Ex LNER U1 Garrett on the Lickey Incline

[brack]

14 hours ago, jools1959 said:

I remember the conversation I had 30+ years ago in the US between myself, a couple of UP and SP retired engineers and someone from South Africa who worked on the railway’s.

 

From what I remember, a Garrett was far better suited to more flatter ground or shallow grade, whereas a Articulated loco handled better on grades

With all due respect (and acknowledging that the gentlemen in question likely knew far more than I regarding their charges), I'm not sure any up or sp retired engineers could have any relevant comparative experience to base their opinions on, given they would have no experience of garratt operations. The south african experience might be slight more informed, but their last 2 mallets went in 1960 (MJ1, Canadian built lighter locos for branch line service, the mainline mallets went 20 years earlier), so it's likely there was little comparative experience to inform his opinions either.

Fortunately there was a proper trial.

The South African Railways had quite a number of hefty mallets which they used over steeply graded mainlines (they had previously tried kitson meyers), the biggest of which, the MH was the most powerful 3'6" loco in the world when built in 1914. Post WW1 they ordered a garratt (class GA) with the same effective wheel arrangement (2662), boiler capacity, grate area, axle load and tractive effort as the mallet, specifically to compare the 2 types (the garratt was however 46 tons lighter - no tender). Whilst the GA garratt was found to need stronger frames and carrying wheels on the inner ends of the power bogies (it was somewhat experimental, and 50% bigger than any previous garratts), it handled heavier trains than the mallets at faster speeds and used less coal and water when tested on the Mallet's duties on the Natal main line.

While the MH was not a bad design at all (they were highly regarded by the SAR) and lasted another 20 years before withdrawal, SAR never ordered another mallet or simple articulated, but instead ordered large numbers of garratts until the end of steam. They were fooled into buying a few of the inferior fake lookalikes buy persuasive prices and salesmen (modified fairies and union garratts- fairly transparent attempts to get around patents), but the conclusion to the experiment of garratt vs mallet in mainline service on steeply graded lines was emphatic. Yes the late american articulated locos were more advanced than a ww1 era mallet, but the garratts BP turned out at the end were vastly improved too, using similar technology, the cast steel engine beds even coming from the same foundry in the US.

 

The mallet design forces you to build a long thin boiler, which in thermal terms is the worst possible shape - indeed some of the larger us locos essentially used the front half as a pre heater or long smokebox as there wasnt much heat transfer going on that far from the fire. A garratt pushes you towards a shorter, fatter boiler, which is the optimum shape, and has clear space around and underneath the firebox so it can be as large as required within the loading gauge. A double bogied loco will ride better than a semi-articulated one (eg. mallet) and on curves the garratt will have greater stability as the boiler goes inside the curve, whereas on the mallet it overhangs the outside of the curve at the front, which gets worse the bigger it is (hence the santa fe 2-10-10-2s with the ball and socket and concertina joints in the boiler barrel, which were predictably unsuccessful).

The only advantage the mallet has is one fewer set of flexible steam joints.

The point is sometimes made of adhesive weight reducing as fuel/water is used up on garratts, but this is only to the extent it occurs on a tank loco, and (as on the GMAM, though this was more to reduce axle loading) you can get around this by reducing water supplies on the loco itself and having a tank wagon as auxiliary water tender.

Edited June 17, 2020 by brack

[jointline]

2 hours ago, lanchester said:

That's not quite as true as usually thought. Indian railways were buying German and American in the 1890s because British manufacturers couldn't cope (and of course the Midland, GNR and others also bought American at the same time for the same reasons); Borsig sold 32 goods engines to East Indian Railways in 1902/3 (assembled in India); Herschel supplied ZF class narrow gauge 2-6-2T in 1935 and I am sure there were many other examples. 

 

(The Indian railways also got many locomotives from the US and Canada in the Forties but of course there was a war on and the UK wasn't in a position to supply).

 

After all, if India (and other dominions) had been automatically buying British anyway,  'Imperial Preference' wouldn't have become a political issue.

