First things first: a few definitions.

• Local service: used here to encompass all heavy-rail short-distance passenger transport, be it urban, suburban, exurban, interurban, regional mainline or countryside branchline; and be it all-stopper, limited-stop or zoned (also see my Local Rail series at dKos)
• Railcar: a self-powered single-carbody railway vehicle that carries passengers (or cargo)
• Multiple unit: a group of permanently coupled vehicles or a multi-carbody (articulated) vehicle that carries passengers (or cargo)
• EMU: self-powered multiple unit with electric traction
• DMU: self-powered multiple unit with diesel traction

In practice, the distinction between railcar and multiple unit has become ever less significant, and modern railcars tend to belong into modular families with 2(-3-4-5)-car multiple units. Thus, I will adopt the imprecision common even in standards(!) of using EMU/DMU and railcar interchangeably.

The emergence of regional multiple units

Industrial development is usually evolutionary, that is: even new classes of products tend to emerge as adaptations of existing designs. In the case of regional multiple units, there was not one but three, radically different origins:

1. Passenger coaches: fitting them with a (steam) motor was how the very first railcars were born, and this idea of cost-saving development persisted (later with gasoline, diesel, electric motors).

BR class 423 3498 left the spectacular portal of Shakespeare Tunnel in Dover, UK, on 11 May 1996. Photo by Martin Loader from Hondawanderer.com.
Class 423 was an outer suburban third-rail DC EMU with carbody adapted from standard Mark 1 coaches, built at British Rail's own works

2. Buses: the idea is to replace the lightweight mass-produced (and thus cheap) rival's rubber tyres with steel wheels. In many languages (but not English), "railbus" became common.

3. Long-distance multiple units: their more elaborate designs were sometimes adapted for regional sisters, or the express DMUs and EMUs themselves were refurbished in old age.

DB 612 502 (front) with 613 601 and its middle car (back) leaves Harriehausen, Germany, on an all-stopper service on 12 June 1982. Photo by Günter T from Drehscheibe Online Foren.
Young West German DB's first new post-war express DMU was the 1st-class series VT 08.5. With extra mid-car doors and 2nd-class seating, the semi-fast sister class VT 12.5 (later 612) was created. (The carbody design was also adapted for two suburban EMUs with even more doors.) The VT 08.5 were soon eclipsed by newer trains, so with replaced interior (VT 12.6, later 613), they joined (and mixed with, check mid-car doors on photo) their sisters in semi-fast, later all-stopper regional service

All of these pioneering ideas had their intrinsic problems:

1. standard coaches were not built to bear the weight and vibrations of a motor or to bear wind resistance and a cab;

2. buses don't have the structural strength to withstand longitudinal forces in trains hundreds of metres long, so railbuses either needed special regulations to not meet upon normal trains – or a strengthening that adds weight;

3. local trains need more doors (for a higher & quicker passenger flux) and start/stop much more often (with connected wear) than long-distance brethren.

The refurbished railbus MÁV-START Bzmot 297 with trailers near Drégelypalánk, Hungary on 10 May 2009 (own photo).
Bzmot belongs to a big family of railbuses built from 1973 by Vagonka Tatra in Czechoslovakia. Its mass of 20 t is 50% more than early railbuses, and 150% more than the bus type it shares main parts with

All of the pioneering ideas were also obviously meant to save costs by using standard parts. That was because in most local service, advantages and drawbacks relative to the locomotive-hauled alternative (see the beginning of Railjetting into Red Bull Country) balanced out and only cost counted. With this last benefit also constrained by the above mentioned issues, not much room was left for high-tech. Thus, for most of the past century, multiple units for local service were low-tech products with an arrested development.

An SNCF X 2200 with trailer passes Mouleydier, France, 15 July 2005. Photo by Patrick Meunier from RailFanEurope.net.
SNCF's series X 2200 is a single-car DMU produced by ANF Industries (today part of Canadian-owned but mostly Europe-based market leader Bombardier) in the eighties. Like its contemporaries, it is stronger, faster, made of better materials than predecessors, but still a thoroughly conventional design

Back to the multiple unit future I: urbanisation

Many multiple units have distributed traction, that is, more than one car has underfloor motors and driven wheelsets. Others are short but multiple multiple units can be coupled together. The resulting advantages of high acceleration, passenger utilisation of the entire length of the train, and low maximum axleload count most in high-frequency urban service.

