• Baldwin vs. Boxpok driver centers

  • Discussion of steam locomotives from all manufacturers and railroads
Discussion of steam locomotives from all manufacturers and railroads

Moderators: slide rules, Typewriters

  by Pneudyne
Returning to the original question, and looking more closely at a selection the pertinent Baldwin-built locomotives:


1938 4-8-4 Baldwin disc
1938 2-10-4 Baldwin disc
1941 4-8-4 Baldwin disc.
1943 4-8-4 Boxpok
1944 2-10-4 Boxpok


1935 4-8-2 Conventional spoked
1937 4-8-2 Baldwin disc main only
1940 4-8-2 Baldwin disc main only
1941 4-8-2 Baldwin disc


1944 2-8-8-4 Boxpok


1937 4-6-6-4 Baldwin disc
1938 4-8-4 Baldwin disc
1942 4-6-6-4 Baldwin disc


1944 4-8-2 Boxpok (clone of 1941 B&M design with Baldwin disc)


1942 2-8-4 Baldwin disc
1944 2-8-4 Boxpok
1949 2-8-4 Boxpok


1943 4-8-4 Boxpok (clone of the 1938 D&RGW design with Baldwin disc)


1946 4-8-4 Boxpok


1937 4-8-4 Baldwin disc
1937 4-8-4 Baldwin disc
1942 4-8-4 Baldwin disc
1945 4-8-4 Baldwin disc


1942 4-8-4 Baldwin disc
1943 4-8-4 Boxpok


1940 4-6-6-4 Baldwin disc
1941 4-6-6-4 Baldwin disc
1947 4-8-4 Boxpok

This I think shows that the “critical” years were 1943 and 1944.

Up to and including 1942, it seems that those roads who wanted Baldwin disc drivers on their Baldwin-built locomotives got what they wanted. That included two new designs for 1942, the L&N 2-8-4 (OK, it was a Baldwin version of the VS rather than a completely new design) and the Frisco 4-8-4.

Then in 1943 and 1944, roads whom one would have expected to buy Baldwin locomotives with Baldwin disc drivers instead had them with Boxpok drivers. These included the AT&SF, L&N and STL-SF. This strongly suggests that there was external influence at work, with the WPB as the prime suspect.

The MoPac and L&HR cases are less certain. Whilst they had clones of designs originally fitted with Baldwin disc drivers, it is also possible that those roads would have selected the Boxpok type had they had a free choice.

The B&O case is weak evidence. Its 1944 2-8-8-4 was a new design, so could have gone either way in respect of drivers had there been a free choice. The DM&IR 2-8-8-4 had Boxpok drivers, as did the later SP 4-8-8-2, although the SP had had a prior commitment to the Boxpok type.

Evidently Baldwin disc drivers were available again by 1945, as evidenced by the RF&P case. But that did not seem to have much influence post-WWII. WM chose Boxpok for its 4-8-4 despite previously being in the Baldwin disc camp, and it looks as if L&N, having had Boxpok on its 1944 batch probably as a forced choice, stayed with that for its 1949 batch of 2-8-4s. NdeM’s choice of Boxpok might have been to achieve standardization, as the other half of its 4-8-4 fleet came from Alco, who fitted the Boxpok type. Also, it was a clone of the NC&StL design, the later builds of which had Boxpok drivers. Still, that the Santa Fe was able to retrofit its 2900 class with Baldwin disc drivers (on the 4th driver set only) indicates that Baldwin was still producing the type post-WWII.

