Atlas number vs radius turnouts

I'm curious too, especially in light of Matt's comment about moving crossovers to the last curved track in one of pro hobby's threads.

Can someone please explain what a number turn out is. 

Numbered turnouts have straight track through the frog in both the straight and diverging routes.  Radius turnouts have curved track through the frog on the diverging route.  The number is a measurement of how fast the straight route and the diverging route separate.  In a number 4 turnout, the rails will be one inch apart when you are 4 inches past the frog.  In a number 6 turnout, the rails won't be an inch apart until you are 6 inches past the frog.  Thus, the higher the number the slower the divergence or the gentler the turnout.

Atlas curves are on 4.5" centers, but their O72 switches aren't if used on a crossover.  But their #5 and #7.5 switches are.  Can the large scale locomotives go over the numbered switches used in a crossover? 

Yes, the #5 and #7.5 switches can handle any O-scale locomotive.  Additionally, in a crossover with numbered switches, the locomotive does a nice smooth transition from one track to the other.  The appearance will be better with the #7.5's but they take a lot of room.  A crossover made with radius switches means the locomotive goes through an s-curve when it goes from right hand radius in one switch to left hand radius in the other.  While most 3-rail equipment will tolerate this (stay on the track) it will look terrible.  Some scale locomotives might derail when the locomotive is on one switch while the tender is on the other.

Is a #5 better than an O72 switch for big locomotives? 

A #5 switch is much gentler than an O72 and is a better choice than O72, again if you have the room. 

Not sure if this shows anything worthwhile, but I thought I'd do a little comparison to see the effect on track alignment, etc., using different switches and configurations.

Atlas O layout using #7.5 HS turnouts for double crossovers. A Parts List is also attached.

Yes, I've been messing with track spacing in SCARM, and I think I'll go with #5 switches for the Big Boy, and O54 switches for everything else.  All my current equipment is rated at O36 or tighter, and I'd be planning to stay around that if it weren't for the Big Boy.  I asked about the O72 vs #5 switch because I thought I had read once on here that someone's scale articulate had issues on a small numbered switch but not an O72 switch.

When it comes to numbered switches, I was hoping someone would explain the turnout curve equivalence of the points (I guess that's what they are called). I understand that the diverging track of a numbered turnout goes straight vs continuing in an arc. What I'd like to know is the arc of the points between the rail track and the diverging track. It seems like a #4 switch would have a much tighter turn, maybe like O31 or O42, whereas a #6 might be closer to O72. If a #5 is tighter than an O72, you might still run into problems with the Big Boy.

Dave can you show me what the Ross #6 with 052 and 056 look like?

Don't have O52 or O56, so I did O54. Pretty much the same as ScaleTrax, only a longer footprint.

A #5 switch is about O100 at the points.

Wow! That wide huh?

The NMRA has an RP (Recommended Practice vs. "Standard") that gives the dimensions of various parts of a turnout. A #4 normally has a "radius" of 30" (O-60) at the points, but in talking to Steve (Ross) he built his #4 points to 36" radius (O-72) for compatibility with larger equipment (a very wise move). I'm pretty sure Atlas builds out their points as well. The NMRA RP is here:

http://www.nmra.org/standards/sandrp/rp12_1.html

Interesting read if you decide you want to "roll your own."

If you are going to get a Big Boy, you will need O-72 curves minimum, and O-72 switches or #5 switches minimum (Atlas).  I would stick with #5s for crossovers and for your siding, although O-72 would work for the siding depending on available area.

If you are looking at a crossover on a curve, then I would use #5 switches. The below attachment uses Atlas #5 switches, with O-72 on the inside curve and O-81 on the outside curve.  You will need a 10" and two 1.5" straights to keep the centerlines at 4.5 inches before and after the curve.  You can set up a similar crossover using O-63 inside and O-72 outside to fit your layout.

Ron

if you want to compare their operation, it's a lot easier to do a quick and dirty calculation for curved vs. numbered switches.

first take the radius of the switch you're interested in, say O72.  the circumference of a 72" circle is 72" x pi = 226.2".  since there are 360° in a circle, for every inch in length of the points, the track will diverge 360/226.2 = 1.59°.  ...let's say 1.6°

number switches diverge at a fixed rate set by the number which is the angle starting at the frog.  for argument sake, let's say switch designers will never exceed angle of divergence at the frog within the transition between the points and the frog.

a #4 switch will diverge 14.0° at the frog.

a #5 switch will diverge 11.3° at the frog.

a #6 switch will diverge  9.5° at the frog.

now measure the curved distance from the outer point to the frog of the O72 turnout and multiply by 1.6° which should give you the angle of divergence at the frog.  (i don't have an O72 turnout handy, so i have no idea).  comparing that angle to numbered switches...

if it is 6", the angle at the frog will be about that of a #6 switch.

if it is 7", the angle at the frog will be about that of a #5 switch.

if it is 8.75", the angle at the frog will be about that of a #4 switch.

cheers...gary

Thanks for the info and discussion.  I don't have any O72 switches to do the math, could someone that does follow Gary's math and post the result?

