Steam speed records including City of Truro and Mallard

The critical comment to me is that included on Mr Andrew's third slide above:

"Always look at the measurement system, what accuracy are you entitled to?"

My distinct sense is that had Mallard done 120mph, we wouldn't be having this conversation. Equally, had it done 130mph, we also wouldn't be discussing tenths of a mile per hour. The problem is that its peak speed happens to be pretty much the same as that achieved by a German locomotive, to within a level of precision that I suspect is greater than "we are entitled to" within the two separate systems.

My strong sense is that to within the likely error bars on both measures, Mallard and the German loco can't be reliably separated as to which was quicker - and certainly not the difference between a claimed (by Gresley) 125mph and a claimed (by Deutche Reichsbahn) 200.4kmh. (Those numbers differ by about 0.3%).

The 126mph claim was made post war. I do wonder if there was a deliberate attempt to try to justify that Mallard was indisputably better than a locomotive operated by a regime with which only a couple of years before, we had been in an existential war? That might prompt a search for any small increment that could be had - in which case, would eyes alight on a small blip that could be construed as a whole extra mph, and few questions asked? In the spirit of the era, I doubt too many people would have stuck their head above the parapet to query the process when the end result was to definitively show national superiority over a recently defeated enemy?

In the end, the differences in where the locos achieved what they did (gradients etc) and the loads hauled are of more significance, but the number that sticks in the mind is - well, just the one number: 126mph or 200.4kmh. My personal view is that what Mallard achieved, and what the 05 achieved, are, in speed terms, probably inseparable with regard the cumulative errors on the two measurement systems(*) - but the prevailing spirit of the times meant that the bragging rights of one claim over the other was of profound psychological importance in the prevailing social mores of the country.

(*) By which I mean - you can't say how fast each loco went, but merely plot upper and lower probability bounds; and those bounds overlap sufficiently that there is a low degree of confidence in being able to definitively say either was faster than the other.

Tom

 

I've covered this below.

No, it was not.

Eric Bannister covers this is in his book. He did the official trace and found 126mph the day after the run. He was one of the on board team.

Gresley agreed with his work, but only claimed 125mph as the 126mph was achieved over a very short section (which I think we are all in agreement in) - and he wanted to be assured of the record.

The number put on the plaque was most certainly decided on post war. The news at the time of the run reported both 125mph and 126mph dependent on which source you went to. I very nearly included all of the different headlines in my Gresley book but I ran out of space (and concluded later it was probably better saved for another book on high speed rail!)

I mean, the Germans claimed 200.4kph and that equates to 124.5. Mallard undisputedly is recorded at 125mph. We previously discussed on this thread that .25 of a mph either way would put them close to each other, but it would still make Mallard at least .5 mph faster if we were being consistent and saying they were both consistently slower or faster than their top claimed speeds.

In the end, one was downhill, one was uphill after on the level, both achieved astonishing speeds of over 120mph, sustained. I still maintain we get more out of Mallard's log looking at the run up Stoke bank and looking at the braking characteristics than we do the 126mph.

 

Exactly this. Any paper that claims to be able to do more than this is stretching things too far and personally I'd be surprised if it was to get accepted by an academic institution. The next layer would be to show how various factors could have impacted on the upper and lower measures.

This doesn't mean that the research shouldn't be done, and the probabilities and ranges established.

 

Thanks for taking the time and having the patience to explain things over and again. Those who seem determined to dismiss Mallard’s run will probably never be convinced though.
I have enjoyed reading the results of your studies.

 

You cannot accelerate 400 ton from 125 to 126 with a tired locomotive on very short stretch.
Come up with some numbers instead of I feel ,I think,I hope and I am sure.

 

I will have to make sure to tell my supervisor, the University of Southampton and the Association of Business Historians that...!

I mean, you're not in the room having discussions with other academics on this, with respect. That part of my overall PhD is sat waiting to get published and is pretty much agreed on. With my apologies!

I agree, that is what I have been doing for eight months at this point.

Absolutely: as we said before here, that's the point of research.

You're very welcome.

Hermod, you've provided:

• No primary evidence
• No secondary evidence
• No working out
• No research
• No real calculations

With respect, I am going to leave it at that.

 

Gå venligst væk.

 

OK.

 

It might be that both are seriously inclined, but the criteria that matter are mathematics and engineering. I am confident that Mr. Andrews has a good grasp of both.

 

i think you need to get a life!

 

As am I of @S.A.C. Martin. Which takes us back to the questions of interpretation and evidence, and @MellishR's bemusement that two competent individuals can draw such different conclusions.

 

Perhaps but I think not.

However, @Big Al is correct, more than enough has been written in this thread and I will not add to it.

 

Gents, researchers are allowed to disagree with one another. That’s why we have research!

