Monday 5 July 2021

Kingmaker of American Tank Building

American tank history is strange in some ways. For instance, Creighton Abrams is the best known American tank ace, and few know about tankers who scored more tank kills than he did. The situation with American tank designers is even stranger. Compared to Soviet tank designers, they are like ghosts. The only well known names among them are John Walter Christie and perhaps Joseph Colby, director of the Tank-automotive Center and a key figure in American wartime tank building. Even many specialists would not recognize the name of the man who effectively created the American wartime tank. Harry Austin Knox was better known as a car designer. However, he was also a talented tank designer who had an effective monopoly on American tank chassis from 1935 to the mid-1940s.

First wave motorist

It so happened that the first wave of tank builders came from either those who built tractors or cars. The hero of this article belonged to the latter. Harry Austin Knox was a talented car designer, one of the first who transformed the car from a motorized carriage into the dominant method of transportation. Harry Knox was born on January 19th, 1875, near Westfield, Massachusetts. John Knox, Harry's father, was a plumber, but Harry went a different way. America was already caught up in the automobile trend by the time the junior Knox enrolled in the Springfield Technical Institute. The first American car with an internal combustion engine was built in 1891, and by the time Knox graduated in 1894 there were quite a few enthusiasts who were attempting to get into this business. Knox ended up as one of them.

Harry Knox in the passenger's seat of the four-wheeled Overman Victor, 1897. Like in Knox's other designs, the engine was located under the hull in the front of the chassis.

Knox did not initially intend to build the entire automobile. His passion was internal combustion engines, and not just ones for cars. Knox built his first four-cylinder engine in 1894, and soon sealed his first deal. The young engineer managed to gain the cooperation of the Overman Wheel Company (Chicopee Falls, Massachusetts). The idea was simple: Overman already produced a three-wheeled Victor bicycle. Knox's engine would be installed on this vehicle. Later Overman Wheel Company even tried a four wheel model. Alas, the bicycle maker was a victim of the economic decline of 1896. Overman Wheel Company finished the year with tremendous losses, which led to its bankruptcy. Albert Overman tried to get involved with car building himself without much success, or the support of Harry Knox.
Knox's patent for an engine. Similar horizontal engines were used on early vehicles of the Knox Automobile Company.

Seeing how poorly things were going with Overman Wheel Company, Knox took a risk. In 1898 he formed his own company in Springfield with $50,000 starting capital: Knox Automobile Company. He had not only the means, but experience. What was no less important: he had a successful technical solution in hand (Knox didn't forget to patent his inventions, which allow us to see the extent of his engineering genius).

Thanks to patents we can see what Knox was building, especially in the period when he switched to tanks. In 1898 Knox invented a very simple and successful internal combustion engine. Instead of four bulky cylinders, Knox's invention had just one. The key feature was the cooling system. A series of rods were screwed into the engine block that acted as cooling fins. This gave the engine the appearance of a porcupine (it was indeed nicknamed "old porcupine"), but allowed for effective cooling. This was the engine used in Knox's first car.

A typical early Knox car. In this case this is a model 1901 delivery van. Harry was already trying to enter the commercial vehicle market.

Knox built his first car in 1899. They were initially assembled at the Elektron Company building located near the Knox's college. Later Knox purchased the neighboring building and set up the Knox Automobile Company there. Knox's first cars had three wheels. The single cylinder 3 hp engine accelerated them to a decent 48 kph. The first three cars built in 1899 proved successful, and 53 were built in 1900. Knox moved to four-wheel designs in 1902, but the overall concept was the same. The hull was attached on top of the chassis, but the engine was under the hull rather than in it. The car had a chain driving the rear axle and a two-speed planetary gearbox. The result was reliable, simple, and cheap. A two-man car cost $750 with another $100 for the removable top, a fair price for that time.

Design of a Knox car.

