The Hallade System


You have to remember that in the old days track maintenance was about the status quo. Prior to WWII and once order was again re-established maximum permissible speeds were 60 or may be 70 mph. Curve maintenance was a matter of setting out proper transitions and circular curves to give a comfortable ride, and then keeping them that way. I was interested to find that the method was devised by Émile Hallade of the French chemin de fer de l’Est in the early 1900’s, the English GNR was using it by the grouping in 1923. The LNER extended its use; the LMS issued instructions in 1934, which was the earliest that I had seen of it up to now.

You select a rail as the datum rail, usually a six-foot rail. Mark out and number every half chain with chalk, this distance between marks is a half chord. Using a taut piano wire a chain long, ie the chord length, measure the offset at the centre of the chord successively at each half-chord point. There were special offset tools to hold the wire and an offset brass ruler to do the measuring, which was all off the back of the rail. You realise that this required at least 3 people working bent down in the six-foot, it was faster if there was a fourth standing up and writing the numbers down; all this with just one lookoutman. No wonder that the nearest in my career that I came to be run down by a train was doing this at Langbank, and I was the inexperienced number writer.

Now the offset measured is the versine of the curve, the relationship is v =chord length squared divided by 8 times the radius, all in the same units. If the successive surveyed versines were tabulated a design version could be compared and by calculating successive differences it could be established if a design fits or not, and in doing so what the necessary slues are. It took a skilled eye and a lot of work with a pencil – as soon as computers became available they did it in moments. A circular curve has constant versines, a transition has versines increasing in uniform steps with a slight pockle, which some people worry about, at each end. As an example: versines: straight track 0 0 0 0 5 25 50 75 100 125 150 175 200 220 225 225 225 circular.

These could be sixteenths of an inch on a chord of a chain, or mms on a chord of 20m, which just happens to be almost the same as a chain.

When you had the design pegs were put in at each half-chord, taking account of the necessary slues with the fine position by a nail in the top. Now the nails were surveyed again to check that they were true to the design curve, and any corrections made. Everything was now ready to correct the track, which in those days required a big gang of men. First the sleeper ends were dug out, taking care not to disturb the pegs, then the track was slued with bars, Hey – Hup! The track was checked against the nails with the 3ft Hallade gauge, which you have. Between the pegs the line was by the supervisor’s eye. Only when it was all settled and everyone was happy were the monuments put in to replace the pegs, in those days the expectation was that they were there for ever.

I can recall two types of monument, there were no doubt different ones on each Region. You will have to forgive if my memories are incomplete. The newer type, which may have been a BR “standard” had a big square hat top with chamfered edges and a cross-shaped foot to bury deep in the ballast. All it had on its top was a steel strip on edge, the fine position was struck onto this with a cold chisel, the cant being marked on  plates on adjacent sleepers. The older type, which may have been LNER, was square but with the profile of a traffic cone so that it had a big base buried deep. Its top was quite small and if I recall aright there was a steel pin set in it, so that it had to be positioned very carefully when installing. It did however also have a lead plate set into the top, into which data such as the cant was punched. [Ask about any subject on the railway and it is always turns out to be much more complex than first thought.]

Monsieur H also invented his track recorder to monitor the success of all this work.

Soon after I started the emphasis changed, increased speeds were being sought and all the survey and design work went into those routes, along with the great leap forward of the fit and forget railway. The pegs would go in and the track would be slued to the new line, if you were lucky and the pegs lasted that long. More and more tractor shovels and JCBs were being used on trackwork, to get them to the site they had to run straddling a six-foot rail – bang go all the pegs. On track renewals the survey and scheme were done but it became usual to use the other six-foot rail as the datum and mark the slues on that. On-track machines improved, the lining machine came in, and then combined tampers and liners. The lining machine could do the design slueing for you, and put it to your marks, but they did something else that was considered to be wonderful – the greatest thing since sliced bread- they could “fine line”. They had two wires, a short and a long, with which they measured the track in the same way that the survey party had done. They compared the two readings and split the difference, thus they went along the track averaging out all the errors. So after doing your design lining for you they then ran through your job and fine lined it, leaving it ever so slightly not what you had intended. It might not have mattered except that District Engineers soon found out that the lining machine working on its own was very productive, and that it had an instant effect on improving the Neptune count of miles of track (and thus of course gaining Brownie points with the Chief.

[Neptune = North East Peak Trace Unit & Numerical Evaluator. The Matisa track recording trolley had replaced the Hallade recorder and York had devised this scoring system per furlong to fit to it, higher score= poorer track, plus yellow and red (paint) recorded exceedences. The GWR of course had a different device in a saloon which hurled paint at the track, I was in it once or twice; but I only once had a trip to Largs and back with a Hallade recorder, just for my benefit.]

So the lining machine, and tampers in lining mode were sent rushing out everywhere. But the trouble is that the more often you average the errors, the more you move the track away from its design. After a few years it was most noticeable how transition curves had got longer, particularly at the straight end where they now overshot. It took a long time for realisation to dawn that the machines are only effective if they are working to design rail heights and design alignment. [I gave an address to the Edinburgh PWI once in which I demonstrated that, of all the work done on maintenance by machines for about 8 years on the ECML, 40% had got worse, 45% was unchanged, and only 15% had got better. I was never asked to repeat it. I am sure that most track generally got better after privatisation because machines were no longer being sent out just to give men shifts.] The Edinburgh office had two Hallade gangers, who worked for the Maintenance Assistant. I don’t think they did any Hallade work as such but presumably assisted the machine supervisors in marking out where the transition curves were supposed to start and end [Actually they spent a lot of time delivering seedlings and plants, because one of them was a nurseryman].

Everything is changed now because the modern on-track machines use lasers and can do the surveys for you. No doubt they have software on board to improve on the survey as found but I think that they mostly work to the design given on the CD. On high speed and OHLE lines it is essential.

You are absolutely right to split your monuments in a wide six-foot, they might also overlap if it is sub-standard. In the 4ft on the N Bk Branch may be the LNER way.