Pneumatic Networks.

HOW IT WORKS

Left: the basics of pneumatic messaging The central exhauster sucks the message carriers through the pipes. In this case the whole system is at below atmospheric pressure, so a carrier can be sent on its way simply by pushing it into a tube. More sophisticated sytems used both pressure and vacuum to move carriers; see the work of Varley in PNEUMATIC DESPATCH IN GREAT BRITAIN.

Left: on the left a carrier arrives Its momentum and weight push open the discharge door, the carrier drops out, and air pressure closes the door again. On the right, a carrier is sent by pulling open the inlet door and allowing the carrier to be sucked in. The inlet door is then closed by its counterweight. Note that the carrier is wider at its ends, to prevent the central part binding against the tube wall in curves.

EARLY HISTORY
The first known work on pneumatic transport was done by the Englishman George Medhurst (1759-1827). During a period of a few years about 1810, he invented not only the pneumatic dispatch tube, but also the pneumatic railway, which is somewhat outside the scope of this page. Medhurst's invention pushed a capsule along a tube a few inches in diameter, using positive air pressure; thus was born the pneumatic dispatch tube.

In 1853 J. Latimer Clark installed a 220 yard pneumatic tube between the London Stock Exchange in Threadneedle Street and the Central Station of the Electric Telegraph Company in Lothbury. The Electric Telegraph Co had been up and running long before pneumatics came along; the company was incorporated in 1846. There were similar installations in Berlin (1865) between the Central Telegraph Office and the Stock Exchange, and in Paris (1866) to the Place de La Bourse. Of these, more later. It is important to remember that the pneumatic system was not a crude and early forerunner of the electric telegraph, but came after it; Cooke and Wheatstone opened the world's first commercial telegraph between Paddington and West Drayton, a distance of 13 miles, in 1838.

The various networks are described below in the order of the opening.

PNEUMATIC DESPATCH IN GREAT BRITAIN: OPENED 1853.

This is an edited and revised version of the information contained in the article on "Pneumatic Despatch" in the 1911 edition of Encyclopaedia Brittanica.

"Pneumatic Despatch is the name given to the transport of written despatches through tubes by the agency of air pressure. It was introduced in 1853 by J Latimer Clark, between the Central and Stock Exchange stations of the Electric and International Telegraph Company in London. These stations were connected by a tube 1.5 inches in diameter and 220 yards long. Carriers containing batches of telegrams, and fitting piston-wise in the tube, were sucked through it (in one direction only) by the production of a partial vacuum at one end. In 1858 C F Varley improved the system by using compressed air to force the carriers in one direction, a partial vacuum being still used to draw them in the other direction. This improvement enables single radiating lines of pipe to be used both for sending and for receiving telegrams between a central station supplied with pumping machinery and outlying stations not so supplied."

"In the hands of R S Culley and R Sabine the radial system of pneumatic despatch was in 1870 brought to great perfection in connection with the telegraphic department of the British post office, since that date the total length of tubes (which are employed for telegrams only) has been greatly increased. In 1909 there was in London a total length of 40 miles, whilst in all large and also in very many smaller provincial towns there are installations; these are constantly being added to, as it is found more economical to transmit local message-work by tube rather than by wire, as skilled telegraphists are not required, but only tube attendants.
In some cases only a single tube is necessary, but three or four, or even more, are in use in some towns, according to local circumstances. Short tubes, known as house tubes are in use in a great number of offices; such tubes, which are worked either by handpumps (when the tubes are very short and the traffic light) or by power, are usually 1.75 in. in diameter, and are used for the purpose of conveying messages from one part of a telegraph instrument-room to another, or from the instrument-room to the public counter."

Left: The Central Telegraph Office of the GPO in London, Oct 1932, showing an impressive pneumatic installation. Communications were still labour-intensive; there were 3000 employees. The tubes were used to send telegraph forms from local post-offices to the Central Telegraph Office, where the skilled telegraph operators were concentrated.

