ACL Class R-1 4-8-4's

While the performance of the R-1's initially exceeded the railroad's expectations, shortly after the locomotives entered service, problems with the counter-balancing of the locomotives' running gear became apparent. Although the R-1's were balanced per the Association of American Railroad's (AAR's) latest published recommendations of 1934, at high speed it was found that the main drivers actually left the rails and repeatedly slammed back down, kinking rails and damaging track alignment for miles. Both Baldwin and Atlantic Coast Line engineers spent much time analyzing this issue. High speed cameras were installed on at least one locomotive which was operated on the ACL mainline, and analysis of these films showed that the main drivers were actually leaving the rail when the locomotive was at high speed (~80 MPH). After much investigation by Baldwin and ACL engineers, the locomotives' counter-balancing was reduced by removing a significant amount of balancing material from the drivers and the problem was eliminated. According to former ACL machinist Jerald McGowan, the locomotives were not modified immediately, but one at a time as they came into the shops for overhaul. As a result, these modifications were still being carried out after WWII. The unmodified locomotives were used in fast freight service, where speeds were lower, until the modifications had been carried out at which time they were used in passenger and fast freight service. 6 of the engines were additionally modified with lightweight main rods, pistons, piston rods, and crossheads to further improve their balance.

The root of this problem is that the driving wheels of a 2-cylinder conventional steam locomotive cannot be perfectly balanced. If you'd like to read more about the physics and engineering behind the balancing of a steam locomotive's driving wheels, this "white paper" at the Coalition for Sustainable Rail (CSR), co-authored by me, provides detailed information: https://csrail.org/s/wp_slrwd_2-1.pdf

Although the balancing calculations were correctly made according to the existing engineering standards when these locomotives were designed, forces were at work on the running gear of the locomotives which were beyond the capabilities of the engineering of the 1930's to analyze. The interaction of the driving wheels, machinery of the locomotive, the springs on the locomotive, and the rails and roadbed would be a challenge to accurately model even with modern computers.  Rules had been established for counterbalancing based on trial-and-error. These rules were found to be outdated when it came to large, modern steam power. The Union Pacific, who built many modern steam locomotives in the 1930's and 1940's, ignored the AAR's recommendations and developed their own guidance for balancing and avoided the problems that plagued these locomotives. Unfortunately, as a result of these problems and competitor Seaboard Air Line's success with new EMD diesel-electric locomotives, the ACL's successive new motive power purchases were all diesel. Some other railways with modern steam power built about this time (New Haven 4-6-4's, Pennsylvania 2-10-4's) suffered similar problems. These problems were significant in discouraging further purchases of modern steam power and unfortunately accelerated the change to diesel-electric motive power in the U.S. The photo below, possibly taken on the ACL, shows the results of an improperly balanced steam locomotive at high speed.

Photo of Damaged Rails from Ralph P. Johnson's book "The Steam Locomotive"

The R-1's were overhauled at the ACL shops in Tampa, Florida, as they were too large for the road's backshops in Rocky Mount, North Carolina. Tampa is southwest of the ACL's Jacksonville to Richmond mainline, which required the locomotives to stray from their usual operating territory when heavy repairs were required. The R-1's were limited to 40 MPH over most of this line, and even slower over some bridges, so this must have been a headache for the railway.