|
|
|
© R G Latham 2025. |
In order to meet the transport needs of modern highly urbanised areas of the world there has been a reawakening of interest in high speed railroads and their possible competitors in the last decade. Evidence of this is provided by the New Tokaido line in Japan, developments in France, Germany and Italy, and the North-East Corridor Project in the U.S. In the U.K., the success of the BR Euston - Manchester and Liverpool electrification has demonstrated that there is a positive public response even to quite modest increases of speed between city centres.
Recent engineering research into the dynamics of railway vehicles together with the application of new, but available, technology from other industries has stimulated a re-appraisal of the prospects for rail transport in the future. This has led to the concepts embodied in the BR Advanced Passenger Train (APT) which represents a step towards meeting some of the transport needs of the immediate future.
The APT has been conceived as a high performance railway train capable of running at high speeds on existing tracks and with existing signalling. It is intended to provide a competitive form of inter-city transportation which makes effective use of existing assets, represented by the very extensive and underused rail network in this country. The capital cost involved in providing this significantly better service should thus be principally that of the train.
Initially, the APT is being designed to run at speeds up to 160 miles/h and the average speed between typical city centres can be well over 100 miles/h. The importance of this lies in the fact that the public response to improved inter-city railway service is particularly influenced by reduced travel times.
Whilst average speed is the most important factor, comfort, reliability, frequency of service, fares policy, and safety are, of course, also important. However, history shows that passenger traffic is usually attracted to the fastest mode of transportation available, so that speed is a necessary but not a sufficient condition for survival of a particular mode.
Increase in the average speed has two particular advantages which are worthy of recognition. Firstly, it gives the railway train a definite speed advantage over its present major competitor the private motor car for distances of over about 70 miles; one objective should be to make the railway train more complementary to the private car rather than so directly competitive. Secondly, it increases the radius from any major city centre from which it is possible to conduct a days business or shopping and travel all within one day. This implies a maximum journey time of about 2 hours, one-way. The last factor is relevant to the results of the BR London - Manchester and Liverpool electrification where of the large increase in passenger journeys three-quarters was newly generated rather than captured from other modes. The other quarter was captured from air - thus illustrating that the increase in average speed increases the distance over which rail can successfully compete with air.
It can therefore be expected that with sufficiently high average speeds, railroads can fill a gap in the spectrum of transportation systems which is not adequately catered for at present. In addition to speed, very high frequency is possible which then provides large capacity which, if used, can support the expensive fixed installations associated with track and signalling. The characteristics of this system are, therefore, ideally suited to highly urbanised regions such as the N. E. Corridor, and it is considerations such as these which accounts for the developments of high speed trains and their possible replacements for similar areas in Japan, Europe and the U.K.
The average speeds of conventional trains on most existing tracks are mainly determined by speed restrictions due to curves and other track features. Increasing power weight ratio in these circumstances does little to raise average speeds and so a common method of increasing average speeds is to re-align curves and otherwise eliminate the reasons for the speed restrictions. Civil engineering works of this type are inevitably expensive. An outstanding example of the cost of a completely new line built to a very high standard of alignment for 125 miles/h conventional trains is the New Tokaido Line which cost over $2m per mile. Half of this sum was for right of way and track. It is unlikely that capital sums of this order could be made available in countries where there are a relatively high number of routes radiating in all directions. Consequently, there is an economic limitation to the achievement of high speeds by track improvements alone. The same sort of difficulty of course applies whenever proposals for high speed ground transportation systems of unconventional types are made. In addition, there are further technical difficulties, leading to increased costs, associated with running conventional trains at much higher speeds than those common today. These problems, and others, stimulated the initiation some 8 years ago, of fundamental research into the dynamic behaviour of railway vehicles.
This research on the dynamics of railway vehicles at high speeds, carried out at BR Research Department, has made several important advances in suspension design possible. As a result of this work, it is possible to design railway vehicles to run stably at very high speeds and to negotiate curves at higher speed satisfactorily. The APT is intended to exploit these ideas and will have a maximum speed of 50% higher, and will negotiate curves at speeds 50% higher, than present-day conventional trains. Because the APT will run on existing track and with existing signalling, there will be relatively little capital investment on permanent way associated with its introduction. In addition, the APT embodies lightweight construction and is powered by gas turbine (or, where a route is electrified, by lightweight electric motors). The exploitation of these features depends on the successful resolution of the critical suspension design problem as mentioned above. Owing to light weight the effect of speed on track maintenance is offset. As well as low axle-load, low unsprung mass and stable running ensure that rail stresses and wear will not be increased.
