A reprint from

American Scientist

the magazine of Sigma Xi, The Scientific Research Society

This reprint is provided for personal and noncommercial use. For any other use, please send a request to Permissions, American Scientist, P.O. Box 13975, Research Triangle Park, NC, 27709, U.S.A., or by electronic mail to [email protected]. ©Sigma Xi, The Scientific Research Society and other rightsholders

www.americanscientist.org

2004 January–February

1

ENGINEERING

BOAT LIFTS Henry Petroski ater may run downhill, but boaters sometimes want to go against the current. Paddles and oars might be thought of as boat lifts of sorts, in that they can be employed to raise the elevation of a watercraft from the mouth of a river almost to its headwaters. Where the river ceases to be navigable, perhaps because of shallows or rapids or a waterfall, smaller boats can be lifted physically out of the water and portaged to deeper, calmer, higher waters. In some cultures, boats have been pulled upstream against a strong current by human trackers. Horses and mules have often been used to tow boats along rivers and canals. Sails have long harnessed the power of the wind to drive a boat or ship silently, even upwind and against the flow. Steam engines, gasoline motors, diesels and turbines, although noisier, have made river travel virtually independent of winds, tides and currents. Currents are usually of little concern in canals. However, being artificial waterways, canals can present changes in elevation that are virtually insurmountable by familiar means. Where conditions allow it, locks are frequently employed. A canal lock, the invention of which is said to date from the 15th century, enables watercraft to be raised (and lowered) a significant vertical distance (almost 30 feet in the case of the Panama Canal) in a relatively short horizontal distance— often barely longer than the craft itself. Several locks can be constructed in tandem to raise and lower boats and ships much larger distances in steps. (The three Gatun Locks in Panama raise oceangoing ships a total of 85 feet.) Such an arrangement of locks is understandably referred to as a staircase or a flight. Even locks as large as those in Panama can be gravity-fed, but they require great quantities of water, which in turn can require a great change in elevation for ships using the waterway. This is because more water must be impounded, requiring a taller lock. The design and construction of Gatun Dam, which impounds the water of the Chagres River, was a key factor in making the Panama Canal work.

W

Henry Petroski is A. S. Vesic Professor of Civil Engineering and a professor of history at Duke University. Address: Box 90287, Durham, NC 27708-0287. 18

American Scientist, Volume 92

However, it does take time for a ship to pass through the three-step lock system, which was built to accommodate the enormous vessels crossing the isthmus. When Locks Prove Lacking Before the railroads displaced them in the 19th century, canals formed the principal network by which goods were transported overland. Like the railroads, canals developed locally and regionally first, which presented challenges to their continuous use over long distances. Where two canals were separated by physical barriers, such as greatly different elevations or a range of hills, means had to be devised for moving the boats— or at least the goods on them—through the vertical distance or over the obstacle. The cargo could, of course, be unloaded from one canal barge, transported between the waterways by some other means, and then loaded onto new boats. Transshipment was, however, time-consuming and expensive. Transporting the entire boat over the obstacle was clearly much more desirable. One means of doing this was by inclined plane. Fully loaded canal barges could be cradled onto a rail car, much as a boat is loaded onto a trailer today. The rail car could then be pulled up the incline—typically by means of a pulley system driven by a stationary steam engine—to move it to the higher elevation, where it could be separated from the rail car and refloated. In the 19th century, John Roebling, who designed the Brooklyn Bridge, gained much of his early experience with the properties of wire (later used in suspension bridges) by manufacturing wire rope for, among other purposes, use on inclined planes. This material replaced conventional rope, which was more subject to wear and rupture, thus making the hauling more reliable—and speedy. For example, in England, an inclined plane that began operating in Foxton, in Leicestershire, in 1900, saved almost 70 minutes over the time it took for a boat to be raised the same distance through locks on the Grand Union Canal. Back in the 17th century, not far from Liverpool and Manchester, a situation developed that was at the same time unique and representative. The counties of Cheshire and Shropshire sat on exten-

© 2004 Sigma Xi, The Scientific Research Society. Reproduction with permission only. Contact [email protected].

