OK, I have grown desperate and am open to any suggestions.
When I run my heavy weight Walthers passenger cars clockwise on my layout they perform well, no derailments. However, if I turn the train around on a wye and run counter-clockwise, I get several derailments and shorts, but not always in the same place. Curiously, when I run the streamlined Walther’s 1955 Hiawatha in either direction it is OK except for the dome car being problematic. Like the heavy weight passenger cars, the Hiawatha dome car has six-wheel trucks. The rest of the 1955 Hiawatha passenger train has four-wheel trucks. As a further test, I took out my Fox River Valley 1935 Hiawatha passenger train and it performs well clockwise or counter-clockwise. This streamlined train only had four-wheel trucks.
Let me tell you what I have done to troubleshoot this problem:
Tested with the NMRA standards gauge all the passenger car wheels. Some were out of gauge.
Also used the NMRA standards gauge to check all the flanges and points near the spots where I am getting derailments or shorts.
While the 6-wheel trucks are manifesting the problem, the problem is likely your track work. 6-wheel trucks are much less tolerant of track flaws than are 4-wheel trucks.
I have had similar problems on my new layout until I methodically tested every inch of track. Most of the problems were on curves and on turnouts.
Any humps or valleys, no matter how minor, can raise a truck off the rails.
If one rail is higher or lower than the other rail, that is another problem.
Kinks on rail joints on curves will derail 6-wheel trucks.
Slight vertical rises where track sections connect, caused by poorly fitting rail, joiners will do it.
You need to do some detective work at slow speeds to find the glitches.
sounds like some trucks on the passenger cars are able to swing more one way than the others. From my personal experience the best way to fix this is to record the cars going past various trouble spots in slow motion with your phone and diagnose it that way.
The shorts are likely caused when the wheels touch over various parts of turnouts.
In any case, it just sounds like either poor track work, or overly tight curves. The best idea is to fix your trackwork. That being said, if you want to make your cars perform better, here’s some advice from someone with tight curves:
(1) I had to mod all my heavyweight equipment by cutting down on the centerbeam frames of the passenger cars to allow the trucks to swing more left and right in order for them to run reliably on my 22" radius curves. I know they look terrible on them but you gotta work with what you got.
(2) Other cars, I noticed the diaphrams were too stiff and would push against each other and derail the car. This can be resolved by using longer shank couplers, or bending the diaphragm springs to be less compressive from the inside of the car.
(3) Also make sure the screw that holds the truck on isnt too loose, but also not too tight (there should be SOME body wobble).
That was my first thought, whenever a car derails I generally check out the truck screws. I’ve found that loosening them a bit often fixes the problem. I’d rather have a car that wobbles a tiny bit but doesn’t derail if push comes to shove.
I have a Milwaukee Road Superdome and it was one of my more troublesome cars to get to run reliably. The clearance between the top of the truck and the car floor is very tight.
Besides a very short bolster screw, which doesn’t give you mush leeway in making them just a little “looser” the design of the Walthers cars rely on four screw heads that contact slightly sprung metal strips in the car floor.
I’ve had success at making some cars run better by using a good, fine mill file and knocking off a bit of the “crown” on these screws. Sometimes the slots are a little buggered up so evening them with a file can help.
Here’s a four-wheel truck but the screw heads are similar:
When you put the bolster screw back in I like to put a small drop of PVA or any good white glue (Elmers) to act as a thread locker. Not quite as aggressive as Locktite but enough so that you can loosen the screw without it eventually working out.
I also seem to recall having to file “something” on the truck tops to gain more clearance for the truck to rock. It may have been brake cylinders or equalizer arms. Don’t recall at the moment but I can look at the car later to see if it jogs my memory.
I have also run into some cars running better in one direction over the other. Not just six wheel stuff, either. Many of my main line curves are superelevated and I’m sure the transition is smoother in some places rather than others. As previously m
I should have mentioned the track radii. It is 32 inches on the outer track and 30 inches on the inner track. So far I have done my testing on the outer track.
Also, another frustration is that when I go slow over the same trouble spot, nothing happens. The troubles start when my throttle is set to about 25 percent or higher. This makes troubleshooting more difficult.
Ed, thanks for the tips regarding the dome car, I will have to give them a try.
I am in agreement that track work has to be the issue. Tomorrow I will go over the troubled sections inch by inch. In general, the derailments are near turnouts and curves. I will specifically check for dips and bumps.
By the way, the outer curves are superelevated.
Thanks guys for all the tips, keep them coming. I will update this as I make progress.
Six wheeled trucks (locomotve, passenger car, or freight car) are simply more sensitive to track problems and more prone to derailment.
A model train truck with rigid side frames is only guaranteed to have three of the wheels on the track at one time. Any imperfections in the track can cause a wheel to lift. On a six wheel truck three wheels can lift.
Equalized four wheel freight car trucks solve this problem, and all four wheels will stay in contact with the rails over faulty areas, Unfortunately, equalized six wheel trucks are much more tempermental.
As mentioned by others, the root cause is probably in the trackwork, and the six wheeled trucks just make it more obvious.
Those are the radii on my double track mainline. Those curves are broad enough to support 6-wheel trucks.
That is not surprising. The slower the movement, the greater the opportunity for a wheel that has lifted off the rail to reseat itelf.
When I mention testing at slow speeds, it gives you a better opportunity to see how the wheels are performing around curves and over turnouts. Even at slow speeds, wheels will lift off the rails if not perfectly aligned.
That doesn’t matter. Properly installed superelevated curves will not derailed a truck as long as the superelevation from inside rail to outside rail remains constant.
