Did GN O-8 2-8-2s really have 71" drivers?

716 views
16 replies
1 rating 2 rating 3 rating 4 rating 5 rating
  • Member since
    August 2013
  • 42 posts
Did GN O-8 2-8-2s really have 71" drivers?
Posted by Shrike Arghast on Friday, July 31, 2020 11:17 PM

I've only been able to find stats for this class on one website. I know the O-8s did some dual purpose hauling, but did they really have 71" driving wheels? That sounds too big.

  • Member since
    February 2011
  • 27 posts
Posted by 03 1008 on Saturday, August 1, 2020 2:07 AM

69" will do. # 3380 was an O-7 rebuilt to O-8 specifications. (Drury 2015, p. 181).

  • Member since
    November 2008
  • 1,476 posts
Posted by Leo_Ames on Saturday, August 1, 2020 2:19 AM

The article "The World's Greatest Mikado" from the January 1969 issue of Trains, states 69 inches for the O-7 and O-8 classes.

Did you get 71 inches from Steamlocomotive.com? There's a lot of inaccurate details at that resource, as to be expected for a site that attempts to cover as much as it does. 

I don't know which is accurate, but the article I just read in Trains appears to be written by a source that's very familiar with the subject matter, so I'm inclined to believe they were 69".

  • Member since
    August 2013
  • 42 posts
Posted by Shrike Arghast on Saturday, August 1, 2020 4:54 AM

Leo_Ames

The article "The World's Greatest Mikado" from the January 1969 issue of Trains, states 69 inches for the O-7 and O-8 classes.

Did you get 71 inches from Steamlocomotive.com? There's a lot of inaccurate details at that resource, as to be expected for a site that attempts to cover as much as it does. 

I don't know which is accurate, but the article I just read in Trains appears to be written by a source that's very familiar with the subject matter, so I'm inclined to believe they were 69".

 

69" sounds way more right to me. Thanks much. And yeah, that's where I got it. I guess I won't rely on that website so much in the future...

  • Member since
    November 2008
  • 1,476 posts
Posted by Leo_Ames on Saturday, August 1, 2020 1:48 PM

Don't get me wrong though, it's a great site. Just when it's trying to cover as broad a topic as North American steam locomotives, unfortunately errors are guaranteed to appear here and there.

Some of it is well off the mark though, so it's best to not take it as a definitive source but rather as a jumping off point to deeper research. Here's one such offender that has long annoyed me.

https://www.steamlocomotive.com/survivors/recent-losses.php

Union Pacific alone entered 1960 with 86 steam locomotives still officially on their roster, 50+ of which were scrapped in the end. 

Toss in stuff like the NKP, N&W, Illinois Central, C&O, and so on and I imagine many hundreds of steamers have met the torch in the US since the 1950's ended. I imagine we're looking at 500+. Toss in Canada and Mexico and we're probably looking at thousands.

  • Member since
    September 2003
  • 11,910 posts
Posted by Overmod on Saturday, August 1, 2020 8:12 PM

I actually posted the 71" in a post over on Classic Trains a couple of months ago, but yes, I got it in a fact-check on steamlocomotive.com and yes, I remembered it from many years ago as 69".  I am not much of a GN fan so don't know if there might have been cheating up to a slightly larger pre-wear tire diameter as was done on some of the PRR J-1s (to 70") which is similarly confusing to document.

  • Member since
    July 2004
  • 2,381 posts
Posted by Paul Milenkovic on Saturday, August 1, 2020 9:28 PM

What is the real story on wheel diameter and intended locomotive service?

Generally, you need "high drivers" if you want speed.  One reason for that is with piston valves, there is an RPM level where you get enough pressure loss in the admission and exhaust steam circuits that you lose power.  Another reason is more room for the proper kind of balance weight distribution, cross balancing being an example?

So then, how did a Norfolk and Western J class run on test up to 100 MPH with 70 inch wheels?

Low drivers are said to be better for low-speed freight operation.  So then, how did Golsdorf in Austria get by with a high-drivered mountain engine?

The formula for Super Power in "Big Boy -- Big Mistake" is a high-drivered locomotive, but the pioneer A1 Berkshire, was what, 63"?  Anyway, if you put high drivers on a locomotive, it can pull OK up to the traction limit of whatever weight it has on drivers, but you need to put a big boiler on it so it doesn't run out of steam at speed, but then the locomotive is so heavy that you need to support substantial weight on the engine and trailing trucks, and then you have dead weight to carry around when climbing a grade?

