....I must end my thoughts and words of the importance of gradients on railroads on this thread. The subject to me is and has been an interesting one for decades and I don't mean to take over this "Kinky Subject" thread with another issue.
Believe I've stated my thoughts to be: Gradients are not a minor issue in main line {for the most part}, but other well used branches as well, and we'll let it stand at that.
Quentin
....Yes, the Saluda grade is a fine example of how not to do a railroad on the cheap....It served for some years but at what cost.....??
And as posted above, it's out of service now for several years.
I for one, standing at that site cannot imagine I'm looking at a RR grade....Almost have to pinch myself to make sure it's real.
1435mm wrote:Low grade is hardly the holy grail of railroad location. Many heavy-haul railroads accepted for the long-term some unpleasantly steep grades because the first cost to do otherwise out-weighed the long-term reduction in operating costs obtained by the low-grade alternative. Some classic examples are the DM&IR from Duluth to Proctor; the NYC exiting the Hudson River Valley westward at Albany, New York; SP Palmdale Cutoff and Los Angeles & Salt Lake on the western side of Cajon Pass (both engineered low-grade alternatives and rejected them for excessive first cost, and chose to emulate the Santa Fe or trackage rights on the Santa Fe, respectively). There are recent studies you might not be aware of where detailed engineering studies were made of mountain passes with high operating costs, that came to nothing because the reduction in operating costs wouldn't pay for the first costs.
Low grade is hardly the holy grail of railroad location. Many heavy-haul railroads accepted for the long-term some unpleasantly steep grades because the first cost to do otherwise out-weighed the long-term reduction in operating costs obtained by the low-grade alternative. Some classic examples are the DM&IR from Duluth to Proctor; the NYC exiting the Hudson River Valley westward at Albany, New York; SP Palmdale Cutoff and Los Angeles & Salt Lake on the western side of Cajon Pass (both engineered low-grade alternatives and rejected them for excessive first cost, and chose to emulate the Santa Fe or trackage rights on the Santa Fe, respectively). There are recent studies you might not be aware of where detailed engineering studies were made of mountain passes with high operating costs, that came to nothing because the reduction in operating costs wouldn't pay for the first costs.
On the other hand, there are very few grades on mainlines that are steeper than the 2.2% set by the Pacific Railroad Act. Indeed, two of the most famous ones, Saluda and Tennessee Pass are currently out of service. It would be safe to say that the diesel locomotive has made grades much less important than in the days of the steam locomotive, as well as much stronger cars and couplers.
The SP's Palmdale cut-off is an interesting example, it has pretty much the same ruling gradient as the parallel AT&SF/LA&SL line, but the maximum curvature was 6 degrees as opposed to the 10 degree curves on the original Cajon Pass line. I'm assuming the L/V ratio was of concern to the folks in charge of this project at the SP. The Palmdale cut-off was the site of the most famous runaway of recent memory - both for the initial damage and the indirectly caused pipeline fire of a month or so later.
- Erik
....All serious points you describe.
And in a perfect world the clients would of course like no cost at all....!
I'll just stay with my last example and comment it was the Boswell Branch of the B&O in Pennsylvania and has been abandoned since roughly right after WWII. But much of it is still very much visible.
It was a coal hauler....had passenger trains too, as did many branches back in that era.
The terrain is on the plateau of Somerset Co....roughly 2000' plus in elevation with rolling hills and some more serious restrictions for building RR's.
You can see it by observing from satellite images just north / west of Friedens, Pa.
On the Friedens end of it they used 4 connected horseshoe curves to maintain an acceptable grade for what it was to be used for.
I can't possibly know what the client and engineers declared which was the cheapest to do back about a 100 years ago, but looking at the alignment it's pretty obvious they decided an acceptable grade was the important item to meet {not mileage}, as it is very circuitous in it's route. Sure it followed contour lines....very necessary in that territory to find an acceptable grade. It certainly had it's share of cuts and fills too and brick and stone arches with the fills stacked overhead, etc....
