"The train was negotiating a sharp curve when at least six cars went off the track. Woods said a combination of forces, “the buff-slack push-pull of a train,” caused the rails to spread, causing the derailment.
“The train was in the process of going down one grade and up another and around an 18-percent curve,” he said. “We look at this as an anomaly, and you have to be very careful operating the train.”
http://thegazette.com/2010/12/22/cedar-rapids-derailment-caused-by-combination-of-factors-rail-official-says/
Not being in the business, is there a simple explanation of the forces at play and their effect?
Those forces result when the slack runs in and out. They can lift cars off of the rails when running in, and string line cars off of the rails when running out.
These slack effects can be unpredictable, so that maybe accounts for calling the above event an "anomaly."
Maybe a 'stringlining ' type effect exacerbated by the product carried (shelled corn(?) might have shifted,and caused a car to rock on uneven track, and then when the rails spread to continue to roll over(?)
Ultimately, a derailment on 'bad track" that spread under weight ( was the track rated for the weight of the cars passing over it(?), and movement of the cars movements.
Sounds like the cars going downhill ran in and pushed against the cars going around the curve, and against the outside rail, and the rail gave way.
Had that happen on my layout. Too many free rolling cars going down a 2.3% grade, entering a 28R curve. After 4 cars entered the curve #5 found it just as easy to go straight, over the rail. Yeah, the cross level of the curve and track guage may need to be checked, but the current cure is not to run 20 Walthers passenger cars in a train.
Phil
Timber Head Eastern Railroad "THE Railroad Through the Sierras"
Stresses and strains of the train going up and down hill and on a curvature at the same time also causes stresses ans strains on the track. When stresses and strains work against others stresses and strains, tracks can widen their guage, wheels will jump off rails, couplers uncouple, and thing get messed up in a hurry. I.E. you have a derailment.
RIDEWITHMEHENRY is the name for our almost monthly day of riding trains and transit in either the NYCity or Philadelphia areas including all commuter lines, Amtrak, subways, light rail and trolleys, bus and ferries when warranted. No fees, just let us know you want to join the ride and pay your fares. Ask to be on our email list or find us on FB as RIDEWITHMEHENRY (all caps) to get descriptions of each outing.
First off curvature is not stated in %....it is stated in degrees. a 18 degree curve is a exceedingly tight curve and would most likely be restricted to 15 MPH or less. With the statement that the train was proceeding down one grade and up another all the while moving around a very sharp curve would make effectively controlling the slack in the train a very difficult task depending upon the length and load distribution of the train.
While having brakes applied (either dynamic or air) would have the slack bunched to some extent in the train as it descended the grade into the curve, because the terrain then went up grade, brakes would have to be released as the train began to climb the grade it it would begin stretching out the slack in the train....thus causing changing dynamics on individual cars as the slack change propagates through the train....under such circumstance derailments are searching for the weakest link....car/truck dynamics or weak tie/rail conditions in the 18 degree curve. A very tricky area to operate a train of any REAL length or tonnage.
Never too old to have a happy childhood!
What BaltACD said above about curvature is corrrect - an 18 degree curve is about 320 ft. radius = 44" in HO, if I'm doing the math right. But the length and percentage of each grade is not stated; nor is the speed and length of the train, and where the derailment occurred within that length, nor the number and type/ power/ weight of the locos, etc.
With all those unknowns . . . it's stated that "the rails spread" which would have been due to train forces, but it was likely poor tie condition that allowed that to happen. Most likely, the outer rail on the curve either slid out a couple inches and/ or turned over partially or entirely, the outside wheel followed along, and then the inside wheel fell off its rail and started tearing up the track structure. Everything that occurs after that is pretty much random and unique to the individual circumstances of each derailment location and train.
- Paul North.
(Tie condition may be getting a bum rap)
Look at it this way.....If lateral forces do NOT roll over the rail by pulling up the cut spikes, the next obstacle to be surmounted would be vertical wheel lift only restrained by friction of the wheel flange against a narrow portion of the railhead on the gage side....If the spikes become screws & pandrols, the curve becomes tangent or the curve happens to be lubricated...what variable finally bites you?
