Hump Yard Operation


I think a REALLY fun way would be to use magnets like said.

Line under the tracks a bunch of small electromagnets, and keep one on the bottom of each car, and as the magnets pass each other, the electro switches polarity to either push or pull the car. With the right force, each car would end up rolling at the same, safe speed. All the switches could be operated by computer, and you have a yard operator select what track the car goes to, and the computer sets the switches for you.

Obviously extremely expensive, and very large. But would make for one heck of a show piece.
 
Dan, the magnet idea might work, as long as each car also had a magnet. I guess you'd need some way to calculate the car's rolling speed, how far into the bowl it had to travel, and then have the computer calculate just right amount of magnetic current to apply to keep it moving smoothly. It's probably the easiest solution in terms of technology but I think eveyone having to slap a magnet under their cars might be a deal killer.
 
Last edited by a moderator:
I like this a lot...I really got ya all in this thinking cap mode lol ...I tell ya it has been on my mind for a while. I find it interesting of some the comments gang. I appreciate the participation! I may do this someday I am not sure...but then it would deviate from my main plans of "Green River" to "Soldier Summit" to "Thistle".
 
Jim: You are right about the slapping magnets on to each car. That would be a pain. Would have issues with weight and having to reweight cars and stuff.

Also you could use breakbeam LED "eye" for detection. Would do the same thing as occupancy detection without the need to have to put gaps in the rail.

Another would be a bouncebeam LED, which you could stick between the ties, and when the trains bounces the beam back, it could calculate speed and occupancy.

Those two types of "eyes" have been used for awhile now with high end paintball guns.

Would take a lot of programming, and may not look very prototypical, but it would be pretty awesome.

The only problem I could see, other then cost and space, would be the electromagnets really messing with kadee couplers.
 
Well this thread is helping allot! I'll definitely have to think about the hump and its operations before I model it. Good thing I have some time.

If nothing I'll be one of the last completed sections, and/or it'll just be a non-operational individual hump, and I can just stick to switching blocks of cars...
 
vato loco said:
Give it time guy's, some where out there is a NERD hard at work with his slide rule and computer at hand to come up with a scale working hump yard retarding system! Then will Pat. it and sell at some crazy price that a normal man can't afford! As with all things, the working hump is still out of reach of most! but some day soon!
Click to expand...

lol, i was thinking this exactaly. someone out there is challenged by the fact that it hasnt yet been done successfully. its just a matter of time really. the one in the first batch of photos looks great though. model wise it looks very close to the real thing.
 
Last edited by a moderator:
John, thanks, that is a really great article. It's one of the best I've seen that explains hump yard operation from the perspective of a guy who worked them, not just designed them. I certainly defines the issues facing someone trying to replicate a hump yard in HO. I never even thought about monitoring wind speed in a really hump yard but I imagine a stiff wind blowing against the hump does affect things like backing off the retrarders some to keep the cars rolling. Fascinating stuff.
 
John P said:
I was wondering how fast cars are meant to be coupled in hump yards, and doing a search on "hump yard" +"coupling speed" got to this really well-written piece by a railroader about how hump yards work:

http://www.railroad.net/articles/columns/hottimes/hottimes_20021031.php

One thing I liked was the way he calls the class tracks area of the yard "the garden".

(4 m.p.h. seems to be right. About the same as the max speed over the hump.)
Click to expand...

A long worth while read. Thanks for sharing this info!
 
As a Fairly new Railroader this topic became quite interesting and found a lot of the replies ' head scratching' for an HO solution. I know most Hump Yards use computers in their yards to activate the retarders per a given weight to reduce the momentum of the car or cars and I believe most of them use Electromechanical retarders by clamping the wheel at various pressures.
I know this would be a huge undertaking to bring this design to HO scale but to throw another idea in the hat could a fine 'brush or feather affair' be set up at the lower end of the hump between the rails and wheels to slow the cars down before or after the assigned cut to the yard?
As for the hump incline from the looks of this yard video it seems to have quite a steep angle in real life and I'm more inclined to agree in scale a .25 -.50% angle would be effective in HO although it to would all have to be calculated on how long the spurs were in the yard.
Just Some Thoughts!

