained, 380 km/h was reached in 1981, 408 km/h in 1988 (beating the German ICE-V record of 407 km/h achieved a few months earlier), 482 km/h in 1989, 515 km/h in 1990, and culminating in an astonishing 575 km/h in 2007 (574.8, to be precise), a record for conventional rail vehicles unlikely to be broken for a long time, if ever. Only maglev trains go faster (and then not by much – the world record, achieved in Japan in April 2015, is 603 km/h, just 28 km/h faster (and eight years later) than the record-breaking TGV).

  The remarkable thing about all these records for conventional trains is that they were achieved on 1435 mm Standard gauge track, albeit on dedicated lines built to the very high standards described above (with relatively steep gradients instead of anything but the gentlest of curves). The only other concession to such very high record-breaking speeds was the use of larger than standard wheels to keep rotational speeds within bounds, as well as upping the voltage from 25 kV to 31 kV.

  A TGV, in standard operational specification, has achieved a speed of 380 km/h. The French TGVs are currently routinely achieving speeds of 320 km/h in everyday scheduled passenger service. Plans are in hand to increase this to 330 km/h.

  The TGV’s successor, the new AGV (Automotrice à Grande Vitesse) concept, will increase this to 360 km/h, still running on the same Standard gauge LGVs. Will the AGV eventually run at 400 km/h?

  Europe is not the only part of the world achieving high speed rail travel on a daily basis. Japan can be considered to actually have started the movement towards very high speed trains in regular use, when it opened its first Shinkansen, or Bullet Train, in 1964, in time for the Tokyo Olympics. At that time, the Shinkansen actually held the world railway speed record in scheduled passenger service, of 210 km/h, and reached a speed of 256 km/h on a special test run in March 1963.

  Since then, these trains in regular service have been achieving speeds of 240 km/h, subsequently increased to 260 km/h, day in and day out, for nearly 50 years. Japan has now upgraded these trains to 300 km/h, and is looking to match the Europeans, at 320 km/h. These speeds have enabled a Japanese Bullet train to achieve a world record of 4065 km in one day. Unlike the rest of Japan, the Shinkansen run on Standard gauge tracks.

  China has already built over 8000 km of Standard gauge high speed lines, of which 2000 km have a maximum speed of 350 km/h (the line from Beijing to Shanghai is designed for an amazing 380 km/h), and the remainder are rated for 250 km/h or higher.

  A fully operational train in China (i.e. not a special test train, as was the case with the record-breaking TGV in France) has achieved over 487 km/h, in January 2011, thus paving the way to another round of increases in the speeds achieved by normal trains in everyday operation.

  [Note: previous safety concerns regarding China’s high speed lines, together with a major accident in July 2011, have temporarily curtailed these very high speeds until China re-evaluates the tendering and construction of its railways. China’s trains are now more or less back to normal.]

  Even at gauges narrower than Standard, speeds with everyday passenger trains are being increased far beyond what was thought not only achievable, but safely achievable. Most of Southern Africa, Japan and parts of Australia use a track gauge of 1067 mm. Trains in South Africa were limited for many decades to a maximum of around 90 km/h, this being considered to be the safe maximum at this narrow gauge. Japan and Australia imposed similar speed restrictions.

  Japan, South Africa and Australia have in the last few years drastically increased the maximum speeds on 1067 mm gauge track, up to a maximum of 160 km/h in normal scheduled service, as long as that track is built to the same standards and tolerances as equivalent Standard gauge track. The Queensland Rail Tilt Train is scheduled to touch 165 km/h on occasion and maintain 160 km/h over long stretches (albeit this was temporarily reduced following a derailment at high speed).

  Most remarkable is the fact that this same train reached a speed of 211 km/h, a speed that was achieved on a normal stretch of track (i.e. not a dedicated high speed line). That speed is easily of the same order of magnitude as that achieved in everyday running by Standard gauge trains on non-dedicated high speed lines elsewhere – e.g. the East and West Coast Main Lines in Britain, the North Eastern corridor from Washington to New York City in the USA, and any number of ‘classic’ main lines in Europe.

  The South Africans have however taken things an order of magnitude further. South Africa is of course also a 1067 mm gauge country. A South African Railways Class 6E1 electric locomotive achieved the remarkable speed record on 1067 mm gauge track of 245 km/h on 31 October 1978 – a speed that was achieved by the French as an absolute world record just 24 years earlier on Standard gauge lines.

  So where does that leave Brunel with his philosophy that only his 2140 mm gauge track could safely permit high speeds not attainable with narrower gauges?

  Brunel always argued that his wider gauge would allow much higher speeds in safety, together with physically larger trains that permitted increased carrying capacity, than was achievable with narrower gauges, including Standard gauge. That was no doubt true in his day, and perhaps so for many years until the second half of the 20th century.