 

Can't disagree.  The book I was reading was by someone who was in India over the independence period, and was involved in the cement and chemical industries.  It was certainly his view that the old boy's network of  Brits in the Indian firms ordered from  the UK by default.  When the main players in these firms were replaced  by Indians there was a sea change in procurement.   Could be  of course, that in a newly independent country there was a  feeling that there shouldn't be an automatic continuation of ordering from the UK,  and perhaps the sales reps of other countries quite rightly saw opportunities in a market where they had traditionally experienced difficulties in getting a foothold. 

However I think overall there was a preponderance of UK built stuff in our colonies.  Remember visiting  North Borneo, for example, and being impressed by the completely UK built steam loco fleet, and Ransomes and Rapier  equipment. 

[brack]

48 minutes ago, jointline said:

However I think overall there was a preponderance of UK built stuff in our colonies

It's fair to say that many of the companies were financed from the uk, often headquartered here too and had majority uk shareholders, at least originally. Hence they tended to be equipped by british firms. Quite a few saw themselves as british railway companies, just the track happened to be located elsewhere (and not just in our actual colonies, for example quite a few argentinian lines fit the description too, or several spanish lines such as Alcoy- Gandia).

[jointline]

2 hours ago, brack said:

but the garratts BP turned out at the end were vastly improved too, using similar technology, the cast steel engine beds even coming from the same foundry in the US.

 

Yes, I have long thought that the Beyer Garratt was way ahead of other locomotive designs in terms  of  maximising the loading gauge for fireboxes etc, and  ease of bi-directional running.  

[Michael Edge]

9 hours ago, Frond said:

 

I have a pretty good idea what the respirators would have been having worked with breathing apparatus for over 30 years. If you can imagine a gas mask only instead of a filter there is a long rubber tube, this tube is led to somewhere where the air quality is better, in this case as near to track level as possible. The wearer basically sucks (relatively) clean air through the tube with exhaled air going our through a one way exhaust valve on the mask.

 

Again using the LNER Encyclopedia as reference apparently the crews objected to sharing the equipment and their use was discontinued. Given what the conditions in the tunnel must have been like I am not sure what was the lesser of two evils.

According to Don Townsley recently they were still there in 1948 when he and his friend had a ride on 69999 from Wentworth junction up to West Silkstone, he said they used them.

[Ohmisterporter]

IIRC Beyer-Peacock sold the US licencing rights for the Garratt to the Lima Locomotive Company. As Lima did not want to build Garratts themselves and held the licence nobody else could build them even if they wanted to. At one stage the chief engineer at Durango* toyed with the idea of building a Garratt using the chassis of two K36s and made a sketch; probably over a beery lunch.

As for BP build quality, one of the last classes they delivered was the 20 Class to Rhodesia Railways. There were so many faults with them that RR pursued BP for compensation even after the company had gone into liquidation. A sad end to a great company.

Mention has been made of the general dislike of tank engines due to loss of adhesive weight as the fuel is used up. Just like a diesel in fact and they did quite well in North America.

 

Edit, one of the narrow gauge railroads in Colorado, please correct me if I have the wrong one.

 

Edited June 17, 2020 by Ohmisterporter

[TheSignalEngineer]

10 hours ago, jointline said:

I recall a LMS driver saying what a revelation it was driving a GW pannier and not having to use a Jinty.  The pannier in his view was far superior.

I would not be surprised if that came from a Bromsgrove man considering that their Jinties were a Fowlerisation of a Johnson design from the 1890s, indeed they did have a Johnson original just before the WR took over. The 94xx which replaced them were a more modern post-WW2 design, higher boiler pressure, more adhesion weight, 4F against 3F power class.....

Their opinion of the 52xx tried by the WR as a replacement for Bertha was not good. Bertha was rated 2 Jinties for calculating banking power whereas the 52xx was rated as 1.5 so neither one thing nor the other.

Edited June 17, 2020 by TheSignalEngineer
Typo

[brack]

1 hour ago, Ohmisterporter said:

As for BP build quality, one of the last classes they delivered was the 20 Class to Rhodesia Railways. There were so many faults with them that RR pursued BP for compensation even after the company had gone into liquidation. A sad end to a great company.

So far as I recall, most of the issues were frame cracks. The 20s had bar frames (unlike the similar South African GMAMs).

In the early 50s there was a glut of big loco orders for BP (and other builders) which whilst welcomed did cause some trouble.