In an operation such as this branch was, I'd not consider "grades" minor to move heavy coal drags out of that area.
Modelcar wrote: ....Believe some alignments were chosen on how much grade the operation could handle determined on what was being hauled on it.Coal haulers, {heavy loads}, especially needed grades acceptable to negotiate such grades and had to be built to allow these loads being hauled. Hence were built to accommodate the tonnage.
....Believe some alignments were chosen on how much grade the operation could handle determined on what was being hauled on it.
Coal haulers, {heavy loads}, especially needed grades acceptable to negotiate such grades and had to be built to allow these loads being hauled. Hence were built to accommodate the tonnage.
First cost is always the first consideration for the engineer. Function is secondary. If the function desired by the client can't be obtained within the budget the client has available for design and construction, the project comes to nothing. One might think that operating cost is more important than first cost -- given the examples you cite of low ruling grade being the first consideration -- but since the only operating cost a client ever likes is ZERO, first cost inevitably is trimmed (and the grade stiffened) until the pain of the operating cost becomes too dear. The engineer's job is to find that balance.
Perhaps you're looking at railroad locations as conceived of purity and thus mistaking the chosen location as the ideal location. What you actually are seeing is the result of an economic comparison between first cost and operating costs, and what are generally long-forgotten are all of the engineering studies that came up with lower grade alignments too expensive to build, relative to the operating costs they would save. I doubt any locating engineer has ever had the luxury of realizing construction on the best location at any time in his career; there are always some better ones that cost too much to build.
A. M. Wellington, who codified railway location engineering, defined engineering as follows. "It would be well if engineering were less generally thought of, and even defined, as the art of constructing. In a certain important sense it is rather the art of not constructing; or to define it rudely but not inaptly, it is the art of doing that well with one dollar, which any bungler can do with two after a fashion." (my italics)
Wellington championed top-down engineering, and termed grades and curves "minor details of alignment," which they are -- the economic questions of location are vastly more important.
S. Hadid
BaltACD wrote: jeaton wrote: This is a link to the NTSB report on the wreck of the Empire Builder Capitol Limited on CSX at Kensington, MD on July 29, 2002.http://www.ntsb.gov/publictn/2004/RAB0405.htm According to the report, prior to the accident a short stretch of track at the accident site was being resurfaced when a mechanical failure made the ballast tamper inoperable. The short stretch of track in the elevation transition was tamped by hand and it was at that location that the buckle occured.As noted above, part of the prevention effort requires that the ties are "locked" into the ballast. It is all part of a system and if one aspect is not up to standards the risk of a buckle increases.The Empire Builder has never operated on CSX tracks at Kensington, MD. I believe you are refering to the derailment of the Capitol Limited.
jeaton wrote: This is a link to the NTSB report on the wreck of the Empire Builder Capitol Limited on CSX at Kensington, MD on July 29, 2002.http://www.ntsb.gov/publictn/2004/RAB0405.htm According to the report, prior to the accident a short stretch of track at the accident site was being resurfaced when a mechanical failure made the ballast tamper inoperable. The short stretch of track in the elevation transition was tamped by hand and it was at that location that the buckle occured.As noted above, part of the prevention effort requires that the ties are "locked" into the ballast. It is all part of a system and if one aspect is not up to standards the risk of a buckle increases.
This is a link to the NTSB report on the wreck of the Empire Builder Capitol Limited on CSX at Kensington, MD on July 29, 2002.
http://www.ntsb.gov/publictn/2004/RAB0405.htm
According to the report, prior to the accident a short stretch of track at the accident site was being resurfaced when a mechanical failure made the ballast tamper inoperable. The short stretch of track in the elevation transition was tamped by hand and it was at that location that the buckle occured.
As noted above, part of the prevention effort requires that the ties are "locked" into the ballast. It is all part of a system and if one aspect is not up to standards the risk of a buckle increases.