The attractive economics of the APT therefore depend to an unusual degree on the novel features of the suspension system, and the basic concepts involved are now discussed.
It is perhaps obvious that the design of a new type of high speed vehicle must be based on a thorough understanding of the various dynamic phenomena that the vehicle will experience. Certain problems, such as the response of a vehicle to track imperfections or atmospheric gusts, are common to many types of vehicle, but a vehicle which is closely constrained to follow a predetermined path by means of a guidance system is subject to problems of stability. Such is the case for the railway vehicle. It has long been recognised that a major limitation to satisfactory high speed running of railway vehicles has been the hunting oscillation. Many of the measures adopted in conventional railway technology in order to control hunting ignore the basic fact that the tendency to dynamic instability is a consequence of the guidance offered by coned wheels. As a result, the full potentiality of this guidance is not exploited on conventional vehicles. The design of the APT suspension has as its starting point the recognition that guidance can be achieved by the use of coned wheels acting in conjunction with the tangential forces which are generated in the wheel-rail interface.
This form of guidance involves feedback, and a basic problem in suspension design of a guided vehicle is to resolve the basic conflict between guidance and stability. Because this problem is so fundamental to the APT it is worth explaining it in more detail.
The basic mechanism of guidance may be understood by reference to the behaviour of a single wheelset. If a wheelset is rolling slowly along the track and is slightly disturbed, there is a difference in rolling radii between each wheel due to the conicity. Because each wheel is solidly mounted on the axle, for pure rolling to occur, the subsequent path of the wheelset must be sinusoidal. Thus, when displaced laterally, the wheelset steers towards the track centreline but overshoots and the cycle is repeated. This is termed the kinematic oscillation, and was first described by Stephenson in 1821. An analysis by Klingel in 1883 showed that the frequency of the oscillation is proportional to speed and the square root of the conicity.
Carter pointed out in 1922 that, in reality, pure rolling cannot be maintained in such a motion because of the phenomenon of creep. When tangential forces are transmitted across the contact area between wheel and rail, elastic strains occur which when combined with forward rolling motion of the wheel cause fractional deviations in the motion. These deviations are referred to as creepages and the overall phenomena is termed creep.
Creep modifies the behaviour of a wheelset in the following way. For a typical wheelset, when restrained by lateral and longitudinal suspension springs, the response to a disturbance still consists of an oscillation at, approximately, the kinematic frequency. However, at low speeds, the oscillation decays and the wheels et returns to the centre of the track. This is because the suspension spring forces have generated creepages which tend to reduce the amplitude of the motion. If the speed is increased, the frequency of the kinematic oscillation also increases. Cons equently, the inertia forces increase with increases in speed. As the inertia forces generate creepages which tend to increase the amplitude of the motion, there is a speed at which there is a balance between the stabilising effect of the suspension springs and the destabilising effect of the wheelset inertia. At this speed, the oscillation neither grows nor decays and the speed is termed the critical speed.
Motions at speeds above the critical speed are limited by the wheel flanges and by slipping at the wheel treads and it is this limiting motion that is properly referred to as hunting.
With the inclusion of the creep phenomenon, the guidance mechanism of the wheelset can be described as follows. When a wheelset is laterally displaced when rolling along the track there is a change in rolling radii on the wheels on each side. Because the wheels are mounted on a common axle, and are therefore constrained to rotate at a common angular velocity there is a descrepancy in the velocities at the wheel treads. This discrepancy is equivalent to a creepage in the longitudinal direction and the resulting creep forces (which are proportional to the creepages) impart a steering couple to the wheelset tending to point it to the centre of the track.
