Steam speed records including City of Truro and Mallard

As ever, Tom, you are not wrong in what you are saying. How we establish that crossover point is much harder to obtain. It is fairly easy to do the calculations for the DBHP required because it is simply engineering maths and there is a good amount of information available on the likes of wind resistance and even the effect of streamlining. Where we are short of information is on the locos themselves. In theory the steam loco is essentially a constant torque/TE machine (unlike a diesel which is constant HP) but we all know that, in reality, this isn't the case. As speed increases the ability to use steam in the cylinder reduces due to the internal friction of the steam pipes and valve gears ability to admit and exhaust that steam. We can't leave the loco in full gear because it would soon become self defeating. We have very little information on all this in reality. In the past locomotives have been indicated in an attempt to see what is happening in the cylinder and we have a few in depth reports on this. However, the Dobbie McInnes and other indicators of old used to establish this information have their limitations as speed increases. The springs upon which they rely to follow the rise and fall of pressure and piston travel have to be sufficiently strong to react to the changes in these parameters and that gets harder as the rate of change increases. I hesitate to guess when a steam loco was last indicated in the UK but I would think it was in the early 1960's . It would be interesting to indicate the likes of Tornado today using modern technology and it's something that could probably be done from the comfort of the first coach, perhaps even remotely, rather than being sat in a cramped indicator shelter. What I don't think has been done is try to establish the effect on steam pressure and superheating as speed and steam consumption rise. We only have info on boiler pressure and indicated steam pressure to tell us what the overall pressure drop is. CFD would be a useful tool in establishing this. Then there's the rolling resistance of the locomotive as a vehicle that we have little information on, especially at the higher speeds that are being considered in this topic.

And all this is before we consider the fireman and the locos ability to actually burn coal and turn that energy into steam. We need all this information before we can determine that crossover point.

 

All these questions are answered in BR bulletin 8 class V2 testing.
The only open question is if cylinders of Tornado or Mallard are more efficient than V2.
It cannot be much.
The specific steam consumption of V2 was as good as Class 7 Britania and 9f and according to Cox only Duke was better in UK.

 

01 1104, the former Carnforth resident, was rigged up in the summer and measurements taken for power and valve settings.

https://m.facebook.com/story.php?st...ZywdakedVDuNXRFd1VErsQFEcl&id=100057139942545

 

Thank you .

 

I'm not inclined to agree with that. How do you equate a V2 with Tornado or other locos and, especially with the high speeds that we are discussing. 60845 only achieved 85 mph during the tests so a fair way short of 100 mph. We may have lots of information about 60845's chartacteristics but I'd suggest that none of them are transferable characteristics and i don't think you could extrapolate them to 100mph and more. The conjugated valve gear gave some pretty poor indicator cards although, admittedly, at relatively lower speeds and tending to improve at the higher ones. I would hope Tornado's are better than that but have we any evidence?

Tom's argument was that we can predict a maximum speed by establishing where required DBHP and output DBHP lines cross. Required DBHP is relatively simple to obtain but establishing output DBHP as a desktop exercise is going to be pretty difficult and Idon't think you can infer that information from tests on other, different, locos.

 

A locomotive with 74 inch drivers doing 85 is mechanically behaving as a 81 inch driver thing doing 93.
Mallard once did 124 and has nearly same machinery as V2.
A numerical model that can predict if 126.5 or 124.5 is truest cannot be made without a lot of public money.
A cheap numerical model can illustrate that gradients matter for wild speed claims.

 

A few gentle observations if I may.

Yes the geometry of the V2 boiler is similar to that on Tornado ( as fitted to the Peppercorn A1s)

But from front end differences Tornado differs in other respects from the V2 wrt potential
maximum power eg. boiler operate at 250 psi vs 220. I.e higher saturation temperature,
Cylinders are 19 inch dia compared with 18.5 and of course the larger grate 50 square feet vs 41.25.

( Similarly suggesting 60845 is analogous to an A4 is I suggest an oversimplification.)

As Steve and Tom have said the difficulty with attempting to estimate the power developed by Mallard
at greater than 120mph is what the mechanical, friction etc resistances with the loco were?.

We have thanks to Dynamometer Tests, Indicator shelters etc. over the years and latterly the work
carried out at Swindon/Rugby supported by road Tests ( particularly thanks to Messrs Andrews and Ell )
data which enables either to generate quadratic equations by graphical fit or by ‘long hand’
procedure to estimate total power requirements in moving a loco ( i.e. total loco resistance )

We do not AFAIK have any data wrt locos operating well in excess of 100mph and IMHO
any extrapolation is questionable.

