fredag 11. mars 2011

The Secrets of the Class 37s Success

For the "first generation" of major dieselision, there is no doubt that the EE Type 3, later class 37, was the most successful on many counts, and this is best reflected in their longevity and the usefulness the class offered into the 2000s.

The only other comparable locomotives are ....also English electric, being the venerable 08 and the class 20, although the latter is most often used in pairs, a somewhat wasteful formation both at the time to erradicated steam and later in their lives. However, they too prove to have their niches.

The main failings of the first generation type 4s were the 1CoCo1 wheel arrangement and the overall wieght. Coupled to the class 40 and 44 being underpowered for their intended duties, and the lack of reliability from the class 45, then these locos were a heavy, expensive experiment. The class 46 was reputadely a more reliable unit, as were of course the class 40s, some of which were in completely working order when cut up in 1985-6.

Of the second generation, there were only three classes eventually ordered: 35 being the first, then 47 and 50. Once again it was actually the type 3 which prove to be the most reliable and "designed for task, delivered to design" . Both the 50s and the 47s required extremely expensive refurbishment to overcome their reliability issues, and in effect the 47s in particular did not live up to their design missive.

Brush Traction then had a pragmatic approach to what was still fairly experimental machinery. They had already shown this in re-engining the problematic and underpowered type 2 AIAs, and took a sensible decision to de-rate the new type 4s after rectifiying the terrible problems with the Sulzer units. One of the key issues was most likely caused by sulzer sub contracting production of the LDA series engines out, and thus quality control at various levels was not fulfilled.

In life, however, the sulzer 12LDA twin bank engine prove easy to both rectify and "keep on top of" with regular, specific maintainance. It is a single turbo engine with otherwise good access to components and inspection points. This contrasted with its own cousin, the 12 and 16 LVAs which had "blind" spots in the crank cases which lead to catastrophic failure due to a simple locating pin tool being ill machined! The 12LDA units could be hammered back into service rapidly, and given the amount of idle service duffs seemed to get in some areas when the more powerful 56s and 8s came in or when pairs of class 37s or 31s were needed to start heavy loads the 47 couldn't cope with in its solo existence, then class availability was kept artificially high. A better measure is MPC: miles per casualty, and for the same high speed services as the 50s, the 47s faired no better in fact!

The 47's legacy really lived on because Brush won the lions share of the two next generations of locomotive orders in the class 56 and 60. The nickname "duff" has cast a long shadow in UK locomotive history.

In the mean time however, GEC had gone into Alsthom and had concentrated on mainly AC locos for mainland europe with only relatively small orders for the "prima" units, tha majority of those for France and the UK, with the rebuild of the 333s being just that. It was the USA who had kept advancing diesel loco technology in heavier RA7 + freight locomotives, and actually could offer units down to under 120 tonnes based around the GM 645 line of PUs and CoCo format.

The failure of the 56s to haul the heaviest trains in the UK, lead to the class 59 outshining all but the "tugs" when they came on the scene. Major teething problems once again with the 60s meant that in the soon-to-be private world the EMD class 66 would rule the rails. ( although as a footnote, the USA proiduced Class 70 has had its share of teething problems)

So in a way, the perpetual under-delivery by Brush and compacency by GEC Alsthom opened the gates for the very competent N.American GM-EMD to swing in. WIth the Class 60s likely to be withdrawn or rationalised, we will one day soon see the end of UK produced locomotives on British metals.

lørdag 5. mars 2011

The British Rail Class 38.......that never was

The class 38 was perhaps quite near to being a signed deal, but the new reduced cylinder type 4 or 5 freight loco was what actually transpired from the successful 37/9 trial.

But what would the class 38 have actually looked like? How would it perform? Would it be like the 60 - freight dedicated -or like it's type 3 cousins, a jack of all trades?

Historical Context

All of the type threes and the nearly type three class 31, had shown that their superior punch in a small package made them winners over the under powered type 2s. At this approximate power banding, 1.4kw, it was deemed potentially useful to continue with newer locomotives which would work medium freight services, track maintainance and perhaps passenger trains. Another main purpose of the 37/9 trial was to estabish if reduced number of cylinders could be effective in operation: ie the far higher power per cylinder offerings in fewer and in line cylinder arrangements would be as reliable as the somewhat over engineered V12 and V16s of the 1950s,60s and 70s which were to be replaced.

The obvious choice of test bed for the proposed class 38 was of course the class 37. The engine compartment and radiator surface-area could accomodate a wide range of power units (PU's) : all from the two straight sixes chosen, through the GM645 unit and even up to the v16s in valenta/ Vp 185 or perhaps an RK 270 v16 - remember the portugeuse C1800s where shorter and lighter than the cl. 50.