This steering couple is directly proportional to the cone angle of the wheels so that a large cone angle is preferred for steering round curves. Unfortunately, the critical speed above which instability occurs is inversely proportional to the square root of the conicity. There is therefore a conflict between stability and guidance and the satisfactory resolution of this is one of the fundamental problems in designing the suspension for the APT.
Another aspect of the conflict between stability and guidance is presented by the behaviour of a wheelset on curved track. In response to track curvature, an unconstrained wheelset adopts a radial position on the curve, but is displaced outwards so that pure rolling occurs. In response to centrifugal or gravitational forces in the lateral direction a wheelset yaws and thus generates a reacting lateral force. As mentioned previously, however, elastic restraint is required to provide an adequately damped response of the kinematic oscillation. For a two-axle vehicle, or a bogie, this restraint is provided by longitudinal springs between wheelsets and frame. Consequently, on curved track equilibrium is established by additional lateral displacements of the wheelsets which cause the vehicle to take up a yawed attitude on the track. These further lateral displacements generate a yawing couple due to the longitudinal creep forces. If the creep forces are greater than the maximum friction available on the wheel tread, slipping will occur and the flanges must come into play. The aim of the APT suspension is to obtain guidance on main-line curves at high speeds from the wheel tread creep forces without the need for wheel flange to rail contact.
The basic concepts of stability and guidance as applied to a guided vehicle have been described and illustrated with reference to a single wheelset. Clearly a wheelset is a small sub-system of a single vehicle which in its turn is a sub-system of a complete train. The development of knowledge of the behaviour of this complete system has involved the detailed analytical study of appropriate sub-systems, digital and analogue computer studies of larger systems and the verification of the theory by experiments with models and full-scale vehicles in both laboratory and on track. Moreover, a wide-ranging examination of the consequences of the theory has led to the discovery of optimum combinations of suspension parameters which are necessary to resolve the conflict between guidance and stability for high speed vehicles. It is this background of recently acquired knowledge which is being exploited in the APT, and the APT in its turn will stimulate further improvements in the theory.
It has been shown that an important parameter in the stability and guidance of a railway vehicle is the conicity or cone angle of the wheel tread. However the treads of wheels are first manufactured they eventually wear to a hollow tread profile - further wear is then uniform and the profile is maintained.
In conventional railway practice the basic form of guidance is considered to be the wheel flange. Coning of the wheel tread was originally introduced simply so as to prevent the constant rubbing of one or the other flange against the rail which occurs with a cylindrical tread. The customary coning of 1 in 20 or 1 in 40 usually ensures a sufficiently high critical speed though this can only be maintained by frequent reprofiling of the wheel treads. This is because as the wheel tread wears the effective conicity increases thus reducing the critical speed. These rather low cone angles are less effective in providing guidance in curves and for conventional trains the basic form of guidance has always been considered to be the wheel flange.
The suspension of the APT has been designed so that stability will be obtained with a wheel tread profile similar to that of a fully worn profile. Not only will this obviate frequent reprofiling of the treads, but the higher effective conicity will improve the guidance of the APT on curves. Furthermore, because of the better conformity of the wheel-rail profiles, the contact stresses are reduced, and because reprofiling is unnecessary, the work-hardened tread surface is maintained thus increasing wheel life.
In the previous sections, the problems of guidance and stability have been discussed in general terms. In addition to achieving dynamic stability and guidance on curves, it is necessary to provide a suspension which provides a good ride for the payload and minimises stresses and wear in the vehicle and track structures. Thus an acceptable dynamic response to discrete track features such as curve entries or exits, switches and crossings and randomly distributed irregularities is required.
For predictions of vehicle response it is necessary to specify the shape of track in terms of curvature, cant and variations in cross-level and alignment. This, together with aerodynamic forces, provides the input into the dynamical system representing the train. The outputs are the stresses in vehicle and track structures and accelerations experienced by passengers, etc.
For a specific stretch of line, the track geometry can be specified by continuous records, but the description of a complete route or network can only be done on a statistical basis. Very comprehensive studies of track irregularities on existing BR routes have been carried out based on the idea that irregularities are continuously and randomly distributed. These measurements have been made with a specially developed track measuring machine and sets of power spectra derived from these results have been used in the design of the APT suspension.
