This is copied from the 3rd page of the thread titled “Compaction” started by Murphy Sidings. That thead drifted way off topic and it appears that his question was not answered, so I am creating this new thread.
Murphy Siding asked:
We’ve all seen photos from 100 or more years ago, showing men with teams of horses grading a future railroad ROW. Once they heaped up the prairie to the correct height, and leveled it, how was the dirt compacted to carry the weight of a train without uneven settling?
DSchmitt answered:
There are books on railroad construction from the early 1900’s. One example is Construction and Maintenance of Railway Roadbed and Track by Federeck J Prior, copyright 1907. I browsed several of them, looking at their Table of Contents , Index and reading sections on roadbed construction and embankments. I did not find any mention of compaction. Using suitable materials was emphasied, with discussion of what is suitable and where to use various materials. Also discussed was estimation and allowing for “shrinkage” and settlement by building embakments initially higher than the finished grade.
Railway Track and Maintenance by by E E Russel Tratmam, copyright 1926. I have the NMRA 2003 reprint. It say: " It is very desirable that the roadbeds should be compact enough to provide a firm foundation for the track, in most new construction this condition is not obtained, and much ballast is lost by being driven into the comparitively soft and loose surface. In exceptional cases, the roadbed has been compacted by 10-ton steam rollers, with very satisfactory results." This book was initially published in 1897 but there were many editions and some major revisions between then and 1926.
That’s a good question, and perhaps eventually will bring a good answer from qualified people.
Even today, I wonder how tracks remain in good consistent elevation and tangent with so much weight and forces against it…especially where braking might be involved, and curves, etc…Driving across a highway crossing and sighting down the ROW, I am amazed how good the alignment seems to be in most places on main lines, as desired. I’m sure there are locations {problem spots}, that react more than most of it and requires fixing…
One spot here in Muncie I note driving coming south into town on route 3, and sighting down a tangent stretch, it has a very slight “bow” in it where it is supposed to be straight…I’ve watched it for some time…months, a year or more, and wonder if that situation will eventually be pushed back to being straight again…It’s very little by sight, and most would not notice it but being a rail fan, I seem to notice conditions such as this, etc…
Something I see admire is the culverts and small bridges built over a century ago out of Sioux Quartzite. Those all seem to be in really good shape after 100 years, so I assume there was at least some standard of compaction used under them.
Our experience leads us to compare line & level of track to that of highways - because we drive highways day in and day out - and experience the bumps, pot holes and poor state of repair of the highways.
Railroads (even poorly maintained ones) are maintained to a much higher standard of line and level than highways as railroad equipment is much more unforgiving to line and level defects than trucks and automobiles.
Structures such as bridges and stone culverts require a lot of attention to proper load bearing conditions for soil. Railroad tracks can be resurfaced to proper alignment by adjusting them in the ballast if the roadbed settles. But structures break up if their foundation bedding settles. So structures were placed on properly tamped soil or bedrock even in the 1800s while railroads of that era often were built on poorly compacted roadbeds.
Another thing to consider is that in cold climates, even with a properly compacted track roadbed, the surface rises as it freezes up in the fall, and drops back down as it thaws out in the spring. Any variation in soil composition causes a variation in the rise and fall of ground due to the freeze/thaw cycle.
Also, bridges and trestles do not rise and fall with the ground frost while the rest of the track does.
This thread reminds me of a story Maury Klein tells about E. H. Harriman.
As I recall, Harriman purchased controlling interest in the Union Pacific in 1898 or 9 when the railroad was in its second bankruptcy. He was inspecting the tracks with a UP management official. He asked why the ballast ran as far as it did on either side of the track–18 inches, as I recall. The management official replied this was the company standard. Harriman ordered him to reduce it in order to save money on ballast. I always wondered exactly how much money it did save.
And today the prevailing thought is to increase the width of the ballast shoulder to provide additional resistance within the track structure to prevent ‘sun kinks’ within track laid with welded rail. Thus saving money and service disruptions by not having to fix the kinks and the occasional resulting derailment
From my own perspective, Balt, which is that I know nothing about laying railroad track, is that Harriman was making a shoot from the hip decision that was probably pretty dumb.
No doubt you are right, Mudchicken. But Maury Klein does not suggest he did any such thing. His whole objection was based on how much he could save buying less ballast. Harriman’s background was in financing railroads where he had real expertise but Klein does not report he was ever involved in actually building one. Since there as an explicit policy about ballast width the issue had been considered by people who were involved with laying track and this was there best considered judgement. Had I been in Harriman’s place I would not have given such an order.
In order for the entire tie to be supported, the ballast must extend the full length under the tie. Since the ballast has height, and requires a slope at its ends, the base of the ballast must extend past the ends of the ties. If the ballast slopes 45 degrees (for instance), and is 12” thick beneath the tie; then the ballast has to extend 12” beyond the ends of the ties. How far did Harriman want to cut it back?
All I can recall, Bucyrus, is that Harriman believed 18 inches was too much and wanted it cut back. Whether or not he gave a new dimension I do not recall.