"The underground, or street tubes are mostly 2.75 in. in diameter, but there are also a number of 3-in, tubes in use; those in the large provincial towns (Birmingham, Bradford, Cardiff, Edinburgh, Glasgow, Grimsby, Liverpool, Manchester, Newport, Leeds, Newcastle, Southampton and Swansea) are 2.75 in. in diameter; but in Dublin, Gloucester, Lowestoft and Milford 1.75-in. tubes are employed. There are fifty street tubes in London, varying in length from 100 to 2000 yards. (Central Office to the Houses of Parliament), and also seventy-five house tubes; the pumps for the whole system are worked by four 100 horse-power steam-engines.
At Cardiff, Edinburgh, Gloucester, Leeds, Lowestoft, Newport, Southampton and Swansea the air-pumps are driven by electric motors; at Bradford and Grimsby gas engines are used, and at Milford an oil-engine.

"The tubes are in all cases of lead, the 2.75-in, tubes weighing 8 lb/foot, and being made in lengths of 28 ft; they are enclosed in 3-in, cast-iron pipes made in lengths of 9 ft.
Great care is exercised in making the joints in the lead pipes. Before the tube is placed in its trench a strong chain is passed through it, and a polished steel mandrel, 6 in. long and slightly less in diameter than the diameter of the tube, is heated and attached to the chain, and pushed half its length into the end of the tube already laid; the new length of tube is then forced over the projecting end of the mandrel until the tube ends (which have been previously cut flat) butt perfectly together; an ordinary plumbers joint is then made. By this means the tube is made perfectly air-tight, and the mandrel keeps the surface of the tube under the joint as smooth as at any other part of its length. After the joint is completed the mandrel is drawn out by the chain attached to it, the next length is drawn on, and the above process repeated. The tubes are laid about 2 ft. below the surface of the ground."

Left: A tube drop point at the Central Telegraph Office in the 1930s

"The tubes radiate from the central to the branch offices, the principal offices having two tubes, one for inward and the other for outward traffic. At the smaller offices both the inward and the outward traffic is carried on through one tube. The carriers are made with guttapercha bodies, covered with felt, the front of the carrier being provided with a buffer or piston formed of several disks of felt which closely fit the tube; the messages are prevented from getting out of the carrier by the end being closed by an elastic band, which can be stretched sufficiently to allow the message forms to be inserted. The 3-in. carriers will hold 75 ordinary message forms, the 2.75-in, carriers hold 25 forms, and the 1.5-in. carriers 20 forms. The carriers are propelled from the central office by pressure, and drawn in the opposite direction by vacuum, the standard pressure and vacuum being 10 psi and 6.75 psi respectively, which values give approximately the same speed."

"For a given transit time the actual horse-power required is much less in the case of vacuum than in the case of pressure working, owing to the density of the air column moved being much less: thus, for example, the transit time for 10 lb pressure is the same as for 6.75 lb vacuum, but the horse-power required in the two cases is as 1.83 to 1. A tube 1 mile long, 2.75 in. in diameter, and worked at 10 lb per square inch pressure, will have a transit time of 2.75 minutes, and will theoretically require 335 horse-power to be expended in working it, (Editor's Note: I think this must be a misprint, as it seems a quite excessive amount of power) although actually 25% more horse-power than this must be allowed for, owing to losses through various causes. The transit time for a 2.75-in, tube is 16% more than for a 3-in, tube of the same length, when both are worked at the same pressure, but the horse-power required is 50 % less; it is not advisable, therefore, to use a tube larger than is absolutely necessary to carry the volume of traffic required."