It might be worth looking at the huge SNCF 141R fleet in this regard, built I think by Alco, Baldwin, Lima and MLW in the years following WWII. Some of the later examples had disc drivers, Boxpoks from the pictures I have seen. But perhaps some of those built by Baldwin had Baldwin disc drivers. Editions Cabri I think did the definitive book on the 141R, but I have not seen it. Some, maybe all of the WP class 4-6-2s exported to India had disc drivers. From the pictures I can find, they were of the Boxpok type, but it seems possible that some of the Baldwin builds had Baldwin discs.

  by Pneudyne
A couple more patents:

US2106566, assigned to GSC, appears to cover the “big hole” version of the Boxpok, and built upon US1862157, assigned to ASF. So perhaps that was the ASF input. Given that there were several disc driver patents assigned to GSC, perhaps the individual contributors came from some of the other participants. I am not sure if the locations for the individuals, such as Drexel, PA, Wallingford, PA and St. Louis, MO give any clues.

Perhaps the Union Steel web-spoke driver centre should also be mentioned here. Some of the NYC L-3a/b fleet were fitted with this type from new, and I think that NYC also used it for retrofits, including on the 4-6-4s, which thus sometimes ended up with mixed sets of drivers. So the web-spoke driver was seen as being “competitive” with the disc types.

  by Allen Hazen
Thank you again for your systematic search for disc-driver users!
-- the L&N's 1949 2-8-4 order came from Lima(*), not Baldwin, so would have been very unlikely to use Baldwin Disc Drivers (I assume there would have been some sort of licensing fee).
--I tried to find out a bit more about Harry M Plager. He lived from 1866 to 1951. He had lots of patents: mostly, I think, for locomotive and car parts. He was made an "Honorary Member" of ASME for his contributions. And one possibly informative bit: in 1928 (when he patented something) he worked for Commonwealth Steel Castings. So he probably came to GSC with the 1929 merger, and may well have stayed at the Illinois plant, not at the GSC Eddystone location next door to Baldwin.


(*) Somewhere-- I suspect in a "Trains" article on the L&N M-1 (2-8-4)-- the story is told that after the BLH merger an ex-Lima engineer was quizzed by ex-Baldwin types who wanted to know how Lima had managed to "steal" what Baldwin had obviously hoped would be a repeat order. Reply: "Lima INVENTED the Berkshire-- it wouldn't have been right for us not to build at least some of L&N's!"
On which not: MERRY CHRISTMAS!
  by Pneudyne
Thanks, Allen. I had forgotten that Lima built the last L&N 2-8-4 batch - another senior moment. That was, I think, Lima's last domestic steam locomotive order.

Merry Christmas!
  by Pneudyne
Correction – the last NKP 2-8-4 batch was Lima’s final domestic steam production.

  by Pneudyne
The subject of disc drivers per se seems to have had minimal coverage in the literature, at least insofar as I can find. Surprisingly, Bruce (1) did not mention them in his landmark book. Drury (2), in his excellent little book, did not have a separate section for them, even though he covered many aspects of steam locomotive design and development with succinct commentary (and which did not always respect established illusions). Swengel (3) provided no detail, but did note that the 1934 DL&W 4-8-4 was one of the first types to use disc drivers. On the off-chance that there might be something there, I found my copy of O’Connell (4), and whilst he called out the fitment of disc drivers to NYC’s 1937-38 4-6-4s, there was otherwise no detail.

Reed provided a little more information in his Loco Profile #2, “New York Central Hudsons”, and I’ll refer to that in a following post.

From looking at the patent documents, one may see that the Baldwin disc driver was somewhat different, being of a double Y-ended shape in cress-section, double-disc at the hub and rims but single-disc in between. The lips around the holes then acted as stiffening flanges over the single-disc section, and further stiffening was provided by ribs between the holes. Possibly it was a bit lighter than the Boxpok design; if so, that might explain why the Santa Fe retrofitted the Baldwin type to the 4th driving axle only of its 2900-series 4-8-4s. That axle might have had the smallest section rods and smallest balance weights to the point where it was worthwhile to fit a lighter driver centre.

Re the Boxpok type, if one finds a picture of a Boxpok-fitted locomotive, in the rods down pose, and taken from an angle and not face-on, and then stares at it for long enough, one sees spokes rather than holes, and the said spokes are of box form. This seems to be particularly so for the common variant with seven holes around the top semicircle (when rods-down).