But Klein presents Harriman as a man who believed in building a railroad to the highest standards of the day. Yet hearing about his concerns about not wasting ballast I have to say “You sure could’a fooled me.”
Just one or two comments. If the ballast is 12" thick beneath the tie, then the total depth will be more like 18" since it will also be around the ties. And of course you need enough beyond the ends of the ties to provide the necessary lateral support.
Secondly (wandering somewhat from the original thread) although the ballast is under the full length of the tie, tamping should be done only around the rails and not in the central area. The reason is that you don’t want the ties to become center bound. On one subdivision the TEC train was always warning of wide gauge defects, but the gang sent out to fix them could not find any problem. It was eventually realized that the ties were center bound. When the train passed over the ties bowed down under the rails, which then angled slightly outwards causing wider gauge. Naturally when the track forces arrived there was no load and the track gauge was within tolerance. The solution is either a small ballast lift, or a more major undercutting project, but I can’t recall which was used.
Thinking back in history, the 19th century locomotives and freight cars were only a fraction of today’s weights. Wi
As I stated in the original post on this thread, books on railroad construction in the early 1900’s do not mention compaction.
It is clear that good compaction, when achieved, was because of good choice of materials based on experience, the compaction caused by the draft animals and equipment working on the roadbed and also through initally building the roadbed higher than the final grade. The weight of the extra material aided compaction.
The first reference to compaction I found was in the book by Prior.
I have since found this information from Wilipedia:
" The Proctor compaction test is a laboratory method of experimentally determining the optimal moisture content at which a given soil type will become most dense and achieve its maximum dry density. The term Proctor is in honor of R. R. Proctor, who in 1933 showed that the dry density of a soil for a given compactive effort depends on the amount of water the soil contains during soil compaction.[1] His original test is most commonly referred to as the standard Proctor compaction test; later on, his test was updated to create the modified Proctor compaction test."
As is usually the case Proctor built on the work of
Concur. Since each Fresno scraper likely carried and placed not more than about a cubic yard (27 cubic ft.) of material, and there were usually many of them operating in a repetitive linear or circular pattern, each load was likely spread pretty thinly, and then and tromped on by many following teams and wheels as the fill was built and the full and empty Fresnos passing over it to add the later upper layers. See this photo for an illustration:
To attempt to put some numbers to this: Consider a horse that weights about 1,600 lbs., and likely has only 2 of its 4 hooves on the ground most of the time (the other 2 are in mid-air moving forward). So each hoof on the ground is supporting 800 lbs. of horse. If we say a typical horse’s hoof will fit in a shape about 6" long x 6" wide, that’s about 1/4 square foot (1/2 ft. x 1/2 ft.). So, the unit pressure would be around 800 lbs. / 0.25 sq. ft. = 3,200 lbs./ sq. ft., or about 22 lbs. per square i
You sure do like to trash the great men who built the best railroad network in the world.
Harriman ask a pertinent question. “Why are you doing this?” He got a BS answer: “It’s company policy.” (AKA “We’ve always done it this way.”)
When someone answers like that it means they’re not thinking. Harriman certainly did bring the UP up to first class standards so he could not have been totally ignorant of what was needed (or not needed). If the railroad wasted money on ballast too wide it couldn’t spend the money on heavier rail.
A good company will be full of managers asking “Why are we doing this?” That’s one thing that made Harriman so good. He would ask questions like that.
Engineering science does progress. When the 18" standard was set it may have been the best practice of the day. 30 years after the railroad was built the standard best practice could well have changed. Harriman, a thinking man, knew this while his manager didn’t seem to think too much at all.
People fall into ruts and follow routines. A good corporate leader, such as Harriman, will challenge people to think and innovate. And always ask “Why are we doing this this way?” OR “Why are we doing this at all?”
Now I am a little confused. When Harriman asked why the ballast ran as far as it did, and was told it was the company standard; and when he order it reduced to save money— was this the correct thing to do, or did Harriman make the wrong decision based on his wrong assumption or ignorance about the ballast use being wasteful?
On each side of the ties the ballast is tapered down from the hinge point (approximately at the level of the top of the tie) to the toe at the level of the roadbed.
What was the 18" dimension? Was it from the edge of the tie to the hinge point or the edge of the tie to the toe?
I suspect it was to the hinge point, in which case ballast was being wasted. If the dimension were to the toe there would not be enough ballast.
UP Standard Plans show 6" tie to the hinge point on branch lines (adopted 1904, revised 1982) and 1 'tie to hinge on main lines (adopted 1927, revised 1982). The slope hinge point to toe is 3:1 in all cases.
Southern Pacific Standard Plans for heavy traffic branch lines (adopted 1958) show 3" tie to hinge point with 2:1 slope hinge point to toe or 2" tie to hinge point with 2-1/4:1 slope to hinge point to toe depending on the Class of ballast.
Pennsylvania Railroad Standard Plans dated 1917 show the hinge point at the edge of the ties.