Railway Technical Research in Asia

Feature

Japan Railway & Transport Review 36 • September 200348 Copyright © 2003 EJRCF. All rights reserved.

Aiming at World-leadingMaintenance Technology

Hiroaki Inazu

Introduction

This article explains some of themaintenance-related research conductedat the Technical Center in the Researchand Development Center of JR East Group.In particular, I would like to explain somethemes related to ‘interface technologies’into which we have been putting a greatdeal of effort recently. The main interfacetechnologies are found at wheel/rail,pantograph/catenary wire, and switches/crossover.

R&D Aiming at Innovation inMaintenance

JR East operates more than 12,000 trainservices each day on both shinkansenlines and conventional lines. To keepthese trains running, there is a hugeamount of infrastructure that must bemaintained in addition to rolling stock.Figure 1 shows the breakdown of currentrolling stock and infrastructure. There aremore than 13,000 pieces of rolling stock,5,500 km of electrified sections andalmost 12,000 km of track. Some of theinfrastructure such as tunnels and bridges,has been in use for just 30 or 40 years butsome parts are more than 100 years old.JR East is required to perform maintenance

in a diverse number of fields, includingmechanical, electrical, civil engineering,etc., on a variety of items built in variousperiods since the first railways in Japan.It is the responsibility of my TechnicalCenter to conduct research that supportsJR East’s maintenance staff in keeping thishuge railway business operating smoothly.The proportion of maintenance costs inJR East’s business expenses as well as theproportion of maintenance engineers inJR East’s total labour force are both nearly30%, clearly a relatively large proportion.Obviously maintenance costs are a veryheavy burden for railway operators andany technical advances in railwaymaintenance would play a major role incost reduction. Figure 2 shows the mainthemes in maintenance today. Reductionof maintenance costs is a major item inleading to better competitiveness ofrailways, but creating a better maintenancesystem structure that would lead to bettersafety and business stability is alsoimportant. In addition, another majortheme is the number of people performingmaintenance and the declining workforcein Japan. These themes are important inmaintaining a competitive business andcreating a railway that is kind to theenvironment and people.Our Technical Center undertakes R&Dbased on these important considerations.

First, our research aims to createinnovation in maintenance that must cutcosts, improve reliability, and support on-site maintenance (Fig. 3). Second, ourresearch a ims to s t rengthen thefundamental concepts of our railwaymaintenance and establishing world-leading evaluation, maintenance andmanagement techniques. Both these aimscome within the premise of the Frontier21 plan forming the medium-termbusiness concept for the JR East group.Our research approach to cuttingmaintenance costs, improving reliability,and supporting on-site maintenance is todevelop systems for maintaining rollingstock and infrastructure that minimize theamount o f manua l work . As aconsequence, a major thrust of our R&Dis developing new intelligent technologiesfor application to inspection and workprocedures. In addition to analyzingvarious older themes in a variety oftechnical fields, such as rolling stock,infrastructure, power supply, etc., we arealso putting a great deal of emphasis onmaximizing the system as a whole basedon the concept of interfaces betweendifferent technologies.Our Technical Center is staffed byspecialist engineers with experience in avariety of technical fields who worktogether on research themes related to

Figure 1 Current Rolling Stock and Infrastructure

Stations, railway system and civil engineering

ing StockRolling StockElectricity

June 2003

Total 13,327 • Shinkansen 1,117 • Electric cars on conventional

lines 10,687 • Diesel cars, etc. 1,523

• Electrified sections 5,504 km• Switches About 11,400• Signals About 13,500

• Stations 1,695• Track length 11,588 km• Tunnels 1,263• Bridges 14,853

Figure 2 Maintenance Themes

Competitivebusiness

People friendlyand environment

friendly

Reduce maintenance costs

Improve safety and reliability

Respond to smaller availableworkforce in future

Japan Railway & Transport Review 36 • September 2003 49Copyright © 2003 EJRCF. All rights reserved.

interface technologies while realizing thegreat importance of producing results thatare applicable to actual maintenance.

Wheel and Rail

The point of contact between the wheeland rail has been a fundamental issue inall previous railway work but there arestill a number of major issues that remainto be tackled. A good example is wheelsqueal that occurs on sharply curvedsections of track due to the contactsurfaces between the wheels and rails.Moreover, corrugation along the rollingsurface of the rail head (Photo 1) have amajor negative impact on passenger ridecomfort as well as on rail and carriagemaintenance. Other examples are wearoccurring on the inside edge of theoutside rail on a curve (Photo 2), andheavy wear causing a change in theprofile of the inside arc of wheel flange(Photo 3). Another example is fatigue-

induced cracking at the rail head calledshelling (Fig. 4), which is one of the majorreasons for replacing rails. If we couldunderstand and solve these problemsoccurring in wheels and rails, the amountof maintenance required would begreatly reduced, having a major impacton cost reduction.Among the annual maintenance costs ofreplacing carriage parts, the expense ofrepairing wheels is quite high but the costof replacing rails is close to four times thatof wheel repairs. Some of our researchaims to optimize the entire system bycutting rail replacement costs.

