Bogie performance through curves can be improved by careful control of the friction between wheel and rail. A group of Japanese engineers has developed a friction control method for the Tokyo Metro using an onboard rail spray to control the friction coefficient and the friction characteristics of the wheel-rail contact area

Kosuke Matusmoto and Masao Tomeoka are respectively Engineer and Manager in the Rolling Stock Design Section at Tokyo Metro Co Ltd. Dr Yoshihiro Suda is Professor at the Centre for Collaborative Research and the Institute of Industrial Science at the University of Tokyo, Takuji Nakai and Masuhisa Tanimoto are Engineer and Manager in the Bogie Design Section at Sumitomo Metal Industries Ltd in Osaka.

JAPAN'S highly-developed urban rail networks are constantly striving to improve performance, and one target is to shorten journey times.

Because many lines feature large numbers of sharp curves with trains operating at relatively high speeds, an improvement in curving performance is a worthwhile achievement to which many operators aspire. Tokyo Metro, which is Japan's largest underground railway operator, has developed an onboard friction control system to improve bogie performance through curves.

When a vehicle runs through a sharp curve, several undesirable phenomena related to the wheel-rail contact can be observed. These include high-frequency squeal at the wheel-rail contact point on the low rail, loud noise from the wheel flange-rail contact on the high rail, excessive rail and wheel wear, together with noise and vibration generated by rail corrugations. The principal cause of these phenomena is the large lateral force which is related to the friction between wheel and rail, and a common method of mitigating the problem is to apply grease to the rail immediately ahead of a sharp curve using an onboard or wayside device. However, the grease sometimes causes wheelspin or wheelslide as it reduces the coefficient of friction between wheel and rail.

The Tokyo Metro network has some routes where there are numerous tight curves, and a joint development group was set up to tackle the problem. The group investigated the use of different substances to control friction levels in curves, and among the most effective was a water-based friction modifier developed by a Canadian manufacturer. Kelsan Technology's Keltrack?? HPF has the advantage that it can be easily supplied as an atomised spray to the top of the rail.

The friction control method

Controlling the friction between wheel and rail is the most effective way to improve the wheel-rail contact condition, and by limiting application of the friction modifier to the curved section there is no effect on stability when the train is running at high speed on tangent track between curves. In terms of installation, use of a friction modifier is advantageous compared with other methods as an application device can be easily fitted to rolling stock.

The developers chose to test an onboard applicator (Fig 1), from which liquid containing the friction modifier is sprayed as a mist on to the top of the low rail from a nozzle located behind the last axle of the train. The following train then receives the benefit. The effectiveness of the friction modifier applied in this way means that one application for every three to five trains passing through the curve suffices to achieve the desired effect.

The system has several advantages. The friction modifier can be applied to the rail uniformly through the curve by controlling the spray volume in proportion to the speed of the vehicle. Not only that, but any train fitted with a friction control device can immediately supply the friction modifier to any curve on the line where wheel-rail contact condition needs to be improved unexpectedly. Thirdly, it is a simple matter to top up the friction modifier liquid and to service the applicator when the train visits its normal maintenance depot.

Initial tests

In the first phase of the development, experiments were carried out with a twin-roller rig. These proved that the friction modifier had the correct friction coefficient for lubrication between wheel and rail for a bogie negotiating a curve. It demonstrated a positive friction characteristic in accordance with the increase in slip rate between two rollers that prevents a wheel spinning or skidding when the vehicle accelerates or stops (Fig 2).

The tests also revealed that the balance between supply and consumption of the friction modifier was very important. This was ascertained during tests in which the state of the wheel-rail contact condition on a line in commercial service was simulated with repeated application of the friction modifier to the same section using two trains. Even though the characteristics of the friction modifier proved better than expected, the tests showed that the volume of modifier applied to the rail had to be appropriate for the number of trains operating. In other words, too much spray or low consumption of the modifier mean that the coefficient of friction is too low and the anticipated positive characteristics do not materialise.

The next step was to carry out a series of tests using scale models. By applying the friction modifier to the top of the low rail and comparing the results with a rail in dry condition with no lubricant or friction modifier, good results were obtained.

Field trials and evaluation

The developers then decided that enough progress had been made to proceed to field trials, and the necessary equipment was mounted on vehicles used on the Chiyoda Line and the Marunouchi Line of the Tokyo Metro.

Special wheelsets able to measure the forces acting on the wheel from the rail were then fitted to the trains, and these indicated that the friction modifier was having a significant impact in reducing the lateral forces exerted on both high and low rails (Fig 3). Specifically, the value obtained from the lateral force divided by the vertical force acting on the wheel was reduced to almost half in comparison with rail in conventional dry condition.

The final phase of development saw friction modifier sprays fitted to eight trains running in commercial service on the Marunouchi and Chiyoda lines for a period of five years, enabling the effects to be fully evaluated. On the Chiyoda Line the trials were carried out on a 2·1 km section with four curves whose radius ranged from 147m to 351m. A 3·2 km section of the Marunouchi Line was chosen, with the sharpest curve having a radius of 127m; other curves had a radius of up to 301m.

Long-term observation of the 147m curve on the Chiyoda Line was carried out with several measurements taken. These included the forces exerted on the rail by the wheels, lateral acceleration of the rail and the level of noise on the test curve. The results were convincing, with a dramatic improvement in curving performance (Table I). Both noise and vibration of the rail fell significantly using friction control.

There was another important benefit. Observation of the low rail in the curve 45 days after the friction modifier was first applied showed that the corrugations present previously had almost disappeared. The development team has confirmed that no new corrugation appeared on this curve for more than four years after the field trials began (Fig 4).

The long-term trial proved beyond doubt that the onboard friction control system was very effective in improving bogie performance in curves. Further research and development is now in hand to develop a more economical system with better performance that can be made available commercially.

  • CAPTION: Fig 1. Trains are fitted with a spraying nozzle (below left), control unit and a reservoir for the friction modifier (below right); an inverter powers a pump for the spray
  • CAPTION: Fig 2. Wheel-rail creep characteristics with the coefficient of traction plotted against the slip rate. The friction modifier eliminates slipping problems which occur with grease
  • CAPTION: Fig 3. During curving tests, lateral forces exerted on the high rail were halved, with an even better result for the low rail
  • CAPTION: Trains on the Chiyoda Line (p605) and the Marunouchi Line (left) were fitted with friction control applicators

Table I. Comparison of lateral rail forces on the high and low rails of a 147m radius curve on Tokyo Metro's Chiyoda Line

Dry Friction control Reduction %

High rail 24·0 kN 12·3 kN 49

Low rail 21·7 kN 8·8 kN 59