Trains.com

Knuckle Coupler

73846 views
220 replies
1 rating 2 rating 3 rating 4 rating 5 rating
  • Member since
    April 2003
  • 305,205 posts
Posted by Anonymous on Monday, August 31, 2009 11:24 AM

Ed,

 

The photos clarify and remind me that those “force ridges” are not anywhere near perpendicular to the line of force, so the coupler cannot function as I previously assumed, pulling though those force ridges alone.  Even if the angle of contact with the force ridges were closer to perpendicular, it would require a binding friction to keep the force from transferring to the locking pin.  That friction would not be consistent or reliable, considering that oil can find its way to the contact surfaces either intentionally or not.

 

In the fifth photo you posted, I notice that there are some shiny areas around the coupler pivot bosses.  This suggests that although the coupler pivot pin does not carry any pulling force, the pivot bosses might.  Clearly they do transmit shoving force as evidenced by the shiny areas.  What the photo does not show, and what I wonder about, is what the boss surfaces that would be positioned to carry the pulling force look like.  Specifically, that would be the surface on the opposite side of the lower knuckle circular pivot boss in photo #5. 

 

I suspect that these bosses do not carry any pulling force and the surface hidden in the photo is not shiny.  If it were, it would seem to be a redundant force bearing point.  It would seem to be impossible to equalize (in this assembly) a pulling force between one contact point and another one that is redundant.  The force would go through one or the other. 

 

The compression force of shoving, however, does not have an effective load bearing point within the force ridges because such a force would cause the tang to deflect to the left.  And in doing so, it would pull away from the locking pin.  So there would be no "bottoming out" for compression through those features. 

 

Thus the knuckle pivot bosses transferring the compression force (as evidenced by the shiny spots) apparently are the only features transferring that force.  To verify this assumption, it would be interesting to see more clearly the backward facing side of the knuckle tang groove and discover whether or not it is shiny.  I suspect that it is not shiny.  In fact, with its large fillet radius, it does not appear to be developed to transmit the compression force of shoving.  In addition, the forward surface of the raised boss feature of the coupler housing does not look shiny, as it would be if it were carrying the compression force.

 

Also, I looked at your other photos, and that is some fine work you are doing with your pen making.  That looks like more than just a hobby.  It seems like there would be a lot of potential to expand that type of craftsmanship and art into other products.  I have a little wood shop, and have been thinking about producing some kind of saleable product. 

  • Member since
    April 2003
  • 305,205 posts
Posted by Anonymous on Tuesday, September 1, 2009 5:19 PM

Here is my revised current analysis of the coupler function, as I understand it, as clearly as I can express it.  This replaces my understanding as described in the second post of page 11.

 

HOW PULLING FORCE IS TRANSFERRED THROUH AN AUTOMATIC RAILROAD COUPER

 

All designation of right / left, and forward / backward is based on viewing one uncoupled coupler from the top, facing the knuckle.

 

On a single coupler with an open knuckle, the knuckle is resting its weight on the coupler body through the lower circular boss of the knuckle pivot pin, and the knuckle is loosely attached to the coupler body by the knuckle pivot pin.

 

Both the top and bottom of the knuckle tang features a groove that follows a radius that has its center on the axis of the knuckle pivot pin.  Also, above and below the knuckle tang, there is a raised ridge boss feature on the interior surfaces of the coupler body that matches the knuckle tang groove in a loose fit.  The raised bosses are positioned within the tang grooves when the knuckle is closed, and the grooves swing clear of the raised bosses when the knuckle is open. 

 

The knuckle tang has a flat surface that contacts a corresponding flat surface on the locking pin.  The contact plane of these two surfaces is a vertical plane lying parallel with, and approximately coincident with the pulling force centerline of the coupler.

 

When the knuckle is closed, the locking pin drops into position, and the knuckle tang grooves come into alignment with the corresponding ridge bosses on the coupler body.  At this point the knuckle fits loosely with its pivot pin, with the locking pin, and in the relationship between the tang grooves and the corresponding ridges on the coupler body.