Thus it came that the swing to a general preference of multiple units in recent decades originated (and developed to its fullest) in highly urbanised areas. The prime exhibit is Japan: locomotives were gone from local passenger service by the eighties already. A similar highly urbanized area in Europe is South East England. On lines south of London, an additional factor of early multiple-unit-isation was electrification with the third-rail system (as used on the London Underground), which is not conductive to high power uptakes on one spot as required by locomotives.

In 1982, the one-time state railway BR grouped its passenger services in the area into the Network SouthEast sector, with semi-autonomous management. The latter kicked off a consequent programme of purchasing multiple units only, banishing locomotives from local passenger service by the early nineties.

NSE's purchases, mostly from BREL (a British maker that went up successively in ABB, ADtranz and currently Bombardier) also pioneered a high degree of modularity. Both a Mark 3-derived family and the Networker family (a proper multiple unit platform specifically for NSE) grew to include DMU and EMU versions; in the latter, DC, AC and dual-system versions; suburban, semi-fast (and express) versions; and different numbers of cars.

BR 165027 and 165011 in NSE colours pass Overthorpe, UK, on 14 June 1994. Photo by Martin Loader from Hondawanderer.com.
Class 165/0 "Turbo" is a suburban DMU from the Networker platform

The multiple-unit-isation of South East England was noted across Europe, where parallel developments make urbanised areas ever more relevant:

• the networks serving metropolitan areas (with integrated timeplans, high frequency, high station platforms) are expanding ever further, absorbing local services on old mainlines and even branchlines, which thus have to be served by the same rolling stock;
• from the fifties, passenger service on most branch lines has sadly been killed off.

Back to the multiple unit future II: going low

Urbanisation made multiple units fashionable, but it wasn't in itself a reason for re-inventing the technology. Innovation happened elsewhere: this was the time of the light rail revival in Western Europe. But a lot of those innovations were just waiting for an application in heavy-rail multiple units:

• various types of articulation;
• miniaturisation and condensation of traction equipment into modules with standardised size;
• similar condensation of heating, ventilation and air conditioning into HVAC modules;
• removable seats fixed to consoles hanging from walls (enabling quick replacement, configuration changes, as well as easy floor cleaning);
• new floor materials;
• lighter carbody (construction using new design methods, better welding and also glues, use of aluminium and fibreglass);
• finally, above all: low-floor layouts (in Europe, platform heights range from 0 to 1,100 mm).

Note that all of the above listed innovations are conductive to modular construction: they allow for variability to meet customer wishes regarding train length, motorisation, seat arrangement, climatic conditions, and even platform height, with off-the-shelf parts.

New DMUs with the above innovations and conscious modularity emerged in the middle of the nineties, on a very small sub-market: regional semi-private (local government owned) railways operating branchlines in Germany and Switzerland.

Vogtlandbahn VT 42 nears Zwotental, Germany, 28 June 2010. Photo by Kevin Friedel from Bahnbilder.de.
This DMU is a RegioSprinter, Siemens subsidiary DUEWAG's 1995 novelty. The short middle car is a trailer with two wheelsets, the motorised end cars sit on it – a light-rail-inspired articulated concept that allows for very high acceleration but is a recipe for unstable dynamics

Due to the comfort difference relative to conventional DMUs, the first few new-technology prototypes were an instant success with passengers, so all major producers rushed onto this market (also hoping for bigger orders once rail privatisation rolls on). The boom phase was of course followed by consolidation, which saw most of the more exotic technical solutions thrown out.

Back to the multiple unit future III: going sexy

In 1996, Talbot, an Aachen, Germany based traditional factory then just acquired by Bombardier, presented its own prototype for a new regional DMU: the TALENT (Talbot leichter Nahverkehrstriebwagen = Talbot light local transport railcar[sic!]). It made an impact with its appearance: although its top speed was just 100 km/h, it got a streamlined nose fit for a high-speed train.

The TALENT prototype during its first test in regular traffic, on the bridge near Bareveld, Netherlands on 12 April 1997. Photo by Jan Beuker from RailFanEurope.net

Not long after, German Railways DB, which was preparing for its privatisation (via floating on the stock market – which didn't happen to this day...) and future competition (which did happen) with major new rolling stock orders, began to order new DMUs.

DB's first order was for 120 TALENTs in two versions. Eventually followed by orders from all of the rival product lines – with one extra wish: boxy fronts shall be replaced by new streamlined designs. The designs were to loosely follow a company 'standard look', with the band of the sloped roof edge continuing in a curve down the nose, making the windscreen and front an extension of the roof. Producers then maintained the 'DB look' in subsequent deliveries for other state and private railways.

