Norm48327 There are simple solutions for complex problems and most of them are wrong.
There are simple solutions for complex problems and most of them are wrong.
Off topic, but the simplest solution seems to solve the complexest problem:
Overmod
I conclude that maintaining sufficient clamping pressure is not the key to preventing lateral movement of the tie plate. I don’t believe that it would be possible to maintain sufficient clamping pressure with lag screws threading into a timber tie. As you mentioned a while back, the slightest relaxation of the clamp effect loses all of the pressure.
So, I think the key to preventing lateral movement of the tie plate is to have the bolts functioning only as solid pins perpendicular to the tie plate resisting the shear force. Properly designed, that function will persist reliably over time.
Right now, there is a combination of clamping pressure to create shear friction, and lag screws to act as shear pins; but neither one, nor both combined, are adequate to do the job. So I would get rid of relying on the clamping pressure function, and strengthen the bolt shear resistance by adding the big bushing.
The big bushing would have a pilot to fit the hole in the tie plate, and a bore to fit the bolt shank. These fits would not need to be especially tight. They could be loose enough to accommodate the manufacturing tolerance variation of the hole patterns of the plate and tie.
But the point of the bushing is to increase the effective diameter of the screw shank so it has a much larger load bearing surface area on the bore surface of the tie hole. That larger load bearing surface would overcome the destructive potential of the limited plate shifting allowed by the clearance of the fits of bolt-to-bushing and bushing-to-plate.
Instead of the elaborate scheme of lag srews and bushings where you would have to drill for the screw and counter-bore for the bushing, why not just go with considerably larger lag screws. The larger diameter would have an effect similar to the bushing, not to mention the much increased strength in tension and the greater area for the threads to hold in the wood.
_____________
"A stranger's just a friend you ain't met yet." --- Dave Gardner
Maybe add a lock-washer to maintain tension as the tie shrinks.
Paul of CovingtonInstead of the elaborate scheme of lag srews and bushings where you would have to drill for the screw and counter-bore for the bushing, why not just go with considerably larger lag screws. The larger diameter would have an effect similar to the bushing, not to mention the much increased strength in tension and the greater area for the threads to hold in the wood.
Paul,
A larger bolt might be the solution. I think it depends on how much larger the diameter needs to be in order to provide a sufficient lifespan. The overall problem seems to be a lifespan that is too short, and not a matter of unpredictable, random failure along the way. When the life runs out, there is mass failure.
At this time, the screw diameter is 15/16”. If sufficient life could be achieved by increasing the diameter to say 1-1/8”, maybe making the screw that diameter would be the best choice.
But if the shank diameter needs to be 1-1/2” to provide sufficient life, then a bushing of that O.D. would be the better choice, because an entire lag screw of that size would provide threads that would probably be overkill in holding power, and their cutting the corresponding thread into the tie might split it.
I believe that the present screw size is sufficient in terms of thread holding ability, but the upper unthreaded shank needs a larger diameter. This would stiffen the shank directly to reduce bending and provide more bearing area in the tie to further reinforce the shank against bending.
There are many different combinations of features that might solve the problem. These need to explored and tested. I don’t think that task has been completed. People who did not like the detailed engineering minutia probably decided that pursuing the matter to its intelligent conclusion amounted to beating a dead horse; so they just cut off and went with a half-baked idea.
Overmod had posted this link earlier to demonstrate an interesting new approach to rail anchors. It is called the SELF-SECURING RAIL SPIKE.
Interestingly, the benefit claimed for this fastener is better holding power. It features a rather ingenious anchoring principle by which the legs of a two-legged pin are wedged to spread apart when driven into the tie. The legs have bevels on their ends, and the wedging action of the bevel passing through the wood forces the legs to spread. In spreading, the two legs act as a loading spring. So when fully spread, the legs maintain an active grip on the wood between them. Additionally, the legs have small barbs on the inside of the legs.
The idea seems that it would be very effective.
But is added holding power really what is needed? I don’t think so.
The problem with lag screws is that they don’t have the column strength; that is the ability to withstand the shear loading perpendicular to their axes without bending.
So they bend and the bending also induces vertical pull. They bend and pull until they fatigue and break. There is no shortage of holding power. This is proven by the fact that they always break above most of the screw threads. The screw threads are doing their job just fine. That part of the screw stays anchored in the tie when the screw breaks.
