The Physics of a Derailment - Explain

[CSSHEGEWISCH]

Mookie

Help?  "You need a chain"..... 

As usual, I am lost....

A chain is a bit more flexible and would have to be used when the curves are too sharp to allow the couplers mounted on a rigid drawbar to line up.  If my memory serves me right (questionable at my age), the Illinois Terminal had some special coupler links that were used to get freight cars through the curves in Bloomington, IL.

[Mookie]

Are the chains the ones I see on the cars around the couplers.  That is where I am confused.

[mudchicken]

ChuckCobleigh

mudchicken

Tightest freight curve I ever remember laid out was 38 degrees at Pueblo (Target Warehouse - Architect pulled an incredibly stupid stunt)

Looks like they "erased" that mistake some time ago, as the spur has been nipped off about half way through the curve.

The mistake was "erased" when they extended the building in the late 1990's. Before then, Santa Fe had already refused to switch the track after (not taking into account the 38 degree curve) the same arrogant bozoes placed a pair of clearance violations on the side of the building ( a dock extension/cover and a fire control valve (with a huge red wheel) reducing side clearance to less than 7 feet. After that, ATSF shoved a car past the clearance point on the track and told Dayton Hudson/ Target to "come get it".

(End of useful life of the industry track, a lesson in an architect/engineer operating outside their limit of expertise)

Mooks - The chain replaces the coupler because it (the coupler clasp/ knuckle) won't work, won't couple and is a lovely illustration of bypassed couplers. Shoving cars, the knuckles head-butt each other (why old cars had poling pockets). Pulling cars, the chain holds the cars together (coupler is useless)....At Pueblo Target, if a car was to be switched in there, the trainmaster/supt. would ask that a GP7/ GP9/CF7 be sent west of Dodge City because the newer power did not have the coupler swing (removable stops on the older power) to handle the curve)

[Deggesty]

Dayton Hudson/Target warehouse architect: non-useful idiot?

[friend611]

I have been studying the wreck of N&W 611 on January 23, 1956 for several years. The wreck was noted as being caused by excessive speed, both N&W and ICC investigators coming to this conclusion. It happened on a 13'13 curve restricted to 30 mph for passenger trains. In the ICC report, the train passed on in approaching the curve, a 578-foot tangent, a 5-degree curve, a 169-foot tangent, and a 213-foot spiral. The overturning speed of the locomotive was calculated at 53.6 mph, though I suspect the 611 was traveling at a higher speed. due to her clearing the eastbound tracks (she was traveling westbound) . She then slid about 300 feet before ending up on a river bank with her tender and the first four cars of her train. It is my feeling that high speed would only allow for the momentum to carry her and part of her train over the bank like this.
What do you think?

[worker]

friend611,

It is hard to determine the velocity by the distance that the train goes after it derails, because it is almost impossible to determine the frictional lost for the sliding.

What do you mean by “It happened on a 13'13 curve”? I know we have a 5 degree curve.  13'13 looks like 13feet 13 inches.   Is it for the POD (Point of derailment)?

 Do you have the data for the spiral, trucks and carbodies? I like to do some calculations for the derailment.

[BaltACD]

The curve was 13 degrees 13 minutes

All that is available to know -

http://ntl1.specialcollection.net/scripts/ws.dll?file&fn=6&name=S%3A%5CDOT_56GB%5CRailroad%5CWEBSEARCH%5C3676.PDF

[friend611]

As for the cars- the train consisted of a heavyweight baggage car (60 feet or thereabouts), a RPO about the same length, four Pullman-Standard 85"streamlined coaches, a "tavern-lounge" rebuilt from a Pullman-Standard coach similar to the other coaches, a heavyweight dining car and three Pullman sleeping cars, one or two which may have been Budd-built 10-6 cars. Do not have data on the sleepers, though they may be similar in length to the coaches. As to the trucks I do not know. I suppose the Pullman-Standard cars used a standard truck, though I may be wrong on that. As to the spiral, it was 213 feet long. The superelevation of the curve went up to 4 1/2 inches, so that may give you some idea of the spiral. The 611 herself is well known, though she is (with tender) 109 1/2 feet long and 16 feet high, and had a calculated center of gravity of 77 inches.
lois

[Deggesty]

At the most, there was one heavyweight sleeper on the Pocahontas, a Norfolk-Chicago car; the other two sleepers were lightweight PS built 10-6's. I'm not at home, so I am working from memory. 

As a result of this wreck, the 611 was extensively rebuilt so that it was in better shape than any other J.

