Steam speed records including City of Truro and Mallard

Simon, if you haven’t read it, my friend John Heaton’s article, based on work by Bill Hemstock,
in the May 2019 Railway Magazine re City of Truro might interest.

Michael Rowe.

 

I assume that IHP figure for Tornado is for a particular section of a particular run. A loco with 50 square feet of grate should be able to sustain at least that IHP for a prolonged period (fireman permitting) and considerably more for a limited period.

 

The max steam possible for V2 when tested and improved postWW2 was 31000lbs/h and ihp 1990.

If Tornado boiler was like Peppercorn A1, it has bigger firebx and grate area,but same tube heating surface and gas cross section as a V2
My models estimate of 2400 for tornado 102 mph run need not be miles of,and most likely to high.

Mallard had same grate area as V2 but sligthly more tube and firebox surface so it will be interesting to see where my model land when I have the nessecary information.
I need total train length,mass and gradient for the Mallard all out power

 

Yes, I have read it. They are about 10mph north of where I am currently expecting CoT to have likely achieved an absolute maximum of. But that'll be for me to explain when I publish the PhD in about four years time.

Put it this way, it's not looking good for any non streamlined locomotive with less than a 42 square foot grate...!

 

I assume your 100mph and 42 sq.ft grate refers to performance on the Level. A light Bulleid Pacific
has achieved 100mph on favourable gradients twixt Basingstoke and Woking, similarly a Royal
Scot between Luton and Bedford.

It also obviously depends on the weight of the consist. Assuming the BR derived figures for
Resistance of MK1 stock are correct ( i.e. 20lbs/ton ) then a three coach consist ( say 110 tons)
would require c. 600 HP ( the arithmetic gives 586, allowing for measurement accuracy
570-600 ) If six coaches 1200 HP. ( These figures represent DBHP. )

A medium size large dia. coupled wheel 4-6-0 would require c. 1100-1300 HP to maintain
100 mph on the level. A Castle, probably a Jubilee, with three coaches could briefly hold
100mph on the Level. Obviously no wind and a higher calorific value coal assist.

And yes at this speed streamlining the loco helps ( despite this I doubt one of the
two streamlined B17s would cut the mustard. )

COT carried, albeit in saturated form, a Swindon Standard No.4 boiler, capable of
producing 1500 IHP with the standard Swindon front end. With the lightweight
train, a falling gradient of 1/80 ( let alone the possibility of a south west trade wind
at the rear ) speeds IMHO well into the nineties become feasible even allowing for the
rapid acceleration required after exiting the tunnel.

Yes grate area is a limiting factor, but only when the corresponding total evaporative
area reaches a certain level . Why specifically 42 sq. ft ?

Michael Rowe.

p.s Don’t forget the 5at.

 

I seem to recall that the basis for BR power classifications distinguished narrow from wide firebox locomotives favouring the former. If that's relevant?

 

But you have not accounted for drag on the locomotive and coaches by way of air resistance, which was shown very prominently by Frederick Johansen and Professor Dalby at the NPL in their research. I think you will find that City of Truro is a VERY draggy locomotive design, to say nothing of the coach design of the era. That behind the locomotive actually matters more.

I am hoping to get the use of a wind tunnel to re-run the NPL tests and CoT is one I'd love to put through the same exercise. Watch this space!





(Bert Spencer's paper of 1947 gave these stats, which I then put into a graph).

I will be honest Michael: I forgot the Bulled light Pacifics were 38ft! So quoted what I misremembered was the smallest pacific grate of note, the A1/A3 at 42 square foot.

Hmmmm....!

 

What is the UK steam speed record, authenticated, on level track?

 

The BR MK1 R is the total resistance to movement i.e.air ( actually in this case a wind speed of
7.5 mph at 45 degrees ) friction and internal . Similarly the figures quoted for the mid
sized 4-6-0 represent the total energy required to move the loco.

My calculations, are based on BR data, whether Rugby or Swindon, wrt the total amount of
power exerted. As a matter of note the specific Royal Scot ton was with a330 ton consist and
the effort with 34102 c.250 tons. It was for that reason I offered figures for a three coach
consist.

Bert Spencer’s figures are for air resistance only I.e c.40% reduction at 100mph
but of the total loco resistance this will be a c.15-20% reduction.