A HL class 112 on the streets of Helsingør, Denmark. Photo from OEA.
This two-car DMU belongs into the LINT family (later renamed Coradia-LINT) produced by French rail giant Alstom's German subsidiary LHB. Unique visual mark: the concave top of the coupler opening on the nosetip, connecting the lower roof edge lines

The new-technology vehicles I spoke of so far were all DMUs. However, some families were designed from the start with enough modularity to allow exchanging diesel power packs for electric equipment.

The Kamnik-Savinja Alps tower up to 2,558 m behind an SŽ 312, near Kranj, Slovenia, on 20 May 2010. Photo by user LJ363 from VLAKI.INFO.
This regional EMU belongs into Siemens's Desiro family, the more conventional successor of the RegioSprinter. The unique visual mark on the 'DB look' front is the curve above the windscreen and the large destination display below it

ÖBB 4024 110 arrives in Bad Vigaun, Austria in the -15°C morning of 9 January 2010 (own photo).
This EMU version of Bombardier's TALENT platform is for rapid urban commuter networks called S-Bahn in German, in this case Salzburg's

The first of the new lightweight modular families to get an EMU version (1998) was not that of one of the then four (now three) Europe-based rail tech giants, but that of a newcomer: Swiss maker Stadler. Stadler entered this race as a small company new to train building, survived the initial phase of adventurous but problematic designs, then gobbled up some production capacity Bombardier had to shed when it ate ADtranz. Maturing its product line, on the local EMU/DMU market segment, Stadler established itself as an equal competitor to the three remaining giants.

MÁV-START 5341 049 reaches Lébény-Mosonszentmiklós on 24 June 2010 (own photo).
This EMU belongs into Stadler's second platform, the FLIRT family. 'DB look' (at least modified at bottom), standard articulation – the visual mark remaining is the high flat sidewall (compare the boxy look, compact middle motor car concept of the first Stadler EMU)

Teething problems?

Readers would be justified to eye the above shiny new high-tech modular wonders in corporate image with suspicion. All of the trains mentioned above suffered various problems, some teething, some persistent, at any rate passenger-angering.

• The rule still applies that multiple unit production has to be cheap. Thus come numerous quality problems, especially if parts of production are outsourced.

• The designs themselves weren't fine-tuned: this can result in anything from annoying carbody swings through software breakdowns to braking problems.

• Carbody lightness was pushed to the extreme, resulting in very complex structures designed close to strength limits, thus one sloppy welding can lead to critical weakening, not to mention susceptibility to felt vibrations (again annoying the passenger).

• It is great that a modular platform makes hundreds of variants possible without any need for further development – but all the variants still need to be tested to be approved, and types proliferate just when the EU and national authorities are getting serious about strict application of approval rules.

The Alstom Coradia Continentals just started regular service when I photographed the one shown above the fold, months behind schedule due to the tests. Bombardier's new generation Talent 2, ordered in four versions, was suffering from all of the problems I mentioned at the same time.

Gimme a platform!

You may have noticed that the story thus far was rather Germany-centred. This has structural reasons: not only did prototypes emerge there and are the big multinationals concentrated around there (even French Alstom with its subsidiary LHB), but Germany was a consequent pioneer of the recent transformation of local rail services across the EU.

In the current system (which may or may not change again into some form of open-access), (subsidized) services are franchises awarded to operators by the regions. In Germany, the regions have full autonomy to choose vehicles for operators to run, creating a much bigger open market than that of regional semi-private railways a decade earlier. With the off-the-shelf products they could market in Germany, the makers could later do business from the Netherlands to Romania, from Norway to Greece (not to mention the USA).

However, in most other major EU countries, the regions' rolling stock orders have to fit within national framework agreements, be it because the state railway retained its near-monopoly, or (as in rail privatisation pioneer Great Britain) because rolling stock orders are centralised by leasing companies or the Ministry of Transport itself. Now, the attitude of the monopolistic orderers was: modularity is all fine and well, but I want a family designed just for my needs! And I include local makers I'm familiar with! (Quite understandably so.)