  But is that still true today? A record speed approaching 600 km/h, as achieved by the French in 2007, surely proves that extraordinarily high speeds are quite possible by both test and regularly scheduled trains on 1435 mm gauge track. In fact, is 600 km/h a speed that is actually useful in everyday railway operations, even on dedicated high speed lines? Or even 500 km/h? It could be argued that speeds beyond, say, 350 km/h to 400 km/h already exceed practical constraints imposed by factors other than the track gauge.

  For a start, the time savings become incrementally less and less as speed increases. A segment of, say, 300 km (about the distance from London to Leeds) will take 1 h 12 min at 250 km/h, 1 h exactly at 300 km/h (a saving of 12 min – reasonably worthwhile), 51 min at 350 km/h (a further saving of 9 min – barely worthwhile), and 45 min at 400 km/h (a further saving of just 6 min – rather marginal). Going from 400 km/h to 500 km/h – a jump of 100 km/h – would produce a saving of just a further 8 minutes, something that will have cost enormous amounts of energy and infrastructure costs to achieve, compared to slower speeds.

  As can be seen, the 350 km/h mark is the point where further incremental time savings can be said to be hardly justified, especially when compared to the incremental increase in costs. (Even the fastest commercial maglev operation, in Shanghai, China, has a maximum speed of 430 km/h, well below the maximum achieved to date, while even though the Japanese have achieved almost 600 km/h with their maglev trains, they have admitted that a maximum of 400-450 km/h is a much more practical limit, and then only on specific routes where the speed advantage can be maximised.)

  There are also physical constraints. Air resistance, braking performance in regard to headways (distance between trains on the same track), and increasing dead weight are three such constraints.

  Power consumption increases exponentially with speed. The proposed 360 km/h speed of a full length fully loaded AGV will push to the limit the capability of the train’s pantographs to draw sufficient power from the overhead catenary. The record-breaking TGVs in 2007 used two locomotives and just three coaches, as that low gross weight still required the maximum power that the two locomotives could provide, limited if nothing else by the electrical capacity of the overhead catenary.

  Speeds of this magnitude will require such large headways as to impose a maximum capacity on the line, in terms of the number of trains running at any one time. If a line was heavily used, these increased headways could limit the use
fulness of such speeds in practical terms.

  Likewise, wheel diameters are another practical constraint – the 575 km/h record (and all of the previous TGV records) was achieved only with the use of greatly oversized wheels. Otherwise centrifugal forces due to high rotational speeds impose excessively large track forces, and would likely simply blow the wheels apart – not to mention that the smaller wheels currently in normal use limit the size of brakes (and hence braking performance).

  All these factors are not improved one bit by increasing the track gauge.

  In fact, Brunel’s physically larger trains could increase both dead weight (which would adversely affect braking performance and power consumption), as well as air resistance, to the point where electrical power requirements could very well exceed that which conventional pantographs and overhead catenary could manage at these very high speeds; while larger wheels would compromise vehicle design in other ways (such as reducing available floor area for seating).

  Finally, even dedicated high speed lines are being used to carry some slower trains in order to improve their economics – 220 km/h commuter trains (and even, starting in November 2011, 160 km/h freight trains at night) on HS1 in Kent, UK, and 250 km/h postal trains in France. The use of these lines for these slower speeds means that they cannot be completely optimised (such as incorporating very steep gradients rather than introduce more curvature) for speeds of much beyond, say, 350 km/h.

  In other words, the 1435 mm track gauge is no longer the limiting factor in the railways’ quest for speed – technically or economically.

  Maybe the British Parliament, in outlawing Brunel’s broad gauge in 1846, got it right after all, even if they didn’t know it at the time.

  WHEELS AND RAILS:

  This book would be incomplete if there wasn’t at least a brief discussion on the subject of the wheel-rail interface. For while it is critical that the rails are maintained at the correct distance apart, it is equally critical that the wheels running on those rails are also the correct distance apart. Not only that, the shape, or profile, of the wheels must also be correct.

  The above statements are of course ‘givens’ – if they weren’t we couldn’t have railways that work as they do! But there are many technical considerations that come into play when we have many types of railway vehicles running, often at very high speed, over what may be very complex trackwork, and often over very large distances in more than one country.

  First, any specified track gauge, and the railway wheels that run over this track, must allow for some tolerances. Tolerances vary between many jurisdictions, but there has to be a minimum level of tolerance that usually permits railway vehicles to traverse tracks of a slightly different gauge with no problems – e.g. Russia’s 1520 mm gauge vs. Finland’s 1524 mm gauge, where trains from both countries travel over each others’ rail systems. Likewise with some of the early railways in Britain – trains ran without difficulty over both 1435 mm and 1448 mm gauges. There are other places where adequate tolerances mean that minor differences in gauge, or even where rails have worn away to a certain extent (particularly on curves), do not cause a derailment on a railway, or prevent interoperability between two different railways.

  The Federal Railroad Administration in the USA specifies, for 1435 mm Standard gauge, a minimum of 1422 mm (13 mm less than Standard), and a maximum of 1460 mm (25 mm more than Standard). These tolerances hold for speeds up to around 100 km/h – above that, the tolerances get a little tighter.