Isn't velocity and kinetic energy and potential energy wonderful for all calculations ??
Cannonball
Y6bs evergreen in my mind
Something doesn't sound quite right. Any number of things can cause the rail to spread. I'm sure Paul can dig up the final report when it's published. There are several unanswered questions:
I never trust what someone says to the newspaper about the cause of a derailment.
Nick
Take a Ride on the Reading with the: Reading Company Technical & Historical Society http://www.readingrailroad.org/
nbrodar I never trust what someone says to the newspaper about the cause of a derailment. Nick
Yes, there are many reasons and ways rails can become spread, and each derailment case is different. But your final sentence, Nick, is not quite accurate. It depends on who makes the statement and when during the event. But more importantly it depends on how much the reporter/writer understands the dynamics of railroading and accidents.
Unless the reporter is a former railroader and has had time in train service they have NO understanding of the dynamics of railroading and the accidents or their causes. No on in the industry can state with any specificity the cause while the cars are still piled up and wreck clearing crews are still doing their thing and the investigation process is still taking place.
Cause allocation in a significant derailment is a tri-part process involving the Operating officers concerning train handling and rules compliance, the Car Department personnel concerning the condition of the cars and any defects that may have had a causative factor and the Roadway personnel to answer for any potential track defects that could have had a causative factor. If it is a REALLY significant derailment and causes loss of life the NTSB will interject itself into the investigation and their cause may or may not agree with the Carriers cause assessment.
"
This derailment was caused by poor train handling. No doubt."
-I don't agree with this commentor's statement. While handling is important, this could happen even with your best engineer at the head. As trains are now quite large in size, controlling all the forces is quite difficult. Stuff like this is more likely to happen due to train size and weight.
Victrola1 "The train was negotiating a sharp curve when at least six cars went off the track. Woods said a combination of forces, “the buff-slack push-pull of a train,” caused the rails to spread, causing the derailment. “The train was in the process of going down one grade and up another and around an 18-percent curve,” he said. “We look at this as an anomaly, and you have to be very careful operating the train.” http://thegazette.com/2010/12/22/cedar-rapids-derailment-caused-by-combination-of-factors-rail-official-says/ Not being in the business, is there a simple explanation of the forces at play and their effect?
If you want more detailed accounts of various types of derailments and other railroad mishaps, visit the NTSB and TSB (Canada) accident websites. Both go into great detail as to track structure condition, mechanical condition of locomotives and cars, weather conditions, negligence, if any, etc.
John Timm
The derailment happened on former Milwaukee Road trackage the Crandic now owens. I have a link to a pdf of the Milwaukee Road Iowa Division 1973 track chart.
http://milwaukeeroadarchives.com/Construction/Milwaukee%20Profiles/IowaDivision1973Part2.pdf
You need to go to page 9 (of 52) and the top of the page, between Vera and Cedar Rapids. More specifically between MP 85 and MP 86. I think the derailment would be at or about MP 85.75 on the track chart. (The Crandic may not use these mile posts. I know on the rest of this line to Amana they have renumbered the mile posts.)
I have no idea on the size or make up of the train that derailed.
Jeff
Yes, there is a simple explanation of the forces at play and their effect.
Most likely the forces are from jackknifing (or stringlining). With the basic data( track, truck and carbody, etc.), the calculation of forces (or displacements) can be done.
The old theory about jackknifing (or stringlining) has many errors or is incomplete. That is why it looks like an anomaly.
"First off curvature is not stated in %....it is stated in degrees."
To state a curve in degree is not so correct either or is somewhat misleading. To do a curving analysis, only the radius of curvature ( R ) of the curved rail is needed. (not the degree).
Certainly, there is a 100-foot-chord method to measure the degree to determine R. However, this method has huge problems:
1). The data measured have very large errors because the center of the curvature of the curved rail cannot be easily located exactly in the field.