 

When a pulling load is applied, the knuckle is at first, momentarily inclined to rotate back open because the pull is on the centerline, and the knuckle pivot pin is offset to the left.  Moreover, the knuckle is resting on the lower circular boss of its pivot pin.  But that rotational movement is arrested by the locking pin, which blocks the rightward swing of the coupler tang.  Upon the tang being obstructed by the locking pin, a very slight rotational reaction force may be created, which forces the knuckle to bear leftward against the knuckle pivot pin.

 

However, both the primary force of the tang against the locking pin and the reaction force of the knuckle against its pivot pin are very small because they only result from holding the knuckle against the rotation of its own weight at the instant that the pulling force is applied. 

 

At that point, while the knuckle is locked from rotation, it is free to slide forward slightly in line with the direction of pull.  Because the contact between the tang and the locking pin is in a plane parallel to the pulling centerline, the locking pin cannot prevent the knuckle from sliding forward.  So it will slide forward until the slack is pulled out of the sloppy relationship between the tang grooves and the corresponding bosses on the coupler body. 

 

More specifically, when pulling the coupler, the rear surfaces of the raised bosses contact the rear surfaces of the grooves.  The contact between the bosses and groove walls is vertically cylindrical, following the arc shape of these features.  These features, when engaged with the knuckle closed, lie to the left of the locking pin, and in the same general front-to-back proximity as the locking pin.  The chord of the arc of these features, beginning at its far left end, comes forward, and converges on the coupler pulling centerline at an angle of approximately 45 degrees.  Therefore, when pulling on the coupler, the knuckle tang is crowded to the right by the wedging deflection of the bosses and grooves.

 

Up until this point, the knuckle has been a rotating element.  But when solid contact is made between the tang grooves and the corresponding boss features on the coupler body, the knuckle ceases to be a rotating element, and becomes a linear link, transmitting the pulling force in a straight line like a link of a chain.  Once this function change happens, the pulling force no longer induces the knuckle to rotate.  The rotation function of the knuckle is not a part of the pulling function of the coupler.  Rather, the rotation function of the knuckle only serves to install or un-install the knuckle as a linear, straight line, pulling link.

 

The rear surfaces of the raised bosses on the coupler body and the tang contact surface (left side) of the locking pin together constitute two opposing sides of a tapered pocket that narrows as it comes forward.  Upon pulling the tang forward, its rear surfaces of the grooves and the surface that contacts the locking pin constitute a wedge that that bottoms out in the tapered pocket and transfers its pulling force to the features that constitute the tapered sides of the pocket. 

 

~ ~ ~

  • Member since
    March 2002
  • 9,265 posts
Posted by edblysard on Tuesday, September 1, 2009 6:07 PM

Uhhh...

Yup.

What he said.Big Smile

23 17 46 11

  • Member since
    June 2008
  • From: Salem, Oregon
  • 189 posts
Posted by NP Red on Tuesday, September 1, 2009 6:34 PM

 http://www.google.com/patents/about?id=0tstAAAAEBAJ&dq=2948414

Here is a site that has diagrams from the original patents of several applicable models.  Click on the individual patent numbers to see the diagrams.  Note the dates off to the right for reference.  The Metzger model is one of the most revealing diagrams, IMHO. 

  • Member since
    August 2005
  • From: At the Crossroads of the West
  • 11,013 posts
Posted by Deggesty on Tuesday, September 1, 2009 7:35 PM

Bucyrus
Here is my revised current analysis of the coupler function, as I understand it, as clearly as I can express it.

Having read Bucyrus' dissertation on the operation of the coupler knuckle, I move we grant him the degree of Doctor of Philosophy in Couplerology.

Johnny

Johnny

  • Member since
    April 2003
  • 305,205 posts
Posted by Anonymous on Tuesday, September 1, 2009 7:42 PM

Deggesty

Bucyrus
Here is my revised current analysis of the coupler function, as I understand it, as clearly as I can express it.

Having read Bucyrus' dissertation on the operation of the coupler knuckle, I move we grant him the degree of Doctor of Philosophy in Couplerology.

Johnny

Thanks Johnny.  I would be honored.