[BaltACD]

Another N&W derailment with a J in the lead

http://ntl1.specialcollection.net/scripts/ws.dll?file&fn=6&name=S%3A%5CDOT_56GB%5CRailroad%5CWEBSEARCH%5C3168.PDF

viewphoto.php?id=40477.jpg

viewphoto.php?id=40479.jpg

viewphoto.php?id=40480.jpg

viewphoto.php?id=40481.jpg

viewphoto.php?id=40482.jpg

[friend611]

These were the cars that followed the 611 onto the bank during the 1956 wreck: heavyweight baggage car 123 , heavyweight RPO #94 and P-3 class (Pullman-Standard, 1949) 539, 539 now in use in excursion service by the Watauga Valley Chapter NRHS. Finally,PM class (Pullman-Standard, 1941) 1727 was the final car on the bank. Two sisters of 1727, 1729 and 1734, derailed but did not go onto the bank. One of these straddled the double track main line while the other had one truck off the tracks with the rear truck remaining on the tracks. What surprises me is these last two coaches absorbed the momentum of the wreck, or the momentum died out so that no further cars were pulled off the tracks. All these cars appear to have had four wheel trucks (the P-S cars having roller bearings). However, the first two cars did not have tightlock couplers, a lack shared by another car in the train, the heavyweight diner.
lois

[worker]

Friend 611,

For the 13 degrees 13 minutes curve, R=435.717ft and speed=30mile/hour. Without super-elevation, vehicle (CoG: 77in.) overturn moment by centrifugal force only is about 2.648m and anti-overturn moment is about 7.007m. (m is the mass of the vehicle) With super-elevation, vehicle (CoG: 77in.) overturn moment by centrifugal force only is about 2.493m and anti-overturn moment is about 8.534m.

However, according to my analysis, there are other lateral forces will cause an overturn. The wheel/rail interaction (somewhat like steering) force, which is the same magnitude as the centrifugal force, will cause an overturn moment in curving. Also, the wheel set interaction force, which is about the same magnitude as the centrifugal force, will cause an overturn moment in curving entrance.

I still cannot perform an analysis in stringlining/jackknifing because of lack of data, to account for the adverse effect.

Nevertheless, we can say that a derailment is very possible.

I could not open the links. It says access denied in my computer.

http://ntl1.specialcollection.net/scripts/ws.dll?file&fn=6&name=S%3A%5CDOT_56GB%5CRailroad%5CWEBSEARCH%5C3168.PDF

[url]http://ntl1.specialcollection.net/scripts/ws.dll?file&fn=6&name=S%3A%5CDOT_56GB%5CRailroad%5CWEBSEARCH%5C3676.PDF[/ur]

[BaltACD]

http://ntl1.specialcollection.net/scripts/ws.dll?websearch&site=dot_railroads

 

Pick your year and then Railroad to find incidents.  Railroad is Norfolk & Western and incidents concerning the overspeed derailments of the J class engines occurred in 1947 & 1956

[mudchicken]

 13 degrees 13 minutes curve, R=435.717ft close(within a couple feet on radius), but no cigar.

Most likely, during the derailment, the curvature was decided by stringlining existing conditions.(three guys, a roll of string and a measuring stick/tape are a lot more available than a survey crew) Even running unbalance =0 and putting 6 inches of elevation (unheard of for that tight a curve), no passenger train can run that fast.

Any surfacing irregularity and that train is headed for the fences long before the rail rolls over, breaks or the cars overturn.

Uniform center of gravity? - Not a chance (even though a passenger train is a little more uniform than a freight train)

[cx500]

Or was it a 5 degree curve with spirals, total turned angle 13 degrees 13 minute?

John

[friend611]

I am not sure if the level of water in the 611's boiler would have had any effect on this or not. I suspect with the curve, the water may have sloshed to the inside, as this was a right hand curve. I do know the amount of coal and water remaining in the tender would have had some effect on the stability of the tender. The weather conditions, being the wreck happened past midnight in January, were very cold and snowy. The ground was frozen hard and icy, which resulted in the appliances as well as the running board being stripped off the fireman's side of the locomotive as she slid along the ground during the wreck.
lois

[worker]

Friends 611,

The water in the tender, I think, will not affect the train stability too much because the angle of super-elevation is very small. In some cases, if treated right, the water can be a damper to vibrations.

For the wreck on January 23, 1956, I could download the Report (No. 3676). From the sketch in the report, point of derailment on the curve is 66ft to the spiral. That means leading truck of the car was on the curve while trailing truck was on the spiral. Thus, there are three adverse effects to the derailment to be included in the analysis. 1). effect of spiral, 2) effect of stringlining/jackknifing and 3).effect of grade (0.03%) of the curve.

The data needed to do the analysis are 1). detailed radii of the spiral, 2) train marshaling, especially the cars connected to the derailed vehicle and 3). Truck and car-end (coupler) data. But data not available.

[Anonymous]

Wouldn't flange wear on a wheel possibly contribute to the wheel climbing the outer rail of a curve?