2).The center point of the curvature of the curved rail must be at the same elevation as the rail. But most rails are higher than the ground. Even worse, in some cases, the center point is inaccessible, making the measurement impossible. For instance, a curved rail on a dam, a levee, a bank or a cliff. If a curved rail goes around a mountain, the center point is inside the mountain.
worker "First off curvature is not stated in %....it is stated in degrees." To state a curve in degree is not so correct either or is somewhat misleading. To do a curving analysis, only the radius of curvature ( R ) of the curved rail is needed. (not the degree). Certainly, there is a 100-foot-chord method to measure the degree to determine R. However, this method has huge problems: 1). The data measured have very large errors because the center of the curvature of the curved rail cannot be easily located exactly in the field. 2).The center point of the curvature of the curved rail must be at the same elevation as the rail. But most rails are higher than the ground. Even worse, in some cases, the center point is inaccessible, making the measurement impossible. For instance, a curved rail on a dam, a levee, a bank or a cliff. If a curved rail goes around a mountain, the center point is inside the mountain.
Just waiting for our resident railroad surveyor to handle this one!
I don't think Mr. Moderator wants me too have a war of wits with an unarmed individual. Hopefully worker hops on the next rubber-tired turnip wagon and prattles on.
I needed a good laugh this morning.
YIKES!
Seems like we are on a carnival ride, with a visit into the Zombie Apocalypse,(circa: 2010) zone
OP was from December22, of 2010 Questions were answered pretty susccinctly at that time. By several of our resident experts, those folks, who earn their livings by figuring this stuff out. in real terms ( MC and PDN, and commented on by BaltACD as well as nbroder and jeffherget ),Qualified individuals; whose opinions are always well thought out, and backed by their railroad experience.
I am a guy who is trying to correct many misleading theories and practices for the railroad industry. Curve- degree-radius measurement is one of them.
Please point out one thing that is wrong in my statements. I bet you cannot do it.
What I can say is that the measurements you did before were wrong. If you continue to do that you should feel sorry for your pay check.
worker I am a guy who is trying to correct many misleading theories and practices for the railroad industry. Curve- degree-radius measurement is one of them. Please point out one thing that is wrong in my statements. I bet you cannot do it. What I can say is that the measurements you did before were wrong. If you continue to do that you should feel sorry for your pay check.
worker,
What measurements are you referring to as being wrong? Who are you addressing?
Bucyrus:"Who are you addressing?"
Worker: My message is to Mudchicken's message :"
I needed a good laugh this morning."
Worker's answer:"
What I can say is that the measurements you did before were wrong. If you continue to do that you should feel sorry for your pay check"
worker Bucyrus:"Who are you addressing?" Worker: My message is to Mudchicken's message :" I don't think Mr. Moderator wants me too have a war of wits with an unarmed individual. Hopefully worker hops on the next rubber-tired turnip wagon and prattles on. I needed a good laugh this morning." Worker's answer:" I am a guy who is trying to correct many misleading theories and practices for the railroad industry. Curve- degree-radius measurement is one of them. Please point out one thing that is wrong in my statements. I bet you cannot do it. What I can say is that the measurements you did before were wrong. If you continue to do that you should feel sorry for your pay check"
Worker,
I generally understand what you said about curves, but what do you mean when you said this?--
"What I can say is that the measurements you did before were wrong. If you continue to do that you should feel sorry for your pay check"
What measurements are you referring to that Mudchicken did?
Bucyrus What measurements are you referring to that Mudchicken did?
There will be no answer to that, since the poster complaining has evidenced little or no knowledge of surveying so far, among other things. I too am a little in the dark as to what worker's actual point is, though his or her confidence seems unbounded. I have seen that phenomenon before from time to time, to which my response has always been "make your case" with specifics.
With that observation, I will make some more general observations which will be familiar to some forum participants in one manner or another.
Chuck, you did intend to say a 100 link chain, did you not? For the benefit of non-surveyors, I will remind them that each link in a surveyor's chain is 7.92 inches long, which makes a chain 792 inches (66 feet) long.
