Flying Scotsman Speed

This graph compares the speeds calculated from the dyno roll in 1938 (when the LNER claimed 125mph) and those from a 1964 study at Doncaster when Mallard entered the National Collection. The 1938 data is based on measurements over 5 seconds and the 1964 data is based on measurements every second. On a piece of paper moving at 24" per mile one second at 125mph is 0.833" and at 126mph it's 0.840"


And the half second on, half second off pulses are not that 'sharp'...


My own measurements show the same general speed curve shape as the 1938 measurements (but not exactly the same speeds).

 

I shall dig-out Yeadon's Registers for the A4s and the A3's to see if any of the preserved locos' boilers ever ran with 60096 - A preservation project! (Incidentally one of only four A3's which I never "copped".)

Slightly OT, and related to the V4 thread, 60881 (with double Kylchap) was officially recorded at 101½mph down Stoke Bank in October 1960, leading to the fitting of Kylchaps to a further 5 V2s for inclusion in the Pacific fleet. More were to have been fitted, but the Deltics started to come on-stream.

 

As has been noted by many people over the years - the Flying Scotsman and Papyrus curves look "dodgy" and the Mallard one is a bit odd.

 

To what extent have the curves been correlated with the detailed gradient profiles and the changing drag of the train from the gradient? Some of the reconstructions I've seen seem decidedly sketchy, maybe even weak on would seem to me to be a critical factor.

 

This is what would a speed curve look like from 68mph at mp100 to 100mph at mp91 - includes effect of gradient and rolling resistance - assumes constant cyl power - which is a simplication - and can't take account how rolling resistance might change going from a cutting to an embankment (tho I think that would be a minor effect) - so I'm not saying the black curve is right - but it gives an idea of what a realistic curve might look like.

So yes there would be a blip in speed as gradient becomes level near mp91 - and it might be more than the black curve - but not as pronounced as FS or Papyrus curves show.

 

.

 

can't say where I saw this , but allegedly the Swindon works manager would always take the Kings over the ton,light engine , before they were released to traffic , again allegedly , 112mph + on occasion

 

Isn't one of the prime doubts about CoT's claimed top speed (no denying it was b' fast) that the reciprocating masses would have derailed it at the circa 100mph mark. Given comments (IIRC from Bill Hoole) about the way a flat out A4 moved in a way differently from any instance he'd previously felt when he pushed one well over 100 perhaps CoT's brakes were applied for a similar reason. The driver felt the loco behaving unexpectedly and no longer felt it was a safe speed and he didn't want to crash it as a previous poster also stated. Hoole's comment I read recently but regrettably can't remember where.


Sent from my iPhone using Tapatalk

 

Yes - wheels would have lifted from track. However only the wheels on the axle directly driven by the con rods (as they had the big balance weights) - not the other pair of driving wheels and of course not the bogie wheels. Also LH and RH wheels didn't lift at the same time. So doesn't mean instant derailment. Lady of Lyon's 1906 run down to Little Somerford would also have been with wheels lifting from track.

Both Cities and Saints (and other GW classes) had balancing distributed over other driving wheels to reduce this effect after the report on bridge stresses in 1920s.

 

A bit off topic, but hey-ho. In the investigation of the Sevenoaks derailment, it emerged that the leading wheels had been off the track for 560 yards before the final derailment, which occurred when passing over a catch point; moreover, there had been another incident in which a loco had become derailed at its leading wheels and run for 3/4 mile before re-railing itself on a crossover. So wheels off doesn't instantly lead to a full-on derailment (if that is not an oxymoron).

Tom

 

[edit]Great work plotting this stuff out! Better to see real data than opinions and arguments from authority![/edit]

Is this plotted with a smoothed curve? When reviewing data, abrupt anomalies and data spacing can be disguised when Excel (or whatever chosen package) plots with smoothed lines; was it one point that caused the line to spike up, which could easily be an instrumentation error, or several points, which would be more likely to be a reflection of something real.

 

The Mallard curve is plotted from speeds annotated on the dyno roll. (As mentioned previously I get slightly different speeds when I run a ruler over the dyno roll - the line in this graph is based on the LNER figures)

The Papyrus curve is created by me manually measuring off a speed curve in the NRM (a curve drawn by an LNER draughtsman from measurements off the dyno roll) and entering it in Excel. The Excel curve looks very similar to the LNER graph.

The Flying Scotsman curve is drawn the same way - however in Excel it does not come out as a smooth curve and doesn't look like the LNER graph. Perhaps if I spent longer on it I could get a better result. Here is a snapshot of the original LNER graph:


You can see it is a much smoother curve - BUT it does show an unbelievable acceleration from 95mph to 100mph in less than a mile. In a nutshell that is why there is doubt about Flying Scotsman's 100mph.

 

4472 meccano magazine January 1935 FISH7373 81C NFP

 

Thanks for that. The 100mph sustained for 600 yards mentioned in the article corresponds to the "plateau" in the speed curve - so the article and the LNER speed against distance curve agree. However it is very hard to believe that curve is feasible - and hence the suspicion that the LNER were a bit optimistic in how they interpreted the time/distance data on the dyno roll. Unfortunately the original dyno roll no longer exists.