In 1901 Knox started to experiment, feeling out the commercial vehicles market. He built a delivery van, although the three-wheeled chassis was delaying his expansion, and so he transitioned to four-wheel designs in 1902. His business was growing, and he built about 500 cars in 1902. The variety of products was growing: by 1902 he had not only a single cylinder 10 hp engine, but a 2-cylinder one. Knox successfully took part in races. Driving the Knox Model R he won the New York - Boston - New York race in the fall of 1902. This car survives today. By that time his cars could hold four people. Instead of an engine, the front had a collapsible seat for two.

"Waterless" engines were Knox's main feature. However, after his departure from the company, it quickly transitioned to water cooled engines.

Work was going well, and Knox paid more and more attention to the commercial vehicles market. Trucks entered Knox's catalogue in 1903. The transition to water cooled engines already began, but Knox stood by the benefits of air cooling. His commercials stressed this fact, marketing the engine as the Waterless Knox. The situation began to change and in 1904 Knox left his own company. He decided to focus on trucks. Meanwhile, Knox cars gained first four-cylinder engines (with them, a large hood), then water cooling. Knox's cars continued to participate in races and gain prestige, which killed the brand in the end. Production of Knox cars ended in 1914. 

Bet on trucks

As mentioned above, Harry Knox was becoming more and more interested in trucks. This led to arguments between him and his business partners, who preferred developing lighter cars with the idea of transitioning into luxury cars. Many shared this opinion, and in the end the market was oversaturated. Henry Ford came out with his Model T in 1908. It was initially not seen as a threat, but Ford's bet on a cheap and simple car paid off. Meanwhile, the market for commercial vehicles was ripe for the taking. While cars replaced light carriages rapidly, cargo was still largely transported by horse.

The first Knox fire engine appeared in 1905.

Knox founded a new company in Springfield in 1904: the Knox Motor Truck Company. Its primary product would be trucks, but Knox also expected to build light cars. His companions once again opposed him, and so another brand was born: the Atlas Motor Company. Like Knox it gradually moved towards the luxury car market and in 1909 one car competed for the Vanderbilt Cup with Elmer Knox at the wheel. Knox had poor luck with cars and Atlas Motor Company shut down in 1913. Its descendant, the Lyons-Atlas Company, produced rear axles and other components for cars.

Knox's trucks went a different way. In 1905 Knox built fire engines, later his spectrum of products grew. In the early 1910s Knox competed directly with a future competitor in tank building: John Walter Christie. Christie was also a racing enthusiast, but his racing career almost ended with is death and his cars weren't doing well either. However, he found success in a line of motorized tractors designed to tow fire engines previously pulled by horses.

The Knox Martin Tractor Model 31, nicknamed Snake. This was a popular and successful tractor in its time.

Christie faced a dangerous competitor in 1911: the Knox Martin Tractor Model 31. This was a rather strange looking but successful three-axle machine. It could be used both as a tractor and connect to a trailer, either a specially made one or one that used to be horse drawn. The Knox Martin Tractor 31 had a wide spectrum of applications. It pulled fire engines and lumber trailers. The maneuverable tractor was nicknamed Snake and became a big problem for Christie's business.

Knox Model 35, Harry Knox's first creation for the army.

A brand new vehicle appeared in 1915: the Knox Model 35. This was a four wheeled vehicle, but the front and rear wheels had dual tires. The 7 L engine put out 40-60 hp, lifting 10 tons. This was just the nominal value, in reality the Knox Model 35 could carry up to 40 tons. World War I had already broken out, and Knox's creation did not go unnoticed by the military, especially the French, who needed artillery tractors. A second purpose soon turned up for Knox's tractor: carrying tanks. The American military was also interested and found common ground with Knox. In 1918 he sold his company and began working for the Ordnance Department. Knox Motor Truck Company experienced a decline which only accelerated in the early 1920s, leading to it shutting down in 1924. This year coincidentally marked Knox's rise as a tank builder.

An alternative to Christie

At first Knox built what he knew best for the army: trucks. The American military wanted to have its own trucks, including all wheel drive ones. Various work in that direction was done in the early 1920s by many engineers, for instance Arthur Herrington, the same one who went on to form the Marmon-Herrington company. However, the situation was changing by the mid 1920s. The drop in orders after the end of the war led to stagnation in a number of fields. Private companies managed to strangle the army's design projects, after which Herrington departed for the private sector. Knox was nearly 50 by 1924, but he finally found a new direction himself: tank building.