  • Member since
    January 2004
  • From: Frisco, TX
  • 483 posts
Posted by cordon on Tuesday, September 1, 2009 8:28 PM

Smile

Thank you, Bucyrus.  I agree wholeheartedly.  Last week I drew a schematic diagram to show the "wedging" of the coupler tang between the lock and the bosses on the coupler body.  For clarity, I left out the body of the knuckle and the ridges next to the knuckle pin.  The arrows indicate the forward and backward forces.  "Ridges" are as indicated in my photos way back on Page 4 (I think).

 

I hope this accurately illustrates your description.

Smile  Smile

 

  • Member since
    April 2003
  • 305,205 posts
Posted by Anonymous on Tuesday, September 1, 2009 9:11 PM

cordon

Smile

Thank you, Bucyrus.  I agree wholeheartedly.  Last week I drew a schematic diagram to show the "wedging" of the coupler tang between the lock and the bosses on the coupler body.  For clarity, I left out the body of the knuckle and the ridges next to the knuckle pin.  The arrows indicate the forward and backward forces.  "Ridges" are as indicated in my photos way back on Page 4 (I think).

  

I hope this accurately illustrates your description.

Smile  Smile

 

cordon,

I think your sketch shows it perfectly.  I have been referring to the force ridge features as a boss on the coupler body and a groove on the knuckle tang.  However it is not really a groove on the tang.  It is actually more of a boss with a hook-like configuration to it.  At one point, I was thinking that it was a groove on the tang that could transmit both pulling and shoving force to the boss on the coupler body, but it is not a groove.  So I don’t see these features transmitting the shoving force.  Even if it were a groove on the tang, the features could not work in conjunction with the locking pin when transmitting a shove loading.

  • Member since
    February 2002
  • From: Muncie, Indiana...Orig. from Pennsylvania
  • 13,456 posts
Posted by Modelcar on Tuesday, September 1, 2009 10:15 PM

Bucyrus
 

HOW PULLING FORCE IS TRANSFERRED THROUH AN AUTOMATIC RAILROAD COUPER

Wow.......Talk about a complete technical explanation.....I actually was able to {very slowly}, work thru those paragraphs and get to the end and really see the line of pulling force that comes thru the coupler assy.

Now I wonder if I can remember it, but that was excellent.

The designer must have been some sort of a genus to figure /design and bring it to reality.

Quentin

  • Member since
    August 2008
  • From: Calgary AB. Canada
  • 2,298 posts
Posted by AgentKid on Tuesday, September 1, 2009 11:33 PM

Deggesty

Having read Bucyrus' dissertation on the operation of the coupler knuckle, I move we grant him the degree of Doctor of Philosophy in Couplerology.

Johnny

I heartily second that motion.Thumbs Up

I too, mostly understood what was said if I read it slowly, and if there were parts I didn't agree with I wouldn't know how to express myself that well anyway. Well done.

AgentKid

 

So shovel the coal, let this rattler roll.

"A Train is a Place Going Somewhere"  CP Rail Public Timetable

"O. S. Irricana"

. . . __ . ______

  • Member since
    March 2002
  • 9,265 posts
Posted by edblysard on Wednesday, September 2, 2009 7:07 AM

What he described is a double ended hook or S hook...the knuckle is basicly that...under load the interior part, the tang on the knuckle, hooks into a pocket inside the coupler body and the exterior hook, the coupler face, hooks into the opposing knuckle and coupler....under tension they lock into each other and their respective coupler bodys, under slack they can rotate and open when the lock pin is removed/lifted.

A rather simplistic description, but thats the basics.

23 17 46 11

  • Member since
    April 2003
  • 305,205 posts
Posted by Anonymous on Wednesday, September 2, 2009 9:37 AM

AgentKid

Deggesty

Having read Bucyrus' dissertation on the operation of the coupler knuckle, I move we grant him the degree of Doctor of Philosophy in Couplerology.

Johnny

I heartily second that motion.Thumbs Up

I too, mostly understood what was said if I read it slowly, and if there were parts I didn't agree with I wouldn't know how to express myself that well anyway. Well done.

