Kinky, man

3592 views
38 replies
1 rating 2 rating 3 rating 4 rating 5 rating
  • Member since
    May, 2005
  • From: S.E. South Dakota
  • 11,991 posts
Kinky, man
Posted by Murphy Siding on Monday, January 01, 2018 9:13 PM

     The February issue of Trains Magazine has a good article about the 7 physical phenomena that cause derailments... They tried to make the explanation straightforward but some things seem to have been glossed over a bit.

     Question- the part about sun kinks says that a mile of welded rail can grow 17" on the hottest days, causing potential problems. Were sun kinks not as big a problem with 39' stick rail? The way I figure it each bolt would have had at least 1/16" of play in order to get the bolt to slide through the hole. That would mean 1/8" of play every 39 feet, which equals 16.92 inches of *growth* before running into problems?


Thanks to Chris / CopCarSS for my avatar.

  • Member since
    June, 2001
  • From: US
  • 13,488 posts
Posted by Mookie on Monday, January 01, 2018 9:20 PM

I just never know what to expect on the other side of your initial posting...

She who has no signature! cinscocom-tmw

  • Member since
    May, 2005
  • From: S.E. South Dakota
  • 11,991 posts
Posted by Murphy Siding on Monday, January 01, 2018 9:46 PM

     It's that internet thing... 7 physical phenomena that cause derailments...   #4 will amaze you!!! Mischief

Thanks to Chris / CopCarSS for my avatar.

  • Member since
    December, 2001
  • From: Sunny (mostly) San Diego
  • 1,664 posts
Posted by ChuckCobleigh on Monday, January 01, 2018 9:53 PM

A few years ago there was a picture in the magazine of a steam-era heat kink, which was offered as an exemplar of such behavior in jointed rail.  According to the 1950 edition of Hay's Railroad Engineering text, joint bars could "freeze" to the rail from corrosion and thus not allow thermal expansion/contraction to be properly compensated for, thus resulting in kinks or pull-aparts.  (As an aside, in another post today, Jeff Hergert noted a number of pull-aparts in and around his territory.)  According to Hay, strategies to combat joint-bar freezing involved some form of lubrication.

  • Member since
    May, 2005
  • From: S.E. South Dakota
  • 11,991 posts
Posted by Murphy Siding on Monday, January 01, 2018 9:57 PM

     Interesting. You'd think the corrosion between the joint bar and the rail would be a pretty weak link that would let go pretty quickly when the rail grew or shrunk.

Thanks to Chris / CopCarSS for my avatar.

  • Member since
    September, 2003
  • From: California - moved to North Carolina 2018
  • 4,197 posts
Posted by DSchmitt on Monday, January 01, 2018 9:59 PM

the track structure lomits expansion and there was a discussion of welded rail vs jointed rail in at least one previous thread. 

I remember that the total epansion in a given length of connected jointed rails is more than than in the same length of welded rail because the welded rail is forced to expand in width and height instead of length. The track structure limits longitional expansion and there are only two ends to expand instead of many.

I tried to sell my two cents worth, but no one would give me a plug nickel for it.

I don't have a leg to stand on.

  • Member since
    September, 2003
  • 5,913 posts
Posted by Overmod on Tuesday, January 02, 2018 12:54 AM

Murphy Siding
Interesting. You'd think the corrosion between the joint bar and the rail would be a pretty weak link that would let go pretty quickly when the rail grew or shrunk.

Do not forget that the joint bars are secured laterally with clamping tension by the bolts, which perhaps themselves are rusted so they don't self-loosen.  The effect is to key the joint with corrosion, making it far more resistant to longitudinal motion, until some of the clamping force is relieved.  This was commonly done by section crews in the days of more hands-on line maintenance, with reference to the crack of rail ends coming together when enough bolt tension on the bars was relieved.

The 'lubrication' Hay mentions is partially to prevent intercorrosion in the joint, partially to assure movement when the bolts are loosened. 

  • Member since
    December, 2007
  • From: Georgia USA SW of Atlanta
  • 8,752 posts
Posted by blue streak 1 on Tuesday, January 02, 2018 3:43 AM

Dorango & Silverton NG had a sun kink a whil back.

Right now with the bitter cold pull aparts are becoming very prevelant.

