I’ve noticed this a few times now, most recently when my SD40 derailed at two spots in what seemed to be a +24" radius curve.
Careful examination revealed slightly tighter kinks in the curve.
I have not experienced this with the other three brands of flex track I have used.
Easy to straighten the kink out though, which is good.
Maybe MRR could do an article on kinkiness. Take all the brands of flex track, bend them around a fixed object going to smaller diametres each time until WHAMO! We get a kink. Kinda like when Boeing bends the wings on their airplanes until they break. Now that’s a test worth watching.
I take it you are not using things like track alignment gauges.
Or taking care to lay out the curves accurately before laying track, perhaps using something like a trammel:
Ed
As far as the kinkiness of the various HO Code 83 flex-track lines, I think most of that is due to the pudginess of the rail cross-section.
Microengineering is thinnest, and close to scale. Then a bit fatter is Shinohara (and likely Walthers). And the chunkiest of all is the Atlas.
And the disinclination to bend sideways goes upwards in that order.
Ed
This suggests that Walthers flex track produces slightly tighter kinks in the curve than Atlas flex track, based upon the title to your thread. Rest assured that Atlas flex track does not produce kinks in the curve at all. I cannot speak for Walthers flex track - - never used it.
Kinks develop on curves with flextrack where adjoining pieces connect without a soldered joint on the outside of the connecting rails.
They also develop when there’s a sloppy installation–not so abrupt, though.
For that matter, they can also develop if there’s thermal expansion joints missing.
Ed
I have no problem with track that gets kinky so long as it does so in the privacy of its own home.
Thanks for the advice but I was merely making the observation as this Walthers product is fairly new and differs from the existing brands. Maybe nobody else has encountered this issue but maybe somdone has and this information may prove useful to them.
One factor will be the thinner tie moulding.
Another factor will be the rail profile: close to Peco and ME as compared to the heavier Atlas rail.
Walthers claims a particular nickel silver alloy is used for better conductivity which may be a factor.
While I have trammel points they aren’t very useful for narrow shelf layouts. I have a few track gauges and ribbon rail products, particularly the straight ones which are handy for straightening flex track back out if I change my mind.
For curvature checking I simply use Atlas 24" sectional track to verify that my minimum radius is at least 24". As long as the radius is greater than that I experience no difficulty laying out and aligning track curves by eye.
Curving flex track is a trial and error process if your objective is to prebend the curves so the track lays in the desired alignment. If you curve it and then pin it or glue it to hold it where you put it then that’s a different matter. I prefer track to be laid and affixed in place without any residual tension.
I have used Peco and Atlas, Code 83 for Peco, and both 100 and 83 for Atlas. The only times I have kinked them is at the joint. At least, until I learned how to make proper joints. Strong ones.
I do recall our late friend tomikawaTT saying he routinely imparted a permanent bend into his flex so that it would stay curved, even if not quite to the radius he intended for it. Took some doing, but it can be done. It’s still not a kink.
The word “slightly” hints to me that the track under discussion might have been laid with an uneven radius, as opposed to having a kink at a rail joint.
Still, this all points out that trackwork should be done to high standards, with a resulting lack of derailments.
Ed
I do not mean bent rails. I mean bent curves. The radius was uneven. I thought I had applied consistent force in the same fashion as for other brands.
My impression is that Walthers flex track may not bend evenly despite using consistent bending force. It’s as well to check as you form the curve and realign any sections that seem to have an incorrectly tight radius.
Or you can make your own for cheap by buying a 4’ long piece of wood at your local Lowes or Home Depot.
As long as you draw a centerline with a trammel, it should be an even radius. Laying track on that centerline guarantees a smooth curve.
While the trammel method for laying out a curve, as pictured above, works in many places, there’s two other methods I use when that doesn’t “fit”:
I cut a plywood template off the layout, and then put it in place to trace the curve.
AND
I use the Ribbonrail guides.
I also pre-solder the railjoints in the curves, and design in expansion joints to eliminate permanent or temporary “heat kinks”.
Using these tools and methods, the only reason left for kinks in trackwork is simply sloppy installation. Of course, it might be a good idea to use more “forgiving” track if one isn’t up to the task.
Ed
Curious what approach you used when laying this curve that may have contributed to this happening. It would seem from your comments you didn’t use a trammel, check with your sectional track, or align by eye. Did you use some sort of template that may be faulty or at least contributed to the issue?