"The somewhat complicated pattern of double sluice valve originally used at the central stations has been superseded by a simpler form, known as the D box, so named Despatching from the shape of its cross section. This box is of and cast iron, and is provided with a close-fitting, Receiving brass-framed, sliding lid with a glass panel. This Apparatus, lid fits air-tight, and closes the box after a carrier has been inserted into the mouth of the tube; the latter enters at one end of the box and is there bell-mouthed. A supply pipe, to which is connected a 3-way cock, is joined on to the box and allows communication at will with either the pressure or vacuum mains, so that the apparatus becomes available for either sending (by pressure) or receiving (by vacuum) a carrier. Automatic working, by which the air supply is automatically turned on on the introduction of the carrier into a tube and on closing of the D box, and is cut off when the carrier arrives, was introduced in 1909."

"On the long tubes (over about 1000 yds) a modification of the D box in its simplest form is necessary; this modification consists in the addition of a sluice valve placed at a distance of about 9 in. (i.e. rather more than the length of a carrier) from the mouth of the tube. The sluice valve, by means of an interlocking arrangement, is so connected with the sliding lid of the box that the lid cannot be moved to the open position unless the sluice valve has closed the tube, nor can. the sluice valve be opened unless the sliding lid is closed. The object of this sluice valve is to prevent the back rush of air which would take place into the tube when the sliding lid is opened to take out a carrier immediately on the arrival of the latter; for although the vacuum may be turned off by the 3-way cock, yet, owing to the great length of the tube, equilibrium does not immediately take place in. the latter, and the back rush of air into the vacuum when the lid is opened to extract the carrier will cause the latter to be driven back into the tube. The sluice also prevents a similar, but reverse, action from taking place when pressure working is being carried on."

"As a rule, only one carrier is despatched at a time, and the second carrier is inserted in the tube until the arrival of the first one at the farther end is automatically signalled (by an electric apparatus) to the despatching office. On some of the long tubes a carrier, when it passes the midway point in the tube, strikes a trigger and sends back an electrical signal indicating its passage; on the receipt of this signal a second carrier may be despatched. This arrangement has been almost entirely superseded by a signalling apparatus which by a clock movement actuates an indicating hand and moves the latter to tube clear a certain definite time (30 to 40 seconds) after a carrier has beer inserted in the tube. By this arrangement carriers can be despatched one after the other at comparatively short interval~ of time, so that several carriers (separated by distinct intervals may be travelling through the tube simultaneously. It is necessary that the carriers be separated by a definite interval otherwise they tend impact each another and may become jammed."

Left: How the London system worked It is interesting to see that the system wholly depends on the use of the electric telegraph so the sender can tell the receiver a carrier is on the way, and so the receiver can tell the sender is carrier has arrived. This corresponds with the absolute block method of railway signalling, in which only one train is permitted in a block. It does not correspond with the methods described above; the clock method is equivalent to the time-interval system of train signalling, which caused many accidents and was eventually abandoned. At bottom right a man at the central station is receiving a carrier drawn to the station by the vacuum. It is not currently clear why there is both a vacuum main and a much smaller vacuum pipe; possibly the 'pipe' carried less vacuum so carriers would not be sucked too violently into the apparatus, endangering the operator's fingers. The original source of this fine drawing has now been identified as The Engineer, for 18th December 1899. Info courtesy of Tom Bates.

Left: The boilers at the General Post Office: 1899 Six double-furnace Cornish multi-tubular boilers are shown here. Each has a large weight poised above it, which presumably weighted the safety valve. The boilers were 6ft 6in in diameter and 20ft 2in long, with two flues 2ft 6in in diameter and 14ft 6in long, terminating in 74 tubes 3in diameter and 5ft 9in long. Each boiler is fitted with a single water-gauge (locomotive practice was to have two) a pressure-gauge and a clack-valve for boiler feed at the bottom of the vertical pipe. The valves with their handles tilted downwards are presumably the steam stop valves for each boiler. At the left is shown one of two 15HP beam-engines for feeding water to the boilers. The bevel gearing at extreme left drove two three-throw pumps which drew water from a well, hinted at in the bottom left corner. The text gives some details of the main steam engines. Image from The Engineer, for 18th December 1899, courtesy of Tom Bates.