  Even those tolerances may be too much on some railways – when a London Underground engineering train derailed a little while ago, it was found that excessive lateral movement, due to inadequate or incorrect maintenance, of the outside rail on a curve permitted the inside wheel of the train to fall into the rail bed between the rails. The movement of the outside rail was less than 25 mm, yet it allowed a derailment to occur.

  Check rails on curves would prevent inside wheels from falling into the rail bed when the gauge is wider than specified. While the London Underground uses check rails on most curves (but not the one where the derailment referred to above occurred, as it was not a particularly sharp curve), the Docklands Light Railway (DLR) in general does not. It would be interesting to know the DLR’s thinking behind the omission of check rails on an elevated railway that has an abundance of sharp curves (in mitigation of this to some degree, the DLR does use wheels of a different profile to normal that are said to better cope with tight curves, albeit with some loss of straight line stability at higher speeds).

  During both the First and the Second world wars, both British and American Standard gauge railway locomotives were shipped into France and other Allied countries, as part of the fighting against Germany. It was assumed that they could start running without problems, and this they obviously did. (The US locomotives were modified export versions, involving smaller outlines that would fit the European – and even the British – loading gauges. See later chapter on loading gauges.) Wheel profile differences at that time were minor, and had little effect, most problems in this area involving a few of the US locomotives, which did tend to foul frogs on wider angle facing turnouts in the UK (the American wheels had wider treads and slightly smaller back-to-back measurements).

  Certainly when the first cross-Channel boat trains started running in the 1930s, prior to World War II, the French railway authorities were extremely concerned that the International Sleeping Car railway carriages had the correct wheel profiles for running on French tracks. After extensive investigation, it was found that they did, and within the permitted tolerances, and so the trains were allowed to cross the Channel and continue their journey on French rails.

  There were the same considerations when the British Royal train was shipped across the Atlantic and undertook a tour of America and Canada in 1939. North America, while using the same gauge as Britain, had certain track peculiarities and rail profiles at turnouts and shallow angle crossings that could have been a problem for British wheel profiles (one of the peculiarities of American – and Canadian – trackwork was the use of a flange on outside of the ‘frog’ of low-speed turnouts in freight yards and the like, that operated on the outside of the wheel crossing the frog to guide the wheels through the turnout – rather than use check or guard rails operating on the inside of the opposite wheel, as is normal practice).

  Fortunately, the Royal train’s wheel profiles were within North American tolerances, and again a train from one continent was able to travel successfully over the tracks of another without derailing through turnouts and crossings.

  But things can get interesting when you start to consider trams (streetcars) and trains, as the wheel profiles of these two types of vehicles can be quite different. Now this is not normally a problem, as usually trams run on their own dedicated and exclusive routes and trackage, with no connection to mainline trackage. But not always. There are now many cities and areas, especially in mainland Europe, where trams (often now called Light Rail Vehicles) and trains actually share the same rails, at least for parts of their journeys. And the incompatibilities between tram and train wheel profiles consequently have to be dealt with. Let us look at typical tram and train wheels.

  The above diagram shows a cross-section of a grooved tram rail, and it can be seen that, while a tram wheel fits in it with plenty of clearance between the wheel and inside rail flange (thus allowing for that all-important lateral play), a train wheel is a tight fit. As the rail surface gets worn down, the train wheel flange would then start to ride on the inside flange of the rail, which would likely lead to frequent derailments.

  This difference in wheel profile becomes even more of a problem at switches and crossings, where you don’t want lateral play – and this is nearly all due to the difference in the back-to-back measurements, and flange depths, between the two profiles.

  The larger train wheel profile has a smaller back-to-back measurement (the distance between the two inside faces of the wheels) than a tram wheel – typica
lly 1360 mm (generally +/-2 mm) for the train wheel, compared with 1380 mm for the tram wheel (although these can vary quite a lot between different systems and in different countries). That 20 mm difference causes problems at switches and crossings, tight curves and other locations where, on a rail line, there are check rails.

  Check rails, as their name implies (called guard rails in North America) provide additional guidance at breaks in the running rails (such as at a switch or turnout where two rails have to cross each other), as well as on sharp curves where there would be a risk of derailing. Check rails act against the inside face of the wheel – and are positioned such that they will just contact those inside faces for train wheel profiles (when the outside wheel flange is in firm contact with the outside rail) and thus provide the necessary additional guidance.

  But the 20 mm wider back-to-back measurement of tram wheels means that there would be, each side, a gap of 10 mm between the inside of the flange and the check rails (when the outside wheel flange will be in firm contact against the outside rail), rather than the inside flange be in contact with the check rail as a train wheel would be, thus rendering the check rails at least partially, and likely totally (depending on how worn things are), ineffective. Without the additional guidance provided by check rails, derailments are more likely.

  Europe has come up with a solution to this problem that allows trams to run on train tracks (though not the reverse). Most European railways already have a key advantage in that their track standards require the tops of the check rails to be at a higher level than the surface of the running rails (unlike Britain and many other places in the world, where the top of the check rail is usually at the same level as the surface of the running rail).