Director & Senior Consultant Alstom Transport
High Speed Project Manager Alstom Transport
YEARS OF RESEARCH, studies and simulations culminated last month when the experimental Automotrice à Grande Vitesse took to the rails. Forerunner of Alstom's next generation of high speed trains, the AGV introduces distributed power to the French TGV family.
No less than 30 years have elapsed since the TGV001 prototype left the Alstom factory in Belfort. Since then more than 500 TGV sets have been delivered, most of them to European railways, but some to South Korea, where numerous test runs at 300 km/h or more have been made during the last two years.
It is the experience amassed over 20 years of commercial service that allows us to launch another high speed train concept with confidence. The two pre-production trains are now approaching their 25th birthday, and the rest of the first-generation fleet has been refurbished for service on TGV Méditerranée. Although built originally for a maximum speed of 260 km/h, SNCF has modified them to run between Paris and Marseille at 300 km/h. Major changes have not been necessary, and the original bogies, traction motors, transformers and other electrical equipment remain in service.
The second-generation trains, one of which hit the headlines in 1990 when it set a new world speed record for a passenger-carrying train of 515·3 km/h, have accumulated 12 years of commercial service. One unit established another record on May 26 this year, by running non-stop for nearly 1 100 km from Calais to Marseille in 3 h 30 min. SNCF plans to run these trains at 320 km/h on a section of TGV Méditerranée in preparation for TGV Est, on which major construction work should start next year.
The double-deck TGV Duplex design formed the third generation of TGVs, and its success has led SNCF to confirm follow-on orders that more than double the size of the initial build.
What then, with three generations of TGV in successful service, has led Alstom to develop another high speed train concept?
First, although the AGV is innovative in many ways, the main design principles developed by Alstom and SNCF remain intact. Articulation has advantages in terms of aerodynamics and comfort, and it has also proved to have major safety advantages under extreme conditions. The TGV bogie has proved to be exceptionally stable, thanks to its specific design: a 3 m wheelbase, light weight (8 tonnes), carefully chosen longitudinal and lateral stiffness of the primary suspension, and the location of the traction motors. Similarly, modular components and duplication of equipment ensure economic operation and good availability.
The AGV concept
AGV's principal innovation is to move away from power cars to distributed power, with traction motors and other electrical equipment spread along the train. We should stress that this change has not been made to improve adhesion, as there have been no problems with existing TGV trainsets developing 1 100 kW per motored axle nor with power cars which were producing 1 500 kW per axle when they set new speed records during trials in 1989-90, albeit on track of exceptional quality.
Neither has this change been forced by the need to develop a TGV with traction equipment for 15 kV 162/3 Hz that respects the 17 tonne maximum axleload specified by European high speed interoperability requirements. The technical feasibility of a fourvoltage TGV power car capable of operating at up to 300 km/h under 25 kV 50 Hz and 15 kV 162/3 Hz catenary has already been demonstrated by the Thalys PBKA trainsets.
Rather, in developing the AGV, we have used advances in electronics and power conversion, such as IGBTbased inverters and asynchronous traction motors, to create a new train design offering significant advantages. By eliminating power cars and increasing the number of passengers that a single-deck train can carry, the AGV should reduce investment, power and maintenance costs per seat, also helping to distribute passengers more evenly along station platforms.
The modular design should enable trainset lengths to be closely matched to the operator's requirements, with distributed traction offering significant advantages over power cars on less busy routes. After developing in the TGV Duplex the rail equivalent of the high-capacity wide-body jet with over 500 seats, the AGV should enable us to develop high speed trains for intermediate or regional applications. AGVs will be made up of modules of three or four cars with two motor bogies, giving trainsets ranging from 140 m to 200 m in length with between 280 and 430 seats, depending on the chosen interior configuration. And although it is not an immediate concern for the principal European railways, the AGV offers the prospect of raising maximum speeds to 350 km/h without major changes to the train design.
Dramatically reducing the number of train subassemblies, the AGV is formed of only two types of car, an end vehicle with a cab and an intermediate car. Similarly, there are only two types of bogie, powered and trailer. This simple format has enabled us to test the AGV concept with one prototype of each car and bogie, marshalled for dynamic testing into a representative seven-car formation with four TGV Réseau trailers and a TGV Réseau power car. The AGV cars share the AGV motor bogie, while the intermediate AGV car shares a TGV Sud-Est trailer bogie with its Réseau neighbour.
With the same total number of cars as planned for the shortest production AGV trainset, this formation is capable of running at speeds up to 200 km/h on conventional routes and up to 320 km/h on high speed infrastructure. As presently configured with six out of 18 axles motored, the AGV test train does not have enough installed power to reach 350 km/h and will therefore be coupled to a modified TGV Réseau trainset for trials at this speed.
Trials will focus on those issues raised by the choice of an all-articulated format for the AGV, in contrast to previous builds of TGV where motor bogies have been mounted conventionally under the car ends. The dynamic behaviour of the AGV with a leading trailer bogie will be tested at up to 350 km/h in accordance with the standards laid down in UIC leaflets 518 and 513.