I know that when building the East African 59 frames BP struggled to get steel in the right quantities and quality and had to send some batches back. Also by the early 50s the size of the locos was getting such that the planing machines for frames weren't big enough and some was subcontracted out (new machines were ordered, but delivery was so slow one of them didnt come until after the end of steam).

I wonder if the issues with the 20s stemmed from either of those sources?

Once the initial troubles were sorted the class ran into the 21st century (upgraded in the early 80s by a rebuilding program when parts and fuel for diesels became unavailable through sanctions).

[Ohmisterporter]

One issue, also found on the EAR Class 59, was that many pf the nuts and bolts on the water tanks and bunkers shook loose during running.  Maintenance fitters replaced all plain washers with spring and shakeproof washers to cure the problem. I cannot believe that this was a new fault; it seems more like a cost cutting exercise in a company that was looking to save itself. 

Edited June 17, 2020 by Ohmisterporter

[brack]

I thought at that stage the finances weren't yet so bad (but might be wrong). Perhaps a change in supplier of nuts/bolts forced on them by shortages of raw materials?

[The Stationmaster]

9 hours ago, Lantavian said:

The Malayan Railway ...   ...   even had a Mallet that was originally destined for Russia until the revolution intervened.

 

 

 

That must have been a fascinating exercise in re-gauging, reducing the gauge by c.18+ inches suggests that either the Russian's used very narrow frames or that some very serious surgery took place to get from a nominal 5ft gauge down to metre gauge

[Aire Head]

44 minutes ago, The Stationmaster said:

That must have been a fascinating exercise in re-gauging, reducing the gauge by c.18+ inches suggests that either the Russian's used very narrow frames or that some very serious surgery took place to get from a nominal 5ft gauge down to metre gauge

 

Russia had a few 3ft 6inch gauge lines.

[34theletterbetweenB&D]

On ‎17‎/‎06‎/‎2020 at 08:32, Lantavian said:

East African Railways used loads of Garratts and its lines were pretty steep. This is the gradient profile of the the Mombasa-Nairobi-Kampala line in Kenya and Uganda, compared to the weedy hills of England and Scotland...

Meanwhile,  further South in what is now Zimbabwe 3'6" gauge Garratts were running heavier trains than operated in the UK up the escarpment to Victoria Falls, 1 in 40 for many miles; don't have the route profile, but it's a long grade.

 

I was lucky enough to footplate (or possibly 'footstep' staying a little cooler by hanging outside the oven like cab) on a 15th class 4-6-4+4-6-4 on this route with about 1,200 tons on. The driver had the regulator full open, increased the cut off as we came on the grade, and the two firemen instead of spelling each other, both fired continuously, and up we went. The driver's only visible action was part closing the regulator when the front engine occasionally slipped. All very competent.

[Compound2632]

One thing that puzzles me about the Garratt design is the decision to place the cylinders at the outside ends. The resultant long steam passages must result in a considerable pressure drop between boiler and cylinders.

 

Why was the U1 not a success when tested on mineral trains? The LMS garratts didn't exactly sparkle but they trundled up and down the Midland main line on mineral trains for getting on for three decades so they evidently weren't complete failures for such work.

[LMS2968]

Inability to make steam on a continuous basis, much like Big Bertha, or so I'm led to believe. Worsborough was seven miles long, three of which were at 1 in 40, while the Lickey was two miles at 1 in 37.7. This demanded a high output but over a short(ish) time scale, whereas a train going a longer distance required a lower but continuous effort. Some mortgaging of the boiler could be used as banking engines with the water right at the top at starting and around the bottom nut at the summit, but there was a limit to how long you could do this in ordinary traffic.

[Compound2632]

27 minutes ago, LMS2968 said:

Inability to make steam on a continuous basis, much like Big Bertha, or so I'm led to believe. Worsborough was seven miles long, three of which were at 1 in 40, while the Lickey was two miles at 1 in 37.7. This demanded a high output but over a short(ish) time scale, whereas a train going a longer distance required a lower but continuous effort. Some mortgaging of the boiler could be used as banking engines with the water right at the top at starting and around the bottom nut at the summit, but there was a limit to how long you could do this in ordinary traffic.