We can of course hazard a guess of R for the carriage stock at 125mph based on Tests on latter day
stock, wind resistance etc. I am however always doubtful about estimating power downhill,
Whilst simplistically, forces are equal and opposite, and therefore if I can estimate EDHP
ascending; by estimating the work done against gravity by locomotive and consist ( A ) plus work done by
the consist in overcoming resistance ( B ) , I am less certain that Downhill the EDHP is
(B) - (A) Let alone the internal resistance of the loco.

All I know is that I think the maximum observed IHP for an A4 is c.3000-3100

Michael Rowe.

p.s have corrected deliberate error

 

Michael, we don't need to estimate the power developed by Mallard. The Dynamometer car records drawbar horsepower on the same graph it records time and distance.



I have been able to, by use of a simple spreadsheet, calculate the speeds and compare to what was given officially against the milepost records.

The original roll records quarter mile distances: you can then calculate to the half mile and match it against the official record.

Seconds between each quarter and half mile is the distance recorded on the graph. Drawbar horsepower is given in horsepower and shown on the graph accordingly.

This in part why I am more confident in giving my answers, and supremely confident that Mallard achieved 126mph beyond all reasonable doubt.

 

Simon,

Many years since I examined a reproduction of the said Dynamometer role but is it not a measure
of drawbar pull i.e. enabling DBHP to be calculated. It is the difficulty in estimating the 125mph IHP that
Steve, Tom and I have been discussing.


Michael. Rowe

 

Hi Michael,

I’m going off what is physically written on the dynamometer graph roll. I can’t get any more “primary evidence” than that.

best wishes

Simon

 

Simon, you have primary evidence for the power exhibited at the drawbar i.e the power
to haul the carriages DBHP. There is no evidence AFAIK wrt the power ( Indicated Horse Power )
developed in the cylinders,

Whatever, the performance was astonishing, particularly IMHO the acceleration from 104 to 109 over the
essentially level 1.5 mile between Posts 97 and 95.5. This probably represents an IHP greater than 3,000.

Michael Rowe

p.s this is where the A4 external casing ( stream lining) helped significantly !

 

Simon - you need to be careful to distinguish the whole train (including locomotive) from just the carriages. The dynamometer tells you the power available at the drawbar, after all the internal resistances in the locomotive, but the physics problem you are trying to calculate includes the locomotive!

At a basic level, you can ignore the difference between the loco and the carriages since they are moving as a single object, and just express the problem as "how much power is required to sustain a given speed of a given train on a given gradient". Or to put it in terms that are of interest in a practical sense - "we know that a 400 ton combination of locomotive and train could do 126mph while moving down a 1 in 200 gradient. How fast could the same locomotive and train sustain on the level?"

The issue, as @Steve has pointed out, is that there is some empirical data for train resistance (i.e. the carriages), but not much for the loco's internal resistance, nor much data on how much power the locomotive can produce at high speed. What you can say is that a 400 ton mass moving at 126mph down a 1 in 200 slope is getting close to 1500hp just from gravity. So if @Maunsell907 is correct that an A4 can produce 3,100 indicated horse power (or thereabouts), then you would need somewhere well north of 4,500 indicated horsepower to do 126 mph with a 400 ton train on the level, assuming the locomotive was packaged into a Mallard-shaped exterior as far as aerodynamics was concerned. That's clearly beyond the wildest excesses of any British locomotive, but it doesn't answer the question about how fast an A4 with a 240 ton train could run on the level, except that the answer would be considerably less than 126mph.

Tom

 

What's being said is that the dynamometer car measures the force and / or power put into the coaches to move them at the measured speed. It cannot and does not measure the force / power being used to move the engine at that speed, since that has already been used before it reaches the drawbar. You can to some extent extrapolate the figures to give you an approximation - an educated guess - at the indicated horsepower being produced, which is what the LMS did for 26 February 1939 for the high power trials with 6234 and came up with 3348 i.h.p. It was probably in that area, but only probably.

 

Exactly that.

What is interesting from the numbers posted by Simon is that at the point the train stopped accelerating, only 1265hp was needed at the drawbar to sustain the carriages at that speed (assuming that they weren't gradually being left behind ) If the 3,100ihp estimate for an A4 is close to reality at that speed (a big assumption I know) then it suggests something like 60% of all the power available was being absorbed by the loco, either in it's component of the overall drag, or in the resistance in the machinery. Perhaps not a big surprise if the lubrication was insufficient to provide the necessary cooling to dissipate all that heat!

