First, a bit of my

First, a bit of my philosophy.  Any 'invisible' track gets at least as much consideration, planning, and building, as any track that I hope to see running well out in the open.  In fact, I go especially carefully on hidden track, and I do not compromise a bit on my curvature minimums.  If I want fast reliable running on my mainline 36" curves, the hidden tracks get the same configuration and finishing...except for ballasting and painting, of course.

For me, I run long heavyweight passenger cars, and the incredibly huge and long Rivarossi H-8 Allegheny 2-6-6-6. On curves with radii smaller than 36 ", I feel (my opinion...) it looks almost silly the overhang is so bad.  The diaphragms between the long pax cars, their close couplers, and the overhang on boiler fronts and cab roof extensions just spoils the illusion for me.  So, I adopted an absolute minimum of 33", and that was not in deference to my Allegheny.  It was in deference to a brass Sunset CPR Selkirk 2-10-4 which really is good, confirmed by me, down to the claimed 30" radius curve.  I wanted an extra 10% to help reliability at track speed.

On my last layout, my 28-ish" curves were hard on the Walthers' heavyweights.  I got derailments.  The literature claims 24" is the minimum for those cars, but I call BS.  If you want reliability, pulling and shoving strings of them around a curve, and it's far worse up a grade, you must be closer to 28-30".  If you are willing to clean up the bolsters, file down contact heads, even trim stirrups or whatever, you can get strings of the Walthers' heavyweights to run well down to about 27-28".  That's my experience...I'm sure some reading will have done much better.

For my helix and mountain hidden tracks, nothing less than 33", and the outer curves (the main is twinned throughout) are 36" minimum.  Most curves out in the open are close to 36-40".

I use either Peco Streamline Code 83 #6 turnouts and hand-made #6 and #8 turnouts.  I recovered and reused a curved #9-ish that I hand-laid for my last layout.

Staging gets the same 'hard-to-get-to' rail philosophy and engineering as the other hidden rails...nothing less than 33".

My longest cars are the Walthers NYC heavyweights which run about 80 scale feet, or close to 11", plus couplers and diaphragms.

12.5" and 14"

On my CNJ Bronx Terminal layout, the radii are 12.5" and 14".  This is in HO scale, and has been scaled precisely from the prototype drawings.  (90' and 104' radius).

-Tim

My design thoughts

On previous layouts the thought was how much can I squeeze in this given room...started with the the "standard" 18 inch curves with a few "broad" 22 inch curves with the inevitable results...when I planned this layout I looked at the space available...a lot...and decided on the maximum curves I would like in an ideal situation...48 inches, Of course this cuts down on the amount of railroad that will fit in the space, but so be it....so instead of starting from the point of view of how much can I fit, I looked at it from the opposite end as to how big can I make those curves....Then came the aisles etc. and before I knew it I connected the curves with tangents and there it was...now for a confession...to maintain aisle width I did have to tighten one curve to 40" but hey we all squeeze a little and I had left room to tighten it up a bit. I haven't done the mountain area yet and by choice I might tighten up to 40 inches again,,,depends on the look I'm going for... but whether it's a 4 8 4 Pocono with a string of 80 foot coaches or if I slip into my 70's mode with the EL SDP 45's on the head of a hot shot TOFC train those sweeping curves do it justice...

Staging is the same as track inside a mountain...why create a problem where it is difficult to get to it?

Now I fully realize that for most these cures are out of the question, but the philosophy is not...don't try to put 10 lbs of four in a 5 lb sack...better to comfortably put 4 lbs in and have a little wiggle room left...My 2 cents...Joe

30" branch and 36" mainline

I set my minimum radius on my HO Siskiyou Line to 30" on the branch (lower deck) and 36" on the main (upper deck).

I rely heavily on the scale-independent curve radius guidelines in issue 1 of MRH (see back issues).