Video Hump Yard-http://www.youtube.com/watch?v=udu1bYJHkO0
 
Smitty, in the real hump yards you hear the flange squeal because a high pressure air jet is being shot up between the flange and the track. This air jet is calculated, either by a human watching a speed readout, or a computer, to keep the car between 3 and 4 mph. Real railroads have some of the same issues that we do with trucks that roll less freely as well as flat wheels that can make it hard to get the car down the hump at a high enough speed. The car will also have to come off the hump at higher speed to the central retarder if it has a long way to travel down the bowl track.

I seems that air actuated retarders are a non-starter in HO because any air jet that would slow down the car would also likely derail it.

There have been hump yard set up with fine bristle bushes between the ties that raise and rub against the axle enough to slow the car down. I think that system shows the most promise. The devil is in how to calculate exactly the speed down the hump, the speed at the central retarder, and what speed you need to have the car rolling at so it reaches the cut of cars at less than 4 mph. A difference of 2 or 3 mph is fairly significant in prototype railroads but almost unnoticeable in HO. I'm sure this can be done but it takes a person with a lot bigger brain than me to overcome all the possible permutations of rolling a car down a hill.
 
I'm certain that the prototype cars are not being slowed down by air jets. Air-operated brake shoes, sure. But air alone wouldn't have the force to slow a moving railroad car. On the other hand, I think it could be done on a model, where the momentum is hugely lower--if you can stand the noise, and not overdo it so as to shoot the car back up the hill, or off the rails. I think this was first done back around 1960 by Ed Ravenscroft. But I can't stop asking the practical questions:

If the cars need to be slowed down, why were they going too fast in the first place? If there's no need to go so fast, use a smaller hump. It seems as though a lot of the design work here is too solve a problem (excessive speed) that never needs to exist. And if the cars need to go fast so that the slowest ones will move at all, maybe new wheelsets ought to be the priority.

Does the user have the information available in a form that's useful? If a car goes over the crest evey 10 seconds, is there a destination established for it, and for the next one, and the one after that? That's what the whole game takes place for, to send cars to destinations.

Does the user really have enough cars to make the whole setup worth building? Enough cars coming in, enough cars going out, one every 10 seconds as long as the yard is operating?

Hump yard? Bah, humbug.
 
Last edited by a moderator:
John, my last post about retarders was poorly worded. The high pressure air is used to actuate a retarder plate that squeezes the flange against the rail, hence the flange squeal in that one video.

There are plenty of modelers with more than enough cars to populate a hump yard. The issues are still the same as have been stated by me and other posters to this thread. Find a way to overcome the same issues the the 1:1 railroads deal with when using a hump yard in 1:87. It can be done, I'm sure, but not cheaply and it will take a lot of planning and engineering.
 
New ideas for hump yard control -- LEGO robotics for control

Sorry to revive an old topic, but I'm a new forum member and currently working on a small hump yard.

My background includes physics, programming, and robotics so I've taken up the hump yard automation challenge a bit.

I've been working with LEGO NXT robotics a lot lately, so I was wondering whether anyone had tried using the easy-to-use, and easy-to-program LEGO NXT robotics system (includes sensors and servo motors) to control the retarder system ?

The system has many kinds of sensors including light sensors that can be used as break-beam sensors to calculate speed. There are also ultrasonic sensors that could be used to measure available distance on a bowl track.

the system can control servo motors to operate retarders for friction -- to control enough servo motors (more than the basic LEGO system), there are add-on servo controllers such as the Tetrix brand ones

the problem can be made simpler (but less prototypical) by measuring the speed at the bottom of the hump (hill) -- and having retarders for each of the final tracks in the bowl -- that way the weight of the car no longer matters -- all of the PE has already become KE at the bottom of the hill -- so the max speed can be measured to calculate how much to retard the speed

more prototypical would be to use a pressure sensor placed under the track section at the top of the hill to weigh each car at the top to predict the speed at the bottom -- but there's more room for error there due to differences in the rolling friction of each car
 
ThirteenHorsehoes said:
Sorry to revive an old topic, but I'm a new forum member and currently working on a small hump yard.

