There doesn't seem to be a US-wide forum, so I'll have to post here I'm afraid! Another thread got me thinking.
Can somebody tell me what the pros and cons are of staggering rail joints on jointed track? European style is to have the joints parallel to each other; US style seems to be to stagger them. From a comfort point of view I tend to feel a little queasy with staggered joints due to the constant swaying. Beyond that I can't think of any pros or cons of either method.
How is US track tamped? Why are the rails spiked to the ties instead of using removeable clips? Tamping machines are common in the UK - these travel along the track to be tamped, unclip the rails, slide out the ties, either filter or replace the ballast, slide the ties back in (or replace with new), clip them back in, all automatically at walking pace. A further machine comes along and pounds the track, simulating the passage of dozens of trains, allowing the track to be restored to normal line speed right away. I'm sure these must exist in the US but I've yet to see one!
Thanks
Geoff M.
Posted by RRRICH (Member # 1418) on :
This sounds like a good question for George Harris......... come in, George!!!
Posted by Geoff M (Member # 153) on :
Interesting - thanks. Although the total force wouldn't be any different, whether parallel or staggered, the fact that both happen at the same time is what counts, I guess.
Geoff M.
Posted by George Harris (Member # 2077) on :
Geoff, I could just about write a book on this question. RussM's answer is like a lot of internet stuff, a light once over partially correct but with some erroneous information.
I will try to be brief. Track design is my main thing for the last 30 odd years.
Axle load really has nothing to do with it. Rocking of the cars is primarily at low speed as at higher speeds the car body does not really have time to fully follow the up and down motion of the wheels going across low joints.
In the very early days everybody used squared off joints. I have tried to find when staggered joints became the practice in the US, and so far all I can say it is before 1900 and the early 1900's writings I have found treated it like a long closed issue. Remember that a lot of early US and Canadian railroad mileage was built into relatively undeveloped areas, built cheap with lots of curves. To deal with the constant need to slightly shorten the inside rail to keep the joints squared up is a real nuisance in that instance. I have a feeling that at some point, and probably at several points independently of each other the track builders said in essence, forget it, or more likely some other expression that would turn into a series of asterisks here, and decided to not try to keep the joints square. They found this actually worked better.
Here are the advantages of staggered joints:
The line of the track stays better on curves. Since joint bars are less stiff laterally than the rail, on curves with square joints the tendency of the track is to take on the form of a series of straight chords with angles at the joints instead of staying in a smooth arc. I have seen this in the flesh. In addition, when this happens the rail develops an uneven side wear pattern with deeper side war in the middle of each rail length getting lesser toward the ends. This can be asymmetric if the track in normally used in one direction only.
The end batter and dipping occurs more slowly if the joints are staggered. This is because only one wheel at a time on an axle goes across the joint so it actually unloads slightly due to continuous support under the other wheel. The car body will also drop less, normally. Think of the difference when driving your car between hitting a pothole with one wheel versus opposite holes or a ditch of the same size with both wheels at once. You will have a much lesser jar with one wheel only dropping instead of both at once.
With wood, the ties (sleepers since you are in the UK) under or adjacent to the joint will last longer with staggered joints than they will with squared joints. There is usually not the need to use special larger joint ties (sleepers) as is commonly done with squared joints.
A pair of insulated joints that are staggered will outlast a pair of insulated joints that are squared off, generally by quite a lot. Have no specifics, but I would think around 50% longer. Properly, insulated joints should be staggered two tie (sleeper) spaces so that there is no commonality of support. Note that when setting this stagger, the stagger should not match or be close to the axle spacing in the bogie. With a normal US 19 to 21 inch tie spacing, this makes the stagger 38 to 42 inches, son one meter is a nice number for those in the metric world. For the sake of machine tamping, I do not like to insert random odd tie spacings.
The common American practice was to have 39 foot long rails and stagger the joints ½ rail length. It might have been smarter to have staggered them 1/3 the rail length, instead. Current mill practice is to roll rail in much longer lengths, usually 85 feet, although some like 78 as if you set things up based on 39 foot lengths this is easier to deal with.
These long rails are obviously now all welded up, which gets us to another point. It is also a good idea to avoid having the welds squared, as well, particularly the thermite welds. No particular value for this, just not opposite and not at the axle spacing in the bogies. I rode on a particular transit system that was put together with squared up thermite welds a few years ago, and you could feel the welds almost like joints. It is impossible to completely avoid hardness variations in the heat affected zone of the weld, but their contract did a less than outstanding job in this area.