The thing I really like about the curve guidelines in issue 1 (they come from the Layout Design SIG) is they're:

1. Scale independent
2. Are based on rolling stock performance and operating realism
3. There are several levels of guideline, you just pick the trade-offs you can live with
4. The guidelines also include the effects of radius on couplers, which is crucial if you're into realistic ops

I find the MRH Issue 1 guidelines answers everything for me, and I go into a given radius choice knowing exactly the trade-offs I will have to deal with.

I personally have not seen a better set of curve radius guidelines than what we presented from the LD SIG in issue 1.

Did I mention the guidelines are scale independent?

The best radii for running

The best radii for running long cars at high speeds seems pretty clear, but what about the other end of the spectrum? What is the real minimum radius that a small wheelbase loco with short cars can be run ? How about at slow speeds? I am in the planning stage right now and these questions are at the top of my mind, as it dictates what scale and gauge I will go with. For reference, my chosen time frame is 1880s, modeling a narrow gauge "make-do" mining operation. Small locos (0-4-0)  and critters derived from diesel switchers only. Single truck or very short dual truck ore cars are the primary loads, with very short flats and boxes (definitely under 36'- thinking more like 24' max) to haul equipment and supplies. Leaning towards Sn42 using HO track , but might go HOn30 due to space restrictions.  What would be a minimum radius for HO track with these criteria? Ideally I want to make a 35" wide cabinet based shelf "U" around the room , which would require a ~14"-15" radius turn around loop on the end. Is it realistic to bend HO flex to this ? What if I want to make it even tighter, say turning around a mountain at the end of the layout? We are talking about creeping slow and steady here, no Casey Jones' on this crew...

Curve guidelines in MRH issue 1 to the rescue

The curve radius guidelines in MRH issue 1 will get you darn close to the bare minimum, and then you need to do some tests with the actual equipment to be sure.

Basically, you look at the length of what you want to run and you compute various radii numbers for it. The numbers are:

  I think Joe hit this right

I think Joe hit this right on the head.  And Crandall made a very good point of non-visable trackage.  Your minimum should be determined by starting with the LD Sig guidelines Joe listed.  This is the absolute minimum for the 'on' or 'off' stage layout trackage.  If you can swing a more prototypical 'wide' curve in a viewed area of the layout, that is a plus(as long as you do not compromise the rest of the layout).

Branch lines can be a special case.  If you are only going to use shorter locomotives that what you are running on the mainline, and the rolling stock that will be running on this branch also will fit into this smaller radius - them it can make sense.  In my case, the same locomotives & rolling stock are run on my Pecatonica branch as on the mainline - Thus no compromised radius trackage.

My original layout had 22" radius curves(HO), and this worked as I was running 4 axle GP9's, 40-50' freight cars and a few 60' passenger cars(old modified Hobby Line or English cars).  Guess what, SD7/9's became available, then BLI dropped USRA 2-8-2's on me, and Walthers offered correct HW passenger cars(that will not run on a 22" radius curve without heavy modification).  I made up my mind that a new layout was going to be my 'Retirement Project'. - And I am very close to having the room prepared.  I hope to go from 'Layout-Less' to having a layout again over the next year.  At least I can run my trains at the club.

Jim

Absolute minimum radius

Thank you for that thorough answer Joe, I appreciate it. The actual length of car and loco in relation to minimum track radius was exactly what I was wondering about. My last layout was 35 years ago, and I just recently caught the MRR bug again. I have a test loco and some flex on the way, so now I have a good baseline to try and figure out just how tight I can make the curves on my "Cabeza Dura Corporación Minera de Oro" layout.  The tighter and funkier the curves the better, but I do actually want it to function nicely, if perhaps a lot slower than a normal line.