Johnny
Deggesty Chuck, you did intend to say a 100 link chain, did you not? For the benefit of non-surveyors, I will remind them that each link in a surveyor's chain is 7.92 inches long, which makes a chain 792 inches (66 feet) long.
What I referred to was something to measure to the next 100 foot station interval, which appears to have been pretty much standard procedure in staking a curve. Of course, these days that would be accomplished with the precision distance measurement capability of a modern surveyor's station; how specifically railroad surveyors worked it in the past I don't know.
A chain is four rods in length, 66 feet, as you note. Land surveying (boundaries and area) is likely somewhat different from engineering surveying as the goals are not perfectly aligned and I would suspect the tools of engineering surveying might differ a bit from meets and bounds surveying. That being said, I was referring to a generic hundred-foot chain, not a surveyor's chain.
What is to this day utterly amazing to me is how accurate many of the 19th century surveys turned out to be when resurveyed with modern instruments of nominally much better accuracy.
BTW, I'll add that I am flabbergasted by the capabilities of the modern total surveying station and what an advantage that gives to the construction engineer. Progress that really helps!
ChuckCobleigh A chain is four rods in length, 66 feet, as you note. Land surveying (boundaries and area) is likely somewhat different from engineering surveying as the goals are not perfectly aligned and I would suspect the tools of engineering surveying might differ a bit from meets and bounds surveying. That being said, I was referring to a generic hundred-foot chain, not a surveyor's chain.
I may be on thin ice here...
Land surveying with a chain was a way of getting away from "meets and bounds" and thus closer to engineering survey than you think. The 100 foot baseline for degree of curvature was likely derived from the push to decimal measurements.
FWIW, the 66 foot chain is the origin of the size of an acre (10 square chains) and the statute mile (80 chains).
Very true.
- Erik
ChuckCobleigh:
Your message clarifies the misunderstanding among us and makes the discussion productive. I am talking about using AAR’s (Association of American Railroads) method (100-foot-chord) to measure the DEGREE or Radius of curvature for a curved rail with zero information about the curve. See sketch below for AAR’s definition. What I found is that the method is a misleading.
(Note: the sketch cannot be shown out here, it the common 100ft method)
What you are talking about is constructing or measuring a curved rail on site with the DEGREE and radius of curvature given. With my limited knowledge in survey, I know there are three methods that can be used to build stations on the curve, such as the tangent offset method. That explains why “they are fanatical about spirals” because the radius of spirals are complicated by spiral design. Am I correct?
I am not a surveyor and you know it. My trouble begins when I want to know if a curved rail is really constructed to the design. That is, if the radius is really the one specified in the design.
Anyway, see if you can do this. If we have a curved rail with a very large radius of curvature (say is about 7,000 meters), and all the records for the curved rail are lost. We don’t know nothing about the radius or degree. Can you (using your method) measure the unknown curved rail to determine the radius of curvature with ease and accuracy? I will certainly learn something from you if you can do it because I came up with a simple method to measure the radius of curvature for all the unknown curved rails.
Certainly, there are tangents, curves and spirals in the rail system. However, a spiral is another misleading issue in the railroad industry. The reason for a spiral to be used, I found, is that there is a jump in radii (R) of curvature from tangent to curve (∞ vs. R). A spiral is used as a transition from ∞ to R. However, my calculations show that there is no jump in radii in curving. That is to say, the transition from tangent to curve is smooth and monotonous. Furthermore, a spiral would cause two problems in curving: 1). truck vibrations and 2).a larger centrifugal force.
Using R in curving analysis is clean and concise. As you said, R is used in centrifugal force calculations. In Super-elevation, degree is used because they played the mathematical game to do that based on the relation between R and degree. Actually, besides the centrifugal force, I have found at least five other lateral forces in curving. All of them depend on R (not degree if you don’t want to complicate the expressions).
Reading the above posts about railroad track that curves leaves me puzzled. There seems to be (or at least no one admits to knowing) how such a track can be built. In fact, the most logical conclusion from the posts themselves would be that such a track is impossible. But of course we do have railroad tracks and some of them are curved.
Our community is FREE to join. To participate you must either login or register for an account.