Locomotive weight. How accurately measured is it?

[melmerby]

Hi all

A lot of loco drawings both steam and MI have axle loads shown under each axle.

How accurate are these figures?

 

e.g I saw recently a claim that the ex GWR "King" class locos were well over the published weight at the time of their withdrawal.

It claimed that there was as much as 25tons on a driving axle well over the published 22.5T and the supposed limit of the civil engineering structures.

It said that the fitting of larger superheaters and strengthening plates on the frames had contributed to this and was one of the reasons why they were withdrawn somewhat earlier than they need have been due to the lack of routes over which they could be used when they were displaced from front line duties.

 

Are these claims true?

If so how are the preserved locos now allowed on the main line?

 

Cheers

 

Keith

Edited September 19, 2017 by melmerby

[Davexoc]

Hi all

A lot of loco drawings both steam and MI have axle loads shown under each axle.

How accurate are these figures?

 

e.g I saw recently a claim that the ex GWR "King" class locos were well over the published weight at the time of their withdrawal.

It claimed that there was as much as 25tons on a driving axle well over the published 22.5T and the supposed limit of the civil engineering structures.

It said that the fitting of larger superheaters and strengthening plates on the frames had contributed to this and was one of the reasons why they were withdrawn somewhat earlier than they need have been due to the lack of routes over which they could be used when they were displaced from front line duties.

 

Are these claims true?

If so how are the preserved locos now allowed on the main line?

 

Cheers

 

Keith

The Swindon weighbridge was at the end of the A shop, or through the tunnel at the end of Dean Street by the Apprentice Training School. It would have been able to record the individual axle weights but I wouldn't know how accurate it was, but would have thought better than cwt resolution. Whether every loco was checked, who knows? But I would have thought it was checked pre-release from Works.

That weight of course is static and presumably could be done either dry or in a fully prepped ready to roll state, but fully watered and coaled would be the only way to get the real weight distribution. Dynamically that axle weight is going to vary due to acceleration, braking, cornering, gradient and the load being pulled.

If a King weighs in at 89T (according to Wiki) it is unlikely the weight distribution was such that there was 25T per driver as that only leaves 14T over the bogie in total.

 

Dave

[EddieB]

The overall weight would not have increased by 7 1/2 tons (at least) such that each driving axle exerted a static load of 25 tons.  Besides, I think the figure quoted is for one of the driving axles, not all three.  However, as Dave says, axle weight varies dynamically.  It has been remarked that the GWR 4-6-0s tended to "sit back" on their rear drivers when pulling away - increasing adhesion and thereby performance (in contrast to the LNER Pacifics, where any similar extra load was applied to the trailing axle).  A bit of speculation, but I'd hazard that the 25 tons wasn't measured, but an estimate of the total load on that axle under certain dynamic conditions.

[Gwiwer]

As a general point of order locomotives were always weighed in working order i.e. with a full supply of water and coal.  Without that a false impression would be given of the load on any axle and the weight overall.  I would have expected locomotive weighing apparatus to be accurate to within tens of pounds though the quoted weight is usually only given to the nearest cwt.  Any locomotive would tend to sit back on its rear wheels at start because it has to overcome physical forces and lift the weight of a stationary train but to what extent and upon which wheels - therefore affecting the adhesive weight - depended upon loco design, wheel arrangement, size and to an extent the unique local conditions at any given point where a start was made.

[melmerby]

The article I saw (unfortunately I cannot now remember where), plausibly claimed that the locos had gained weight due to bigger superheaters plus frame patches which were not there when the loco's published weight was originally given.

When new in 1927 this was 21.5T (bogie) + 22.5T + 22.5T + 22.5T = 89T which is the weight of the loco full, as built.

 

As the maximum permitted static axle load on a double red route was 22.5T they were (conveniently) right at the maximum.

I note that no other GWR loco has such a balanced load on the driving wheels which seems, to me, more than just a coincidence.

(Mind you Stanier managed the same feat of 22.5T with the Princess locos on the LMS, although the "Turbo" exceeded it on a couple of axles)

 

Is it possible they were always over the published weight on one or more driving wheels which was done to keep the Civil Engineers happy (at least on paper) but nobody raised the point as they were doing their job without a problem?

 

Any weight added for additional parts loading the driving wheels will inevitably put them (more?) over the limit.

As they had larger superheaters fitted and (likely) patches to the frames as well as double chimneys/blastpipes I can see no way they remained at 89T for the whole of their lives without removing mass from elsewhere, which seems unlikely.

 

The question is just how overweight were they?

However 25T on an axle seems a bit fanciful unless the suspension compensation was grossly maladjusted.