But the ultimate speed of an engine is determined by the RPM where it self-limits power by the pressure drops in the steam circuits.  Suppose you have a high-drivered engine and just put a placard in the cab, "You will run out of steam if you operate in full gear past 30 MPH"?  Is there some reason you need to put low drivers on such an engine so the 30 MPH limit is automatically enforced by pressure drop in the steam circuit, or could you just train the crews that the locomotive has a speed limit past which it will run faster than the boiler can make steam?  

If GM "killed the electric car", what am I doing standing next to an EV-1, a half a block from the WSOR tracks?
  • Member since
    January 2009
  • From: Maryland
  • 9,305 posts
Posted by ATLANTIC CENTRAL on Saturday, August 1, 2020 10:24 PM

The difference between 69" drivers and 70" drivers is pretty small. It accounts for aproximately an 4" difference in circumference, about a 1.5% difference at that diameter.

So in general terms, we could put the driver size of the O-8 in the same range as the NKP/C&O/PM 2-8-4 and the N&W J. And, all three classes have a similar weight on the drivers, all between 280,000# and 292,000#.

As for top speed, obviously the J has the advantage with the 4 wheel lead truck, but in terms of getting TE to the ground, the O-8 was just as effective as the J, and more powerful than the 2-8-4's.

The O-8 also had just as much grate area as the 2-8-4's

Suggesting that 2-8-4's are just bigger Mikado's and lighter track was the biggest justification for the 4 wheel trailing truck. And that drivers bigger than 63"/64" worked fine for pulling freight under the best most developed boilers.

I'm just a freelance model train guy, but I built five of these from Berks for my freelanced model railroad - picture taked before the paint shop:

 

And then there where these, baby brothers of the LIMA Berkshire, the DT&I 800 class Mikes - on 63" drivers, and smaller grates, but still pretty impressive:

http://steam.wesbarris.com/whyte/2-8-2/USA/photos/dti800-dickenson.jpg

It was this DT&I 800 class that inspired me to learn more about 2-8-4's vs the biggest 2-8-2's, coming to the conclusion that my freelanced LIMA Mikes with 69" drivers could have been built, even if only for roads with top notch trackwork.

Sheldon

 

    

  • Member since
    July 2004
  • 2,381 posts
Posted by Paul Milenkovic on Sunday, August 2, 2020 8:58 AM

Atlantic Central:

Way, cool!

If GM "killed the electric car", what am I doing standing next to an EV-1, a half a block from the WSOR tracks?
  • Member since
    September 2003
  • 11,910 posts
Posted by Overmod on Sunday, August 2, 2020 9:16 AM

ATLANTIC CENTRAL
As for top speed, obviously the J has the advantage with the 4 wheel lead truck, but in terms of getting TE to the ground, the O-8 was just as effective as the J, and more powerful than the 2-8-4's. The O-8 also had just as much grate area as the 2-8-4's

This is the stuff that happens when freelance model railroaders start doing steam design -- a bit like crayonistas revising London transit routes.

There's a whole lot more to achieving high speed than putting a four-wheel pin-guided lead truck on a locomotive, and considerably more in the case of the N&W J class, which uses special balancing methods derived by Voyce Glaze (and which are applicable, with some care, to engines with good modern 2-wheel lead and trailing trucks).

A major reason for the two-wheel trailing truck on a Berk is weight distribution.  Another is proper arrangement of the ashpan. These are relatively unimportant on a model railroad where the firebox structure is a molded plastic shell and the frame a 'casting of convenience' but in real life the combination of a good cast bed, a full set of circulators, and metallurgically-suitable material for high pressure on a locomotive with large grate area will be difficult to accommodate on a single axle that also guides and steers well.

(This is not to say that good modern high-power locomotives can't be developed on two-wheel trailing trucks -- they certainly could, and Richard Leonard among others has speculated that they are 'all that's necessary' for practical reciprocating power that is reasonably thermodynamically improved (e.g. with Snyder or other primary-air preheating and Cunningham circulators to chamber and legs) but this goes hand in hand with reducing required GA...)