As for switches on the curve, not sure where that came in.  I don't plan on having any switches on curves, but that could still change.

Sorry, the point about switches on a curve was just another option offered as something to consider to avoid creating a reverse (S) curve depending on the layout design.

8.75" is close enough.

Don't get too lost in the math here.  The point you're missing is that the O72 switch has a complete O-72 "curved" diverging leg, whereas the any of the numbered turnouts has a STRAIGHT diverging leg.  That makes a HUGE difference when operating articulated locomotives and/or very long passenger cars that have minimum curve requirements.  As Bob highlighted earlier, that's what makes the numbered turnouts MUCH more appropriate for parallel track cross-overs and avoids the "S" curve that comes when attempting to use O-XX turnouts for cross-overs.

David

David

At risk of sounding like a bit of a crackpot:

A while back I worked out the math to find the effective radius and angle of the Atlas #5 and #7.5 switches, using the fact that they can be positioned in a 4.5" center distance configuration, and that the curve portion has the same horizontal length as the straight, if you project it on the straight portion.  Thus my calculations are simplifying the 'curve' of the turnout by assuming it starts and ends at the beginning.  In reality there is some straight lead-in before the points.

For the #5 switch I found an equivalent radius of 90", with an angle of 12.84 degrees.

For the #7.5, 134.51" radius, 10.49 degrees.

Edit: By angle I mean the angle of the curve, in the same way that the O-72 curves are 22.5 degrees.

I'd really like to figure the same out for the Ross numbered switches but to my knowledge they don't have the same geometric properties.

Adam

Just wondering going with atlas 081 and 072 4.5 center rail would be safe for too  scale articulate Big Boy to pass each other. thanks

Seems to me we've already determined that a #4 is about equivalent to an O72, so I'm not sure what good all this math is. I admit I'm not sure exactly what we're looking for or what problem we're trying to solve. I guess it might be useful in some cases where manufacturers deviate from standards. Like David said, the math is great for determining the "effective curve" for numbered switches, but just because a #4 is close to an O72 doesn't mean a large engine is going to work well with O72 or larger switches forming a crossover when compared to using numbered switches.

FWIW, I put together this little chart to show some differences which I think might be important, mostly the resulting track separation. I find it interesting that Atlas seems to have standardized this separation and this seems to partly address CHARLES7777's comment.

Here are some different track arrangements from Atlas, using their track, that show different turnout, crossover and crossing possibilities with some measurements. I found them helpful and just thought I would throw it out for reference if anyone is interested.

Not criticizing the math. Totally agree that the math is fun, but I thought Bob answered the initial questions in the 3rd post and the math backs up what he said. Once the point was made that a #4 is basically equal to an O72, only with a straight diverging track, it seemed pretty obvious that any numbered switch is better than an O72 switch, and so on. So, I'm left wondering what question we're still trying to answer, that's all.

Of course, there are always other considerations that the math itself can't answer. For example, just because a numbered switch is "better", does that make it the right choice depending on cost, footprint, planned engines/rolling stock, and most of all, availability. FasTrack doesn't offer numbered switches, so should one change brands just to get a numbered switch? And then there's always the fact that most people will be just fine using O72's. Personally, the lack of numbered RealTrax switches is what got me interested in ScaleTrax and now I won't use a brand that doesn't offer numbered switches, mostly because of the track separation widths. I can always increase this numbered switches, but I can't decrease it with regular switches.

I found the info rtr12 offered to be very useful, especially page3 where it shows the track separation widths and wish other manufacturers offered similar illustrations. I hate having to do it in RR-Track, though I'm thankful I can.

I use Atlas #5 for my yards and Atlas #7.5 for mainline and nothing else.   I run all scale equipment with KD couplers and I'm a "snob" and also cut the drawbars of my steamers for a more realistic distance between the engine and tender.  My min radii is 090.   It's a difficult spec to build to but I like the look.   Also, my newly delivered GDD SF SuperChief with KD couplers and all the diaphragms up snug just has such a fantastic look to it (to me!).  

I would go with Bob's advice. O72 xovers are just gross looking no matter who makes the turnout.

I should have used #7.5s on the mains like Trevize here but used all #5s instead. Still 1000X better than watching things contort themselves through a sharp S turn.

Here are the approach angles for the Atlas switches. I am using one O72 switch on my layout for a spur but that may get swapped out for another #5.

Another reason to avoid O72 switches is the tendency for long rigid wheelbase steamers to short out when trying to negotiate the diverging route. Blind drivers on Decapods and duplexes can contact the center rail when doing so.