A patent for a tank suspension dated 1925. This is likely how the Light Tank T1 looked at first.

By the time Knox started building tanks American tank building was in a difficult spot. The aforementioned Christie who in 1924 had a row with Major General Clarence Williams (the head of the Ordnance Department from 1918 to 1930) was not always to blame. He built nearly 20 types of tanks and SPGs for the department from 1917 to 1924. Some of them were quite good, better than both foreign analogues and vehicles developed in the United States. However, the military often didn't know what it wanted. Despite his hot-headedness and opinion that he knew better than his customer, Christie delivered fighting vehicles that were superior to those of his competitors. Most tanks designed by other companies were a sad sight to behold. The best thing the American military managed to do was put a copy of the Renault FT into production under the index M1917. 952 tanks of this type were built in total and they were the backbone of the American tank force for a long time. Attempts to build a faster version of the M1917 were unsuccessful. There were also a hundred Mark VIII International tanks built for a British order.

Medium Tank M1924, the initial state of the Medium Tank T2.

The American military began to look for an alternative to the Christie M1919 in 1919. The idea of building an American analogue of the Medium Tank Mk.D, a product of British engineer Phillip Johnson, seemed promising at the time. The tank managed to reach a speed of 30 kph on tracks, faster than the Christie tank on wheels. This time work wasn't contracted out, but conducted internally by the Technical Department of the Tank Corps at the Rock Island Arsenal. This resulted in the Medium Tank M1921, the first tank developed directly by the American army. The tank was also assembled at the Rock Island Arsenal. However, development dragged on for too long and the tank had a great number of drawbacks. Development of a medium tank was taking a lot of effort, but the army also wanted a light tank on top of that. Harry Knox's debut as a tank builder took place at this difficult time.

The T1 self propelled chassis designed to test track links.

Knox didn't forget to patent his inventions even while working for the Ordnance Department. Each patent stated with the disclaimer that the US government could use his work without payment of royalties, same as with any other Ordnance Department employee. Judging by the patents, KNox initially worked on tracked and halftrack tractor chassis, but proposed a tank suspension in the fall of 1925. This is around the time that the Technical Department began working on a light tank. This was a tank with a classical layout with a mass of about 5 tons and a crew of two. The tank had a few features in common with the Medium Tank M1921, unsurprisingly.

Officially, work began in February of 1926, and work on a medium tank with a mass of about 15 tons began in March. The Rock Island Arsenal was the center for tank development, but not for long. On September 1st, 1926, the Technical Department was moved to Fort Meade, Maryland, where the Tank Department was located. The Ordnance Consulting Committee (a part of the Society of Automotive Engineers, SAE). The SAE deserves a separate introduction. This organization formed in 1905 and became the basis for the rapid militarization of American production in the first years of WWII.

The first Light Tank T1. The caption calls it Light Tank M1, but that title didn't stick around for long.

The result of these discussions was the development of a new light tank concept that was named Light Tank T1. The Ordnance Department and SAE came to the conclusion that a rear fighting compartment was preferable. There were already examples of this layout: the Medium Tank Mk.I and Mk.II. As a result, work on a light tank took a new course on September 15th, 1926, and Knox effectively became the lead engineer of the project. Work on a special chassis, the "self propelled tracked chassis T1", was also launched to test different types of tracks. Test chassis T2 was designed to test suspensions. Both chassis had a front engine, but the T1 had a transmission in the front and the T2 had it in the rear.

A patent for the Light Tank T1E3. This is proof that the tank was developed by Knox, and that James Cunningham Son & Co. were merely subcontractors.

The concept of the Light Tank T1 was finalized by mid-March of 1927. James Cunningham Son & Co. enter the picture around the same time. A contract with the company was signed on April 12th and a prototype was built by August 1st. There is a misconception that the Light Tank T1 was designed by James Cunningham Son & Co., but that is not the case. The tank was developed and patented by Harry Knox. This is also true for the aforementioned tracked chassis, suspension, and other elements.