Some of the early Berks in fact used an amusing truck that was designed to act either like a long Delta (with booster) or the American Arch articulated design with the weird inability to reverse corrected.  The result could not get the rear axle to 'track' correctly with the correct action to center and steer the rear of the locomotive chassis (remember the booster required it to be fixed with respect to the frame) if any other axle(s) were present in conventional pedestals -- so for a couple of years the forward axle in the truck floated laterally as 'weight-bearing' only, supported by transverse hardened-steel rollers with only its own flanges for horizontal location and centering.  (I don't know if this action was ever modeled...)

 There's more to driver diameter than just 'larger is faster' and 'diameter speed', too.  It pays to study Golsdorf's passenger 2-6-4 as in a number of respects the detail-design choices that seem so weird turn out to be instructive.

The choice of 68"-69" drivers in the '30s over 63" is largely an issue of sufficient room for in-plane balance installation in an otherwise strong-enough wheel center.  Note that even as early as the conversions of the T&P 2-10-4s from dogs into reasonable road locomotives, adoption of lower rodwork inertia removed much of the 'problem' with that diameter; there was little augment issue with the P&LE A-2-As even at their 50mph road speed, and of course the ACE3000 used 58" nominal with Withuhn conjugated duplexing.

(I'm sure the issue with cast driver centers is going to come up at some point; remember that by far the fastest engine on 70" drivers did not use them even on the mains... while the ACL R-1s were a balance disaster even with them.  And contrary to what I would have believed if not shown, the late Web-Spoke design became a functional replacement for cast centers on a number of roads, including NYC.

The difference between 69" and 70" is indeed small, and of course the diameter may decrease sustantially as the tires wear to design limit (in the case of the ACE3000, I recall this being 6" (or 3" radius) as they thought government single-wear wheel requirements would apply to them).  The problem is that these wheels have flanges, and the spacing between axles in most reciprocating locomotives, in trying to reduce the rigid wheelbase without too many expen$ive lateral-motion cannon boxes, will have those flanges nearly touching behind the brake hangers.  The Niagara was a special case, eith the frame spacing (and expensive Timken rod length) intentionally long enough for the 8" upsize when desired -- and at the time, 79" over 75" was enough of a no-brained that the change was promptly done.  On the other hand I never saw even the shred of a proposal to put 79" drivers on a Mohawk, even though it might make 'design' sense a la Golsdorf to try, because 72" was already perfectly ample with correct balancing for any sustainable speed rationally expected of a 4-8-2 at its rated load.

  • Member since
    January 2009
  • From: Maryland
  • 9,305 posts
Posted by ATLANTIC CENTRAL on Sunday, August 2, 2020 9:58 AM

Overmod

 

 
ATLANTIC CENTRAL
As for top speed, obviously the J has the advantage with the 4 wheel lead truck, but in terms of getting TE to the ground, the O-8 was just as effective as the J, and more powerful than the 2-8-4's. The O-8 also had just as much grate area as the 2-8-4's

 

This is the stuff that happens when freelance model railroaders start doing steam design -- a bit like crayonistas revising London transit routes.

 

There's a whole lot more to achieving high speed than putting a four-wheel pin-guided lead truck on a locomotive, and considerably more in the case of the N&W J class, which uses special balancing methods derived by Voyce Glaze (and which are applicable, with some care, to engines with good modern 2-wheel lead and trailing trucks).

A major reason for the two-wheel trailing truck on a Berk is weight distribution.  Another is proper arrangement of the ashpan. These are relatively unimportant on a model railroad where the firebox structure is a molded plastic shell and the frame a 'casting of convenience' but in real life the combination of a good cast bed, a full set of circulators, and metallurgically-suitable material for high pressure on a locomotive with large grate area will be difficult to accommodate on a single axle that also guides and steers well.

(This is not to say that good modern high-power locomotives can't be developed on two-wheel trailing trucks -- they certainly could, and Richard Leonard among others has speculated that they are 'all that's necessary' for practical reciprocating power that is reasonably thermodynamically improved (e.g. with Snyder or other primary-air preheating and Cunningham circulators to chamber and legs) but this goes hand in hand with reducing required GA...)

