The heaviest rail ever produced in the US, and most likely in the world is the 155PS. "PS" means Pennsylvania Standard. This section was designed by and used by the Pennsylvania Railroad. It was designed in 1946, but production had ceased before 1962. It is unlikely that there is any left in main track. The next heaviest was the 140PS also designed in 1946, which was later adopted by the American Railway Engineering Association (AREA) as a "recommended section" and its designation changed to 140RE. It was the primary section used by PRR, then PC, then in the early years of Conrail. Production went to under 50 track-miles per year after 1979 and ceased in 1992. About two years ago the AREMA proposed a new section, the 141AB which is essentially a modification of the 136RE with an imporved shape and slightly larger head. I have no information on who is using this or the quantities since it came in to being after the end of my collection of statistics. From the late 1940's in the east and mid 1950's in the west the primary mainline sections used were 136RE (originally 136CF&I) in the west, 132RE in the east with the exception of 133RE on the Union Pacific, 122CB on the C&O/B&O 127DY then 136NYC on the New York Central and 140RE on PRR/PC. Where they bought new rail for them at all, on medium traffic lines most roads used 115RE but some western roads used 119RE (originally 119CF&I). Most rails lighter than 100 pounds per yard have been produced in negligible quantities for a number of years.

Quantities are in track-miles. Sections for which the 35 year total was less then 200 track miles are not listed. Numbers are rounded to the lesser of 100 miles or two significant figures.

Section----Total 62-96----per year 94-96
140 RE--------4,700--------------0-
136 RE-------35,900-----------1,990
136 NYC---------510--------------0-
133 RE-------10,900-------------380
132 RE-------37,300-------------140
130 RE-HF-------210--------------0-
122 CB--------3,400--------------0-
119 RE--------5,700---------------7
115 RE-------29,200-------------475
100 RE--------2,300--------------21
100 RE-HF-------840--------------0-
100 ARA-A-----6,200--------------23
100 ARA-B-------390*-------------35
100 ASCE--------220--------------0-
90 ARA-A-----1,500--------------0-
85 ASCE--------540--------------0-
85 CP----------520--------------0-
80 ASCE--------280--------------0-
(some sections are not listed, therefore sum is greater that the list of items)
Total all---142,000-----------3,100
*This quantity is significantly under-reported since this is the proimary sections used by NYCTA and some other northeast transit systems, and on the only time I was there, I saw a pile of 100ARA-B rail fresh out of the mill at Steelton in 1988 that amounted to about 100 tons.

The average weight of all rails for 1996 was 131.8 pounds per yard.

As to the statements on rail temperature: There seems to be a little misunderstanding here. Theere are a number of issues involved in rail life. Shape, support, hardness, brittleness, grinding program, control of wheel flats, etc. Where temperature comes in is the issue of setting the zero-stress temperature. In other words, laying the rail so that a particular temperature the rail is in neither tension nor compression due to thermal effects.

If this temperature is set too low, the rail may buckle due to high compressive foreces when the rail gets hot. If it is set too high, there may be pull-aparts at welds when the rail is cold. A pull apart can be less dangerous and is more easily detected in signalled track, as it breaks the track circuit. A buckle is usually only found in operation, and can result in derailment. Since a long steel rail is almost like a rubber band, keeping the rail in tension most of the time is desirable. That has become the normal American practice. It requires good steel, good quality rails, and storng insulated joints to work. The temperature to consider is rail temperature, not air temperature. While rail temperature will not go below air temperature, it can go above it, particularly in desert climates. Extensive calculations of stresses in rails and buckling forces have been done to detemine the reasonable range of zero stress temperature. Much of this is to counter that ingrained part of engineering education which says you must allow for thermal expansion and contraction of steel. Actually, in rail your basis is that the rail will not move and temperature effects are turned into internal forces. It has been the general conclusion in the US (and Canada and Mexico) that the zero stress temperature should be toward the high end of the temperature range and a formula of (2H+L)/3 + C is normally used, where "H" is the normal high rail temerature and "L" is the normal low rail temperatue, and "C" is some added constant, which is usually the Chief Engineer's opinion, and is normally in the range of 15 to 25 degrees F.

The Europeans, which generally use lower strength rail steel and smaller rail sections, tend to keep their zero stress temperatures lower and then go to great lengths to provide huge ballast shoulders to prevent buckling. By so doing, they appear to be wasting a lot of money for no appreciable benefit.

