3.1.1 Original technologies of Japan and of China
It can be thought that all or almost all technologies of HSR are of Japanese origin because Japan was the pioneer in the world for HSRs, but actually this was not the case. When Japan decided to construct a Shinkansen between Tokyo and Osaka, the conventional Tokaido line, which was and still is a double tracked electrified railway with comparatively short quadruple track sections near Tokyo and near Osaka, was approaching its limit of transportation capacity and the quick addition of two more tracks was an absolutely necessary requirement. The majority of the then Japanese National Railways (JNR) managers intended to add a double track route alongside the conventional narrow gauge tracks. To add two more tracks on the same route was thought extremely difficult in the already densely populated area in a short period of five targeted years, so an alternative idea to construct a separate route with necessary interchanges was the most viable. But there were a few people who had strongly wanted to improve the JNR by changing the narrow gauge to standard gauge since the end of 19th century, less than 30 years after the first railway opened in Japan. The then president of JNR, Shinji SOGO (十河 信二) was one of them, but all other influential JNR people were not. Shinji SOGO and Hideo SHIMA (島 秀雄) who had been retired after severe accident on the JNR and recalled by SOGO as Chief Engineer, together proposed strongly to add a standard gauge new double track line instead of a quadrupled conventional line. SOGO and SHIMA were able to narrowly persuade the rest of JNR headquarters and the Shinkansen was therefore successfully initiated. In this situation, the Shinkansen had to realise enough traffic capacity for both expanding passenger and freight traffic together with the conventional line with enough availability and by the targeted opening date. Thus, all construction and trains together with the operating systems design were very conservative without any important failure allowed. Conservative design means not too innovative, rather heavy cars, less speedy movement, and a tendency to have an excess margin in vehicle safety and a small potential to increase further the train speed in future track design.
3.1.2 Original design of Japan’s cars: development of light-weight bogies with stable operation at high speed (Sone, 2014)
The Shinkansen’s car was based on a very innovative super express (SE) trainset of the Odakyu Electric Railway (OER), which established a world speed record on a 1067 mm gauge railway of 145 km/h in 1957 and a tare weight of 147 t with 108 m length. The Shinkansen’s Series 0 cars were comparatively heavy with an average 56 t tare weight and each was 25 m long. To avoid a further increase in mass, a wheel base of 2.5 m and a wheel diameter of 0.91 m was chosen and proven to have enough stability margin to run up to 210 km/h. JNR intended to reduce each car by 4 t, to realise a maximum axle load of 15 t, but failed mainly because of the mass of the traction equipment with enough margin. This meant an axle load of 16 t was narrowly met including the fully seated passengers.
3.1.3 Original design of Japan’s cars: distributed traction system with all axles motored
Based on the then Japanese Private Railways (JPR) standard design of high performance electric multiple units (EMUs), the traction system of paired motor cars with eight traction motors controlled by a traction controller was a natural and unique solution and never a result of the optimised design of a number of cars, motored and non-motored, i.e., trailer cars. After Japan’s success with the HSR, Britain and France followed in 1976 and in 1981, respectively, but with concentrated traction systems of different types. Until the 1990s, when AC motor tractionFootnote 1 was available, concentrated traction systems were thought by many railway engineers other than Japanese to be superior to distributed traction with less maintenance of traction motors and more maintenance of non-motored axles’ braking system. But in 1992 Japan’s Shinkansen succeeded in introducing the world’s first high-speed EMU with AC regenerative braking, as the Shinkansen’s Series 300 (Fig. 1). In R&D works for this innovative trainset (the author was one of the team), a very important discovery was revealed; a non-motored axle with a necessary brake has more mass (total and unsprung mass) with more heat and more maintenance works required than a motored axle mainly because of the theoretical results of the braking system requiring kinetic energy to heat energy conversion. Thus, in Japan the next Series 500 was changed from 10M6T formation to 16M(0T), mainly to reduce the unsprung and total mass. This important knowledge was presented in 1994 at the World Congress on Railway Research (WCRR) in Paris by the author (Sone, 1994).