As to the derails only in one direction, all the trailing point turnouts become facing point turnouts when going the other way. Facing point turnouts are more derailment prone if anything is even slightly wrong. Trailing point turnouts allow the wheels to push them up against the stock rails. The one thing you did not mention checking on the rolling stock is truck free swing. Be sure that the trucks swing freely in both directions and don’t get hung up on coupler boxes, centerbeams or underbody details. Do all your Kadee gladhands clear all the turnouts and crossovers? Is one truck on each car tightened up so the truck swings freely but does not rock? And is the other truck given a tad more slack so that it can rock a little bit to keep the wheels on the track should the track be a bit off level?
Try keeping notes on your derailments, which car, which end, where on the layout. A pattern may emerge.
Check your track work. Track gauge. Kinks. Is the track level from side to side? Do you have grades? If so, is the “vertical curve” leading into and out of the grade smooth, or is it an abrupt kink? Are the points of all the turnouts pressed hard up against the stock rails, turnout set in both diections? Do all the turnouts feel smooth to the finger?
Here is the thing about turnouts. They often don’t derail trucks on their own. Typically, by time the offending truck has reached the turnout, it has derailed. The turnout simply becomes the fait accompli. For that reason, if it appears that the truck is derailing on the turnout, check the track for a few feet before the turnout, particularly any nearby curves.
As gmpullman said, the usual culprit on Walthers passenger car derailments is due to the screwheads having burrs on top of them. When they assemble these trucks in China, they use powered screwdrivers. If the clutch isn’t set right, the screwdriver can “cam out” and lift a little bur out of the Phillips screw top.
Whenever I have a derailing Walthers passenger car, the first thing I do is file the heads of the screws on the trucks. Not a lot, but just enough to smooth it all out so it slides real easy on the contact pad.
One way to eliminate one possibility is to reverse the cars before you reverse direction of the train. Run them again and see if the derailments are now when running in the original derailment free direction. Running the same train the reverse direction but with all the cars oriented in the same direction relative to direction of travel as the derailment free direction should narrow it down to either the track or the cars. Which end of the car is leading may make a difference especially if the looser truck leads. Typically, one truck is generally set loose to pivot in both directions and the other tighter to pivot only in one direction. It may matter which is the leading truck.
There are two constraints to these six wheel trucks: massive coupler boxes that interfere intermittently with truck rotation, ironically these contain the compound swing couplers intended to reduce derailments, and very little end float available for the axles. Cutting away at the center beam only gets you a tiny amount of extra truck rotation until the front of the truck then hits the coupler box.
Most model railroad cars have significant axle end float. Most also have plastic side frames whereas the Walthers heavyweights have metal sideframes, or at least the older ones I have do. I’m going to experiment with putting a short axle wheelset in the center position and see if the extra compliance gets me no derailments.
It’s not a radius problem per se. Ive checked mine by hand and the trucks will travel around a 24" radius. Just not when towed by a locomotive.
I notice that ‘truck tuning’ isn’t on the list of improvements tried. I recommend that it should be…
My understanding of stable running was that one end of the car be set to smooth pivoting only, and the other one allowed a little ‘twist’. If that is not done, then some form of constant-contact side bearing would be needed on at least one truck, and that contact area would be something that might preferentially hang up in one direction as noted.
Just for grins, take the center axles completely out (thereby simulating long-wheelbase 4-wheel trucks) and see if the issues persist.
This only worked well on shorter freight cars, and it never worked as well as a pair of equalized trucks that pivoted only.
On longer passenger cars both trucks should be set with just a little twist. I think this might be why most passenger car trucks have such a larger contact area where they meet the bolster.
In three dimensional terms, all trucks need to yaw. When you set a truck loose you desire pitch to help the car maintain more wheel contact, unfortunately loosening the screw also allows roll, which is less desirable.
I now have the Chippewa heavyweight passenger train running counter-clockwise without derailments and shorts about 90 percent of the time. What have I learned so far?
Richotrain is correct that it is usually not the turnout that is the problem. The derailment starts before the turnout, it just does not manifest a short when it hits the frog.
Slowly watching the train go by especially before the turnout reveals the problem. I can see the problem wheels lift off the track.
In most cases it is because of subtle dips and bumps. Once I fix these that section of track there is no longer a problem. Loosening the screw holding the truck has also helped on one car.
It is not just a case of bad track work; it is an interaction between problematic cars and bad track work. Lastspikemike’s suggestion of reversing the problematic car often solves the problem. Of course, I do not want to have to remember which direction to place a particular car when running clockwise or counterclockwise, but reversing the car helps identify which truck is the most problematic.
I want to thank all of you for your advice. This is clearly going to take me some time to fix. Apparently, my track laying is not perfect. I plan on testing, testing, testing until all my passenger trains run smoothly, in both directions, on the mainline, the sidings, the yard, and on the wye. I will then test my freight trains again to verify they still run smoothing, especially in the sidings and yard. I will not be doing any ballasting until the passenger and freight trains run smoothing 99 percent of the time.
We sometimes forget or overlook the fact that long wheelbase locomotives and rolling stock are rigid in both horizontal and vertical directions. In fact, that’s why road vehicles use the suspension they do, roads being much more variable in “flatness” than railroads are. Shay and other truck type steam locomotives were designed to allow relatively large driver sets to work on very poorly laid track. Tighter radius, steeper grades and not very flat.
At 1/87 any variations in flatness are going to be huge compared to prototype even though they don’t look it. Model trains handle poor track much better than any prototype could. But as we strive for more accuracy these particular cows may come home to roost.
Those two statements seem contradictory. Variations in flatness are going to be huge at 1:87 scale compared to the prototype. For that reason, model trains do not handle poor track as well as the prototype does.
That has to be the biggest frustration for anyone new entering the world of scale model railroading, poor track work leading to derailments and unintended uncouplings.
