Chapter 11: Production of Tanks
In a general way, tank building followed the methods of automobile production. Major components were produced in widely separated plants and then brought together and assembled at an assembly plant. Although some tank contractors made more components in their own shops than did others, none made them all. Armor plate and castings, for example, came from Pittsburgh or Chicago steel mills and foundries in a rough or semifinished state. The guns were supplied by Ordnance arsenals or commercial producers. Rubber-bushed tracks came from one of the major rubber companies in Ohio.1 Within the tank’s enveloping armor the two most important major assemblies were the engine and the transmission, but there were also radios, periscopes, ammunition racks, and countless other items. Most tank parts had two things in common—they were very heavy, and they were made chiefly of steel. More than any other factors, these two determined the pattern of tank production. They required plants with big cranes to handle heavy assemblies, ingenious fixtures to hold parts in position, and a great variety of huge machine tools for cutting and shaping the material. (Table 19)
The Schenectady plant of the American Locomotive Company was such a plant. Its tank assembly line—adjacent to continuing locomotive production areas—was a series of seven stations at each of which a major component was added. Starting with the lower hull, or chassis, the gas tanks and the mount for the big gun were first put in place. At the next station the giant transmission was added. At the third stop an overhead crane lowered the engine into place and the drive shaft was connected with the transmission. As the hull moved slowly from station to station it gradually took on the appearance of a fighting tank, finally rolling onto its tracks and receiving its big gun and turret.2
Although the locomotive companies were able to use much of the equipment they had on hand, many new machine tools were required, as well as additional handling equipment. American Locomotive not only needed over one hundred new machine tools for its first tank order but also had to rearrange its entire plant layout to make room for them.3 For its first light tank order, American Car and Foundry installed seventy-five new tools and a series
Table 19: Tank production by facility, 1940-1945
Breakdown of light, medium, and heavy tanks by facility but does not include experimental tanks by development division.
| Facility | Total Production Through 31 Dec 45 | Percent of Total |
| Total Tank Production-1940-1945 | 88,410 | 100.0 |
| Detroit Tank Arsenal | 22,234 | 25.2 |
| American Car & Foundry | 15,224 | 17.2 |
| Fisher Tank Arsenal | 13,137 | 14.9 |
| Cadillac Motor Company | 10,142 | 11.5 |
| Pressed Steel | 8,648 | 9.8 |
| Pullman-Standard | 3,926 | 4.4 |
| American Locomotive Works | 2,985 | 3.4 |
| Baldwin Locomotive Works | 2,515 | 2.9 |
| Massey Harris Company | 2,473 | 2.8 |
| Ford Motor Company | 1,690 | 1.9 |
| Lima Locomotive | 1,655 | 1.9 |
| Montreal Locomotive Works | 1,144 | 1.3 |
| Marmon-Herrington | 1,070 | 1.2 |
| Pacific Car and Foundry | 926 | 1.0 |
| Federal Machine | 540 | 0.6 |
| Rock Island Arsenal | 94 | * |
| International Harvester | 7 | * |
* Less than 0.05 percent.
Source, Tabulation dated 14 Jan 46 in folder, Col. Colby charts, OCO Detroit files.
of heat-treating furnaces.4 The Detroit Arsenal required over 1,000 machine tools and some 8,500 specially designed jigs and fixtures. All the companies had to pioneer in developing new techniques as well as new tools and fixtures, particularly for
Overhead crane lowering a General Stuart M3 light tank onto its tracks at Rock Island Arsenal.
welding, cutting, and straightening heavy armor plate.
The Detroit Tank Arsenal stood in sharp contrast to the locomotive plants, for it was built from the ground up for the sole purpose of building tanks. But it nevertheless went through a series of rather drastic changes. Before the arsenal was built, Knudsen’s idea of having it produce its own armor plate—and practically all other parts—was abandoned, the first step in a long process of decentralizing tank production. The year before, American Car and Foundry had installed furnaces for face-hardening its own plate because other sources were not readily available, but the planned production schedules for the Detroit Arsenal were so high —and the use of thicker homogeneous plate was rising so fast that Chrysler decided to buy its armor plate and heavy steel castings from other firms. Nearly all other parts, except guns, were made at the arsenal during 1941, including the famous 30-cylinder engine built by gearing five Chrysler truck engines to a single drive shaft, and hundreds of extra transmissions for other tank contractors. The arsenal became a well-integrated basic pilot plant.5
In 1942, as tank requirements zoomed and automobile manufacture stopped for the duration of the war, Chrysler began
farming out its operations. This was in accord with Keller’s original plans for expansion in time of a real emergency. Between February and September more than 700 large machine tools were moved to other Chrysler plants, some of which machined gears and suspension wheels while others welded hulls and still others performed a variety of machining, forging, and assembling jobs. More and more, the tank arsenal became a final inspection and assembly plant supplied by scores of other production units.6 Looking back on the experience from the vantage point of the year 1948 Keller observed that
... the job experienced all the standard hardships of World War II production. The first design was scrapped before we could begin. Despite the early start made, the value of priorities for machine tools and equipment quickly melted away like snow on a hot day. Frantic calls for increased production alternated with drastic cutbacks. Disappearance of critical materials held it up. Sudden changes in design upset ability to deliver, and broke the planned flow of operations. We never once had all of the machine tools and equipment that our schedules called for.7
In tank production, as in other phases of Ordnance procurement, industry integration committees played an important role in bringing manufacturers together to eliminate bottlenecks and speed production.8 Beginning with the medium tank committee’ in 1942, a total of twenty-seven were created by April 1943 when the last one was organized. Many committees were active only for short periods because the problems they dealt with were successfully solved. But some lasted all during the war years. Speaking of the medium tank committee, one Ordnance officer aptly described the work of them all. “You might think of this committee,” he remarked, “as being a great merger of tank plants all combined under the trade mark of the Ordnance Department and all making the same product—the American medium tank.”9
Engines