The tank was far from ideal. The driver gripped the engine between his legs, and visibility was poor. The commander, doubling as the gunner and loader, sat on the gearbox. The Light Tank T1 looked like an anachronism in general. However, looks are one thing, but function is another. The M1917 needed refurbishment after 260 km of driving, while the Light Tank T1E1 drove for 3232 km over 57 days without serious issues. The top speed was 28 kph and the average speed was 16 kph, a phenomenal result for the late 1920s. No other light tank manufactured at the time could reach this speed. It is not surprising that infantry command, namely Major General Robert Allen, insisted that the tank be standardized as the Light Tank M1. This index was only used from January 14th to March 30th, 1928.

Trials of the Medium Tank M2 resulted in disappointment.

The Ordnance Department pushed to accept the tank into service. A contract for 250 tanks was at stake. This was when Christie struck back. The Convertible Medium Tank M.1928 departed from Fort Meade on a test run on November 28th, 1928. This tank was a test bed built by the US Wheel Track Layer Corporation. The tank's designer referred to his creation as the M.1940, suggesting that this tank was the way of the future. The tank's main feature was the independently sprung suspension, today known as the spring suspension or Christie suspension. The tank travelled the same route that the Light Tank T1E1 did, but its Liberty L-12 engine allowed it to travel with a much higher average speed: 45 kph. The top speed was even higher: 68 kph on wheels and an unthinkable 112 kph on wheels. The Light Tank T1E1 needed trucks to travel long distances, but the Christie tank did not. Putting the Light Tank T1E1 into production was no longer considered.

Comparative trials at Fort Benning showed the advantages of Christie tanks, but this was only the beginning.

Work began on an improved version, the Light Tank T1E2, but this tank also had problems. The suspension was overloaded. Despite Knox's efforts, Christie was the clear leader in two aspects. As mentioned earlier, work on a medium tank began in March of 1926. This tank, named Medium Tank T2, was essentially an enlarged Light Tank T1. Since the light tank was a priority, work on the Medium Tank T2 began only in 1929 and it was built in 1930 by the same James Cunningham Son & Co. The 14.2 ton tank had the same engine as the Christie tank, although reduced in power.

The T2 had better armour and armament, but a few months later the Convertible Medium Tank M1931 made it look like an anachronism. This was especially true for the suspension, which prevented it from reaching its designed speed of 40 kph. Despite powerful pressure from the Ordnance Department, namely Captain John Christmas (future Major General and a key designer of all American tanks during WWII), Christie had won that round. A contract to build five Christie M.1931 tanks was signed between The US Wheel Track Layer Corporation and Ordnance Department on March 25th, 1931. The order was expanded to seven vehicles three months later, on June 12th, 1931. As further events showed, this was victory in the battle, but not the war.

A well deserved monopoly

Christie celebrated victory too soon. First of all, the financial crisis hit the army's budget hard, and the order for 250 tanks would remain a dream. Secondly, even if the money existed there is serious doubt that Christie would have been able to take on such an order. The Soviet commission that visited Christie's workshop noted that he was only capable of producing small scale orders. Thirdly, even the Convertible Medium Tank T3 had its drawbacks. This was a two-man tank with very slim modernization potential. The infantry was satisfied with the results of the trials, as was the cavalry. Four out of the seven Christie tanks went to Fort Knox where they were indexed Combat Car T1 (the cavalry was not supposed to have tanks). In 1932 infantry command composed requirements for an improved tank: the Medium Tank T3E2. This was a four-man tank with a larger turret and other changes. Christie thought that he knew better than the American military, but his gamble did not pay off. The contract for five tanks worth $200,000 went to American LaFrance, another former competitor of Christie's on the fire engine market.

Combat Car T2, Knox's poorest design.