 

Keith

Edited September 20, 2017 by melmerby

[Miss Prism]

I suspect Chief Civil Engineers took a somewhat philosophical view toward some loco weight diagrams, knowing CMEs were prone to take an occasional liberty. CCEs were usually tolerant to slight overweights on axles - what they did not like was any increase in reciprocating mass. In that respect, 3- and 4-cylinder designs were better than 2-cylinder ones.

[LMS2968]

The loco weights as given to the public (and the CCE) were not always the same as those known the CME. One reason is that the static loading was only part of the equation; the dynamic loading, particularly on the driving axle, could be substantially higher. This was particularly so with two cylinder engines; multiple cylinders tended to balance out the reciprocating masses. Many CCEs were either unaware of this or refused to acknowledge it, hence C.J. Bowen Cook's having to reduce the weight of the Claughtons because E.C. Trench refused them due to their static axle loading. In this case, all cylinders drove on to the leading coupled wheelset so the reciprocating imbalance, and the need for hammerblow-inducing weights, was nil.

 

The first Stanier Pacific 6200 turned the scales at 109 tons 18 cwt against an estimated and published weight of 104 tons 10 cwt. Great care was taken that the CCE never discovered the true figure. 6202 was designed with the axle weights above the 22.5T limit; being a turbine, there were obviously no reciprocating masses, or balancing weights to cause hammerblow.

[OFFTHE RAILS]

The article I saw (unfortunately I cannot now remember where), plausibly claimed that the locos had gained weight due to bigger superheaters plus frame patches which were not there when the loco's published weight was originally given.

When new in 1927 this was 21.5T (bogie) + 22.5T + 22.5T + 22.5T = 89T which is the weight of the loco full, as built...............................

 

 

The information you are quoting probably came from the excellent book "Great Western Locomotive Design - A Critical Appreciation" by The Rev. John C. Gibson, AKC published by David & Charles.

I quote:

"There is little doubt that the Kings were significantly overweight. Oficially, the weight of the engine alone was always quoted as 89 tons in working order, with 22 tons 10cwt on each axle. These figures are suspiciously exact, especially as they were still quoted after a number of additions had been made, such as four row superheaters, strengthening pieces welded to the bogie frames and patches on the mainframes. The weight as first built is believed to have been around 93 tons, which could have grown to 94 or 95 tons in their old age.

Swindon seems to have been adept at 'cooking' the weights for the benefit of the civil engineer and the board....................................Many of the Welsh engines taken over at the Grouping were quoted as being three to five tons lighter, even after being fitted with  new boilers carrying higher pressure, enlarged bunkers and other features which would be expected to increase the weight. All rather suspicious.

 

John Gibson was an apprentice on the MSWJR, worked at Swindon in the 1920s and had a personal friendship with Harold Holcroft, (who had been) Churchwards assistant.

 

I seem to remember that there were discussions on the Southern Railway as to why the 2 cylinder N15 4-6-0s were heavier than the GWR 4 cylinder 4-6-0s and the reply was "perhaps the specific gravity of steel was greater at Swindon..."

 

Ian

 

 

[Miss Prism]

For more 'suspiciously exact' axle weights, see Bulleid pacific diagrams!

[peterfgf]

Even with the best endeavours and equipment, it is very difficult to achieve a measurement which is closer than +/- 5% uncertainty of the "real" value.  A figure for the axle loading is unlikely to be any closer to the "real" value than +/- 10% with the mechanical weighing machines used in the industry at the time.  [Note that the "real" value is never actually known - all you can do is get more and more approximations to this value].  Any locomotive weight diagram that gives measurements to one decimal place (one tenth of tonne) is suspect and the originators are deceiving themselves with a false sense of accuracy (uncertainty).    I believe wheel loads are now measured using strain gauges attached to the web of the rail which ought to achieve at least an uncertainty of better than +/- 5%.  I read that the A1 Steam Trust used this method to measure the wheel loads on Tornado but unfortunately I can't find any reference to it at the moment.

 

There is actually quite a lot behind your apparently simple question.  If you really want to know more, then I suggest reading Hayward, Repeatability & accuracy, published by the I.Mech.E. in 1977. The National Physical Laboratory has a Measurement Good Practice Guide No. 11 (Issue 2) Stephanie Bell, A Beginner's Guide to Uncertainty of Measurement, available as free download at http://www.npl.co.uk/publications/a-beginners-guide-to-uncertainty-of-measurement.

 

Peterfgf

[peterfgf]

I noticed the other day that the weights quoted on "Desiro" coaches are apparently accurate to the nearest gram - i.e. 52.329t

It appears that there is a distinct lack of good engineers at Siemens if they are letting stuff like that go out.  It means that they have absolutely no idea of what they are doing, why they are doing it or its relevance to anything.  

