Track Standards

Model Railroader General Discussion (Model Railroader)
1

It’s true, typed message often convey or don’t convey what was intended by the author. Emoticons were invented partly to help in that regard. And to be fair, communicating in ways to overcome lack of facial or body language can be a challege on forums and in emails. It can take extra effort to avoid being misconstrued. It a can be learned but even then unintended consequences still happen. [B)]

Re: Track standards and John Armstrongs book, things have evolved in the hobby since Track Planning for Realistic Operations was last revised, but much of the information is still fundementally sound and worth digesting and putting into action. Very much worth reading.

It has, but curves need to taken in context too. For those running shorter rolling stock, 24 inches may be totally fine. OTOH, those running 89’ auto-racks or flat cars or 85’ passenger equipment, may be less than happy with 24-inch curves.

John Armstrong does discuss those kinds of compromises in his book and that full length passenger cars (very much relevant in his time too) may need modifications or special consideration if to be reliably run on 24-inch curves.

I’ve read that Walthers passenger cars, even though the recommended minimum radius is 24-incyes, it is reported they don’t work very well on those curves and many recommend 30" radius or more.

A growing number of modelers are interested in post-1990 model trains and

2

I’m not sure changing the definition of very sharp/sharp/conventional/broad would be helpful.

First, who is going to set the definition and maintain it as necessary.

Second, it will be confusing because older models/articles and newer models/articles will be using different values which will confuse buyers/readers.

Interestingly, the NMRA RP11 for curvature has 13 different classes of curves for model railroads.

When discussing curves, I think a lot of us tend to forget that a lot of folks have small layouts. The hobby press is full of large layouts. But I suspect there a lot of casual hobbyists who have a 4x8 or little larger. For them the current definitions are good.

For those of us a little more involved, perhaps we should the RP 11 definitions and call our curves Class F or L or P, etc.

Paul

3

Yes true. And for that matter, why would John Armstrong even set such a labeling scheme in the first place? (Sharp, Conventional, Broad). Conventions of the time.

My only point is that what some consider to be sharp, conventional or broad has changed a bit over time. If any terminology is meaningless to you, well, blame John Armstrong for bringing it up in the first place - he was very naughty!

Never-the-less book and much of the content is still very useful. His discussion on “squares” didn’t really work for me either, so I discarded that part. Not every part of Track Planning for Realistic Operation will be useful to everyone, but a lot of it will be.

To answer the first question, apparently JA had the audacity to set the definition. But since he is no longer alive, he cannot maintain or justify it, apart from what he wrote in the book.

As I mentioned to Keven, the question of curve radii boils down to application. Obviously the smaller the layout and/o

4

I don’t think anyone is in a place to convince railroaders to uniformly reorder the Armstrong curve classification. Also, I suspect the majority of layouts -especially beginner and casual hobbyist layotus- operate in a limited space and still would find his standards represent their reality.

For those who have progressed into large personal or club layouts, they dont need to use the Armstrong curve recommendations anyway. What they need instead is for producers of railroad equipment to simply be simply more forthright about what curve measurements their rolling stock will reliably operate on. I suspect they don’t care much what Armostrong thought a conventional curve was, but they do want to know what curve their latest, super-detailed 89’ rolling stock is going to require.

5

@Elif, as always YMMV.

Regarding JA’s. curve recommendations, one of the more important things is to read the discussion on rolling stock vs. curves, which is still pretty relevant as cautionary items. Basically, as you get to sharper curves, you may need to take special consideration with respect to interference of things like underbody details and possible modifications, if necessary. That is going to be more salient than the labeling of sharp, conventional and broad curve labeling that I’m going to be made very sorry I mentioned even mentioned. [U] Anyway, y’ll enjoy yourselves doing the 20 lashes with a wet noodle thing. [:o)]

As for forthrightness, that’s a discussion you’ll need to have with manufacturers. As I mentioned, Walthers recom

6

True. However, just focusing on mainline, modern modeling tends to distort the historicasl use of these terms. Yep, everything’s bigger, needs broader curves, etc.

But that is essentially reducing the argument to modeling the present or very recent past.