AgentKid

 

Thanks AgentKid.  I think Ed has summed it up nicely in the preceding post.  This thread had covered a lot of ground in developing an understanding how a coupler actually functions.  I have learned a lot here after beginning with a considerable misconception of the coupler function.  My main surprise was that the coupler pivot pin does not carry any pulling load.  Upon learning that, I have been struggling to understand how that is possible.  In the beginning, the first Janney coupler did involve the knuckle pivot pin with carrying the pulling load.  I don’t know when this basic principle changed in practice, but it does seem like an ingenious development. 

  • Member since
    October 2006
  • From: Allentown, PA
  • 9,810 posts
Posted by Paul_D_North_Jr on Wednesday, September 2, 2009 10:01 AM

Concur.  Bucyrus has been a real 'bulldog'/  'pit bull' on this one - just hasn't let go.  We've all benefitted from his inquisitiveness, persistence and diligence with the various questions, explanations, and revisions of same.  Thank you !   Thumbs Up  Bow  

And thanks too to all of those who have taken the time to reply and explain to us uninitiated fans, esp. edblysard

- Paul North. 

"This Fascinating Railroad Business" (title of 1943 book by Robert Selph Henry of the AAR)
  • Member since
    September 2009
  • 1 posts
Posted by charpent on Wednesday, September 2, 2009 12:59 PM

In the august 2000 issue of Trains magazine is a very good article called Making the Connection by Richard Dawson. His authorship is adapted from the "Coupling Systems" section of the 1997 edition of "Simmons-Boardman Car & Locomotive Cyclopidia". The article is extremely informative about the evolution of coupler technology. Some of the descriptions are quite scientific especially how energy absorption is dealt with, but you'll get the idea. Buff (push) and draft (pull).

  • Member since
    April 2003
  • 305,205 posts
Posted by Anonymous on Thursday, September 3, 2009 11:27 AM

Charpent,

 

I almost seem to recall seeing that article by the title you mention, but I don’t recall anything about the content.  I might even have that issue, so I will start digging for it.  It is interesting that, although we have resolved the mystery of how a coupler transfers pulling force, there is an entirely different mechanism and set of principles that are dedicated to the coupler transferring shoving or buff force.  It is not just a reversal of the pulling mechanism.

 

There is some other functionality pertaining to the locking pin and knuckle thrower that does not seem too mysterious, but I could not explain it or draw a picture of the details.  When you pull the cut lever, it pulls the pin (locking pin).  When you let go of the cut lever, it drops back down, but the pin stays pulled.  If you pull the cut lever up with an energetic jerk, the knuckle thrower mechanism opens the knuckle.  However, if you pull the cut lever up slowly, it will pull the pin without throwing the knuckle open. 

 

If a knuckle is open and you close it by hand or close it by coupling to another coupler, the pin will drop. Here is something that I have never seen demonstrated, so I don’t know what would happen:

 

Say you have a cut of cars with the slack bunched and you pull a locking pin on one coupler.  Now say you pull the cars just about an inch or so, but not enough to begin to open the knuckle with the pulled pin, and then stop.  Now push back and bunch the couplers back together.  Will that pulled pin drop?

 

It is easy to see the knuckle tang holding the pin up once the tang swings with the opening of the knuckle.  Then when the knuckle would close, the tang would shift back and allow the pin to drop.  But that cannot be the way it works.

 

The knuckle tang does not shift its position upon pulling a pin if the couplers are coupled and the knuckle cannot open, yet the pin hangs lifted when pulled.  Therefore, I assume that the pin must shift its position when pulled, so it can go into a suspended condition by shifting onto some kind of hanging feature. 

 

Since the pin hangs without the knuckle opening, it is hard to see what knocks the pin off of its hanger when an open knuckle closes and the tang returns to the position where the locking pin originally hung lifted.  Occasionally, a locking pin will not hang pulled, so you have to ride with it and hold it up for a kick.  I assume something must be worn out for this failure to occur.  

 

Thanks NP Red for posting that link to coupler patents.  I agree that the METZGER patent drawing of the top view clearly illustrates the relationship of the mechanical features that we have been discussing here.  It shows the exact configuration of the features that cordon shows in his drawing above.  However, the patent itself appears to deal with an improvement in the knuckle thrower. 