Optimizing contact shapesThe tread of a wheel on conventionalcarriages is in the form of a series of arcsthat are designed to prevent derailing. Inour research, we used wear tests toexamine the variation in wheel treadwear. In addition to developing a newwheel tread profile, we discovered a wearprofile that provides better stability in

actual use. On the other hand, for therail cross-section shape we tested tworails with changes to the shape of the railhead. Running wear tests using differentcombinations of these wheels and railsproduced an optimum rail designsuffering much less wear. At the moment,we are still performing quantitativemeasurements of rail wear under varioussevere test conditions.

LubricationFigure 5 shows some of the problems thatoccur on rails from curves over time.Since the wheel tread contacts thesurfaces of the heads of the inner and outerrails at different positions, there is clearlydifferent wear on the inner and outer railsof curves. We are developing rail-headlubrication to solve these problems.Currently, each time a train passes througha difficult curve, a greasing mechanismmounted on the track lubricates the innerrail head to reduce wear (Photo 4).However, this method has problems with

Figure 3 Roles of Technical Center

JR East Group

e@train as future railway systemJR East Research and Development Center

World-No. 1 Maintenance

Technical Center

Internalization of railwaytechnology

Internalization of railwaytechnology

Build up fundamentalconcepts and evaluation

techniques

Build up fundamentalconcepts and evaluation

techniques

Build up maintenance andmanagement techniques

Build up maintenance andmanagement techniques

Aim for maintenance innovationAim for maintenance innovation

Supportmaintenance

Improvereliability

Cut costs

Figure 4 Fatigue-induced Shelling at Rail Head

Photo 1: Corrugation wear along rolling surface of railhead (JR East)

Photo 2: Wear on inside edge of outside rail on curve(JR East)

Photo 3: Rectangular wear at inside arc profile ofwheel flange (JR East)

Japan Railway & Transport Review 36 • September 200350

Railway Technical Research in Asia

Copyright © 2003 EJRCF. All rights reserved.

wheel slip, and sliding, etc., and also withcontamination such as grit sticking to therail and causing more wear. As a result, itcan only be used on some sections. Futuremethods for lubricating the inner rail willhave to be easy to use, environmentfriendly and minimize the wear factorwhile reducing corrugation wear along therail head as well as wear caused by lateralthrust. Conversely, they will have tosecure sufficient adhesion between thewheels and rail.On the other hand, for the outer rail, agreasing mechanism installed on theground lubricates the inside edge of therail while another greasing mechanismmounted on the bogie (Photo 5) of thecarriage lubricates the inside wheel flangeeach time a train passes through the curveby spraying lubricant onto the face of theflange running along the inside of theouter rail. In this case too, there areproblems with contaminants such as gritsticking to the lubricated rail as well asproblems with wheel slip, etc. The newlubricant developed for this purpose is

similar to current lubricants but is moreeffective at reducing the wear factor whilebeing water-soluble to reduce anyenvironmental impact. In addition, weare also looking at developing greasingmechanisms that apply a constant coat oflubricant reliably.

Rail surface fatigueWhen a wheel runs repeatedly over a railas shown in Figure 6, a fatigue layer isformed the point of rolling contactbetween the wheel and the rail. As thefatigue leads to rail defects, the surface ofthe rail starts shelling, then metal scalesform internally. In the serious cases, therail may actually break. When shelling isdiscovered the usual procedure is toreplace the rail, but if the damage is nottoo serious it is possible to grind the railsurface and remove the wear. Thisprocedure can be quite effective and isexpected to play a large role in cuttingcosts and preventing accidents. Animportant R&D theme is assessing whento perform grinding and how to perform

it. Another aspect is analyzing data onthe weight and speed of different types ofpassing trains as well as on the trackgeometry in relation to the occurrence ofshelling. One method is to intentionallygenerate shelling using a rolling tester andthen to confirm the effect of grinding onthe fatigue layer, the best frequency forgrinding and the required amount ofgrinding.