Left: Vacuum and pressure pumps at the General Post Office: 1899 The vacuum and pressure pumps were installed in the basement, underneath the steam engines that drove them. Each pump could either draw from an 18-inch vacuum main, or force air into two 15-inch pressure mains, depending on the setting of the valves shown on the sides of the pump cylinders. The general air of steampunk nightmare is augmented by the menacing-looking horns; these were actually entry diffusers to reduce losses when the pumps sucked in atmospheric air. The handwheels, one of which is being clutched by an intrepid worker, determined if the flow of air was from atmosphere to the pressure main, or from the vacuum main to the pressure main. This picture (probably drawn from a photograph) shows the nearest main to not be in use as its branches are unconnected to anything and not blanked off. Image from The Engineer, for 18th December 1899, courtesy of Tom Bates.

Left: A section of one of the vacuum/pressure pumps at the General Post Office: 1899 The control valves can be seen between the mains and the pump cylinders. Note there is what appears to be a safety-valve on the right side of the pump cylinder. Image from The Engineer, for 18th December 1899, courtesy of Tom Bates.

Left: Hose connections between send/receive terminals and the vacuum and pressure mains in the central hall of the General Post Office: 1899 Once again a smaller 'vacuum pipe' is shown. Image from The Engineer, for 18th December 1899, courtesy of Tom Bates.

Left: The London system: from The Engineer, for 18th December 1899 The local engine referred to was at the GPO Foreign office at top right. Presumably it was a steam engine.

Left: The London system: 1948 The Post Office pneumatic tube system was badly damaged by enemy action in December 1940. The Central Telegraph Office (CTO) was practically destroyed by fire and all telegraph traffic was transferred to C.T.0. 'R' and other offices in London. In 1943 work on the reinstatement of the main tube system was commenced. In 1948 there were about seventy tubes in all ranging in length from approximately 300 yards for the tube to Headquarters Building, to approximately 4,500 yards for the two tubes to the Western District office. Source: The Post Office Electrical Engineer's Journal, for Jan 1948. Vol 40, Part 4

Left: The London system: 1948 The outgoing carriers are dispatched from the tube table shown here. Source: The Post Office Electrical Engineer's Journal, for Jan 1948. Vol 40, Part 4

Left: The London system: 1948 There was an innovation in the method of receiving incoming carriers. All the local tubes which carry 'up' traffic to this room from the pneumatic switchgear are terminated in a nest at high level so that carriers are ejected from the open ends of the tubes into a large hopper from which they then fall down a zig-zag chute to table level. With this arrangement, which is shown in Fig. 3, it is possible to bring all the carriers to one point so that they are within reach of one operator without the use of a conveyor. The incoming carriers fall down the chute to the right of the operator. Source: The Post Office Electrical Engineer's Journal, for Jan 1948. Vol 40, Part 4

Left: The London system: 1948 This automatic pneumatic switchgear was used to transfer carriers from the 'up' or vacuum street tubes to the local pressure tubes and from the local vacuum lubes to the 'down' or pressure street tubes; it also controlled the supply of pressure, or vacuum, to those street tubes which are worked alternatively in either 'up' and 'down' direction. The relays and other apparatus used were all of the standard 1948 Post Office telephone type.

Left: Compressor for the London system: 1948 New compressors and vacuum pumps were installed to move the carriers. Positive-displacement rotary compressor were obtained from Messrs Hick Hargreaves & Co Ltd, of Bolton. The compressors were driven by 80HP Crompton-Parkinson motors over a speed range of 300-600 rpm, and the vacuum pumps by 65-h.p. motors over the range 250-500 rpm. "The solid cast-iron rotor A is mounted eccentrically in the cylinder B and rotates in the direction of the arrow, carrying with it the steel blades C which slide in milled axial slots. These blades are kept in contact with the cylinder by centrifugal force during rotation of the rotor and sweep through the crescent-shaped space between the cylinder and rotor. The portion of this space between any two adjacent blades is at a maximum at the trailing edge of the suction port and is gradually reduced as the rotor turns. The air in this space is thus compressed until the leading edge of the delivery port is reached, when it is discharged at the higher pressure. The blades are not actually in contact with the cylinder wall but are held a few thousandths of an inch away by two restraining rings D. These rings are free to rotate in slots in the cylinder wall and are carried round by the blades. The space between the edges of the blades and the cylinder wall is sealed by a film of oil, the friction. between the blades and the cylinder wall being reduced appreciably by this device." Source: The Post Office Electrical Engineer's Journal, for Jan 1948. Vol 40, Part 4