 

The S&DJR 2-8-0 tried on Toton - Brent mineral trains was similarly unsuccessful, though adequate on its home turf - the differences in working conditions are similar; the 2-8-0s were of course designed by the same team as produced Big Bertha - indeed, the larger boilers subsequently supplied for the 2-8-0s were made using the flanging plates made for the 0-10-0's boiler. 

[brack]

2 hours ago, Compound2632 said:

One thing that puzzles me about the Garratt design is the decision to place the cylinders at the outside ends. The resultant long steam passages must result in a considerable pressure drop between boiler and cylinders.

 

Why was the U1 not a success when tested on mineral trains? The LMS garratts didn't exactly sparkle but they trundled up and down the Midland main line on mineral trains for getting on for three decades so they evidently weren't complete failures for such work.

K1 has the cylinders on the inner ends. The cab floor gets rather toasty from the pair of cylinders directly beneath I'm led to believe, plus putting them on the outer ends means crew dont have to climb over the motion or around hot cylinders to get in and out.

I rather suspect that the kitson meyer experience ties in too - if the con rod drives on the leading couled axle it causes excess wear, whereas cylinders at the outer end puts the drive (on 4 or 6 coupled locos) on the inner powered axle of the bogie.

 

Both LMS and LNER garratts weren't terrible, just heavily compromised by railway company interference in the design process, producing something pretty mediocre when they mightve been excellent had beyer had a free hand. The LMS garratts had about the highest axle load of any garratts built, but their t.e. was only 7000lbs more than a 9F (approx 46000). At the same time as they were being built BP produced the GL class for south africa - a loco about 25% heavier but for 78000lb te and nearly 5t lower axle load.

Would a more successful garratt have caused repeat orders? More types, change in traffic policy towards heavier trains? Who knows, but instead the LMS and LNER locos reinforced the notion to our CMEs that such things were for "over there" rather than our domestic railways. The spanish 282-282 and 462-264s were built around the same time, were pretty successful and lasted til the seventies, showing that they could work in european service.

 

[Michael Edge]

2 hours ago, Compound2632 said:

One thing that puzzles me about the Garratt design is the decision to place the cylinders at the outside ends. The resultant long steam passages must result in a considerable pressure drop between boiler and cylinders.

 

Why was the U1 not a success when tested on mineral trains? The LMS garratts didn't exactly sparkle but they trundled up and down the Midland main line on mineral trains for getting on for three decades so they evidently weren't complete failures for such work.

The first Garratts did have the cylinders at the inner ends of the bogies but this was very quickly changed. Positioning of larger cylinders and accessibility came into it as well - cylinders under the cab and near the front of the boiler would have been more restricted.

The LMS Garrats weren't so bad for steaming, they were let down by the puny standard LMS 4F axleboxes fitted at Derby's insistence. In the end they were hardly a match for the BR 9F.

[LMS2968]

It was more than just the axleboxes which were an issue. They steamed well enough but ate almost as much coal as the 3F plus 4F combination which preceded them. They were also heavy on maintenance costs which, as was usual with all locos, rose to even more as they got older. These two issues effectively meant that the economic gains were the wages for one set of footplate crew and far less than hoped.

[Ohmisterporter]

On 19/06/2020 at 10:05, Lantavian said:

I find modelling in the UK bizarre at times.

 

There is so much excitement and delight when some starts selling a model of a failed loco or train (APT, any number of post-1955 diesels) or yet another version of a 16t mineral wagon or Mark 1 coach, and yet there is little interest in British successes abroad. 

 

 

 

DJH sell several kits of British built locos for SAR in H0. Lima also produced rtr models for SAR locos and stock, but I cannot recall any that were British built.

[The Stationmaster]

2 hours ago, brack said:

K1 has the cylinders on the inner ends. The cab floor gets rather toasty from the pair of cylinders directly beneath I'm led to believe, plus putting them on the outer ends means crew dont have to climb over the motion or around hot cylinders to get in and out.

I rather suspect that the kitson meyer experience ties in too - if the con rod drives on the leading couled axle it causes excess wear, whereas cylinders at the outer end puts the drive (on 4 or 6 coupled locos) on the inner powered axle of the bogie.

