Steam speed records including City of Truro and Mallard

I stand corrected on that point.

Nope. That’s just what has been reported by secondary sources.

The actual dyno roll shows Mallard continued to accelerate on the flat section. You could only know that by examining the roll and realising where the mileposts are in relation to the actual quarter and half/full miles attained.

Michael. I have the actual roll in front of me. I paid a significant amount to have it scanned and can report on it accurately.

I am confident that much of the discourse around Mallard’s run is actually down to people not understanding what the 125mph or 126mph records actually refer to.

Particularly that the originally recorded 125mph was claimed by average speeds over 5 second intervals taken from the graph. The 126mph was instantaneous speed at one point on the graph.

What hasn’t been done - and it is something I have done - is to work out the instantaneous speed according to the distance travelled/time taken from the graph, from which you will get slightly different results, but nevertheless showing Mallard’s supreme achievement was 125mph sustained for nearly a mile. This is clear from the dyno roll itself.

More troubling - particularly given the desire to beat the Germans absolutely - is that graph shows a different story to many of the secondary sources. I am not convinced that people like Nock or Allen actually looked at the original roll. If they had, they would have had questions of it, and not because Mallard potentially went slower than 126mph.

 

As you say, IHP has to be measured at the cylinders and not anywhere else using an indicator (hence the 'I') to produce an indicator diagram from which you can calculatethe mean effective pressure. Mallard certainly wasn't indicated during its record run. Indeed, I have my doubts that any A4 was ever indicated out on the road as I have never seen a photo of one with an indicator shelter at the front and that would be a necessary pre-requisite in the days of steam where indicators were purely mechanical contraptions. 60007 may have been indicated at Rugby on its visit there but I have my doubts. It was only there over three days for the opening of the Testing Station and was probably only on the rollers for demonstration at the opening and nothing else. Certainly the Pathe News film shows no indication of any of the necessary pipework from the cylinders to the indicator bank.

BHP on its own generally refers to brake horsepower, brake being a reference to the dynamometers used for measuring the output of engines as, essentially they are putting a brake on the output and measuring the power absorbed by this brake. It can also refer to boiler horsepower, which is an entirely different thing and was a term used to describe a boiler's output of steam. A boiler horsepower is approximately equal to 13 horsepower. This equates to an overall efficiency of about 7% so, if you wanted (say) to drive a machine requiring 100 hp you would use a steam engine of 100 hp and feed it with a boiler that would produce 100 boiler horsepower of steam; actually about 1300 horsepower. To add to different horsepower abbreviations, you can also have RMSHP (root mean square horsepower) which is used with electric motors that have a cyclic or similar load where the speed is a significant variable, usually involving frequent starts and stops. RMSHP is the equivalent constant horsepower and is generally around 50% of the peak horsepower. It is related to the motor's ability to dissipate heat.

 

I think my question to that would be whether all the data presented by your derivations from the dyno roll is internally consistent with the rest of the information the dyno roll provides and consistent with the external environment. So Mallard accelerated on the flat, having just lost the equivalent of 1500hp of gravitational assistance. Does the drawbar pull spike to match this behaviour, is the acceleration profile broadly consistent with the drawbar pull profile?

I don't like over-extrapolation and theoretical models of coach resistance too much, but I do like physics, and while there may be other components, F = ma never goes away (at least at locomotive speeds).

 

How much can specific steam consumption vary between A4 and V2 and how much can max steam production between their boilers vary relative to driver diameter?
V2 with all Swindon blastpipe technology was indicated and published in BR days.

http://users.fini.net/~bersano/english-anglais/BR-tests/BR_P&E_No8_LNER_V2.pdf


A lot of knowledge about behaviour of steam in cylinders can maybe be picked up from

http://users.fini.net/~bersano/english-anglais/The development of Locomotive Power at speed 404.full -O.pdf

 

If that is what the roll shows, it demands some very thorough investigation.

 

There’s no “if”. And that is precisely what I am doing with it.

 

Very easy as mr Martin has it in digitized form.
An periodic step length variation of speed as postulated by mr Courier must be first target.
True or false
Next step will be to construct an energy model of the last ten or five miles from geodetic data for whole train combined with an air resistance model that need not be exact but non- physics violating either.
Lots of fun.
Can we have max -minimum of water in boiler to estimate how much extra heat can be stored?

 

All to be performed within the constraints of the NRM's copyright of the original roll and the images made of it - IIRC, the reason why @S.A.C. Martin has not shared the images.

 

Fair enough! Your table does show reduced acceleration from mp97, which is about where the level section starts (actually, if the published gradient profile is to be believed, a short level section followed by a short falling gradient and then another short level section). What was the DBHP on that stretch? The steeper falling gradient starts again about 95¼ but the acceleration seems to stay about the same, possibly because of the rapidly rising losses at those speeds. As you have said, the dynamometer car does tell us how the loco was actually pulling the train. Can you show us the DBHP figures? (Aplogies if they're already further up this thread or one of the others.)

 

Simon, I have been wondering why you are insistent re 125mph was sustained on the
Level, even though it would require a power output way beyond any British steam loco,

AFAIK the Dynamometer Roll included work done and speeds. It was annotated by hand
with cut off, boiler and steam chest pressures, location and gradient data.

i.e gradient data was added by hand.