Tank engines constituted one of the worst bottlenecks early in the war.10 Before 1940, Ordnance tank designers had planned to use both diesel engines made by the Guiberson Company and “Whirlwind” gasoline engines made by the Wright Aeronautical Corporation, but as the defense program got under way these firms could not meet the tremendous demand for tank, plane, and ship engines. As airplanes and ships had top priority, Ordnance had to look for other sources. In the fall of 1940 Ordnance contracted with the Continental Motors Company to rehabilitate its old Detroit plant to produce the Wright aircraft engine, under a license arrangement, at the rate of twenty engines per day. Early in 1941 Ordnance felt that, with more than six thousand engines
contracted for, it faced no real problem, but as the tank program was doubled and redoubled in ensuing months, the need for engines far exceeded earlier calculations.11
In the winter of 1941–42 there was no time to design and test a completely new tank engine, and build new plants for its manufacture. Substitute engines that could be produced at once, using tools already at hand in existing plants, had to be adopted, including the Chrysler multi-bank, the GM twin diesel, the Ford V-8, and the Caterpillar RD-1820—an air-cooled radial diesel. The Guiberson diesels were manufactured for a time at a new plant in Garland, Texas, but the contract was later terminated and the plant taken over by Continental Motors to make gasoline engines.12 The picture was further complicated by the preference of the British and Soviet governments for diesel engines in lend-lease tanks, and by differences of opinion among U.S. Army authorities as to the relative merits of gasoline and diesel engines for tanks. As early as April 1942 Maj. Gen. Jacob L. Devers of the Armored Force had urged elimination of both the Guiberson and Chrysler engines.13 But the demand for engines was so great that every reasonably acceptable type had to be used, even though this practice played hob with field maintenance and spare parts supply.
Meanwhile engines scheduled for tanks were diverted to the Navy or the Air Forces, sometimes without consulting the Ordnance Department. At a production conference in June 1942 General Christmas reported that since the first of the year, 2,500 GM diesel engines had been diverted to the Navy, and General Clay added that 1,100 more had recently been diverted. General Christmas bluntly warned the conference as follows:
We cannot stand any more diversions of engines and still meet the 1942 tank objectives. ... They are continually calling me into meetings where they want to take the Wright engine and put it into training planes. Fifty percent of the tanks made this year will have Wright engines, 34 percent will have General Motors engines, izo percent will have Chrysler and six percent will have the Ford. So if they start taking away General Motors’ and Wright’s engines, they are taking away the foundation of the program.14
The supply of engines improved gradually during the year, but remained a problem even in the early months of 1943.
In June 1943 General Christmas reviewed the whole engine problem in a
letter to General Clay. He pointed out that the War Department goal was to use only one type of medium tank engine and then outlined the reasons for continuing with the half-dozen existing models. He cited the “wide divergence of reliable opinion” on two points: the relative merits of diesel and gasoline engines, and the merits of liquid-cooled versus air-cooled engines. Furthermore, he pointed out that none of the American engines had yet seen extensive battle service, and all were still in various stages of engineering development.15 Considering all factors, ASF approved the continued use of all existing engines, but their number was soon reduced by elimination of the less desirable types.16
Transmissions
Transmissions and final drives—described together as power trains—were, at the start of the rearmament program, as troublesome as tank engines, but they were well under control by the first anniversary of Pearl Harbor. The gears and castings needed for tank transmissions—of special Ordnance design and much larger than commercial products—were not easily manufactured. When American Car and Foundry began production of the light tank in 1940 it obtained transmissions from the Timken-Detroit Axle Company, a firm that had shared in their development and had specially equipped itself for their production, and from Spicer. In 1941, when production began on the General Grant, the railway equipment companies, as noted in the preceding chapter, obtained transmissions from the Mack Manufacturing Company, a pioneer in this field, and later also from the Iowa Transmission Company. The Detroit Arsenal not only made its own transmissions but also supplied other contractors. At the outset, Ordnance purchased transmissions from the producers and furnished them as “government free issue” to tank contractors, but later Ordnance stepped out of the picture and let the tank builders buy transmissions direct. During the winter of 1941–42, as requirements mounted, Ordnance took steps to create monthly capacity for over five thousand medium tank transmissions. The Buick Division of General Motors, the Ford Motor Company, the Reed Roller Bit Company, and Caterpillar Tractor all came into the program and by the end of 1942 production had caught up with demand.17
Lack of machine tools was at the root of the delay in transmission production
during 1941. In the case of one contractor, the Mack Manufacturing Corporation, production was hindered by the firm’s inability to get delivery on four vertical boring mills. In the fall of 1940 Mack had made the first two M3 medium tank transmissions for Ordnance by hand, but volume production had to await the installation of special equipment. The first two Mack transmissions were described at the time as “the most traveled transmissions in the history of Ordnance,” for they were sent to tank plants all over the country as pilot models. Mack and other contractors could not get the machine tools they needed before Pearl Harbor because their contracts carried a relatively low priority, but the high tank production goals set early in 1942 brought higher priorities and eventually eased the machine tool problem.18
Armor
Pre-1940 tank production did little to prepare American industry for manufacture of tank armor in World War II. The few tanks built in the 1930s were made with comparatively thin steel plates of high nickel content, face-hardened on the outside. To protect light vehicles against small arms fire—the function of armor in the 1930s—these face-hardened plates were the best known material, for they had a hard, bullet-resistant surface and a tough back. Because the metallurgical composition of face-hardened plates made welding extremely difficult, the plates were riveted or bolted together. As a result, tanks of the 1930s were not only lightly armored but had a boxlike shape, were studded with rivets—two thousand in every light tank—and offered many flat surfaces to enemy fire.19
