Discussion in 'Steam Traction' started by Reading General, May 5, 2017.
As an aside, I wouldn’t fancy doing 125-126mph in that, pretty though it is.
Human reaction time, yes, agree, but how were mileposts laid out? If it was just chain and chain then I can imagine cumulative errors...
That makes CoT a single cylinder loco as a Saint only has two cylinders.
108mph on just one cylinder then? Wow!
Wipes egg off face.
It rode better after 1928 - Gresley requested it to be placed on a set of his bogies and it was arranged for the vehicle to be overhauled more fully at that time.
You have to remember that Gresley was anticipating higher speeds and his carriage and wagon departments and its heads would have been engaged in looking at bogie design together with other factors (including articulation) towards higher speeds.
Hence my scepticism, really, given what we know broadly about milepost locations.
This has been an interesting read, it's not a topic I know much about but I happened to look up Flying Scotsman's 100mph claim on Wikipedia (yes, I know, not the best source perhaps) and there are two quoted:
30/11/34 - 100mph
5/3/35 - 105mph
So on the second occasion there is no doubt that it did 100mph even taking your 2% error margin.
Papyrus is not listed on Wikipedia, what speed did it do and was it before or after 5/3/35?
On the subject of accuracy of setting out of quarter mileposts, as a signal engineer these tend to get surveyed each time we draw up a scheme and I recall when working on West Coast Main Line a few years ago that one in particular was 15m adrift of where it should have been, presumably having fallen over at some time and being stood back up in the wrong place. But this was unusual, most were fairly accurate.
However, you have to ask how you measure milepost distance on a curved railway. Are they measured along the cess in which they're planted, along the adjacent line (nearest or furthest rail?), or along the centre line of the formation? And even if the latter, which would make sense from an initial construction point of view, if the railway is later quadrupled the centre line moves but you can bet that the mileposts don't.
Over a longish distance the curves balance out (well, maybe not on the Circle Line) but over a given quarter of a mile the Up and Down lines might be a few feet different.
I'm guessing most of these speed records were set on straight bits? (Sorry for lack of geographical knowledge.)
Papyrus's 108 mph run was also(?) on 5/3/35.
Mallard's dyno roll & graph in Alfred Gottwald's Baureihe 05 - Schnellste Dampflok der Welt
Be careful - the only definition of a mile that you can use for Mallard's run are the mileposts - otherwise how do you define a mile. (It's can't be whatever the dyno car says it is - it has to be something external that you can check the dyno car against)
Now, as it happens the difference between the mileposts (taken over 15 miles) and the dyno car was only 4 parts per ten thousand, which suggests that both dyno car and mileposts were pretty accurate. (That statement applies to whole miles, as mentioned earlier the error within miles as the paper sped up and slowed down was more than that)
FYI Carling stated 1/10,000 for the wheel and 1/10,000 for the rollers that drove the paper. He didn't say anything about the rest of the mechanism, slippage of wheel on rail or paper on roller. I suspect the chains used by surveyors were also of the order of 1/10,000.
Some of that variation has been introduced by my measurements - so don't hang your hat on that 0.3 seconds.
Rous-Marten would have agreed with the risk that a single timing would be in error - that is why he published a series of timings (which you can plot a realistic curve through) and why he took an additional timing of the whole mile (37.2 seconds). And of course there was a second person timing the train.
Outside of Thomas the Tank Engine this isn't a competition between CoT and FS. If you look at the history of technology there is a big difference between 100 mph in 1904 and in 100 in 1934.
In 1904 German electric railcars had achieved the ton a year or two previously - but no other manned vehicle had done so. For any type of vehicle of any nationality this would be a significant milestone in human achievement: 100 mph achieved during a revenue earning journey (the German runs were experiments) - and you should spend more time considering the evidence.
Slightly confused by that statement. I assume you mean specifically rail vehicles - by 1934, the aeroplane record was 440mph, and the car record 272mph - and those were proper records, too! (As in, set in controlled, standardised conditions with defined measurement methodologies).
have a read now, should make more sense
All this is very amateurish. When you have corrected for track curvature, cant, gradient (real), variable atmospheric pressure, air temperature, wind speed and direction, Coriolis effect, variable acceleration due to gravity, curvature of the earth, tides, elliptical orbit of earth and time dilation, then we can have a debate about accuracy.