sive salt deposits, which had been exploited since Roman times. Transport of salt and other minerals found in the region had long been by cart or pack animal over poor roads. The River Weaver offered a preferable alternative, but it was normally accessible only at high tide. Prompted by the needs of salt merchants, Parliament passed an act in 1721 to canalize the river in order to achieve more efficient and reliable regional transportation. In 1732, what came to be known as the Weaver Navigation was opened (navigation meaning simply “a river that has been engineered so that it is navigable”), and boats carrying as much as 38 tons could travel 20 miles upstream. In the meantime, pottery making had developed from a cottage industry into a major commercial enterprise in nearby Staffordshire, which was the site of deposits of the principal raw material—clay—and the customary fuel to fire kilns—coal. In time, ready sources of local clay became scarce, and the material had to be imported. Improved transportation links became necessary, and Josiah Wedgewood, among other pottery manufacturers, called for a canal connecting the River Trent, which empties into the North Sea, to the River Mersey, which empties into the Irish Sea. The 93-mile-long Trent & Mersey Canal was completed in 1777 and was hailed at the time as England’s greatest-ever work of civil engineering. Siphoning Off Business—with Water After the Weaver Navigation was established, it attracted the transportation trade of both the salt fields in Cheshire and the pottery works in Staffordshire, which over time came to rely on this waterway. The Weaver trustees, seeing their business threatened, had opposed proposals to build the Trent & Mersey Canal. With its completion, however, they looked to developing a link between the two waterways, by means of which they hoped to siphon off business. The Weaver Navigation and the Trent & Mersey Canal roughly parallel each other for about five miles in the vicinity of Anderton, where they are separated by only about 400 feet horizontally but by 50 feet vertically. Because locks were not thought to be practical for such a steep gradient, alternatives were sought. To facilitate the transshipment of goods, trams and chutes were constructed at a site that became known as the Anderton Basin. The chutes were especially effective for salt, which could be efficiently offloaded from canal boats to larger riverboats below. In time, cranes were also built at the Anderton Basin to haul freight readily from the river up to the canal. Such efficient means of transshipping made the Weaver Navigation the preferred route for getting goods between Anderton and Liverpool. By the middle of the 19th century, Anderton was a major locus for water traffic in the region. Still, transshipment of cargoes was costly and time-consuming, and a www.americanscientist.org

Figure 1. Originally completed in 1875 to connect the Weaver Navigation and the Trent & Mersey Canal, the Anderton Boat Lift was fully restored in 2002. (© 2002. Photograph courtesy of Chris Sykes, Shefield, United Kingdom.)

means of transferring fully laden boats between the two waterways was sought. Around 1870, Edward Leader Williams, Engineer to the Weaver Navigation Trust, after studying alternatives and precedents, conceived of a boat lift—essentially an elevator (or “lift” in British English)—operated by a hydraulic ram. Williams turned for advice to Edwin Clark, who had served as resident engineer for the Britannia Bridge, one of the most ambitious engineering projects of the mid-19th century. Williams’s idea developed into the Anderton Boat Lift, which Clark designed. Rather than having a single container into which a boat could be floated and lifted to the higher elevation, the Clark design employed a pair of tubs, or caissons, which by counterbalancing each other greatly reduced the amount of power needed to raise the weight of a boat, water and the container. The Anderton Boat Lift, completed in 1875, was a masterwork of Victorian engineering. It consisted essentially of a pair of identical gondolas—each 75 feet long, 15.5 feet wide and 9.5 feet deep—into which boats could be floated. Once a boat from the Weaver was inside, the gondola gate would close behind it. Whether or not the upper gondola had a corresponding boat would not affect subsequent operation, since any boat merely displaces its weight in water. With both gondolas closed, some water from the lower one would be siphoned out, making it lighter than the upper one. Since the hydraulic rams beneath the gondolas were interconnected, the descent of the heavier one would drive the ascent of the lighter one. Because of the difference in their weights, the gondolas would stop about six inches off local water level, and the final stage of the operation consisted of letting the appropriate amount of water out of or into the respective gondolas, after which the gates could be opened and the boats—complete with cargo—floated out. It took about three minutes to lift a boat from the Weaver to the Trent & Mersey—and simulta-

© 2004 Sigma Xi, The Scientific Research Society. Reproduction with permission only. Contact [email protected].