Until recently the British standard was 113 lb/yd and was called 113A. I understand that they have recently adopted the UIC60 section, which makes no sense at all to me because it has a 150 mm wide base compared to the BS113A's 5.5 inch = 139.7 mm wide base. One thing for certain, if you have a system that is for the most part in concrete ties, the last thing you want to do is change the base width of your rail section, because the seats for the rails and attaching points for the rail clips are built into the tie. BS113A is a thick web version of their former main line section, which was BS110, which is essentially identical with UIC54. This rail is very similar to the 112RE.

A lot of the published information is done in such a way as to make comparison difficult. For example, in North America we usually quote freight cars by their carrying capacity, while the Europeans quote them by gross weight. That is, when we talk about a 100 ton car, we mean a freight car with a capacity of 100 tons, which usually means a gross weight of 263,000 pounds = 119.3 tonnes (metric tons of 1000 kg each), while when a European railroader refers to a 100 tonne wagon, he means a car of that gross weight, which would be 220,000 pounds, which was usually the allowable gross weight of what we would refer to as a "70 ton" car. Further, only Britian and a few specialized lines allow loads that heavy. For most of continental Europe outside the former Soviet Union five foot gauge lines, the axle load limit is 22.5 tonnes = 49,600lbs or 198,400 lbs for a four axle car, and most axle loads, including engines are less than that.

Rail: American Rail is specified primarily by metallurical limits and hardness. European, and most others, rail is specificed primarily by metallurgical limits and tensile striegth.

The Metals Handbook by the American Society for Metals provides information on the relationship between hardness and tensile strength for rolled steel sections.

Without going too deep, here are some of the comparisons:

Factor----------AREMA----BS11-WRA--UIC/EN-900A
Carbon, min------0.74%----0.65%-----0.60%
Carbon, max------0.84%----0.80%-----0.80%
Harness, brinell hardness number
BHN, min----------300------263*------263*
Tensile strength in Newton/mm^2, then converted to 1000 pounds per sq inch.
Tensile N--------1000#-----880-------880
Tensile ksi-------145#-----128-------128
Elongation @ break:-9%------8%-------10%
*from Metals Handbook for defined tensile strength
#from Metals Handbook for defined hardness

NOte also that the European definition is for ultimate tensile strength, which is significantly higher than the yield strength more commonly used in the US to define steel strength. If loads stay below yield strength, the material life under repeated load is essentially infinite.

This should make it obvious that rail to the American specification is stronger. It has to be. Our allowed axle loads are much highe than those in Europe. The European and British rails in these comparisons is their middle grade, not their lowest grade.

Shape: This is too complex to go into in any detail, but there are two parts, wheel rail contact pattern and internal stress distribution. A couple of points: The European sections all have relatively small head to web radii, and this is particularly true for the UIC60 section (120 lb/yd) which is the current premier section that they market arould the world as the world class international rail section. Increase in this radius was the main factor in the redesign of the 112RE and 131RE into the 115RE and 132RE sections, respectively. While I do not know of any instances where this small head-web radius has proven to be a problem with the UIC60, there are virtually no installations, if not simply no installation of this rail where the axle loads are heavy enough for it to likely be a problem. Wheel rail contact appeas also to be better with the American 10 inch=254 mm crown radius instead of teh 300 mm crown radius of the UIC sections. In fact, the current analysis suggest that even a 10 inch crown radius is probably too large.

The Russians use a rail sections that is approximately equivalent to the 132RE that they call the P65. Remeber that in the Russian alphabet, Russia starts with a P. I do not know anything about its metallurgy and strength, either theoretically or in practice, nor about the allowed and normal loads carried by the Russian system.

The 136RE or a slightly modified metric derivitive thereof is use in heavy haul lines in Australia and, I think, Brazil.

The main reason that European tracks tend to look better and ride better can be summed up in one word: Money. The European systems spend money on track maintenance and upgrades that is in the level of died and gone to heaven for an American track engineer. But these systems do not pay taxes or make a profit, instead they consume huge amounts of government money. Given the materials and techniques developed over the years to get the best bang for your buck and a track budget per passing ton of somewhere over half that spent in Europe, we could have the best track in the world in a few years.

After all, you can't run 186mph trains on shoddy track. Okay, so a Eurostar isn't as heavy as a double stack container or tanker, but axle weight is not the only factor in rail design.