I am disappointed to note a couple of members would like to throw shade on my abilities and work. This isn’t a schoolyard, gents. It’s serious research. Neither myself nor Mr Andrews should be personally reduced by way of criticism.

I disagree with his interpretation and I would like to see his working out for his graphs and reasoning more, but I will defend the right of the man to disagree with me or anyone else without things being turned personal.

God knows we have enough of that in the world already, thank you.

 

What are the error bars on those figures?

Isn't Hermod's assertion justifiable by simple physics? The kinetic energy of the train at 125 mph is so-and-so. The kinetic energy of the train at 126 mph is so-and-so plus some more. That much more, delivered in a period of a few seconds, is surely more than the possible power output of the locomotive.

But has anyone actually claimed that such acceleration took place? Every point on the speed graph has some uncertainty.

 

I disagree that any of the discussion has been getting tedious. This has been a fascinating debate, but one slightly reminiscent of all those alleged arguments among medieval bishops trying to establish just how many angels could dance on the head of a pin.

Possibly one can only safely conclude that:

Inevitably, the speed measuring technologies used by mid 20th century locomotive testers were subject to slight errors and uncertainties. So they were probably incapable of the absolute accuracy required to prove or disprove speed claims quoted as 124 versus 125 versus 126 mph. Especially as the German measuring apparatus was not calibrated directly against its British equivalents.

Using averaging, to smooth a spiky curve of "instantaneously" measured speeds, is a useful technique. And it can arguably be shown by analysis that the power to accelerate from 124 to 126 mph in the 4.5 sec time available (in the Mallard claim) could not have been provided by an A4 locomotive. But all this doesn't prove that 126 mph was never attained. Given the inevitable slight inaccuracies in measurement, isn't it possible that the true course of the smoothed, averaged speed curve could have been maybe one mph higher, throughout? (Or indeed lower!).

It would be useful to work out just how big the error bars should be on all these old measurements.

Regarding Mallard and the DR 05, we will never know which was fastest; or indeed which design "was best". Both were exceptional locomotives. On test they both attained very, very similar top speeds. Mallard on Stoke Bank had assistance from gravity so the DR 05 performance was arguably the more impressive. On the other hand, the A4 was designed as a practical high-speed workhorse; and was good enough for the LNER to build a large fleet of them, and use them intensively in regular service, as indeed did BR. The DR 05 was designed more as an experimental record-breaker. On the other hand again, we cannot know what use would have been made of the DR 05s - and whether more might have been built, and perhaps further improved - had WW2 not intervened.

Finally, there is a comment up-thread that "even 100 mph is actually very difficult to achieve with a steam locomotive". If one is talking about mid 20th century trains and track, it is indeed hard to argue against that; though some in the USA claim that certain of their high speed services regularly ran at 100+ over significant distances. However, care is necessary about taking things out of context. In principle, steam locomotives could have been developed much further, for example to run modern services at 125 mph or more, if no alternative traction had been available. (While it might be hard to envisage circumstances required to rule out diesel or electric traction, I recall there was a Kingsley Amis science-fiction story set in an alternative British future. In his scenario, all electric power had long been made illegal for religious reasons; express railway services ran at over 150 mph, with triple-headed steam locomotives).

 

We have research in order to advance the state of knowledge. I agree with Simon that the gold standard is publication in high quality peer reviewed academic journals where issues of inference and confidence intervals will be gone through with a fine toothcomb and just as important the interpretation of the results in words will be put alongside the data itself to check for presentation bias.

One question I would be asking is whether the data collected as it was 86 years ago slightly pre GPS is even capable of being interpreted in probabilistic terms or not.

I am sure we can rely on Simon to provide his supervisors with a concise digest of the comments in the last 1117 posts which he regards as material to his argument. I don't necessarily buy the proposition that they are more knowledgeable on this particular question than some (not all) of the commentators here.

 

125 miles per hour was what Sir Nigel Gresley publicly stated.

There are indications he was of some personal probity: literally interpreting the speed as achieved over a mile - not a shorter distance. He had hoped for a faster speed: the instructions to the footplate crew were to the locomotive crew to give the engine her head and 130 mph was mentioned. Afterwards the driver - Duddington - regretted to his grand child that the speed was not quite that.

I would be quite content to take the speed as 125 mph out of respect for him and in particular his honesty, not rounding up the recorded information he had to hand.

 

I would like to agree wholeheartedly with this especially the very last sentence.

There was an unexpected speed limit that day through Grantham station at the foot of the climb up to the top of Stoke Summit before the racing ground down through Essendine to the level before Peterborough. Had Mallard been able to start the climb 10 mph faster and come out of the tunnel, say, 5 mph faster this might have pushed the maximum speed up into the high 120s or even more.

But going faster on the climb might have been more than the boiler could sustain and the speed two thirds of the way down the descent not so remarkable.