(If you read the literature on high performance aero engines of the period, it is amazing just how much thought was given to heat dissipation - not for nothing were the Supermarine Schneider Trophy racers of the era known as "the flying radiators!)

Tom

 

And my hero mr Cox wrote that it was probably overestimated from dynamometer DBHP value.
He gives 36000 lbs of steam per hour as max new production and 14.2 lbs per ihp so max 2535 ihp.
Higher values means heavy short time mortgaging maybe like the peak shown for britania.
So steep and abrupt that it migth as well be some coupling jerk.

WP_20180722_003[1] by Hermod posted Jul 22, 2018 at 4:46 PM

 

I am always wary of estimated and extrapolated values for ihp; however when actual indicator diagrams are involved they're handy in this context, because they give an absolute ceiling beyond which no further power can be generated. Holcroft plotted curves of ihp against speed for a Lord Nelson operating at various cut-offs, although the curve is only complete for 25% cut off. The curve indicates a peak of 1450 ihp at about 65 mph (at 25%), from which point the ihp rolls over and starts dropping. From the incomplete curves you can extrapolate that for shorter cutoffs the peak is later but lower, and for longer cutoffs the peak will be at lower speeds. The Southern didn't have a dynamometer car, so indicator tests were their most thorough form of investigation. From these various investigations, Holcroft concluded that ihp tended to peak on most designs at a piston speed of around 1000 feet/min.

This of course is not much use for considering ihp in an A4 at 125mph, but let's take a couple of assumptions: The cause of reducing ihp with speed is throttling in the ports reducing the pressure doing useful work in cylinders. Note that even with Chapelon's suggested port cross section one fifth of piston cross section, the steam velocity in the ports has to be five times the piston speed to maintain the cylinder pressure constant. The peak piston speed in an A4 at 125mph is circa 3570 feet/min, mean piston speed 2270. This is obviously well past Holcroft's suggested 1000 feet/min peak, and the pressure drop will go with speed squared. Lengthening the cut-off doesn't necessarily help, because it means the required port flow velocity gets higher as the valve stays open into the part of the stroke with the highest piston speed, rounding off the corner of the indicator diagram where the valve closes. So even if you assume the steam circuit for an A4 was 50% better at speed than that in an early LN, it's probably developing its maximum indicated power at around 90 mph (50% per cent more flow area means the flow velocity matches at 50% greater speed, leading to basically the same pressure drop, assuming flow through an orifice is much of a muchness in these terms). So by 125mph, whatever it was generating in the cylinders was I suggest unlikely to be higher than the 3100ihp as per @Maunsell907 and probably lower.

The vast heat rejection requirements of interwar aero engines was more about getting the combustion heat out of the walls at a rate that allowed the temperature to be kept below the oil decomposition temperature. Since the working fluid temperatures in steam locomotives are much lower, most of them would have benefitted from better cylinder insulation, but in an IC engine the high combustion and exhaust temperatures mean you've got to keep pulling heat out to maintain the wall temperatures low enough. Even then, it was often a bit overkill; you've got to keep the coolant temperature below boiling too and that meant some engines had to be over-cooled for the sake of their coolant systems, a definite case of cart before horse. The pressurised radiator systems typical of Rolls Royce aero engines allowed them to run the coolant hotter than German and American equivalents, which did slightly reduce the power lost to the cylinder walls.

I will add a side note that drawbar horsepower requirements to keep the train moving go down if the locomotive is aerodynamically filthy; a while back somebody showed figures from wind tunnel model tests suggesting that the small tenders on Fowler Royal Scots developed thrust, not drag. Of course, they were behind a great big engine creating low pressure in front of them and in front of a larger coach generating high pressure behind them; this is not a recommendation for small tender propulsion since both the other vehicles experienced more drag as a result. The total aerodynamic drag still goes up with an aerodynamically dirty locomotive, but you won't see that in the dbhp.

 

I have resisted sticking my neck out, but here goes
.
At 100 mph a 240 ton train on the level ,using BR MK1.R=20.0 lbs/ton for total Resistance, will require 1280 DBHP.
At 100 mph a 160 ton loco will require c. 1500-1600HP
i.e. total ( cylinder HP ) IHP to sustain 100mph on Level c. 2800-2900
That is why Tornado, quote from an account at the time “struggled to reach 100mph ?
If assume A4 casing saves 150-200 HP then c. 2600-2700 IHP

At 105mph on level with 240 tons
By extrapolation R = 23.9 lbs/ton then DBHP = 1400
160 ton loco by extrapolation at 105mph 1700-1800
Allowing for casing 1500-1600
IHP = 2900-3000
Hence my previous comment re Mallard acceleration from MPs to 97 to 95.5 needed
the savings of the casing ( “stream lining “ )

I putatively think the boiler was being mortgaged over the 1.5 miles and the
109mph at the commencement of the 1/200 made the 126 possible. I also suspect
this is why Duddington immediately eased/shut off afterwards.