An important application of these dynamic response studies has been in the formulation of rational loading cases for the structural components of the APT. As these design cases are further refined, they will make an important contribution to improving the structural efficiency of high speed railway vehicles.
In order to reduce the stresses experienced by the track, and to improve the riding characteristics of the vehicle, a low value of unsprung mass is being aimed at. This involves the consideration of unconventional wheelset designs. In addition, low axle-loads are being achieved. A major contribution to the achievement of low unsprung mass would be the use of resilient wheels in which the major part of a wheelset is elastically supported by the rims. Studies are being carried out to assess the value of this feature.
A further application of response studies is of fundamental economic importance and will provide information on the capability of wheels on rails to achieve speeds much higher than initially aimed for with the APT. By establishing relationships between loads in vehicle and track and irregularities in the track as affected by track usage and maintenance procedures, it should be possible to optimise the economics of the vehicle-track system. This is particularly important at high speeds, and there is an obvious trade-off between track alignment and suspension sophistication.
A vital feature of the APT is the ability to traverse existing curved track at significantly higher speeds. Guidance in curves has already been discussed in relation to wheelsets and it will be sufficient to state that the articulated configuration of the APT has been evolved in order to maximise the inherent guiding capability of the coned wheelset. As previously mentioned, it is intended that main-line curves will be negotiated at high speeds without flange contact, thus reducing the dynamic stresses and wear on the track and improving the quality of the ride. On sharp curves negotiated at low speeds, such as those in sidings and stations, the flanges will operate in the normal way.
The most important limitation to the speeds in curves is that of passenger comfort. Present speed restrictions on curves are set in relation to cant and curve radius so that the lateral acceleration applied to the passenger is limited to about 0.07g. Existing track cants are normally limited to about 100 mm (the height of the outer rail above the inner). The body of the APT will be tilted up to a maximum of 9° relative to the track about a longitudinal axis. As the total cant plus the cant equivalent to the lateral acceleration which can be comfortably experienced by the passenger (the 'cant deficiency') is thereby increased to 17° the speed through curves can be increased by √ 17/8 i.e. about 50%.
This increased speed in curves is, of course, still well below any limiting speed necessary for safety reasons on account of deformation of track or flange-climbing and derailment.
Because of the length of transition curves on existing track, a purely passive pendular type of tilt system is not suitable and an active system is being developed. In this system the body is tilted by an electro-hydraulic servo-system responding to sensors measuring the lateral acceleration of the body and other motion variables. In order that the transient accelerations, due to the rotation of the body, applied to the passengers are minimised and maximum use is made of the existing clearances a suitable apparent tilt axis has been chosen down the centre of the vehicle body.
By tilting the vehicle instead of canting the track, modifications to the track are avoided. Such modifications would be expensive and highly canted track would also be inconvenient when the same tracks are used by slower trains.
The braking system of the APT must be capable of stopping the train within existing signalling distances, and this means stopping from 160 mile/h within the stopping distance of existing 100 mile/h vehicles. For a given train weight, the energy that must be dissipated during a stop varies as speed squared whilst the mean rate of working or power varies as speed cubed. Even though the reduced weight of the APT represents a significant alleviation of the braking requirement, braking by purely frictional means by using conventional techniques does not seem to be an appropriate solution.
Consequently, a hydrokinetic brake is being developed which will be mounted on the axle. When filled with fluid a braking torque is developed in the same way as in the familiar engine dynamometer. Associated with each axle brake is a body mounted tank and radiator so that fluid, having been heated by the braking action in the axle, is cooled after being transferred to the radiator. As this form of brake is not effective at low speeds a friction brake is provided which has, however, a comparatively light duty.
In parallel with the development of the hydrokinetic brake, research and development work on alternative braking systems is being carried out. These systems include various kinds of both liquid cooled friction brakes and electric brakes.
Though the average braking rates of the APT may be higher than for conventional trains, the rate of change of acceleration will be reduced, and more effective use will be made of existing signalling distances, by an electronic brake control system. This will also provide protection against wheelslide.
The braking performance is ultimately limited by the maximum value of adhesion that can be exerted between wheel and rail. It is expected that the improved vertical and lateral dynamics of the APT will result in improved adhesion at high speeds, compared with conventional trains. In addition, various devices which clean the rails and thereby promote increased adhesion are under development.