 

Regarding the pulling features we have been discussing, the key feature is that the knuckle pin carries no load.  This represents a big departure from the original Janney coupler of the 1800s where the knuckle pin did carry pulling load.  Today the knuckle pin is an unused, load-carrying backup feature.  I wonder when this sea change occurred.  It would be interested in hearing how the inventor described the reasoning for it.

 

Overall the patents linked by NP Red make one realize that for as sophisticated and long developed the standard automatic coupler is, there are still a lot of people trying to improve it.  It is hard to analyze the patents and determine whether they have, or have not, been embodied in today’s coupler.  Most probably some of them or parts of them have been. 

 

It is mind boggling to consider the amount of deep thought that has gone into the development of couplers from the link and pin coupler forward.  There were lots of designs and patents on just link and pin couplers.  Other than the well known safety hazard with link and pin couplers, one of their biggest problems was that they were not standardized.  Their parts were not necessarily interchangeable.  Another problem was that they had loose, separable parts.  Links and pins walked away, and some of them are still hiding out there today. 

 

Crews hoarded links and pins because they never knew what non-interchangeable hardware might be needed for cars that they had to pick up.  And such cars often were stripped of their hardware by the crew that left them.  Once the railroads began looking for a new coupler, there was a blizzard of patents on new versions of automatic couplers.  It took a great effort for the industry to arrive at, and settle on the Janney design.  And that design has evolved ever since.   

  • Member since
    March 2002
  • 9,265 posts
Posted by edblysard on Thursday, September 3, 2009 6:45 PM

Look at the diagram Larry posted on page 10...note the lock pin is "wedge" shaped, rather top heavy, and has a lip on the back side.

The cavity in the coupler body is close to the same shape, the lock pin leans backwards at a slight angle, when you lift the cut lever, the lock pin rides up and the lip on the back side slides back and rest on a corresponding shelf in the coupler body and it stays up...the lip on the lock pin can wear down, and this causes the lock pin to slip or fall back forward and drop, so it wont stay up on its own.

Hold the cut lever up and open the knuckle, and the knuckle tang will hold the lock back and up.

The thrower hook serves two purposes, it throws the knuckle open, and when the knuckle is shoved close, the thrower rotates around and the back side or rear part of the thrower flips the lock pin forward, gravity does the rest.

Even minus the thrower, momentum will cause the lock to fall forward and drop...when you bring a 70,000lb hopper against a 70,000lb hopper, there is a good amount of force involved.

 

23 17 46 11

  • Member since
    April 2003
  • 305,205 posts
Posted by Anonymous on Thursday, September 3, 2009 9:17 PM

Thanks for that explanation Ed.  So, as I understand you, the lock pin hangs up because its head drops over a shoulder when the lock pin is lifted.  And it would hang up when pulled, even if the knuckle did not open.   But the knuckle does have to open and then close in order for the knuckle to work through the knuckle thrower to dislodge the hanging pin and cause it to drop.

  • Member since
    March 2002
  • 9,265 posts
Posted by edblysard on Thursday, September 3, 2009 10:20 PM

Yes, although a hard impact can cause the lock pin to fall forward without the thrower function and drop, which is why cars can sometimes couple back into each other at real slow speeds.

23 17 46 11

  • Member since
    June 2005
  • From: Phoenixville, PA
  • 3,495 posts
Posted by nbrodar on Friday, September 4, 2009 6:10 PM

 If you slam them too hard the cars can ricochet away from each other without dropping the pin. 

Sometimes just the opposite happens.  You make a cut, and breath on the coupler wrong and pin drops on you.  This is particularly annoying when you pull the pin on a cut of cars before coupling up to the track.

Nick

Take a Ride on the Reading with the: Reading Company Technical & Historical Society http://www.readingrailroad.org/

  • Member since
    March 2002
  • 9,265 posts
Posted by edblysard on Friday, September 4, 2009 8:15 PM

And every yard has one or two of those tracks where you kick a car in there, and it just dosen't sound right when it couples up...but it looks like it is staying put...right up to the time you have to shove a cut in the track right next to it, then it decides to start rolling back at ya!

23 17 46 11

  • Member since
    June 2005
  • From: Phoenixville, PA
  • 3,495 posts
Posted by nbrodar on Saturday, September 5, 2009 8:12 AM

 I hate it was that happens...