Edit to add: I have begun my new layout dubbed "Isla de Caldera Cabeza Dura" . Click here to visit the thread

Often, what comes out from

Often, what comes out from all the forms of advice expressed in lengthening threads like this one, and there are literally hundreds scattered here and on other forums across the www, it is really the items themselves that determine what will work and what won't.  In other words, the owner of the rolling stock, with his/her skills and materials, is responsible, and advised, to find out what works.  That usually means mocking up curves and grades, using turnouts of various frog angles and configurations, and so on, and finding out what you can really get away with if you must keep your track plan in a tight space.  The locomotives and the trailed items they tow will let you know quickly what they can tolerate.  Then, back it all up through the same track, shoving the formerly trailed items this time.  Often, that's where you learn that rolling stock behaves differently on a curve when the couplers are bunched.  You'll be picking stuff up that derails or rolls over.  Better to learn that on a mocked-up track temporarily laid than six months later after you have done all the ballasting, weathering, and scenicking.

Joe mentioned that, but I wanted to reinforce it.  Do trials!!  Dance with the locos ya brung, in other words.

Crandell

22"

Bill,

I'm running 22" minimum mainline radius for the visible and staging sections of my layout. I know this goes against the grain on a lot of layout planning, but I wanted as much operation potential as I could get out of the limited space I hade. This radius works for me, as I am modeling a freelanced N&W layout. The shorter coal hoppers that make up most of my rolling stock do well with these curves. The longest engines I have are some U30c's. They operate just fine on these curves, but would look a little better from the outside of the curves if the radius was larger. That's a trade-off I'm willing to live with though. Overall, I've been very pleased with how things have turned out. 

I read the article

Hi Joe,

I read that article about a year ago and again today before I posted this topic. I think it is very well written, totally scalable, and makes perfect sense. Well done.

Based on the article, I should be looking at a 36" minimum radius on my layout using the 3.0 calculation.

I have 2 spots where I have a ~33" radius, but it is the one 30" in the staging that has me concerned.

I'll be running mostly 50 and 60 ft equipment except for those wonderfully long 86' to 89' flats that will make the occasional appearance.

I posed the question to see what people are doing in the real world

Cheers!

Radii

I'm in On30 in a spare bedroom, modeling a (very) narrow theme - a Maine 2 footer.  The largest radius I could comfortably work with while planning was 24", so I set that as my minimum and only had to cheat to 21" in a couple of places.  Visually, it works OK, though bigger would be better.  Absolutely no problems with any equipment, but then my biggest engines are little 2-4-4 Forneys and a 2 truck Shay, and all my freight cars are under 30 scale feet long.

At the club (HO, with a Southern railway mainline theme), we use a 34" mainline minimum, but allow passing tracks on the inside of a curve (32").  On branches, we allow 30", down to 24" on one turn back curve at the end of a peninsula.  Again, we have no problems with any equipment, but of course I'd prefer even broader curves if possible, just for the visuals.

In my experience, the real answer is to use the largest radius you possibly can in your space, and still get in the scenic and operating elements you want.  That does not mean you should try to execute a modern Class 1 mainline theme on a 4x8 sheet of plywood!  Joe's guidelines above from MRH are a great starting point for thinking about what kind of theme might be possible in your available space.     

22 inch minimum

I run 22 inch minimum radius. Would I like to have larger?  Sure.... but I don't plan on running passenger cars and my freight cars are mostly going to be 40 foot boxcars and USRA two bay coal hoppers.  Locos will be Consolidated's and maybe a Mikado or two plus some short early diesels.

FWIW

30" min on mainline (with 2 exceptions of less-than-90-degree curve locations 24")
24" min on spurs
44" min in yard
30" min in staging and hidden
4 axle diesel
50' freight
HO

No passenger, no steam - can't make money hauling people, steam is obsolete in 1969

"what others are doing

in the real world" is always going to be subjective. Space, space, space, layout goals (operations or scenic), space and space are always going to be the guidelines for this question.  I myself have a kind-of limited space (as defined by the CFO of the Meramec Scenic....also happens to be my beautiful and loving wife...) of 11 x 16 plus an extension 'south' for a service/turntable yard of about 6 feet off of the 16 at the right side...I have a big "L").  I wanted the whole basement....ahem....