Of course. Since I made that post just before lunch time, my brain must have been short on vital nutrients.
"We have met the enemy and he is us." Pogo Possum "We have met the anemone... and he is Russ." Bucky Katt "Prediction is very difficult, especially if it's about the future." Niels Bohr, Nobel laureate in physics
ChuckCobleigh wrote: It's been a million years since I took the solid mechanics (odd place for a double-E to be, but I needed a certain number of out-of-major engineering electives) but it seems that the moment around the vertical axis of the rail cross-section would be less than the moment around the horizontal axis of same, which suggests that the deflection would more likely be horizontal than vertical. (This without (a) digging out the old text book and (b) without actually doing the calculations, mind you, so it is only a semi-educated guess.) An interesting problem, nonetheless.What is your take, Mr. Hadid, being a more polite (civil) engineer than I (decidedly uncivil) am?
It's been a million years since I took the solid mechanics (odd place for a double-E to be, but I needed a certain number of out-of-major engineering electives) but it seems that the moment around the vertical axis of the rail cross-section would be less than the moment around the horizontal axis of same, which suggests that the deflection would more likely be horizontal than vertical. (This without (a) digging out the old text book and (b) without actually doing the calculations, mind you, so it is only a semi-educated guess.) An interesting problem, nonetheless.
What is your take, Mr. Hadid, being a more polite (civil) engineer than I (decidedly uncivil) am?
Can't say that many people think I'm civil, either, but then they get to know me and they're quite definite about it.
Yes, the deflection of a rail is more likely to be horizontal as there is a smaller moment around the vertical axis.
jeaton wrote: This is a link to the NTSB report on the wreck of the Empire Builder on CSX at Kensington, MD on July 29, 2002.http://www.ntsb.gov/publictn/2004/RAB0405.htm According to the report, prior to the accident a short stretch of track at the accident site was being resurfaced when a mechanical failure made the ballast tamper inoperable. The short stretch of track in the elevation transition was tamped by hand and it was at that location that the buckle occured.As noted above, part of the prevention effort requires that the ties are "locked" into the ballast. It is all part of a system and if one aspect is not up to standards the risk of a buckle increases.
This is a link to the NTSB report on the wreck of the Empire Builder on CSX at Kensington, MD on July 29, 2002.
Never too old to have a happy childhood!
1435mm wrote:Homogenous materials typically expand equally in all axis; steel rail is not purely homogenous because the rolling and cooling process tend to orient the crystalline structure, but it's close enough. For steel rail, in two of the three axis (vertical and horizontal) the steel is free to expand and contract at all times. In the third axis (longitudinal) because the material is constrained, the steel goes into compression as it expands, or deflects slightly, between each pair of rail anchors (wood tie) or resilient fastener (concrete ties).S. Hadid
Homogenous materials typically expand equally in all axis; steel rail is not purely homogenous because the rolling and cooling process tend to orient the crystalline structure, but it's close enough. For steel rail, in two of the three axis (vertical and horizontal) the steel is free to expand and contract at all times. In the third axis (longitudinal) because the material is constrained, the steel goes into compression as it expands, or deflects slightly, between each pair of rail anchors (wood tie) or resilient fastener (concrete ties).
Mr. Hadid has furnished with his several responses the most complete analysis of this issue. I would choose to expand on the alignment issue as railroads were/are originally built. Until the latter part of the 20th century alignments were chosen which minimized the initial costs and then as certain locations became major coridores with increasing traffic density line changes were made to accomodate the need for higher speeds or lower grades. This has been described as " there is not enough money to do it right but there will always be money to do it over." That description is not always fair because many of the original projections did not materialize and the original location was sufficient to handle the business generated.
As major line segments became strategic (the BNSF Transcon for example) new construction concepts were introduced by civil engineers and accepted by those who controlled the finances. Then construction of new lines and line changes on existing segments were instituted to embrace not only the then accepted optimum operating conditions, but also to anticipate what the future might hold. Perhaps the 1960 forty four mile Williams to Crookton, AZ line change by the Santa Fe is most illustrative of this and is a major reason why the Transon is todays example of how a railroad can adapt to current transportation needs.