I haven’t as yet laid any track down permanently, it’s just sitting on the foam riser system.
If you use the Woodland Scenics foam riser system you don’t need to lay out the track first. The bare risers will show the alignment of the track.
You lay out the foam risers instead. You check the radius of those. You can’t easily put a too sharp bend (which is what “a kink” means to me: a sharper than expected bend) into these foam risers. The turnouts all have to be located before you know where the track goes. Those sit on foam blocks or joined risers. The risers connect the right of way between the turnouts. The risers take up a naturally eased curve when you bend them and pin them into place.
After laying out the foam risers I pin them down and only then form the track. The track is connected and short trains are run. If trains run reliably on track that is not fixed down then those trains will run reliably when the track is fixed down. That’s how I found the very short bits of track that were not correct radius even though they appeared to be. Nothing gets fixed down until the trains run reliably. Very simple experimental technique and foolproof by definition.
Once your risers are where you want them to be you glue them down. I then relay the track onto the now fixed in place risers and test it again. I can run trains during construction which is good because the hobby work is slow. I have a “day” job.
There’s no point drawing a layout full scale for a shelf layout and there is no need to do so. The shelving dictates where the track must go. The challenge is fitting the curves only down to your minimum radius where the shelves force the track to go. Any radius broader than the minimum doesn’t have to be measured for radius, that curve just goes where it goes. You don’t need trammel points. This is just a model railroad not an engineering project. A yard stick held in the hands is sufficient to "su
I agree. I don’t know how alignment by eye can work, except by chance. On my early layouts, I used a trammel which if used correctly produces a reliable centerline.
But the laying of track on a centerline drawn by a trammel is less than precise. So, on later layouts, I used Ribbonrail Metal Track Alignment Gauges. Initially, I bought two 10" straight gauges and one each of the curved gauges, ranging from 26" radius to 32" radius. This works far better and more precise than a trammel in my experience.
On my current layout, realizing that a minium 32" minimum radius would be ideal for my purposes, I bought another five 32" radius track alignment gauges and another four straight alignment gauges.
The beauty of the metal track alignment gauges is that they “lock” in between the rails and hold the track in position until the track is secured by nails, caulk or whatever.
Rich
Absolutely!
My 25 x 19 U shaped layout has two broad sweeping quarter-turn curves. I have no idea what radius they ended up being, nor does it matter. I don’t need to know.
I settle on the final radius, whatever it is, after seeing it in place. One curve took about 4 re-lays to get the exact angle correctly into the switching district for appearence purposes (and I don’t know the degree of the angle.)
The other curve took about 5 re-lays over the course of 6 months, simply to create a little more room for an industry, changing the location of one industry to the other side of the layout, or a few more inches of scenery.
I’d be drawing a dozen centerlines on the plywood if I used center lines. A dozen holes in the trammel. A dozen ribbon rails, LOL.
But, in the end, a consistent radius throughout the curve is not the goal. A gradual hyperbolic curve is the goal.
How does a ribbon rail or a trammel help to make a gradual hyperbolic curve?
My curves aren’t long enough to have a fixed radius within it for very long, the total curve is generally only 90 degrees at most. I don’t have peninsulas because they always include nonprototypically sharp turnback curves.
For avoiding kinks in laying the hyperbolic curves, its pretty easy to just run a car over the track to see if there are abrupt changes in the spacing between the bottom of the sill (a fixed point) and the rail as the car moves along. The spacing should change as the car moves from the easement to the sharper peak of th
Da Truth.
The amount of planning that goes towards building a shelf around the walls layout into and average room seems like a hobby by itself. But it is not necessary unless you are compelled or simply enjoy it.
If I had a barn to fill and it was like an open canvas that I could put a bunch of turnback blobs in it, I would want to draw up some sketches first.
The photo above was only to show what the trammel I have looks like. I’ve set it up on the floor to draw centerlines and subroadbed edges so they are pre-cut to install.
In a different situation where Homasote was installed and I could not use a trammel, I made a template using the trammel and then drew the centerline from the template.
Here is an example where it was not possible to use the trammel but 95% of the time I can use the trammel.
I’ve always been a little leary of Ribbon Rail because the center-line could wanter a bit off vs. a centerline drawn with a Trammel. I’m sure it’s “close enough” but I’m anal about track geometry so choose not to use them.
Presoldering would work good with Atlas flex but I’ve found Peco would not work so well that way so I solder as I go.
