4 Japan Railway & Transport Review / October 1994

SPECIAL FEATURE – 30 YEARS OF HIGH-SPEED RAILWAYS

Copyright © 1994 EJRCF. All rights reserved.

Future of High-Speed Railways

1. Significance and Merits/Demerits of Shinkansen

The technical content of theshinkansen was first made public at alecture held by the Railway TechnicalResearch Institute (RTRI) of JNR cel-ebrating the 50th anniversary of itsfoundation in 1957 when Japan stillhad yet to recover completely from thedefeat of World War II. In the sameyear, the Odakyu Electric Railway andRTRI jointly developed the first real su-per express car in Japan establishing anew Japanese record of 145 km/h in atest run executed by JNR.

The successful development of theshinkansen owes most to two greatmen: Mr Shinji Sogo, the President ofJNR, who believed in the RTRI groupand showed strong leadership to pro-mote its development under difficultcircumstances; and Mr Hideo Shima,the Chief Engineer, who coordinatedtechnical affairs under the leadershipof Mr Sogo.

From the technical viewpoint, theshinkansen owes its success to selec-tion of a completely different gaugefrom the narrow-gauge (1067 mm)conventional lines which are unsuit-able for high-speed running, as well asto design concepts combining proventechniques as far as possible.

The technical and business successof the shinkansen came as a greatshock to the conservative railway in-dustries in Europe and America whohad long believed that the maximumworking speed of railways was about160 km/h. It also led to the revival ofEuropean railways. High-speed pas-senger trains were even tried in theUSA, where inter-city railway passen-ger transport was already in declineand where railways were no more thana freight transport business.

Five years after the shinkansen com-missioning, a plan was developed in1969 to operate a super express non-

stop between New York and Washing-ton in 2 hours and 30 minutes. How-ever, the plan failed due to technicalproblems. It was later modified to intro-duce Swedish technology, and today,the trains connect New York withWashington in about 2 hours and 50minutes.

On the other hand, after the TokaidoShinkansen, high-speed railways in Ja-pan meant no more than the TokaidoShinkansen, making real technical de-velopment—developing and applyingdifferent techniques in different condi-tions—difficult. Concretely speaking,almost the same system as the TokaidoShinkansen was used for the TohokuShinkansen despite their great differ-ences in the magnitude and seasonalvariation of demand. This promulgatedtheir characteristically-poor currentcollection by multiple pantographs formore than 20 years. In addition, therewere almost no speed increases in areaswithout shinkansen lines.

The Tokaido Shinkansen was con-structed on a very favourable routewith the largest and most stable de-mand in the world. Therefore, despitethe huge investment, it soon became

profitable as long it offered high-qualityservice. However, constructingshinkansens does not pay elsewhereunless the construction costs can be cut,and improvement of conventional linesis usually more efficient than construct-ing a new line. It is regrettable that thegreat success o f the TokaidoShinkansen nipped diversified techni-cal development in the bud.

2. Quick Response inEurope

European railways, which had a longexperience of higher-speed operation atup to 160 km/h, responded quickly tothe shinkansen. France and Germanyquickly developed locomotives withhigh power and excellent performanceas well as new passenger cars with railbrakes, new signalling systems, etc.Services soon started at 200 km/h onfavourable lines. Germany planned toreach service speeds of 200 km/h asearly as 1965, but high-speed operationwas forcibly suspended due to track de-terioration. However, as a result of the1966 French success, achieving 200km/h with a train hauled by an electric

Satoru Sone

� British Rail HST at Paddington Station (T.SUGA)

Japan Railway & Transport Review / October 1994 5Copyright © 1994 EJRCF. All rights reserved.

locomotive was no longer such a mile-stone success.

On one hand, 200 km/h operation wasachieved on conventional lines by ex-tending conventional technology, andon the other hand further speed in-creases were tried by developing newtechnology and constructing new lines.The success of the shinkansen appar-ently started revitalisation of railwaysworldwide.

3. Shinkansen DrawbacksAnswered in Europe

The British HST (High Speed Train)was the first to achieve high-speed op-eration. Trunk lines in Britain suithigh-speed running because the curvesare gentle and there are almost nocrossings. The British planned to in-crease the speed of the HST using theAPT (Advanced Passenger Train) with-out constructing new lines. The HSTreached 201 km/h on non-electrifiedsections, while the APT reached 250km/h on electrified sections. HST opera-tion started successfully in 1976.