Some good information on the Scullin disc driver is provided here: http://www.martynbane.co.uk/modernsteam ... cullin.htm" onclick="window.open(this.href);return false;.


(1) A.W. Bruce; The Steam Locomotive in America: Its Development in the Twentieth Century; Bonanza; 1952 (originally published by W.W. Norton).
(2) G.H. Drury; Guide to North American Steam Locomotives: History and Development of Steam Power Since 1900; Kalmbach; 1993; ISBN 0-89024-206-2.
(3) F.M. Swengel; The American Steam Locomotive: Volume 1 The Evolution of the Steam Locomotive; Midwest rail Publications; 1967; LCC 67-29846.
(4) J. O’Connell; Popular Mechanics Railroad Album; Popular Mechanics; 1954; LCC 53-9420
  by rlsteam
Strictly a non-expert here, but still fascinated by the whole disc driver question. Regarding smaller power, I have seen a Green Bay & Western 2-8-0 with a disc main driver at the Mid-Continent Railway Museum, and I have seen online images of a Chicago Great Western 4-6-0 with all disc drivers (both Boxpok, I believe).

Now my main question: where does the "web-spoke" driver fit in this picture? Is it a variation of the BDD? I photographed a few examples myself. One of the GTW K-4-b Pacifics, 5633, had web-spoke drivers (http://www.railarchive.net/rlsteam/gtw.htm, scroll down) but not its two classmates. Also, Illinois Central's 4-8-2 2603 appears to have web-spoke drivers (http://www.railarchive.net/rlsteam/ic.htm, scroll down) as do several other of the lower-numbered engines in this group according to images I can find, although other higher-numbered examples appear to have Boxpok centers. (The 2600s were built new by IC in 1942-43.) I have seen images of other instances of the web-spoke driver including, I believe, a NYC J-1 Hudson and a PRR K-4. Since the second group of L&N 2-8-4s seem to have the same kind of driver (per builder's photo of the 1967), is this the BDD you are discussing?
  by Allen Hazen
No, the "web-spoke" is a different design, sort of a 'semi-disc' driver I think. But clearly relevant to the general question of the pros and cons of these late driver-center designs. (So is the SCOA-P, about which there is a separate string…) The Baldwin Disc Driver design is distinctive: I suspect the best place to look for a good image would be a side-view of a Santa Fe 3460 class 4-6-4, or the ACL 4-8-4.
I'll see if I can find out more about the web-spoke design.
(B.t.w., the idea of Santa Fe putting a warbonnet colour scheme on the streamlined 3460 is intriguing: some one should do a COLOR rendition, or a model so we could see it from different angles. (Grin!))

You seem to have a better reference library than I do, but there does seem to be disappointingly little discussion of Disc drivers: I had looked at Bruce's book before starting this string… and kept looking, sure that I must have missed the relevant paragraph! … The Scullin stuff from Martyn Bane's(*) site would appear (from format, and a page number in the 600's) to be a Scullin ad from some edition of the Simmons-Boardman "Locomotive Cyclopedia": my guess would be, since I think the "LC" was quadrennial, 1937. The 1941 edition (reprinted by Kalmbach in the 1970s-- I have a copy of the facsimile) doesn't have a Scullin ad, but has a similar ad from GSC touting the virtues of the Boxpok. (In the next few days I will try to scan and post-- but neither my scanner nor my skills are great.) … I had had the impression that better balancing was the major benefit of Disc drivers, but I don't now see why pockets for lead counterweight couldn't have been placed just as advantageously in a more conventional wheel. The biggest claim in the Scullin and GSC ads seems to be that the disc designs give structural strength more evenly around the rim: conventionally spoked drivers apparently were prone to deformation (flattening) in the gaps between spokes. (And the test results I found somewhere -- Wikipedia? -- that were cited in favour of the SCOA-P design seemed to be about resistance to this sort of deformation.)
(*) A lot of Martyn Bane's site seems to be relevant to a project-- now apparently defunct-- to build a new-design, advanced technology 4-6-0 for use on excursion trains in Britain. For some reason-- MAYBE it is simpler, so easier to reproduce, than Boxpok-- they seem to have envisaged the use of Scullin-design driving wheels.
  by Pneudyne
The attachment shows the Baldwin disc drivers on an ACL 4-8-4.

