Minimum radius for non-mainline 2-rail tracks & minimum turnout number

I have always felt that 48 inches was a good working minimum for most diesels.   I am not familiar with the new modern large 6 axle units.     They may not work as well with that.

I have Sunset 3rd E7s and an old MG SD9.    They will both handle such a curve.   My mainline curves are 52 inches minimum but I have a branch with 48 that I test them on.    A lot of the new stuff from Atlas and MTH have 2 truck mounted motors and a lot of truck swing.    My guess is they would work pretty well as far as movement on 48 or even possibly smaller radius.    There might be a problem with coupler swing however.   You would have to test that and maybe make some modifications.      I did have a pair of OMI E8s at one time to work on for a friend.   The chassis did run on 48 inch radius.    So Tom's assessment sounds good too.

I have a lot of 4 axles diesels and steamers as large as mountains and mikados and they run well on 48 inch or even less.    Also all my 40 ft and 50 ft would probably work on 40 or 36 inch radius.

I have mostly #6 turnouts on my layout.   I have no problems with that.    I have not tried #4s.   i think they might work for slow speeds and shorter trains.     I do a lot of switching and backing of long trains through switches into industries and in yards.  

By the way, I do have 2 passenger trains that use sunset 3rd rail GGD cars.    These are 5-7 cars long and run well.    They have 2-3 coaches, some head end, and a pullman.     The coaches are the P70s and P70FAR.    All these cars are full length.

I use Atlas #5s all over with my MTH six axle diesels. I run them into my engine yard with non-powered frogs and they are fine. I bought some OP switches that are very tight (3s or 4s?). I put them in a car yard and the long cars don't like them. They were in my engine yard and I yanked them out.

I have a couple of tight bends in certain places. The larger steamers always show me them before any MTH diesel will.

I had a dip in my largest curve and a newer steamer would derail. The MTH diesels never have. I didn't know how bad it was and after the steamer derailed, I put a long straight edge to it. It was quite a dip almost 3/16" in a small area! I had been running over it for years. This new Texas type needs big smooth curves.

My Alleghenies do better with the articulated motion. My tighter one ( I converted) still likes bigger than 60" radius curves. The first, will do almost whatever the diesels will. They all run my Atlas #5 switches that I feel would be a good minimum for U. I now have their #7.5s on the mains.

 I don't have any (larger horizontal) drive diesels except for some (smaller) Atlas switchers. Originally, 60" was among my largest curves. I would guess the smallest are in the 40's " (46"?) used only in the yards. I now have some 72" curves and wish they all were larger on the mains. I have four axle diesels to run in the yards if ever needed.

I'm not that experienced in O scale layout building yet, but I would strongly advise to get the layout smooth. Sharp dips and bumps are worse than tight radii! I've altered my bench work and didn't realize how bad some areas became. I rush things and wish I did it right. I have some handicaps that limit what I now can do. I used to be good at carpentry. Maybe it's my old eyes too??

#6 Switches should present no problems at all for 2-rail, fixed pilot 6-axle diesels; they don't for my Atlas SD40, which will actually take a curve as tight as 24inch radius!! It won't pull any stock around such a corner though. It can pull stock round a 36in radius curve though, & I think that is the minimum recommended by Atlas themselves.

Here's a video clip of my Atlas SD40 taking the 2ft curve on my small layout "Schiller Point".

ALL of my MTH six-axle scale-wheeled diesels have been able to sneak through 36" radius and go through #5 turnouts easily. Even my scale-wheeled Big Boy pulled it off. I run them on the club's 3-rail layout (Gargraves/Ross/Atlas). What causes problems for me is when they hit uneven track and the truck rides up on the pickup rollers allowing a wheel to derail (lately we've had some temperature/humidity-related distortions in the benchwork). My build design is for 42" radius minimum with #4 and #6 turnouts.

I think I figured a work-around for the roller problem (other than the obvious of going full 2-rail) but it's going to take time to modify all of those pickup rollers.

Hi Naveen,

One thing you want to be aware of modeling modern era equipment versus steam era is that modern era minimum radius can be more defined by the rolling stock rather than the locomotives.  Six axle locos do require some decent curves to operation reliably but here’s an example where the rolling stock “can” define the minimum radius.  In my case, I want to be able to run the new Lionel 86’ Auto Parts boxcars (converted to 2 Rail) in a mixed freight, when I couple one of these long boxcars with its long swing out to a shorter piece of rolling stock and go around a tight curve, one is going to pull the other off the track.  My test case was a 50’ boxcar coupled to an 86’ boxcar and it made it around a 60” radius, but you can see any tighter of a radius and one them is getting pulled off by the other.  Had I coupled a modern Atlas 34’ single bay air slide covered hopper to the boxcar and it wouldn’t make it around that same 60” minimum radius and I would have to use maybe something closer to a 66” or 72” minimum radius.  However, a 6-axle Overland SDxx pulling the same train could probably navigate a 50” radius with no problem.  Now if you want to run those same 86’ boxcars in a unit train or couple them, to say, some converted Lionel 89’ auto racks or Atlas 89’ piggyback flat cars, you could probably get away with, say a 55” radius (just guessing here and you would have to test to confirm) since all the cars in the consist have large overhangs so they can work together going around a little tighter curve.

