It is interesting that some things return full circle to the way engineers thought in the late 1940s and into the 1950s when presented with a problem to solve> Back Then it was replacing steam with quite literally,. a turn key solution, while now it is replacing the second generation of 1970s&80s Diesel locomotive with locomotives which can comply to the outrageously strict emissions and noise legislation laid down by the EU commission and also lead by California who have pointed their judicial compliance bow to zero emission locos. This in itself is a farce becuase even diesel loco hauled trains are many hundred times more efficient over any long haul route than the equvalent 20-50 lorry loads or 200 - 300 personal car journeys. Rather than painting power units in locos with the same black brush as mass anarchy transport, they should be setting a sensible level of progress on these fronts, all be that very much quieter locos than we enthusiasts may like to hear pounding the metals..
Today Seimens present a modular power unit loco with possibility for dual power (over head or maybe third rail units top be popped in) and / or battery packs. One key benefit they quote for today's operations is that the locomotive can dial in and out power as it needs it, with the power units then working at their most efficeint peak range more of the time when they are on, thus reducing fuel consuimption and emissions,. In their Marketing PR launch discussion they talk with no reference to the long history of multiple power working doing just this, dialing in more power when needed, while then saving fuel by cutting back on parts of the route which do not need so much horse power. Rail actually requires a large tractive effort to start a train and to take a train up a steep or continous gradient, but since the days of Rocket it has been known that there is a lot of coasting and low tractive effort haulage going on due to the inherent efficiency of running on rails, particularily with roller bearings and optimally loaded axles. Thus you are actually dialing in power a lot more in a train than you are in a lorry which has a lot more relative wind and rolling resistance than a train, and you can times that by as many as 60 in europe for the biggest 2000 tonne trains
I can't remember if the PR release mentioned engine redundancy in case of power unit failure, but this has to some extent been quoted as the reason for multiple engined locos. This may have been at some point someones 'also ran' selling point, be that point made by a supplierr on internally to British Rail western regiona and the Deutsche Banen in particular. In fact the notion is largely a red herring : in the case of the German multi engined V200s/220s and the British Deltics, the attraction was more horsepower per se from a single locomotive within some technical limitations. Interestingly those limitations were quite similar and both technical. In the 1950s there was a desire for light locomotives with 2000kw output approximately to sustain speeds of over 90mph/150 kmh and run ideally cruising at 100mph//161kmh. The limitation were on both diesel electric and the hydrualic mechanical transmissions at the time.
Both systems were limited to about 2000hp per power unit by technical limitations. Firstly in the electrical generators, where English Electric had a size limitation and a knowledge of flash over (which would later plague the class 50 and competitor's class 47 locomotives ) in the dirty railway environment for the then DC generators. The biggest EE could offer was about 1.75 Mw , brush and AEI offered slightly higher while the American locomotive manufacturers could offer their home market around 2.2 by the early 60s with the larger loading gauge. The DB in Germany had opted for Diesel Hydraulic for their higher speed services and much of their other locomotive provision in the 1950s. The limitation to them was that at that time for the footprint required at least, neither Voith nor Mekhydro could offer power transmissions of over about the same 1.7 mw, or actualy at rail about 1500 hp.
One efficiency advantage of diesel electric in this respect is that a single engine can be employed to deliver the maximum applicable kw to the traction motors at start and low speed in a light weight express locomotive. In the Deltic this means that the power control handle does not activate the second engine until approximately 18 - 22mph is achieved. With the secondengine revving up to deliver power to the same level as the first and then both going onto rev further to deliver 50mph
, the system as a whole is very efficient for passenger workings because they require exactly this type of acceleration avoiding a field diversion(electrical gear change). A typical 350 tonne passenger express of the late 50s'1960s requires light starting effort and offers actually not very much momentum to push through field diversion . This explains why the class 50 out accelerates a deltic to over 30 mph, while a class 37 will out accelerate a deltic to 18pmh hands down because it can lay down far higher amps as its single engine is delivering maximum mechanical effort in the first field which dies at around that speed. Both these classes however could be plagued at speeds of under 50mph by trains not having enough momentum to push through the diversion, then meeting a gradient, or by a speed limit or signalling which meant the engine hunted between two gears as a manner of speaking about the automatic detection and switching equipment. The twin engined deltic and presumably Brush "Falcon" overcame this issue by dialling in the power and avoidimg low speed diverts, while the former offered quite low starting effort and maximum continuous effort due to having a lower amps/ higher voltage system (amps are a measure of torque where as higher voltage relates to spinning if you like, groossly over simplified)
Diesel hydraulics however have a different advantage in having their torque convertor 'half gears' ie slippage in the fluid coupling followed by the married phase, as well as having a range of 'hard ' gears to change up to. Correctly engineered this makes for a very smooth acceleration with virtually no wheel slip and quite a high starting effort. Both engines can power their respective bogies from a standing start without the typical electric overload of 1950s first generation DE locos, they are designed to be geared correclty for this tractive effort and the limiting factor is how long you can cool the hyrdaulic transmission oil when it runs in fluid unconnected phase (slipping stator relative to rotor). For all types of trains in fact, a twin engined diesel hyrdaulic has another slight advantage over DE designs of the late 50s at least, and that is that one complete transmission bogie system can be left on maximum power while the other system reves down and changes gear for its next bit of the cherry so to speak, thus the train can be kept at a constant speed if not on a steep gradient. Field weakening on DE is a just a natural physical barrier for the entire system if it is to run smoothly, although a complex out of phase double system coudl be concievable, where the two bogie/traction motor sets are tuned to different field weakenings. Instead as in the deltic, the number of gear changes to 100mph is just three versus effectively six for diesel hydraulics even when they are more powerful as in the german V classes.
