I'm having a devil of a time with six wheeled diesels on curves

[voldemort]

I have Kato and BLI six wheeled diesels.  The track is all Kato.   

The problem is with derailments occurring around curves.

Its not a problem with the four wheeled diesels.  Also not a problem with most steam locos. 

I've checked the wheel gauge of each locomotive.  They're all within range.  I've checked the tracks with my NMRA gauge.  They're all within range.   I've fiddled endlessly making sure there aren't changes in grade occurring at any of the unitrack joints on the curves. 

I'm wondering if the track gauge, or perhaps the wheel width needs to be adjusted to run the diesels.

Geometry wise, I think that the middle axle on the three cannot be tight with the track on both sides if it is in line with other two axles.  The geometry would be like pushing a chalkboard eraser (Rectangle) between two curved parallel lines.  The front and back would fit, but the middle would cut off the inside of the curve.

Any suggestions on how to fix this?  All the track is properly glued down, probably a 100+ hours have gone into fiddling with it trying to end the derailments.  It is all properly aligned, and, as I said the grade changes have been eliminated on the curves.  i don't think its related to the presence or absence of easements- whether one is there or not doesn't seem to affect it.  It also happens both in the middle of a curve as well as the beginning and ends of the curve (I've spent a lot of hours making and looking at close up videos to spot where it comes off),

I'm debating whether I need to somehow (ie, with a file) widen the gauge on the curves.

Most of the radius is 13+ inches, nothing anywhere less than 11

Ideas?

[brokemoto]

What models are you operating?  That might be the key to understanding the problem. 

I do not have many six axle diesels, but I have a few:  Kato RSC-2, Atlas/Kato SD-7. C-C PAs and a pair of Erie-builts on C-C chassis with V-:Line shells.

The Erie will derail on anything less than the Kato 13-3/4.  That is because the nose coupler is on a tongue on the front truck.  The opening in the nose of the V-Line shell restricts the movement.

I do not run the PAs on anything less than the 13-3/4.

I do not run the SD-7s on anything less than the eleven inch, but I have run them on the PECo "match cover" turnouts and they will clear them.

The RSC-2 will clear an eight  inch curve.

I had both LL and C-C DL-109s at one point.  The C-C was more tolerant of the sharper curves than was the LL.  At times, the LL would derail on eleven in ch curves and Kato #4 turnouts.

I had BLI Centipedes.  They would operate on thirteen and three quarters.

My track is  a mixture of Kato UNITRACK, B-mann E-Z TRAK, PECo turnouts, Atlas turnouts, sectional and flex plus one Shinohara turnout.

I have had other first generation six axles, so if you are running first generations, I might have some experience with them.  I have no experience with any prototype later than an RSD-5.

[nickelplate759]

Quick question - do they derail running light (not coupled to anything), or only when coupled to something else?  If the latter, then the problem is likely due to coupler overhang mismatches. If the former, then it likely has more to due with gauge.

[thomasjmdavis]

Question- do the locos derail at all speed ranges?  That is, do they derail at very slow speeds?

I have found it useful in some cases to run a loco through at a crawl and watch it closely.  At very slow speeds, it might not derail, but you will see a "bump" at particular points indicating that might be a place where it could happen (if the track) or allows you to see the truck getting caught on a ladder or step or other hindrance on the frame or shell, or a wheel riding up on a miscast "spike head" or other obstruction.

Another thought- have you tried rolling a 6 axle freight or passenger car through the same curves- and do they have any issues negotiating the trackage?  If the problem relates to the number of axles we might expect such cars to be bothered as well.  When testing my track, I will rest my fingers on top of the car and roll it through- which allows me to feel any hitches or bumps that the wheels encounter.

[mmagliaro]

My first suggestion is to back up and start applying basic system engineering principles to this.
Start with one engine that derails, and ONLY one.  Pick one that derails the most.

Run it around the layout at one speed, exactly one speed, and record what speed that is,  with no cars.
Every time it derails, mark the spot where it happens.   Then pick a spot where it happens most, or at least a LOT.
Then we can begin to analyze the problem.

The way you have described it, there are too many variables:
- six-axle diesels - okay, but how many?  which ones?  every single 6-axle diesel you own does this?
- most curves more than 13", none less than 11" - okay, but where do MOST of the derailments happen?

You need to narrow down the problem or you will chase your tail forever.

[bbussey]

Are you using easements between straight track and curved?  Or joining straight directly to the 13” or 11” radius?  Transitioning between the two with a larger radius should lessen the derailments.

Disregard. I reread your post and see you’ve already tried easement.

[u18b]

I feel for you.  Most of us have had to search out trouble spots.

As others have said- more info would be helpful.

List for us the locos that are derailing.
Kato SD70ACe has a much longer wheelbase than a Kato RSC-2- or even SD40.

As Max said, there are way to many variables.  For example, some Kato locos have larger flanges, some have very small flanges.
And some of the more modern Kato locos have a shock absorbing center axle that can be problematic (like SD70ACe and SD40-2 mid version).

[nickelplate759]

Max has outlined the method to follow - find which places derailments most happen at, and which engines the derailments most happen to and focus there.     One issue I've had - a small vertical dip in the outer rail can enable a 6-wheel engine truck to climb the rail.  A small kink might also do it - I unknowingly had one such kink that gave me zero problems until one engine (happened to a Life-Like lead-sled FA with NWSL low-profile flange wheels) found it and let me know about it.