I have been assured with an absolute straight face by a German track engineer that staggered joints cause derailments. Obviously not true. I have tried hard in the 15 years I have worked overseas to understand what the advantages are to using squared joints, and am still coming up dry as far as operations and maintenance are concerned.
The only advantage I can see to squared joints is that of being able to easily lay the track in panels. This also gets us to the error in the internet response. Heavy axle loads will not automatically give you broken joints if they are square. I know of a siding extension near where I am from in the US that was laid with panels that have been in place with the joints still squared for now nearly 20 years and so far no apparent joint failures. It is a 25 mph track, and sees something like 4 to 6 hundred car freights a day.
Heavy rock and roll with staggered joints occurs when the trains are run so that the spring constant of the suspension is near the time it takes to cover the distance between joints. This is particularly a problem if the bogie centers (truck centers to the US side) is about the same as the joint spacing. For 50 foot cars on 39 foot rails this was close to the case, and a speed in the 12 to 18 mph range got you the harmonic speed. That is why most 15 mph speed limits in sidings disappeared and went to 10 mph if they could not be, as was frequently true, raised to 25 mph. Squaring the joints does not cure this. It gives you a harmonic bounce instead, which is, if anything, worse.
Railroads are a system. Anyone not understanding this will frequently learn it the hard way. That is one large reason why divorcing track operators and train operators is a bad idea. Everything on the North American side of the pond is based on the system we use. While American equipment, assuming size and axle loadings permit, could probably operate on European track without difficulty, I doubt that the reverse is true. Our bogies are built to tolerate a lot of differential movement due to variations in cross level. The fully framed freight bogies I have seen on European equipment likely would not take so well to American track.
George
Posted by RRCHINA (Member # 1514) on :
Mr. Harris has , again, provided engineering expertise tempered with the judgement obtained with much experience. We are fortunate to have his participation at this site.
Posted by Geoff M (Member # 153) on :
I had hoped George would answer this. Thanks.
Now you mention it, extremely sharp curves with parallel joints do sometimes look jagged at the joints, ie not a smooth curve.
I think insulated joints are less of a problem these days as we tend to use jointless track circuits - or, dare I admit it, axle counters instead.
I'm not sure why separating train operators and track mainainters is bad. After all, all new trains have to meet standards and have to be fully tested in overnight possessions to ensure they don't interfere with TC operation or signalling equipment.
I agree that European stock probably wouldn't last long on American track. But like I've said before, the trains are built for the track standards upon which they run. Designing it for any other system is a costly over-engineering exercise.
Any ideas on the tamping machines and spiking?
Geoff M.
Posted by MILW (Member # 2538) on :
Isn't the weight per axle a lot more on American Frieght trains then on European freight trains and if that is true, wouldn't running an average American Frieght train on European track cause problems with the European track.
Also, I believe that American Passenger trains because they share tracks with frieghts are built a lot higher buffer and strength standard than European versions, which might make them heavier as well.
I used to live in Europe and I noticed while living there that across the board the American standards on safety are higher overall. I noticed that with both Automobiles and Trains. Not sure as to the reason.
Posted by Mr. Toy (Member # 311) on :
Once again Mr. Harris gives us a technical lesson in easy to understand terms. Though the technical aspects of railroads are not my main concern, I read his entire explanation, and learned something interesting today.
quote:Originally posted by MILW: I used to live in Europe and I noticed while living there that across the board the American standards on safety are higher overall. I noticed that with both Automobiles and Trains. Not sure as to the reason.
Perhaps we have a lower tolerance for death and injury? Actually, I believe it was the Europeans who pioneered many of the safety features we now take for granted in automobiles, such as rounded fenders (which do less damage to humans hit by cars), side impact beams, and crumple zones. But I don't think they have as high standards for their trains, which takes me back to my original answer.
Posted by George Harris (Member # 2077) on :
Taking the parts I left off in Geoff's question first: Tamping in the US is virtually always by machine and has been for at least the last 30 plus years.
Use of spikes and tieplates on wood ties (dogspikes and rolled baseplates on wood sleepres in English English) continues to be the majority of trackage in the US for the simple reason that it works well. It is the Rodney Dangerfield of the railroad world. It never gets the respect it deserves. It is simple, cheap, and reliable. Good quality hardwoods are much cheaper in the US than in Europe. Southern Railway did a 30 year experiment on treatment processes that ended in the early 90's in which they found that an improved treatment could make the wood last well beyond the 30 years normal in their territory.
