Get best of the week

Gears of war: when mechanical analog computers ruled the sea


Advanced Gun System (left) was created as a replacement for 16-inch guns of battleships (right). Apart from the GPS-guided missiles, the AGS fire control digital technologists perform the same task as the Iowa Rangekeeper Mark 8 battleships, they only have less weight and fewer people work with them.

The latest Zumwalt destroyer, currently undergoing acceptance testing, has a new type of naval artillery on board: the Advanced Gun System (AGS). The automated AGS is capable of firing up to 10 precision shells with rocket acceleration per minute at targets at a range of 100 miles.

These shells use GPS and an inertial guidance system to increase the accuracy of the gun to the circumference of a possible error of 50 meters (164 feet). This means that half of these GPS-guided projectiles will fall within this distance to the target. But if you remove the fancy shells with GPS, then the AGS and its digital fire control system will become no more accurate than the mechanical analog technology, which has almost turned a century.

I mean electromechanical analogue fire control computers like the Ford Instruments Mark 1A Fire Control Computer and Mark 8 Rangekeeper. These machines could continuously and in real time perform calculations with 20 or more variables long before digital computers made their way into the sea. When I served aboard the Iowa battleship in the late 1980s, they were still in use.

During my lifetime, several attempts have been made to combine or replace these legacy digital systems. Noteworthy was one of them (Advanced Gun Weapon System Technology Program), which looked like an AGS projectile with a range of 100 miles: an 11-inch dart-shaped projectile with GPS and inertial guidance, enclosed in a detachable 16-inch case (pallet), capable of due to the large caliber of the guns of the battleship, flying almost the same distance without rocket acceleration.

So why did the Navy take the path of "digitalization" of large guns of battleships? I asked this question to retired Navy captain David Boslow- Former Director of the Navy Tactical Embedded Computer Program Office. If anyone knows the answer, then this is Boslow. He played an important role in the development of the Navy Tactical Data System, the forerunner of modern Aegis systems, the forefather of all digital sensor and fire control systems.

“Once, my committee was commissioned to study the prospects for modernizing the fire control systems of Iowa class battleships from analog to digital computers,” says Boslow. “We found out that digitalization of computers will not increase either the reliability or accuracy of the system and issued a recommendation not to make changes.” Even without digital computers, the Iowa could shoot with deadly accuracy 2,700-pound (1,225 kg) "stupid" shells about 30 miles with a diameter of a probable error of 80 meters. Some battleship shells had a larger diameter of the lesion.

But how was the box with gears, cams, racks and pins capable of performing ballistic calculations in real time based on differential equations with dozens of variables? How did a colossus with a Volkswagen Beetle manage to hit a target beyond the horizon? And why did these metal and grease devices outperform digital systems for so long? Let's start with a short excursion into the history of ballistics of battleships and training films of the Navy, demonstrating the process of operation of analog computers.

Along the trajectory


Shooting a gun from a ship is not an easy task. In addition to the usual problems that ballistics encounters - calculating the power of a shot, aiming altitude, wind correction and the Coriolis effect - the fact is added that the firing is from a platform that constantly changes pitching pitch, yaw and position. If you are lucky and the target is motionless, then, due to the number of variables, this is still comparable to trying to hit the target with a ball of water sitting on the back of a jumping kangaroo.

Shooting at targets in the radius of view of a ship is a feedback loop. We aim, calculate the relative motion of the target and other ballistic conditions, shoot, see where the projectile hit, and adjust the parameters. Shooting targets beyond the horizon is even more difficult. An observer is required, giving accurate geographical coordinates and corrective fire, depending on where the projectiles hit.

In the era before the invention of gun turrets, ships fired guns from the sides. The adjustment was mainly carried out depending on the place where the shells hit and waiting until the side looking at the enemy did not lift upward. But with the advent of dreadnoughts and battlecruisers at the beginning of the 20th century, the range and lethality of the ship’s guns increased significantly. However, now they needed much more accuracy.

This need was consistent with the development of analog computers. Mechanical analog computers have been used for centuries by astronomers to predict the location of stars, eclipses, and moon phases. The very first mechanical analogue computer known to us is the antikythera mechanism , dating back to about 100 BC. But until recently, no one had any idea of ​​using computers to kill people.

To perform the calculations, analog computers use a standard set of mechanical devices - devices of the same type that convert the torque generated by the car engine to the rotation of the wheels, the movement of valves and pistons. Data in analog computers is "entered" continuously, usually by rotating input shafts. The mathematical value is tied to one full rotation of the shaft 360 degrees.