Therefore, the above applies to yards and industrial spurs too.  Classification yards “can” end up being even more of a radius hog if you put your yard on a curve and you want to automatically couple on a curve without using your fingers to help line up the couplers.  Automatically coupling long cars on curves can easily require well over a 100” or more radius.

Track configuration can affect your minimum radius (at that location), as you have to be careful on reverse curves.  Reverse curves can easily derail two long cars coupled together as one car swings one way and the other swings the other way as the two cars are in the middle of the reverse curve.  So NMRA recommended practice recommends at least one car length of the longest car between reverse curves.

Another “reverse curve” gotcha is when two turnouts are used together to form a crossover from one parallel track to another.  This too creates a reverse curve as the train crosses over and in cases with long cars and/or large locomotives, you need to use larger turnouts (say two #8’s instead of two #6s) even if all your equipment can easily navigate a #6 turnout on its own with no problems.  Now again, these are isolated cases an don’t need to be applied to the entire layout, just saying that a certain minimum turnout frog number that works in some cases, may not apply everywhere.  You have to watch out for reverse curves on certain types of yard ladder configurations too.  As a rule, I would bump up the frog number at least 1 or 2 sizes when laying out crossover to improve reliability.

As a rule with modern 2-rail:

I like to use the following:

#6 on yard ladders (you are going to be shoving long cuts of longer cars back and forth through these turnouts so an extra frog size (#5 up to #6) really pays dividends in reliability).

#7.5 or 8 on passing sidings and crossovers

#6 on industrial spurs but can go with a #5 (again you will be shoving cars through these switches, but usually shorter cuts and quite often at very low speeds so #5’s can work, but #6s do make it look and operate so much better if you can swing it).

One thing that people seem to forget with O scale is less is sometimes more so don’t overdo it with track and switches.  You don’t need a lot of switches and big yards to have fun with today’s O scale.  With wireless walkaround DCC control, great DCC sound decoders (which is improving all the time, by-the-way), and the great slow speed and tracking capability of O scale, it is a blast to run trains even if the layout doesn’t have an industry for every car type you own or a 20-track classification yard with full engine facility.  Today’s modern railroads are lean and mean anyways, a four track yard and handful of industries can be just as fun (and prototypical), create plenty of operation to keep you and some friends busy for hours and you don’t have to cut corners on radius or turnout sizes, let alone shell out a bunch of cash to buy all the materials (that flex track and those turnouts ain’t cheap, no matter who you buy them from) to cover up every inch of bench work.  Then don’t forget you have crawl under the layout and wire all that track up too.  Less can be more in the long run…

Sounds like things are coming together for you.   Can't wait to see those big SDxx’s in action.

Scott K.

Austin, TX

Those are very good points, Scott

"Less is more" definitely something to remember with O Scale. Also consider how much time you have - both to build your layout, & then actually run it!! These are points made by Lance Mindheim, a well known HO modeller, but equally applicable to O, if not more so.

Here's a picture (sorry it's not top quality) of the Industrial branch on my under-construction loft layout. Length viewed here is about 14ft (behind the view here the switching lead curves around to the other side of the loft where it joins the Main) There's just two #6 switches (second one is in the distance!!), so three spurs, that comfortably hold 3 or 4 cars each. It looks stupidly simple on paper, nothing complex or 'exciting', but to switch it, pulling 'empties' first then returning with 'loaded' cars, can take upwards of an hour or more, taking things steady like the real thing.

Track is hand spiked; Code100 rail on ties cut from coffee stirrers. Turnouts hand built with home-made self-guarding frogs, Peco point rails and Caboose Industry N-scale ground throws. Not yet ballasted.

I get around the large car problem by dedicated track in my yard, reserved for the larger cars (unit trains). When U layout your yards, there will be tracks that are easier for the larger equipment. I believe that's what the OP is asking when he said what is the minimum? So just like the engines, we should ask what are the cars? If you want all larger (longer) rolling stock, the minimum curve would go up. Short rolling stock can usually handle any 2 rail stock radius in the yards. Mix them with longer cars, and the problems show up.