Both AGV motor and trailer bogies have been fitted with special wheelsets to record track forces across a range of speeds, rates of cant deficiency and train weights, and operation under reduced power will also be simulated. Rotational torque and bogie behaviour on uneven track will also be investigated, under the direction of Eurailtest which is conducting the trials.
This full programme of dynamic testing may seem superfluous, given that the AGV motor and trailer bogies have been derived from TGV designs whose behaviour in terms of ride comfort and track forces has become well known after several years in service. Nevertheless, our testing programme aims to explore the dynamic behaviour of the AGV as fully as possible with future operation at up to 350 km/h in mind. We should also mention that measures to improve lateral ride quality at speeds above 320 km/h have been investigated using a TGV Réseau set fitted with active lateral suspension on some bogies (RG 7.99 p426). In terms of improved passenger comfort, the results were promising.
Interior noise levels will be measured at precise locations in order to test the effectiveness of sound insulation fitted to the AGV driving car and installed around the vestibule of the intermediate car. A number of measures have been taken to reduce noise and vibration, given that passengers will now ride in the leading car and that the motor bogie is closer to the passenger saloons, even if the articulated TGV format means that there are no seats directly above the bogies. A TGV Réseau car with a standard interior will serve as the benchmark during trials.
Train noise will also be measured, in order to assess the potential of mitigation measures should these be required. Such measures could include shrouding the leading bogie and optimising the location of secondary suspension components. Compared with other high speed train designs, the articulated TGV format offers significant advantages in terms of train noise, given the smooth and uniform transition between cars and that the overall number of bogies per trainset is reduced.
A major strand of the testing programme will comprise trials of the AGV's single 'Europantograph' under high speed catenary at 25 kV 50 Hz and 15 kV 162/3 Hz, as well as 1·5 kV and 3 kV DC on conventional routes. The effect on interior noise levels of the pantograph located at the end of the car above the vestibule area will also be measured. Having only two pantographs on an AGV, irrespective of the total number of cars, should prove to be a significant competitive advantage in the future development of multi-voltage trainsets.
Given the rapid development of power components, we have not chosen to test a complete drive on the prototype AGV. Instead, the more long-lived components and subsystems are being tested in order to reduce the technical risks faced by future builds. Dynamic trials will focus on the traction motor (already subjected to static testing), braking rheostat and traction equipment ventilation. The traction motor is a proven self-ventilating asynchronous design, and the AGV trials will focus on the levels of air pressure in the motor bogie relative to speed and direction of travel.
Having retained the TGV's final drive, the AGV trials will need to confirm that mounting the traction motor on the vehicle body will not have an adverse impact on noise levels within the passenger compartment. The traction motors will be supplied with current having a wave form similar to that produced by future inverters, and motor flux levels will be varied across the full range of speed and torque values to determine how these influence the level of vibration transmitted to the vehicle body.
The electrical braking system follows recent TGV practice, and the AGV is intended to have a powerful rheostatic brake developing 1·2 MW per motor bogie. An initial series of static trials has proved the feasibility of a smaller, roof-mounted braking rheostat, indicating that sufficient volumes of air would pass over it to ensure the correct temperature levels for its active parts as well as exhaust air. In addition, these tests raised the possibility of varying the flow of cooling air in proportion to the amount of power being absorbed by the rheostat, which would enable noise levels to be reduced significantly as the train slows down.
The second phase of testing to be carried out using the test train will seek to verify that changes to air flows due to the speed of the train and the direction of travel will not have a significant impact on the amount of air passing over the braking rheostat and the intakes for air-conditioning equipment.
Traction equipment has been installed beneath the floors of both AGV vehicles. Even if the amount of cooling air required can vary between different types of equipment and at different train speeds, the underlying technical principles remain the same for transformers, traction motor inverters and those that supply ontrain auxiliaries. Transformer cooler groups have been installed under the AGV driving car. Trials will seek to determine that they do not increase interior noise levels as they take more air in, and that these flows are not adversely affected by train speed or direction of travel.
Following the installation of measuring equipment, the test train was due to begin trials on conventional infrastructure during the second half of October. The first runs at 320 km/h between Lille and Calais are scheduled for the start of November, with testing at 350 km/h to follow in January next year. The current AGV testing programme is due to finish in March 2002.
At the same time as the AGV trials, studies are continuing into a number of emerging issues that will shape the development of high speed trains in Europe. Until comparatively recently, the behaviour of high speed trains in cross-winds was not a major concern, although it had been considered by those who worked on the first TGV designs. It was not felt to pose a major risk due to climatic conditions on the first TGV lines and the adoption of power cars (for both TGV and the first ICE builds). Thus the leading vehicle most exposed to cross-winds was relatively heavy. As it carries no passengers, the power car has a constant mass and therefore enjoys excellent stability in cross-winds.