Some of the early Berks in fact used an amusing truck that was designed to act either like a long Delta (with booster) or the American Arch articulated design with the weird inability to reverse corrected.  The result could not get the rear axle to 'track' correctly with the correct action to center and steer the rear of the locomotive chassis (remember the booster required it to be fixed with respect to the frame) if any other axle(s) were present in conventional pedestals -- so for a couple of years the forward axle in the truck floated laterally as 'weight-bearing' only, supported by transverse hardened-steel rollers with only its own flanges for horizontal location and centering.  (I don't know if this action was ever modeled...)

 There's more to driver diameter than just 'larger is faster' and 'diameter speed', too.  It pays to study Golsdorf's passenger 2-6-4 as in a number of respects the detail-design choices that seem so weird turn out to be instructive.

The choice of 68"-69" drivers in the '30s over 63" is largely an issue of sufficient room for in-plane balance installation in an otherwise strong-enough wheel center.  Note that even as early as the conversions of the T&P 2-10-4s from dogs into reasonable road locomotives, adoption of lower rodwork inertia removed much of the 'problem' with that diameter; there was little augment issue with the P&LE A-2-As even at their 50mph road speed, and of course the ACE3000 used 58" nominal with Withuhn conjugated duplexing.

(I'm sure the issue with cast driver centers is going to come up at some point; remember that by far the fastest engine on 70" drivers did not use them even on the mains... while the ACL R-1s were a balance disaster even with them.  And contrary to what I would have believed if not shown, the late Web-Spoke design became a functional replacement for cast centers on a number of roads, including NYC.

 

I understand all that, I was just trying to offer some thoughts without writing a book.

But the fact that the O-8 worked, deminstrates that more than one approach works, depending on the axle loads you can tollerate, the speeds you need, etc.

Maybe the main point I was trying to make got lost in there, it was about driver diameter. A Mikado with 69" or 70" drivers is really not that different from a Berk or Northern with that size drivers - the O-8 deminstrated this.

I seem to remember some story about the B&O considering or experimenting with the idea of larger drivers on a Mike as well?

And the NKP Berks were much lighter than the C&O versions, maybe they could have been suitably balanced on a single axle Delta, maybe not.

But, much as I know about real trains, I play with model trains for fun. So when I came up with this project years ago, I did at least some exploring of its plausiblity.

Based on the DT&I 800's and the O-8, I think LIMA could have built a 69" driver Mike similar to mine for a road willing to tollerate axle loads like the C&O had on a number of locos.

Possible/plausible, that's all.

Sheldon

    

  • Member since
    July 2004
  • 2,381 posts
Posted by Paul Milenkovic on Sunday, August 2, 2020 11:15 AM

ATLANTIC CENTRAL

I understand all that, I was just trying to offer some thoughts without writing a book.

Someone beat you to it writing such a book.

David Wardale writes in "The Red Devil" of wanting to replace the 4-wheel trailing truck on the 25-Class with a 2-wheel trailing truck and tinkering with the equalization to put a bit more weight on the drivers.  I have to search for that section, but I think he states that the added driver axle loading, beneficial to tractive effort, would have restricted the route availability of the modified locomotive that became the Red Devil 26-Class locomotive.

If GM "killed the electric car", what am I doing standing next to an EV-1, a half a block from the WSOR tracks?
  • Member since
    January 2009
  • From: Maryland
  • 9,305 posts
Posted by ATLANTIC CENTRAL on Sunday, August 2, 2020 11:59 AM

Paul Milenkovic

 

 
ATLANTIC CENTRAL

I understand all that, I was just trying to offer some thoughts without writing a book.

 

 

Someone beat you to it writing such a book.

David Wardale writes in "The Red Devil" of wanting to replace the 4-wheel trailing truck on the 25-Class with a 2-wheel trailing truck and tinkering with the equalization to put a bit more weight on the drivers.  I have to search for that section, but I think he states that the added driver axle loading, beneficial to tractive effort, would have restricted the route availability of the modified locomotive that became the Red Devil 26-Class locomotive.

 

Well, I looked that up, it's interesting.

It is way outside my knowledge base because I have never followed trains outside Norh America to any degree.

It is an interesting idea, converting a Northern into a Mountain.

The B&O never had 2-8-4's or 4-8-4's, they were always happy with their fleet of 2-8-2's and 4-8-2's.

Sheldon

    

  • Member since
    September 2003
  • 11,910 posts
Posted by Overmod on Sunday, August 2, 2020 12:33 PM

ATLANTIC CENTRAL
But the fact that the O-8 worked, demonstrates that more than one approach works, depending on the axle loads you can tolerate, the speeds you need, etc.

