Look forward..

I look forward to your comments as well as others. Like i said earlier in 22 Stories up, I am about to starting getting into doing some for my layout. I will most likely be building some at my workbench (#6.5 curved), but I am assuming There will need to be some done on the layout itself. I ordered everything Friday from Fast Tracks, and the CVT bases from my LHS, he wasnt sure when they would be in. Actually got them for $7.75 for Central Valley HO #2602 Curvable Switch Tie Strip #6R and the same for the #6L. So well see. 

Not looking so good

Well, the demonstration only shows that you can't bend a straight turnout to make a curved crossover and hope to get the geometry right; it just can't work.

The thing is that a straight turnout and a curved turnout which use the same frog angle can't have the same lead length.  The lead is lengthened proportionally to the reduction in radius; in other words, the tighter the radius, the longer the turnout.  If you don't follow this principle, you just end up with a badly bent turnout.  The photo shows that you can't maintain a consistent parallel radius with these premade ties.  There is really no substitute for properly laid out, handlaid turnouts in this situation.

Drawing the template

Here is a quickly thought up sequence for drawing a template on your roadbed for curvable turnouts that is geometrically approximate, but probably still good enough to make sure that you end up with a smooth, and good running, curved crossover.

You can choose your own desired any curve radius and frog angle, subject only to the limitation that the outer turnout will have a much sharper inside radius, that its outer curve, especially if you use smaller frog numbers. I have used 24 and 26 inch radii for this example, and #8 frog angle.

So lets see if this works for you.

.

1. First draw both curve radii, as center lines, but then add the rail lines as well

2 Draw a vertical line where you want the center of the crossover, and an exactly  "true" horizontal mid way line across it.

3. Now we get cunning. Through the intersection of the center lines draw a "tilted" Midway line at the exact frog angle of the turnouts you want to use for the crossover.  You can probably concoct that by using one of the turnouts bases themselves, rather than measure it with a protractor.

4. This will need a little more care. A "tee" square is also a handy tool for this. Draw another line offset from the "vertical" center line, also by the frog angle. It should end up at exactly right angles (90 degrees) to the first tilted midway line. And it needs to be long enough to reach past the center of the curves (OR, if not possible - see later comment "A")

5 Now this is getting exciting! Using a yardstick compass, draw a circular arc of the average of the inner and outer radii, through the center intersection, with it's center on the 90 degree offset midway line.

6 And now the template is almost finished. Offset the yardstick radius to draw the rail lines of the midway curve. Where they cross the rail lines of the inner and out tracks, is where the frog vee points of the two curvable bases should be placed.

All you need to do now is test fit the two bare bases in those positions, curved to fit the curved tracks,  and then just fit a short piece of flex track between the turnouts to see and make sure if everything smoothly lines up. Rubber cement is a handy temporary fixative if you run out of hands at this point.   If not quite perfect, you can move the turnout bases slightly, but only ALONG each curve to adjust them. Just don't move them off their respective inner and outer curves. 

Note (A) If for any reason (e.g. no room. or the center is over the aisle), you can't use a compass to draw the midway curve on site, you may be able to make a separate paper template of the just the curve to use instead. You just need to draw a tangent to the curve on it as well, so you can line the tangent up accurately with the tilted line. A commercial curve radius template may also work here.

Note (B). If the turnouts do not fit on your ideal track plan where the template puts them, you would be wise to reconsider placing the curved turnout in that place at all, or moving something else instead . A squashed up curved crossover is likely not to work well enough anyway.

Note (C). The frog sections of the curvable bases should be left straight. Just best fit the turnouts to the curve with that section straight and don't worry about it. Both model and prototype "standard" frogs are always straight and trying to make curved ones isn't within the scope this method. And if your crossover will only fit with the frogs being curved, you are on your own!

Not so fast! You spotted the wrong mistake!

Jurgen,

The end view of the top photo only shows the inner turnout clearly. To which your comment obviously doesn't apply.  It's actually less curved . Here is the top view of the curved outer turnout and you can see it's still in good shape. Although the outer curvature is quite large in this practical case.  (around  42")

However there is a different type of construction error that you can see related to the top view of the inner turnout below.

If you carefully compare the two turnout lengths, you should spot that the inner turnout has a sharper frog angle that the outer, so has a shorter lead and thus causes the inner track to turn too sharply. The inner track just past the turnout then has to stay straight for a short distance to compensate. Someone must have opened the wrong turnout box the second time!!! (And didn't use my method!)

The (unplanned) lesson here is that the simple CV "distance" method of placing each turnout the same distance back from the center line, does not work if you have two different frog angles.

I'm not sure if the template drawing method I described instead, can be adapted easily to using two different frog angles.  And in the interests of "KISS", I'm not going to try.

Finally, for comparison the other end view from the far (outer turnout)  end looks very presentable, even though the naughty inner straight section and sharper, shorter turnout can now be more easily seen.

Just adding a link

I'm just adding a link to the other thread, so future readers can take a look.

Thanks Andy.

spotting the wrong mistake?

Using two different frog angles on a crossover is improper as well, and explains some of the geometry issues, but my comments on lead vs. curvature are still correct in any case.  You can get away with curving a turnout a small amount (we are talking around 100" radius here,) but doing it on a 24" radius is just a bad idea.  One look at some prefabricated turnouts such as from Walthers will show you that curved turnouts are much longer than straight ones.

As for that CAD drawing, marking the frog angle in the middle of the two curves doesn't make sense.  The frog angle is where the inner rail of the inner turnout meets the outer rail of the inner radius, or basically the opposite on the outer turnout.  The radius of the crossover track isn't arbitrary, but has to be such that it will make the curve allow both frogs to be in the proper position without needing any tangent track to line them up or having to vary the radius of the crossover significantly.  It can be done by eye to some extent, but if you want to lay it out in advance, you need to work out all the specifics.  In any case, the crossover radius should never fall below the inner track radius.

Curving any turnout shrinks its lead - use a larger frog #

Always use the largest frog #'s you have room for, for all curved turnouts, in order to keep the lead (and internal turning radius) as large as possible. The 24" inner double track radius I quoted above was just an arbitrary example to test the method. 

Jurgen, if you didn't quite understand why I drew the CAD that way, try starting from a straight crossover, measuring the angle of the tangent track in the center, (it is at the frog angle for its whole length),  and then imagine the complete crossover section being gradually curved in very small steps at a time (like calculus). All the same frog angles I mentioned are of course the angles measured relative to the tangent of the curved main track where they are located. So the relative frog angle stays pretty constant along the length of the "ex-tangent" track, as it becomes more and more curved, at the the same rate of turn as the whole crossover is becoming more and more curved.

As a practicing  engineer, I have learned the hard way to discount many non-technical modeling articles, and so now always get my track information directly from the prototype and prototype drawings.  I do understand what makes a turnout lead longer or shorter. And increasingly curving a turnout of the same frog number will definitely drastically shorten the possible lead. (It's basic geometry). What I think you'll find with some commercial model curved turnouts is that they ignore or guess the true frog number and just supply a longer turnout specified as a pair of radii that works instead. AKA "marketing license"

If you want to provide an example of a lengthened model turnout, I'd be happy to put it on CAD so we can see and check it out here.