Why won’t these Self-Securing Rail Spikes break off just like the lag bolts? They have no enhanced features that would make them less prone to breaking off above the grip of the two spring loaded legs.
The SSRSs do have a slight advantage over lag bolts in that they compress the wood rather than subject it to tensile stress as a lag screw would do in tending to split the tie from the expansive force of the interfering threads. But, again, lag screws do not lack sufficient holding power, even if they do tend to cause splitting.
As I mentioned earlier, I believe that lifting is a consequence of bending. If bending occurs, it will lift these Self-Securing Rail Spikes despite their barbs gripping the wood. It will either tear the wood to release the barbs, or it will bend and pull until the upper portion of the device breaks off. So it will do no good to add features that discourage lifting while doing nothing to discourage bending. The problem will be solved only by increasing the resistance to bending.
The Self-Securing Rail Spike:
http://www.pages.drexel.edu/~garfinkm/Spike.html
Bucky, you are hopeless. The lag screws DO NOT BREAK ABOVE THE THREADS. They break IN THE MINOR DIAMETER. That is the weakest point of any threaded fastener and is worse here because of the coarse pitch of the threads. The lag screw is a bad idea.
The SSRS might work. It does not have a minor diameter as a weak point but it still offers no real resistance to lateral movement. The greater mass of metal will likely be stronger but only testing will prove that.
If I was UP I would get them in place somewhere like powder river area with lots of heavy trains and no or little hazmat and find out how godd they are. Lag screws would be out, cut spikes in.
tdmidgetBucky, you are hopeless. The lag screws DO NOT BREAK ABOVE THE THREADS. They break IN THE MINOR DIAMETER. That is the weakest point of any threaded fastener and is worse here because of the coarse pitch of the threads. The lag screw is a bad idea.
Yes they do break in the minor diameter which is the weakest point. My point was that they do not strip out of the wood, so a lack of grip is not the problem. I have edited my above post to address your concern about my point which I characterized as "above the threads."
Is the Self-Securing Rail Spike a flawed concept?
The claimed benefit of the Self-Securing Rail Spike is that it squeezes a portion of wood for a grip rather than inducing a splitting stress as is the case with a nail, for example.
However, each leg of the SSRS does induce a splitting stress. And since there is no pre-drilling, I suspect the two legs might actually cause two splits in the wood. The two splits would follow the legs back up to where they converge at the head of the spike on the surface of the tie. So the double split would be an inverted “V” shape showing on the surface of the tie as a single split, but having two diverging branches of the split extending down into the tie.
The length of this “V” shaped split would be in line with the direction of rail/tieplate force. So the SSRS fastener would have very little ability to resist the rail/tieplate force. To visualize this, consider a nail standing in a split. It would be rather easy to tip the nail over in the direction of the split. This is because it would in an open crevice with no wood fiber to directly constrain it from tipping in the direction of the split.
So the SSRS would tend to have less ability to resist the rail/tieplate force than a lag screw or even a conventional cut spike. Yet, resisting this lateral force is the number one requirement. Holding power or grip is number two.
The SSRS claims that its grip is not loosened by tie splitting. That may be true as far as it goes. But the SSRS will tend to cause tie splitting. And that tie splitting will encourage the rail/tieplate force to bend and tip the SSRS out of vertical. Once the fastener tips out of vertical, and allows the tieplate to shift, the holding grip is irrelevant.
Maybe we should go back about 130 years or so:
https://www.flickr.com/photos/paulofcov/28258154210/in/album-72157626021256880/
Text is here:
The American Railway at Archive.org
https://ia802606.us.archive.org/22/items/americanrailway00clargoog/americanrailway00clargoog.pdf
"The American Railway" (about 1889), H. G. Prout, editor, "Railroad Gazette"
Interlocking rail bolts-- book page 220--PDF page 255
I don't know how they insert the bolts past each other unless they drill oversize holes
That is an interesting concept. It shows that they understood the problem, but I am not so sure about the execution. But the concept makes sense. Instead of relying on spikes to hold vertically, just set bolts in diagonally so they have a better advantage. But the details as drawn are somewhat puzzling, particularly those square nuts angled to the axis.