...I suggest "a reduction of grade somewhat secondary" would not be so if said line under construction was competing with an established line with moderate grades already in place.
Case in point: Vanderbilt's South Pennsylvania Railroad across Pennsylvania. The Pennsylvania RR was already established across that territory and making it's way over the mountain ranges with just a hair over 1.8% grades.
Vanderbilts line was being graded with massive cuts and fills and making it's crossings of the up to 7 elevation rises {mountain elevations}, and keeping the grade to roughly no more than 2%. In order to do such it was quite circuitous in climbing the rises, such as the Laural Hill west of Somerset, etc....But hold to the moderate grade, it did. Of course we know his project was abandoned before it became a reality....an agreement had been reached...{long story}, and grading was stopped but some areas can still be seen very easily.
About 53 years later the Pennsylvania Turnpike {original part}, was constructed roughly on Vanderbilts's original route....but modified and straightened a bit and that brought about an increase in grades to 3%.
MP173 wrote: Hold on, got my slide rule out. So, as the sun warms up the rail it tends to expand. As it expands, it is held in place by all of the hardware. Where does it go, or is it held in place by the pressure of the hardware? I tend to agree, when it snaps outta place, I wouldnt want to be around.ed
Hold on, got my slide rule out.
So, as the sun warms up the rail it tends to expand. As it expands, it is held in place by all of the hardware. Where does it go, or is it held in place by the pressure of the hardware?
I tend to agree, when it snaps outta place, I wouldnt want to be around.
ed
Usually the "meander" is to minimize first cost -- the alignment follows contour lines to minimize cut and fill, and bridging. Reduction of grade is somewhat secondary to that purpose and many first locations incorporated repetitive heavy momentum grades as well as stiff ruling grades. But in no case is location made to accommodate expansion of rail. Historically with jointed rail shims are inserted between the rail ends when laying rail, the thickness according to temperature that day (thinner = hotter) to leave a gap to accommodate expansion. The rail joints were maintained in such a fashion to allow the rails to move past the joint bars to open or close the gap; "frozen joints" were a common cause of sun kinks in jointed rail territory.
Semper....No, in my opinion RR track, is layed according to the design of the civil engineers / Surveyors to find {like you suspected}, the best grade over a rise, etc....Hence the circuitous routing .
Down in NC we have the longest tangent track in the USA at just a bit over 78 miles....Totally straight.
Real life example of the POP QUIZ... just much shorter section of track and not 1:1 scale.
I used to have just 40-ft of straight plastic "G"-gauge track on some 1x6 boards on the ground in my backyard to run my locomotive on. I could only run forward and back as I had no curved track that the loco could traverse.
The plastic track I had purchased "used" and it was not designed for outdoor use but had been on the ground for a couple of years at a friends garden RR. It was not UV protected and the sun had deteriorated the plastic and made it very brittle. I had noted that the molded joiners on the ends of the 1-ft sections would break easily if the track was twisted in any way.
Once Sunday afternoon after I had run my loco for a while I noted a storm was brewing and I was afraid the wind would blow the track off the boards and break it. So, I nailed it down on just one end. I figured that would be enough to keep it from blowing away and yet allow for expansion and contraction.
In the morning, on my way to work I glanced at the track and it was right where I had left it.
When I came home that afternoon it was very hot (above 90-deg.) and when I looked over at the track I got a big surprize!!!
I don't know the coeffecient of expansion of that plastic, but the track had taken on a "bell" curve and the middle of the 40-ft was about 3-ft in the air!!!!! I looked at the end I had not nailed down and the last tie appeared to have snagged on a knothole in the wood.
I walked to the middle and, like a fool, I barely touched the track. The "hill" fell over and broke most of the plastic joiners clean off.
