HO- inexplicable derailings

[tc1961N]

I have begun replacing some of my suspect Atlas turnouts with Pecos and have virtually eliminated derailing at those areas. The rest of the Atlas turnouts were ok and cars and engines would pass through with no issues. Yesterday I ran an engine with 5 cars attached around the same loop 5 times with zero derails. This morning I tried it and had 3 or 4 derailings before getting around one lap with the same train. None on the Peco turnouts. Weird. My layout is in my basement and I have a dehumidifier running all the time to keep humidity constant and there were no large temp swings in the basement that I was aware of.

 

[RhB Michael]

Everything else being equal, it usually is a temp or dewpoint swing when track has otherwise been behaving normally.

Check for any buckling and use a track gauge at the trouble spots and see if the track is still in gauge.

 

[rgeiter]

I had a similar problem. Certain cars would derail on my Walthers switches but not my Atlas or Peco. Turns out that it was the cars wheel sets were out of gauge.

Get yourself an NMRA gauge and check your wheels and turnouts.

 

[Rico]

Just yesterday I had a train running on its own while I was working at the desk.
Must have made a few dozen laps at least then one car kept derailing at just one spot, no switch anywhere near.
After checking the track work and wheels/trucks/couplers etc. and making zero adjustments to either it ran just fine.
Gremlins.

 

[santafewillie]

Check the wheel gauge as Bob suggested.

 

[Bigboy57]

Wheel gauge can definitely be the culprit, especially as humidity levels fluctuate. Sometimes a quick solution can be as simple as rotating the car end for end.

 

[RhB Michael]

I have a flat with a log load that refused to go through a turnout. I moved it to a different order and it's been running fine for three years now.

You just never know.

 

[tc1961N]

I checked all the wheels- ok. I changed the order, now ThermIce is in the rear and 7 laps around, no derails. weird!

Also noticed if I flip the car around it can work differently.

 

[jward]

Changing the order a car is in, or its orientation does not solve the underlying problem.

Does your car have truck mounted couplers and wheels with pizza cutter flanges? If so, The first step should be to change out the wheelsets for something with an RP25 contour, like the ones you'd find in an Athearn kit. This is one of the simplest fixes you can make that will improve tracking. Another one is to loosen the bolster screws that hold the trucks to the car frame by a quarter turn. But I suspect your car does not have theseand if it has the snap in type of truck like Tyco or some AHM there isn't alot you can do about that.

Did you watch the car carefully to see where it actually derails? This may be a completely different spot than where it hit the ties. Wheel flanges can ride the rail tops for quite a ways before dropping over the edge. With a bright light like your phone's flashlight, watch the wheelset that derails as it runs around the layout. Once you find the trouble spot, look at the track for things like a misaligned rail joint, narrow or wide track guage, or a twist in the track.

Finding the little problems like these and correcting them will greatly improve the reliability of your railroad. And in the process you will learn alot about what works and what doesn't.

 

[Selector]

Wheel gauge for sure, but also run one of those loooonng, thin, needle files on edge through the frog routes. scrub, scrub....not hard, and not for an hour. Just two/three swipes in each direction, change the route and repeat. Sometimes flashing is in there from the mold extraction if it's a black plastic insert for the frog. Also, NMRA gauge comes with flange path clearance nubs. Put those nubs through the points, facing down into the gauge, gauge held transversely square to the axis of movement, and held very close to vertical. If it catches, you have a gauge problem there, either in height, causing the flanges to lift or buck, or in gauge and you have either a pinch there, lifting the odd axles, or the wheel might drop. Points and frog, please. And do gauge every axle.

One last check: a poorly prepped roadbed might have the turnout appliance bowed upward or downward. If it rotates the frame at all, it can lift one wheel and its guiding flange out of the gauge. So, take a 12" known flat like a small level or something and place it over the turnout so that it also covers at least one approach to it. Can you see daylight where there shouldn't be any?

 

[rgeiter]

Also check your “air hoses” on your couplers. They can be an issue too.

 

[NVSrr1500]

Could also be the truck has to much or not enough play in it. It could be the other truck on the car as well.

 

[J.Albert]

OP wrote:
"This morning I tried it and had 3 or 4 derailings before getting around one lap with the same train. None on the Peco turnouts..."

It looks like you have more "replacing" to do...

 

[tc1961N]

I'm amazed at how replacing 3 Atlas turnouts with Pecos have nearly eliminated the derailings. Now it's coupling airhoses on cars hanging toolow so when I check those and use the coupling height gage to verify, I have almost zero derails. And that's running 12 cars over 10 laps. I do see where I am eventually going to be replacing all the atlas turnouts but for now I'm just happy to be running mostly reliably and now attention turns to rail car coupling issues.

 

[Snowman]

There are other possibilities too. Can you show us a trackplan or photos of exactly how you built it? Not just scribbles on paper but exactly how it fits together and where the derailments are happening? It could also be due to a planning error, which would be harder to fix, but which still might be fixable.

 

[tc1961N]

Some of the issue is the airhose on a few of the couplers need replaced or adjusted, Ran one real slow and watched the air hose clip part of the turnout because it was not centered. I'm also learning that if I run the train real slow at problem areas I can identify the problem easier. Also can't throw(figurative) any old car on the track and expect it to work perfectly, except maybe most new cars. I guess the best way to make sure the car is not the problem is to take each car, get the weight right, examine the wheels(gauge), repair/replace, trucks repair/replace and with preferable body mounted couplers adjusted to NMRA height.

 

[santafewillie]

No freight cars ever go on my layout without a complete documented inspection first. It includes where applicable, Kadee couplers at correct height, metal wheels, wheel gauge, coupler hose adjustment, all trucks get the "truck tuner" treatment and correct weight with pennies added as needed. Saves a lot of headaches later.

 

[tc1961N]

No freight cars ever go on my layout without a complete documented inspection first. It includes where applicable, Kadee couplers at correct height, metal wheels, wheel gauge, coupler hose adjustment, all trucks get the "truck tuner" treatment and correct weight with pennies added as needed. Saves a lot of headaches later.

That's a good practice, I plan to start using it. I have the truck tuner, but I need to get the Kadee pliers for the air hose adjustment.

Thanks!

 

[RhB Michael]

You can pick up a cheap kitchen or postage scale at Walmart. I have a kitchen scale I use for a variety of weighing, but as for rail cars, every one of them has been in NEM specifications right out of the box. The Europeans follow the NEM like it's Gospel.