 

Both LMS and LNER garratts weren't terrible, just heavily compromised by railway company interference in the design process, producing something pretty mediocre when they mightve been excellent had beyer had a free hand. The LMS garratts had about the highest axle load of any garratts built, but their t.e. was only 7000lbs more than a 9F (approx 46000). At the same time as they were being built BP produced the GL class for south africa - a loco about 25% heavier but for 78000lb te and nearly 5t lower axle load.

Would a more successful garratt have caused repeat orders? More types, change in traffic policy towards heavier trains? Who knows, but instead the LMS and LNER locos reinforced the notion to our CMEs that such things were for "over there" rather than our domestic railways. The spanish 282-282 and 462-264s were built around the same time, were pretty successful and lasted til the seventies, showing that they could work in european service.

 

I suspect the real handicap on wider UK use of Garratts, just as with the booster fitted P1s on the LNER, was the nature of the wagons they had to haul and the railway over which they had to haul them.  Most mainline gradients where rear end assistant engines (bankers)  were used would still have require the assistant engines to be used.  Thus moving to a more powerful loco would no doubt have allowed heavier trains but would not necessarily had the right sort of impact on operating costs and might well have increased them.  In addition of course there was the other problem of there being teh right infrastructure to enable longer trains to be run.

[Compound2632]

32 minutes ago, The Stationmaster said:

I suspect the real handicap on wider UK use of Garratts, just as with the booster fitted P1s on the LNER, was the nature of the wagons they had to haul and the railway over which they had to haul them.  Most mainline gradients where rear end assistant engines (bankers)  were used would still have require the assistant engines to be used.  Thus moving to a more powerful loco would no doubt have allowed heavier trains but would not necessarily had the right sort of impact on operating costs and might well have increased them.  In addition of course there was the other problem of there being teh right infrastructure to enable longer trains to be run.

 

The Midland main line, with a pair of lines more-or-less set apart for slow-moving mineral trains, was probably their best chance, in terms of making use of line capacity. As @LMS2968 notes, the LMS Garratts were a double-engine doing the work previously done by a pair of engines, so there was no increase in train length or weight that was already at the maximum for the infrastructure.

[brack]

I might also question the use of 5'3" wheels for a loco intended for mineral haulage, though I know they were part of the same standard parts problem as much of the rest of the loco.

LMS2968 mentions they were heavy on coal. It wasnt just the axleboxes, but spring arrangements, short lap valve gear (from the s&d 7F) and maybe cylinders and valves came from derby's standard parts bin too.

A couple of years later the LMS enquired about a 442-244 express passenger garratt with the cylinders and front end standardised with the 3 cylinder compounds. Beyer Peacock offered them a double pacific as a counter proposal - I wonder if that was a polite way of declining the kind offer of having another design hamstrung by derby parts and outdated design (the compounds weren't bad, but the design was 30 years old at this point).

 

2 hours ago, The Stationmaster said:

I suspect the real handicap on wider UK use of Garratts.... was the nature of the wagons they had to haul and the railway over which they had to haul them.

I fully agree, my speculation is that with highly effective garratts it might have pushed things more in that direction, at least for block movements.

 

There were a lot of garratt proposals for the big 4, often for passenger machines over steeply grade/slightly laid (or both) lines, such as the highland main line or the SR west of england lines. I can see an express passenger loco would have less need for change in infrastructure compared to a heavy freight loco.

[Compound2632]

1 hour ago, brack said:

I might also question the use of 5'3" wheels for a loco intended for mineral haulage,

 

That was the standard diameter for Midland goods engines since antiquity - most recently on the 4F. Johnson had built some 0-6-0s with 4'11" wheels, classified as mineral rather than goods (many of these were rebuilt to 3F in due course) but they were rather the exception. So the LMS Garratts had the same diameter wheels as the engines they replaced.

[MidlandRed]

Notwithstanding company parts bin issues, by no means all mineral train locos had relatively small driving wheels. The LNER O1 2-8-0 had 5ft 6", and the BR 9F

2-10-0 had 5' 0".

 

GWR practice seemed to be rooted in Churchward's 28xx 2-8-0 of 1912, with smallest driving wheels of this type of loco at 4' 71/2". Presumably they had different traffic needs from other companies, not having these Midlands/Yorkshire to London block coal workings requiring relative speed also. 

Edited June 21, 2020 by MidlandRed