At the time though the principle cost concern was what diesel mechanics call "unit man hours", a unit being a single cylinder and thus a multiple of these being considered the key cost driver. Heavy maintainance on the RK16 was costly, and the newer RK 12 has often been quoted as "over powered" and relatively unreilable by drivers and enthusiasts.

The eventual move to fewer cylinder "units" prove not to be quite as smooth as the 37/9 trial, with the expensive mistakes made on the class 60 with the MB275T/ 8 PU. However, no matter the teething issues with the PU in the "tugs", man hours and replacement components are indeed the major cost in maintaining a PU. Thus both straight 6 PUs chosen fulfill this criteria and establish if higher pressure PUs could operate at least as reliably as the best examples of the class 37s at Cardiff Canton where the trial was initiated.

What Form for the Class 38 ?

The class 38 would of course have been a standard "slab end" loco and would have followed the conventions at the time in being either similar or even standard in some elemetns to the two other UK locos viz. 56/58 or the US built class 59 perhaps. Given the contract had been placed with Brush Traction Ltd, as per the class 60, then the appearance would have been no doubt standard to the front end of the 60s. In essence the class 60 body is just a reworking of the 47 and 56 stressed frame.

One variation could have been employing a "Bone" format akin to class 20 and 58: at the time of the 60 order, the bone format had been clearly shunned on BR with a preference for walk through bodies allowing all weather access for inspection. With the whole design strategy being based on in line PUs, then a good deal of inspection and repair could be achieved in situ within a full width body. Furthermore, this would allow for more space for inclusion of a (redesigned!) silencer unit.

To Co or not To CoCo....

The MB275 unit is despite being only 5 mm wider in cylinder bore, considerably heavier than the RK270-6Tt unit, and in fact heavier than the original 18 tonne 12CSVT fitted in the Class 37. Being two stroke, the engine has longer stroke than the RK 270 (mm) and in general is far sturdier built to accomodate higher compression. In addition the crank case is a larger volume with large inspection ports for each "big end/con rod" connection to the crank shaft.

Therefore if the MB275 had been chosen then the loco would almost certainly have been a CoCo.

Power Unit Deciding Ultimate Bogie Arrangement

BRB had been somewhat dissatisfied with the RK reliability and service costs and also the delivered performance in classes 56 and 58. The PUs required more servicing than desirable, and it was seen that less cylinder units could offer cheaper servicing. Dissatisfaction was also all too evident at some customers like Foster Yeoman who ordered the GM 645 based class 59s as a result of poor class 56 performance, the road to GM EMD being paved and the UK industry's fate being more or less sealed as 'also ran'. I actually saw a FS stone test train with double class 56s in 1985 at Bristol Temple Meads Station, a service which a single 59 could tackle. So the Ruston RK 6 (or eventually 8) engine was not perhaps the lead runner.

In the final reckoning of course,m the superior power rating of the the MB 275 stratight 8 over the RK 8 lead to the former being chosen for the type 5 project, the class 60. The RK 8 in traction set up would be at a maximum rating of about 2800hp in the late 1980s, probably being favoured at a lower 2500hp to extend service interval. Coupled to SEPEX or GEC-Alsthoms latest power delivery & wheel slip control at the time, then a 2500 rating would have outperformed all the 12LDA locos and probably the standard class 56s and 58s.

On the other hand, in defence of the Rustons offering, there were several classes existing at the time which had a longevity expected with their RK / C/ SVT power units: 08, 20, 31, 37, 50, 56 and 58. Many components were "backwards compatible" eg new pistons from 58s into 37s, engine blocks from withdrawn 56s in class 50 life extension and so on. The MB 275 was completely unproven in a traction environment at that time, and Mirrlees had not enjoyed a very enviable reputation for their other traction power units, being swapped out of at least two types of brush locomtives and not favoured in the HST/ IC125 contract despite being trialled.

The RK 6 is four stroke and therefore uses less lubrication oil and probably has a longer heavy-refurbishment programme than the 6 cyl, MB275. Class 37s had proven to be masters of the type 3 power range, over-shadowing performance of their heavy and sluggish big sisters the 40s and providing more reliable service than all other type 4s. So this also stood in favour of Ruston's offering.

Details of the 37/9 trial are not fully released to the public, but however several commentators at the time stated better delivery with the MB but better service interval on the RK. The RK unit was fitted with some complex turbo lag synchronisation control, delivering fuel slowly until the turbo pressure was achieved. Hence the name "slugs".

