Alternative High Speed Steam?

Rous-Marten was aware of the existence of milepost errors and mentioned it in one of his articles, which is why he tried to obtain a series of readings between consecutive posts; indeed in (I think) the same article he mentioned recording 90 mph but commented that this was not certain because he was unable to time the quarters before and after.

 

A bit late responding to this one, but ...
I think we are entitled to say that CoT definitely did go very fast, with a maximum of about 100 mph before the brakes came on. We cannot be sure of the exact figure. I agree that the onus is on those wishing to claim >100 to prove their claim, but the same applies to those (you?) wishing to claim <100. Post #62 is revealing of both the likely figure and the uncertainty.

If there were indeed no speeds close to that for another 30 years, then CoT well deserves its place in history, regardless of the actual exact figure.

 

Could also be the move to inside frames and outside motion, lots of things flailing about, or just exposed wheels are much worse for aerodynamics than when those parts are largely covered.

 

No, the onus is on those claiming it to prove it beyond reasonable doubt. I have don't have to disprove something that has not been proven!

That's never been a bone of contention for me, though, and the LNER thought it a significant enough locomotive that they subsequently bought it for preservation...!

 

I feel this pretty much sums up why I am dubious.

 

Not to take anything away from recognising and being grateful for the vision of the LNER in both realising that a railway museum could be a viable proposition and setting it up, I don't believe they purchased her. GWR Locomotive Committee minutes, 12th March 1931
"Under circumstances which were represented the Committee approved the presentation, after renovation at a cost not exceeding £65, of the engine "City of Truro" (No 3717) and tender to the London and North Eastern Museum at York, on the understanding they would be returned if required in the future."

 

My apologies. I think we both agree that without the LNER's subsequent inquiry, CoT might have been scrapped.

 

They do indeed, but if you study the detailed history, as opposed to the rosy tinted press accounts afterwards, a large part of the story is usually a constant struggle for finance after a charismatic individual gets the speed record bug. There have of course been times when generous government or corporate funding has been made available, but not many! A mate was part of the team that currently holds the outright record for a sailing craft on water, and that was very much their story.

Well no, but that comes down to what I had to say before about granularity. 112.5 is an 8.0 second 1/4 mile. With a 1/5 second stopwatch (unusual) that's between 8.1 (111.1) and 7.9 (113.9) , and the more usual 1/2 second stopwatch between 8.25 (109.1) and 7.75 (116.1). Given a watch with a second hand the range is between 106 and 120mph!

 

Back in the day, unless there was a reason to impose a speed limit (curves, condition of track, etc) there was generally no upper limit on speed. It was pointless having (say) a 100mph limit if none of your trains ever reached that speed. Thus, there was nothing to stop a driver pushing his loco to the limits of its ability and there were people who would do this. I think that a 100 mph limit on the ECML was the first and only came in with the advent of the Deltics when they first went into service. (Happy to be corrected on this statement, though.) Bill Hoole was only stopped from exceeding 112 mph on 60007 because the inspector on board said that was fast enough.

 

I had the impression that you were claiming that CoT definitely did not reach 100 mph. If that is your claim then the onus is on you to prove it. If you are only saying that we can't know for sure, that's fine.

 

I most certainly am sceptical that it did 100mph, but again it's up to those who make the claim to prove it, not for me to disprove it.

 

Splitting hairs, the claim is out there and widely accepted as a truth. I think the onus is on anyone seeking to either reaffirm that truth, or to challenge it, to prove their case.

 

Well over a hundred years since the event it is most unlikely that any additional evidence will appear beyond what has been known since the story eventually came out. We will never know for certain whether CoT's maximum speed was a little above or a little below 100 mph, but we have the evidence that it was about 100 mph. Can we not settle for that?

 

I don't think it is accepted as "truth" - most railway historians don't treat it as proven. Scotsman's 100mph is the "first authenticated" - that is the widely accepted record worldwide.

I have challenged it.

I am not convinced by the evidence provided in this thread that CoT definitely did 100mph or over. If anything it's made me disbelieve it that little bit more (sorry Jim but your graph and analysis of the Rous Marten figures was instrumental in making my mind up that CoT probably didn't achieve 100mph).

CoT is a lovely 4-4-0 that is famous for a high speed run that is claimed to be over 100mph. It's unproven that she achieved it. Therefore it's an unauthenticated record. That seems fair.