Loco Profile #20 fc.jpg
  by Pneudyne
I had a closer look at the L&N 2-8-4 case.

I found in Swengel a picture of #1967 from the second series, and this is clearly shown with web-spoke drivers.

Farrell (1) stated that the first series had disc drivers. But from a re-examination of the pictures of that series, I think that it actually had web-spoke not disc drivers. Nevertheless, from some angles those web-spoke drivers do look like a Baldwin disc variation, and it would have been easy to make the assumption that they were, as they would not have been at all unexpected on a 1942 Baldwin product that was mechanically state-of-the-art. That’s my excuse, anyway, and I’ll happily lend it to Farrell. Staufer and May (2) made the same mistake. On page 77 of the said book is a picture of NYC J1d #5391 with PT-type tender and with web-spoked rear drivers, the caption for which includes the comment: "..and a sort of modified Baldwin rear driver.” But then in fairness, that was not intended to be a scholarly tome.

Anyway, the data for the L&N 2-8-4 now appears to be: web-spoke drivers for the first two (Baldwin) batches and Boxpok drivers for the final (Lima) batch. Web-spoke drivers as original equipment from a commercial builder seem to have been quite scarce.


(1) J.W. Farrell; North American Steam Locomotives: The Berkshire and Texas Types; Pacifica fast Mail; 1988; ISBN 915713-15012.
(2) A.F. Staufer & E.L. May; Thoroughbreds: The Most Famous Class of Locomotives in the World: New York Central’s Hudson; Staufer; 1974; LCC 74-75354.
  by rlsteam
Yes, I thought of looking up builder's photos of the AT&SF 3460 and the ACL 4-8-4s, but didn't take the time to do it (was working after midnight here). But mention was made of L&N second group of Berkshires having BDDs, and (as pointed out by Pneudyne above) those drivers are the web-spoke type in the Baldwin builder's photo of the 1967 erected in 1944. BTW, Allen Hazen, thanks for the comment on the "war-bonnet" AT&SF 3460 (http://www.railarchive.net/fantasysteam/index.html). I think it would have been much harder to digitally apply the "war-bonnet" livery to a color photo of the "Blue Goose," but it proved doable in B&W.
  by Allen Hazen
(Something I probably should have looked up earlier…)

"Trains," issue of December 1972, has a BIG article (32 pages, plus a colour reproduction of a painting originally done for an L&N publication as cover art) article on the L&N M-1 Berkshires.(*) It has good side-view builder's photos of all three orders of M-1, confirming that the 1942 Baldwins (1950-1963) and the 1944 Baldwins (1964-1969) had Web-spoke drivers, and the 1949 Limas had Boxpok. Text says that, among the "minor" design changes of the Lima batch was that "driving and trailing wheel centers were General Steel Casting's Boxpoks (instead of the Baldwin spoked)": it's not clear to me whether this is intended to mean only that the spoked wheels were used on the Baldwin engines, or that the Web-spoke design was a Baldwin proprietary design that Lima would have had to buy from Baldwin.

And the article WASwhere I had read the BLH story recounted earlier:
"…naturally after the BLH merger the boys in Sales and Engineering compared notes. An Eddystone man, still bothered by the 1948 L&N order's going to Lima, asked a former Ohioan, "How come?" After a bit of hedging, the Lima man's reply was "Well, after all, we [Lima] invented the Berkshire, and it just wouldn't have been right to let you guys [Baldwin] build <i>all</i> of L&N's 2-8-4's!""
(But the next paragraph suggests that Lima had underbid Baldwin by $10,000 per engine, which may have helped ensure that justice was done.)