Pantograph and Catenary Wire

Figure 7 shows the relationship betweenthe catenary and the pantograph. Thepantograph slip plates press directly onthe catenary trolley wire to collect current.To keep the trolley wire horizontally flatwith no sag, it is hung on hanger wiresfrom a messenger wire running in a seriesof long arcs between support structuressuch as poles and beams, etc. On theother side of the interface, the pantographis constructed of an arm and a collectorshoe or wing carrying the slip plates forcollecting the current. The figure showsa modern single-arm design. If the slipplates separate from the trolley wire,arcing damage can occur and there arealso problems with frictional wear due tomechanical abrasion as the slip plates runalong the wire at high speed. A furtherproblem is the temporary loss of powerwhen the slip plates separate from thetrolley wire. In terms of maintenance, themain work involves managing wear

Figure 5 Optimizing Wheel and Rail Contact Figure 6 Managing Shelling by Rail Grinding

Photo 4: Track-mounted greasing mechanism(JR East)

Photo 5: Bogie-mounted greasing mechanism(JR East)

Japan Railway & Transport Review 36 • September 2003 51Copyright © 2003 EJRCF. All rights reserved.

occurring on the catenary trolley wire andpantograph slip plates (Photo 6). The mostdifficult and most costly work involvesreplacing the trolley wire. For example,replacing the trolley on the Yamanote Linein Tokyo has to be performed once every6 or 7 years and in terms of annualmaintenance costs, trolley wire repairs areabout 20 times more expensive than slipplate repairs. Consequently, our researchat this interface is endeavouring to cut thecosts of trolley wire maintenance.

Contact measurement pressureAs shown in Figure 8, the pressure of thepantograph slip plates on the trolley wirefluctuates greatly as the train runs along thetrack. The highest pressure occurs at thecatenary supports (points A and C) and it isalso reasonable to expect large variationsin the pressure at locations where there arefaults such as disconnected hangers, etc.As a consequence, we are proceeding withresearch into a system that can be used tomeasure the contact pressure automaticallyfor troubleshooting the catenary.We have taken the first step in developinga system for measuring contact pressureautomatically by installing battery-powered telemetry test equipment on theroof of a Series E3 shinkansen carriage(Photo 7). This sends measurement databy radio transmission to in-carriageequ ipment fo r subsequen t da taprocessing. We have performed two testruns between Sendai and Kitakami inJanuary and November 2002 and Figure

9 shows an example of some of theobtained data.Looking at this chart, we can see thereare particular places where the contactpressure drops, in other words places,where the pantograph separates from thetrolley wire. In addition, there also pointswhere the contact pressure spikes quite

Figure 7 Pantograph Contact Figure 8 Changes in Pantograph Contact Pressure

Photo 6: Missing and damaged slip plates (JR East) Photo 7: Catenary measuring equipment on roof ofSeries E3 shinkansen test train (JR East)

suddenly. These might be explained bythe so-called ‘overlap section’ where aterminating wire runs in parallel forseveral meters with a beginning one inseveral meters. Clearly, by using this typeof contact pressure measurement systemwe can quickly get an understanding ofthe condition of the catenary and we

Figure 9 Pantograph Contact Pressure Data

Japan Railway & Transport Review 36 • September 200352

Railway Technical Research in Asia

Copyright © 2003 EJRCF. All rights reserved.

believe that this will play a major role incutting the catenary maintenance costs.Although the system is already producingresults, there are still some potential areasfor improvement; in the future we intendto estimate of the aerodynamic lift forceduring running to obtain an even moreaccurate contact pressure.The future development involves testingthe bas ic charac te r i s t ics o f themeasurement system using opticalsensors, installing the system on JR East’sinspection car, testing the installation withtest runs and diagnosing the catenaryusing the obtained contact pressure data.

Next-generation Crossoversand Switches

Current problemsThe direction of a train passing throughpoints is controlled by the movement ofthe tongue rails and the structure can beextremely complex, which is a weak pointin these types of large ground infrastructure.Certain points error can result from poormating between the tongue rail and mainrail due to deformation like that shown inPhoto 8.Another weakness of points is that theiroperation can easily be disrupted. Thepresence of a foreign object such as aballast stone between the main rail andtongue may leave the tip of the rail open,causing a points error.Moreover, the signal failure tends to occur

due to the complicated crossovermechanism. Generally, the left and rightrails are isolated to detect the position ofa train. When a train comes, the left andright rails are shorted electrically with awheel and an axle, and detect a train by atrack circuit.However, sometimes even when there isno train coming, slippage of the left andright tongue rails may cause contactbetween the switch bar and switchadjuster rod creating a short.Moreover, the presence of an electricallyconductive foreign object, such as anempty metal can may cause a short at theisolator, forcing the signal to remain at redand causing disruption to operations.In addition, the maintenance of pointsrequires tremendous manual labour.Typical examples might be lubrication ofpoints or adjusting the switch barmechanism.We are currently involved in R&D into next-generation points and switches for the 21stcentury. Our development basis is fault-free and low-maintenance structures.