With the smaller and much lighter RK6 being chosen form of the locomotive, the class 38 layout could then be left a little bit more open: Co Co configuration with the compact RK270 6T would have necessitated ballasting for heavy freight allocation, or indeed in any intended guise over six axels, as it weighs several tonnes less than the 12 CSVT, being essentially a single bank of that v12 engine and more like a class 08 RK footprint! This ballasting could have been a usefulk feature, with variable axel loading facilitated by lifting out the ballast depending on route and train type allocations.

An alternative just for the sake of arguement, could have been an AIA configurations: this could free up the trailing axel for advanced hydraulic braking or a dedicated rheostatic brake "motor". There are some electrical advantages in having only four traction motors to feed at this ampage/ voltage range in traction power application: these are much speculated upon by class 31 enthusiasts in comparison to the more powerful 37s. Also given the advanced SEPEX traction motors being a likely part of the order, then this electro-mechanical advantage could have rendered CoCo uneconomic: ie utilising a maximum kilo newton starting and continuous effort for the total power out put and axel loading over two SEPEX thus negating using standard TMs or bothering to employ six such advanced TMs.

The obvious option for reduced cost though would have been to produce a BoBo locomotive: Less materials in a shorter locomotive. The RK 6 engine is only slightly larger and heavier than the 280mm bore 6LDA 28 engines engines fitted to so many type 2s on BR and further a field. Although I find no interwebie quote for the type 3 rated 8LDA unit from the class 33, It is probably slightly heavier than the RK 6.

The success of the class 33 could have pointed to a SEPEX based locomotive being a sensible BoBo "Little sister" to the class 60 project. In fact the entire class 38 project had already been achieved by Brush traction when they re-engined their own "mini grids" : a class of BoBo locomotives made for Rhodesia / Zimbabwe. These were fitted latterly with RK 8 cylinder engines with a rating IIRC of around 1800 to 2000 hp and a tonnage under 80 gross. These suprising cousins to the 56s did have some good web listings and images a few years ago, an excellent british export locomotive site being amongst them. However, due no doubt to both copyright issues on images used and sources, and the closed nature of Zimbabwe as far as the internet goes, details of these locos are beyond my patience to find. I would be grateful for any links to provide some providence in this case. I would be very grateful for any links placed in "comments" below, by means of providing providence for this class of loco!

Total wieght at under 78 tonnes would later be advantagous for track access charging. However, there are several stumbling blocks for this BoBo arrangement: firstly the type of duties expected of a type 3 would include the torturous routes negotiated by 37s:

One reason 37s were introduced to the WHL and FNL/Kyle routes was to both accelerate trains while reducing corner wear over the type 2 bobos. By 1986 though, articulated CP5 bogies had been deemed a success though so a BoBo could have been practical in a new build. So this is one factor to consider: route availability and effective wear and tear.

The RK 270 is a longer unit than the 6lda, thus taking up much of the space in a locomotive as in the class 33. This would mean that features like rheostatic heat dissapators or indeed an auxilliary ETH generator would be excluded. One final factor I come upon to negate the use of Bobo is that a modern CoCo is in effect an AAA configurations: each traction motor can be independently removed or supplied less power ( in fact any 37 TM could be bumped out of connection, so they were in fact not a series connected CoCo) Heat dissapation is also a major consideration in a 1.4kw plus PU system and 37s had proven that their high ampage and resulting power delivery at low speeds was ideal for the CoCo configuration. So given the demands of heavy frieght and gradient work as a replacment to the 37s and 40s etc, and supplement to type 4s and 5s, the 38s would have done well to benefit from wheel slip / traction control in CoCo set up.

Final comment on CoCo: if the loco was to be a true mixed traffic entitity then the fitting of a 400- 600 hp auxiallary would have rendered a 2000hp prime mover with SEPEX TMS in a modern loco more effective than any of the then current sulzer type 4s when hauling ETH stock. A CoCo would allow for both this and space to carry ballast, while most likely offering better wheel slip and power delivery management than a bobo at the time of potential ordering in the late 80s.

Conclusion on Format:

At the end of the day, given the class 38 order went ahead in the same time
frame as the actual class 60 order, , then the Mb engine would be the choice to offer standardisation between the type 5 and type 3 erm, 3.5. Hence the loco would really need to be a CoCo to carry the weight and no doubt tackle the thumping torque -amps delivery of the big two stroke PU. It may have even been desirable to use the exact same body shell to give even more standardisation.

Had the order been put out to tender a little later in hypothetical time, then one tendering partner would have been the GEC altsthom with their Prima format, which has by in large been manufactured in BoBo format in both electric locos and the GM 645 based builds including the class 67s. With the light RK in their portfolio, then the loco would almost certainly have been a bobo and perhaps run alongside the royal mail /BRB build of the rather horrid class 67s.