 

One further observation: 52 tonnes and 52.0 tonnes are not the same thing. The former implies that the value is to the nearest tonne, whilst the latter is to the nearest 0.1 tonne.

 

Peterfgf

[Trog]

I thought that you had to weigh loco's wheel by wheel to ensure an even weight distribution, adjusting the springs as required. As having a wheel taking very little weight because it has a soft spring while an adjacent wheel is overloaded by a hard spring is an invitation to derailment. 

[eastglosmog]

Another example is the Midland Singles, which had a published driving axle weight of 18t 2cwt and an actual weight of 19t 2cwt.  The LSWR Civil Engineer refused to accept the loan of a Midland siIngle for trials because of this and effectively scuppered Adams intended single (see Bradley's LSWR Locomotives, The Adams classes).

[Northroader]

The weighbridges at a steam loco works could be lined up to be under each wheel, and then transfer the weight from being supported on the tread by the rail, to the flange on the weighbeams, which you could adjust to find what load was being carried. You then went round adjusting the spring hanger nuts, a bit here , a bit there, until you got to near what you wanted. The hanger nut spanner was big, one guy was in the pit holding it steady on the nut, and another two guys heaving on a rope attached the the end of the spanner, standing outside the pit. These days the technology allows load cells to be used.

One of the main problems is that steam locos had leaf springs, with a lot of leaves. These deflected to smooth the ride, and the friction in the leafends damped the bounce out. If you applied a load you would get a deflection, and if the load was reduced the spring could hold at the same deflection, due to this friction binding the spring, and it would take quite a big reduction before the spring started to move, this being known as hysteresis. Consequently a lot of loading could be lost in the hysteresis effect, and you could adjust quite happily and the loading could be still out. One way to help correct this was do an adjustment, send the loco round the yard and bump over some rail crossings, then come back and try again. Not a precise science by any means.

[LMS2968]

The weighbridges at a steam loco works could be lined up to be under each wheel, and then transfer the weight from being supported on the tread by the rail, to the flange on the weighbeams, which you could adjust to find what load was being carried. You then went round adjusting the spring hanger nuts, a bit here , a bit there, until you got to near what you wanted. The hanger nut spanner was big, one guy was in the pit holding it steady on the nut, and another two guys heaving on a rope attached the the end of the spanner, standing outside the pit. These days the technology allows load cells to be used.

One of the main problems is that steam locos had leaf springs, with a lot of leaves. These deflected to smooth the ride, and the friction in the leafends damped the bounce out. If you applied a load you would get a deflection, and if the load was reduced the spring could hold at the same deflection, due to this friction binding the spring, and it would take quite a big reduction before the spring started to move, this being known as hysteresis. Consequently a lot of loading could be lost in the hysteresis effect, and you could adjust quite happily and the loading could be still out. One way to help correct this was do an adjustment, send the loco round the yard and bump over some rail crossings, then come back and try again. Not a precise science by any means.

It was common practice put pieces of steel on the rails as the loco was moved to shake up the springs and hopefully that they would then be relaxed by the time the second weighing was done. Crewe Works weighbridge, by the way, allowed for only ten wheels, so Pacifics had to be weighed in two goes anyway.

 

Weighing was not a popular job: it was hard and uncomfortable work, especially as Crewe's weighbridge was, for many years, outside in the open air.

[Siberian Snooper]

I have stood on one of the elements in the Swindon weigh shop and it was pretty close to what I thought my weight to be. This was when the DTG kept their 1000 in there, back in the eighties.

[melmerby]

One of the main problems is that steam locos had leaf springs, with a lot of leaves.

Not entirely.

Whilst the normal springing was leaves, coils or volute springs were not unknown on some axles, especially on engines with a set of driving wheels under the cab. eg. 57XX (but not 54XX), Dean Goods & 1366

 

Keith

[brack]

Trouble is, although loco weights were often conveniently undermeasured, nobody actually knew how much weight the line would take from a loco actually working over it due to the factors discussed above, which the CMEs often seem to have understood but the CCEs seem to have wilfully misunderstood on quite a few occasions - you'd think that understanding hammerblow and balancing would be a critical part of their job, and they were educated men. Often it seems they're just choosing to be awkward deliberately - see Newlands Vs Smith at Lochgorm. although the Rivers had a lower axle loading than the last Big Bens and weighed less than the 064Ts, Newlands wouldn't let them run and demanded they be sold at a time when the line was crying out for them. Sadly he had the ears of the board (I wonder if this was because he was a local boy from Elgin and had been with the company a dozen years longer than Smith, who was a Geordie), the highland made a profit on the rivers, the Caley got their only good 4-6-0s and poor Smith got sacked. Post grouping Newlands became CCE for the LMS and allowed the Rivers to run on the highland line without any trouble. The fact that Newlands and Smith were never on speaking terms seems the most pertinent detail and to my eyes makes the CCE's pettiness rather shameful. Perhaps other railways sided with their CME rather than the CCE when there were disputes? Or had better relationships between the two men.

