truck tuning: consistent delrailment on a tight radius

[peteski]

Late to this party, but in looking at this photo, the first thing I saw was that the inner rail to the left of the joiner has popped free, or the "spikes" have been sheered off, for at least 6-8 ties.

But he had to cut them off to install the rail joiner.  And yes, the rail will then straighten out, since nothing is holding it to the curved ties.

[Point353]

Would gluing a short section of guardrail next to the inside rail eliminate the derailments?

[mmagliaro]

If i solder PCB ties under the joints, don't I loose my expansion joints there?

Right now it's reasonable to access that track, and in the future it will only be slightly less so.  I'll be filling in the space between the two levels with a vertical wall, but the track will always be easy to get to from the top.  I'll just have a bit less finger room on the outside edge of the lower level.  The upper level is wide open right now, but eventually i'll be adding a piece of facia there which will make accessing the outside edge of the upper level a bit harder as well. 

I do want to perfect this piece of track before moving on.

Put a couple on either side of the actual gap, and don't solder across the actual joint itself.   Just be sure to pull that joint out into alignment while you solder in the PC ties.

There's another possibility.   You say the radius is 10".  Atlas makes 10" sectional  half-curved pieces.  Maybe you can cut out a short section and splice in a rigid sectional.    That will tend to keep the ends of the flextrack that attach to it in line.

[boisecity]

I had a similar problem on my layout. I fixed it by using a pair of long nosed pliers and gently bent the rail so that it was curved not straight at the joins.
JohnF

[DKS]

But he had to cut them off to install the rail joiner.  And yes, the rail will then straighten out, since nothing is holding it to the curved ties.

You only have to cut off two at most to do the rail joiner. Look at the photo again--they're missing waay far back.



Maybe I missed a post explaining this. At any rate, this needs to be addressed, because not only will the gauge be off, but the rail may be lifted slightly, resulting in a vertical curve, which will exacerbate any other issues.

 

[peteski]

You only have to cut off two at most to do the rail joiner. Look at the photo again--they're missing waay far back.

Maybe I missed a post explaining this. At any rate, this needs to be addressed, because not only will the gauge be off, but the rail may be lifted slightly, resulting in a vertical curve, which will exacerbate any other issues.

You're right.

Hmmmm . . . in reply #13 fewwdj wrote "I tried removing the around 4 "spikes" on the outside of the inside rail where the kink is to let that smooth out, but it didn't seem to help.".  Hmmmm . . .

Unrestrained flex-track rail will tend to get back it its relaxed (straight) state.  Not very desirable when it is supposed to be a 10" radius curve.

Besides, a 10" radius helix in N scale seems like asking for trouble in the first place (unless the trains are very short). I think that it might be a good idea to leave it easily accessible.

[randgust]

This is about as contrarian as I get, but because I make some pretty tiny modules with some rather absurd curves, you learn things.

If I'm doing curves 11" or under, and particularly if they are hidden, I'm simply soldering up sectional track, Atlas C80.   It's rock-solid for gauge and the rals are pre-bent.  Not only that, but if you leave an occasional joint unsoldered, you've got expansion/contraction covered.  That's an issue, particularly on portable modules, as they are subjected to a lot of heat and humidity variation.    I've generally found that kinking is rather deadly on anything under 11", period, so keeping a tight curve in gauge is a challenge.

To match exact radius that's not 9 3/4 or 11", I'm cutting the between-tie section on one side of the sectional pieces with an Xacto knife so that you can bend it tighter or wider, your choice.  For 10" radius I'd use 9 3/4 and stretch is slightly wider.  But if you're soldering up sectional it's going to stay in curve, in alignment, in gauge, and can be rather lightly secured.

Your choice on rail size is a personal issue, but in the battle of what looks better, a derailed train on C55 or a running one on C80, unless it's clearly visible from the side, I pick the latter..

I'm using C55 wherever I can, but on my logging modules, any hidden is still C80 sectional, soldered up.

[DKS]

If I'm doing curves 11" or under, and particularly if they are hidden, I'm simply soldering up sectional track, Atlas C80.   It's rock-solid for gauge and the rals are pre-bent.  Not only that, but if you leave an occasional joint unsoldered, you've got expansion/contraction covered.

This!

[Point353]

1) Has the OP tried running the loco (that derails) without its shell on to determine whether or not the issue is simply interference between the trucks and the steps?

2) Atlas makes code 55 sectional curve track in a 10" radius (and up from there in 1.25" increments).