[basementcalling]

Are you using easements between straight track and curved?  Or joining straight directly to the 13” or 11” radius?  Transitioning between the two with a larger radius should lessen the derailments.

Disregard. I reread your post and see you’ve already tried easement.

Uni-Track so probably no easements.

Are the engines derailing or the car behind them? With most newer 6 axle offerings having body mounted couplers, long wheelbase modern engines like SD70s or SD80MACs and GE units can pull the lead car off the tracks because the couplers don't swing as far to the side if body mounted as they would if they were truck mounted.

Also, 11 inch radius curves are pretty tight for long modern equipment. While most N scale equipment USED to be engineered to take 9 3/4 inch radii, not all new engines are, and even some cars like the late Tangent coal hoppers with body mounted couplers stated clearly they were best run on 18 inch min. radii.

[NtheBasement]

What is the radius of your curves?
Check the coupler swing and make sure the trucks swivel freely.
How flat is your track at a failure point?  If you put a steel ruler on the rails, can you see daylight anywhere between the rails and the ruler?
A lot of light helps - point one of those goose necked desk lamps at the suspect area.  I had a derail problem that actually occurred a good foot before the derail became apparent.

[peteski]

Uni-Track so probably no easements.

I had to read the initial post 3 or 4 times to find that info.  For some reason "all Kato track" was not registering as "Unitrak".  That limits the curve radius to just few fixed values.

[Point353]

Using Unitrack doesn't necessarily imply a lack of easements.
I have a test track loop made with Unitrack.
The curves begin/end with one section of 20-150 28-1/4" radius curve track, then one section of 20-160 19" radius curve track and the rest from 20-180 15" radius sections. 
 

[nkalanaga]

The BLI locos I can see being a problem.  But I would think that Kato locos, even the modern ones, would run on Kato track.  After all, they sell a lot of American prototypes in Japan.

[robert3985]

Welll....since nobody's said anything about it, the very FIRST thing I'd do is check the gauge on all the engines' wheelsets.  "Within range" isn't good enough.  Make sure they're "perfect".  I have NEVER had any diesel engine come brand new properly gauged...they've ALL required re-gauging. 

The only engines I ever bought that had all the wheelsets properly gauged from the factory were all of my Kato FEF's and all of the brass LIK GS's my son had.

If you properly re-gauge your engines carefully, getting the gauge perfect, and they still derail, run them one at a time where they're derailing, getting down low and watching where the trucks jump up before they come off the track and put your finger on the side of the track at exactly that spot.

Take the engine off the track and then visually inspect your track to see if there are any misaligned track joiner mistakes, debris, kinks...and then wipe your finger along the top of the track and see if you can feel something that you can't see.

If your engines are all properly gauged, and they're still derailing in the same places on your curves, then the problem is your track.

Another suggestion would be to increase the coupler draw-bar lengths between engines for your 11" curves, which is quite sharp for long passenger equipment.  I don't know if your models come with both long and short couplers, but if not, then find Micro Trains couplers with long shanks to replace the Kato couplers between engines. t's good when running long passenger equipment, to look at the space between engines or cars from the top while the train is running, to see if the ends of the engines or cars are touching each other when going around sharp curves...that's what the longer couplers will do...make the ends of your engines less likely to touch and bind on the sharp curves.

That's it.  Use these techniques and the order they're in to find your derailment problem and fix it.

Cheerio!
Bob Gilmore

[OldEastRR]

I had a problem with a KATO E8 derailing at a certain turnout every time it came through it. Another 6-axle engine would derail in the middle of a curve. I did everything I could to find out what the heck was derailing the locos at those points and found nothing. So then I slowly ran the locos over the entire track routes, carefully watching the trucks to see where they came off. Turns out the derailing happened feet before the engine actually derailed. For the turnout derail, the loco's front wheels slipped off the rails in the middle of a curve before the turnout -- and when it hit the frogs of the turnout it flipped off the track. For the other the loco front wheels slipped off the rails  just as it left a bridge and then partway around the next curve at the tightest part the whole truck slid off. The problem was in both cases was a very short section of track where one rail dipped slightly below the other for only a fraction of an inch.
The thing was these were 6 axle locos whereas the 4-axle FAs and RSes had too short a wheelbase for an axle to float over a slight dip in the track. With 3 axles, two of them on the level track would allow the 3rd (front) axle float above the dip in one rail just long enough for the flanges to slip off the rail and wind up slightly derailed. Then that misplaced axle would derail the entire engine on the next tight curve or frog. What made it more frustrating was that some of my 6-axles would sail through with no problems. I realized it was because my older locos had bigger flanges than the more modern ones. So an old KATO E8 would run fine but a very recent one would derail. That was the wackiest part of all -- only SOME Kato E8s derailed!!
This of course is a drawback to having very small flanges on N locos nowadays. It means people have to lay track as perfectly as possible. Obviously with pizza cutters the oversized flanges could never be dislodged by slight dips or flaws in one rail. If you do have access to any of the big-flanged older 6-axled locos you might see if they also have derailing problems. If not, then uneven parallel rail heights might be the problem.