[DCB]

According to K. Cook Swindon Steam      Following the Midgham derailment when a King Bogie was derailed  The Swindon weighbridge had a section 1" deep ground out with a removable section which was removed to check that each wheel carried very nearly the full weight even when a dropped an inch into a track irregularity. Coil springs were added to the bogie and the number of spring leaves increased and individual springs leaves made thinner and more flexible on the coupled wheels.   The sideplay on the trailing axle was also done away with and King riding was then exemplary.    I read when KGV was overhauled in preservation she came out under the quoted weights.

The increase in King weight from the new boilers with four row superheaters, double chimneys , new front end frames etc in the 1950s would have been at the front end over the bogie in any case.

There can be no doubt that locos with springs changed at running sheds in a rush to get locos back into traffic would sometimes have been way over the stated load, probably on one wheel, and the others underweight as a consequence.

[Bernard Lamb]

One further observation: 52 tonnes and 52.0 tonnes are not the same thing. The former implies that the value is to the nearest tonne, whilst the latter is to the nearest 0.1 tonne.

 

Peterfgf

Steady on, that is far too sensible an approach.

Such remarks will upset the vast majority of those who claim to model a gauge of 18.83mm.

Another observation. For those not involved with engineering or material specifications. Take a look at the tolerance on steel thickness. It will show a sizeable difference in the potential total weight if you take enough material to build a locomotive.

Bernard

[melmerby]

 

The increase in King weight from the new boilers with four row superheaters, double chimneys , new front end frames etc in the 1950s would have been at the front end over the bogie in any case.

 

IMHO that is a too simplistic view of weight distribution. It depends where the balancing point is which would usually be behind the front drivers, in which case any weight added would also increase the load on the front driving axle. That could then be distributed more evenly by adjusting the compensation. The King is quite clearly tail heavy as there are 45T behind the front driving axle and only 21.5T on the bogie (using published figures).

Most, if not all, 4-6-0 & similar locos would happily sit on the track without a front bogie without tipping forward, however the front driving axle will have much more load on it possibly damaging the springs etc. Certainly props would usually be used in this case.

 

See this A4:

https://c1.staticflickr.com/8/7148/6735903959_661a577cc0_b.jpg

http://www.davidheyscollection.com/userimages/0001-k-long-60023-Ferryhill.jpg

 

Keith

Edited September 21, 2017 by melmerby

[LMS2968]

Most, if not all, 4-6-0 & similar locos would happily sit on the track without a front bogie without tipping forward, however the front driving axle will have much more load on it possibly damaging the springs etc. Certainly props would usually be used in this case.

This was common practice, and I've seen several locos with the leading bogie or pony truck removed. Generally, packing is placed between the frames and the tops of the leading coupled axleboxes to maintain some height at the leading buffers to allow to loco to be moved from that end, otherwise the application of an extra 20-odd tons to the leading coupled wheelset would cause the boxes to bottom out against the frames, with the springs to go to full deflection, and a noticeable front end droop. In this condition, it would be moved slowly around the works or shed yards.

[Skinnylinny]

 

 

I noticed the other day that the weights quoted on "Desiro" coaches are apparently accurate to the nearest gram - i.e. 52.329t

It appears that there is a distinct lack of good engineers at Siemens if they are letting stuff like that go out.  It means that they have absolutely no idea of what they are doing, why they are doing it or its relevance to anything.  

 

One further observation: 52 tonnes and 52.0 tonnes are not the same thing. The former implies that the value is to the nearest tonne, whilst the latter is to the nearest 0.1 tonne.

 

Peterfgf

 

Um... Isn't a tonne 1000kg, rather than 1000g (=1kg)? Otherwise I've been buying sugar and flour by the tonne!

I would believe that modern electronic weighing facilities can probably measure that sort of weight to the nearest kilo.

[LMS2968]

Um... Isn't a tonne 1000kg, rather than 1000g (=1kg)? Otherwise I've been buying sugar and flour by the tonne!

I would believe that modern electronic weighing facilities can probably measure that sort of weight to the nearest kilo.

They possibly can, but it's rather pointless: even the weight of someone leaning on the side would alter it.

 

There's no point in weighing to that accuracy as its a transient figure.

[melmerby]

They possibly can, but it's rather pointless: even the weight of someone leaning on the side would alter it.

 

There's no point in weighing to that accuracy as its a transient figure.

The vinyl overlays would weigh more than a kilo!

 

Keith