All these characteristics went out the window before the war was over. Suddenly faced in 1940 with a demand for armor protection against artillery fire, Ordnance had to develop and produce a radically new type of tank. One-inch armor gave way to 3-inch and 4-inch steel hulls, and by 1945 study was being made of armor from five to ten inches thick. Face-hardened plates were supplanted by homogeneous armor that permitted the welding of joints and speeded production.20 Flat, angular surfaces gradually disappeared as cast hulls and turrets with rounded con-tours—less vulnerable to enemy fire—came into production. Meanwhile the use of nickel and other scarce alloys was reduced, and new techniques were developed for welding, casting, and heat-treating tank steel. The armor on 1945 tanks was as different from that on 1939 models as the 90-mm. gun was from the 37-mm. on prewar tanks. More than any other factor, it accounted for the doubling and tripling of tank weights, for armor accounted for more than half the weight of World War II tanks.21
As early as 1938 Ordnance had achieved some success, after years of experimentation, in welding a medium tank turret of face-hardened plate. The advantages of welding over riveting were readily recognized, but the difficulty of the process delayed its adoption until 1941. Early that year two welded M3 medium tank hulls were produced, one by the Carnegie-Illinois Steel Corporation and the other by Rock Island Arsenal.22 The most troublesome problem at the start was finding a way to keep the plates from buckling during the welding process, and to eliminate cracks that appeared in the armor. Working on the light tank, the Cadillac plant tried reinforced fixtures but the plates buckled even in the strongest frames. To aid in overcoming these difficulties the Ferrous Metallurgical Advisory Board formed a subcommittee on the welding of armor. Composed of industry and Army members, this subcommittee studied methods, procedures, and specifications and periodically submitted its recommendations. It was only after countless experiments by all the producing firms that a complicated procedure of welding operations was devised that partially solved the problem, but the real solution lay in switching to homogeneous plate.23
The switch to homogeneous armor was prompted by several factors in addition to the welding problem. For one thing, face-hardened armor was so difficult to produce and hard to machine that its use in the expanded tank program of 1941-42 was out of the question. Building thousands of tanks with face-hardened plate would have made impossible demands on an economy in which manpower and machine tools were at a premium. Homogeneous armor was not only easier to produce but could be produced either by rolling or casting. The case for homogeneous armor was further strengthened when test firing showed that, if properly sloped, it had resistance to penetration substantially equal to face-hardened armor. An additional advantage was that homogeneous armor had less tendency toward “back spalling,” i.e., splintering of the back under impact of a projectile. As a result, homogeneous armor was authorized for all areas where a sloped surface could be presented to the enemy.24
The next step in the process of tank armor development came with the use of cast armor. Before 1940 neither the U.S. Army nor any other army in the world had made use of such armor except at points, such as the transmission housing, where the shape and contour were such that plates could not readily be used.25 In 1939 the General Steel Castings Corporation of Eddystone, Pa., designed and produced for Ordnance a one-piece cast upper hull, claimed to be the first of its kind ever produced.26 In June 1940
procurement of a 6-ton cast upper hull for the new M3 medium tank was authorized. When ballistic tests of the first models—described as resembling “inverted bathtubs for elephants”—showed excellent results the cast hull was adopted as an alternate type, and further development was pushed. In September 1940 the Ferrous Metallurgical Advisory Board formed a subcommittee on cast armor to draft specifications and to advise on production techniques. The cast hull, formed all in one piece, not only eliminated the need for riveting or welding together over one hundred separate plates and castings, but also facilitated the production of hulls with rounded contours.
In the fall of 1941, and again in January 1942, when tank requirements were doubled and redoubled, the chief armor procurement problem was enlarging production capacity for armor of all kinds—face-hardened and homogeneous plate and castings of many shapes and sizes. Some armor-making capacity had been created in 1939 and 1940 but it was wholly inadequate to meet the needs of 1941-42.27 Beginning in September 1941, Ordnance arranged for the expansion, with Defense Plant Corporation financing, of nearly all existing cast armor plants, notably American Steel Foundries, Continental Foundry and Machine Company, and the General Steel Castings Corporation. The Ford Motor Company built a foundry with a capacity of ten thousand tons per month, of which nearly half was to be cast armor.28 Two safe manufacturers, Diebold and Mosler, produced face-hardened plate. Early in 1942, when production of rolled armor also had to be increased, leading steel producers, such as Republic Steel, Henry Disston and Sons, and Carnegie-Illinois, expanded their rolling mills. The latter corporation not only operated the government-owned Gary Armor Plant in Indiana next to its Gary steel mill, the largest of its kind in the world, but also enlarged its plant at Farrell, Pa. American Car and Foundry, a pioneer in the field, expanded its capacity for making armor plate for light tanks and eventually became the largest producer of face-hardened armor plate in the United States. The Pacific Car and Foundry Company of Renton, Washington, enlarged an existing foundry to make its own armor and thus avoid heavy shipments from the Chicago area to the West Coast.29 The Standard Steel Spring Company of Detroit contributed greatly to the program when it became the coordinating agency for a pool of firms that normally made automobile springs, bumpers, and related equipment. The need to build a new plant was avoided when the facilities of these firms were used
to cut, harden, straighten, and machine rolled plate received from steel mills.30
By the time plans for armor production were well in hand the first cutbacks came in September 1942. The Army Supply Program published at that time dropped the requirement for heavily armored assault tanks, thus reducing the cast armor requirement from about 77,000 tons per month to 57,000. When the over-all program for tanks and other combat vehicles was further cut in November the armor-producing plants felt the effect immediately. Expansions under way were abruptly canceled, and Ordnance began a detailed review of its future needs.31 In selecting plants to be closed down, Ordnance was guided by the desire to retain in production three types of plants: the older facilities, those in advanced state of completion, and those with excellent production records. Wherever possible, excess plants were converted to other war production through transfer to the Air Force, Navy, or Maritime Commission.32