Many high volume tracks are having concrete ties installed, but only where wood tie life is shortened because of either heavy traffic or not so heavy traffic on lines with heavy curvature. Only in these conditions do concrete ties make economic sense under US conditions. You will see quite a bit of track where the straight parts stay in wood and the curves have concrete. You will also notice that it is the railroads such as the Florida East Coast in the high humidity warm climates of the gulf coast areas use concrete everywhere. Florida termites use chain saws. These lines also tend to be further from the source of good hardwoods.
The European wood tie fastening systems that I have seen seem excessively complex and expensive. (Screw spikes, cast base plates, bolted rail clips) Use on wood ties of rolled baseplates with screw spikes and pandrol clips is freqently found in turnouts in the US, but these are rolled baseplates, not castings.
There are a couple of things that boggle my brain completely about European track. First, why do you/they persist in keeping the end bolt holes in rails so close to the end? It used to be in the US that the standard was to have the end hole 2.5 inches from the end of the rail. Back in the late 1940's it was recommended that this be increased to 3.5 inches. Since that has become standard, rail end brakes on these rails is very rare. Yet, when I see the British standard for fishplate (joint bar on US side) holes, they are even closer to the rail ends (2.25 inch I think it is) than the number that was decided to be too close to the end in the US over 50 years ago. Why? I read all kinds of things done to prevent rail end breaks in Europe such as cold expansion of the hole and also the development of the thicker web BS113A rail to replace the BS110A, when all that was really necessary was to move the hole back. This is also not to mention that one of the significant tidbits they hit us with in the one Steel Design course I had to take was that structural bolts should be set so that there would be about 3 bolt diameters width of metal between the bolt and the end of the section. This is what you get exactly with the one inch bolts used in 115RE rail, and you get almost that much with the 1.25 inch damter bolts in 132RE and 136RE rails.
George
Posted by Geoff M (Member # 153) on :
Firstly, saying that American railroads are "safer" than European is a broad, sweeping statement that doesn't take the facts into account. For example, signalling (my daytime job) is far safer in Europe than in the US from the knowledge I've gained both in UK signalling centres and in US dispatching centers, not to mention from my own IRSE qualifications.
European passenger trains share the rails with freight trains, just like in the US. And, just like in the US, passenger vehicles are built to standards that take that fact into account.
One thing I have learnt and now concede is that bridges over US rails are safer than in Europe.
Sleepers/ties: we use whatever is appropriate, same as you - hardwood, concrete, or steel. I believe recycled polywhatever is being trialled in certain areas. But I believe that hardwood sleepers are the most common in the UK.
I'm surprised about the statement that spikes last a long time. I had heard (unsubstantiated) that removing and replacing them can only happen a few times before new holes need to be drilled.
Jointed rails / fishplates: I'm not an expert but what I do know is that we don't have a problem with this. Again, lighter axle loads may be a factor. Mind you, welded rail is more normal these days anyway.
So, yes there are different ways we do things. But saying "US is safer than the UK" or vice versa is nonsense. Both have their pros, both have their cons. We learn from each other.
Geoff M.
Posted by George Harris (Member # 2077) on :
Geoff, I am not going to get in an argurement with you on signalling, but really can't see the basis for your statement. I would say on both sides of the Atlantic, both signals and track properly installed and maintained are as safe as it is humanly possible to make something safe.
European track generally looks better because, 1. more attention is paid to some of the more cosmetic aspects, and 2. low speed tracks appear to be grossly overbuilt. A really horrible looking track can be quite safe at low speeds.
Admittedly you are there and I am not, but I was of the impression that installation of new sleepers had been almost 100% in concrete for quite a few years now, and virtually all pictures I have seen of UK tracks that are of recent vintage show concrete sleepers.
One thing that I realize makes US signal and train control practices look unsafe to European eyes is that US practices are much more dependent upon the integrity and correct action of the train crew. There is much less in the way of ATS and other forms of "big brother" hardware, and there are still a lot of lines, including at least one currently used by Amtrak that have no signals at all, but depend upon track warrants and manual block procedures for safe separation of trains.
Generally when a tie is considered "spike killed" it is far from new and spikes have been redrived quite a few times. To avoid this problem is part of the reason that tie plates normally are opunched with 8 holes but at the most you only need four. Secondly, there are wood tie plug that look a lot like a spike shank, except for being wood. When a sopike has to be redriven into a previously used hole, you are supposed to put one of these in first. There is also a liquid product that can be inserted into a previously used hole with a caulking gun to improve the holding of a redriven spike.