So let me ask both of you then how do you configure a spur or siding? Especially on smaller layouts? If you just add a half O72/O81 curve to the diverging track, isn't the spur/siding track then too close to the main? Do you then add several straight tracks to increase the separation? Doesn't all this minimize the useful space on the spur/siding?

Obviously, a #5 is better than a #7.5 for this purpose, but most of us with smaller layouts don't have the large engines anyway, so using O72 turnouts seems to be an acceptable compromise.

I use #5 for a siding and I don't use fixed curves either.   I use all Atlas flex track.  That way I can gracefully create a curve off the switch for the siding at any spacing I choose.

Having a 072/081 curve immediately after the switch is also creating another "S" curve so I would avoid that.   

Also, it is a compromise where you end up choosing on a small layout.  Luckily I have a decent amount of space and Norm even more so.   

I just don't think it'd be practical to use flex when my curves have to be O54 or less, though do I plan to use flex to add gentle curves on my straight sections. Maybe I could also use if for the spurs/sidings. Would I need a bender for that?

So let me ask both of you then how do you configure a spur or siding? Especially on smaller layouts? If you just add a half O72/O81 curve to the diverging track, isn't the spur/siding track then too close to the main?

Dave, your diagram shows 5.5" track spacing.  If your passing siding is on straight track then 3.5" track spacing (14 scale feet) is prototypical and works well in O-gauge.  There is enough space between trains on those tracks to get your fingers in between them if you need to.  I use #5 or #6 switches for my passing sidings and #5 for most industry spurs.  Due to space constraints, some industry spurs use #4, O96, O72, O63, O54 or even O42 switches.

Bob, I did that only because that is what the O72's worked out too, but I get your point that my siding might not be as short as it appears if I shortened the width between tracks for the numbered examples. To be honest, I originally showed the narrower gaps without any straights, but they were less than 3" and I didn't feel like playing around to get them to 3.5" or so. I'm going to go back and do that though for my own education and as a reference for my design project. Thanks for pointing that out.

I use all flex as my min curve is O90.   Atlas flex is very easy to work with when the curves are greater then O90.   Atlas flex and a rail bender don't mix.   Basically with Atlas flex you just lay down the track and screw it in place.   Very easy.   However, NOT for O54 and less.   The min I would attempt with Atlas flex is O81 and that would even be a bit painful.  

Like I said above, I'm running scale GGD passenger cars with scale diaphragms, with KD couplers.  The diaphragms on the cars are touching on straights and mash up prototypically on curves.  Also, my steamers have their drawbars cut down to bring the tender to a more realistic distance from the tender.  Many of my 3rd rail steamers won't even make an O72 curve which is why my min radius is O90.   

So anyway, I'm using #5 switches for sidings and my yards.  My sidings have to hold a GGD 8 car Superchief plus the ABBA units, which is a good 16 feet long.  My freight train lengths are 18-20 cars which works out to be about the same.  

If your planning on curves O54 and less then your equipment should handle the S curve.   You may even be able to use the Atlas O54 shorty switch and make a nice easy S for a siding.   Something IMO to test out.   On an old layout I used O54 shorty switches and had very good luck with them.  

Thanks, Trevize, you're certainly giving me things to think about. So often people just say use this or that track (which is how this thread got started) without considering how it's going to be used. If Atlas flex is only good to O81, then maybe it's not going to work for me. I'll redo my "straight" sections with O81/O90 to see how the sweeping curves might work with flex. I've been using O72 to mimic things, but if flex will not work, then I need to rethink things. All I really want though is to avoid the straight look. I wonder how far and easily ScaleTrax bends? As I said, I've kind of nixed ScaleTrax because I don't think I can reasonably stack O54 loops and it doesn't have a curve in the 40's. Even though all my gear will run on O31, I'd like to avoid it if possible. I still need to experiment with an O31/O54 combination, but I think that might look too funny even before trains get run.

 Old layout with a yard using O54 shorty switches

New layout with #5 switches

Track spacing is much closer with the #5 switches which allows for more tracks in the same space.  Also, long cars with KD's can easily pass through the yard.   This was not at all true of the yard built with O54s.

Trevize, thanks for the photos. While it's a little hard to visualize the overall space difference because the layout is so different, I find the new yard much more interesting. Do the yard tracks dead-end in the back?

But, there are sooooo many nice LONG places to park trains!!!

Looks real nice to me!

It dead ends at the back.   It is actually 2 yards split with a walk-way down the middle.   It is approximately 25 feet long by 8 feet wide.   The width being split by a walk way with 4 feet on each side.   I also have the structural support poles I had to build around which are giving "interest" into switching operations.   

I've got an O31 loop going around mine as part of a reversing loop. It allows me to enter the yard from one side, loop around and then back trains into their slots. It's 3x10 with a 1x10 shelf at the front off to the left. I've switched to curved turnouts because of the "L". They give me more storage than a head-on yard with regular turnouts and the Ross 3-ways/4-ways have too big of a footprint.