Let’s not forget that there are still plenty of people doing narrowgauge. And it’s not that everything NG works on 18" radius. It doesn’t. But industrial lines do even sharper in the normal course of business. Traction is not so much modeled these days, but also fits in the JA system quite nicely. That’s leaving aside those who model the 50s and earlier, where - excepting many passenger cars - lots of stuff was fine with 18" curves.

JA also notes his reduction of curves to 3 broad classes is an attempt to simplify the 14 varieties of curves found in the NMRA Recommended Practices. Since this is an RP, even the NMRA’s set of definitions is only a suggestion. In ant case, JA notes that selection needs to be based on a number of factors besides simply measuring and labeling.

Then there’s the fact that even modern day lines have certain limits below the allowed minimums, given the life span of many lines. Industrial and branch line traffic is often fine with more restrictive practices, so long as operators know that restrictions are in place and what cars and locos are allowed on certain lines.

How JA defines what works on which of his broad classes of curves snd this info still seems pertinent. JA also supplies examples of these definitions for scales larger and smaller than HO.

Importantly, JA ends his discussion by noting that “Experimentation [i]s an aid to selection of curves” so he’s well aware of that defining the changes needed to get certain gear to operate properly is an important part of the process. We can offer all the h

7

Aye.

8

Confused. Is this a continuation of another thread?

9

I’ll wager that in Armstrong’s time as is the case now, some models will run on 18" curves, some models require 30" curves and most models will run on 24" curves. This makes 18" for tight, 24" for conventional and 30" for broad as accurate now as it was then.

Keep in mind also that as equipment has gotten longer, models have more often been altered to run on tighter than normal curves. I’d say it’s balanced out.

10

Did a bit of thinking and looking at the Armstrong standards and I think you’re absolutely correct. I assume we look at the standards as denoting reliable running and in that respect I think it’s absolutley right. Sure even bigger curves look better, but for actual operations most will run on 22"/24" and virtually everything will run on 30". Nothing preventing folks who have the space for going for wider, grander curves, but his numbers still encompass virtually all HO rolling stock and thus are still useful today.

I’m only a couple years into the hobby, but is there anything that won’t run on a 30" curve?

11

Who is volunteering to do the testing? And what happens if the manufacturer’s test track is significantly better in curve joints and lack of kinks than the average Joe’s trackwork? Should the manufacturer degrade the minimum radius because Fred’s trackwork is too sloppy to reliably run at the manufacturer’s minimum radius?

The club I participate in was trying to establish what the minimum reliable radius for HOn3 operation should be. Some HOn3 brass locomotives were/are known for being “stiff” and requiring more than 18" radius. But we couldn’t test everything in advance. So the club settled on eased 24" minimum radius. Is this enough? So far, it has been. But I can imagine there could come a time that a “stiff” K-36 or K-37 or large Shay would struggle with these curves.

Fred W

12

With more and more rolling stock being made to a higher degree of fidelity, some have even reported issues with interference of underbody detail on even a 50’ box car!

Here are a few anecdotal experiences I had:

Before BLI came out with their HO CZ passenger cars, I had a set of Kumata (Oriental Ltd/High Country Brass) passenger cars. One would work on 32 inch curve but another would derail and needed a bit larger radius.

Before the above, Chuck Macklin who is a train book dealer, brought some brass CN&W heavy weight passenger cars over to my garage layout, which had 30 inch minimum curves. Those cars shorted out on the curve due to the detail on the truck side frames hitting the body of the cars.

So 30-inches seems generous but I discovered that it’s rather tight when it comes to brass some passenger cars. Of course brass steam engines can have issues on 30-inch curves too or even larger.

Most mass produced plastic rolling stock is manufactured or has compromises to allow operation on 24 inch minimum curves. Some six axle loco’s will operate on 18 inch curves.