Note the continuous oil supply to both bearings and the two ends of the cylinder. Two manual lubricators were provided for the restraining rings D. There is a pressure gauge and a non-return valve on the output. The non-return valve was to prevent the rotor from reaching an excessive speed in reverse if the power failed.

Left: Compressor for the London system: 1948 One of the compressors seen from the inlet side, paired with its electric motor. The oil pump is at the extreme right on the end of the shaft. Air was drawn in from the engine room through a filter of the Visco oil type and a silencer, and the compressed air in each header is passed through a cooler before distribution to the pneumatic tubes. Source: The Post Office Electrical Engineer's Journal, for Jan 1948. Vol 40, Part 4

Left: Air cooler for the London system: 1948 When air is compressed it gets hot, so it had to be passed through a cooler before distribution to the pneumatic tubes. The compressed air was passed through tubes about 2 inch diameter with external gills and cooling air was drawn from the engine room by a centrifugal fan and blown over the gilled tubes. The fan and driving motor can be seen in the foreground of Fig. 7 with the cooler housing behind. The compressed air passed four times through the cooling air stream and was cooled from about 180°F to 100°F for air at 12 psi, and twice through the cooling air stream while being cooled from about 140°F for air at 6 psi. The air coolers were manufactured by Messrs Hunt & Moscrop Ltd, of Apex Works, Middleton Junction, Manchester. Source: The Post Office Electrical Engineer's Journal, for Jan 1948. Vol 40, Part 4

Left: Used in the air cooler for the London system: 1948 Gilled tubes by Messrs Hunt & Moscrop Ltd, of Apex Works, Middleton Junction, Manchester.

Left: Compressor controllers for the London system: 1948 "The pressure-sensitive device on each controller, which is visible through the lower windows in the controller housing (see Fig. 8), is a small transformer, the primary of which can move relatively to the secondary. This primary winding is mechanically coupled to a bellows arrangement connected to the source of pressure and is loaded by a dead weight applied through a lever so that it stabilises at the required pressure. An increase in pressure causes the coupling between the primary and secondary winding to increase, and thus, with a constant voltage applied to the primary, the secondary voltage, over a small range, is proportional to the applied air pressure. This secondary voltage is rectified and compared with another unidirectional voltage and the difference, if any, applied to a sensitive galvanometer. When operated the galvanometer closes one or other of its two contacts according to the direction of the difference, and, by means of various relays and the contactors in the starter already referred to, corrects the speed of the associated compressor by causing a movement of the shunt field regulator." It all sounds a bit complicated. The control equipment for all the machines was supplied by the Watford Electrical & Manufacturing Co Ltd.

PNEUMATIC DESPATCH IN GERMANY: OPENED 1865
Berlin's pneumatic dispatch system began in 1865 and closed in 1976. It was a gigantic pneumatic post system 254 kilometres in total length at its maximum extent. (NB some sources put it as high as 400 km,which is quite a discrepancy) The first installation in 1865 consisted of two tubes, each 2835 feet long and 3.5 inch in diameter, running between the Stock Exchange and Central Telegraph stations in Berlin. It was implemented by Latimer Clark, working with the Siemens telegraph company. At its peak, Berlin had 27 tube lines in a 254km long system, connecting Berlin post offices. Carrier speed was up to 10 metres per second. In 1898 alone more than 2.3 million mail items passed through the tubes.