An investigation on the effect of contact wire irregularity on current collection in high speed lines M. Papi (*), A.G. Violi (*), G. Diana (**), A. Collina (**) (*) Trenitalia S.p.A. - UTMR Sperimentazione.- Prove Elettriche V.le Spartaco Lavagnini, 58-50129 Firenze, Italy, papi@asamrt.interbusiness.it (**) Politecnico di Milano, Department of Mechanics, Campus Bovisa, Via la Masa 34, 20158, Milano, Italy, andrea.collina@polimi.it Introduction The operation of a high speed line requires a high level of performance of both overhead line (OHL) and pantograph, because of their mutual dynamic interaction. The disturbances to the current collection mainly come from the dynamic response of the coupled system (pantograph+catenary) related to the catenary structure (such as periodicity of the span, droppers, transition sections), aerodynamics forces acting on the pantograph (due to mean and turbulent flow), and contact wire irregularities (such as height errors and localised wear). Aim of the paper is to investigate, both experimentally and by means of simulation, how contact wire irregularity influences current collection quality. There are several motivation for the irregularity of contact wire, such as: - errors in dropper length - non correct tension of the wires, due to malfunctions of the tensioning devices - differential creep between contact wire and messenger wire - non constant height of the contact wire of the line above the track plan due to errors of mast positioning. All the above mentioned defects cause a variation of the trajectory of the collectors of the pantograph, inducing a supplementary acceleration and related inertia forces, which can lead, in some cases, to contact loss, and, in any case, causes in increase in the dynamic variation of the contact forces, which, in turn, excite the catenary to a higher dynamic level. Since the amount of the defects that can be found on a Overhead Line (OHL) depends on the level of maintenance, it can be said that the actual maximum speed that can be achieved on a OHL is also related to the level of maintenance. In fact, the influence of contact wire irregularity is more pronounced at higher speeds, depending on the typology of the catenary, the kind of pantograph and the mean contact forces. This topic requires new consideration in view of interoperability of European network. The balance between the costs required by a high level of maintenance, and the convenience in obtaining the full maximum speed from the OHL, are then to be compared. In this paper, in particular, the attention is focused on the effect of positioning error across the span: - investigating the experimental evidences, obtained from on-board measurements of pantograph dynamics; - interpreting the results of the simulation of pantograph-catenary interaction; - using the simulation to forecast the effect of different levels of the same defect. Aim of the investigation is also to assess up to which level the considered defect can be tolerated Experimental results The experimental results have been obtained from a series of tests carried out by UTMR (Traction and Rolling Stock Unit of FS Trenitalia, based in Florence), on the 3kV line Roma Settebagni-Orte, equipped with the new 540C catenary, developed in previous tests [Diana 1999]. The test train was an ETR500 train-set, fitted with two ATR90 pantographs (with double collector), equipped with a sophisticated monitoring and acquisition data system, able to measure all the quantities relevant to the current collection:
To perform an extensive analysis, the following quantities are measured, for front and rear pantograph [Papi 1997, 1998]: - Mechanical: dynamic vertical component of contact force (global pantograph force and individual collector head force, and vertical motion of pantograph members (collector heads, frame); the presented results refers to a set-up where the contact forces where not measured, while the contact losses where detected by means of an innovative device [Papi 1999]. - Electrical: voltage and current supplied by overhead line; loss of contact of the pantograph with the contact wire (number and duration); - Thermodynamic: strips temperature and air pressures on pneumatic circuits. Auxiliary quantities are taken: TV camera image of the pantograph aimed to control its vertical movements. All signals are digitized and managed by a single dedicated software application, distributed on a local area network of personal computers, with a stratified architecture consisting of a first layer (Physical Channel Abstraction Layer: PCAL) that provides the software interface toward the individual instruments and converts the electric signals, acquired from the various transducers and instruments, into virtual channels. The second layer (Virtual Channel Layer: VCL) consists of a direct representation of the acquired quantities in the corresponding physical units, regardless of the physical chain of acquisition involved. Management of virtual channels is one layer of the application (VCL) in which the measurements, handled in real time, are: elaborated: giving rise to new virtual channels (quantities) stored: digitally as files in the mass memories of the network displayed: numerically and graphically on monitors or printers transmitted: to other computers through the local network. The most interesting quantities available for the presented investigation are the pantograph movements and the contact losses detection. The contact losses has been detected by means of the current in the equipotential cable that connects the front and the rear pantograph [Papi 2001]. Figure 1. The ETR500 train set: the equipotential cable connecting the front and the rear pantograph, also used to detect contact losses.
Figure 2. Action of the EW in the case of a break arc Figure 3. Experimental results at 200km/h.
Figure 4. Experimental results at 252km/h. Figure 5. Pantograph motion at different train speed.