My track now is nickle-silver rail on UV resistant plastic ties, buried in "chicken grit" ballast. The layout is about 150-ft composed of two 16-ft diameter loopbacks connected by 50-ft of tangent track. All of it is elevated about 3 ft in the air on pressure treated wood posts and stringers. The loop-backs will contract and overhang the inside edge of the stringers in the winter and expand to overhang the outside edge of the stringers in the summer.
I did tie it down a few places by loosely passing a piece of fishing line under one rail over a tie and under the other rail and then tied to the other end of the line under the wood stringer. I have had to cut some of the tiedowns because, even as loosely as I tied them, they still caused some strange kinks in the track at places.
I am preparing to re-layout my pike and intend to have more "S" curves in places to absorb the expansion/contraction.
I had always figured a real railroad track meandered around because they were seeking the lowest percent of grade to get from one place to the next. Could it be that some of the meandering is to compensate for the expansion/contraction issue?
Semper Vaporo
Pkgs.
I finally received this month's issue. The photo is worth a thousand words. It really gave a good meaning to what happens. I can now understand what has been discussed on this thread and also the STB report. I never quite "got it" about the tamping until I saw the picture.
Since this is now the puzzle thread:
Imagine you have a very long piece of string. It's long enough to stretch all the way around the earth. Since the circumference of the Earth is about 40,074 kilometres, that's how much string you would need.Imagine that you lay out this string along the ground and on top of the oceans, all the way round the Earth. That's one long piece of string!Now pull it tight, so it lays flat. The string makes a big circle that is 40,074 kilometres, or 4,007,400,000 centimetres long.Then you realize that you have an extra metre of string in your pocket, and you want to add it to the string around the Earth ...
You'll have to cut the circle of string somewhere, as it passes in front of you on the ground. Then you'll add exactly one metre more string. Now you want to spread out this extra bit of string around the Earth, supporting it somehow, so that the string forms a circle off the ground, all 40,074 kilometres (plus one metre) around the world.This may take a while! But eventually, you've smoothed the string out into a perfect circle, all the way around the Earth, and slightly off the ground because of the extra 100 cm you added.
You would probably say that the string will hardly be off the ground at all. After all, you only added 100 cm, and the string's length was 4,007,400,000 cm to start with. Most people guess something like 0.00005 cm.
The very surprising answer is that the string will be 15.9 cm off the ground, all the way around the Earth!If you think about it, this is a very large distance for so little change to the total length of the string!Another way to look at the problem may make the answer seem more reasonable. The height of 15.9 cm is in addition to the radius of the Earth. Since the Earth's radius is about 637,800,000 cm, the change in height is in fact very tiny.But wait, there's more ...: http://www.worsleyschool.net/science/files/string/aroundearth.html
gabe wrote: mudchicken wrote: You can't cool the rail (about all you can do is bury the rail in wet ballast, which works only in limited cases and now you can't see the rail rising/running or falling/contracting in the tie plates.Might I suggest an excursion of the Coors Silver Bullet?Gabe
mudchicken wrote: You can't cool the rail (about all you can do is bury the rail in wet ballast, which works only in limited cases and now you can't see the rail rising/running or falling/contracting in the tie plates.
You can't cool the rail (about all you can do is bury the rail in wet ballast, which works only in limited cases and now you can't see the rail rising/running or falling/contracting in the tie plates.
Might I suggest an excursion of the Coors Silver Bullet?
Gabe
Not a good idea...Probably in a heap at Globeville Yard with the two destroyed BNSF locomotives and the wrecked asphalt & stinky beer tank cars.
The number assumes the track is completely unconstrained except at the ends, that it refuses to compress, and that it's infinitely stiff except at the center point where only there it magically turns an angle, none of which happens in reality.
You folks might find this web site informative.
http://www.catskillarchive.com/rrextra/tkpage.Html
Seems to me that sometime last year, Marilyn had close to this same math problem in her weekly Parade column. I think she said the answer was around 50'.
.