Light Tanks: M2A4 to M24
In terms of numbers produced, light tanks led the procession in 1939, 1940, and 1941. These 13- to 18-ton machines mounting 37-mm. guns were the first American tanks to come into production in 1940 and in 1941 outnumbered medium tanks by nearly two to one. They were used effectively by the British in North Africa in 1941-42, particularly as reconnaissance vehicles. But in 1943 they fell behind as the demand for more powerful tanks continued and production of Grants and Shermans gained momentum. In 1945 the number of light tanks produced was less than half the number of mediums.33
In the spring of 1941, while American Car and Foundry was producing early model light tanks at its Berwick plant in Pennsylvania, the Cadillac Division of GMC proposed to Ordnance that a light tank be built with twin Cadillac engines and automatic transmission, then a new development in the automotive industry. Ordnance was reluctant to change from the air-cooled engine, but the need for tank engines was acute and test reports on a Cadillac-powered model were favorable. Furthermore, the Cadillac engine was easier to start; it operated better at idling speeds; and the hydramatic transmission made the tank driver’s job much easier. In October 1941 a Cadillac-powered tank proved its durability by running under its own power all the way from Detroit to Aberdeen, a distance of over five hundred miles. In the course of installing its engine and transmission in the standard M3 light tank, Cadillac had made so many design changes that, when adopted, the tank was given a new model number, M4, later
changed to M5 to avoid confusion with the M4 Sherman tank.34
The M5 was the first combat vehicle to use the new automatic transmission, which soon gained widespread acceptance.35 The first tank came off the Cadillac assembly line in Detroit at the end of March 1942, and in July the Cadillac plant at Southgate, California, turned out its first tank. Meanwhile another producer, the Massey-Harris Company, a farm implement firm, came into the picture. It took over the former Nash-Kelvinator plant in Racine, Wis., rounded up the needed machine tools, and, with the aid of Cadillac, got into production as an assembly plant for the M5. In October 1943, American Car and Foundry switched to the M5 and production of M3’s stopped altogether.36
The changeover from automobiles to tanks at the Cadillac plant was accomplished speedily, but not without the usual conversion problems. Makeshifts were the order of the day, for new equipment specially designed for tank production was virtually unobtainable. Because jigs and fixtures, so essential to mass production, take a long time to make, Cadillac did without them at the start, building its first tanks almost by hand. The company sent representatives all over the country to look for used machine tools, and, as it did not itself plan to manufacture scores of tank parts, to discover sources of parts supply. In January 1942 it set up a “parts clinic” in its new car show room exhibiting 189 tank parts and inviting potential suppliers to examine them and quote prices on such items as oil pumps, axle housings, clutch drums, herringbone gears, and axle shafts.37
In the post-Pearl Harbor drive to build twenty-five thousand light tanks a year, Ordnance created another new facility, the Quad Cities Tank Arsenal at Bettendorf, Iowa. Purchasing adjoining plants owned by three private firms—one in bankruptcy—Ordnance contracted with the International Harvester Company to operate them as an integrated unit. The roof was repaired, new concrete flooring laid, and new wiring installed throughout so that International Harvester could build a new model tank known as the T-7 at a rate of 750 per month. The arsenal was intended to be purely an assembly plant, with engines, transmissions, final drives, and all other components coming in from subcontractors, but the arsenal did some machining and welding of hulls, turrets, and rings.38
The history of the Quad Cities arsenal during 1942 shows the tank program in its worst light. At the start there was great demand for speed, high rates of production, and a “cost be damned” attitude. The company placed orders for two thousand
new machine tools with firms that were already swamped with tool orders and could not promise delivery for months. More than three thousand drawings were needed for the complete tank but, as Rock Island was still at work on the pilot models, the drawings were not available. As the priority rating for the plant was not high enough to give it a green light, countless hours were spent in trying to expedite the project. Then in the middle of the summer Ordnance decided upon a major change of design—equipping the tank with a 75-mm. rather than a 57-mm. gun. This meant redesigning the turret and adding to the over-all weight of the vehicle. Even before the gun was changed, the tank, originally expected to weigh between eighteen and twenty tons, was up to twenty-five tons. The added weight was too much for the engine and made the tank slow and hard to steer. To remedy the deficiencies as they appeared, Rock Island issued hundreds of revised drawings every month, and continued to do so until near the end of the year when the contractor and Ordnance representatives finally agreed to freeze the design. The first essentially complete tanks were shipped in December 1942 just as cutbacks in all tank production were taking effect. Scheduled production at Quad Cities was then scaled down and discontinued completely in April 1943, after completion of only thirteen tanks.39
Meanwhile the Marmon-Herrington Company of Indianapolis undertook production of the M22, an 8-ton airborne tank, and Ordnance began planning to switch from the M5 to a more powerful light tank, the M24. In the spring of 1944 manufacture of the M5 at ACF, Cadillac, and Massey-Harris was discontinued. ACF dropped out of tank production at this time after having produced over fifteen thousand tanks, more than half the entire wartime output of light tanks. In 1944 and 1945, Cadillac and Massey-Harris, the only two producers, turned out a total of 4,731 M24 tanks mounting the 75-mm. gun and weighing approximately 20 tons.40
The Shift From Grants to Shermans in 1942
The most far-reaching change in production plans for medium tanks during 1942 was the shift from the Grant (M3) to the Sherman (M4).41 From the very start, the Grant had come in for a lot of criticism. It had been hurriedly designed in 1940, after the German offensive had demonstrated the unsuitability of existing mediums. It went into production in spite of inadequate test and development because it was more advanced than the design that later became the Sherman.42 During its initial production, when the