  • Member since
    December, 2001
  • From: Northern New York
  • 18,590 posts
Posted by tree68 on Tuesday, January 02, 2018 8:26 AM

One must also consider the track bed.  Back in the day, as I understand it, ties were only half-deep in the ballast - longitudinal support was less than we find today.

Nowadays, the ballast is kept right to the top of the ties, which helps keep the track in alignment, which is one reason that welded rail is forced to expand outward, as discussed.  

Every now and then the expansion overwhelms holding ability of the ballast and you get one of those carnival ride kinks that show up occasionally on social media.

The slack in jointed rail may help deal with normal expansion, but provides than many more places where a pull-apart can occur.  It just depends on which set of bolts fails first...

LarryWhistling
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...

  • Member since
    May, 2003
  • From: US
  • 14,582 posts
Posted by BaltACD on Tuesday, January 02, 2018 8:35 AM

Welded rail is a technology that the railroads have yet to MASTER in all of the various weather conditions.  When I was working, the first cold snap of the season ALWAYS produced a high number of broken rails and pull-a-parts when it hit.  Trains moving over the territory were constantly leaving on track occupancy indicators on CTC model boards all over the system - when Signals & Roadway investigated the track circuits they found broken rails and pulled apart rail joints.  In unsignalled territory the signal maintainers got called out for crossing protection operating without the presence of trains.

When the heat would hit during the summer you would end up with Sun Kinks - and the signal system did not help identify them, they would have to be seen.

         

Never too old to have a happy childhood!

  • Member since
    September, 2003
  • From: California - moved to North Carolina 2018
  • 4,197 posts
Posted by DSchmitt on Tuesday, January 02, 2018 8:51 AM

https://www.railpage.com.au   has discussions of welded rail .

 https://www.railpage.com.au/f-p1189924.htm 

https://www.railpage.com.au/f-p1284213.htm

                       

 

 

 

I tried to sell my two cents worth, but no one would give me a plug nickel for it.

I don't have a leg to stand on.

  • Member since
    December, 2006
  • 1,135 posts
Posted by diningcar on Tuesday, January 02, 2018 9:23 AM

Will MC or Paul please furnish their analysis of the mitigating effect of rail anchors.

  • Member since
    November, 2013
  • 1,180 posts
Posted by VOLKER LANDWEHR on Tuesday, January 02, 2018 9:40 AM

How do you get to 16.92 feet?

1 mile is 5280 feet, is about 135 pieces of 39' length. Thus the total play on one mile is about 17 inches.

Here is an interesting presentation about the differences in thermal behavior of jointed and welded rail following European standards:
https://pwayblog.com/2017/03/01/rail-thermal-forces-for-jointed-and-cwr-track/

I'm not sure how it translates to American practices. The physical pasics are the same but the fastener systems differ and thus the longitudinal resistance might differ.
Regards, Volker

Edit: Rails are usually fastened with elastic clips to the ties in Europe. These clips provide a vertical pressure between rail and tie and by resulting friction a resistance against longitudinal movement of the rail by changes of temperature or outer forces.

In spiked track the rail anchors are intended to do the same. How effective they are I can't answer

Here is some information about thr requirements:

www.rta.org/assets/docs/TieReports/tiereport6.pdf

https://www.arema.org/files/library/2008_Conference_Proceedings/Resistance_to_Rail_Creep-What_Do_Rail_Fastenings_Really_Have_to_Do_2008.pdf

 

  • Member since
    May, 2005
  • From: S.E. South Dakota
  • 11,991 posts
Posted by Murphy Siding on Tuesday, January 02, 2018 11:35 AM

VOLKER LANDWEHR

How do you get to 16.92 feet?

 

Oops! Typing error (and proofreading error) on my part. That should read 16.92 inches. Thanks for pointing that out. I went back and edited it.

Thanks to Chris / CopCarSS for my avatar.

  • Member since
    October, 2006
  • From: Allentown, PA
  • 9,383 posts
Posted by Paul_D_North_Jr on Tuesday, January 02, 2018 9:13 PM

Overmod
Murphy Siding
Interesting. You'd think the corrosion between the joint bar and the rail would be a pretty weak link that would let go pretty quickly when the rail grew or shrunk.