In Bert Spencer's address to the iMechE in 1947 he produced a set of figures which showed the required horsepower output to overcome head on air resistance for the LNER's streamlined and unstreamlined Pacific locomotives. See below:



I have been playing around with this and I have produced a graph for my book, as follows:



When you look at the required horsepower for the non streamlined locomotive to achieve 100mph compared to the streamlined locomotive it tells its own story.

 

My suggestion for inside cylinders was as much for the reduced side to side thrusts, thus leading to a more steady ride at high speed, obviously there is also a benefit of being able to get a cleaner aerodynamic side covering. But I still think that the limiting factor was the convoluted steam passageways associated with the slide valve layout. Gresley had large dia piston valves and smoothed steam ports on the A4 locomotives.
I am also influenced by the developments in internal combustion engine design. Sleeve valves were also a very significant design element in the 1930's and 40's....Bulleid's "Leader."...

 

450 hp seems achievable, especially with a gradient, but these figures aren't very helpful, unless you are running light engine.

As Gresley seemed to understand it was streamlining of the entire train not just what the shape of what is on the front, it's impossible to consider things in isolation.

Also are these figures produced with a stationary windtunnel model, a model with the wheels and motion moving, or "on the road" testing? The latter is more likely to prove most accurate.

 

Although I have read all this I am still not sure whether City of Truro reached 100 mph or not, but I am very glad a number of people thought it did.

 

Good question, I will find out for you.

 

Since both the Salisbury accident, and City of Truro's exploits on Wellington Bank have been mentioned, some notes about the relevant train working at that era.

In 1904, the American Line started a weekly cross-Atlantic steamer service. In England, the main arrival point was Southampton, however, as the liner passed Plymouth, it would be met by a tender to transfer the mails. Doing so saved about 12 hours relative to waiting until the ship docked at Southampton, and a few passengers would generally avail themselves of the service so as to arrive at their destination sooner, albeit in rather less comfort.

By arrangement, the GWR took the mails while the LSWR would take any passengers (on the day of the Salisbury disaster, there were 43). The bulk of passengers of course stayed on the ship and disembarked at Southampton some hours later. The Royal Mail preferred the GWR for the mail contract because the route (via Bristol) gave better onward connections to the Midlands and the north.

The general timing for the GWR was 4h 15m to Paddington with a loco exchange at Bristol, and for the LSWR 4h 20m to Waterloo with loco exchange at Templecombe. The GWR route was longer, but the LSWR had to reverse where the Stonehouse Branch met the mainline, which served to equalise the times. For the first few weeks both services ran close to booked times; however, there was a marked acceleration from the LSWR on 23 April 1904, which was followed the next week by the GWR and lasted for about four weeks. For example, on 7 May 1904, both companies ran in 3h 59m.

The inaugural run (on 9 April 1904) from the LSWR was not without incident. The time taken was 4h 23m (against the 4h 20m schedule); however, it was punctuated by a 7 minute stop near Whimple; and a 90 second stop near Wilton, just short of Salisbury. It was to be over 80 years before the reason for the stop at Whimple became known - the loco had in fact collided with a p/way trolley on the line! No derailment occurred and, despite the presence of the press on the inaugural service, including "Sekon" (G.A. Nokes), the affair was hushed up. The stoppage at Wilton was to allow Drummond (who was travelling on the train) to issue an instruction to the crew; it appears that this may have been to convey a specific instruction to moderate speed through Salisbury station, there being no explicit limit at that time.

On 20 April 1904, eleven days after the inaugural run, Drummond issued a specific instruction to locomotive crew setting a 30mph limit for trains passing through Salisbury non-stop.

The LSWR accelerated their train (in practice, not in timetable!) on 23 April 1904, while the GWR ran to time, such that the LSWR time that day was 4h 01 (+19 minutes on schedule) against the GWR's 4h 14m (+1 minute). The LSWR's train that day had Rous-Marten on board, and his timings indicate that the train picked up 10 1/2 minutes between Templecombe and Waterloo. He did not at the time give detailed timings through Salisbury, just recording that they went through "at full speed", but two years later, after the Salisbury accident, he claimed a speed of 75mph for the 1/2 mile before the station "with no appreciable slowing" and a speed of 60mph exiting Salisbury Tunnel. If Rous-Marten's account is correct, the loco - a T9 on that day - must have traversed the curves at the eastern end of the station at around 65mph, in complete contravention to Drummond's instruction which had been issued only days before. However, the crew had the Outdoor Running Superintendent F.C. French on the the footplate and actually lost two minutes to Salisbury, so it seems perhaps more likely that they did in fact slow somewhat through Salisbury and Rous-Marten was "sexing up" his description of the 23 April 1904 train after the 1906 accident, for which he perhaps had motive. (His contention was that the Salisbury accident was not primarily caused by excessive speed).