(*) A "Railroad News Photo" in the same issue shows something with almost as much nostalgia-value as the steam article: Louisville and Nashville's GE U30C-XR diesel locomotive 1499, which was put into service on September 24, 1972, wearing a special paint scheme celebrating GE's "XR" improvements to its locomotives.
  by Pneudyne
Pneudyne wrote:Reed provided a little more information in his Loco Profile #2, “New York Central Hudsons”, and I’ll refer to that in a following post.
Page 40 of Loco Profile #2, copy attached, includes a picture of part of an NYC 4-6-4 with a full set of Scullin disc drivers, with the caption including the sentence: “First installation of the Scullin double-disc wheel on the N.Y.C., applied later to 25 of the J-3a Hudsons.”

So from that wording, one deduces that the picture was of a J-1 that had been retrofitted. Interestingly, the rod arrangement appears to have that for the 3rd driver coaxial with the main rod on the main pin, with the 1st driver rod connected by a knuckle joint. The J-3a with Boxpok drivers at least was, I think, the other way around.
Loco Profile #2 p.40.jpg
Loco Profile #2 p.43.jpg
Also on page 40, continued on page 43, is a commentary about cracking problems with the disc drivers that appeared to affect the Scullin type more than the Boxpok.

Page 42 included diagrams of the both the Scullin and Boxpok types.
Loco Profile #2 p.42.jpg
  by Pneudyne
[quote="Allen Hazen]Text says that, among the "minor" design changes of the Lima batch was that "driving and trailing wheel centers were General Steel Casting's Boxpoks (instead of the Baldwin spoked)": it's not clear to me whether this is intended to mean only that the spoked wheels were used on the Baldwin engines, or that the Web-spoke design was a Baldwin proprietary design that Lima would have had to buy from Baldwin.[/quote]

Those web-spoke wheels on the L&N 2-8-4 were probably not a Baldwin proprietary design. A web-spoke design – which looks just like those on the L&N locomotive - was covered by US patent 2222982 filed 1939 June 08, and assigned to Blaw-Knox.

The RME 1941 January article describing the NYC L-3a and L-3b class 4-8-2s, as reprinted in TSC #50, shows Union Steel Castings division of Blaw-Knox as the supplier of the wheel centres for the 10 (out of 50) locomotives that had web-spoke drivers, the other 40 having the Boxpok type.

So the web-spoke driver was relatively late on the scene as compared with the Scullin, Boxpok and Baldwin disc types. The NYC L-3 was probably an early application, in which the NYC perhaps wanted to see how it performed in practice. And the NYC used it as a retrofit on some of its 4-6-4s, as well. One imagines that to have gotten in the door at that stage, it must have challenged the established players on cost.

  by Pneudyne
To add to the complexity of this topic, it would appear that there was a variant of the Baldwin disc driver that did not have the inter-hole stiffening ribs. For example, these are lacking in the close-up pictures of the drivers of the WM 4-6-6-4 in TSC #47, and the Pennsy T-1 prototypes in TSC #56. Possibly that was an evolutionary change.

But then to add some confusion, Farrell & Pearsall (1) included a picture of AT&SF 4-8-4 #3780 (1941 batch) sans ribs and a picture of #3767 (1938 batch) with ribbed 1st, 3rd and 4th driver sets, but non-ribbed main drivers. Of course, the main drivers could have been changed.

Then Farrell & Pearsall (2) show B&M 4-8-2 #4115, R-1d class (1941) with all non-ribbed Baldwin disc drivers. #4110, R-1b class (1937) had ribbed Baldwin disc main drivers, the other being conventional. There is not a clear picture of an R-1c class, 1941, and apparently with disc main drivers only, like the R-1b.


(1) J.W. Farrell & M. Pearsall; North American Steam locomotives; The Northerns; Pacific Fast Mail; 1975; LCC 74-33883.
(2) J.W. Farrell & M. Pearsall; North American Steam locomotives; The Mountains; Pacific Fast Mail; 1977; LCC 76-13756.