The 'lubrication' Hay mentions is partially to prevent intercorrosion in the joint, partially to assure movement when the bolts are loosened. 

I'm not aware of any modern-day practice - say, after 1940 - to loosen and lubricate bolts or joints generally so as to allow the rail freedom to move.  That might have been done in a specific instance to relieve the accumulated thermal forces or slippage of the rail (sometimes called "running") - either tension or compression - in the rails at a certain location at a particular time, but not elsewhere in the absence of an evident problem.  Work on bolts and joints in my experience and knowledge was always to tighten them.  Lubrication was generally applied only at the time of laying of the rail, rarely afterwards.  

All that said, others may have had different experience and knowledge. 

As to rail anchors, their purpose is to keep the rail from sliding past a tie by use of a mechanical compression (only) fixture - it clamps around the base of the rail and bears against the side of the tie.  Without them, there's very little resistance to the rail sliding at a tie - mainly the friction from whatever weight of the rail is resting on the tieplate and hence that tie, plus perhaps a little bit from the spikes if they're tightly driven.  With a rail anchor, any motion of the rail then also pushes against the tie and essentially has to drag the tie along with it.  The tie moving encounters resistance from the ballast that it and the weight of the rail sit on, plus the ballast across most of the length and the entire depth of opposite side of the tie.  With a full-depth ballast section up to the top of the tie - a configuration not common with jointed rail back in the day - that resistance can be considerable, certainly enough to keep the rail from moving most of the time, as miles of CWR demonstrate every day.

- PDN.     

"This Fascinating Railroad Business" (title of 1943 book by Robert Selph Henry of the AAR)
  • Member since
    December, 2001
  • From: Sunny (mostly) San Diego
  • 1,664 posts
Posted by ChuckCobleigh on Tuesday, January 02, 2018 11:38 PM

Paul_D_North_Jr
All that said, others may have had different experience and knowledge. 

That being said, this accident report concerning the Amtrak Auto Train hitting a sun kink in 2002 near Crescent City FL has a detailed discussion of track dynamics and structures.  I think I've linked to it before but it is an interesting read on how track can "go bad" unexpectedly.

  • Member since
    December, 2001
  • From: Denver / La Junta
  • 9,289 posts
Posted by mudchicken on Tuesday, January 02, 2018 11:40 PM

Ya know how hard it is to find a working mutiple bolt/ joint tightening machine anymore? - Those rascals are dinosaurs in a CWR world. (I had some 90 MPH territory with 132 CWR (laid 1954) in western KS that was a constant headache - no lubrication there unless you were backing off the bolts with penetrating oil to change out bars or a dutchman/bolt hole break.)

Anchors also solve the skew tie problem that used to be a big problem. rails tend not to run in anything resembling a uniform manner. However the anchors have to work and finding anchors for less than 112/115# rail is a problem. Those rails and turnouts have bigger headaches these days because they do not make the smaller anchors and used anchors are shot. (just slide along the B/R and fall off eventually).

Mudchicken Nothing is worth taking the risk of losing a life over. Come home tonight in the same condition that you left home this morning in. Safety begins with ME.... cinscocom-west
  • Member since
    December, 2006
  • 1,135 posts
Posted by diningcar on Wednesday, January 03, 2018 9:17 AM

Question - if rail anchors diminish or somewhat control the linear movement of rail  caused by temporature change then rail must adjust by - getting taller and fatter within the confines of the area between the applied anchors?  Further analysis please MC, Paul and any who choose to contribute. 

  • Member since
    December, 2005
  • From: Cardiff, CA
  • 2,890 posts
Posted by erikem on Wednesday, January 03, 2018 10:30 AM

From Armstrong's book, The Railroad, what it is, what it does, the "secret" for CWR is laying it at near max temp so it is usually under tension. As temperatures drop, the height and width are reduced slightly (IIRC subject to Poisson's ratio to exactly how much they are reduced). Conversely the rail should expand in cross section when the rail is above neutral temperature.

I suppose it would be possible that the neutral temperature could be set lower if there was a reasonable way to constrain the track structure in the same way that guy wires keep a guyed radio tower from buckling. (Had an 80' foot tower in the back yard of my first house.)

And please keep in mind that I am not a structural engineer.