In the days of the ancient Greeks, data entry was performed by turning the wheel. In more modern analog computers, the sensor data variables — speed, direction, wind speed, and other parameters — were transmitted via electromechanical connections: synchronization signals of gyrocompasses and gyroscopic gyro-verticals, tracking systems, and speed sensors. Constants, for example, elapsed time, were introduced by special electric motors at a constant speed.

To turn the shafts into a continuous set of calculation output, I connected them all together a set of gears, cams, racks, pins and other mechanical elements that convert motion into mathematical calculations using geometric and trigonometric principles. Also, “hard-set” functions were produced that store the results of more complex calculations in their precision-made forms. When working together, these details instantly calculated a very accurate answer to a specific set of questions: where will the target be when it reaches a huge bullet that I push out of a 68-foot (21 meter) rifled barrel, and where I need to aim so that it goes there hit?

With perfect assembly, analog computers can answer such questions much more accurately than digital computers. Since they use physical and input data rather than digital, they can describe curves and other geometric elements of calculations with an infinite level of resolution (however, the accuracy of these calculations depends on the manufacturing quality of parts and decreases due to friction and slippage). At the same time, no less significant digits are discarded, and the answers are given continuously and do not depend on synchronized for-next calculation clocks.

Coding in metal


The most fundamental part of any mechanical analog computer is its gears. Using combinations of gears of different types, an analog computer is able to perform such simple mathematical functions as addition, subtraction, multiplication and division.

Gear ratios - the use of two gears having a specific ratio of circumference - this is the simplest way to perform calculations using mechanisms. They can be used to increase or decrease input or output values, or to apply constant multipliers of input data to other calculations. For example, if you rotate a shaft, the ratio of which to another shaft is 2 to 1, then the output shaft will rotate half as many times.

Rack-and-pinion transmission systems, such as those used in driving a car, are also used in analog computers to convert rotational motion into linear output data; they geometrically move read data or components to solve other types of calculations in a ballistic task.

You can understand how similar gear systems worked in analog computers, from a fragment of the 1953 training film on the Navy dedicated to fire control computers:


Shafts and gears of a fire control computer.

The gears of the differential of cars are designed so that the wheels in turns rotate at different speeds. But in analog computers, they perform a different function: they provide the ability to perform mechanical addition and subtraction. A set of differential gears installed between two input shafts with the same gears will always make turns, which are the mathematical average of the turns of two input shafts; if we multiply this average by two, we get the algebraic sum of two input values. For example, if one input shaft turned three times forward and the other turned once forward, then the differential gears rotate the shaft connected to them two times, that is, half of their sum - four.


, .

All this is wonderful when it comes to simple mathematics. But for functions of a higher level, for example, for calculating the curves of a ballistic trajectory or the influence of the Coriolis effect on long-flying projectiles, analog computers require more complex details. Some of these functions can be performed by cams - rotating surfaces made in such a way as to “store” answers for a range of values. Simple cams can store the range of responses depending on one variable, for example, turning the input rotation into trigonometric or logarithmic output data using a pin connected to the rail. More complex three-dimensional drum cams can store responses to complex functions with two variables, such as rotational volume calculations. An example is shown in this movie clip:


Cams are stored functions of analog computing.

All these components were well known to the creators of the first astronomical calculators, however, the method of their manufacture could not provide accuracy even close to the accuracy that tools of the industrial era can achieve. But there is another mechanical component that combines everything you need for the complex calculations required to predict the position of a target in ballistic computing: an integrator. This device uses various rotational speeds of the rotary disk, used as a continuously adjustable differential gear.

The integrator, first developed by Professor James Thompson of Belfast in 1876, was perfected by his brother Lord Kelvin as an element of a “harmonic analyzer”.


“Harmonic Analyzer” by Lord Kelvin with disk integrators.

Lord Kelvin used a harmonic analyzer to isolate various factors that influence tidal patterns so that they can be predicted in the future. The computer received two input values: time was represented as rotation at a constant speed, and the tide height was monitored from the recording using a mechanical needle. Ropes and pulleys generated output by drawing a curve on a paper roller. The British Navy fell in love with Kelvin's tidal computer because it allowed it to collect historical tide data recorded anywhere in the world and then create tide tables in much less time. More than half a century later, Lord Kelvin's tidal computers helped plan the Allied landings in Normandy.thus making a direct contribution to the outcome of World War II.