Had BR decided to be rid of the underpowered type 2 fleet at an earlier date when their obselecense was clear, say 1976, then they could have looked to rationalise over to 8lda and ETH for a great number of class 25s and then build a new BoBo with brush at the time of the 56 order, as per the Rhodesian mini-grids discussed above.

Workings for the Proposed Class 38

What would they have worked? Well of course they would have worked all those mid wieght, mid speed services and track maintainance duties which class 37s have only now been by-in-large withdrawn from. Thus the class 38 would have lead to an earlier demise of the majority of the class 37 and probably all other type 2 and 3s, if they lived up to their promise of longer service hours, lower MPC and faster HGR turn around.

The 38s would have offered the opportunity to replace the earlier sulzer 45/46s from freight duties and supplement 47 freights, perhaps offering a heavier over all standing lift. In pairs they would have outperformed all the single type 4s and 5s of the early 1980s ( a pair of 37s would out perform 56s and 58 in sinlge unit operations according to motherwell depot). In the rush to dieselise in a post war britain still having materials rationing, pairing locomotives was seen as wasteful. However it is common practice in most long haul countries, where by larger "robot" multiples are often only engaged for gradients or actually switched off when running the wagons on "light" return journeys. Given that a pair of 38s would put out "type 6" power with 12 cylinders and be able to run light journeys on one unit, there may well have been an ongoing appeal from managers at depots like motherwell, Eastfield, Immingham, gateshead and Cardiff Canton.

Passenger wise, the 38s would have seen some employment even if they were nominally a freight ony locomotive given the same developments in the private era as transpired. However any potential order would probably have kept an eye on WHL sleeper services, routes like bangor, scarborough, Hull, Aberystwyth, Welsh Border Route and the potential for electric units / forwarding services to be operated by a light and fuel efficient loco.

If the class had been ordered as a freight only, this would have been under the BR sector scheme and so there would have been some debate upon privitisation and the ubiquity of class 66s as to the class 38's utilitity.

Now we see that the class 15x series multiple units are seeing the end of their effective lives and indeed the later generation units are showing signs of deminishing reliability, the 38s could have picked up where the 37s left off, working long distance semi expresses and hauling both de-engined 156s etc and things like 318s on services to say Lincoln, Hull, Bangor, Abersysthwyhth and in Scotland.

Fleet Size and Order-Number ?

Given BRB were serious in replacing earlier loco types 2 and -3, then the order could have been in the region of 150 to 200 examples. The author considers that both a super low geared and a higher geared "stroke 4" passenger dedicated variant sub class could have been subsequent orders upon successful experiences in core frieghts.

However the type of sub 1000 tonne traffic was drying up and becoming uneconomic. Also passenger traffic was strategically set over to multiple units: there was hope on a larger degree of standardisation, economic viability and reliability, all of which have proven dubious and probably not of any benefit if you compared the second or third generation DMUs with running class 31,33 and 37 (with 35 in there for historical noeworthyness!)

In the privatisation period, then private owners would have been very happy to exclude everything from class 20 to class 37 and rather own a single type 3 for lighter workings where a class 66 burns a lot of diesel! However, they were otherwise delivered fleets of pretty reliable "legacy" locos at bargain basement prices: some speculate that the acquisition of locos was less than scrap value in the total equation. There was no impetus for a new type 3 post privatisation due to this and now with network rail being national they struggle on with legacy 37s etc while 66s and 60s are often seen on rather puny track maintainance or even weed killer trains, and class 67s have appeared on many services which were otherwise completely adequately delivered by class 37/4s. In 2008, a large number of class 60s were stored as were many 67s and even 92s. It was claimed that there was not the demand for these locomotives.

As of time or writing, with no credible incentive to reduce axel weights or introduce locomotives on semi-fast services poorly served by DMUs, then there is little sign of any possible replacement to the recently pensioned type 3s.


The worrying issue for us at the time of the 37/9 trial, was if class 37s would routinely become "slugs". However, the refurbishment and "reallocation" programme showed the 12 CSVT to be a venerable and over engineered item which would last at least another ten years as planned, and of course nearly three decades in profit making service in reality.

If the RK 6 was chosen, then the loco would have almost certainly taken a BoBo format, while the MB275 lump would necessitate six axels. Intriguingly, the RK 8 could also be fitted into an RA6 / 7 Bo Bo loco as brush had done some years earlier to their Zimbabwe-Rhodesian "mini grid" bobos. My speculation ends there.

From all accounts however, the straight sixes were always planned to go in a new loco or be a test bed for reduced "unit" cylinder practicality for the larger "8s", and the latter was the project goal which prove desirable and came to fruition.