inevitable “bugs” were being eliminated, engineering changes were authorized at the rate of three thousand per month. They ranged from minor modifications in track design to major changes in the shape of the hull or turret. The riveted hull of the early Grants gave way to a welded hull on later models, and the welded hull on the first Shermans eventually gave way to a cast upper hull.43
The shift from Grants to Shermans was gradual, starting in July 1942 when the first Shermans were made by the Fisher Tank Arsenal, which, unlike the Chrysler arsenal, had been designed with production of Shermans in mind. Chrysler produced its first Sherman in July 1942 and closed out production of Grants on 3 August and the railway equipment companies followed suit. The nearly five thousand Grant tanks built in 1942 remained in service during 1943 as “limited standard” and were not declared obsolete until early in 1944. They posed a major problem of disposal for Ordnance. “We are beginning to run into the motor car dealer’s problem,” Colonel Christmas commented. “Our customers, the fighting men, want only the latest models.”44 Some Grants were converted to tank recovery vehicles, i.e., their heavy guns were removed and replaced by powerful winches for towing disabled tanks. Others were used for instructional purposes by Ordnance evacuation companies and by Engineer training centers. The rest of the Grants were dismantled, with usable parts salvaged and the remainder disposed of as scrap.45
Discontinuance of the Grant models did not completely simplify the matter by any means, for there were five ‘different models of the Sherman tank in production in the United States by the end of 1942.46 The essential differences were in the engines. The original M4 was powered by a Continental radial aircraft engine. The M4A1 also used the Continental radial engine but had a cast rather than a welded upper hull, and improved surface contour. The M4A2, shipped in large quantities to the Soviet Union and the British, was powered by twin General Motors diesels. The M4A3 had a Ford GAA, and the M4A4, which went to the British, had the multibank Chrysler engine. The designation M4A5 was assigned a model with a 57-mm. gun produced in Canada for Canadian use, often called the Canadian Ram. The M4A6 was powered by a radial air-cooled diesel-type engine manufactured by the Caterpillar Tractor Company.47
The armament of the M4 tanks introduced further complications. The original design mounted a 75-mm. gun as its main weapon, but later models were equipped with the high velocity 76-mm. gun, and a few were supplied with 105-mm. howitzers. Early in 1944 limited procurement of a heavily armored “assault tank” known as
the M4A3E2 was authorized.48 As production of the new heavy tank M26 got under way early in 1945, production of mediums was gradually scaled down and ended in July 1945 after some 57,000 had been built.
Tank Depots
The experience of 1940-41 demonstrated that it was not practical for the tank plants to install the scores of minor accessories—radios, spare parts, small arms, first aid equipment, interphones, and Chemical Warfare items—needed to make a tank fully ready for battle. Unavoidable delays in getting delivery on all such items slowed down production at the factories, and the spectacle of scores of nearly complete tanks standing for weeks outside factories waiting for some small parts had a bad effect on labor morale. Further, as supply of many of these accessories was the responsibility of the Government, not of the contractor, Ordnance considered it more sensible to install this “On Vehicle Matériel” itself rather than disperse it to the contractors’ plants. It was for this purpose, along with others, that the tank depots were established.49
The depots were intermediate facilities that received tanks from manufacturers in a reasonably complete condition, installed required items, made any special modifications needed, prepared tanks for shipment, and stored them until shipping instructions were issued. In January 1942 Ordnance took over the New York Central Railroad shops in Toledo as its first tank depot, and soon contracted with the Electric Auto-Lite Company for its operation. Two others were opened shortly, both operated by the Ford Motor Company. The first was at Chester, Pennsylvania, and the second at Richmond, California. Both were Ford assembly plants and were well located for both rail and water shipment. In December 1942, when work on a proposed gun plant at Lima, Ohio, was discontinued, the plant was taken over as a tank depot, soon replacing Toledo. It was operated by the United Motors Service Division of GMC. The Longue Pointe depot at Montreal, Canada, was a Canadian Army installation used by Ordnance primarily for processing tanks and other vehicles shipped to the United Kingdom on lend-lease.50
Early in the war, when most overseas shipments of tanks went to other nations on lend-lease, it was discovered that some standard U.S. equipment was not suitable for those countries. British rather than American radios had to be put in tanks going to the United Kingdom or Russia, and all tanks destined for British Army use were equipped with sand shields, smoke generators, and a smoke bomb thrower. To avoid confusion at the plants, the depots were given responsibility for installing this special equipment. Meanwhile field reports on defects and proposed modifications were received by Ordnance, and, on tanks already built, the approved changes were made at the depots. In addition, major modifications were made on certain vehicles to meet special needs, as when some
medium tanks were converted to mine exploders and other tanks and gun motor carriages were converted to prime movers.51
The early experience with shipment of vehicles overseas “in accord with best commercial practice” was disastrous. The tanks arrived in badly damaged condition after a long sea voyage. Special techniques had then to be worked out to protect the engines, fire control instruments, and other parts from damage by rough handling or exposure to salt water. Accessories were individually packed in boxes and stowed inside the vehicles. Engines were protected with an internal coating of preservative oil and an external rust-preventive spray. After a desiccant was hung in the engine and crew compartments, these sections were sealed with a waterproof tape. The guns were treated with a heavy rust preventive and sealed at the muzzle. To permit the vehicle to be towed and steered without breaking the sealing tape, cables were attached to the steering levers and brought out through the bow gun mount.52
The worst problem for the depots at the peak of tank production was the failure of accessory shipments to keep pace with tank shipments. Tanks reached the depots with innumerable shortages of tools, equipment, and supplies. The situation grew more critical all during 1942, with more than ten thousand tanks deadlined in November when tracks were in extremely short supply. The shortage of tracks stemmed chiefly from lack of alloy steel and the doubling of demand for spares. The tank producers appealed directly to the WPB for more track steel, and Ordnance assigned expediters to follow up all types of parts.53 In January, General Christmas forbade manufacturers to ship tanks to depots unless completely equipped according to the latest On-Vehicle Matériel List. By March 1943 the shortages on vehicles at depots, which had averaged forty items per vehicle in November, had dropped to three.54