Apart from improvements that improve their reliability in harsh marine environments, the fire control computers used until the late 1990s, in fact, functionally remained the same as those used by Lord Kelvin. They are shown in the video below. Hannibal Ford, who developed the Rangekeeper and Mark 1 fire control computers, invented this improved integrator, which used a pair of balls in the running gear to transmit rotation information from the turntable.


The disk-type integrator, similar to that used in the Mark 1 fire control computer, is similar in function and design to Lord Kelvin's integrator.

Computer network (fire control)


The “fire control systems” of the First World War were mostly separate devices connected by people shouting out information by telephone and intercom. The only data that entered Rangekeeper Mark I automatically was the heading of the ship transmitted by the gyrocompass repeater. The situation changed in the next decade, when the fleets of the world better mastered with a new product called "electricity".

The Washington Maritime Agreement of 1922 for nearly a dozen years limited the further development of the fleet, but throughout the 1920s Ford continued to improve its Rangekeeper, culminating in the 1930 Rangekeeper Mark 8. Mark 8 became the pinnacle of fire control systems of large naval artillery. This system was used on Iowa class battleships and controlled 16-inch guns of all four vessels from the moment they were put into operation during the Second World War until the bombing of Iraqi forces in February 1991 during the Persian Gulf War.


The central artillery post of the Missouri battleship battery, which housed the Rangekeeper Mark 8 and its analog computing equipment. Wall-mounted switchboards made it possible to switch towers and guns controlled by the system.

Rangekeeper Mark 8 also provided operators with the ability to manually enter data in the event of a connection failure with sensors; in addition, they could modify the data based on observation of shots and make other adjustments. The machine could even work without electricity due to the manual rotation of the flywheel. The bearing of the target and the distance to it now came in the form of electrical input from an artillery fire control device. The speed of the ship was transmitted automatically based on the data of its speed sensor, and the wind speed - directly from the anemometer.

After “pointing” the system to the target, Mark 8 transmitted signals to the gun turrets and installations through the switchboard to maintain their correct aiming, and then sent stabilization data to adjust the elevation of the guns in accordance with the yaw and trim of the pitching of the vessel. Mark 8 itself had an electromechanical network. It consisted of five cases of analog computer equipment, fastened together in a single module.

Mark 8 was designed for large guns, which, because of their size and rate of fire, were used only for shelling surface and ground targets. Smaller guns, such as the 5-inch 38-caliber twin mounts on the Iowa and many smaller warships of the Second World War era, should have been able to aim at faster and smaller targets in three dimensions - simply put, in airplanes. This required much more complex calculations, which led to the creation of the crown of electromagnetic analog computing: the Ford Instruments Mark 1 fire control computer.


The Mark 1A Fire Control Computer is the processing power of 3,000 pounds of aluminum alloy.

The Mark 1 weighed more than 3,000 pounds (1,360 kg). Like Rangekeeper, he received input from artillery fire control devices - “turrets” with an electromechanical drive and optical sensors (and later radars) that continuously transmitted information about the bearing and distance through electrical synchronization signals.

The computer took into account the displacement between the control device and the instruments it controls. He also needed to calculate the burning time of mechanical fuses, so that the shell exploded near the target. (However, there were several instances in practice shooting in the 1980s when the Iowa directly hit a towed aerial target, albeit unintentionally.)

Mark 1, considered the most accurate anti-aircraft computer during the war, still had some pretty serious limitations. To explode shells near air targets, he used mechanical fuses and was able to perform calculations for air targets moving at a speed of less than 400 knots relative horizontal and 250 knots relative vertical speed. Because of this, he was ineffective against jet aircraft and kamikaze attacks.

Goodbye gears



The Mark 48 computer for “coastal attacks” is an electrical analog system with electromechanical input data. He had a light table for cards projecting a position and target data from below.

So why are we even moving away from using these mechanical masterpieces in aiming and undermining targets? Despite its high accuracy, mechanical analog computers had limiting factors. They are heavy and take up a lot of space. Even when they became more automated, they still needed a large staff. The torque required for their operation, including all servo drives that convert electrical signals into rotation, required a lot of electricity - 16 kilowatts at peak load.

And despite their overall reliability, the most serious enemies of electromechanics are friction and mechanical fatigue. Providing sufficient lubrication and monitoring the wear of the gears of the fire control computer is a much more serious task than a visit to the nearest car service to change the oil. In addition, there is the problem of “reprogramming” an analog computer. If you want to change the range of the input they receive or change the output so that they take into account the new variables, then this will be like rebuilding a transmission.