This is very true, and one thing it points to is demonstrable fairly dramatically in English practice.  A steam locomotive's indicated power is a function of its cylinders, rodwork geometry, and wheel diameter... whether those are on a 4-6-0 or 4-8-2 is circumstantial ... and the available TE then just a function of axle load.  This avoids a major actual determinant of performance, economical long-term steam generation at lowest net cost, and it flirts with another, which is practical adhesion and slip recovery as usually bundled into a discussion of FA, but it explains why a relatively small and light 4-6-0 (say, a Castle or King class) can happily do the work of a much larger Pacific, on a comparable train, at comparable timing.  And it's instructive to see the ways in which the feat requires special attention or care...

Maybe the main point I was trying to make got lost in there, it was about driver diameter. A Mikado with 69" or 70" drivers is really not that different from a Berk or Northern with that size drivers - the O-8 demonstrated this.

And from the standpoint of indicated performance this is again quite correct.   Where I want you to start looking at the difference is in some of the early reported Berkshire testing vs. the excellent Lima/Woodard 8000 NYC H10 Mikado that preceded it, which has very similar constructional details.  The advantages of the bigger boiler and chamber far outweigh the added weight, length, and cost to provide them on a locomotive optimized for long heavy service -- but if you are not needing the sustained high speed, then there is only the guiding advantage between a large Pacific and a proper Mikado... and that, we can agree, is relatively easily handled for a Mike just as it would be for a Berk if desired.

For fun, look at Deem's version (in the '70s) of Chapelon's "North American" 2-10-4 of the latter 1940s  This is not at all a large engine, comparable in size essentially to a Mike stretched by one driver pair with the two-axle trailer only for reducing axle load (as on the 242A1) and it is instructive how it makes enough steam for its three fairly commodious cylinders... and how the balance arrangement makes it less difficult for the two-wheel lead truck to do the necessary business without a Krauss-Helmholtz or similar arrangement.

  • Member since
    January 2009
  • From: Maryland
  • 9,305 posts
Posted by ATLANTIC CENTRAL on Sunday, August 2, 2020 1:12 PM

Overmod

 

 
ATLANTIC CENTRAL
But the fact that the O-8 worked, demonstrates that more than one approach works, depending on the axle loads you can tolerate, the speeds you need, etc.

 

This is very true, and one thing it points to is demonstrable fairly dramatically in English practice.  A steam locomotive's indicated power is a function of its cylinders, rodwork geometry, and wheel diameter... whether those are on a 4-6-0 or 4-8-2 is circumstantial ... and the available TE then just a function of axle load.  This avoids a major actual determinant of performance, economical long-term steam generation at lowest net cost, and it flirts with another, which is practical adhesion and slip recovery as usually bundled into a discussion of FA, but it explains why a relatively small and light 4-6-0 (say, a Castle or King class) can happily do the work of a much larger Pacific, on a comparable train, at comparable timing.  And it's instructive to see the ways in which the feat requires special attention or care...

 

 

 
Maybe the main point I was trying to make got lost in there, it was about driver diameter. A Mikado with 69" or 70" drivers is really not that different from a Berk or Northern with that size drivers - the O-8 demonstrated this.

 

And from the standpoint of indicated performance this is again quite correct.   Where I want you to start looking at the difference is in some of the early reported Berkshire testing vs. the excellent Lima/Woodard 8000 NYC H10 Mikado that preceded it, which has very similar constructional details.  The advantages of the bigger boiler and chamber far outweigh the added weight, length, and cost to provide them on a locomotive optimized for long heavy service -- but if you are not needing the sustained high speed, then there is only the guiding advantage between a large Pacific and a proper Mikado... and that, we can agree, is relatively easily handled for a Mike just as it would be for a Berk if desired.

 

For fun, look at Deem's version (in the '70s) of Chapelon's "North American" 2-10-4 of the latter 1940s  This is not at all a large engine, comparable in size essentially to a Mike stretched by one driver pair with the two-axle trailer only for reducing axle load (as on the 242A1) and it is instructive how it makes enough steam for its three fairly commodious cylinders... and how the balance arrangement makes it less difficult for the two-wheel lead truck to do the necessary business without a Krauss-Helmholtz or similar arrangement.

