What would cause a car to travel over a switch (turnout) fine 4 out of 5 times, but on that one time, jump a little and the cause a derailment?
Probability, and poor tolerances. It is likely that the turnout only loosely conforms to the NMRA standards…it is probably out of gauge. However, it could just be that the points are loose, or that they need to be sharpened so that they don’t encourage the flanges to pick them, or that the frog filler is too high, or that the flangepath in the frog is too narrow or wide, or that the guard rails are not set properly…there are many things that can account for your troubles. I forgot about the wheels on the axles, too…they may be out of gauge.
If the axles can slip back and forth in the side-frames, all it takes is a certain vibration caused by a certain speed, and the axle comes far enough one side to allow the flange to get in the way of something…or the other way around, really.
You should have a series of slow runs through that turnout, and watch for the derailment. Stop the loco, and using a bright flashlight, have a close look at which wheel seems to be derailing. Maybe that will be someplace to start.
selector did a good job of an explanation. I too have some problems with turnouts, and am thinking of replacing them with a higher quality turnout, I currently run Atlas, not sure what kind of quality I can get from a code 100 equivalent that will be compatible with the rest of my Atlas track.
-beegle55
These are Atlas Code 80 for N Gauge. Thanks for the tips. Sometimes this hobby can drive you crazy…but I love it!
The laws of probability. That’s 20% derailments!
Trackwork. 1. Oversized flangeways 2. ‘Bargain’ switches 3. bad track joint.4. running too fast.
Is it ONE car?
- wheel alignment 2. truck binding (horizontally) (vertically)?
START with an NMRA Gage.
May I recommend KATO track? Your probability will drop cosiderably (like 10 times).
Seems like the joint heading into the switch was just a teeeeny bit off. Straightened it out and seems to be running fine. Amazing wht a tiny bit will do. Thanks for the help…
Then maybe the change is not meant to be…now they jump the opposite direction. Time to go to bed…time change is catching up with me. Thanks again…
The wheels on the car could be picking the switch. I’ve had this happen, and it would be just one particular car doing it. Usually a change of wheels solved the problem. Some things you might try are:
Using a fine file, file the switch points so that they blend smoothly with the stock rail.
Check the gauge of the wheels on the trucks.
Check to make sure the wheels are centered on the axle. If they aren’t, you could have a dog-leg truck. This will force the truck against the rail and cause derailments, usually at a turnout.
Check to make sure the point rail of the turnout isn’t sticking up a little. I have many cars that will go right over something like this and some that won’t. I have a P2K PA1 that’s downright allergic to this problem and will jump the track every time it hits a raised point rail.
Check the frog. The wheels on the car could be dropping into the frog, causing them to jump up when they come out.
It’s also possible that there is debris at the points, ballast or tight throw for the points or the switch machine doesn’t have enough throw to firmly plant the point against the rail.
Alas, I had this problem on a pair of IHC steamers I have (the tenders). Seems the wheel guage was a bit tight, and the backs of the wheels were hitting the guard rails. A simple re-guage of the offending wheelsets (all 4 axles, on 2 tenders), and all was well.
I have noticed that entering the curved part of one turnout (Atlas, code 83), there is a bit of a “sharp place” where plastic wheels hop (and sometimes derail) where the point rail ends, and the rail to the frog starts. Apparently, this joint isn’t quite perfect, and I can fell a slight snag when I run my finger across it. I’ll find my files and fix this, one or 2 passes across the offending corner should do the trick.
Brad
You’re in N scale, right? Try body-mounting your couplers…Truck-mounted couplers tend to cause these sorts of derailments at turnouts since the pull/push force against the coupler also directly dictates the turn of the truck.
I too have a similar turnout problem. Would a Dremel be a good tool to file the points or would that risk damaging the turnout.
I had a similar problem that didn’t involve a turnout, but a slight dip in the track. My E7’s never had a problem but some SD70’s would trip and derail in one direction only! I found the dip using a 6 inch steel ruler set edge on the rail, with a tiny flashlight behind it. I shimmed it up and did some filing on either side of the dip. At our next get-together we will see if I was successful, eh? jc5729
Only if you are a brain surgeon, Friend! [:D]
Seriously, what I do is take a small piece of tie and place it between the point to be shaped and its adjacent stock rail. The tip of the point should extend about 1/16" beyond the piece of tie so that when you file you don’t keep causing the tie to scoot out from behind the point rail.
Now, take a needle file with a suitable profile in cross-section, and run it back and forth against the points rail, not too much force, and make about 10 passes each direction. Concentrate your pressure on the top half of the point tip because that is where the flanges pass, not lower. Also, it is better to keep as much metal and “foot” under the point as possible for strength generally and for maintenance of posture so that it doesn’t dip when it gets pressed by the flange.
Remove the tie piece, and let the action bring the point rail flush to the stock rail. Does that look sharper, and is the back of the point lying nice and flush against the stock rail’s side still? Run a car through it by hand, and see how it looks and feels. Repeat as necessary.
Don’t forget to vacuum filings before you pass a locomotive overhead.
I hope that works for you.