To all you engineers out there (and especially 1435mm): what geographic conditions or other considerations merit the construction of a RR “loop” (i.e., Tehachapi or Williams in CA)? Isn’t the economic and catastrophic potential of one of the bridges washing out or collapsing onto the top of another part of the train one that would discourage too many of them from being built? While we’re on the subject, are there any other “loops” out there, either in the States or abroad? Have there ever been any really serious accidents of the type I’ve described?
Actually, the Tehachapi Loop has a tunnel. I would think that the main reason why there are not many loops is that the train is doing quite a bit of traveling but not making much progress towards its destination.
I never thought about it before but I guess the Tehachapi Loop tunnel made it through the 1952(?) Tehachapi earthquake with little damage.
The one’s I have seen (CP spiral tunnels at Yoho in British Columbis and the spiral tunnels on the Alishan forest railway in Taiwan) are also both tunnels, so no bridges are involved.
The purpose of them is to give a gentler grade between 2 points. When the CP was originally constructed the grade through that part of the mountains was 4.5%, by putting in the spiral tunnels the grade ws reduced to 2.2%. http://www.kohlin.com/can-97/canada-4.htm shows a diagram, and you can see that if the tracks just went from the bottom of the diagram to the top the grade would have been horriffic.
Other loops are Williams CA on WP, also a tunnel, and Georgetown Loop on the touris Narrow Gauge at Gorgetown. I can not remember original builder. This one has a bridge at significant elevation over the lower level. None of them have had any particular problem that I know of.
…I see no increased danger of a bridge collapsing on a loop as any other bridge being used on a railroad. Either one failing would shut down the line.
Sometimes you can’t reach from point A to point B in any other fashion except to make the route more circuitous and of course making it a complete loop is the ultimate to do that.
Many former and current railroads have loops in their route designs…Loops of the type that almost cross to complete the circle but not quite. All in the effort to gain altitude at an acceptable grade. All variations of the loop add mileage of course but is necessary to get up the grade in those cases.
A loop is a device to increase the numeric value of RUN in the equation GRADIENT = RISE/RUN. If minimizing distance was the pre-eminent concern of a locating engineer then a railway route confronted by a ridge across its path would attack the ridge straight-on. But while distance is a fixed cost that the locating engineer must seek to constrain gradient is always the controlling concern and if the topography in the direction of desired progress rises faster than the maximum acceptable gradient, the engineer chooses to (1) diverge the route around the topography; (2) tunnel through the topography; (3) introduce length into the railway route in order to surmount the topography without exceeding the maximum acceptable gradient. All three methods are routinely employed separately and in combinations on almost any route one cares to examine. A textbook example because there’s little vegetation or intervening points of economic interest to obscure the case is the Western Pacific between Portola, Calif., and Salt Lake City, Utah, where there are found two summit tunnels and a whole lot of diversion around the mountain ranges that furrow the Great Basis in perpendicular opposition to the WP’s desired route. Because there were no traffic sources of consequence in this region the locating engineer could choose his route unencumbered by considerations of the route having to pass through these sources regardless of the cost of construction or later cost of operation.
In the third case where length is introduced the locating engineer seeks to recognize from the topography a “supported grade” which is an alignment that where the topography itself supplies the desired rise and run with a minimum of earthwork and not exceeding the desired maximum curvature nor introducing excessive curvature. The easiest way to envision a supported grade vs. an unsupported grade (and both can be found, again on the same line) is to consider the case of a low, smooth-sided ridge lying perpendicular to th
One is still in operation on the Chihuahua & Pacifico (“Chepe”, now part of Ferromex) through the Copper Canyon in Mexico.
There also was one on the old narrow gauge Newfoundland Railroad, the “Trinity Loop”. Physically it’s still there, although the road was abandoned back in around 1988 and most rail pulled up. The high level crossing there was originally a trestle, which was later filled, leaving only a girder bridge at the crossing point.
CP has a tunnel that is a loop somewhere in BC I think. Ideally a railroad would like to go straight and level but that isn’t the way the world is. Switchbacks, loops, horseshoe curves and all kinds of oddities are part of the engineers toolbox. Whatever the solution you can rest assured it is the cheapest feasable one and will not collapse upon itself.
You’re referring to the Spiral Tunnels on the west slope of Kicking Horse Pass, as did Mr. Jampton earlier in this thread.
Actually the chosen solution is seldom the cheapest feasible solution if one refers only to the cost of construction. The locating engineer’s charge is to achieve the desired outcome between initial cost and operating cost. In general terms early construction in North America favored lowest initial cost due to the high cost of capital and low cost of labor and fuel whereas later construction favored lowest operating cost as these fixed inputs of capital, labor, and fuel reversed places. Exceptions exist and serve to prove the general rule.
