The Chinese railway industry first developed a number of multiple units. The high-speed push began with trains for the upgraded Guangzhou-Shenzhen[Hong Kong] line: the DDJ1 prototype (1999), and a year later, applying lessons learnt with an X-2000 rented from Sweden since 1998, the eight DJJ1 "Lanjian" ( = Blue Arrow) trains, one of which achieved a national record of 264 km/h. However, in regular service, they only did 200 km/h.

A DJJ1 at Ex-Tutang station on 17 December 2006. Since then, these trains have been replaced by CRH1 units (see below). Photo from Rick W @Flickr

Seeing the difficulty and complexity of the technological challenge, a number of not much publicised prototypes followed, including "Xianfeng Hao" ( = Pioneer). Then, in 2002, came the intended real thing, the DJJ2 "China Star". After achieving only 321.5 km/h in trials, planned service top speed was scaled back to 270 km/h. Then it entered regular service in 2006 with just 160 km/h, and not long after, the apparently failed trains were mothballed.

A "China Star" disused already in 2007, left exposed to the weather in Shenyang. Photo by undiyz @ Hasea.com, taken from the blog of a Hong Kong netizen

Then the leadership chose another option: import foreign high-speed trains, with technology transfer. From 2007, the new subsidiary China Railway High-speed (CRH) began to run the following types:

• CRH1: a 200 km/h train from Bombardier's Regina platform (which originates in what used to be ABB's Swedish branch, maker of the X-2000)
• CRH2: the train whose origin is never mentioned in official Chinese media, being a spinoff of JR East's E2-1000 Shinkansen from Japan; first batch 250, rest 300 km/h
• CRH3: Siemens's Velaro CN, a wide-bodied export version of Germany's ICE-3 high-speed train, for 350 km/h
• (CRH4: it was speculated that this was reserved for an upgrade of the "China Star")
• CRH5: a 200 km/h train from the now Alstom-owned Pendolino platform

What's more, European and Japanese firms and technology was involved in line construction. Which is on a scale unthinkable elsewhere: the 935 km from Wuhan to Guangzhou and the 1318 km from Beijing to Shanghai are just the beginning. The first leg of the latter, China's first true high-speed section, the 117 km long Beijing-Tianjin high-speed line, is (also) fitted with Europe's long troubled signalling system ERTMS. On trials on this line, first a CRH2 reached 370 km/h (22 April), then two months later (24 June), the CRH3 set the current speed record for China: 394.3 km/h.

The line was opened on 1 August, a week ahead of the Olympics. The leaders didn't miss the opportunity to boast with the fastest (conventional) train service in the world, thus the CRH3 were ran at their top speed of 350 km/h. Where I note that

1. design line speed itself was pushed up progressively from 200 to 350 while the line was already under construction!
2. The line being relatively short, the CRH3 reach top speed only for a short time.
3. The Spanish version (which holds the rail speed world record for a series train without any modifications at 403.7 km/h) of the train doesn't yet do 350 km/h because the operators want that only when ERTMS Level 2 becomes truly reliable. (Also see: High-speed to Barcelona)

Show the best image for the Olympics: cleaners remove the stain of unlucky insects from a CRH3 train in Beijing Central Station, while a CRH2 waits on another track. Photo from SPIEGEL

:: :: :: :: ::

Now, all the above is very impressive. So, why am I sceptical about 380 km/h trains produced in 2010 and ready in 2012 when Beijing-Shanghai opens?

It's not power and speed. Even accounting for the rule that trains have to run safely at 10% above their service top speed (that would be 418 km/h for our Chinese Superstar), the goal is not out of reach. And last year, a TGV test train had several runs above 500 km/h culminating in a new rail speed world record of 574.8 km/h (I diaried). However, there are other factors:

1. Economics of enegy use. At high speed, air resistance is the overwhelming factor, and its force increases with the square of speed. You need to multiply force with speed to get power, so power goes with the cube of speed: that means +28% just from 350 to 380 km/h, +103% compared to the 'standard' 300 km/h!

2. Noise emissions. Unsurprisingly, the relationship with speed is almost the same as for traction power. But noise emission limits are not flexible, unlike train power and ticket prices, so this is a more pressing concern.

3. Ride comfort. The stronger forces mean more carbody motion, which is fine on a test run, but regular passengers might not like it.

4. Track wear. Stronger forces mean more strain for rails and trackbed, which means they have to be replaced quicker. High-speed tracks are rather expensive. (Just the other day, on the Cologne-Frankfurt line that sees 300-330 km/h traffic, it was found that rails have to be replaced much earlier than planned.)