Michael Rowe.

 

There’s a few points bothering me about the discussion thus far on what, really, should probably be back in the Gresley thread (or in a Mallard thread - we do probably have one somewhere), so I’ll bullet point these as best I can:

• Use of the BR Mk1 as a vehicle isn’t comparable to the actual makeup of the record breaking train on the day (Mallard plus dynamometer coach plus six streamlined coronation coaches. There is air resistance to consider for the whole of the train, including the significantly draggy Dyno plus the genuinely streamlined coaches.
• So IBHP isn’t the same as Drawbar horsepower - totally appreciate that. However, we do have the latter from Mallard’s dynamometer roll, so using for the basis of an informed more accurate estimate for IBHP, if we wanted
• The comment Tom has made about an A4 on the level can achieve has bothered me, because it doesn’t match what the dyno roll gives us, so as below:
• The little Bytham stretch the run was done on is 1 in 200 at the start, then flat on the level for around a mile, then 1 in 240 (126 mph was actually achieved directly after the flat stretch, and this is reflected in the dyno roll)
• Therefore I can confirm, after checking, that Mallard with its train on the level achieved a sustained 125mph for almost a mile, which frankly is probably the achievement of the record aside from the aforementioned standing start to 80mph acceleration up Stoke Bank
• Duddington shut off steam as there were speed restrictions ahead, together with the smell from the middle end bearing having alerted the test crew in the dyno to advise of easing off too (Bannister)

Far too much theorising going on when we have factual records for Mallard.

 

Was the A4 indicated during this record run? I've seen nothing anywhere else to suggest it was. But that is the only way you can MEASURE, as opposed to extrapolate or estimate, the indicated horsepower, which is what you seem to be suggesting. It cannot be done from within a dynamometer car. Nor can it be done on rollers such as at Swindon or Rugby, come to that; it has to be done at the cylinders and deduced from the resulting diagrams.

I agree with 8126: I am unconvinced by extrapolations of calculations based on resistance formulae. It assumes all the stock is conforming to an average, which is unlikely, or the plusses and minuses cancel out at the end. Aerodynamic resistance? How many windows were open? And these days how many heads were protruding from them? They'd make a big difference.

 

Simon, I have never hidden theorising, extrapolation etc.

If I may take some of your points;

Why Mk1 Because originally my thoughts were in relation to ‘Hermod’ posting re Tornado
twixt Darlington and York, followed by queries concerning 100 mph records on the Level. I
appreciate that the consist behind Mallard was made up of an elderly Dynamometer Car with various
protrusions and ‘ stream lined stock’ . (I suspect the Dynamometer mitigates against any
advantages ‘stream lined’ stock might have had over MK1s at speeds greater than 100mph)
I have attempted to make allowance for the effect of streamlining compared with data for
a non ‘cased’ 4-6-2. I doubt my overall figures are any better than plus or minus 2.5%.

Please its IHP , the B refers to bar in drawbar. IHP is the Indicated Horse Power. The power
developed in the cylinder, ( DBHP drawbar horse power, EDHP equivalent drawbar horse
power is DBHP plus any work done by the loco climbing a gradient. )

Re Tom’s comment. I leave it to him.

My understanding re the last mile or so of Mallards record.
The 1/200 Down finishes at MP91, there was then c.700 yards level, on which the roll showed a
falling off in speed, then c. 1200 yards down at 1/240 on which for c.200yds 126 mph
was shown. At MP 89.25 steam was shut off and brakes partially applied. At Peterborough the
loco was taken off the train and one of the Peterborough Ivatt Atlantic pilots continued on
to Kings Cross.

There is IMHO nothing to confirm “a sustained 125 mph for almost a mile” not least because
there wasn’t nearly a level mile and on the short stretch of level there was an apparent decline
in speed. ( as it would require something over 4000IHP, even mortgaging the boiler, I think not)

None of the above diminishes the outstanding performance, although in terms of repeatable operation
I have always thought the performance of Silver Link some two years previous with 280 tons of train
and speed remaining above 110mph for a considerable distance past MP 89.25 was a true indication
of the A4s ability to operate comfortably at very high speeds. There are many other examples of their
high speed capabilities, but 125mph on the Level with 240 tons is not one.

Michael Rowe