The first commercial HST, or 253 Se-ries, was made up of two lightweight(67 tonne), high-power (1680 kW) dieselelectric motive power units with asmall baggage room and a small con-ductor room as well as 7 trailers, or pas-senger cars, between them. Currentknowledge of high-speed railcars sug-gested that with this formation, the

train should not have been able to ac-celerate when the adhesion was poor,and the braking distance should havebeen poor. But, in practice, the trainshowed better acceleration/decelera-tion performance than Japaneseshinkansen and ran stably at highspeeds. Japan could have learnedmuch from the HST including its high-speed adhesion performance, and re-duced running resistance achieved bycovering under-floor equipment, etc.

The TGV (Train à Grande Vitesse)was developed by France in 1981 as avehicle dedicated to new high-speedlines. It resembles the Japaneseshinkansen in purpose, but differsgreatly in design philosophy. The dif-ferences are attributable to overcomingthe disadvantages of the shinkansen.Namely, the TGV was completed bylearning the lessons of the shinkansen.Table 1 shows the main differences.When France worked out these solu-tions in 1981, it was Japan's turn tolearn from the TGV, but JNR did notevaluate the TGV soon enough or cor-rectly. Since then, we have learnedmany things from the TGV and have atlast succeeded in improving the notori-ous current collection system.

4. Unfortunate JNR EraLooking back over 30 years of high-

speed railways in Japan, the unfortu-nate JNR era cannot be forgotten.While the South-East TGV line was un-der construction in France, the Tohokuand Joetsu Shinkansen were underconstruction in Japan. However, it isdifficult to say that 20 years of experi-ence of the Tokaido Shinkansen was re-flected in any technical or functionalimprovements. Neither were the eco-nomics of these lines thoroughly evalu-ated considering the comparativelysmall demand. In fact, their construc-tion costs were too large, compared tothe functional improvements. This isattributable to JNR's critical situationregarding labor-management relationsand finance. This situation was re-flected in the technical field, making itimpossible to develop new technology.

When the JNR div is ion andprivatisation was finalised, the situa-tion was gradually resolved, and thestagnating railway service and technol-

ogy began showing rapid improvement.New speed records and new rollingstock are now being developed eachyear.

5. Different Japanese andEuropean Styles

Even now, the technology and servicediffer greatly between Japan and Eu-rope. It is useful to compare these dif-ferences, even if the conclusion con-firms that Japanese style matches Ja-pan, and European style matches Eu-rope.

Re-recognition of RegenerativeBraking and Distributed Traction

For comparatively long trains usingDC motors, concentrated traction isprobably more economical than distrib-uted traction, based on the car con-struction and maintenance costs. How-ever, in Japan, because of the quality ofcouplers and shock absorbers and be-cause the push-pull system is uncom-mon, most passenger trains includingshinkansen have distributed traction.Shinkansen used fully-motorised axlesas a matter of course in the early stage,because of insufficient knowledge of ad-hesion limits. However, when the HSTand other new trains showed that fully-motorised axles were unnecessary, con-centrated traction slowly progressed aslong as floor space was not sacrificed.The Series 100 shinkansen manufac-tured in 1985 has 12 motor cars and 4trailer cars (12M4T). Two or fourtrailer cars are double-decker to in-crease seating capacity. A feature ofthe Series 100 trailer car is its brakes;an eddy-current disc brake activated bythe motor current of the adjacent M caris used to lower the maintenance costs.

The Series 300 was developed withthe same philosophy in a 10M6T forma-tion. However, while the weight of themotor and driving gear was greatly re-duced, the weight of the eddy-currentdisc brake could not be reduced asmuch, resulting in heavier trailer carsthan motor cars for non-suspended andtotal mass.

Since AC motors require hardly anymaintenance, fully-motorised axlesshould have been used for the Series300 . As a result of learning this lesson,the Series 500 had fully-motorised ax-

� French TGV Sud-Est at Paris (T.SUGA)

6 Japan Railway & Transport Review / October 1994

SPECIAL FEATURE – 30 YEARS OF HIGH-SPEED RAILWAYS

Copyright © 1994 EJRCF. All rights reserved.