Given the current disciplinary bent of my employer, I shove every car to couple.  It's a real pain and takes forever, but I'm not getting 30 days for having a car roll back on me.

I find overzealous engineers are the biggest hindrance to good couplings. 

Nick

Take a Ride on the Reading with the: Reading Company Technical & Historical Society http://www.readingrailroad.org/

  • Member since
    April 2003
  • 305,205 posts
Posted by Anonymous on Saturday, September 5, 2009 9:39 AM

nbrodar

 

I find overzealous engineers are the biggest hindrance to good couplings. 

Nick

What do you mean by that?

  • Member since
    March 2002
  • 9,265 posts
Posted by edblysard on Saturday, September 5, 2009 6:16 PM

Because if you couple up too hard the knuckle bind the pin up, and one or both fail to drop and lock.

You need just a little slack after the coupling to let everything "relax" into place.

Then you have to take it ahead, make a one car gap between the tow car, open one of the knuckles and couple back up.

Thats why we always stretch the joints, (couplings) to make sure they hooked up...nothing is more of a pain in the keister than making a joint, lacing the hoses and cutting in the air, then heading back to the locomotive, get the highball from the yard, and going into emergency as soon as you move three feet.

A good engineer will bring 'em back, let the slack run out and coast to the joint, timing it just right with the conductor so all you hear is the knuckle closing and the pin dropping...takes a little skill and working with the same conductor a while to get it down pat...a overzealous engineer will simply back up till he runs into the cars and then stop.

23 17 46 11

  • Member since
    June 2005
  • From: Phoenixville, PA
  • 3,495 posts
Posted by nbrodar on Saturday, September 5, 2009 6:59 PM

 You beat me to it Ed!   Exactly.

Nick

Take a Ride on the Reading with the: Reading Company Technical & Historical Society http://www.readingrailroad.org/

  • Member since
    April 2003
  • 305,205 posts
Posted by Anonymous on Saturday, September 5, 2009 7:42 PM

Nick and Ed,

I see what you mean by "overzealous."

  • Member since
    October 2006
  • From: Allentown, PA
  • 9,810 posts
Posted by Paul_D_North_Jr on Thursday, February 4, 2010 1:58 PM

 Time to revive this thread . . .

Here's a link to a neat photo that shows a the body of a coupler in cross-section - because the durn thing broke while switching !

 http://www.railpictures.net/images/d1/0/2/5/5025.1264712710.jpg  

The comments underneath are interesting, too, as a 'post mortem' of why it failed.

- Paul North.

 

"This Fascinating Railroad Business" (title of 1943 book by Robert Selph Henry of the AAR)
  • Member since
    April 2003
  • 305,205 posts
Posted by Anonymous on Thursday, February 4, 2010 3:11 PM

That's a great photo, but I don't see any comments that open with it.

  • Member since
    May 2003
  • From: US
  • 25,292 posts
Posted by BaltACD on Thursday, February 4, 2010 5:13 PM

Paul_D_North_Jr

 Time to revive this thread . . .

Here's a link to a neat photo that shows a the body of a coupler in cross-section - because the durn thing broke while switching !

    

The comments underneath are interesting, too, as a 'post mortem' of why it failed.

- Paul North.

 

I notice that the drawbar cross key seems to have been recently installed as the locking parts and the cotter key holding it all together appear fresh without any rust or road grime on it.  I suspect this car has been a problem for a while and it's handling during these problems fatigued the coupler itself causing the fracture.

Never too old to have a happy childhood!

              

  • Member since
    July 2003
  • From: Elmwood Park, NJ
  • 2,385 posts
Posted by trainfan1221 on Thursday, February 4, 2010 6:28 PM

That is an interesting picture..takes broken knuckle to the next level.

 

  • Member since
    February 2002
  • From: Muncie, Indiana...Orig. from Pennsylvania
  • 13,456 posts
Posted by Modelcar on Thursday, February 4, 2010 6:46 PM

Paul_D_North_Jr

Here's a link to a neat photo that shows a the body of a coupler in cross-section - because the durn thing broke while switching !

.....Sure is a good large, clear photo of it.

Quentin

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