But the curve thing reminds me back when I was gaming and remember reading about minimum computer specs recommended. A lesson I took from that is that you generally don’t want to run computer games on the absolute minimum computer specs. Likewise with model trains, operating them on the minimum recommended by manufacturer may be ok, or as happens with Wal

13

As it evolves, the hobby moves on several fronts. Brass has effectively been supplanted as the go-to for high details and accuracy, and for realism. Not entirely, not by quite a long shot in fact, but you still pay about three to five times the price for brass (new). A toy J1 2-10-4 in times of old, IF it were available commercially and not scratch-built in John’s day, would decidedly have been seen as a rougher approximation/toy, and not so much as a fine replica in scale the way the Paragon version is. That model needs 22" minimum, according to BLI, although our hosts and I found through trials that it could creep through 21" curves.

John’s somewhat arbitrary range for the smallest type of curvature still encompasses 22". I don’t know that it should change, or that his category range should change. If in 2026 nothing runs on curves less than 24", his lower range will be almost meaningless in the context of the day’s modeling, so I agree with Fred just above.

John’s contribution was to craft a framework for track plan design. It was a tool set, with rationale provided in his running commentary. Maybe the word ‘dean’ is too strong, or hard to appreciate in modern parlance. Nowadays, the buzzword is ‘influencer’. Whatever the term one uses, he encouraged a lot of nice layouts by making the thinking behind their creation lucid, methodical, and efficient.

14

Armstrong innovated or popularized so many things beyond curve radius classifications that we use today – including things like narrow spiral footprints and mushroom benchwork. One or two people not understanding the use of the term “Dean” is not reason enough to scramble for a different word. IMHO, YMMV, etc.

15

For whatever reason, dean of layout design has been coined for years regarding John Armstrong. As Cuyuma pointed out, we don’t need to find a new title because it seems to strong to the newer enlightened among us. Perhaps history is judging him more harshly as time goes by, and that is often the case as we watch the world around us evolve.

I used to be an avid reader of MR magazine in the 70’s, 80’s and 90’s. and by the mid-1980’s I did notice a trend of moderate sized layouts and that was that the standard minimum radius I kept seeing over and over and over was 30 inches. It was a pattern that repeated and stood out. But that “trend” that I was seeing seemed to translate in my mind, after enough time and articles, to “conventional” in John Armstrongs nomenclature.

con·ven·tion·al

/kənˈven(t)SH(ə)n(ə)l/

16

Statistically, quite a few of the mid-sized layouts I design for others use less than 30" radius in HO (or the equivalent in other scales).

Just for a quick check to see if my clients’ requirements were unusual, I looked at the HO layouts that come up in the MR Track Plan Database for sizes 100 sq. ft. to 300 sq. ft. (I think that most would consider that “Mid-Sized” or even large)

I looked at the first 25 that turned up (eliminating Narrow Gauge and purely logging/quarry layouts).

For those 25 layouts, Average (Mean) minimum radius is 24.5". Median minimum radius is 24". Suprisingly, a handful at 18"!

Maybe “Conventional” is, well, still kinda conventional.

Edit: That said, I often recommend 28" with easments in HO for decent performance with most all equipment in a mid-sized space. Broader curves are always preferred when they can work for the concept in the space, of course.

Byron

17

I have worn out two copies of “Track planning for realistic operation”…

And I think JA brought science and planning to layout design in ways never considered as a whole process before him.

On this topic of curves…and yes I have a nice size space, for my last two layouts and my next one…many decades ago, based on my experiances at the Severna Park Model Railroad Club, I decided any attempt to model a Class I railroad in HO scale required 36" radius curves.

It should be noted here that the Severna Park layout is not all that large, it is only 12’ x 36’, yet it has 36" minimum mainline curves.

Is it the best track plan in the world? No. Does it work for the goals and space of that group? Yes.

Pictures and a track plan can be found in the June 1973 issue of MR for those having archives access or your own library.

The exact nature of compromise in this hobby is left up to each modeler, as each has a different set of goals, different space, different skills, different resources, etc.

Can really nice layouts be built with 30" radius? or even 24" radius? Sure.

Many consider my standards extreem, maybe they are. I model 1953, my longest freight car is 75’, and there are not very many of them. More on passenger cars later.

My mainline minimum, hidden or visable, is 36" radius. If passenger cars go there, it has 36" or larger curves.

Since the mainline is double track, That means that many curves are 38" or larger.