5. Safety and train frequency. A lot of potential accident factors (side winds, track fatigue, etc.) are more severe at higher speeds, and there may be unknown new factors, ones you would rather discover in multi-year top speed test runs. One of these factors is signalling. It is critical because of the opposed needs of having a safe stopping distance between trains, and running trains as frequently as possible. No system proved itself yet at 350, not to mention 380 km/h, so you either risk accidents during signal trouble or have to run much less trains than possible at somewhat lower speeds.
   

Chinese decisionmakers might well ignore the first three and half of the last: prestige objects don't have to be economic, the regime won't be troubled with protests from noise-bothered locals, ride comfort can still be much much better than on old Chinese trains, and accidents happen. But no way around maintenance needs and the need for high train frequency, I believe. (In fact, I have conflicting reports about whether the CRH3 truly do even the announced 350 km/h, or, in case they did it during the Olympics, whether they kept it up after.)

Anyway, for a serious go for higher speeds, the way to go is to optimise technology, and that's a serious job. Some examples of what European and Japanese developers are up to.

A Chinese official flattered German reporters on a CRH3 by saying that

"The Japanese technology is more economical and energy-efficient, while the German type is bulkier and more than 10 times higher in price. But we wanted our passengers to sit in a sturdy, secure Mercedes rather than a cheap, light Honda."

Well, that sounds compelling, but it aint'. For one thing, Siemens's wonder is more susceptible to flying off the rails due to side winds. Aerodynamics researchers in Japan long discovered that a duckbill-like nose shape is better, even if ugly - most prominently on the Series 700. The newest high-speed train on Japanese rails is a thoroughly optimised version, the N700.

Duckbill nose of the first true high-speed tilting train, the 300 km/h Series N700 Shinkansen, run by both JR Central and JR West. Photo from The Japanese Railway Society

Changes include

• nose shape (with much more complex curvature);


• shrouding for all bogies (such shrouding has been frequently tried but frequently abandoned in the past 70 years, due to difficulties for maintenance workers and stuff getting stuck in it while aerodynamic benefit was little due to poor design);


• complete covering of the joints between the cars (diaphragms);


• aerodynamic design of the underside of cars (important in noise emissions);


• pantograph held by a single bar between noise-insulating spoilers [photo] (for an ICE-3 leading car, the pantograph causes one third of the air resistance!).

Together with weight savings, energy consumption was cut by 19%(!) compared to the Series 700. Yet, this wonder still only runs at 300 km/h. The third big Japanese company, JR East, wanted the successor to its decade-old 275 km/h E2 units (progenitor of that "cheap, light Honda") go one mayor step further.

Looks like out of a Sci-Fi movie, does it? The two Fastech 360 prototypes were meant as testbed for 360 km/h traffic with noise emissions, track wear and safety not worse than current trains at their much lower top speeds. The many innovations are on the same fronts but more extreme than for the rival N700. However, in testing, the trains came short of expectations: in particular on noise, and braking distances. JR West decided to order a new generation of trains for only 320 km/h.

Meanwhile in Europe, many of the same problems were attacked by the engineers at Alstom, when they designed the TGV successor, the AGV. I introduced it before (in High-speed to Barcelona), so I only note that it is in some respects a less radical design. For example, noise requirements aren't as strict as in more crowded Japan, and the AGV lacks diaphragm and full bogie shrouds; then again, it has car-connecting bogies, and a narrower body (which reduce the benefits of shrouding), and reduced cross section compared to the Chinese and Japanese rivals means less drag and noise by default [if speed and drag coefficient is the same].

However, this very advanced train is still 'only' meant for 360 km/h, and that after heavy testing: some parts were tested years ago, the train itself from this year, but type approval is expected in 2011.

Above: the AGV prototype during testing at the Velim test track in the Czech Republic. Photo from a helicopter from NTV's Gallery Below: model displaying the AGV in the livery of its first purchaser, Italian private operator NTV. Photo from Railway Gazette

How can the train be made to cut air even better? One possibility is to learn from nature: that shark skin is not totally smooth is a feature. Small dimples influence turbulences in a way that improves air [resp. water] flow.

You can already see this on golf balls. But researchers want to try it on airplanes - and trains. Tests in 2004 (see SPIEGEL article [pdf, in German!]) showed that a dimpled surface can reduce high-speed train air resistance by up to 16%.

Wind tunnel test of the 1:20 model of a dimpled ICE train. Photo from inventors network

If it takes such effort to advance this little, those Chinese developers have a monumental task ahead of them. Either that, or these ambitions will be buried, like the "China Star".

Don't get me wrong: I do think that China will become one of the cutting-edge high-speed rail developers. I also do believe that rail passengers will one day travel at 380, even 400 km/h. But I think that may take fourteen, not four years.

:: :: :: :: ::

To be updated complete list of past train diaries on ET: Train Blogging