The 1943–45 Period
By the spring of 1943 light and medium tanks were rolling off the assembly lines of sixteen plants at the rate of about four thousand per month. This was roughly half the designed capacity of the plants, not counting Quad Cities Arsenal, which stopped producing tanks in April. The over-all capacity of 7,705 tanks per month had been created during the preceding two years at a cost to the government of approximately $250,000,000 for tools, equipment, and buildings. In May 1943 a representative of the Bureau of the Budget was able to report that, all things considered, the tank program had “gone very well,” with most of the contractors getting into production with surprising promptness. But he raised serious questions on two points: the wide variations among the
producers in the cost of tanks, as shown in contracts, and the need for closing plants no longer necessary to meet falling requirements.55 General Glancy and other officials of the Tank-Automotive Center strongly objected to this report on the ground that it contained factual errors and drew unwarranted conclusions.56
The estimated costs of medium tanks under contract in the spring of 1943 showed a wide spread, from $33,500 for the Grant tank and $42,400 for the Sherman tank at the Chrysler arsenal to $70,000 for the Sherman tank at the Federal Machine and Welder Company. Both firms had so-called fixed price contracts, but, General Glancy pointed out, the prices were actually not fixed at all because they were subject to redetermination and were, in fact, little more than estimates. Further, the operating conditions of the two firms were not comparable as the arsenal was completely government-owned and the Federal plant was privately owned, and the prices cited were for different models of tanks. The contract price for the Sherman tank at the Fisher arsenal was high, $67,173, but was undergoing substantial reduction to bring it into line with cost data resulting from actual production experience. The Fisher arsenal soon proved to be one of the lowest-cost producers in the whole program. The other medium tank producers, holding cost-plus-fixed-fee contracts, showed estimated costs per Grant tank ranging from $58,850 at Baldwin to $67,860 at Lima. The fixed fees on these contracts ranged from a low of $2,860 per tank at Lima to $3,850 at Baldwin. The cost figures in these contracts—but not the fees—were clearly estimates and had no binding effect. They were generally based on toolroom production experience only, and the assumption was that they would be revised later as assembly line production brought the cost down. The companies with cost-plus-fixed-fee contracts were eventually reimbursed for all approved costs incurred in producing tanks, plus the stipulated fee for each tank. With both CPFF and fixed price contracts the problem of keeping costs down was extremely complicated and continued so to the end of the war when final adjustments were made.57
The underutilization of plants was one of the most striking features of tank production all during 1943 and early 1944 “It has been evident,” wrote a Bureau of the Budget representative, “that too much production capacity was provided. ...”58 The sixteen plants placed under contract in 1941-42, when tank requirements were sky high, continued in operation until the last quarter of 1943 when four were eliminated—Lima, Pullman - Standard, Ford, and Pacific Car and Foundry.59 These cancellations reduced over-all capacity from approximately 8,000 to 6,600, but capacity was still more than double the
rate of production. In 1944, four more plants stopped building tanks—American Car and Foundry, Baldwin, and Marmon-Herrington in April, and Fisher in October. By the end of the year capacity was down to about 4,000—half what it was at the start of 1943—but still more than double the rate of production.60
The excess of production capacity during 1943–44 eliminated some problems and created others. By and large, shortages of machine tools and materials—the two major bugbears of 1941–42—eased considerably in the spring of 1943. So did the manpower problem in the tank industry, although there were always difficulties in some areas. The output of spare parts rose rapidly in relation to complete vehicles.61 Among the new problems was that of arranging for the orderly transfer of plants and workers to other war production. Similarly, the accumulation of surplus parts and raw materials was becoming a problem at plants where schedules were cut back. Efforts were made to divert such matériel to other tank manufacturers, but a report in the spring of 1944 on one plant that had been closed for months showed that millions of dollars’ worth of critical matériel—guns, gun mounts, cable, tools, motors, welding rods, and so on—was still on hand.62
Production declined during each of the first five months of 1944, but invasion of Europe in June reversed the trend. The rise in output during the latter half of the year was gradual, for, in spite of excess capacity, manufacture of specific models could not be increased overnight—or even over a period of several months—to meet sudden increases in theater demands. In May 1944, for example, the ASF Requirements Division declared that restudy of the tank picture had shown that, “we should push at once for as many additional medium tanks as we can get in 1944.”63 The change in calculations stemmed chiefly from two factors: an unexpected increase in the overseas replacement rate, and the large proportion of old tanks counted as resources in the February 1944 supply program. After conferring with Ordnance officers in Detroit, the head of the ASF Production Division, Brig. Gen. Hugh C. Minton, reported that from three hundred to five hundred additional Sherman tanks could be produced in 1944 but only “by applying all possible pressure to the producers.”64 Ordnance was immediately directed to apply the necessary pressure, but, because of changes in design and the need for retooling, production rose slowly.65 The measures taken by General Campbell to increase output included personal visits to the tank plants, publicity, recruitment of labor, and “every other known means of stimulating production.”66 To break the bottleneck that developed in production
Remanufacturing light tanks M3A3 at Red River Arsenal, Texarkana, Tex.
of wide tracks, the “special directive treatment” was resorted to, giving wide tracks an overriding WPB priority that put them in the same class with landing craft, heavy artillery, and the Dukw.67 Meanwhile, as theater commanders called for more and more tanks, Ordnance was instructed to launch a supplementary program for tank “remanufacture.”