FYI.http://www.cbc.ca/news/canada/montreal/rail-safety-federal-dot-111-lac-megantic-tankers-1.3693564Thank You.
NDGFYI.http://www.cbc.ca/news/canada/montreal/rail-safety-federal-dot-111-lac-megantic-tankers-1.3693564
From the link:
Accident investigators have said for decades that the DOT-111 railcars are easily punctured or ruptured, even in low-speed impacts. A new class of tank car, the TC-117, was unveiled in May 2015 and is described as having a thicker steel hull, thermal protection to increase the ability to withstand fire, a full head shield, protective valve covers and a bottom outlet valve for safety.
"This type of tank car will be much more able to resist puncture," Garneau said.
The same was said of the new and improved 1232 tank cars, and yet they rupture in derailments at speed as low as 22 mph. Has the actual crash rupture resistance of the 117 cars ever been quantified in terms of speed? Or are we to suffer the nonsense of “much more able to resist puncture”?
Paul of Covington Maybe we should go back about 130 years or so:
I dunno - seems to me that arrangement would actually be weaker - the tie would likely fail horizontally. And that would force replacement of the entire tie.
The weaknesses of the lag bolt heretofore discussed notwithstanding, that concept makes sense.
Spikes, however, seem to have stood the test of time
Larry Resident Microferroequinologist (at least at my house) Everyone goes home; Safety begins with you My Opinion. Standard Disclaimers Apply. No Expiration Date Come ride the rails with me! There's one thing about humility - the moment you think you've got it, you've lost it...
Bucky will never agree, but the most expedient solution would be to install concrete ties so the tie plates can't move.
Norm
tree68Spikes, however, seem to have stood the test of time
That's what I've been thinking, too. One other idea I was thinking of is to have oversized holes in the plates for the lag screws and use conical washers. Then use good ole spikes to hold it in place laterally. But really, why, after all these years, is it important to hold the rail tightly in contact with the tie? Could it be that someone thought, "This is the way we do it with concrete ties, so it is the new way, and we ought to it with wood ties."?
Norm48327There are simple solutions for complex problems and most of them are wrong.
Norm,
It is not hard to come up with a correct solution. The way it’s done is to consider many solutions and then discard the ones that are wrong. This will lead to the right solution to even complex problems. You won’t know a solution is wrong until you carry through in developing it, and trying it in practical testing. But even before getting to that point, ideas can often be discarded simply by what is learned in the cad modeling stage.
So complex problems do indeed inspire many simple solutions that are found to be wrong on the way to finding the right solution. But that is not the issue with this lag screw problem. First of all it is not a particularly complex problem.
The industry has probably done sufficient trial and error to arrive at the perfect solution long ago. That is almost inevitable in such a large standardized industry. They have a lot of time to get it right. The problem is that with the ability to find the perfect solution that is inherent with the industry, they learn exactly where limits lie.
So they seek solutions that go just to the threshold of success and not one bit further. You can’t afford any excess in the solution because the industry is big and standardized, so a little excess adds up to too much excess to tolerate. So everything has to be just enough and no more. The problem is that the threshold of just enough is constantly creeping out of reach as loading and traffic levels creep upward. So new solutions are needed on a frequent basis. Yet each new solution requires exhaustive development to find just the perfect one that provides just enough merit to fix the problem and no more.
The problem that has now emerged is holding the gage on curves. For this task, the lag screw systems had been applied in lieu of standard cut spikes and tie plates. Since that decision, however, the problem has either grown a tiny bit worse with factors such rising tonnage, train frequency, or speed; or some combination. Or it may be that the “just enough solution” of lag screw systems was not quite enough, and it has taken 10-15 years of empirical experience to reveal that. But whatever the reason, the gage holding problem is back.
So the U.P. has decided that the painfully tedious task of finding just the right tiny increment of the next improvement is less preferable than the immediate solution of reverting back to the tried and true cut spike/tieplate method. It may require more maintenance, but at least it does not hide its impending failure like lag screws have proven to do.
Lag screws hid their relatively short life span until recently. Not only have they now revealed that, but they may have also revealed that the life of lag screw systems is not even as long as cut spike/tieplate systems that it replaced; but nobody knew that going in.
Special track work and the use of lag screws
Never too old to have a happy childhood!
rdamonI think it is important to look at how things fail. A cut spike will loosen when pulled, but a screw with over 2x the holding force will break. A loose or partially pulled spike will still provide lateral stability.
