After reviewing the development of railway sleepers in Southern Africa in Part 1 – the next in the series of three articles, compiled by Mr Robertson and published in 1957 – discusses the methods and economics of prolonging sleeper life and reviews some of the research carried out in this field in South Africa.
Note that this is reproduced so tone and language is identical to what was published in 1957.
The world shortage of timber led to a closer study of sleeper requirements, manner of deterioration and methods of preservation. Timber became something of value and not just timber to be cut from the surrounding forests. Every effort was made to prolong sleeper life.
The first improvement came with supervision and inspections.
Inspection Programmes
In order to eliminate hastily conceived and unnecessary resleepering schemes some sort of inspection “drill” is advisable. Good practice is to open up a set number of sleepers per week. If 40 per week are opened up about a mile of track will be covered in a year. If this is done on every mile of track a complete inspection is done annually.
The inspection report is best entered on a set form and the remarks of the inspector can be checked up from year to year. The report is best submitted in a simple form. For example, classify only three types:
- Those needing replacement at once;
- Those good for another year;
- Those good for longer.
If reports for the same section are compared from year-to-year results are sometimes most informative. General resleepering programmes are not economical. This is because good sleepers are removed and wasted. It is better to carry out “Spot renewals” certainly more economical.
The engineer must be careful that the general standard of the line is not dropping due to the falling off in conditions of the sleepers due to too much attention to economy. Economy can be overdone when dealing with sleeper maintenance.
Steel sleepers may be inspected in exactly the same way but the defects in steel sleepers are not always as apparent as those in wood when a superficial inspection is made.
Classification of Sleeper Defects
It is wise to try to catalogue sleeper defects. This has been done extensively in the United States of America. Here it was found that plate cutting and shattering were the more serious defects. In South Africa the following have been classified.
- Splitting of timber due to the drying out of resins by the hot sun, particularly noticeable in teak and pine sleepers.
- Breaking at the centre or at the rail seat. This is due to bad packing as a rule.
- Insect and fungus attack (dry rot).
- Plate cutting.
- Screw killing.
Steel Sleepers
- Corrosion due to locomotive ash or chemically active soils.
- Wear and corrosion in the slots.
Creosoting
It is found that if timber sleepers are creosoted, they last longer. Actually, a mixture of creosote and tar may be used. Railway sleepers should be treated with about 10 lb. of creosote mixture per cu. ft., but hardwood sleepers can absorb only 3 lb. per cu. ft. The full cell process is usually used. This is fully described by N. A. Richardson in the Proceedings of the Institution of Civil Engineers, Part 1, November 1953.
In this process the timber is first subjected to a vacuum and then the preservative is applied under pressure of 180 lb. per sq. in. Creosoting stops all forms of insect attack as well as fungoid attack. It also provides additional oil to the timber, thus preserving the side bond in the fibres and hence lessening the tendency to split.
Creosoting generally pays as the following calculation shows. If a loan of £1 is repaid in equal instalments over “n” years at a rate of interest “r” then each instalment
= r(1+r)n ———— (1+r)n-1
This is known as the Capital Recovery Factor.
If a sleeper costing £1 untreated lasts 10 years, then an amount of £0.1297 must be paid annually to pay the sleeper off in 10 years. If we can double its life by creosoting, then we pay back only £0.0802 annually for 20 years. Hence there is a saving of £0.0495 annually for 20 years. This represents a saving in capital cost of:
£ 0,0495 ———— = £ 0,61 or 12/3 0,0802
This figure of 12/3 is the justifiable expenditure on creosote treatment.
Plate Cutting and Screw Killing
Plate cutting is a symptom of age. It may be reduced by using larger soleplates or chairs, but as these are usually standard size there is little that can be done about it.
Screw killing can be very troublesome and the number of screw holes often determines whether a sleeper is serviceable or not. An excellent advance has been made in the use of compounds which close the existing holes and allow them to be used again without reboring.
Thus, loose coach screws can be inserted in the same holes which saves a considerable amount of time and often the life of the sleeper.
Ballasting
Much sleeper damage is caused by poor ballasting. Centre bound sleepers are very prevalent in narrow gauge track and break upward at the centre. Hand tamping or packing with beaters generally results in the edge of the sleeper becoming damaged and boat shaped.
The use of pneumatic tie tamping tools is recommended under certain circumstances, but their cost is not justified on little used lines.
