Cars not uncoupling at crest of the hump

[nkalanaga]

"that should not be a problem in a hump yard unless you intend to pull the cars out of the yard back over the hump after they are classified."

The NP/BN did that quite often at Pasco.  After humping an arriving train, they'd pull each track and rehump it, into empty tracks, to "block" the outbound trains.  The first humping sorted the cars by outbound train, the second blocked cars within the trains.  If a train didn't need to be blocked at Pasco, one with all eastern cars that could be sorted out in Minnesota, for instance, that track wouldn't have to be rehumped.

It was easier to pull the tracks back over the hump than to pull the track down into the departure yard, across the receiving yard, and up the "running rail", to the hump leads.

[bbussey]

My old club layout from decades ago had both a working hump yard and a working rotary coal dumper. For the hump yard, the long magnet (or magnets end-to-end) must extend from just before the crest (to create the coupler slack) to beyond the crest (so the lead car begins to roll away with the couplers still magnetically activated). It still took some practice to perform the operation smoothly, especially with shorter cars such as the two-bay hoppers on the coal dumper. An inexperienced operator would tend to push two cars over the crest, or push the second car slightly off the rotary to make the lead car clear the crest on the far side of the rotary.

Today, there are flat N52 magnets much more powerful than the Kadee/MTL above-track magnets that can be buried under or on either side of the track and placed end to end to create long magnets of any length. They work well with both MTL couplers and the less-responsive copycat couplers.

[videobruce]

I will be pulling the cars back over the hump, but that was never a problem with my hump 25 years ago and it doesn't appear to be now either. About the only thing that isn't a problem.

[nkalanaga]

If the cars roll smoothly, and the loco runs smoothly, there shouldn't be a problem pulling.  On the uphill side, the slack is stretched, so the uncoupler won't work.  Going down the other side, the slack may be bunched, but as soon as the cars are past the uncoupler, the couplers will latch again.  The first part of the train will still be stretched by the weight of the cars on the uphill side.

[Maletrain]

As I posted before, my experience trying to pull cars over the Bachmann under the tracks magnets has problems when using cars with steel weights or axles.  Those particular magnets are strong enough to actually pull the cars with steel into the magnet, creating slack over the magnet, so the car uncouples when trying to pull a string of cars over it. 

If you use magnets that strong, and you have steel in some cars, then don't be surprised if the ones at the end of a drag do that.  There is probably enough rolling friction of even up-hill drag to keep the slack out at the beginning of the car string being pulled.  But, those particular magnet will pull a car with steel in it up a grade.

[learmoia]

I agree you want some length of magnet on the uphill side of the hump, there the couplers are compressed, and extend the magnets until just over the hump (so the cars are still on the 'uncoupled' state as they start to roll down the hill.)

That will allow you to hump one car at a time.... but what about multiple car cuts...
Electro-magnets? and implement an 'uncouple button' that activates the magnets when your ready to uncouple a cut.
.. But on the flip side, you could only get away with 1 or so cars over the hump before that car starts pulling up hill car back to compression.
Having some kind of retarder at the top of the hill that is just string enough to hold cars on a slight grade would help that.

~Ian

[nkalanaga]

What you really need is a magnet, on the uphill (hump lead) side, long enough to hold a little over half of your longest cut-to-be-humped.  But, that would also uncouple every car, unless it was a series of electromagnets.  Possible, but hardly practical.  Easy on the prototype, as the pin-puller could run back and forth as needed. 

Humping one car at a time, even if they go to the same track, isn't unprototypical, just inefficient.  Since model yards seldom have to worry about maximum throughput, I'd say do one car at a time!

On the other hand, I have personally seen the results of trying to hump too many cars at once.  When they computerized the Pasco hump, shortly after the BN merger, the programmers decided that the computer should be allowed to decide where to make the cuts.  My father was the retarder operator, and had no use for, or understanding of, computers, but did know the yard - and the retarders.

He told them they couldn't hump more than three 100-ton grain hoppers in a cut.  The computer said they could.  He refused to do it.  So, the yardmaster told him to let the computer run the hump.  OK....

The first attempt was six cars.  The retarders are about 120 feet long each, a master retarder near the crest, and group retarders just before the switches for the individual bowl tracks.  That would be roughly two 100-ton grain cars.  Three cars could be controlled, with two in the retarder at any given time.  With six, four of them are pulling or pushing, with braking only on two.  The master retarder hardly slowed them down.  The computer, sensing that they were going far too fast, locked the lead cars in the group retarder, which should have stopped the cut.  Instead, the remaining four pushed hard enough to pull the "brake shoes" out of the retarder.  The yard was out of commission for several days while repairs were made.

The computer was reprogrammed!

[Maletrain]

The computer was reprogrammed!

Seems like the yardmaster also needed to be reprogrammed.

[nkalanaga]

Possibly, although I suspect that he was working on orders from higher up.  Management probably figured that they'd paid for the computer, it should be allowed to do its job.  Of course, they didn't pay much for it, as it was second-hand.  They could have gotten a second one for the same price, but decided they didn't need two of them.  Spare parts, anyone?

But, remember, back then management at most companies didn't know much, if anything, about computers.  So they really were making it up as they went.  It worked surprisingly well, overall, in spite of some glitches.  One of those was that, to do "housekeeping", the computer really needed to be shut down every day, for half an hour or so.  Modern computers do "that stuff" automatically, while everything else is running, but this one could barely manage to run more than one program at a time.  So, since "they" expected the yard to run 24/7, and the computer to do the same, it had a habit of shutting down at about midnight to do its memory and disk cleanup.  They finally figured out that it made sense to do that at the same time the overnight shift (my father) took lunch.  After that, it was happy to run the rest of the time.

[videobruce]

Seems like the yardmaster also needed to be reprogrammed.

Yardmasters have little power these days, they are too busy doing clerical jobs since the carriers cut the clerks jobs years ago and more recently.

[mark.hinds]

Most cars are MT truck mounted couplers, the others are Atlas accumate.
(remaining content removed)

I've never been able to get Accumates to work properly with under-track magnets; they are too stiff.  Step one is get rid of them. 

Then you need to adjust all your MT couplers to work correctly on a section of test track.  You also need to position all the under-track magnets on your layout so that couplers on normally centered cars (in their natural position, not hand held) will swing to opposite sides.  When both swing to the same side, they won't typically uncouple hands-free.  There are other adjustments you can make as well, but these 3 are the main ones. 

[NtheBasement]

I've tried other couplers too, and though far from perfect, MTLs perform way better than the rest.  Except for my rotary coupler hoppers, pretty much every car I have is now riding on MTL short shank roller bearing trucks.

My biggest mistake was buying a dozen Bluford hoppers of which I could get two to uncouple.  Their body mounts were molded into the frame and required delicate carving to convert to MTL trucks.

[Maletrain]

For what it's worth, I don't have problems uncoupling Atlas couplers with the Bachmann under-track permanent magnets.  Those magnets are wider than track gauge, and pull the pins far to the sides.

The problem is when those couplers drop their steel pins and no longer react to any magnet.

Microtrains couplers work fine on the Bachmann under-track magnets, too.

[videobruce]

I moved the magnet up again another 1/2 to 3/4 inch just past the crest and it improved uncoupling. Up from around 25% uncoupling to around 75% of the cars.
With the MT magnet 50x7x3mm (measured) cars now separate at the very end of the magnet meaning to move it up further will require a longer or just more short magnets.

[nkalanaga]

I'd add another short magnet.  Much easier than replacing the one you have.