If the timings of that run are correct (reported at the time in the Western Daily Mercury in a report by "our London Correspondent" and repeated in the South Western Gazette) then apart from the headlong rush through Salisbury, the train then ran at an average of 74mph between Salisbury and Woking; and no less than 87.5mph average between Woking and Clapham Junction. The writer then went on "The train dashed through Clapham Junction at 7.59 and drew up with a flourish at Platform No. 5 in Waterloo Station exactly as the hands of great clock pointed to 8.3". Allowing for the relatively short distance of that latter section leading to a wide error bar, it nonetheless points to a nominal average speed of 56mph, clearly in contravention of a 40mph limit at Clapham Junction.

(As an aside - the Clapham Junction passing time was clearly a stopwatch on the train, set against a clock at Waterloo, so the times are not exactly robust; but even an extra minute still only takes that last average down to 45mph. The florid writing style of the Western Daily Mercury piece suggests that the "London Correspondent" may have been Rous-Marten writing anonymously. What is clear is regardless of the precise details, the train ran very fast that day, particularly between Salisbury and London.)

Shortly afterwards, on 3 May 1904, Drumond issued another instruction specific to the boat trains. "This train must run to scheduled time in the future". At the Salisbury accident enquiry, French explained to Major Pringle that that instruction had been explicitly issued because of an incident of excessive speed in the Basingstoke area - presumably the April 23rd run. Drivers arriving at Waterloo more than five minutes in advance of time were subject to being removed from the footplate.

After June 1904, there were no further examples of particularly fast runs, and it seems that Drummond's instruction, coupled with the very severe sanction that crews running faster than scheduled would be removed from the footplate, served to take the heat out of any notional "racing" that was taking place.

Just to complete things from the GWR side, City of Truro's fast run - with Rous-Marten on board - was on 9 May 1904. This was an Ocean Mails Special, though not in that case in connection with the American Line service, i.e. not running in direct "competition" with an LSWR service. The combined time was 3h 47m and the LSWR must have realised they would never get close to that time; after one last flourish on 14 May, when the LSWR ran in 4h01m and the GWR in 4h02m, the racing seems to have stopped - or at least, there are no published timings. In June 1904, Drummond issued another warning, and there were no further complaints of excessive speed.

Taken together, it seems that there was a very brief outbreak of "racing" between the GWR and LSWR for a period of a few weeks in April - May 1904, but very firm management action by the LSWR stopped it.

Skipping forward two years, the Railway Engineer carried a pithy article after the Salisbury Derailment, in which:

"Only a few years back when a train left a station late it arrived at the next equally late. But a band of self-appointed railway experts arose and found it profitable to make 'logs of runs' for the daily and other papers, and they began, in the language of The Jungle to 'speed up' one railway after another.

The arch speed expert, Mr Rous-Marten, was soon in print with the object of showing that speed was not to blame [at Salisbury], but that appearances pointed to some 'hidden flaw in a wheel or axle or in the permanent way' and to denounce those who dared to express a different opinion as 'irresponsible and ill informed'".
​

That latter comment probably refers to Rous-Marten's more detailed description of the 23 April 1904 run - published after the Salisbury derailment in apparent argument that high speeds were not dangerous - and in particular his alleged 65mph through Salisbury. Given the findings of Major Pringle's investigation in to locomotive centres of gravity, his estimate is that the train loco - an L12 4-4-0 - would overturn at Salisbury at around 67mph. There is no such calculation for a T9 but with a lower boiler, it would probably scrape round at 65mph, but it would have been a close run thing, apart from being profoundly uncomfortable. The enquiry, incidentally, found no flaw with the locomotive or carriage wheels or axles in the Salisbury accident.

Tom

 

Following on to the circumstances around the actual derailment, an obvious question is what specific instructions were issued about speed through the station.

The station had been extensively remodelled in 1902, and it appears that no instructions were given as to acceptable speeds over the junction at the east of the station. The crossover there from the up through to up local line had a radius of 7.5 chains over a reverse curve. It appears that initially, Jacomb-Hood's view (the Civil Engineer) was that no limit was required, since most trains stopped at Salisbury. The only instruction was therefore a general rule from the rule book that:

"Engine-drivers of trains, when running through Junctions to or from Lines diverging from the straight road, must so reduce their speed as to ensure a steady passage for the whole train through the Junction Points and Crossings."
​

In other words, the speed limit was not absolute, but instead at the discretion of the driver as to what they felt would give a "steady passage" to the train.