Just out of interest, any idea why US rails are spiked whereas the UK (and possibly Europe) tend to use Pandrol clips?

141AB rail is now standard for main line installations on UP and BNSF, and others are beginning to use it as well. I saw my first 141 on a former Conrail line near Albany, NY in the fall of 2002. It, like 136RE, is essentially a "fattening up" of the 132RE section. Base and "fishing" (web) are the same as 132, with more metal in the head.

A note on elastic fasteners: they are necessary on concrete ties, which are much more common in Europe than North America.

The following may be getting a little too technical for this venue.

A comment: Since rail design is more experimental than theoretical, generally rail shapes are changed as a result of problems with existing shapes.

Another comment, primarily to goeffm, yes I do think that American rail in both shape and other factors is better, in fact head and shoulders above, that of Europe. And that is mainly because American rail is worked a lot harder so more problems present themselves on the one hand, and because everyone who works with it has their own ideas and with multiple private companies have had more chances to try them out, and therefore there has been much more experimentation with various shapes in the US and Canada. For other factors in track, more later.

To rresor: Rail head radius: It appears that you may have misunderstand the point I was trying to make.

First: radius at the bottom of the head. The European problem as far as load capacity is the radius at the bottom of the head where it meets the web of the rail. This is the primary factor behind the 112lb section being redesigned into the 115 and the 131 into the 132. If this radius is too small it results in high internal stresses at the intersection of the head and web. This was discovered due to rail failures in this area in worn rail in the later part of World War II. This radius was increased from 1/2 inch to 3/4 inch, Problem solved, even though average and maximum axle loads have gone up significantly since. In the UIC60, the radius at this location is 7 mm. The head depth is greater, the bottom slope greater, and rails are pulled out of track with much less wear than in the US, so they have as yet had no problem. However, it will bear watching in the high speed lines as the rails wear, as wheel rail impact is a V^2 function, so a 300 km/h light axle passenger train may be as hard on rail as a 50 mph coal train. As yet the European high speed lines have not had a high amount of passing tonnage on their rails. The Japanese, which have had a large volume of tonnage on their high speed lines use a rail that has a large head to web radius.

Now: radius at the top of the head. Here we have the wheel contact point. The other major change in the 1946 section redesign was to change the crown radius from 14 inches to 10 inches. Not everyone was convinced, as the CF&I designs, 136, 119, and 106, kept the 14 inch crown, but with a larger gauge corner radius. I think these sections were developed sometime in the mid 1950's but do not know exactly. Can somebody tell me when these sections were first introduced and first used? First a comment, there are actually usually three radii on the rail head: A central large radius, with a width of about 1.25 to 1.4 inches, second a shoulder radius on each side of 1.25 inches, and finally a corner radius from this second radius to the side which is 9/16 inch on the 136 and 119 and 3/8 inch on 115, 132, 133, 140. As noted, since rail shape design is more experimentally modified than theoretically modified, serious re-analysis of head shapes began with the advent of grinding programs to try to extend the life of rail as more and more traffic became to be concentrated on less and less track. As a result, the 136 pound rail was produced for a while with four different crown shapes. There was the original with the 14 inch crown radius; a 10 inch crown radius version, which was later became the standard without any change in section name; an 8 inch crown version developed in Canada, first used by British Columbia and then adopted by Canadian Pacific; and a 4 inch crown version ordered by CN for use on curves only. The new 141AB has an 8 inch crown radius, so even though it seems intuitively illogical, it appears that a smaller crown radius actually performs better under heavy axle loads. This may be true virtually regardless of axle load, as use of a 200 mm (almost exactly 8 inches) crown was found to be the optimum in at least one high volume but low axle load passenger line.

As the man said, I am not sure that I understand all that I know about this subject.

To goeffm: High Speed Track: While a good wheel rail contact path is necessary for high speed, it is also very important and ordinary railroad speeds. What is necessary at high speeds is a good quality alignment both macro and micro. by macro, I mean very larger radius curves, use of variable rate spirals, large vertical curves. By micro I mean careful attention to deviations from line, level, and cross level in the track. A very wide range of track forms can be and are used. You will usually find that each developer swears that their own is best and all others are inferior. Such items as specific components and even track gauge are not really that important as to the possibility of running high speed. Proper component selection can have a lot to do with the maintainability of the system. A lot of the same considerations apply wheter you are running 50 mph or 200 mph. There are some dynamic factors that come into play at higher speeds, but it does not appear that even when the French got to 515 km/h (320 mph) that these were coming into play. They did not go faster simply because thay did not have enough track length to speed up and slow down. By the same token, when riding the Japanese Shinkansen at 285 km/h (177 mph), there is no sense that they are in any way pushing the envelope of normal wheels on rails operation.