In running on existing tracks the accelerating capability of the train exerts a major influence on average speed and this usually decides the amount of installed power. By adopting a light-weight form of construction it is possible to achieve a high power/weight ratio with a relatively modest size of power plant with consequent advantages in lower capital cost, operating cost and weight.
For the APT a form of light alloy stressed skin construction has been adopted which is related, in part, to aircraft structural practice. This, combined with the articulated configuration and light weight power plants, means that the structural weight per passenger seat is about half that achieved in conventional trains. Further significant reductions in weight would be possible except that the risk of overturning in an exceptionally strong side wind would be increased and the cost of the structure would be very much increased.
The configuration of the APT is articulated, as already mentioned, and the leading and trailing cars are power cars containing power equipment and drivers controls. The power cars have three axles and the trailer cars two axles. The trailer cars forming the intermediate cars contain the passenger accommodation. A high degree of environmental comfort will be provided with low levels of noise and with air conditioning, whilst present standards of passenger space will be maintained. In a typical case there could be two power cars with eight trailer cars with a total all-up-weight of about 250 tons. Each power car would be rated at about 2000 H.P. so that in this particular example the power/weight ratio would be about 16 H.P. per ton.
As light weight power plants are required, gas turbines are a natural choice. Because of the potentially attractive operating characteristics of the automotive gas turbine with heat exchangers which is now being developed, the design of the power car is such that various engines of this type, such as the Ford 707, can be fitted. These will be installed in a multiple installation in each power car, and an additional engine will be used for auxiliary power for services.
For electrified routes, an electrically driven APT can be provided.
A number of different power transmissions are being studied, including mechanical, electrical and hydraulic.
At higher speeds the airflow round a train assumes much greater importance to both design and operation. Both aerodynamic forces acting on the train and the airflows in the vicinity of the train can exert a significant influence on the motion of the train and on objects nearby. Aerodynamics as applied to aircraft is a highly developed discipline, but its application to vehicles operating on the ground is a challenge to both experimental and theoretical approaches. This is because the effects of ground proximity complicate both the actual flow pattern, and methods for its analysis. The train itself has a complicated geometry, its form being determined by many factors which are not exclusively aerodynamic in nature, and possessing by aeronautical standards an extremely high fineness ratio.
At high speeds aerodynamic drag represents a large proportion of the total resistance to motion of a train. Careful design of nose and tail and the avoidance of unnecessary protuberances are necessary to minimise drag and hence the installed power.
The behaviour of trains in cross-winds introduces a number of problems. If the vehicle is too light, overturning may be caused by a strong gust. For existing gauge tracks and with a limitation of axle-load, a knowledge of the statistics of extreme gusts at or near ground level and the choice of an acceptable probability of occurrence provides a minimum weight for a high speed train. In a steady cross-wind, there is the possibility of excitation of motions of the body on the suspension in response to vortex shedding and other periodic peculiarities of the flow round the body. In addition to the effects of the quasi-steady cross-flow, the dynamic response to lateral gusts may contribute as much to the acceleration level experienced by the passenger as the effects of track roughness transmitted through the suspension.
When a train passes another train or a lineside object, rapid pressure changes occur. These consist of a compression followed by a rarefaction corresponding to a head wave attached to the front of the moving train. A similar though smaller tail wave is attached to the trailing end of the train. When a train enters a tunnel at high speed there is an initial pressure rise as the air ahead of the train is compressed, followed by a gradual decrease in pressure as the pressure wave which is reflected from the tunnel mouth meets the train. Several pressure peaks may occur as successive waves travel up and down the tunnel.
In both the case of passing trains and tunnel entry the most important variables are speed, the transverse gap between passing vehicles or between the vehicle and the tunnel wall, the shape of the nose and the pressure attenuation between the outside and inside of a vehicle. A number of aspects of the whole problem are being studied at present. Theoretical methods of predicting pressure transients due to trains entering tunnels or passing other trains are being developed in conjunction with the University of Liverpool and the University of Leeds. Experiments with full-scale vehicles are currently taking place. The indications are that the pressures experienced by the APT will be no worse than existing trains at conventional speeds. This is partly due to the smaller cross-sectional area of the body and its tapered nose and also to the sealing of the interior of the bodies. This latter feature imposes requirements on the air conditioning system and on the air-tightness of the structure.