2004 January–February

19

James Gentles

Figure 2. Part of the Millennium Link, Scotland’s Falkirk Wheel connects the Forth & Clyde and the Union Canals.

neously to lower another. To lift just one boat through a staircase of five locks would have taken about an hour. Old and New The success of the Anderton Lift led Edwin Clark to design other boat lifts on the Continent, including one for wide barges on the Neufosse Canal in northern France. It operated until the mid-20th century, when it was retired as a “static monument.” The Belgian Canal du Centre has four boat lifts whose design was also influenced by Clark. They were put into service in the latter part of the 19th and the early part of the 20th century, and they remain active. All told, six major boat lifts are thought still to be operating in Europe. No engineering system is without problems, however. One of the Belgian lifts was recently damaged in an accident, and a modern boat lift was scheduled to replace it until it could be repaired. The Anderton Boat Lift was not without its problems, either. On one occasion, a hydraulic ram ruptured, and corrosion plagued the pistons, conditions that required frequent repairs that in turn necessitated interruptions in service. This led to the recommendation that the lift be converted from hydraulic to electrical operation, a changeover that was completed in 1908. The conversion involved the addition of a new superstructure, supported over the old by inclined columns (A-frames) and topped by prominent sets of pulleys over which wire-rope cables were attached to massive cast-iron counterweights. A routine safety inspection of the “Cathedral of the Canals” conducted in 1983 found the structure seriously deteriorated, at which time it was shut down. However, in the meantime, dedicated fundraising and commitment by the Trent & Mersey Canal Society and British Waterways (the authority responsible for the 2,000-mile network of canals and rivers that crisscross the kingdom) has led to the restoration of the boat lift to its 20

American Scientist, Volume 92

original hydraulic operation, a conversion that was completed in 2002. The now-nonfunctional pulleys remain in place, but the massive counterweights have been removed and rearranged nearby into a maze for the amusement of children who visit this tourist attraction. Pleasure boaters on the canal can now relive an experience from another era. That is not to say that boat lifts are simply preserved things of the past. Among the purposes of Three Gorges Dam, the gargantuan project on China’s Yangtze River, is to make it navigable for large cargo ships all the way to Chongquing. This is being achieved by creating a great reservoir behind the dam, of course, but the 185-meter-high dam also presents a great obstacle to shipping. A series of five Panama-scale locks will enable ships to be raised and lowered around the dam, but the process would naturally be frustratingly slow for the smaller craft that will continue to ply the Yangtze. To speed their passage, an alternative to the locks—a boat lift—is being incorporated into the design of the Three Gorges Dam. It will operate on much the same principle as the original Anderton lift, employing a pair of gondolas acting as mutual counterweights. The electrical power used will be generated by the dam itself. And Very New Another new, modern boat lift was recently completed in Britain, but it no more resembles the Anderton lift than a canal boat does a hydrofoil. The innovative system is the brainchild of the people at British Waterways. Most of the waterways they manage are in England, with a few reaching into Wales—most notably the Llangollen Canal, with its famous and dramatic twocenturies-old Pontcysyllte Aqueduct, which carries canal boats in leisurely fashion high over the Dee River Valley. Like Wales, Scotland has few canals compared with England, but one of the most famous is the

© 2004 Sigma Xi, The Scientific Research Society. Reproduction with permission only. Contact [email protected].