The Ordnance Department didn't waste time either. It remembered that Christie sold the right to use all of his patents, including future ones, in 1924 for $100,000. As soon as it turned out that the spring suspension worked well, several teams inside the Ordnance Department began working on Christie tanks. This was the start of the tank building career of Joseph Colby, a key American tank builder. Gladeon Barnes, another key figure in American tank building, also got his start on Christie style tanks. Barnes is better known as a talented artillery designer, but he started working on tanks in 1930. Finally, Knox also took up this new direction.

This was the initial concept for the tank. The Combat Car T2 was quickly forgotten.

A rather strange vehicle indexed Armored Car T5 and quickly renamed to Combat Car T2 made its way to trials in September of 1931. This vehicle is often called "Christie style", but it had nothing in common with his tanks. The suspension was built with leaf springs and only the front wheels had some manner of coil spring. This was a bad design from the start. Work on the Combat Car T2 lasted for two years after that, and it went down in history as the worst of the Combat Cars. It had no future at all, and lost not only to Christie but to the Combat Car T4 developed under the direction of Gladeon Barnes. It was Barnes, not Christie, who became Knox's rival.

Another suspension concept, this one with British roots. It was not built in metal.

In parallel with a leaf spring suspension, Knox worked on coil springs. His coil spring suspension had nothing to do with Christie either. Its inspiration came from across the ocean. The British were responsible for some of the most progressive tank technologies at the time, chiefly Vickers-Armstrongs. Due to the situation that formed in the early 1920s, the British mostly worked on the export market. IT was these export tanks that first used the horizontal suspension that later inspired Tomio Hara's suspension used on Japanese tanks. Knox also tried a similar suspension, and like Hara he used fairly large road wheels. This idea went nowhere, but Knox still patented it.

It was not enough that Knox copied the Vickers Mk.E concept and its suspension, but he also patented it.

The situation with copying another tank, the Vickers Mk.E, went in a different direction. This was the best light tank in the world at the time of its creation, but it was rejected by the British military. Their myopia is confirmed by the history of this vehicle. The Vickers Mk.E and its foreign derivatives became the most numerous pre-war tanks in the world. The level of influence John Carden and Vivian Loyd's tank had on tank building worldwide is enormous. This tank also had an influence on American tank building. One Vickers Mk.E Type A was delivered to the Aberdeen Proving Grounds in July of 1931 under the pretext of trials. The British made a huge mistake. The tank was closely studied and various components were copied wholesale. Knox even patented them! 

The Light Tank T1E4, a dead end for the T1 program but a huge step forward for American tank building.

The Americans developed their own version of the Vickers Mk.E called Light Tank T1E4. It was converted from a Light Tank T1E1 at the Rock Island Arsenal. This tank was not very good and further work didn't do much to improve its characteristics. However, this tank was a step in the right direction. The driver's compartment now housed two men instead of one. The assistant driver later also received a machine gun. Observation ports in the front and the sides improved visibility. The layout of the tank was also good, but it still had to be polished. "Help" came from above: the War Department limited the mass of the light tank to 6804 kg. The Light Tank T1 program was buried only to be reborn.

The Light Tank T2 was essentially a revised Light Tank T1E4.

Requirements for a new tank were composed on June 3rd, 1933. It weighed 6300 kg and had a crew of 4, a top speed of 48 kph, and bulletproof armour. It was armed with two Browning M1919 machine guns plus one high caliber Browning M2HB. Two versions were built at the same time: a light tank for infantry and a combat car for cavalry. The infantry's Light Tank T2 was built conservatively with a suspension similar to the one used on the Light Tank T1E4/T1E6. This tank had a two-man turret. The Combat Car T5 had two turrets, but more importantly a new suspension. The Vertical Volute Spring Suspension (VVSS) was radically different from what Knox used earlier. Fairly large road wheels were coupled into bogeys. A volute spring acted as the suspension element. This was a compact design with long travel.

Harry Knox's most important invention: the VVSS suspension.

Trials showed that putting a different suspension on two tanks was the correct approach. THe two tanks were very similar, but while the Light Tank T2 accelerated up to 43.4 kph, the Combat Car T5 accelerated to 68.8 kph. The calculated speed was exceeded by a third. The British suspension was only holding it back. There were, however, complaints about the metallic track with a metal-rubber joint designed by Gladeon Barnes.