Another aspect of the aerodynamics of high speed trains is that involved in the airflows into and out of gas turbines, air conditioning systems and cooling systems. As each end of the train is powered and the train is bidirectional, the problems of avoiding recirculation are many. Wind tunnel studies on this important aspect are proceeding at Derby.
Great importance obviously attaches to these, and other, aspects of aerodynamics. Both theory and wind-tunnel model testing have their place, but in view of the complicated nature of the airflows involved, a heavy reliance will be placed on full-scale testing with the experimental APT. The role of the wind-tunnel in studies of aerodynamics of trains is at present being studied by Loughborough University of Technology.
It has been indicated that there is considerable technical innovation in the APT. The successful implementation of these ideas obviously requires a balanced research and development program and this is being carried out by the British Railways Board with the active and practical support of the Ministry of Transport. Since this program commenced in 1969, a team has been built up and extensive facilities have been created. A new laboratory has been built at the Railway Technical Centre, Derby, which houses a very large roller rig, a brake test dynamometer and a wide range of vibration equipment for developing suspension systems.
A test track has been built near Melton Mowbray which is 15 miles long. This contains examples of all track features likely to be met in practice. It has many curves, reflecting the importance of the APT's high-speed curving capability. It also includes a section on which the APT can attain its maximum speed of 160 miles/h.
Laboratory rig work is now underway on various subsystems for APT-E, the experimental 4-car train which is under construction. This experimental train will run on the track in 1971, and it will be used for experimental and development work. As this work proceeds, prototype trains will be designed and built which, after a development phase, will enter revenue earning service in 1974. The production of further trains will be determined during the progress of the work on the experimental and prototype trains.
In parallel with the development of the train itself the many operational and commercial problems which arise in a project of this kind are being tackled. The proper ergonomic design of the drivers cab, the effects on and requirements for signalling, the safety of men working on high speed lines are examples of a wide range of problems which must be resolved in advance of commercial operations.
The obvious application of the APT is to the main intercity routes. As already discussed the timings of the APT represent a big improvement in average speed between city-centres and a substantial increase in passenger journeys can be expected as a result. Whilst the APT will clearly have a higher capital cost than existing conventional trains, it will also have higher productivity because of the higher average speed (fewer trains required for a given service) and because it will be designed for high utilisation. Because of the large fixed costs associated with railway operation, the economics of the system is more dependent upon achievement of a high volume than on minimising direct operating costs. It is, however, expected that direct operating costs of the APT will be similar to those of conventional trains at current speeds.
Because of the ability of the APT to run on existing tracks with existing signalling at higher average speeds it is not only the present main inter-city routes that would benefit from its application. Some of the less well developed cross-country routes could be made more useful and competitive by operating this type of train. The crucial point here, of course, is that the capital investment involved here is substantially that of the trains, and this is small compared with possible expenditures on fixed installations in order to achieve the same objective with conventional trains, or indeed any system requiring new or up-graded track.
An important possible application of the APT is to provide airport links between city centres and airports located at some distance. Such sites are obviously more attractive if a relatively inexpensive high speed link can be provided. A further possible advantage is that by using the existing rail network, such airports could effectively serve a larger proportion of the country in a more direct way, rather than being completely reliant on an-interchange in a major city.
A further possible application of the APT is to outer suburban services particularly around a large city. It may be possible to develop a larger market for high speed suburban services covering a rather larger area than is conveniently served at present.
By making full use of the inherent flexibility of APT, an imaginative provision of new services may contribute to a more effective social and economic use of the large investment represented by the existing railway system.
If successful, the application of the APT on existing tracks may demonstrate the value of high speed intercity transport and thus stimulate demand for higher speeds. As mentioned previously, from a purely technical point of view, very much higher speeds than 160 miles/h are feasible using wheel-supported vehicles but these developments would need larger capital investment for the new tracks that would be required.
|
|
|