The tests were carried out at different speeds, from 150km/h to 252km/h, with two pantographs. The attention focuses on the effect, on the quality of current collection, of a variation of contact wire height (i.e. its vertical level) across two consecutive spans. Figures 3 and 4 show, from the top, as a function of longitudinal position along the line: - Absolute height of the contact line, which represents the absolute motion of the pantograph, obtained under quasi static condition (static: low train speed, and a minimum values of pantograph pre-load), and dynamic condition (dynamic: normal operating condition of speed and pantograph pre-load). - Dynamic pantograph motion, obtained as the differenc between dynamic and static line heigth, representing the dynamic motion of the pantograph depurated from the contact wire vertical position, but including the dynamic effect of contac wire irregularity; - Front and rear collector-articulated frame relative motion; - Current in the equipotential cable. The level of change of contact line height is visible in the 32.6/-32.4 longitudinal position in the static/dynamics line height (figure 3). Examining the test results (figure 3, at 200 km/h and figure 4, at 250 km/h), it can be observed the great difference in the dynamic pantograph motion between 200 km/h and 250 km/h, where contact losses clearly appears detected by the current in the equipotential cable. Therefore the current collection was satisfactory up to a limiting speed, beyond which a systematic contact loss pattern appears, in correspondence of the span where the local defect is present, so establishing a threshold for the maximum allowable speed of the train on this section. Considering also the other train speeds (figure 5), the dynamic motion of the pantograph that occurs where the irregularity of the contact wire is located, changes once a limiting speed for the considered change of line height (around 220km/h), is exceeded. Simulation and comparison with measurements. The same situation has been examined through simulation, by means of a computer code for the dynamic interaction of pantograph and catenary [Diana 1998]. The OHL is represented by means of finite elements (tensioned beam elements), considering also the non linearity related to the slackening of the droppers. This last point is essential for a realistic simulation of contact forces. The pantograph is reproduced by means of a lumped parameters model, that takes into account also the kinematics of the suspension of the collector, enabling to simulate also the effect of contact force unbalance between front and rear collector (see figure 6). The simulation of pantographcatenary dynamic interaction is performed in the time domain, with the possibility to include the contact wire irregularity.
Figure 6. Lumped parameters model, including the kinematics of the collectors suspension Figure 7. Simulated results at 200km/h.
Figure 8. Simulated results at 250km/h.
Figure 9. Simulated contact force at 200km/h and 250km/h. The effect of the same kind of contact wire irregularity found experimentally has been simulated with the mathematical model, in order to reproduce the measurements, and subsequently to evaluate the effect of different amount of contact wire height change. Figures 7 and figure 8 reports the results of the simulations pertinent to the experimental ones shown in figures 3 and 4. Similar features can be found: relative motion of the front collector head show, in both cases (measurement and simulation), the tendency to approach the upper end of suspension excursion. Moreover, the total contact force reaches a low value in correspondence of the absolute maximum height of the line, approaching the contact loss condition, similarly to what happens on the experimental case (Figure 9). A full contact loss is not obtained from the simulation, even if the contact force shows the tendency to it. It should be noticed that other disturbances are present in the real operating condition (such as shorter wave irregularity of contact wire, aerodynamic disturbance due to turbulence of the flow incident on the pantograph, etc.) which are not included in the simulation. These factors increases the variation of contact forces, and, superposed to the main variation due to the change of contact line height, would lead to a full contact loss. Evaluation of effect of irregularity by means of simulation Since one of the aim of the numerical analysis is also to correlate the amount of the defect to the maximum speed compatible with current collection, the mathematical model has been then used to evaluate the effect of the same kind of contact wire irregularity, considering different amount of vertical line height change across the two considered spans (Figure 10). Table I reports besides the results with the condition corresponding the real one, others with different values of Dz1 and Dz2 equal to zero
It can be noticed that: a) the main effect on current collection is due to the first line height change (Dz1); b) Considering the speed of 250km/h, there s a significant change in the minimum value of the contact force, from10cm of defect to 15cm of defect, where an evident contact loss appear, with a duration of about 12ms. Figure 10. Simulated contact force at 250km/h with different amount of Dz1 defect Table I - Summary of the simulated contact force results. Condition speed [km/h] Dz1 Dz2 Fcmin [N] 1 200 +12cm -3cm 35 2 250 +12cm -3cm 17 3 250 +5cm 0cm 34 4 250 +10cm 0cm 19 5 250 +15cm 0cm 0 Effects on maintenance costs As an application of the previously reported analysis, the consequences on the maintenance strategy of the OHL are here briefly outlined. As known, the presence of systematic contact losses in a section leads to an accelerated wear rate of the contact wire, localised in the same section [Borgwardt 1988], requiring an anticipated substitution of contact wire, and imposing a limitation on the maximum speed of the train. Without any intention of completeness, the following scenarios are examined (see Table II). Three possible maintenance options are compared: a) immediate correction of line height, out of the normal maintenance program. The costs are the