[glakedylan]

been thinking over this all week long
what about a or some "sectional" track in places that the flex is problematic?
IIRC atlas code 55 came in sectional pieces, each radius being 1.25" larger
i know i have used some in the past (allbeit larger radius 17.5" and larger)
yet, i cannot seem to find any atlas code 55 sectional at usual online N Scale
resource sites?
point is...if you know it is 10"R by design, a piece of 10"R sectional track

could be placed in that curve without the worries of gauge or loose, slipping
rail, that could make a good solid soldered joint.
just the results of my mental exercise.


gary

[freedj]

Thanks for all the suggestions!

I spent a while trying to get the shell of the loco and could get the long hood end loose, but I couldn't seem to get the short hood end loose.  After breaking off the bell detail I decided to stop trying

I picked up another two lengths of flex and have ripped up the track back past the kink and will relay the track to the module connection.  The 10" radius snap track is a great idea, but the module is designed with easments and the roadbed has been cut with that in mind.  I don't think the snap track would fit well enough.

[freedj]

Sigh.  I ripped up the "trouble area" and can't seem to eliminate the derailment.  So I went back to getting the shell of the loco.  I got it off and since I now have a great view from the top down of when the derailment happens.. I can see that is happening 2 feet further up track.  I think it was also happening where I highlighted in the image.

I have removed both the shell and the fuel tank and eliminated the truck hitting either as a culprit.  I have also noticed that the trucks are "on the stops" in that curve.  If the trucks had another 5* of free rotation I think it would work.  Here is a list of the evidence:

* loco derails long hood forward going downhill
* loco stays on track long hood forward going uphill
* loco stays on track going short hood forward downhill
* loco stays on track going short hood forward uphill
* both trucks are on the stops or almost on the stops in the tighter section of the curve.

I can not make the curve significantly broader.  I may try picking up the 10" snap track and see if the loco can navigate that without issue.  If it can, i may rip out everything that is supposed to be 10" and replace it with the snap track which would ensure there isn't a spot less than 10".

I am baffled why it only derails in one orientation going one direction.

Has anyone made truck modifications with a file to allow the truck to turn slightly more?  This area is not a viewing area and if it looks clownish but is reliable, i am OK with that.  Would it be more advisable to slightly narrow the gauge of the outer wheels of that truck so as to not mess with gear mesh?

The loco is an atlas gold series sd-35 and is currently my only sound equipped loco, so i really don't want to sell it.

[Maletrain]

I can understand why it is easier for the flanges to climb over the rails going down hill than going uphill.  I really have no idea why the end pointed down hill makes a difference.  Does the same truck derail every time there is a derailment?  Can you swap ends with the trucks on that model?

I would think that you could attach 10" radius sectional track to flex track (solder while flex is straight) and then bend the flex into the shape of your current easements to get the same track configuration as your existing roadbed. 

[Point353]

Sigh.  I ripped up the "trouble area" and can't seem to eliminate the derailment.  So I went back to getting the shell of the loco.  I got it off and since I now have a great view from the top down of when the derailment happens.. I can see that is happening 2 feet further up track.  I think it was also happening where I highlighted in the image.

I have removed both the shell and the fuel tank and eliminated the truck hitting either as a culprit.  I have also noticed that the trucks are "on the stops" in that curve.  If the trucks had another 5* of free rotation I think it would work.  Here is a list of the evidence:

* loco derails long hood forward going downhill
* loco stays on track long hood forward going uphill
* loco stays on track going short hood forward downhill
* loco stays on track going short hood forward uphill
* both trucks are on the stops or almost on the stops in the tighter section of the curve.

I can not make the curve significantly broader.  I may try picking up the 10" snap track and see if the loco can navigate that without issue.  If it can, i may rip out everything that is supposed to be 10" and replace it with the snap track which would ensure there isn't a spot less than 10".

I am baffled why it only derails in one orientation going one direction.

Have you tried narrowing the spacing of the wheels on the center axle of the truck under the long hood (similar to that for the truck under the short hood)?

Is there any relatively more abrupt change in vertical elevation (as opposed to horizontal radius) at the point where the derailment is occurring?

Or, as suggested previously, take a short section of rail from a piece of flex track and install it as a guardrail at the point where the derailment occurs.

[peteski]

Have you tried narrowing the spacing of the wheels on the center axle of the truck under the long hood (similar to that for the truck under the short hood)?

Is there any relatively more abrupt change in vertical elevation (as opposed to horizontal radius) at the point where the derailment is occurring?

Or, as suggested previously, take a short section of rail from a piece of flex track and install it as a guardrail at the point where the derailment occurs.

If the derailment is due to the trucks being rotated to their maximum rotation angle, then the middle axle is not the cause of those derailments.  The loco's wheelbase is just too long for that curve. The truck's wheelbase is ok.