 

When I did my research for my ATLANTIC CENTRAL LIMA Mikes, I did find and get partly into that early Berk testing and the H10.

In fact I considered mentioning the H10 in my earlier post. But, after getting the general feel for where the Berk came from, I simply shifted back to deciding if the Mike would have been plausible.

Since I already know everything about old houses Geeked, and I know a lot about trains, but not nearly what you know, I simply wanted my LIMA Mike to be possible.

There are after all, there only so many hours in the day, and I have a layout to build.

The ATLANTIC CENTRAL has a carefully thought out roster designed to appear plausible for its location and era.

Our primary rigid wheel base power is 2-8-0's, 2-8-2's, 4-6-2's and 4-8-2's.

There are few 2-10-2's, and 4-8-4's.

Larger power includes 2-6-6-2's, 2-8-8-0's, and 2-6-6-4's.

And then there are these new fangled Diesels....... 

Sheldon

    

  • Member since
    September 2003
  • 11,910 posts
Posted by Overmod on Sunday, August 2, 2020 1:47 PM

Here's the thing: a Lima Mike would be immensely possible, and it would have been possible to 'design' a good one, just not with a Berkshire firebox.  (And not with an implied frame design assuming a four-wheel trailer, but that's less important in making a model more 'verisimilitudinous'...)  That box is just a tad long and wide to be 'balanced' over a single axle which then needs leverage to the rear corners of the chassis -- badly, on a large-cylindered Mike -- to reduce 'waddle' and also reduce the overbalance needed.

One thing you could do on the Atlantic Central, although a little radical, is split the tender, and move the lead truck up as far forward as you can, as on some of the late JNR and Reichsbahn power.  Then modify the radial buffer so that the tender actively helps guide the rear of the locomotive chassis, a bit as if there were a Jacobs bogie articulated there.  In ll probability this would give you a water-bottom tender of NYC proportions, with very short cistern, 6-wheel Commonwealth truck front and 2-axle behind, like the arrangement on the smaller MILW A tenders, and a separate 4-axle A-tank (multiples if you want coast-to-coast style range).  An advantage here is that a comparatively very short table will turn the engine, without recourse to exaggerated overhang like on the NYC PT pedestal tanks) and similarly there is less issue with stall length for high-horsepower steam.  This was and is the solution for those otherwise-spectacular RENFE 4-8-4s that could never do much, just as the PRR V1 would never have done much, because the bunker and water capacity that would 'fit' only gives relatively short range before you have to stop.

Some of the detail changes are slight.  Snyder preheaters are like the air-brake radiator 'bundle' of pipe, but arranged along the gap between mud ring and ashpan; they run on the same exhaust steam that goes to the FWH with a little turret tap up and a little trap down so they don't freeze.  C&O testing of these reported produced over 10% efficiency gain (in the service the tests were run for) which is substantial.

Cunninghams tap circulation off a low point in the convection section, usually forward 'enough' from the throat connections to syphons or circulators.  They pass this through a jet pump with no power moving parts, and then into ram's-horn manifolds that look similar to old Chevy 283 exhaust manifolds along the water legs.  Here the pumped flow is directed into upward-facing fan nozzles inside the legs that greatly enhance upward flow in the sides of the radiant firebox.  (There should be arrangements at the top to channel the upward flow, but Cunningham did not detail them).  One thing this allows is thinner waterleg space for the same heat uptake, hence less mass than that added by the manifolds, and much of the advantage of an explicit forced-circulation boiler like a Lamont.  

  • Member since
    January 2009
  • From: Maryland
  • 9,305 posts
Posted by ATLANTIC CENTRAL on Sunday, August 2, 2020 2:14 PM

Agreed, the firebox is larger than it would have been. Only the very informed will notice that.

I did move the Delta trailing truck back to the correct bearing point at the end of the firebox, and added the beam detail.

It is after all just a little 1/87 scale toy.......

And the ACR Mikes exist in three sub classes, with variations in sand box/steam dome configuration, and rectanglular and Vanderbuilt tenders.

No pictures of the others right now......

Sheldon

 

 

    

Join our Community!

Our community is FREE to join. To participate you must either login or register for an account.

Search the Community

Newsletter Sign-Up

By signing up you may also receive occasional reader surveys and special offers from Trains magazine.Please view our privacy policy