  • Member since
    January, 2014
  • 4,938 posts
Posted by Euclid on Wednesday, January 03, 2018 10:44 AM

 

If a single stick of rail is heated, it will expand in all directions by a certain percent.  So it will get longer, taller, and wider according to the percent of expansion. 

 

If you constrain it lengthwise so it cannot get longer due to expansion; does it then get wider and taller than it would have, had it not been prevented from expanding lengthwise? 

 

That has been my understanding, but I am not sure if it is correct.

 

If this is true, then the rail grows in height and width for two different reasons: 

 

  1. The direct thermal expansion of height and width if the length is not prevented from expanding. 

     

  2. Compressive, resilient bulging of height and width if the length is prevented from expanding.

     

 

Under condition #2, if without any cooling, you suddenly removed the constraint of the length; the length would get longer and the height and width would get smaller. 

 

  • Member since
    November, 2013
  • 1,180 posts
Posted by VOLKER LANDWEHR on Wednesday, January 03, 2018 11:03 AM

erikem
From Armstrong's book, The Railroad, what it is, what it does, the "secret" for CWR is laying it at near max temp so it is usually under tension.

If it were this easy. As the rail has to carry vertical loads and according bending stresses additionally to the temperature stresses track welding is a balancing act depending on load and climate.

The temperature at which the rail gets welded is choosen so that tensile strength is not exceeded at the lowest expected temperature and compressive strength not at the highest expected temperature. That neutreal temperature can be different at different locations.

As you said the Poisson ratio rules how much the cross section changes.

The whole track structure (subroadbed, ballast, and ties) keep the rail from buckling. But sometimes the stresses get high and a rail breaks or develops a sun kink (buckles).
Regards, Volker

 

  • Member since
    October, 2006
  • From: Allentown, PA
  • 9,383 posts
Posted by Paul_D_North_Jr on Wednesday, January 03, 2018 10:17 PM

Poisson's Ratio is the amount of transverse 'stretch' to the distance of the longitudinal 'stretch'.  For steel, the ratio is usually about 0.3.  Thus, if a rail expands 0.10% along its length (made-up figure, for illustration only), it will shrink by about 0.03% transversely.  For more on this, see:

https://en.wikipedia.org/wiki/Poisson%27s_ratio 

For steel, the coefficient of thermal expansion is about 7.2 x 10^-6 per degree F.  For a 100 degree change in temperature, the steel will expand about 7.2 x 10^-4 (0.00072), which is about 7.2 x 10^-2 % (0.072 %) of the distance under consideration.  Applying Poissons Ratio means the transverse contraction will be about 2.2 x 10^-4 or 2.2 x 10^-2 %.  For a 6" wide base of rail, the change in width would be 6" x 0.00022 = 0.00132" (if I'm doing my math correctly this late at night).  You'd need a micrometer, fine vernier caliper, or a similar instrument to see that small of a change.  

- PDN. 

"This Fascinating Railroad Business" (title of 1943 book by Robert Selph Henry of the AAR)
  • Member since
    December, 2005
  • From: Cardiff, CA
  • 2,890 posts
Posted by erikem on Wednesday, January 03, 2018 10:43 PM

VOLKER LANDWEHR

 

erikem
From Armstrong's book, The Railroad, what it is, what it does, the "secret" for CWR is laying it at near max temp so it is usually under tension.

 

If it were this easy. As the rail has to carry vertical loads and according bending stresses additionally to the temperature stresses track welding is a balancing act depending on load and climate.

Climate as in the northern High Plains? My dad's home town, Miles City, was -30F the morning of Jan 1, and summer air temps can be over 100F. Coastal southern California is a lot gentler on CWR as the temperature spread is a bit over half of eastern Montana. One could make the rail out of Invar, but that would probably open up an even bigger can of worms, besides being ludicrously expensive...

I've seen a benign case of "string-lining" on the loop line at the Orange Empire Railway Museum, where the corners of the loop pull in by about an inch on summer evenings.

  • Member since
    December, 2007
  • From: Georgia USA SW of Atlanta
  • 8,752 posts
Posted by blue streak 1 on Wednesday, January 03, 2018 11:05 PM

PDN  Would your micrometer have to be at the same temperature for both measurements ? Or can it be calibrated before each measurement ?