3.1.4 Chinese design concept
When China decided in 2003 to introduce HS trainsets from developed countries, it excluded concentrated traction systems such as the French TGV (Train à Grande Vitesse), German ICE-1 (InterCity Express), and Italian ETR500; this sound and clever decision was supposed to be based on the above-mentioned Japanese result as well as China’s own experience of difficulties in the R&D of China Star (中華之星), with concentrated traction, against Changbaishan (長白山), with distributed traction trainsets. All of the introduced CRH1, CRH2, CRH5, and CRH3 have distributed traction systems but there are two different types; CRH2’s with driving trailers and the rest of CRH1, CRH5, and CRH3’s all of which are of European origin, with driving motors (Zhang, 2009a). This difference is, from Japanese experience, very important from the following two viewpoints: from the electromagnetic compatibility (EMC) point of view, a driving motor whose front bogie has traction motors brings much more trouble to signalling circuits, and from the adhesion point of view as well. Wheels of the front bogie tend to slip and slide because they are still wet in rainy condition. On the contrary, a formation with driving trailers and intermediate motors does not require sophisticated adhesion control and if there is no brake applied at the front bogie, the wheels can be used as a reliable speed and position sensor because they do not slide at all (Fig. 2).
At present, Chinese designed HS trainsets, CRH380’s, are of the above-mentioned two types; CRH380A and AL consist of driving trailers with all intermediate cars motored while the rest of CRH380B, C, and D have driving motors and intermediate motors and trailers with the same numbers of motors and trailers: 4M4T or 8M8T (Yang, 2015). If Chinese railways want to standardise the traction systems, the author strongly proposes driving trailers with as many numbers of motors as are economically viable. Even if a sophisticated and high performance re-adhesion control is used, small slip and slide are inevitable under adverse conditions, meaning that the acquired speed and position have many errors. If CRH380AL’s 14M2T formation is thought to provide too much margin and is too expensive compared with CRH380A’s 6M2T formation using the same traction motors, the 12M4T formation may be the best solution for construction costs, but still the running cost may be better for the 14M2T because of much better braking performance.
3.2 Comparison of achievements of the HSR in Japan and in China relating to subsystems used
3.2.1 Power feeding system
Among Japanese original technologies, one of which that has not yet been adopted in other countries is the quasi-continuous power feeding system using two consecutive catenary sections and wayside change-over switch (Ishikawa et al., 2015). The system’s action is as follows (Fig. 3): suppose the train moves from A to C via B, the switch D is kept closed until the whole train enters the section B, when D is off and a very short time, typically 0.3 s, after opening D, switch E is closed. This system together with simple but superior current collection using two interconnected pantographs achieves nearly perfect current collection because of no interrupted acceleration and deceleration using a regenerative brake and stable electrical connection even when one pantograph loses mechanical contact (Study Group of High-speed Railways, 2003). At the moment the European system prohibits the use of a parallel connected two pantograph system to prevent short circuiting from two power sources in different phases. The change-over switch was improved from the original air blast type and then changed to a vacuum switch and now gradually to a semiconductor type.
China studied using a quasi-continuous power feeding system but has abandoned it for the moment. The reason was supposed to be easy acceptance of the European type trainset and changeover switches are still being improved from vacuum breaker type to semiconductor ones.
3.2.2 Electromagnetic compatibility
In Japan there are many special designs for avoiding EMC problems based on experienced difficulties. Contrary to the above-mentioned standard practice, Series N700 of 8 car formation in the Kyushu and the Sanyo Shinkansen are not 6M2T with driving trailers but 8M0T with driving motors to cope with restarting conditions after stopping on a steep gradient section of 35‰ with half the units cut-off (Fukunaga, 2015). To cope with EMC problems due to driving motors, traction motors and bogies are of slightly different designs for the front bogie, rear bogie of the front car and of intermediate cars both from an emission and immunity point of view. As far as adhesion is concerned, acceleration and deceleration forces of the front two bogies are squeezed while the rest, including the rearmost cars, produce full force.
3.2.3 Phase balancing measures after introduction of regenerative trains
The author should touch upon a practice and its changes in phase balancing measures relating to a single phase high power load from a three phase power grid. At the start of the Tokaido Shinkansen in 1964, three phase AC was converted to rectangular two phase AC using Scott connected transformers. One phase was fed to down tracks and the other to up tracks. Thus, on average on most occasions a better balance was established because average power was almost balanced between down and up trains. The situation was changed in two ways: in a change from a booster transformer (BT) feeding system to an auto-transformer (AT) system and the introduction of regenerative trains. A BT was used to pick up return current flowing in the running rails to a negative feeder located about the same height as the catenary wires so that induced electrical noise to nearby communication wires was cancelled. This was good for avoiding EMC but for current collection the BT section was a weak point where the terminal voltage of BT should be broken by leaving pantograph producing electrical arc every time a train passed across the BT section. To cope with this problem, the BT feeding system was changed to an AT feeding system and according to this the power balance to be expected was changed from between up trains and down trains to between up and down trains running on one side of a substation and those on the other side.