driving a tunnel under the Roman Antonine Wall so as not to disturb it. The centerpiece of the restoration was clearly a means of connecting the Union, which would end in an aqueduct, and the Forth & Clyde with a “stunning structure to mark the new millennium.” This naturally left out locks. According to the British Waterways engineering manager: We looked at overhead monorails, tilting water tanks, water cylinders, giant spoons that worked like a seesaw. We had wheels that ran at right angles, and inclined planes and vertical lifts. Eventually we picked on a wheel with four hanging gondolas, two at the top and two at the bottom, with quite a lightweight structure in between. And we envisaged a wheel with a rim, since that is the easiest position to drive a wheel from. Such a Ferris-wheel device became an “exemplar.” In 1999 a sketch of it was distributed to tenderers, with the hope that it would inspire “something much more visually exciting.” Unfortunately the ideas remained “too Victorian,” and the Waterways’ preferred contractor—Morrison Bachy Soletanche—was asked to go back to the drawing board. It did so, with the help of consulting architects and engineers. Tony Kettle, of the Edinburgh architectural firm of RMJM—the designers of the Scottish Parliament building—became involved in the project. Kettle saw the canal route as a spine across Scotland, which was broken at Falkirk. He said he “wanted the wheel to be a celebration of it being joined back together again.” He elaborated on his idea: Think of a spine and you think of elegant, organic structures like fish bones. So I began to develop the aqueduct as an organic spine-like form. I wanted it to seem to float through hoops. And I wanted the wheel to be a celebration of moving between this aqueduct and the basin below. This was done by giving it direction. The hooked leading edges on the arms are all about rotation. It was as much a sculptural idea as a piece of engineering.

James Gentles

Forth & Clyde Canal, which after being completed in 1790 was the first to enable overland passage from the North Sea to the Atlantic Ocean. The Forth & Clyde made it possible to travel between Edinburgh and Glasgow in seven hours by steamboat. The Union Canal, begun in 1818 and intended to transport coal to Edinburgh via an inland route, met the Forth & Clyde at Falkirk, about halfway to Glasgow. There the difference in elevation between the two canals is more than 100 feet, so a flight of 11 locks had to be negotiated, a process that took the better part of a day. The railroad was introduced into Britain shortly after the Union Canal was completed, making it no longer a critical link in the transportation system. A century later, the automobile further threatened the continued viability of canals. The locks connecting the Forth & Clyde and Union canals ceased major operations in the 1930s and were abandoned in the 1960s. At the same time, the demand for more and better highways in Britain struck the final blow. Newly built roads blocked the waterways in almost 30 places, “stopping the canals dead in their tracks and creating stagnation and pollution,” according to the chief engineer for British Waterways in Scotland. The situation began to be turned around in the 1962, with the establishment of British Waterways, whose mission is to restore inland waterways for leisure use and to preserve the canal system and its appurtenances for their historic value. Because the authority has control of canal rightsof-way, it was able to engage in joint ventures to install fiber-optic cable along the canals. Having up-to-date telecommunications links readily available to buildings on abutting property makes this real estate more desirable, which in turn promotes waterway use. The income from such ventures has provided British Waterways with a source of funds to restore, enhance and otherwise improve the image of the canal system. The largest project to reclaim a British canal is the Millennium Link, which is restoring the Forth & Clyde and Union between Edinburgh and Glasgow. This has involved constructing miles of new canal, repairing and replacing locks, building new road bridges and new aqueducts, and

Figure 3. Vessels enter this aqueduct from the Union Canal, and the Falkirk Wheel at the end lowers them more than 100 feet to the Forth & Clyde Canal in about 15 minutes. www.americanscientist.org

© 2004 Sigma Xi, The Scientific Research Society. Reproduction with permission only. Contact [email protected].