Table 1 TGV Improvements Based on Shinkansen Drawbacks and Influence on Japanese Railway

Quoted from introduction "Recent High Speed Railway in Japan and Europe" (by Satoru Sone) to special edition " High Speed Railway—Comparisonbetween Japan and Europe", OHM, April 1994 (partly modified)

Japan Railway & Transport Review / October 1994 7Copyright © 1994 EJRCF. All rights reserved.

les with the motor mounted on eachaxle but centralised power supplies.This system, which centralises powerunits if floor space permits but distrib-utes traction motors, has the advantagethat all floor space can be used as pas-senger room. High-speed operation isfacilitated because the total mass pereffective floor area is small, the maxi-mum axle load is small, and mechanicalbraking is not usually used. When re-generative brake equipment is used forservice braking, the heavy brakingrheostat—a large onboard source ofheat—becomes unnecessary, greatly in-creasing safety.

Fully-motorised axles are more ad-vantageous for AC driven high-speedrailcars and will prevail in the future.

Electrically-Connected DualPantograph

Comparison between the TGV cur-rent collector and the Series 0shinkansen, which uses 8 sets of panto-graphs distributed over the 16-car for-mation and generates much sparking,shows the superiority of the TGV. Toincrease the speed of the Series 200shinkansen which has 6 sets of panto-graphs along the 12-car formation, thepantographs were connected by bus toinvestigate the effect on the current col-lection. It was concluded that two com-paratively-separate pantographs arebest, except for those at the ends of theformation, because even if one panto-graph bounces, current collection iscontinued by the other pantograph.Also the bouncing pantograph does notgenerate much sparking. The Series200 and 300 shinkansen operating upto 275 km/h have three sets of panto-

graphs, but only 2 sets are used. Thenumber of pantographs will soon be re-duced to two. This method of currentcollection seems superior to the TGVelectrically-connected one-pantographsystem and the ICE independent two-pantograph system.

Japanese Advantages in Car BodyTilting Control

Although many countries have beentrying to develop effective car body tilt-ing control systems, few have succeededin improving riding comfort. So far,only Japan, Italy and Sweden havesolved the problem. I have comparedthese three methods on commercialtrains (JR Shikoku Series 2000 DMUrailcar and Series 8000 EMU railcarand JR Hokkaido Series 281 DMU rail-car, FS ETR450 and DB VT610 for theItalian method, and X2000 for theSwedish method). All have no problemwhen passengers are seated, but thereare some problems when they walkalong corridors.

Since riding comfort differs greatlywith track conditions and runningspeed, it is difficult to determine whichmethod is superior. The Japanesemethod is based on a natural pendu-lum. The problems due to time lag inthe pendulum motion on short curveshave been solved by "feed-forward" con-trol of the time lag calculated from cur-vature data stored in an onboard com-puter, based on train speed and currentposition determined by track detectors.This makes passengers feel that thetrain is running on a gentle curve andoffers a comparatively good ride evenunder bad track conditions. From theviewpoint of adaptability to varying

track conditions, the Japanese methodis good. There are other possible advan-tages too such as automatic collection ofdata for track maintenance.

Actual Adhesion LimitHigh-speed running of the TGV-A

with a centralised power plant (4M11Tin bogie units), including operation inrain, showed that the speed limit offully-motorised axles is high enough(approximately 460 km/h). This meansthat adhesion itself does not limit thecommercial speed of wheel-rail rail-ways. But this does not mean that anadhesion railway system running at460 km/h will be practical. Consideringthe weight and size of current electricalequipment, such trains would be nomore than locomotives, offering hardlyany passenger space. In addition, thereare many other problems other thanadhesion to solve, such as noise. There-fore, it is unfair to insist from the adhe-sion viewpoint that development of newsuper-speed railways is useless.

High-Speed Running TestAt high speeds, aerodynamic drag

dominates running resistance. The re-sistance is proportional to the square ofthe speed, while power is proportionalto the cube of speed. From this point ofview, it is unnecessary to establish aspeed record higher than 500 km/h inthe high-speed running test to achievea maximum service speed of approxi-mately 350 km/h. Therefore, althoughthe speed record of 515.3 km/h estab-lished by the TGV is a great record, it isnot indispensable to achieving commer-cial railway speeds of approximately350 km/h.