Further on the change in temperature required to change the length of a mile (5,280 ft. = 63,360 inches) of steel rail by 1 inch:
It's even less than I surmised (above) - about 2.37 degrees Fahrenheit, based on a coefficient of expansion of for rail steel of 6.67 x 10E-6 (10 to the minus 6th power).
So, a change in rail temperature of, say, 50 degrees - like from the "neutral temperature (see above) - would induce a change in length of an completely unrestrained rail of about 21.1 inches.
And, to apply that to the Pop Quiz example above, the deflection would then be about 21 times greater (slightly bigger angle, and I am being mindful of the "significant figures" here to avoid false precision, too), which would be about 310 feet.
I wouldn't want to be standing anywheres near that rail when it "jumps" !
Note: I've omitted most of the math. If anyone is really interested, I'm willing to post it in more detail when I have more time (not during the working day).
-Paul North.
My answer is:
D. 14.8 ft.
Solved by using Pythagoreas' theorem - the lengths of the 3 sides of the right triangle are 2,640.00 ft. for the base = 1/2 of the 1 mile (5,280 ft.), 14.83 ft. for the upward side, and 2640.04 ft. for the hypotenuse - the 0.042 is 1/2 inch in decimal foot format, rounded (up) to the nearest 1/1,000 (0.001) of a ft.
Interestingly, if the question was how far would the track deflect to the side (instead of lifting up) the answer would be about the same - the geometry is esentially the same, and the track isn't much stiffer in that plane.
Further, it wouldn't take much of a change in the rails' temperature to induce a 1 inch change in length over a mile of track. I don't have a coefficient of expansion value handy, but I suspect it's on the order of 10 degrees Fahrenheit or so. Both of these reasons are why sun kinks can be such a problem.
- Paul North.
POP QUIZ!!!!!!!!!
You have 1 mile of track that is absolulely affixed to the ground only at the ends and the track gets 1 inch longer. Assume the track cannot move in any manner except the center can lift to accomodate the additional length.
How high would the center lift?
A. .5-in.
B. .707-in.
C. 1.48-ft
D. 14.8-ft
F. 54.8-ft
Anybody got an answer?
Mr. Hadid:That is a great explanation, thanks so much.
Murphy Siding wrote: 1435mm wrote: 1. Not decreasing neutral temperature in the process of rail changeout, e.g, such as cutting out a defect in the rail and inserting a stick of rail longer than what came out, and welding it all back together.Is that as simple as replacing, for example, a 10' pc. of track with a 10' pc. of track, or is there more involved than that? Wouldn't a new piece of rail have a slightly different profile than an old piece of rail, causing it to expand/contract at a different rate?
1435mm wrote: 1. Not decreasing neutral temperature in the process of rail changeout, e.g, such as cutting out a defect in the rail and inserting a stick of rail longer than what came out, and welding it all back together.
1. Not decreasing neutral temperature in the process of rail changeout, e.g, such as cutting out a defect in the rail and inserting a stick of rail longer than what came out, and welding it all back together.
One mile of rail will expand 16.35" for a 40 degree F increase in temperature. The profile of the rail isn't important -- say you put in a 52.8' piece of rail, the delta for 40 deg. F is .1635" -- not very much!
What's important is that the rail if it breaks in cold weather -- when it's more likely to break -- is under tension and the ends move back, and when the new rail is inserted the rail has to be returned to NEUTRAL temperature (by heating, jacking, etc.). Since that's difficult to do in cold weather especially when the gang is under the gun to get the main track back in service, it's OK to put in a piece of rail that's longer than what came out (and thus "lose" the neutral temperature) so long as before it gets too warm someone comes back and returns the rail to neutral temperature (and cuts out any excess length). The failure to do that is what gets people into trouble. Or the failure to keep proper records so people know what's been done, where and when, and what needs to be done.
.....Yea, that sounds like an interesting question with possibilities.
Our community is FREE to join. To participate you must either login or register for an account.