One you omitted is the old Hiwassee (sp?) loop on the Louisville & Nashville’s “Hook and Eye” line from Etowah, TN down through Copperhill to Marietta, GA.
The 1962 NRHS convention ran a trip from Atlanta to Etowah up the main line and back down the back way over the loop. Not only did we see the scenic attractions, but the power was an Alco FA-FB-FA consist.
If you travel from Holland to Denmark by train all the way you pass 3 loops on route. And this is generaly flat country, no mountains at all.
There is a loop in the town of Osnarbruck where the train stops at the same station twice because that’s where the track crosses over, and then Rendsburg the train loops on a giant bridge over the Keil Canal and the for some reason the train loops AGAIN around the town of Flensburg. Not a real famous route or anything, just 3 unfamous loops (well the bridge loop over Keil Kanal is well known I think).
In order to get over the Allegheney mountains in western Pennsylvania, the Pennsylvania railroad built the horseshoe curve. Actually a loop around the mountain that is a 1.72 percent grade. There was no way that they could go over the mountains without blasting out gigantic tunnels. It was more economical to take the longer route with the easier grade. Furthur west they did have to cut tunnels through the mountains at Gallitizan.
Actually I consider where the train passes over itself is more of a overpass similar to any highway overpass you see in the USA. Yes there are tunnels at the loop but not where the train passes over itself to the best of my knowledge. There is a tunnel as come you out of the horseshoe curve in Caliente & start up hill turning the train 180 degrees
…Back when the Pennsylvania RR put their “Horseshoe Curve” route up and over the Allegheneys they didn’t need to deal with building any tunnels on the 12 miles up from Altoona until they got right to the summit. As a poster indicated above, at Gallitizin. Eventually they had 3 tunnels that were used to cut a few hundred feet off the grade at the summit.
Generally the grade is about 1.8% up from Altoona to the summit with a few spots steeper and around the curve I believe it moderates {to compensate for the curve}, to about 1.45% grade.
1435mm: On another thread, you spoke of some of the engineering on the Milwaukee Road between Milbank, and Aberdeen, S.D. I could never figure out why that line had such huge fills, for miles at a time, in an area with sort of ,gently rolling hills. Given that, and the fact that some lines ended up with s-curves, horseshoe curves, and loops, what was the intention of the engineers? Did they try to get the easiest, straightest, ROW over the lines as a whole? Or did they just do their best as they went along, and note the ruling grade afterward?
Different engineers had different approaches and different instructions from the boss. In some cases it was to build it cheap. The original UP main is a good example. Most of the NP is another. In some cases it was build the cheapest to operate. This is a decision to build better, more money today, to save operating costs. One could argue that this was the objective of the MILW’s Pacific Coast Extension. Problem was the line never had enough traffic to pay for the very high investment the line represented. Line relocations were typically “build better” situations, but they were constrained by budget issues as you are trading dollars today for savings in the future, so there is a limit to what makes economic sense to do.
Very common was to build an acceptible line within some constraints and then upgrade it as funds allowed. The land grant transcontinentals were constrained to 2.2% grade by law. An example of this on the NP is Evaro Hill on the NP west of Missoula MT. Original line went up and over on legal max of 2.2%. Later a longer river grade line was built which became the freight main. Passengers took the old line. Both lines are still in service today, a rarity.
The real answer is they did what made sense at the time, and as time went on the tendency was to reduce operating costs by reducing grade and curves. Relatively little of this type of work has been done since 1910 due to the capital starvation of the industry due to maximum rate regulation which started in 1906. For an overview of the capital starvation issue see “Enterprise Denied” by Albro Martin.
Railways at that time (late 1895-1906) strove to reduce operating costs to the absolute minimum. The primary way to reduce operating costs were to reduce engine-miles per train-mile and increase tonnage ratings for a locomotive on a given division, and the primary way to reduce that ratio was to flatten the railroad out. Thus the big cut-and-fill project between Milbank and Aberdeen. Another way to look at it is “engineer the alignment in order to load the locomotive to the maximum tonnage which it can just barely stagger over its engine district at a speed of 25 mph.” It wouldn’t do at all to have the engine district limited by one or two short hills out of small drainages when on the rest of the district the engine didn’t have to work. The way you get maximum work out of a locomotive is to work it hard 100% of the time.
The last question you ask is whether railways looked at this wholistically. Absolutely! Railways tended to do this work a locomotive-district at at time, using heavier locomotives in the districts where it was economically impractical to reduce the ruling grade to the same as adjacent, flatter districts, or accepting helper grades on top of the ruling grade. Or the railway introduced tonnage set-out points and when enough excess tonnage accumulated at this point an extra train would clean it out for recons