For practicability, a speed 10% to20% higher than the maximum servicespeed is sufficient. It is more importantto perform sufficient running tests un-der various conditions than to aim justat higher speed.

Because of the distributed traction,� JR Shikoku Series 8000 EMU (JR Shikoku)

� JR Hokkaido Series 281 DMU (JR Hokkaido)

8 Japan Railway & Transport Review / October 1994

SPECIAL FEATURE – 30 YEARS OF HIGH-SPEED RAILWAYS

Copyright © 1994 EJRCF. All rights reserved.

unlike the TGV, the Japaneseshinkansen cannot increase the power/weight ratio by removing intermediatecars. Test cars have been designed toachieve the real test speed, or about 1.2times the maximum service speed, byreplacing trai l ing bogies withmotorised bogies to increase power(STAR 21 of JR East) or concentratingthe traction power of four cars intothree cars (WIN 350 of JR West).

Is MAGLEV Development Useless?It is unreasonable to say that

MAGLEV is useless based on the TGV515.3-km/h record, which was achievedon a down grade with a dry track.

Introduction of new high-speed rail-ways is opposed from various viewpoints,including noise, vibration, energy saving,efficient land use and safety. Selection ofwheel-rail or MAGLEV systems is re-lated to all these items. Both systemsshould be evaluated and the superior oneshould be selected. There is an opinionthat high-speed surface transport is notnecessary, because there are strong com-petitors like airlines. However, if it is pos-sible to develop a new system that is supe-rior to aircraft in terms of energy sav-ing, efficient land use and safety, natu-rally, it should be promoted. The mostimportant problem in Japan concerningconstruction of high-speed railways isnoise. In this respect, MAGLEV has ad-vantages because the car body is verylight and the traction motors are installedon the ground, so no noise is emitted bycurrent collection.

However, MAGLEV railways requiremany new developments and are not so

advantageous at low speeds. There-fore, MAGLEV is less attractive forshorter development times, or lowertarget speeds. If it is necessary to de-velop a new railway system comparableto shinkansen or TGV quickly,MAGLEV will not be selected. How-ever, the size of present electricalequipment makes it difficult for wheel-rail type railways to secure sufficientpassenger space. For these reasons,MAGLEV is attractive at 450 km/h orhigher speeds. If such a system canachieve greater energy saving, more ef-ficient resource utilisation and highersafety than aircraft, it will be worth de-veloping.

5. How Should RailwaysImprove Transportation?

Further Speed IncreasesSpeed is the most important factor in

the transport competition. However,speed does not mean the maximumspeed but the total trip time includingaccess time and waiting time. Althoughspeed has universal importance, the en-thusiasm of the transportation indus-try and society for speed increaseschanges with time.

Of course, at some times it may bedifficult to invest large amounts ofmoney just to shorten the travel timeslightly. However, even in recessions,speed increases can be promoted fromthe software viewpoint, for example,by re-examining stops, lines, and con-nections, etc. It is always important tokeep speed increases in mind.

Safety and EnvironmentalFriendliness

Typical transport modes competingwith high-speed railway are aircraftand cars. Although these three modesof transport should each develop intheir own fields to satisfy people's di-verse needs, more passengers should beattracted to high-speed railways, be-cause railway has a decisive advantagein terms of safety and environmentalfriendliness. Railway personnel shouldmore actively appeal this point to gov-ernment, companies and individuals.And railway staff should more activelytackle correction of systems ororganisations obstructing development

Change from National toInternational Railway Technology

Although railway engineers in eachnation share common bonds, internalcooperation is still insufficient. Tomake railways more attractive and tomeet social needs such as environmen-tal conservation, safety and energy sav-ing, we must promote information ex-change and cooperation.

The change from JNR to JR in Japanis one step typified by publication ofthis magazine. However, comparedwith Europe, the attitude in Japan to-wards exchange and cooperation is stillinsufficient.

Readers must cooperate actively byparticipating in the debate in thismagazine. �

� JR Central Series 300 Nozomi at Tokyo Station (T. SUGA)

Satoru Sone

Dr Sone was born in 1939 in Tokyo. He has long been interested in railways and graduated from

university in electrical engineering because it seemed to have a great influence on new railway

technologies. He has been a professor at the University of Tokyo since 1984, engaging in the

study of transport system engineering and power electronics. He plays a positive role in the

government-university-industry debate on transport policy.