The other important change in the situation was the introduction of regenerative trains in 1992. The principle of three to two phase conversion by a traction transformer is that when the power sum of one of the two phases is equal to that of the other phase, the three phase side is balanced. Before the introduction of regenerative braking, the worst case is maximum power at one phase and zero at the other phase. In this worst case three phase to two phase conversion has no effect; the same as taking the total power from one of the three phases alone, but after the introduction of a regenerative brake the worst case is positive power on one side and negative power on the other side; in this case three to two phase conversion has a negative effect; the situation is much worse than taking both positive and negative power from one of the three phases.
To cope with this adverse effect together with a voltage stabiliser both on the grid side and the train side, additional equipment called a railway static power conditioner (RPC) was introduced where it was thought necessary. An RPC can transfer active power from one side to the other and absorb necessary reactive power on each side independently (Ishikawa et al., 2015).
From a purely technical point, an RPC can be seen as sometimes necessary but from a totally economical design of the whole system, it is doubtful whether or not it is justifiable.
3.2.4 Result of light-weight design of cars
Know-how in constructing light-weight cars is by far the most advanced in Japan mainly because of the following three factors: (1) weak roadbed and infrastructures, (2) many privately owned train operators and car manufacturers, and (3) not too much buffer strength required as is necessary for heavy freight trains. After World War II passenger traffic demand expanded very rapidly between big cities and their suburbs, and the majority of this demand was carried on the suburban lines of the JPRs. Thus, light-weight and high performance EMUs were developed and manufactured from the early 1950s by JPRs which are the origin of the Shinkansen basic vehicle, with the other origin being AC electrification developed by the JNR. Some of the European car manufacturers think Japan’s trainsets of small mass means there is not enough strength in a crash and some others think there is not enough rigidity against vibration. The former criticism has been proven wrong on various occasions: one of the typical cases is the Wenzhou collision on 2011-07-23. In this case a light-weight aluminium bodied CRH2 hit a comparatively heavy stainless steel bodied CRH1 and damage was much greater to the CRH1. The latter criticism is half true and half incorrect; it is a matter of the design’s balance between riding comfort and energy saving. The mass per axle of less than 12 t of the E2 of the Tohoku Shinkansen, the origin of the CRH2, is the result of Japanese balance (Yang, 2012) and the roughly 15 t of the CRH380A is the result of Chinese balance (Yang, 2015).
3.2.5 Cab signal ATC
Although there had been some cab signal systems in smaller urban railways, automatic train control (ATC) using a cab signal adopted by a major railway was a world first for the Shinkansen (Kotsu Kyoryoku Kai Foundation, 2015) and since its inauguration more than half a century has passed but still it is not perfect nor is there a final version of such a system. Japan’s ATC is not intended for driverless operation but the braking operation is almost automatic with manual operation left for final stopping at a designated position at a train speed of below 30 km/h. The French TGV has a different philosophy; an automatic train protection (ATP) system should help the driver’s operation by showing a forthcoming situation and only if the driver fails to react safely does the ATP system intervene. German philosophy is a little different to that of France but does not differ that much.
China has introduced several signalling systems from Europe and Japan and at the moment several systems are coexisting on the same track according to the several different trainsets. Coexistence of several signalling systems on the same track is not desirable from the standpoints of drivers’ confusion, more maintenance work, and possible interference between the systems.
To solve this problem in the future, there will be much potential for solutions such as sorting the systems into categories by area, line groups, speed range, or climate conditions, etc. The process and results of the Chinese solution will be paid strong attention to by the developed countries, none of which has had such a need as yet.
3.2.6 Poor tracks and infrastructure in Japan
The Shinkansen’s weak points are often seen in the infrastructure; too little spacing between adjacent lines of 4.2 m (in the case of the Tokaido Shinkansen) or 4.3 m (Sanyo Shinkansen and all thereafter), a small cross section in double track tunnels of about 64 m2, a non-reversible signalling system, unique in the world, and a speed restriction of 70 km/h at turnouts into and out of stopping stations. The poor infrastructure was justifiable for the Tokaido Shinkansen because it was the pioneer, who cannot learn from others, and because of the fact that Japan was not economically strong enough at the start of the construction in 1959. After minimal improvement of the Sanyo Shinkansen, further improvements in line with world standards have not been realised.
In contrast to this, China can choose from many products around the world and a traditionally strong infrastructure for heavy load has been in due course further improved for the HSR. Some Japanese infrastructure specialists say that poor turnouts from the viewpoint of restricted speed should be justified to maintain necessary reliability by prohibiting two or more shifting motors, which is thought the main reason of turnout troubles.