Michael Rowe

 

Speed 55.7msec
I have found lenth of train to be 154m
Gradient 240
Mass of train 407metric ton
Asume air density 1,23kg/cubic m
Fudge factor rolling 0.0015
Fudge factor eqvivalent drag area 0.13


gravity assist 407000*9.81*55.7/240 = 927kW
Rolling loss 407000*9.81*55.7*0.0015 = - 334kW
Air loss 154* 0.13*0.5*1.23*55.7**3 = - 2128kW
Nessecary Cylinder power 927-334-2128= 1535kW= 2058 Ihp

A V2 was tested post WW2 after being optimized by mr Ell and maxed out 1990 IHP.

 

Tornados 102.5mph?

 

If testing City of Truro it might be surprising, particularly if compared with the wheels turning, although all pretty meaningless unless measuring the entire train, although I can't imagine coaches of that era being very smooth, but possibly smaller cross section than say mk1s.

 

I don't believe there is one. It would have been quite a game to identify an area that was flat enough for long enough to eliminate all gravity assistance, and the number that would come out of it would be pretty disappointing compared to the numbers gained from careering down various hills. I suppose someone keen enough could identify if there are any such areas, and then go hunting through published logs to find any such areas, but the numbers you would get would surely be in the course of normal service, and not the result of an attempt at maximum speed anyway.

 

It’s a good question, but almost irrelevant in the grand scheme of things.

Very few places in the UK where railway records have been made are on exactly level track.

In fact it’s my contention that some of the meaningful records were actually not the speed records, but the accelerations up inclines (such as looking at Mallard’s dynamometer car roll reveals some astonishing statistics for horsepower and speed going UP Stoke bank, not down it).

 

I am more than a little dubious about any attempt to calculate what speed a given locomotive really reached, bearing in mind some past attempts have produced results that to my mind clearly fail sanity checks. It's clear it's a very tricky endeavour if one is to avoid making assumptions that have little evidence behind them.

What I will suggest though, is that I believe there are a number of other logs of the Dean/Churchward 4-4-0s running at speeds in the early 90s. I suggest that any calculation should be run against those logs as well. If the algorithm suggests those are incorrect as well then it would be a fair presumption that the methodology is invalid.

Unfortunately the reverse is not true since it's probable such runs were not driven as close to the limits as the Ocean Mails trip, but such an exercise would give a partial control.

 

The problem with this line of thinking is that you’re only looking at the cross section of a carriage and not the really draggy areas of trains, which are normally the connections between locomotive, tender, and between coaches, together with the shape of the rear of the train.

I agree with Jim - some of the previous calculations have some absolutely wild results. The only things we can look at it more closely are any physical records from measuring devices (and by that I mean dynamometer car records or similar, emphatically not timekeepers logs).

The methodology for assessing speed records needs to be consistent across a range of data and events, taking into account the same fundamental issues and applying rigorous (and frankly, as far as is possible, a unbiased eye at the claims).

One of the things which drew me to look at Mallard’s dynamometer roll this year, for example, was realising that it in fact records time, distance and drawbar horsepower all on the same graph, so you do in fact get instantaneous results from a study of the roll.

When I used the measurements from the dynamometer roll I got results I was not expecting, and that was because I was not using the same method as other people have done before. Others followed the LNERs approach of getting an average over the distance for specific five second intervals: the thing abundantly clear to me when I then remeasured this was taking the datum points in that way actually introduced more inaccuracy than using the ones actually recorded.

This gave me absolute maximums for each mile and quarter mile. Given we have distance and time, recorded instantaneously against each other, it makes sense to measure against that rather than what are almost arbitrary 5 second intervals (which don’t reflect the true instantaneous top speed anyway, which appears to be higher).

I need to get it peer reviewed and that will likely come as a result of working on it during the PhD.

 

There are possible sources of error in any measurement method, timekeepers' logs included, but surely the right approach is to take account of the uncertainty, not to dismiss logs out of hand. They do have one huge advantage over dynamometer car records; that they are far more numerous.

 

So this is something for the PhD discussion, but I have done a lot of work in examining specifically LNER timekeepers logs from over a 20 year period and some of the errors in recording are astonishing.

No, not every single log will get dismissed out of hand, but right now for every five log I have examined, I have discounted four. This is on the basis of some obvious errors. So far I am around the 25 log mark across numerous secondary sources. The same names keep cropping up with similar errors, I am afraid!

That is true, they are more numerous; that doesn't mean they hold greater sway over the dynamometer records or similar that we have.