This new process—the complete overhaul of combat vehicles—was, in the summer of 1944, a natural solution to the problem of how to increase the supply. During the preceding months, as one armored unit after another had been shipped overseas, the tanks they had used during long months of arduous training were withdrawn and replaced by new vehicles, with the result that large numbers of used tanks accumulated. Occasional criticism of the Army resulted when irate taxpayers saw these tanks standing idle in storage and concluded that they represented waste of valuable war matériel. As early as the summer of 1943 General Campbell, during a trip to the West Coast, had explored the possibility of contracting with industry to recondition these tanks. In October 1943 he formally recommended to ASF that a reconditioning program be authorized, including a balanced withdrawal of tanks from troops for this purpose. It was not until June 1944, when the quantities of tanks left behind by troops going overseas had reached high levels, that Ordnance was assigned the job of overhauling and modernizing them so they could be shipped overseas in
new-tank condition.68 The process was to start with a thorough cleaning of each tank and partial disassembly, followed by an overhaul of its engine, replacement of worn tracks, reconditioning the guns, adding improvements made since the original design, and giving the whole vehicle a new coat of paint. The total cost of remanufacture was estimated to be about half that of building a new tank. After teams of Ordnance technicians visited Army Ground Forces camps to select the tanks to be overhauled, the work began in August and proceeded at the rate of six hundred medium tanks per month for the rest of the year. Light tanks, half-tracks, gun motor carriages, and scout cars were also remanufactured, bringing the total for 1944 up to more than eleven thousand combat vehicles. The bulk of the work was done at the Quad Cities Tank Arsenal, Evansville Ordnance Plant (converted from production of small arms ammunition), and the Montreal plant of the American Locomotive Company.69
By January 1945, the demand for more tank production became intense and industry was once again called upon to make an all-out effort. Requirements for all types of tanks went up from 18,000 to 22,000, and then to 25,000 in February, including nearly ro,000 heavy Pershing (M26) tanks armed with 90-mm. guns or 105-mm. howitzers.70 In February, Ordnance reported that even though all tank producers were scheduled to capacity the procurement goals could not be reached. General Hayes, chief of the Industrial Service, stated that the tanks required for 1945 could be produced only by bringing back into production all former tank producers—a slow process that would not yield any results until near the end of the year, and would be prohibitively costly. Considering the two most critical components—Ford engines and torquematic transmissions—General Hayes observed that it was “extremely doubtful” if production could be further increased in 1945 by any means.71
By March the storm had subsided. As the defeat of Germany appeared more and more imminent, tank requirements were cut back and manufacturers received cancellation notices. The planned expansions at various plants were halted. From the peak of 2,268 tanks accepted in March, production declined to about 1,800 in both April and May and then dropped to 456 in July. By November it had stopped altogether.72
Shift to Heavy Tanks in 1944–45
Of all the various shifts and trends of tank production in the later war years, both in Germany and the United States,
Heavy Tank M6, mounting a 3-inch gun, produced by Baldwin Locomotive Works, is inspected by (from left) Lt. Col. David N. Hauseman, Brig. Gen. Gladeon M. Barnes, William H. Harmon (Baldwin official), and Capt. Arthur J. Seiler, December 1941.
none was more important than the coming into its own of the heavy tank. And none illustrates more clearly the problems faced by Ordnance in developing new matériel under pressure and meeting rapidly changing requirements. To see the full picture of changes in U.S. requirements we must look at two separate stages in the history of heavy tanks: the 1940-42 period, and the eleventh hour demand for heavy tanks in 1944-45.73
During the 1930s no one had shown much interest in heavy tanks, but in the spring of 1940 Ordnance was authorized to proceed at once with development of a 50-ton tank mounting a 3-inch gun. An appropriation was soon made to build fifty tanks; a contract for their manufacture was placed with the Baldwin Locomotive Company in August 1940; and the pilot model was finally unveiled in a public ceremony on the day after Pearl Harbor.74 The M6 heavy tank, as it was called, had a cast hull between three and four inches thick, a 925-horsepower engine; and weighed over sixty tons. It still needed a lot of development work to improve its suspension, transmission, brakes, and other
vital parts, but time was at a premium in the winter of 1941–42—and the President’s program announced in January called for building five hundred heavy tanks in 1942 and five thousand in 1943. Two models of the tank were quickly standardized and Ordnance contracted with the Fisher Tank Arsenal as well as Baldwin to build them at the combined rate of 250 per month.75
The Army Supply Program of September 1942 took the steam out of this ambitious plan by cutting heavy tank requirements from more than 5,000 to only 115. Ordnance immediately canceled its contract with Fisher and curtailed production at Baldwin. In December 1942, on the first anniversary of Pearl Harbor, General Devers of the Armored Force declared that, because of the M6’s great weight and limited tactical use, the Armored Force had no requirement for it and recommended that its production be stopped.76 The British later agreed, apparently because the M6 had been intended for North Africa, where bridges were not a problem, and by early 1943 the end of the North African campaign was in sight.77 In the summer of 1943 the Armored Force Board reported, on the basis of service tests, that the heavy tanks M6 and M6A1 were not acceptable because they lacked firepower commensurate to their weight, had obsolete fire-control equipment, were equipped with unsatisfactory transmissions, and had awkwardly arranged crew compartments.78 As a result, only forty heavy tanks of the M6 series were built, nearly all of them in 1943. The Germans, meanwhile, were throwing the bulk of their tank-producing capacity into building the heavy Tiger (63-ton) and Panther (47-ton) tanks, in spite of their many mechanical deficiencies.