I realize that there is not much of a problem NOW in Europe with fishplates / boltholes, but there was in the past. The US side solution of moving the hole position for newly purchased joint bars and rails beginning 50 years ago was essentially a no-cost solution. The European side solution of keep things as-is but do special treatmetns of the hole and in case of the BS113A actually redesgn the rail cross section was definitely not a low/no cost solution. On both sides of the ocean, use of welded rail in almost all tracks has essentially relegated the problem to historical curiosity. Also, use of glued joints for insulated joints has eliminated the problem there, where it was the worst.
George
Posted by Geoff M (Member # 153) on :
Agreed, don't want to get into an argument, but I will stand my ground if factually incorrect statements are made by anyone.
Yes, the US do rely more on humans than equipment, and those humans are trained to rely on their own behaviour more than the equipment, but at the end of the day humans are not perfect. That is why most accidents are caused by human error, not equipment malfunction. If you can reduce the chances of a human causing a disaster by improving the safety equipment, then that system is safer than one which relies more on humans. That is the basis of my point that European signalling is safer than US signalling.
But that was only one aspect of rail systems and, as I said before, I'm sure the US will "win" on some points and the EU will win on other points.
Concrete sleepers are more common for new track but there are still a huge number of wooden sleepers left.
Just out of interest, and I have no opinion on this, what do you think of the French LGV sleepers which are two concrete blocks joined with a steel bar, rather than one long block?
Geoff M.
Posted by George Harris (Member # 2077) on :
I have no real opinion on the French two-block sleepers, because I have no first hand experience with them. These things are the French standard for all their lines, not just the LGV. The Mexicans were big users of the French sleepers for quite a few years, but no longer do so. They now use monoblock sleepers that are, I have heard but am not certain of it, based on a German design. There was a fairly negative article about them by one of the top level Mexican track guys that was published in an AREA bulletin several years back, which was followed sometime later by a somewhat haughty French rebuttal that made a lot of disparaging remarks about the quality of the Mexican work.
I will say that they have been used for a long time with very little change in design, which usually means something that is fairly good, but then the French are probably about the worst at admitting that they designed something that did not turn out right.
The "RS" designation for these things is for the designers name, Roger Sonneville. He must have been a fairly young man at the time he did this design, as so far as I know he is still living, though quite elderly.
Mr. Sonneville is also the designer of the version used in concrete slab track called LVT track. Effectively the current version of this design puts a rubber boot around concrete block with a 20 mm thick rubber pad under the bottom. When I first saw this quite a few years ago, I was a skeptic, but it seems to be performing wonderfully well. This is the track form used in Eurotunnel. The first version retained the tie bar between blocks but that has proven to be unnecessary. This old design is still being sold under the name Stedef.
The patent for the current design (LVT) is held by Sonneville himself, or his company, and he and his business are now based in the US. I have met the elder Sonneville several years ago, and seen his son who now runs the business several times.
The LVT system is used in some of the Taiwan High Speed, but only in a few km of tunnels where the speed is under 160 km/h, for reasons I can not go into in public.
Your point on preferring mechanical overrides is well taken and normally correct, however, there are situations where nothing mechanical can substitute for good judgement, particularly if the programming of the control system makes a bad decision. Several years ago there was a fatal accident on Washington metro becuase the train operator was refused permission to go manual, and the programming made a fatal decision. I will look up the accident reference later.
George
Posted by Geoff M (Member # 153) on :
Interesting. Presumably the Taiwan reference is to the Trupo tunnel.
The Washington metro incident sounds nasty, but you're right - garbage in, garbage out as it's still humans designing the thing. Ought to be picked up by the internal testing, independant testing, safety reviews, and product acceptance but there you go. Presumably (without knowing details) this control system was designed to SIL4 standards at least?
I'm about to go through my second SIL2 system and that is painful enough. And that's long before NR let it go anywhere near their signalling centres.
Geoff M.
Posted by George Harris (Member # 2077) on :
I have no idea on the design standards for the WMATA control system.
Yes on Sonneville System in TRUPO. Have to be very careful what I say about this project.
Back to two block sleepers: I can give you a list of presumed advantages / disadvantages of these things:
A1. Four end faces instead of two gives greater lateral stability - this is a so they say, I have no knowledge of any tests to veryify this
A2. Eliminates center binding problem
A3. Lower cost, easier to produce due to no need for prestressing work
D1. Center bar subject to bending and corrosion.
D2. Greater width or closer spacing required to get the same vertical bearing area as monoblock sleepers.
D3. Higher rate of fastener insert problems due to lack of prestress forces holding concrete tight.
I have always worked in places where monoblock was part of the given, so have never done my own analysis of these things.