The speedo I understand was jammed at the end of the dial i.e. 120 mph after MP92.

I have copies of “Gradients of the LNER” published in the early 1930s
and Gradients of the British Main Line Railways published 1947.

They both show, as per my previous posting, commencing MP 91 c. 700 yards level
and then c. 1200 yards down at 1/240.

I suspect the hand written data includes a note re the commencement of the short
level stretch. At 124/125 mph 700 yards takes 11.5 seconds. My understanding is the
roll shows a slight fall in speed initially after MP 91 followed by an increase to 125/126
on the subsequent 1/240 Down.

I also believe the added boiler pressure data stops at MP 92.75. There are no more
figures for the next three miles - there has been suggestions I believe that the
boiler pressure was beginning to fall as the cylinders were beginning to beat the
boiler.

Whilst I recognise the quite correct questioning of HP calculations ( In this case we
have DBHP via the drawbar pull measurements but estimating loco resistance
beyond 100mph may be questionable. I have invariably when writing on
performance limited my thoughts to DBHP and EDHP because of the
uncertainties ) However it seems likely that the IHP was >3,000 from MP 98 to 93,
suggesting a steam demand approaching 50,000lb/hr.

The rapid fall in power after MP94 is perhaps surprising, a fall in cylinder efficiency
perhaps ?

However I do hope we can now forget any thoughts of 125mph sustained on the Level.
The limitations of hand written data addition ?

Michael Rowe

 

Grate area is identical in A4 and V2
Heating area is 7 % greater in A4
Giesel Gieslingen ,who designed locomotives for a living, calculates max reasonable evaporation to 28100 /28700 lbs steam per hour for V2/A4.
BR on purpose made V2 front end limit production when making 31000 lbs/h on better coal.
The value for A4 is anybodys gues but it is likely not more than 32000Lbs/h continiously.
If mortgaging boiler cut-of can be higher but up goes specific consumption and rapidly

 

Hi Michael,

You’ll appreciate that I have the actual roll in front of me.

On that I think I’ll leave it.

Best wishes

Simon

 

Precisely.

 

If we are to be pedantic I think you'll find there will be no copyright on the original roll. Indeed its questionable whether a dynamometer trace is subject to copyright. However before the NRM let you have a copy of anything, under copyright or not, you are required to agree to a contract which limits what you can do with your copy. And unlike copyright, contracts don't expire.

 

And, presumably, anyone else wanting to carry out similar research (or attempt to refute Simon's after he's published) can go to the NRM and request their own copy on the same terms.

 

Yes indeed, that is the point.

 

Thinking of this in terms of forces and the data available.

• Things we know: Newton has been unchallenged for four centuries in terms of macro-scale dynamics
• Things we claim to know: The gradients, the speed at various points, the mass of the train
• Things we can calculate: the corresponding acceleration at various points (i.e. was the train getting faster or slower, and by how much)
• Things we don't know: The propulsive force being applied by the loco; the resistive force of the loco and train (air resistance, friction etc).

What you can look at though is what happens at a change of gradient. Suppose the train is going down a gradient. You know from F=ma what the net force is on the train (from the measured acceleration - which can of course be positive, zero or negative). You also know that that net force is the combination of propulsive (unknown) + resistance (unknown) + gravitational assistance (known). That means you can calculate what the difference of the two unknown forces is (i.e. the propulsion minus the resistance) even though you don't know each of them individually.

If the train then transitions to a level bit of track, before it has slowed by any appreciable amount, the two unknown forces should remain the same (i.e the propulsive force; and the drag / friction). However, the gravitational force disappears. The consequence of that is that you can still calculate the new net force on the train (that will have decreased by the gravitational component). That should directly correspond to a revised figure of acceleration. The new acceleration on the level may be a smaller positive number (train continues to accelerate, but less rapidly); or may be zero (train maintains the speed it was doing when it got to the level) or may now be negative (train slows down). But the data should show the effect: the change in acceleration that can be derived from the measured speeds should adjust by an amount required by a decrease in force due to gravity by the change of gradient. That should be a fairly simple bit of analysis from the known data, and would give a quality check on the data.

For a 400 ton train moving from 1 in 200 down to level, you essentially take out nearly 4,500lbf from the net forces (about 20kN). That should take out about 20,000/400,000 ms^2 out of the acceleration, or 0.05ms^2. In imperial units - suppose the train had just reached a steady speed on the 1 in 200 descent. The effect of going onto the level is that it should lose about 1mph every 9 seconds, so would lose just over 3mph in a mile of level track.

Suppose you did the calculation and found the change of gradient didn't affect the acceleration. Then your options are (1) the speed data are wrong (2) the gradient data are wrong or (3) the driver succeeded in the mythical quest for third valve on the regulator. 3 is unlikely, and I didn't bother with option 4, which is that Newton was wrong.

Tom

 

# 167 propose same in fewer words.

 

This scientific ignoramus has heard of something called "momentum". In his ignorance, and quite possibly wrongly, he understands this to mean that a train travelling at a given speed will take a little time before the forces described above would achieve their full effect.

That leaves two questions of clarification:
1. Is this ignoramus correct in his understanding of momentum?
2. If the answer to (1) is yes, how much can that be expected to skew the expected results?

 

I guess you're right but @Jamessquared's additional words made it far more intelligible to someone like me, who only studied Physics to O-level