When a screw breaks without losing its anchoring power, it tells me that the screw is not big enough to do the job that is assigned to it. That would be my first conclusion with this problem of breaking lag screws on track curves.
Or need to be make of a higher grade metal. But this may only move the point of failure, not prevent it.
I would opine that track structure in general needs to be dynamic. Screws would tend to make it rigid and any point that gets more stress than others will be the first to fail.
Spikes tend to give a little if need be - they don't necessarily pull all the way out - just enough to let the track flex a little.
The puzzle switches in the video have a little tighter tolerance, and would see much more intensive maintenance than a mainline, so the screws might work there.
rdamon Or need to be make of a higher grade metal. But this may only move the point of failure, not prevent it.
If it is made by man - it will fail - where, why and when are the only questions.
This is a real problem that can be solved. It should be solved. I cannot imagine the railroad industry simply dismissing the problem because anything built by man can fail. Lag screws breaking on curves is not what would be called an “Act of God” like a tornado derailing a train.
What I would like to know is how widespread the use of lag screws is on U.S. railroads. Are there hundreds or thousands of mainline miles fastened to ties with lag screws. Or are they just used on curves by some companies as is the case on the U.P.?
EuclidThis is a real problem that can be solved. It should be solved. I cannot imagine the railroad industry simply dismissing the problem because anything built by man can fail.
It's gonna come down to dollars and cents, pure and simple. If lag screws are not economically feasible, why pursue them? As noted before, spikes work, and have worked for, what, 150 plus years?
While the industry may do more work with lag screws to see if they can be made to work, in the end, the accounting department will have the final say.
tree68 Euclid This is a real problem that can be solved. It should be solved. I cannot imagine the railroad industry simply dismissing the problem because anything built by man can fail. While the industry may do more work with lag screws to see if they can be made to work, in the end, the accounting department will have the final say.
Euclid This is a real problem that can be solved. It should be solved. I cannot imagine the railroad industry simply dismissing the problem because anything built by man can fail.
When I said that the lag screw problem should be solved, I meant that in opposition to BaltACD seeming to dismiss the problem without solving it, and doing so based on the recognition that anything made by man will fail.
So when I say the problem should be solved, I am not saying that it must be solved only by better lag screws or some other new idea. I agree that the problem of breaking lag screws could also be solved by reverting back to cut spikes. I also agree that whatever the solution, the choice will be driven exclusively by cost.
In this case, the cost will be comprised of the cost of the following:
Initial fastener application.
On-going inspection cost.
On-going maintenance cost.
The cost of train wrecks caused by fastener failure.
Although cut spikes have worked for 150 years, they have become less workable due to the rising cost of maintaining them on track curves of the busiest, highest tonnage main lines. So the rising track load stress over the last 150 years has reached a “breaking point” in terms of maintenance cost on main line curves. I assume that the U.P. chose to substitute lag screws for cut spikes on curves in order to reduce the cost of maintenance due to the better grip of lag screws.
It appears, however, that the cost of maintenance of lag screws did not account for breakage, nor the inspection required to detect breakage. So the failure to account for those two hidden costs led to a decision to use a fastener that causes derailments, thereby raising the cost higher than the cut spike/tieplate system that lag screws replaced.
So the company has two choices:
Improve the lag screw system to eliminate the breakage, and the need for extra inspection to detect breakage.
Revert back to the cut spike/tieplate system and its higher maintenance cost due to the inferior grip of cut spikes.
Or, they could simply go with concrete ties in such locations...
Concrete ties do seem like a viable solution. I wonder why they have decided to revert back to cut spikes/tieplates instead of concrete ties.
Euclid Concrete ties do seem like a viable solution. I wonder why they have decided to revert back to cut spikes/tieplates instead of concrete ties.
It is what can you do TODAY vs. what can you do in a couple of years. Concrete ties don't grow on trees, wood ties do. Concrete ties and wood ties can't effectively be intermixed on a territory. Today's repairs may not be tomorrow's long term fix.
BaltACD Euclid Concrete ties do seem like a viable solution. I wonder why they have decided to revert back to cut spikes/tieplates instead of concrete ties.
Our community is FREE to join. To participate you must either login or register for an account.