McKenna Process Company
Plant formerly located in Joliet, Illinois - ca. 1920.
Autobiographical Data - The Metallurgy of Steel
by George Langford, Sr.

When I came to the McKenna Company in March, 1898, steel rails were being manufactured by the Bessemer process, but the Open Hearth process was superceding it.  Some rail joints were being made by the Sellers Manufacturing Company out of puddled wrought iron, but the large steel companies were following the general trend of Open Hearth succeeding Bessemer.
Railroads were having trouble with railroad rails, battered at the joints, which made rough riding.  Some railroads removed these rails, cut the ends off at cold-sawing plants, drilled new bolt-holes, and then relaid the rails in track.  Some railroads used the battered rails in secondary track or sold them to independent mills where they were rolled into concrete rods and other products.  The various treatments or worn rails were a more or less thriving business.
Then came the McKenna process of "renewing" worn rails by heating the worn rails and putting them through two passes of rolls which reshaped the worn parts and elongated the rails enough to saw off the worn rails at the ends and preserve the original length and fit of joint.  This was widely heralded as a new and wonderful idea and was backed by men of means, who built a mill at Joliet, Illinois, which rerolled some rails in 1897.
When I took the job in 1898, the mill was operating only part time, and yet a mill was being erected at Kansas City, Kansas, which did some part time work that year and in 1899.  In spite of the fact that neither one of the two mills had reached anywhere near full operating schedule, another mill was being erected in New Jersey with dockage for ocean vessels.  To cap the climax, an English company was formed, and another mill was in process of construction in 1903.  I was sent to each of these four mills when they operated, not one of them operating one-third of the time.  And yet those who furnished the money for these ventures were convinced that the business would prosper exceedingly.  Reclaiming worn rails was a popular theme in those days, and some railroads reported better wearing quality in rails renewed by the McKenna process.
But the worn rails thus treated were of medium carbon Bessemer steel which could stand hardening by rerolling.  The higher carbon Open Hearth rails began coming to the mills, and results after rerolling were disastrous.  Breakage in process of manufacture and in track became prohibitive, and business dwindled beyond hope of profit.  It was not difficult to find the reason.  Overheating made the steel brittle, and when properly heated, the rail was distorted by the strains set up by cooler rolling.  Long lengths of intricate section made proper heating extremely difficult.  Many remedies were tried but with no permanent success.  Reclaiming worn rail joint bars in press and dies provided a respite and did well for a time, but again the proper heating of high carbon Open Hearth steel was the stumbling block.  Eventually all four McKenna mills went out of business, simply because high carbon steel rails and rail joint bars could not stand the heating and rolling punishment to which they were subjected.

Note added by George, III: The rerolling process deformed the steel only a few percent.  It is now known that such a few percent (not more and not less) of deformation, followed by the recrystallization that defines the hot rolling process, will result in extreme coarsening of the grains, embrittling the steel.  Renewing of the rails (and splice bars) by the methods described here might well have succeeded if my grandfather had had the inspiration to use the first pass through the mill rolls to over deform the rail, followed by redundantly deforming the rail back to the desired shape in the second pass.  Sufficient deformation to obtain an smaller grain size and acceptable toughness might then have been obtained.  He came close, first with US Patent No. 1,212,964 for rolling rail of unsymmetrical cross section and then with his US Patent No. 1,724,031, which describes the steps of prequenching a just-deformed renewed splice bar before immersing it in hot, viscous oil and then cooling the bar very slowly by placing it in a stack of similarly treated, hot bars.  The prequenching stops the undesired recrystallization, and the hot oil minimizes the formation of proeutectoid ferrite, permitting the bulk of the austenite retained thereby to transform nearly isothermally to pearlite during the slow cooling in the stack of hot bars.  Modern rails are processed as closely to 100% pearlite as possible, as this gives the best combination of strength and durability.