About pull aparts.  If a train runs over a pull apart or at least part of a train how much damage will happen to the wheels for various distances ?  That of course would depend on how much distance the pull apart is.  Certainly more than a low rail joint.  Wheel rail confence question ?

  • Member since
    November, 2013
  • 1,180 posts
Posted by VOLKER LANDWEHR on Thursday, January 04, 2018 3:38 AM

erikem
Climate as in the northern High Plains? My dad's home town, Miles City, was -30F the morning of Jan 1, and summer air temps can be over 100F.

Yes, as one example. The rail can get even warmer than the ambient temperature. Then take the Mojave desert, in this case Death Valley, with low temperatures of 15°F and highs of 130°F. For both examples it makes sense to use a different neutral temperature at which in-situ welds are made.

The tendency will be to choose temperatures as high as suitable. Tensile stresses (braking) are better controllable than compressive stresses (buckling).
Regards, Volker

  • Member since
    July, 2008
  • 1,093 posts
Posted by rdamon on Thursday, January 04, 2018 5:26 AM

Found an interesting study on how to measure NT done in 2012.

https://www.fra.dot.gov/eLib/Details/L01645

 

  • Member since
    April, 2007
  • 1,584 posts
Posted by Convicted One on Thursday, January 04, 2018 8:43 AM

Murphy Siding
Were sun kinks not as big a problem with 39' stick rail? The way I figure it each bolt would have had at least 1/16" of play in order to get the bolt to slide through the hole. That would mean 1/8" of play every 39 feet, which equals 16.92 inches of *growth* before running into problems?

 

IIRC,  there was a Trains magazine article back during the Hemphill era that pretty much said that

  • Member since
    September, 2003
  • 5,913 posts
Posted by Overmod on Thursday, January 04, 2018 9:20 AM

blue streak 1
If a train runs over a pull apart or at least part of a train how much damage will happen to the wheels for various distances?

I suspect a more significant question is how much progressive damage will happen to the rail in the vicinity of the pullapart.  In at least some cases, the separation would occur at a stress raiser of only nominal added 'severity', meaning the other portions of the rail might be proportionally compromised at low temperature.  Likewise if there is pronounced gauge-corner cracking, corrugation, transverse fissures or other damage to the railhead, or I expect in some head-hardened rail types, the pounding over the separation is likely to induce further breakage either at points reasonably adjacent to the initial pull, or at logical vibration nodes in the balance of the CWR in track.  Since I suspect that many rails that break under trains go unrecognized for at least the passage of those trains' consists, quite a bit of cold-enhanced breakage might result, including unrecognized cracking damage in wheels or axles (much of which might only be detectible by good NDT and not the usual sort of car-knocking inspection)

  • Member since
    December, 2001
  • From: Northern New York
  • 18,590 posts
Posted by tree68 on Thursday, January 04, 2018 9:43 AM

blue streak 1
About pull aparts.  If a train runs over a pull apart or at least part of a train how much damage will happen to the wheels for various distances ?

It would make a difference how far the rails spread apart.  Recall that there is probably an inch gap in diamonds.  Yes, they are being repaired almost daily, but the wheels that are running over them don't seem to have any problems as a result of that jolt.

I think, too, that a pull-apart at a joint might be less trouble, to a point, than a break in a stretch of rail.  In most cases, only one side of the joint comes loose, so the rails remain in alignment.

A rail may break between ties, which is going to expose the failure to vertical flexing, which will compound as more wheels pass over it.

Then, again, recall the video of active efforts to derail a train by outright removing small sections of rail.  In general, it made for a bumpy ride, but they didn't end up with pick-up sticks... 

LarryWhistling
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...

  • Member since
    November, 2013
  • 1,180 posts
Posted by VOLKER LANDWEHR on Thursday, January 04, 2018 10:09 AM

In 1944 the Army tried to derail a train by blowing out pieces of rail. The results are surprising: https://www.youtube.com/watch?v=agznZBiK_Bs
Regards, Volker

Join our Community!

Our community is FREE to join. To participate you must either login or register for an account.

Search the Community

Newsletter Sign-Up

By signing up you may also receive occasional reader surveys and special offers from Trains magazine.Please view our privacy policy