cost of the corrective operation, and the cost of the extra out of service of the line (even if it may be in the night hours). No additional wear of the contact wire is caused in this case. b) Correction of line height in the first available period of programmed maintenance. In this case the cost is only related to the corrective operation, since no extra out of service is caused. The wear rate of the contact is increased until the corrective action is taken: it is assumed this can be undertaken within one year. This cause a certain decrease of the life of the contact wire. c) No correction at all is taken. Therefore, an accelerated wear rate is caused on the contact wire, requiring an anticipated substitution, which is the cause of the extra cost. Table II reports an hypothetical calculation, based on the time-wear matrix reported in [Borgwardt 1988]. A rate of wear of 2% of contact wire section is assumed in the first year of service, this corresponds to a total life of 16years and 3months, as a reference condition. This would be the case if an immediate countermeasure would be taken on the line (option a). If an accelerated wear rate is assumed in the first year (f.i. 4%), because of the increased consumption related to the systematic contact loss without taking any countermeasures, (option c), then a life of four years and six months would be forecast. If, (option b) after one year a corrective action is taken, and assuming that the reference wear rate holds after the maintenance operation, then the remaining 16% of the wire section is worn out after about 12 years, for a total life of 13 years (3 year less than with the normal wear rate) could be supposed. The obtained results are of course only indicative, since the operating conditions of the 3kV dc line of Fs is different from the DB line the wear matrix is based on, nevertheless it is pointed out the interaction among pantograph-catenary dynamics, reponses to wire irregularity, wear of the contact wire and maintenance strategy is very close and should be always considered. Table II comparison among different maintenance actions Option Cost of maintenance operation Life duration of contact wire A Corrective action+out of service 16 years B Corrective action 13 years C - 4 years and 6 months Conclusions This study is a first step in order to assess which is the level of accuracy required in the positioning of the contact wire on a high speed system. This knowledge can be of importance, since the accuracy of the contact wire height requires higher maintenance cost, to be included as a part of the cost-benefit balance, and is related to the maximum operating speed of the line [Resch 1998, 1999]. In a more general viewpoint, it could be affirmed that the balance between higher maintenance cost on one side, and the higher revenue obtained from a line with a higher train speed, which attracts more people, needs to be considered. An effective maintenance of a OHL allows in fact to obtain the maximum benefit from it, especially in the case of a high speed line, which is an alternative to the air traffic. References [1]. G. Diana, M. Bocciolone, A. Collina, M. Papi, A. G. Violi, Optimization of new d.c. catenary by means of measurements and simulation of pantograph-catenary interaction, WCRR 99, Tokyo, Japan, 19-23 October [2]. M. Papi, A. G. Violi, A. Balestrino, O. Bruno, A. Landi, L. Sani, Innovative methods for evaluating the quality of the current collection, International Conference Railway Traction System,14-16 May 2001, Capri, Italy [3]. M. Papi, E. Mingozzi, A. G. Violi, O. Bruno, A. Landi, L. Sani, La captazione di corrente e
l interazione pantografo catenaria - I parte: Metodologie delle misure per il controllo della qualità Ingegneria Ferroviaria - Avril 2001, Italy [4]. M. Papi, E. Mingozzi, A. G. Violi, O. Bruno, A. Landi, L. Sani, La captazione di corrente e l interazione pantografo catenaria - II parte: Innovazioni per il rilevamento della qualità della captazione, nelle catenarie a 3 kv cc. Ingegneria Ferroviaria - May 2001, Italy [5]. M. Papi, P. Masini & G. Puliatti, Virtual acquisition system for experimentation in Pantograph Catenary interaction, Proceedings of Comprail 98, Lisboa, 2-4 September 1998, in Computer in Railways VI, Wit Press, pp. 847-856 [6]. M. Papi, A. G. Violi, Pantograph-overhead line interaction: methods of measurement in experimentation on the line, WCRR 97, Firenze, Italy, 16-19 Nov. 97 [7]. G. Diana, S. Bruni, A. Collina, F. Fossati, & F. Resta: High speed railways: pantograph and overhead line modelling and simulation, Proceedings of Comprail 98, Lisboa, 2-4 September 1998, in Computer in Railways VI, Wit Press, pp. 847-856. [8]. K. Becker, A. Konig, U. Resch, B-W Zweig, Systematic development of a high speed Overhead Contact Line, RTR, No3-4/1995 [9]. K. Becker, U. Resch, B-W Zweig, Optimizing High-Speed Overhead Contact Lines, AEG report, ABF 14.7/0994-944 [10]. Borgwardt, Verschlei verhalten des Fardrathes der Regeloberleitungen der Deustche Bundesbahn, Elektrische Bahnen 94(1996), eb 11/96