Atmospheric Railways: The 19th Century Trains That Ran On Air

Dec 22, 2018 0 comments

The 19th century ushered in a new form of transport—railways. Journeys that previously took weeks were now completed in days. Distances that could only be covered in days on horseback or on horse-drawn wagons were now being covered in a matter of hours. All this was possible because of the steam locomotive, the hulky and bulky “iron horse” that could pull a dozen wagons behind it and cover great distances without fatigue. It was the locomotive that made railways a grand commercial success. Yet, railways had barely penetrated the British countryside when some railway pioneers began to wonder if trains could be made speedier and more efficient by getting rid of the locomotive altogether.


Illustration of Dalkey Atmospheric Railway - Arriving at Kingstown.

It was a fallacious reasoning but not without merit. Locomotives are incredibly heavy, heavier than the wagons they are pulling. Much of the energy generated by the locomotive is utilized to pull itself. So if the source of power could be offloaded to a static location, such an engine could be more robust and with more available space be potentially more powerful. Another limitation with locomotives was the poor coefficient of friction of metal wheels on metal tracks. This limited the amount of incline a train could climb which constrained the building of railroads in hilly terrain. Lastly, steam locomotives produced an awful amount of smoke. It was like a travelling coal furnace billowing smoke all over the countryside. People hated it and it gave the general impression that trains were dirty.

London-based inventor George Medhurst was one of the first to float the idea of using compressed air to move goods and passengers through cast iron pipes. This was before the first steam locomotive was even built. Medhurst proposed two methods by which this could be achieved. In one method he proposed using a tube through which a piston travelled up and down dragging the train the piston was attached to. Pumps ahead of the train would draw air out of the tube creating a partial vacuum and the air pressure behind the piston would push it forward. In another method, he suggested that the passenger train itself could act like a piston travelling through a much larger tube.

Not much happened during Medhurst’s lifetime, but a few years before he died, in 1824, a man called Vallance took out a patent and built a short demonstration line consisting of a 6-feet diameter cast iron tube through which a vehicle moved by air pressure supplied through the ends of the tube. Although Vallance's system worked, it was not adopted commercially.

Several engineers created competing prototypes based on Medhurst’s piston method since it needed little modification to existing railroads, although it did present a tough technical challenge. The piston in the pipe had to be attached to a passenger car by an external arm, meaning the tube had to have a slit running its entire length. Keeping this gap sealed to preserve the negative and positive air pressure on either side of the piston, while allowing the arm to pass unimpeded, posed a problem.


Schematic diagrams from Samuda and Clegg’s patent of atmospheric railway. 


Diagram explaining the different parts of the Samuda and Clegg system. Photo credit

It was a trio of engineers—brothers Jacob and Joseph Samuda, and Samuel Clegg—who successfully overcame the challenges. They devised an elaborate zipper like system made of overlapping leather sealed with beeswax and animal fat that allowed the arm to pass through and close immediately behind it to maintain the pressure. If you are wondering how efficient this system could be, in two full-scale models the engineers built at Southwark and Wormwood Scrubbs, Samuda and Clegg demonstrated that a 9 inch vacuum pipe evacuated to about half an atmosphere provided enough force to propel an 11-ton train at 36 km/h.

In 1840, the first “atmospheric railway“ line was built in Ireland, as an extension of the Dublin and Kingstown Railway, linking Kingstown to Dalkey located some three kilometers away. A steam-engine powered vacuum pump at Dalkey pulled trains weighing up to 70 tons up the ascent, achieving speeds of up to 64 km/h, while the return journey was achieved by gravity.

The success of the Dalkey Atmospheric Railway attracted the attention of many eminent engineers of the day. A similar system was laid on a section of the line from Paris to St Germain crossing the Seine River in 1847, and remained in use for thirteen years until the development of more powerful steam locomotives led to its abandonment.


Brunel's Atmospheric Railway remains at Didcot Railway Centre.

Another atmospheric railway was installed between Forest Hill and West Croydon, in south London—a distance of about 8 kilometers. Unfortunately, the railway link, a part of the London and Brighton Railway, did not enjoy the same success as the atmospheric railway in Dalkey did. There were frequent mechanical breakdowns and delays and the public began to lose faith in the new system. The leather valve system gave particular trouble. That summer of 1846 was unusually hot and dry, which caused the tallow and beeswax compound that was supposed to seal the joint to melt and become runny and unable to keep the leather flaps closed. Besides, the tallow, which is animal fat, attracted rats who nibbled at the leather and got immediately sucked into the pipe once the pumps were turned on. Then came the winter, which made the leather flap stiff and snow got into the tube. The following winter, the atmospheric system was abandoned. The line had remained operational for less than 18 months.

Many of the problems encountered on the Croydon line also plagued the South Devon Railway built by one of the most prolific engineers of the era, an Irishman named Isambard Kingdom Brunel. A 20-km section of the line between Exeter and Newton Abbot used atmospheric propagation instead of locomotives. Pumping stations were installed every 3 miles, and they worked in relays. When a train was due in a section the station of that section began to pump out the air, and then ceased to pump when the train had passed. Strangely, there was no telegraphic communication between the stations, so the pumping stations relied on a strict timetable. Any delays therefore meant that a station began pumping too soon wasting coal in the process. Eventually, the cost of running and maintaining the railway became too great, and this combined with other technical difficulties killed the atmospheric rail service. The line was replaced by regular steam locomotives.


The Aeromovel shuttle at Salgado Filho International Airport. Photo credit: Trensurb/Flickr

Atmospheric railway saw a revival in late twentieth century. A Brazilian company called Aeromovel Corporation installed a driverless atmospheric train service in a theme park in Jakarta, Indonesia, in 1989. It ran three trains in a 3-km-loop using the same principle of positive and negative air pressure developed by George Medhurst more than two hundred years ago, but with an improved and more robust seal than that designed by Samuda and Clegg. Another system was installed at Porto Alegre Airport, Brazil, transporting passengers over a 1-km-long route in 90 seconds.

More recently, a US-based company called Flight Rail Corp. has developed a new design for high-speed atmospheric trains where the piston is magnetically coupled to the passenger train instead of being physically connected. This allows the vacuum tube to be closed, thus preventing leakage—an issue that bedeviled earlier systems.

The company has successfully built a 1/6th scale pilot model operating at speeds up to 40 km/h, but the corporation claims that a full-scale implementation would be capable of speeds in excess of 300 km/h—which sounds plausible, considering that Brunel’s South Devon Railway was able to achieve a record speed of 103 km/h with leaky valves.

It appears the future of rail travel is by air.

More on Amusing Planet


{{posts[0].date}} {{posts[0].commentsNum}} {{messages_comments}}


{{posts[1].date}} {{posts[1].commentsNum}} {{messages_comments}}


{{posts[2].date}} {{posts[2].commentsNum}} {{messages_comments}}


{{posts[3].date}} {{posts[3].commentsNum}} {{messages_comments}}