Thus, Spain's RENFE has the Civia EMU platform, and just like in olden times, it exerted overall control of the design while including Alstom, Siemens and local partner CAF. Even small Portugal had Siemens and Bombardier team up for the CP-2000 "Viriatus" EMU platform (see photo in Local rail (1) @ dKos). Though Stadler's GTW and FLIRT made inroads in Italy, Trenitalia (state railway FS's passenger branch) and new semi-private operators bought most EMUs and DMUs from the Minuetto platform, which is a heavily adapted locally-produced version of Alstom's Coradia-LINT ("Coradia Meridian").

In Britain, the very narrow loading gauge (cross section) and ubiquitous high station platforms were natural blocks to standard imports. Thus Bombardier got to continue the local lineage of the Networker with updated technology: the Turbostar (DMU)/Electrostar (EMU) family. Despite the addition of "Coradia" to the product name, Alstom did the same with the Class 175 DMU and its EMU sisters of the Juniper platform. Siemens could enter the market with the Desiro UK EMUs and DMUs, which have little in common with the original beyond the name.

Even in Germany, until recently, DB insisted on an own platform and involvement of multiple makers in its EMU purchases. (The Siemens contribution again got the Desiro name with little justification in commonalities.)

In France, the main regional train platform of the noughties was the AGC (for Autorail Grande Capacité = high-capacity railcar[sic!]) – a heavily adapted and higher quality spinoff of Bombardier's TALENT platform.

SNCF Transilien's B 82547/548 (set 547N) in Paris Est on 18 May 2010 (own photo).
This is a DC-AC-diesel hybrid version (in French "BiBi" for bi-mode, bi-courant) of the AGC platform

This decade, the all-new wide-body platform Spacium 3.06 ensured that Bombardier got to supply new suburban EMUs. Meanwhile, Alstom found its way back to the home market in a rather roundabout way.

Back in 2000, Alstom had its German subsidiary LHB build a radical technology demonstrator, the LIREX. This DMU, while loosely based on an earlier suburban EMU for Denmark's DSB, included all kinds of funky ideas like single-wheel(!) running gear, roof-mounted traction equipment, flywheels, oval windows, extravagant interior. Most of those were ditched in the first series descendants, suburban and regional EMUs for Sweden. (I showed both the Danish ancestor and Swedish successor in Local Rail (2/5): Rapid transit @ dKos.)

The Coradia Nordic adventure led to the reduced loading gauge (narrower) Coradia Continental, which I showed above the fold. But, this again had to be beefed up when SNCF wanted a new EMU/hybrid platform for outer suburban and regional semi-fast services, resulting in the currently in-construction Coradia Polyvalent (which SNCF calls Régiolis).

In Poland, for almost two decades, local makers PESA and Newag struggled to survive on very small orders from a cash-strained PKP (mostly for prototypes only). But that was enough to develop low-floor modular platforms of their own and be on stage when the regions began ordering new trains: only Stadler's FLIRT could dent their market share a little. Meanwhile, PESA is already exporting DMUs to Italy.

PKP ED74-001 near Mszczonow, Poland, 1 July 2007. Photo by Monsieur Josviaque from RailFanEurope.net.
This PESA-built EMU for regional semi-fast services was converted later for express service with more spacious seating. However, it was the basis for PESA's ELF platform, now in construction

Double-deck trains

This is the one sub-sector of local rail passenger transport where locomotive-hauled trains can still compete: first, a loco can keep up with multiple unit acceleration with a shorter double-deck train; second, there isn't all that much free space to put power equipment into on a double-deck car, so it usually takes away some passenger space in EMUs.

The special placement of power equipment, and rather different carbodies, also mean that double-deck EMUs are hard to integrate into the same modular platform with single-deck ones. Even if they are part of families by brand name, most double-deck EMU types are still unique designs.

However, at least true modularity with the double-deck EMU constraint is found in France. There were three successive families (Alstom-ANF [Bombardier] co-productions Z 2N, TER 2N and the Alstom-made TER 2N NG) that spawned multiple versions in terms of voltage and number of cars, a fourth is now ordered (from Bombardier; Régio2N). The third, called Coradia Duplex by Alstom, also had export versions for Luxemburg and, with more significant modifications, for Sweden.

SNCF's three-year-old Z 24657/658 (set 379) and barely 100 days old sister Z 24781/782 (set 641) at Lille Flandres on 18 May 2010 (own photo).
Both sets are of the dual-voltage, 3-car, 2nd-class-only version of the Coradia Duplex

:: :: :: :: ::

Check the Train Blogging index page for a (hopefully) complete list of ET diaries and stories related to railways and trains.