The Combat Car T5 on trials. The tank was much more agile than the Light Tank T2, but the tracks needed replacing.

The new American tank concept evolved from the improved Combat Car T5 and Light Tank T2. This tank had a rear engine and front transmission. As a rule, the engine was a converted aircraft radial engine. A long driveshaft ran along the floor of the tank to the gearbox. 

Knox's second most important invention: the rubber-metallic track link. Variants of this design are still used on tanks.

The running gear was Knox's design. This included the suspension, the road wheels, and finally the tracks. In 1933 Knox created a successful rubber-metallic track with end connectors. This concept is used to this day, and even tracks of modern MBTs are descendants of Knox's design. The resulting chassis was simpler and cheaper than Christie's system. The top speed of the Light Tank M2A1 and Combat Car M1 was 72 kph, the same as a BT tank on wheels. It's not surprising that the Americans stuck to this concept for a decade.

All vehicles in this photo were built on Harry Knox's chassis. The tanks that were built during WW2 were descendants of these designs.

Harry Knox earned a practical monopoly on tank chassis as of 1935. There was no lobbying involved, it just so happened that the 60 year old engineer reached an optimal design. There were many attempts to create an alternative, but Knox's suspension was simply better. The bogey suspension made room to grow. This resulted in the Medium Tank T5, which evolved into the Medium Tank M2, then the Medium Tank M3, and finally Medium Tank M4. Evolution of the light tank led to the Light Tank M3 and Light Tank M5. Even Marmon-Herrington, who started with their own tank suspension, ended up using Knox's design. This suspension was also used on the Heavy Tank M6

The tally of all the vehicles built on Knox's suspension shows how important his contribution was to tank building worldwide. Even all the Christie tanks put together pale in comparison to Knox's vehicles. There was also a significant difference between the careers of the two engineers. Christie was responsible for only one production tank (produced in a series of seven) and barely made ends meet, Knox occupied an important post in the Ordnance Department and had a direct impact on any tank developed with his suspension. Unlike Christie, Knox didn't frequent the front pages of newspapers and magazines, he merely did his job.

The initial design of the HVSS suspension that replaced the VVSS in 1944.

Nothing remains forever. By 1943 it was clear that the VVSS suspension was reaching the limits of its development. The Americans also familiarized themselves with the torsion bar suspension. It was first tried in the 1930s with no results, but now they had the Pz.Kpfw.III and KV-1 as samples. Gladeon Barnes promoted his own suspension as of 1942, but Knox also didn't sit still. He started on an alternative suspension in 1942: the Horizontal Volute Spring Suspension (HVSS). Initially this suspension was meant for tractors. Knox didn't invent anything radically new. The Hotchkiss H 35 had a similar system, and even that design was beaten by Sidney Horstmann. Nevertheless, the HVSS suspension was not a clone, but an original design.

The HVSS suspension in production.

American tanks gained the torsion bar suspension in 1943. Early variants were nothing but a source of headaches, meanwhile the HVSS suspension worked much better. It proved to be the most reliable suspension tested on the experimental Medium Tank T25. The suspension was also successfully tested on the Medium Tank M4 and approved for installation on all M4 family tanks. The T26 still got a torsion bar suspension, but it took a long time to get it working, meanwhile the M4 continued to use it until the end of production.

The last known patent by Harry Knox. Tracks like these were used by American tanks for a long time.

Knox's star began to set only in early 1945. Nevertheless, he continued to work in the Ordnance Department, creating new running gear elements. Tracks that he developed were used by American tanks and SPGs for a long time. Knox retired in the late 1940s, well into his 70s. Unlike Barnes and Colby he never received a general's rank, retiring as a colonel. Harry Austin Knox died on June 2nd, 1957 at the age of 83. His death was not an event, and the obituaries recalled only his contributions to car building. Knox's enormous contribution to tank design remains unknown to many. Knox's influence is comparable to that of Heinrich Kniepkamp, the man who defined the concept of German tanks for nearly 20 years.

1 comment:

  1. Many thanks for this study on a neglected subject!