I am not down on all things European, it is just that having studied and used some of their materials, I will say that a lot of it looks better at a distance than it does up close. Again, it is money. Economics appears to be much less of a factor in their decision making processes. They spend a lot more on providing a good passenger service and on maintenance of their railways in general. The German turnout geometic concepts and design theories for high speed turnouts are very good but some of their applications of those designs leave you scratching your head. You could take their theories and better their current designs. A lot of their components are somewhat high-maintenance, and a lot of the details of rounding and chamfering that we use they do not. Generaly, I find European turnout designs better geometrically, but more complex in componentry and worse in durability, even theough they do have some very good components. What we want is to pick the items, not buy the store.

Elastic fastenings: That is one excellent deveopment from Europe. Pandrol had the right idea, easy to install and remove, and no bolts to keep tight. Some of these come in two versions: a North American version and an everywhere else version that is not as strong. In Europe they use elastic fastenings even on wood ties because they have a philosophy that spikes into wood are not safe except on slow industrial trackage. When they use elastic fastenings on wood, it is usually with a cast baseplate that is held to the tie with four large lagscrews with the clips each attached to the plate by another bolt, unless the clip is Pandrol. Needless to say it is expensive in both material and labor.

to boyishcolt: Best track: This is like asking the doctor for free medical advice. For starts, good drainage and good subgrade material. In fact the first three rules of good road or railroad design are drainage, drainage, drainage. One man years ago described a lot of what we do in track upgrading and maintenance as trying to build a two story brick house on a tarpaper shack foundation. If you do not have good material under the track nothing you do will stay good. Do your best superhighway type earthwork and have a good ditch on each side, then put on about 8 inches of good subballast about 24 to 28 feet wide and with a good crown plus about a 4 inch layer of a good asphalt base course about 12 feet wide directly under where the ballast will be. On this have at least 12 inches of ballast under the bottom of the ties. If the track is relatively straight, traffic density and speeds are normal range, then the wood tie and spike is as good as any unless the climate is Florida or south Lousisana. Use two rail holding and two plate holding spikes. Have ballast shoulders of 12 to 15 inches level with the top of the ties. Use large tie plates, 141AB rail if you will have heavy freight. Now if we are talking crooked, go to concrete ties with Pandrol Fastclips. Concrete is also a good idea if you want to run 100 mph plus or are building in high rot areas. In general, you should not be using concrete ties if wood ties will perform well. I know this is not the current revealed wisdom, but concrete is much less forgiving of poor support, high impacts from flats or out of round wheels and othes such surprises than concrete. Of course 100% welded rail, glued insulated joints.

1. In your opnion US rail is better than European rail - but you also say that US rail takes more of a pounding. So what I believe you really mean is that US rail is *more suited to US conditions* whereas European rail is *more suited to European conditions*. Why use over-engineered rails for lightly used branch lines that see half a dozen light carriages a day?

2. "Such items as specific components and even track gauge are not really that important as to the possibility of running high speed." - I think you'll find guage is extremely important. Experiments with the TGV trials proved this. With a slightly wider gauge, severe hunting was experienced which resulted in deformed track. I *think* high speed lines are more 1432mm instead of 1435mm. Hazy memory there and I'm no expert.

your point 1: I am talking about shape, hardness, metallurgy in rail in main line srevice in both cases. For branch lines, you use whatever is left over on both sides of the Atlantic.

your point 2. I was not talking about the relation between track gauge and wheel gauge, which can be very important in wear and ride quality issues, but gauge as an absolute value to say that there is nothing "magic" about standard gauge per se. If you want to run high speed service you can whether your gauge is 1435, 1524, 1600, 1067, 1000, or some other number. (4ft 8.5in, 5ft, 5ft 3in, 3ft 6in, 3ft 3 3/8in, naming some of the more common track gauges used around the world)

The UK loading gauge is smaller than the US ones, smaller than even Eastern US, hence smaller trains. Newer lines are built to a larger loading gauge. I can't remember the standard but basically so European freight traffic will fit along key routes.