For most applications for which analog computers were created, this is not a problem. Over the past century, fire control variables have not changed much. The advent of jet aircraft and the need to provide longer-range bombardment of ground targets led to a new cycle of innovations in analog systems that lasted until the mid-1970s: electrical analog systems.

These electronic computing systems were not digital and performed the same functions as gears with cams, but in the form of analog electronic components. However, electronic parts were easier and easier to maintain than full-scale mechanical systems, and allowed integration with mechanical systems using signal outputs similar to synchronization signals used to integrate other sensors into a common system.

During World War II, Bell Labs developed the first fully electronic fire control computer, the Bell Mark 8. Although it was never put into operation, parts of its technology were combined with the Ford Mark 1, known as Mark 1A. An advanced system helped track and aim at faster aircraft.

Mark 1A and Rangekeeper Mark 8 also received additional electrical assistance in aiming at ground targets during the Korean War. Coast Attack Computer Mark 48It was designed specifically for conducting "indirect fire" - firing at targets that the ship could not see, based on information from a spotter plane, reconnaissance spotter or (from the late 1980s) from a Pioneer drone. He used the existing fire control system to aim at a known reference point (usually a relief element indicated on the map). Also, to determine the location of the ship, he could use radio or satellite navigation. Based on the ship’s location and the transmitted target location, the Mark 48 calculated the initial fire control data by transmitting Rangekeeper or Mark 1A data depending on which guns were used to bombard the unfortunate target.

Outdated systems


Four Iowa-class battleships were the only ships to receive the Mark 48. For the rest of the fleet, the transition to digital fire control systems began in the mid-1970s, as ship designers began to strive to create lighter ships with more emphasis on hunting for submarines and flying apparatus than shooting at other ships.


In the photo - the author of the article in his youth, when he was a naval officer aboard the battleship "Iowa" in 1988. The photo was taken next to the armored citadel on the bridge, located below the fire control system instrument, part of which was Rangekeeper Mark 8.

In 1987 and 1988, I served on board the Iowa in the deck crew, nominally responsible for 125 foremen and unskilled sailors. Many people from my division served the second gun turret or one of the ship’s 5-inch gun batteries, so my interest in their device was not idle at all. I often crawled along the shell decks of the gun turret, making sure everyone was in the right places.

During my stay on board, we fired more shells from the 16-inch guns of the ship than the Iowa shot during the entire Korean War. And despite all the experiments to add digital technology to the gun system, the one sensor installed right before I got on board made the guns more accurate than they ever were. This is a Doppler radar sensor capable of detecting the velocity of a projectile upon departure from the barrel.

The radar was installed after the urgent commissioning of the New Jersey battleship (Iowa type) in the early 1980s, when it encountered serious problems with precision guns during the Beirut crisis.. The problems were mainly related to the fact that the powder charges in the bags used on board the ship were mixed and their explosive profile changed.

By accurately measuring the velocity of the projectile upon departure from the gun during the first shot with a specific amount of powder charges, the personnel of the fire control could understand what it would be like with other shots, and change the input speed data for the computer accordingly. I personally saw a couple of times examples of this accuracy aboard the Iowa, including at night gun exercises off the coast of Puerto Rico near Vieques. The gunners ideally hit metal targets with idle training shells, and I could even see a few miles from myself sparks flying apart when hit.

Ultimate evidence of battleship accuracy came during the Gulf War, when Missouri and Wisconsin used Pioneer drones as spotters for attacks on Iraqi artillery batteries and bunkers. It was after the Missouri bombardment that the Iraqi forces on Failaka Island surrendered to a drone launched from Wisconsin, linking its low span with an imminent bombardment.

The real end to analogue fire control came not because of its accuracy, but because of commonplace dollars and cents. For the funds that need to be spent on bringing the Iowa into the sea, the Navy could equip ten Zamvolts, which, moreover, could take a double supply of fuel compared to battleship tanks. In the 1980s and 90s, the Navy spent a lot of time justifying the continued use of battleships, despite their cost, trying to use technologies such as the Advanced Gun Weapon System Technology Program or testing powder charges with more power. The explosion on board the Iowa in 1989, allegedly caused by spontaneous combustion of gunpowder made in the 1930s, put an end to such experiments.

It is ironic that analog computing technology continues to exist at Zamvolte as part of its fire control system. Electronic analog computers are part of a radar station with a phased array antenna, which provides aiming missiles Zamvolta. However, from the point of view of old naval veterans, a control computer cannot be real if it does not have servos.

More articles:


All Articles