As things are, I insisted upon 33"-minimum radii (running heavyweights in the 50's, transition-era so there is steam too).  I went for what looks 'good enough'.  Not perfect (huge radius needed for actual prototypical appearance) and not 'fun' with lots of curves (those passenger-cars just would NOT look right on tight corners, even though I could have fitted a whole lot more onto my layout!!!).  My 'L' area is planned to be the maintenance area.  I could/should have tighter curves because of low speeds used, etc but at this point I guess I just want a consistency throughout my layout.  Perhaps I want to run a heavyweight into the maintenance area and it binds on a 24" radius?  CRAP!!  I have to, in my case, plan for the 'what if I want to...' even though it may not be a realistic prototypical scenario.  It would bug me if I could do something I wanted to only because "X" and "XX" and "XXX" can't make that radius of a turn.

God knows I want to though!  The space I could have used using coal hoppers instead of passenger heavyweights...!!!

Go big if you can

I'm just working on my trackplan, but I am leaning towards 50 inch minimum curves on the visible track wherever possible and 30 inch on hidden tracks.  I'm doing European prototype, so the passenger cars are long (26.4 meters scales out to about 90 feet or about 12 inches long in HO) so they look a lot better on broad curves.  The models themselves are engineered to run on 18 inch radius, believe it or not, and can do so, but everything will work and look better with broader curves.  

I orginially considered using sharper curves, but decided that the room was too narrow for a peninsula with a blob turnback at the end.  That left me with almost 12 feet to turn at each end of the room, so I thought why not go big if I can get away with it.  I will take full advantage of the tighter radii possible on the hidden track to make the staging work the way I want.

Would like to but!

Spiral Easements

There is a lot of great discussion about radius requirements.  I just want to pipe up and mention spiral easement.  Easement should always be used because they improve reliability by helping trains in and out of the curves.  The added bonus is they make the transition from curve to tangent look better and the curves look bigger than they are.   

Easements

And how are Easements properly done?

properly?

On the Sudbury Division we use spline roadbed which will create natural easements.  In fact, the danger with spline is that your curve will be all spiral, which will end up as a parabola; you have to take care to ensure that most of your curve maintains a constant radius.  

On track built on plywood, I make an easement by offsetting the tangent from the radius of the curve by about half an inch and then using a piece of spline or flextrack to find a natural spiral to join the two.

My Curves

The curves on my CNY&NE are 34-36 for the mainline and 24-28 for secondary tracks.  My minimum is 24" and I really try to hold that to just a few places on my layout.  My choice is based on operational reliability and good looks in that order.  I would have increased my curve radii if space allowed.

Easements

In the past, I have also used a piece of "spline" to make my easements, just using it to join the tangent to the curve, but I was wondering if there was a "calculation" or "formula" for it.

Easements

1. As Jurgen stated, spline construction will automatically create them;

2. Many track planning software packages will automatically create them;

3. Calculate the path mathematically like real railroads do it (really tough to do unless you are into that sort of thing);

4. Use a fixture like Sweepsticks for the center of the curve then eyeball each connecting end to tangent track;

5. Use a string wound around a drum as a compass to layout each half of the arc;

6. Let floppy flextrack (Atlas) relax at the ends to automatically create them;

7. Eyeball the whole track laying operation.

Did I miss any?

http://www.nmra.org/member/sites/default/files/datasheets/Trackwrk/d3b3.PDF

Have fun!

There is an easement formula...

There is a simple formula to get a good easement - this page has a decent description:

http://trackplanning.com/easements.htm

The rule of thumb is you want the total length of the easement to be about twice the length of your longest car.  Plug that length into the formula:

Offset = (2 * L) / (24 * R)

and you get the offset from the tangent.  See the illustrations on that page, which make it clearer.

The closer you come to the lower limit of radius, the more important it is to get the easements right - a slightly tighter curve with a good easement will work better than a slightly larger curve with no easement.

- Ken