Otherwise we could accept the results of the best available technology and, at this remove, give them all laurels as they were/are all magnificent.
The experimental German Siemens electric railcars' record was set in 1903. They both attained around 210 km/h. Their speed was controlled by telegraphing the steam engine generating the current to adjust its speed, so hardly a practical system for actual use!
Whilst it shows that the two were accurate in relationship to each other, they could both have had the same sort of error. OK so that may be fairly unlikely but you are making an assumption which we all know is never a good thing.
Er - hang on a minute Courier. Are you saying you have to disregard the automatic mile notches in the dynamometer car's record?
The dynamometer is the known quantity. We know it takes automatic measurements and that those automatic measurements are scaled. That's the point of the whole mechanism: where the physical mileposts are in relation to the actual miles recorded is almost entirely irrelevant. What matters is that the dynamometer car knows it has travelled a full mile and makes that recording.
No difference in principle to the Odometer on a car - and we know from our own experiences, I would hope, that the accuracy of a car's odometer does rely on a couple of variables such as tyre wear, pressure and as intimated by yourself earlier, any binds or similar within the mechanism of the recording device.
In any event:
1) The dynamometer car notches give you automatic measurements of miles the dynamometer car has travelled over a journey. This together with the time and speed recorded can give you the full DST equation and you can look at the measurements and work out the inaccuracies (if they apply).
If you check the automatic measurements of miles and find significant variations between individual measurements, that would indicate the wheel that drives the rollers or another part of the mechanism is not working within the parameters required to give good readings.
You haven't stated that there are any such significant variations, though you have indicated that you suspect there was some. You haven't said to what extent?
2) The human notches give an indication of a milepost sighting
You have confirmed that the human element throughout all of this has a potential inconsistency. My question to you would be whether you trust the consistency of a machine designed specifically to record miles travelled automatically or the sight and recording of a human with a button. Again, I know what my answer would be to this.
But you have just showed us a section of the roll where the human element was inconsistent when compared to the dynamometer's automatic readings. So that suggests that they have different levels of accuracy and again the Dynamometer car is likely more accurate.
There are an awful lot of assumptions being made there and elsewhere in your argument.
If the question is actually: do all of the variables within the mechanism of the dynamometer car add up to make the potential inaccuracies in Mallard's record greater than CoT's, then I would ask you directly how you resolve the big question mark over the timepiece used by Rous Marten?
Nowhere in your argument have you given me a make, model or similar for us to even try and work out the potential inaccuracies arising from that. That device is a big unknown unlike the dynamometer which we can reasonably assess and ascertain some level of inaccuracy.
A new variable. Do we have his records too to compare?
I don't believe anyone said it was.
But your evidence for CoT is as simple as a man (and now two men, apparently, introduced at a late stage of this discussion) watching out for mileposts on a speeding train with a stopwatch of unknown origin and accuracy, and human reaction times to factor in together with any potential errors between the man, the milepost and the sighting.
The evidence for Mallard can be examined, the vehicle in question studied and understood better, and better men than you or I have had the opportunity of using said vehicle and using in daily service during 1948 for exchange trials.
If the vehicle and its mechanism were as inaccurate and flawed as you are trying to claim, I rather suspect that any number of ex-railwaymen from other railway companies or indeed from British Railways would have been lining up to take pot shots at the LNER and the records it claims. Yet I can find no such criticisms from the very men who would have had inaccuracy in data at the forefront of their minds. Cecil J Allen devotes an entire book to the exchange trials of which this dynamometer car was a prominent tool and there are no such qualms with the data presented to British Railways.
There's an awful lot of trust and assumptions made to make CoT's run appear more believable; there is a lot of questioning of the LNER's recording device in order to undermine our confidence in it.
To me, I would like to rationalise CoT's record beyond what has been said but you will not convince me that a man with a stopwatch is anywhere near more accurate than a device specially designed, built, set up and improved over a number of decades to record speed, distance, time and drawbar horsepower.
When all's said and done you've introduced more doubt to CoT's record in my view than you have introduced merit; and in Mallard's case I can accept quite happily that there are more variables I was not aware of than before - but I question if all of the tiny deviances you are describing ascribe to changing the record so drastically as you seem to intimate.
Am I to take it you are using a secondary source to make your criticisms?
Separate names with a comma.