At one spot there are 5 tracks going around a curve as a yard lead joins the mainline and a double track passenger station lead joins the mainline. So the outer most track is in the neighborhood of 46" radius.

Being an around the room layout with two peninulas, there are only three places where major curves are viewed from the outside.

Passenger cars - I have some 80’ passenger cars, they however do not make up mos

18

You have to consider WHEN Armstrong wrote his material. Both the hobby world and the prototype.

Brass? The floodgates didn’t open on that until the early 60’s and PFM.

The prototype was still running mostly shorter cars and shorter locos. Yeah there were giants just retiring, like the Big Boy - but thoose were RARE locos, regardless of how many models of them are made. The everyday loco was the 2-8-0 and 2-8-2. Or 4 axle diesels.

Modern locos are much larger. Modern rolling stock is bigger - 89 foot stack cars and articulated auto racks, and 60 foot box cars. So you need to up the radius for the classes of curves.

My 4-8-4’s don’t look silly on 30" radius curves. The tender can couple to the loco in the closest position. It can handle these curves at warp speed. Most anything I run cn handle #6 turnouts at ridiculous speeds as well, but I am using #8’s for mainline crossovers. I am actively considering #5s in the yard, because no big locos or passenger cars will run there, just mostly 40 foot and shorter cars and small switchers. A random 50 foot car here or there. Engine terminal will have #6s, because the biggest locos I have can handle those no problem. The only reason I even drew it out as #6 was because that’s the size Peco was going to introduce first in code 70. But it’s been over 2 years and all we got so far is the flex track, I’m not holding my breath. If I have to mix rail size then I might as well use the turnout that gets the most out of the space.

In the context of the early to mid 50’s, Armstrong’s classification of curves as sharp, conventional, and broad fits botht he model world and the prototypical world. For MOST modelers. To model lines using some of the true giants of the rails, you may indeed have to adjust those upwards.

Space is usually the limiting factor. Few people have hangers. Everything needs to be compressed, including the track. I

19

My thoughts is if one must use 18" he or she should run small engines like 4 axle diesels or nothing larger then a 2-8-0 and 40 and 50’ cars.

Today’s highly detailed cars including 72- 89’ freight cars and full lenght passenger cars one should try for 30-36" curves even though long cars might still look silly on 30" curves.

As far as that goes and this hit home while designing a new Slate Creek switching layout 63-72’ freight cars requires at least a #6 switch or a long Peco to look right. I suppose a size 8 switch might be better?

Good LDEs and the type of equipment that will be used on the layout should always be considered while designing a layout…

20

[quote user=“rrinker”]

You have to consider WHEN Armstrong wrote his material. Both the hobby world and the prototype.

Brass? The floodgates didn’t open on that until the early 60’s and PFM.

The prototype was still running mostly shorter cars and shorter locos. Yeah there were giants just retiring, like the Big Boy - but thoose were RARE locos, regardless of how many models of them are made. The everyday loco was the 2-8-0 and 2-8-2. Or 4 axle diesels.

Modern locos are much larger. Modern rolling stock is bigger - 89 foot stack cars and articulated auto racks, and 60 foot box cars. So you need to up the radius for the classes of curves.

My 4-8-4’s don’t look silly on 30" radius curves. The tender can couple to the loco in the closest position. It can handle these curves at warp speed. Most anything I run cn handle #6 turnouts at ridiculous speeds as well, but I am using #8’s for mainline crossovers. I am actively considering #5s in the yard, because no big locos or passenger cars will run there, just mostly 40 foot and shorter cars and small switchers. A random 50 foot car here or there. Engine terminal will have #6s, because the biggest locos I have can handle those no problem. The only reason I even drew it out as #6 was because that’s the size Peco was going to introduce first in code 70. But it’s been over 2 years and all we got so far is the flex track, I’m not holding my breath. If I have to mix rail size then I might as well use the turnout that gets the most out of the space.

In the context of the early to mid 50’s, Armstrong’s classification of curves as sharp, conventional, and broad fits botht he model world and the prototypical world. For MOST modelers. To model lines using some of the true giants of the rails, you may indeed have to adjust those upwards.

Space is usually the limiting factor. Few people have hangers. Everything needs to