As touched on in the pre amble above, the redundancy of twin engined locos was a secondary selling point in effect then. However it did mean that the most important express services on the GWR and the ECML could limp home at about three quarters speed in the mid 1960s when the whole genre of diesel was still to be proven as a mass produced item. Further into the 1970s, the HST IC 125 sets had two power cars in order to deliver the magic extra thousand over deltic services to go 125 mph., but the betting on effectively two power units per train proved prudent in terms of limping home as both the Ruston Paxman valenta and the Mirrlees power units prove to be less reliable than anticiapted for their service intervals in the reality of the dusty conditions and high thermal cycling loads placed upon them by the nature of having more stop starts than originally discussed with the manufacturers. The 125 was actually designed with train crew changes en route and fewer stations than the deltic and class 52 services they replaced.
In terms of fuel efficiency then I have not heard of westerns warships havbing one PU shut down, but it happened regularily on Deltics which would have their second engine shut down on the slower Ediinburgh-aberdeen section of the expresses from KX. Presumably after ETH was introduced this required that the second PU be on a heat only selection, revvving not far boave idle I do not know. Deltics did limp home and run light loco on a single PU, and especially with the teething problems with pistons they encouuntered, this became a very fortuitous feature.
In post war Europe and the UK there was still materials rationing and oil was imported from the US and Middle East, so the use of multiple workings, dual or more locomotives is for those historical reasons far less wide spread than it is in the USA/Canada, coupled to a steady decline in freight from the 1960s to the 2000s as the road network improved. The states on the other hand and canada had a heavier loading gauge and bigger, more economic commodity and produce loads to haul by rail and thus it was economic for them to run multiples and 'robot' locomotive arrangements were widespread by the early 1960s with multiples of up to 4 at the head and more down the train. I do not know of any european robot multiple locomotives apart from shunters like the class 13 and some on the continent, robots being a unit which lacks a proper cab and are only used in multiple with cabbed locomotives leading. By 1960 the BTC/BRB had abandonned through connection doors for locomotive design, although many were in still in production then, and were stipulating that type 4 motive power be above 2700bph and not fitted with multiple working ( a decision reveresed in the class 50 due to the gradients of the WCML and the ambitious timetables laid out in the run up to electrification in order to keep the route competitive in respect of the new M6 /M74 and the advent of the Glasgow Airport-London shuttle. Class 50s were not fitted with these as a response to theirr reliability issues, which were in fact somewhat better than their rival mass priduced brush type 4, which required a massive rectification at a cost of over three million pounds back then. That class 50s were delivered without multiple working cables t first has been cited as being because of availability of the materials at the time, safety testing not being complete, the price of the class 50 running into problems and then this being taken as part of the hire-purchase scheme they were bought on, subject to absolute rquirement for the faster timetables north of Crewe and later Preston)
So at the end of the day you have to examine what your actual missive for tractive effort and maximum speed is, and in fact this is where DE wins over DH: For the same horse power and power unit arrangement, " Faclon" was a better performer than the class 52 Western hydraulics with their voith systems. The limitation here was that the v12 MD655 engines were not powerful enough for the third gear on the Voith transmissions which were presumably very similar to those fitted in the more powerrful German V2xx locos. It may have been possible to have fitted tiwn v16 MD870s or alternatively a Mekhydro gear case could have provided a better matched power deliver for running at over 75mph. The Voith transmission is in theory smoother than the Mekhydro due to it working on a priciple of triple, sequential torque convertors and hydraulic actuation of these three main gears. Although Westerns and of course Hymeks saw a lot of working on relatively heavy freight services, the similarily powered Falcon and the Class 37 EE type 3 both out performed their respective DE cousins in standing lift and progress, all be that with some degree of wheel slip on the EE type 3. Deltics did work a very few freights in their lifetime and heavier night mail/sleepers of over 650 tonnes, famously last year D9000 working some heavy lifts to the rail head due to lack of class 66s. Falcon, the class 53 one off from Brush, excelled at both freight and passenger working, and indeed ended its productive life on ore and coal traffic, slugging them around south Wales. Class 50s were designed on out set to be a stop gap passenger loco for the WCML but also then to have the possibility to work higher speed freights of upto 1000 tonnes, being in fact prepared with mounting points for buck eye coupling as if it was going to happen in the 1970s.
We have then come full circle in which transmission system suits which traction missive: For light passenger trains of 40 mph to 80 mph the vast majority of diesel services in the UK are now sloshing oil over convertors with mechanical hydraulic tranmission. This was a decision taken a very long time ago by some engineers, probably ironically enough, about two or three years after Western Region lost their last main line DH locos. For heavy freight you want to slug it out with 25 tonnes per axle at least and over 3000hp in DE, while also the hands down winner for mixed traffic is the DE too because of the controllability and range of electrical traction systems. This is reflected in the delivery of the latest class of locomotive to the UK, the mixed traffic class 68 which is pretty much the type of locomotive the GWR would have ordered if they had been forced to run with a larger proportion of DE locos. The main point lacking here is that we do not see many multiple power unit locomotives with their specific advantages and redundancy, but that in main part is due to the amount of power available in reasonably sized single power units in mid or higher speed * GM versus Caterpillar, MTU and the most powerful per weight, the VP185 from Rustons /MAN, and also that there has been no specific missive for this type of locomotive. Even in DH we have seen several Vossloh locomotive classes and the Anglo Belgian powered freight class emerge with only single power units. It seems the reliability and modern engine management have by in large rendered twin engines obsolete, where as the Seimens loco can carry up to four power units.