2004 January–February

21

I was asked to convert it into a mechanical object in collaboration with [the engineer]. By making the gondolas narrower and longer and using two instead of four as in the exemplar design, I could make the aqueduct much narrower and more elegant. It also meant I could surround the end of each gondola with a large cog and have a simple gear system for keeping the gondolas level. Although simple, the cogs were a difficult concept to understand and my daughter made a little Lego model of it to try it out. The engineers for the project were from Butterley Engineering, a firm older than the Union Canal and one that specializes in engineering solutions that are “bespoke,” that is, custom made. (Among Butterley’s best-known projects is London’s St. Pancras Station.) Kettle’s organic idea, complete with hooklike arms that are said to have been “inspired by a Celtic double-headed axe,” was turned into an engineering reality by, among other things, using bolted rather than welded joints. This minimized the weight of the arms and thus the possibility of metal fatigue as the wheel would go through stress cycles. Such considerations enabled the engineer and Butterley director Colin Castledine to declare that the structure “is built to rotate for the next 140 years.” The Falkirk Wheel, as it has come to be known, is a striking piece of architecture and engineering. The gleaming structure of the wheel is truly a fully integrated extension of the aqueduct, which emerges gracefully from the tunnel under the Antonine Wall. As with the Anderton Boat Lift, the gondolas—whether empty or carrying watercraft—are of equal weight, thus minimizing the amount of power needed to rotate the wheel. (According to the promotional literature, the motors use only “the same power as 8 toasters” to turn the wheel and lift “the weight of around 100 African elephants” through “the height of eight double-decker buses.”) A boat ride on the wheel is silent, slow and spectacular. Beginning in a basin connected to the Forth & Clyde Canal by a single lock, the boat enters the lower gondola as if it were just another lock. After the watertight doors close, the crew of the great wheel conducts a series of checks and double-checks before the system begins to move. When it does, it moves almost imperceptibly in a great arc. After reaching the top, the watertight gates at the other end of the gondola open and allow the boat to float ceremoniously onto the aqueduct, through the graceful tunnel, and into the revitalized Union Canal. The transit between canals takes about 15 minutes. The Falkirk Wheel, which opened formally in 2002, is located in a parklike setting, with plenty of space and distance from which visitors can admire it. At one end of the basin sits a plaque recognizing the significance of the Forth & Clyde Canal to international civil engineering. At the other end rises the wheel, gleaming against the Scottish sky. Be22

American Scientist, Volume 92

side the basin is a visitor center, with a sloping wall of glass that allows the wheel to be viewed in comfort even in inclement weather. The commodious center was designed to accommodate a large number of visitors, but interest in the Falkirk Wheel has proved so great that the visitor center is already scheduled for an expansion. As popular and elegant as it is, the Falkirk Wheel has not been without its detractors and those who would improve it. One reader, writing to The Engineer, conceded that it is a “fantastic structure” but wondered why the leveling gear was exposed. He did not feel that it fit in aesthetically and saw it as subject to corrosion in its bared state. In a manner representative of classic confrontations between engineer and architect, the writer added, “The design was prototyped in Lego, but I wonder if Meccano would have produced a more elegant solution. It is no surprise that an architect would use Lego rather than Meccano.” Meccano is, of course, the British precursor to the Erector set. In 1949, the toy’s publication, Meccano Magazine, carried on its cover a striking portrait of the Anderton Boat Lift. If it were still being published, the magazine might not feature the Falkirk Wheel on its cover, however, because the sleek lines and graceful curves of this bespoken structure represent the architecture and engineering of a new millennium, suggestible but not fully capturable in the angular components of Lego or Meccano sets. British Waterways could not be happier with the Falkirk Wheel, and it is considering a project for a new boat lift connecting the Grand Union Canal at Milton Keynes with the River Great Ouse in Bedfordshire, north of London. The linking canal would be the first built in Britain in 100 years, and its route requires the crossing of the M1 motorway and the scaling of Brogborough Hill, whose 150foot elevation would require a staircase of a couple dozen locks. As an alternative, a design competition for a boat lift is being considered, which would offer engineers a “blank canvas to come up with a really innovative solution.” One thought is that the boat lift might be like the British Airways London Eye—the giant Ferris wheel that now operates beside the River Thames—carrying not people but canal barges over the hill. This sounds surprisingly like the exemplar that yielded the Falkirk Wheel. But a successful design competition will likely result in something that is as different from the wheel at Falkirk as it is from the lift at Anderton. Engineers, like architects, love blank canvases. Bibliography British Waterways. 2003. Assorted folders and brochures on individual canals, navigations, and waterways. See also www.britishwaterways.co.uk. Carden, David, and Neil Parkhouse. 2002. A Guide to the Anderton Boat Lift. Lydney, Gloucestershire: Black Dwarf Publications. 2002. Wheels within wheels. The Engineer, April 5, p. 18. (See also letter, April 19, p. 13.) 2002. Project study: Falkirk Wheel. New Civil Engineer

© 2004 Sigma Xi, The Scientific Research Society. Reproduction with permission only. Contact [email protected].