While the M6 tanks were running into trouble, Ordnance was attempting to give its medium tanks more punch by equipping them with the 76-mm. high velocity gun and the 1 05-mm. howitzer, and improving their suspensions and tracks. It was also trying to win acceptance for the T-20 series of new and more powerful tanks to replace both the M6 and the Sherman, but the Army Ground Forces strongly opposed these efforts.79 Then in June 1944, after a demonstration at Aberdeen before high-ranking War Department officials, Ordnance finally won approval of its plan to mount a 90-mm. gun on the experimental medium tank T26 and reclassify it as a heavy tank.80 This marked
Table 20: U.S. Tank production, 1944–1945
| Type | 1944 | 1945 |
| Total | 17,565 | 11,968 |
| Light Tanks | a 4,043 | b 2,801 |
| Medium Tanks | c 13,468 | d 6,793 |
| Heavy Tanks | e 54 | e 2,374 |
a M5’s and M24’s.
b All M24’s.
c Nearly all M4’s.
d All M4’s.
e Nearly all M26’s.
Source: Summary Report of Acceptances Tank-Automotive Matériel 1940-45, by OCO-D.
the beginning of a new cycle in heavy tank production, inspired largely by the appearance during 1943 in Africa and Italy of heavy German tanks that were more than a match for the American mediums. Thus, long after the M6 series went by the board, the T26 was permitted to take its place in 1944, but only ten of these 45-ton heavyweights were delivered during the first half of 1944. After extensive tests and modifications, the new tank went into limited production in November 1944 and was standardized early in 1945 as the heavy tank M26. Nicknamed the General Pershing, it was probably the best heavy tank to see action in World War II. But it did not arrive in Europe until after the worst of the fighting was over. Only fifty were built in 1944, and, all told, only seven hundred were built before Germany surrendered in May 1945.81 (Table 20)
The Balance Sheet
Arguments about American tanks in World War II will no doubt continue as long as veterans of that conflict survive to continue the discussion. They will continue because the subject is so involved, with so much to be said on all sides, that no simple analysis can encompass the whole. To draw up a balance sheet fairly representing the views of all concerned, and weighing every factor in due proportion, is extremely difficult, if not impossible. But at this point some of the essential data may be mentioned on which there is likely to be general agreement.
Had the war been fought with light tanks the U.S. Army would have been as well equipped at the start as any army in the world. The M3 light tank that emerged in 1941 from the experimental work of the 1930s gave a good account of itself in North Africa, and its successors, the M5 and M24, were well received in the field. But the war was not fought with light tanks. Medium and heavy tanks predominated, and the U.S. Army had no first-rate medium or heavy tanks on hand in 1940. When the 1940 campaign in
Table 21: Comparative table of German, British, and American tank production, 1940-1945
| Date | German a | British b | American c |
| Total | 24,360 | 24,803 | 88,410 |
| 1940 | 1,459 | 1,399 | d 331 |
| 1941 | 3,256 | 4,841 | 4,052 |
| 1942 | 4,098 | 8,611 | 24,997 |
| 1943 | 6,083 | 7,476 | 29,497 |
| 1944 | 8,466 | 2,476 (6 months only) | 17,565 |
| 1945 | 988 (1st quarter only) | – | 11,968 |
a USSBS—Tank Industry Report, cx. A.
b Great Britain, Central Statistical Office, Statistical Digest of the War (London: His Majesty’s Stationery Office, 1951), Table 126.
c Whiting, Statistics, Table PR-7.
d Includes procurement from 1 July 1939 to 31 December 1940.
France revealed the need for tough-skinned, hard-hitting tanks, Ordnance had to rush through a drastic redesign of its existing medium model and begin work on a new heavy tank. Both the Grant medium and the M6 heavy were hasty improvisations that brought little credit to Ordnance, in the eyes of the Armored Force, and were regarded even by Ordnance engineers as makeshifts. The M6 never got into volume production, and the Grant was soon replaced by the Sherman—a well-designed vehicle that could hold its own with any medium tank on the battlefield. The Sherman was more mobile and mechanically more reliable than German medium tanks, and had greater flexibility and rapidity of fire. The trouble was that it too often found itself up against heavy Panthers or Tigers.82
The lack of heavy tanks to match the German heavies was the crux of the problem. In summing up the reasons for this lack, two items are reasonably clear: (1) the using arm’s lack of interest in heavy tanks during the prewar and early war years, and (2) Ordnance’s failure to come up with something better than the M6 in 1941. The two items are closely related and mutually supporting. Because the using arm expressed no desire for heavy tanks in the prewar years Ordnance made no effort to carry on the elaborate design and development work such tanks required. Because Ordnance in December 1941 could produce no heavy tank better than the primitive M6 model, the Armored Force was more than ever convinced that heavy tanks were
impractical and could be dropped. Meanwhile the Germans, who in 1941 had no tank comparable even to the M6, launched a heavy tank program after their encounter with the Russian heavyweights. The result was that the Germans in 1944 were able to oppose American Shermans with heavy tanks which, though far from perfect, had much tougher protective armor and more powerful guns than did the Sherman. Only then was Ordnance given a green light to proceed with production of the Pershing—too late to have much effect on the fighting in Europe.
The best way to take the quantitative measure of U.S. tank production in World War II is to view it in comparison with German and British production. Qualitative considerations aside, table 21 clearly reveals the extent to which the United States outproduced Germany. In every year except 1940 the United States produced more tanks than Germany and in the middle years of the war turned out from five to six times as many. (Table 21) Even Britain, while subjected to intensive bombing, produced almost as many tanks as Germany did in 1940, and in 1941 Britain turned out more tanks than either Germany or the United States. These figures should serve to demolish some of the myths that have grown up around German tanks. They should demonstrate for all to see that German tank successes were due more to skilled tactical use, and the employment of heavy German tanks against Allied mediums, than to any failure of American industry to produce in quantity the tanks desired by the using arms.