Measured shovel packing using specially manufactured sighting boards and void meters reduces the amount of damage to sleepers by ordinary beater packing. In addition, the running top of the track is vastly improved.
Good ballast lengthens the life of all types of sleepers. It provides drainage and hence inhibits dry rot in wood sleepers and being chemically inert it prevents excessive corrosion in steel sleepers. In wet soils steel sleepers can be viciously attacked by electrolysis due to stray currents. Nevertheless, steel sleepers are used extensively in electrified tracks.
Steel sleepers are very susceptible to corrosion due to wet boiler ash which generally contains sulphates subjecting the steel to attack by sulphureous and sulphuric acid. Extra careful coating with tar products do not offer protection.
The retention of worn sleepers in the track places a great responsibility on the maintenance engineer, taxing his experience and judgment. It is advisable at this stage to examine what stresses are produced in railway sleepers.
Sleeper Stresses
The theory of elastic foundations was worked out by Timoshenko in the 1920’s and depends on the fact that upward track reaction is proportional to settlement. A load W carried on steel rail uniformly supported cause the track to deflect an amount y determined by the differential equation:
d(4)y EI ———- = – ky dx(4)
Where “x” is the distance measured along the track, “I” is the moment of inertia of the rail section and “k” is a constant known as the Track Modulus.
For normal 80 lb. track the load is spread in accordance with Fig. 3.
With 96 lb. rails the distance of spread is greater. As locomotive axles are spaced 5 feet apart the load may be considered to be uniform on the sleepers.
Hence with 20-ton axle loads placed 5 feet apart on sleepers spaced 21/2 ft. apart the load per sleeper will be 10 tons, or 5 tons on each side. Add to this the effects of lurching and hammer blow and the load on each side is about 61/4 tons.
A sleeper is thus loaded as indicated in Fig. 4. It is again subject to the general bending equation but this time the value of k is changed and may not be constant due to irregular packing of the track. If the upward reaction is constant, then the maximum bending moment occurs under the rail seat and for a 7 ft. x 10 in. x 5 in. sleeper the upward reaction is 33 lb./inch. Maximum bending moment
333 x 20(2) = ——————— 2 = 66,600 lb. in.
A 10 in. x 5 in. sleeper having a Section Modulus of 41.7 in 3 units will undergo a fibre stress of:
66,600 ———- = 1,600 lb./sq. in. 41,7
and a shear stress of:
330×20 ———- x 1,5 = 200 lb./sq. in. 50
Jarrah and most other hardwood sleepers will accommodate these stresses comfortably. This property can be tested quite easily by supporting the sleeper on a 5 ft. span. If it can support 8 tons centrally it has a factor of safety of 4 for a working stress of 1,600 lb./sq. in. The above calculations may be oversimplified but they represent very closely what happens in the track.
Research
In the Union of South Africa active research is being conducted to find the answers to many of the sleeper problems. This work is being done by the Research Branch of the S.A. Railways and by the National Building Research Institute, a branch of the Council of Scientific and Industrial Research. The main problem is to do with sleeper spacing. When we consider that the Canadian National Railways use 3,200 sleepers per mile and the British Railways 2,200 per mile there seems to be a real need for research here.
Some years ago, the number per mile in South Africa was increased to 2376, representing a capital outlay of about £2 million. A post-war piece of apparatus for testing sleepers and track fastenings is the “Vibrogir”. This consists of two rails to which a vibrating element is attached.
The vibrating element is driven by a strong electric motor and imposes a load of about 6 tons on each rail at the rate of 50 times a second. The running of a train over the rail would be represented by 2 seconds of working. A minute thus would represent a full day in a fairly busy main line. Sleepers, ballast, sleeper fastenings and rail fastenings can be tested in the vibrating machine.
By running the Vibrogir for 36 hours we have the equivalent of 20 years of service. The vibration can be stopped, and the progress of the test examined from time to time. Valuable research data can be collected, especially useful for new types of sleeper fastenings, and new types of sleepers such as plain, reinforced and pre-stressed concrete.
This will eliminate many of the long-drawn-out tests which are continually being carried out in the tracks of most railways. These test sections have to be examined for years and the results are often lost or lose their significance, going almost from one generation to the next. Suppliers of permanent way materials will be able to get quick decisions with this welcome addition in the field of research.
In the next article new types of sleepers will be discussed, especially pre-stressed concrete sleepers.
