I built a rather large HO scale layout using code 100 atlas flextrack with PECO and Shinohara switches including double slips and 3 ways. I ran mostly 4 axle geeps and yard engines and found the track to be pretty much "bullet proof"
Then--I started putting some of my older brass steam and a couple of later Chinese 4-8-2 and 2-10-2 locos on the layout. Then I had to re-define "bullet proof". It no longer applied to my trackwork.
Test your trackwork as you lay it with the longest wheelbase, rigid frame engine you intend to run if possible. Then you will know what "bullet proof" means.
grizlump (grouchy German)
These first two replies cover the universe of issues so well that I am not sure what else there is to add. So, here is what I will do. Without referring back to the first two replies, I will summarize all of my issues in failing to achieve bullet proof track.
1. Curves. You gotta solder all of the rail joiners before installing a curved section of track to avoid kinks.
2. One Rail Higher Than The Other. This is especially a problem on curves.
3. Humps and Valleys. Any dip or rise in the track from level will often cause derailments.
4. Rails Out Of Gauge. This most often occurs on curves where too many ties are removed to accomodate soldered rail joiners. But, it can also occur on straight sections of track. Nailing down track can also cause rails to be out of gauge if the nails are driven to deep.
5. Lack of Easements. All curves should begin and end with an easement.
6. One Rail Higher or Lower Than The Other. This occurs at rail joints and is caused by rail joiners that slip and cause the adjoining rail to be higher or lower.
7. Flex Track Meets Turnout. Improper joining results in a kink.
Hope this helps.
Rich
Alton Junction
Gidday John, having built all my benchwork and tracklaying in a club environment, my definition of “Bullet proof trackwork” is “Track that the most moronic reckless club member can run his prized piece of garbage locomotive at Warp Factor 9 without it derailing”.
As with everything in life, to do well there must first be a good foundation, hence don’t expect good trackwork if the bench work is rubbish, not that that means that the timber has to be of first grade dimensional quality, though, of course, it helps. All the trackwork done at the club has been with predominately Peco Code 100 Flextrack, with some GT Italy Code 100, laid directly on the MDF road bed / bench work, so I can’t comment on any brands or codes of track or other road bed materials.
I draw the track centre and use a homemade radius bar to draw the curves.
I’m old fashioned in that I use track nails, for the reason, especially in laying the curves I may find that I have to remove the nails to relieve any “Oops” that I’ve inadvertently built in!! Also, I shouldn’t be allowed anywhere near caulk or similar substances!!
However, I should stress that while the nail should be hit in snug, don’t over do it as you can narrow up, the gauge!
I’m a fan of the Mk 1 eyeball but I’ve also built a test car that I push around a “hopefully" completed piece of track.
I like to use an easement, if possible, on mainline curves and am also a fan of the use of super elevation. Also, easy transitions on either side of a grade.
To sum up in a nutshell, I strive to make my trackwork “Flow”.
That said I’m always trying to improve.
Cheers the Bear.
"One difference between pessimists and optimists is that while pessimists are more often right, optimists have far more fun."
1. Rails are not separated accurately. That is, they are out of gauge;
2. One rail dips or rises such that they are not at the same height measured at any one point along the centerline;
3. Both rails rise or dip concomittantly, but it's too much for the wheels and suspension of the items meant to roll over them;
4. The joints are not carefully assembled. There is a large gap on the outer rail of a tight curve, or the joint(s) are kinked sufficiently that there is too great a resulting radius at that joint. Or, one rail is outside of/above/below the joiner by mistake;
5. At some point along the general arc of a curve, the radius is sufficiently short over a sufficiently long arc that one truck or an entire wheelbase of a steamer, as examples, cannot safely negotiate the curve at that sharpest point, and there'll be a derailment, or possibly just a stall if the speed is slow and the rails pinch and bind the flanges. Or, with longer cars, or passenger cars with diaphragms, the sharpness of the curve will cause derailments when the couplers cannot swing in azimuth sufficiently or the diaphragms impinge;
6. Turnouts are the wrong diversion angle for the rolling stock's ability to run through them railed, or turnouts' approaches and exits have kinks at their joints with external rail segments;
7. S-curves, caused by turnouts too sharp or set too close to one another, or caused by sinuous geometry on regular non-turnout rails;
8. Improperly prepared roadbed, especially under turnouts, or at vertical curves into and out of grades; and
9. Super-elevation along curves inexpertly fashioned that causes derailments, almost always along the outer raised rail, and usually due to inconsistent grade. A dip on the outside rails of curves, but also super-elevated curves, will almost always result in a derailment.
Over the years there has been a steady reference in these forums to "bulletproof" trackwork. Often accompanied by justifiable pride that it has been achieved. But let me ask this question: Just what constitutes that description? Obviously, bulletproof means trackwork that is never itself responsible for derailments and other operational problems. I may be missing something from the definition here, but if so, please weigh in.
But what I'm really asking for is a list of goofs, mistakes, and imperfections that can pop up in the process of both design and installation - and have to be corrected after the fact. If you wish, please include a tip or two on how to avoid them. I'm not looking for things as obvious as insufficient radii for the length of motive power and rolling stock to be used later - I'd think that's a no-brainer from the start. Let's just make that first on the list and get to the rest of them.
Thanks!
John