[friend611]

I do know that the first, second and eighth cars of the train (all heavyweights) were not fitted with tightlock couplers. It appears nearly all the cars in the train were fitted with four wheel trucks, mostly friction bearing except for the roller bearing Pullman-Standard trucks on the streamlined coaches, and perhaps roller bearing trucks on the Budd sleepers. The heavyweight diner is in question, though it may have been modernized as well.
lois

[friend611]

Something that may be of some help. Pictures of some of the cars that were in the wreck, from the N&WHS archives.
P3 class coach 539:
www.nwhs.org/archivesdb/detail.php?ID=100467
PM class coach:
www.nwhs.org/archivesdb/detail.php?ID=90254
M1 class RPO:
www.nwhs.org/archivesdb/detail.php?ID=17440
lois

[worker]

"Reply by Bucyrus

Wouldn't flange wear on a wheel possibly contribute to the wheel climbing the outer rail of a curve?"

---------------------------------------------------------------------------------------------------------------------------

Wheel flange/rail contact and wheel climb derailments are the very basic but very misleading subjects in the railroad industry. To understand these subjects, we must begin with the relatively new discovery. It is discovered that wheel flange/rail contact is actually a brake or a retarder. That means the contact will consume the pulling power of wheel to slowdown (or to stop) the wheel rotation. It is very dangerous that the train operator does not realize that his train is in braking because he didn’t do it.

With that been said, due to the discovery, it is found that there are two derailment modes in wheel flange/rail contact: derailment due to contact-brake and derailment due to wheel climb.

Contact-brake derailment mode will be initiated first during curving, as shown in analyses and tests. However, L/V (lateral force/vertical force) is not a factor in this failure mode.

Wheel climb derailment mode is the wheel going up not going down on the rail, and is a function of L/V. Therefore, Nadal’s L/V limit cannot be used as a criterion for this wheel climb derailment mode because Nadal’s L/V is for wheel going down on rail. Most importantly, Nadal’s L/V leads to misunderstanding why a train passes a curve with lubrication and fails without lubrication.

 As far as these two derailment modes are concerned, flange wear on a wheel will only change the flange angle and coefficient of friction. Though flange wear may cause some performance problems.

[Anonymous]

worker,

I understand your explanation of the two derailment modes, but I am surprised that the retarder mode is a new discovery.  My question about flange wear concerned the tendancy for the flange-to-rail contact area to increase as the flange wears.  It would also be contributed to by rail wear. 

In other words, the wear will bring the entire flange face into contact with the rail head rather than just the base fillet radius of the flange contacting the rail head.  And the action of the wheel rotation in this full flange contact will, in reaction to the rail, lift the wheel until the flange is able to ride over the top. 

[worker]

"In other words, the wear will bring the entire flange face into contact with the rail head rather than just the base fillet radius of the flange contacting the rail head.  And the action of the wheel rotation in this full flange contact will, in reaction to the rail, lift the wheel until the flange is able to ride over the top. "

Bucyrus,

Flange root will get wore off as the wear process goes on. Thus, the flange angle will increase. For the wheel on the rail, there is a critic angle at (or beginning from) which the wheel will never move up (climb) on the rail no matter how large is the lateral force. As an extreme case, if flange angle = 90 degrees, physics and intuition tells us that wheel will never move up on the rail no matter how large is the lateral force.

As shown previously, wheel flange/rail contact is a brake to the train and will consume the pulling power of the wheel. Thus, an anti-rotation frictional force F will be produced on the flange from the contact. In some cases, force F has a vertical component to lift the wheel.  However, the maximum force F is the force to consume ALL the pulling power of the wheel. Based on calculations, F_max = 0.1V approximately (V: vertical force). While only the vertical component of F_max contributes to lifting the wheel, so the force F is far away from enough to lift the wheel but can stop the wheel rotation. Thus, the wheel will skid or jump due to the inertial force and this is the failure mode due to contact-brake not wheel climb.

I believe that what you described most likely is the failure mode due to contact-brake not wheel climb.

[friend611]

Also of interest is how the 611 was able to jump completely over the eastbound tracks (after derailing on the westbound tracks) and at least for a few brief seconds, was airborne. It is my feeling she, derailing on a curve to the right, cleared the eastbound tracks and went over onto her side upon the ground, where she slid upon the river bank. She may have dropped down a bit, as the baggage car (the car directly behind the tender) was pointed directly vertically down the bank. The tender was turned upside down, with the trucks and wheels torn off.
llois

[worker]

Friends611,

Due to the stringlining/jackknifing, the locomotive was expected to be forced to displace about 5 inches laterally when on the spiral, depending on the train marshaling. Combining with other lateral forces, the overturn most likely occurred on the spiral before the curve. An airborne in the derailment is possible depending on the derailment modes. If the derailment did not occur near the bottom of the grade, 0.03% down grade would only help push the derailed locomotive further into the bank, otherwise, the grade will  play a part in the cause of derailment.

 

Train marshaling data and site-inspection would certainly help further analysis.

[friend611]

That makes sense. However, the line at that point was on a gradual downgrade.
lois