In contrast, the effects of crosswinds have been studied more extensively in Japan due to climatic conditions, accidents involving multipleunits on the narrow gauge network where stability due to gravity is less than on standard gauge, and the choice of distributed traction for high speed trains. The topic has become something of an issue in Europe as high speed development in Germany has produced trains with distributed traction and leading cars significantly lighter (especially when empty or lightly loaded) than the ICE1 power cars. Studies were undertaken by SNCF prior to the opening of TGV Méditerranée, as this serves a region with stronger winds than are found elsewhere in France and has many exposed, tall structures such as viaducts and high embankments.
In the absence of dedicated power cars, stability in cross-winds is more of an issue for trains with distributed traction, and it is our intention that the AGV should enjoy the same stability in cross-winds as the TGVs that formed the basis of the TGV Méditerranée studies. Articulated trainsets enjoy an advantage over conventional rolling stock as most bogies are loaded by two separate cars, which do not react as a single body when they encounter a gust of wind at high speed. In addition to work being carried out as part of the test programme, the AGV design is being refined to take advantage of this phenomenon, and of the fact that the aerodynamic force of each car decreases between the lead vehicle and the middle of the train.
Trials with TGV trainsets ranging from the Sud-Est type with eight trailer cars to the Eurostar with 18 and the double-deck TGV Duplex have confirmed excellent ride quality at up to 300 km/h or 320 km/h on high speed lines currently in service. Test runs undertaken with set No 325 prior to its successful attempt on the world speed record, and more recently with TGV Réseau sets at up to 360 km/h on TGV Méditerranée, have demonstrated that lateral ride quality remains excellent at 350 km/h and above on track maintained to the highest standards.
In order to ensure that this ride quality can be maintained at 350 km/h and above where track geometry tolerances are less exacting to allow cost-effective maintenance under heavy traffic, we expect further studies and trials of the active lateral suspension that showed promise when tested under a TGV Réseau trainset. We intend to draw on the lessons learnt during trials with the TGV trainset modified for tilt, which produced an interesting development of control software to incorporate detailed knowledge of track geometry, even including long-wave defects. Combining this control software with active lateral suspension should ensure high ride quality at 350 km/h under all circumstances, eliminating the effects of bogie hunting and track defects, and ensuring that the same excellent levels of lateral comfort are maintained right from the front of the train to the rear car.
One of the advantages of the articulated format, particularly evident in the TGV Duplex, is that with the passenger saloons positioned between the bogies, a relatively low floor height can be achieved, improving accessibility.
Although detailed designs for the various types of AGV have yet to reach the definitive stage, it seems possible that this advantage can be maintained despite the added difficulty of accommodating more underfloor equipment.
Although it is too early yet to go into detail, the constant evolution of on-train technology means that the future development of the AGV has already begun to be mapped out, so that the product will adapt itself to change in the manner of the TGV. We should not therefore rule out the possibility that, in due course, the AGV family may include an updated version of the high-capacity TGV Duplex.
- CAPTION: Dynamic testing with the two experimental AGV cars began last month
- CAPTION: A traction motor assembly is slung under the body of the driving car, driving the articulation bogie
- CAPTION: The power bogie awaits mounting under the two AGV cars in Belfort
- CAPTION: The AGV cars are married with four trailers and a power car from a TGV Réseau set CAPTION Instrumentation work under way on the 'Europantograph'
L'AGV fait entrer la motorisation répartie dans la famille des TGV
Des essais menés avec deux véhicules expérimentaux à grande vitesse jettent les bases d'une autre étape dans le développement du concept français de TGV. En tant que concepteur de l'Automotrice à Grande Vitesse, Alstom saisit l'occasion d'explorer une série d'innovations en vue de réaliser une version commerciale capable de circuler à 350 km/h. Grâce à l'utilisation d'un équipement de commande léger basé sur l'IGBT, l'AGV comportera une motorisation répartie plutôt que des motrices indépendantes, comme pour les TGV construits précédemment
AGV bringt verteilten Antrieb in die TGVFamilie Versuche mit zwei experimentellen
Hochgeschwindigkeitsfahrzeugen legen die Basis für einen weiteren Entwicklungsschritt bei Frankreichs TGV-Konzept. Als Entwickler der Automotrice à Grande Vitesse nutzt Alstom die Chance zur Erforschung einer Reihe von Innovationen mit dem Ziel, eine kommerzielle Ausführung mit Spitzengeschwindigkeiten von 350 km/h zu bauen. Dank dem Einsatz leichter auf IGBT basierenden Antriebssysteme wird die AGV einen verteilten Antrieb aufweisen, im Gegensatz zu den Triebköpfen der heutigen TGV-Konstruktionen
El AGV hace entrar la tracción distribuida a la familia TGV
Las pruebas con dos vehículos de alta velocidad experimentales están echando los cimientos para otro paso en el desarrollo del concepto TGV en Francia. Como desarrollador del Automotrice à Grande Vitesse Alstom está aprovechando la oportunidad de explorar una serie de innovaciones con vistas a construir una versión comercial capaz de alcanzar los 350 km/h. Gracias al uso de un sistema con IGBTs, el AGV contará con tracción distribuida en lugar de las cabezas tractores de los trenes TGV ya en servicio