Instead, it was Drummond - the Locomotive Superintendent - who on 20 April issued a specific instruction about observing a 30mph limit through Salisbury. Rous-Marten's description of the 23 April run says that was widely exceeded that day, against which there is the fact the loco had French on the footplate and it seems reasonable to assume that, if not right down to 30mph, it did at least slow considerably from Rous-Marten's inferred 65mph. Major Pringle, at least, was fully aware of the non-standard situation in which a speed limit had been imposed by the running department rather than by the civil engineer.

The day of the accident was the night of 30 June / 1 July 1906. Significantly, the GWR were due to start running passenger services over the Castle Cary cut off, eliminating the "Great Way Round" via Bristol, on 2 July 1906. That has led to some speculation that LSWR were out to have one last fling before the GWR gained an unassailable advantage of a shorter route. However, the first section of the LSWR boat train that night - Plymouth to Templecombe - arrived only one minute early. Given the known capabilities of the locos shown two years before, that was hardly racing speed, and clearly Drummond's injunction that the train must not run ahead of the scheduled times was being scrupulously observed.

At Templecombe, Driver Robins attached, on L12 No. 421. This was a bigger loco than the normal T9, with a high-pitched boiler. Robins had joined the LSWR in 1884; by 1906 he had 8 years experience as an express driver and was fully familiar with the route between Exeter and Waterloo. His fireman was named Gadd. Together, they had departed Salisbury at 8.12pm that night, running in reverse light engine to Templecombe, where they arrived at 9.06pm. At Templecombe, they were shunted into a siding to await the boat train, which was not yet due (indeed, it hadn't even have left Plymouth by time Robins and Gadd arrived at Templecombe, not leaving Plymouth until 11.01pm).

Robins and Gadd thus had a long wait at Templecombe, during which Robins oiled round his engine. All who spoke to him noted that he appeared sober and alert. A shunter who spoke to him remarked on the running of the boat train, to which Robins replied "Yes, I shan't get into Waterloo before time, else I shall have to go up and see the governor" [i.e. Drummond].

So it is clear that Robins had no intention of racing, and that is further supported by his timekeeping when he did eventually depart - on time - at 1.24am. Between Templecombe and Dinton - a distance of just over 20 miles - Robins lost 4 minutes on the schedule. (The fireman in me can't help thinking that having over 4 hours wait at Templecombe can't have done wonders for the fire, and the loss may well be attributable to the fireman struggling until the fire heated up). After Dinton, he began gradually to recover time. Timings are only from signal boxes, so subject to inevitable error over short distances, but it appears he was running normally at around 70mph through Wilton. Proper practice was to slow down so as to pass the Salisbury West signal box at 30mph, continuing at that speed through the station before accelerating again up towards the Tunnel. Instead, he continued at high speed and derailed at the east end of the station on the 7.5chain cross over, striking a stationary Beyer-Peacock goods engine, with the carriages piling up behind. Major Pringle calculated that a speed of at least 67mph would be required to overturn on the curve in question.

The question is therefore what had caused Robins not to slow down. His record showed he was clearly experienced and knew the road well. It is also clear from his words at Templecombe, and his loss of four minutes in the first twenty miles, that he was not intent on record breaking that night, despite later speculation.

The most likely cause, in my view, was set out by Norman Pattenden in a South West Circle Monograph. In it, Robin's hours of duty in the lead up to the accident were set out.

29 June - Book on 5.30pm, work through to 4.00am 30 June

[Then followed 9.5 hours rest - the standard at the time was a minimum 9 hours]

30 June - Book on at Nine Elms 1.30pm, work the 2.50pm Waterloo - Exeter express as far as Salisbury, where he arrived at 4.42pm. At that point he booked off immediately (after just over 3 hours work), before signing on again at 8.00pm. He then took the light engine to Templecombe at 8.12pm, arriving at 9.06pm. He departed after a long period in the sidings - in which he oiled the engine - at 1.24am; the crash occurred just before 2am.

The suggestion in the monograph is that Robins probably had a "micro sleep" just on the approach to Salisbury, caused by a combination of fatigue and changing shift patterns. Seemingly his approach to the normal braking point was normal (in terms of sounding the whistle etc) but at the moment he should have braked, he didn't. That state of affairs may only have last 30 seconds or so, but at a speed of around 70mph, that was plenty to cause him to run through the station at full speed to the inevitable crash.

Tom