My background includes physics, programming, and robotics so I've taken up the hump yard automation challenge a bit.

I've been working with LEGO NXT robotics a lot lately, so I was wondering whether anyone had tried using the easy-to-use, and easy-to-program LEGO NXT robotics system (includes sensors and servo motors) to control the retarder system ?

The system has many kinds of sensors including light sensors that can be used as break-beam sensors to calculate speed. There are also ultrasonic sensors that could be used to measure available distance on a bowl track.

the system can control servo motors to operate retarders for friction -- to control enough servo motors (more than the basic LEGO system), there are add-on servo controllers such as the Tetrix brand ones

the problem can be made simpler (but less prototypical) by measuring the speed at the bottom of the hump (hill) -- and having retarders for each of the final tracks in the bowl -- that way the weight of the car no longer matters -- all of the PE has already become KE at the bottom of the hill -- so the max speed can be measured to calculate how much to retard the speed

more prototypical would be to use a pressure sensor placed under the track section at the top of the hill to weigh each car at the top to predict the speed at the bottom -- but there's more room for error there due to differences in the rolling friction of each car
Click to expand...

I haven't done anything with the Lego robotics stuff since middle school, but I have a general idea of how it works. The sensors that are made by lego are rather spartan. We built something similar to a tank and the setup with the sensors just didn't react quickly enough for the purpose. We eventually made wired controller with 3 push-buttons (left to turn left, center to go backwards, right to go right, press both left and right to go forward)

The motors that come with it are decent though, high torque, and low speed, but they don't always have the same RPM for a given speed level.

You're better off trying to use those surface mount LED infrared sensors that they make for scale grade crossings. You can probably buy an off-the-shelf grade crossing box and instead of hooking up gates that swing down, hook them to some DIY retarders on the tracks.

You will have to standardize on one width of wheelset though unless you find some way for the retarders to adjust to the different wheel tread widths used by different manufacturers.
 
Last edited by a moderator:
Lego NXT is a good approach. The difficult part, I think, will be accurately slowing down the cars without derailing them. Everything else is pretty easy. There are even a variety of ways to identify the cars and sort them according to a predetermined list.
 
I've got two of the first-generation Mindstorms kits, but I've never tried the NXT. I hear they're good, though. I'd be more inclined to use something like an Arduino, just because it's designed for permanent mounting; the Lego connections and all are designed for intermittent use and easy tear-down, neither of which fit what you're doing. It would probably work fine.

The difficulty of course (and probably something to work out before bringing in the control system, in case it requires things the Legos can't do) is in the braking system. The flange-clamp system would cause derailments, if it worked at all (especially difficult unless you replace every wheelset so they match). The air would require a source, so you'd be changing cans or running a compressor. If you're up for an electronics challenge, you could rig up a few coils under the track, and have them pull back against a steel weight in the car... like a multi-stage coil gun, only the coil fires after the car's passed it, slowing the car. Might be a little jerky though, and it might mess with the couplers.

Jim's bristle pad thing seems promising. You could program the system to raise or lower the scrubber as needed. It would probably need some tweaking to avoid couplers etc, but so long as all your rolling stock has similar-sized wheels, it should work fine... and it would only need one servo, so you could rig up as many as you had sensors for.

Low speed might actually screw things up, depending on how much your cars weigh and how well your couplers are adjusted. You might be able to get away with no brakes with a very slight hill and light cars... sounds like you're up for the challenge, though. :) Can't wait to see what you come up with!

Edit: Loudmusic's idea sounds cool... total automation! That would be a heck of a thing.
 
Last edited by a moderator:
You must log in or register to reply here.



Back
Top