Increased rigidity by grid sleepersPoints receive tremendous and repeatedlateral forces each time a train passesthrough them. As a result, as shown inFigure 10, the sleepers tend to be displacedhorizontally, creating changes in the trackgeometry. To prevent this, we havedeveloped and are using so-called gridsleeper track in which the sleepers run

longitudinally bound by crossties. Thisstructure has much greater durability andis able to resists changes in track geometry.

Improved base plate for morereliable switching motionSince the tongue rail has a sliding baseplate, the main rail can only be securedby a tie fastening on the outer side of therail. With our new design, the lowerheight of the tongue rail permits use of ahigher base plate (Fig. 11). As a result,the main rail can also be gripped frombelow. In other words, the main rail issecurely fastened on both sides,suppressing rail slippage. In addition,foreign objects can fall into a small spacebetween the tongue rail and main rail,reducing the possibility of switchingerrors. Finally, since it uses a bearing baseplate, lubrication is unnecessary.

Innovative switch machineTheoretically, we think it is possible toachieve a 75% weight reduction in thisvery heavy piece of equipment from400 kg down to around 96 kg; we arealso achieving a 66% reduction in thearea of the switch drive mechanism.These advances have been achieved byusing lightweight aluminium alloys forthe switching frame as well as smallhigh-performance servomotors and anew link design.With this development, the status of thesenew switches can be easily confirmed by

Photo 8: Deformed main rail next to switch tongue rail (JR East)

Figure 10 Grid Sleeper Track at Switches

Deformationand damagedtongue rails

Today Repeated lateralforce

Future

7 times greater resistance to lateral forces

Japan Railway & Transport Review 36 • September 2003 53Copyright © 2003 EJRCF. All rights reserved.

Hiroaki Inazu

Mr Inazu is Director of the Technical Center in the Research and Development Center of JR East

Group. He joined JNR in 1973 after graduating in engineering from Tohoku University. He has held

various posts in the operations and rolling stock departments of JR East.

anybody using a number of lit LEDs.Furthermore, the switch controller has beenchanged from the previous type of relaycircuit to an electronic circuit with datatransmission using optical fibre cables. Asa result, problems with the relay contactsare eliminated by digitization that alsopermits remote monitoring of faults. Inaddition, the main body of the older switchdesign was used for mounting otherancillary equipment, increasing theinstallation footprint area and making itdifficult to secure a switch to a sleeper usinga multiple tie tamper. Similarly, anyshortage in the amount of ballast stonecould easily lead to movement anddeformation of the track geometry. Thesmaller lighter simplified structure makesit possible to install the entire switch bodyand peripheral equipment on a singlesleeper, permitting much easier mountingusing multiple tie tamper and ensuringsufficient ballasting.

Improved switch ancillarydevicesAs I explained previously, currently thereis the problem with shorting if aconductive foreign object such as anempty metal can falls centrally betweenthe rails. This has been solved relativelysimply by installing insulators at twolocations, greatly reducing the incidenceof track shorts. We have made otherrevisions but probably the most effectivein simplifying difficult work has been

Figure 11 Improvement of Base Plate

High base plate

Fastening Base plate

Space

New tongue rail

Main rail Tongue rail

Base plate

Today

Future

Figure 12 Pin Coupling at Switch Bar

Today

Future

Coupling plate

Double locknuts

Nut

Linking plate

Safetyfastening

Switch bar

Coupling plate Welded pin

Switch bar

lightening the weight. As a consequence,the ancillary equipment can be installedat the work preparation stage and coarseadjustments made at the same time. Thismeans that the on-track work time can bereduced while cutting the number ofdifficult operations too (Fig. 12).Our above-mentioned next-generationpoints have been installed in OmiyaStation yard on the outskirts of Tokyo sinceFebruary 2002. We hope this encouragingresult will soon lead to their fullcommercial introduction throughoutJapan and help reduce track-relatedaccidents and maintenance costs.

Summary

As described in this article, our TechnicalCenter is pursuing R&D in a diverse rangeof fields related to solving maintenanceproblems especially at the interfacebetween different technologies. But ourR&D is not just related to technologyinterfaces and we are working hard toinnovate railway maintenance in many

other areas. As other articles in this issueof JRTR have mentioned, this workrequires efficient introduction of world-leading, advanced and new technologies.To this end, the Technical Center is alsoconducting joint research with railwaylabora to r i e s a round the wor ld ,universities, rolling-stock manufacturers,other industries, etc. This type of R&Dcan only be conducted in concert withother experts in the field and we areactively considering how to create newinformation systems for sharing our resultswith others. I hope that readers of thisart icle around the world wil l beencouraged to contact and work with uson ways to improve the world’s railways.

