Gateways to Commerce: The U.S. Army Corps of Engineers' 9-Foot Channel Project on the Upper Mississippi River

All of the 9-foot channel installations along the Upper Mississippi River include observation decks. These decks offer scenic views of the river valley, as well as an opportunity to view the operations of the various lock and dam structures. And even the most casual observer of the 9-foot channel can see the differences between the dams, particularly between the dam gates. Some dam gates look like cylindrical tubes; others are pie-shaped wedges. While some dams have only one type of gate, others combine different systems.

During the course of the 9-Foot Channel Project, the U.S. Army Corps of Engineers developed technological innovations so quickly that, in some cases, structures were out-of-date almost as soon as they were built. Nowhere is this more evident than in the design of the dam gates. At the beginning of the 9-Foot Channel Project, Corps engineers designed dams that were equipped with non-submersible roller gates, then considered a state-of-the-art technology. These gates were soon followed by submersible roller gates that, in addition to having the capability of being raised, could also be lowered below the water's surface to allow for easier passage of ice and debris. During the project's middle stages, Corps engineers experimented with combination roller gate/Tainter gate dams. By the project's end, the Corps had designed new submersible and elliptical Tainter gate systems that made the combination gate systems obsolete. Indeed, the evolution of Tainter gate design is arguably the most significant technological development of the 9-Foot Channel Project. But it was not the only innovation. During the construction of the 9-foot channel on the Upper Mississippi River, the U.S. Army Corps of Engineers also generated innovations in construction techniques, foundation pilings, lock design, operating machinery, and a host of other technical details.

Thus, a unique construction document developed as the project progressed, imbedding itself into the concrete and steel of the 9-foot channel lock and dam system. From the beginning to the end, as Corps engineers designed and refined the technology of the 9-foot channel, those technological changes were manifested in the 9-Foot Channel Project's locks and dams. When viewed as a whole, the entire system affords a look at the dynamic evolution of American river engineering in the first half of the twentieth century.

The Locks and Dams of the 9-Foot Channel The Corps of Engineers typically designed each lock and dam complex in the 9-foot channel to include four basic components: A non-navigable movable dam. Each dam holds back enough water to create a navigation pool, or slack-water lake, of a minimum 9-foot depth, while ensuring that the water level does not raise so high as to create flooding. The dams are "non-navigable" because vessels cannot navigate directly over them or through the dam gates; all river traffic must pass through the locks. The dams are "movable" because they have gates that are raised or lowered. Lockmasters adjust the flow of water through the gates. The Corps equipped each dam structure with roller and/or Tainter gates, as well as dikes, spillways, submersible dams, and other auxiliary installations. Locks. Locks provide navigational access through the dam complex. Locks are the steps in the "aquatic staircase" of a lock and dam system, by which vessels are lifted or lowered from one pool to the next, while the pools themselves remain level. Corps engineers built the Upper Mississippi locks to standard dimensions of 110 feet wide and 600 feet long. Foundations for future auxiliary locks. Most complexes include an auxiliary 100 by 269-foot lock. Although the auxiliary locks were seldom completed, the Corps always built the auxiliary lock foundations, equipping them with emergency gates. The foundations allow for future expansion. The Corps opens the gates of the auxiliary lock when the pool is drawn, allowing river traffic to pass. Buildings and esplanades. The Corps built several buildings and structures at each complex site. These include lockmasters' and assistant lockmasters' dwellings, main control stations, stage recorder houses, garages, outbuildings, and various esplanades or greenspace. [1] Lock and Dam No. 24, c. 1980. (U.S. Army Corps of Engineers, St. Louis District)

Roller Gates

Lock and Dam No. 15, located near Rock Island, Illinois, was the first complex constructed as part of the Upper Mississippi River 9-Foot Channel Project, and its non-navigable roller gate dam served as the prototype for the project's subsequent roller-Tainter combination dams. The Corps of Engineers released the plans of the complex for public review in December 1930. When it was completed in March 1934, Dam No. 15 became the largest roller gate installation in the world. Individual gates had been built in Europe that were both larger and longer, but never had a single dam incorporated so many gates of such an aggregate length. With Dam No. 15, the Corps of Engineers developed construction methods and techniques that served as models for the remainder of the Upper Mississippi 9-Foot Channel Project. [2]

Dr. Max Karstanjen, director of the Maschinenfabrik Augsburg-Nurnberg (MAN) in Germany, had developed roller gates at the turn-of-the-century in Germany. European engineers, particularly those in the Scandinavian countries, adopted the design almost immediately. Two German companies, the Krupp Company and the MAN Company, controlled basic patents for the gate. By 1930, European engineers had been using roller gates in dams extensively for over 25 years. However, only 10 such structures had been built in the United States. [3]

The first roller gates on the Upper Mississippi were non-submersible. Non-submersible gates are raised above the water's surface to allow the river to flow freely beneath them. Submersible roller gates are lowered beneath the water's surface, allowing the river to pass over them. Submersible roller gates were seen as an improvement over non-submersible gates because they allowed for the almost unobstructed flow of floodwater, ice, and debris. Dam No. 25 Construction Drawing, January 1937. (U.S. Army Corps of Engineers, St. Louis District) (click on image for a PDF version)

When the Corps of Engineers decided to utilize roller gates on the Upper Mississippi River dams, that decision represented a massive commitment to a familiar technology in an unfamiliar context. However, the Corps refined and improved the design and operational characteristics of roller gates throughout the course of the Upper Mississippi River project. These innovations and improvements resulted in the development of a decidedly American style of submersible roller gate.

Still, American engineers understood the importance and influence of German engineering technology on the 9-Foot Channel Project. In a 1938 article published in The Military Engineer, Lieutenant Colonel P.S. Reinecke noted parallels between the Mississippi and the Rhine regarding commercial navigation. Reinecke delineated the similarities and differences in the two "regulated river" projects, including the advances made in American technology. He concluded his article on a positive economic note: "After considering the example of the Rhine . . . it is reasonable to forecast a highly successful and satisfactory development of navigation along the Upper Mississippi in a few years." [4]

Simply defined, a roller gate is a cylindrical, metal tube that lays across the water between two concrete piers. The first roller gates on the Upper Mississippi were non-submersible. When lowered, a non-submersible roller gate rests directly on the dam's concrete sill, holding back the water. When raised, the roller gate allows water to flow freely beneath it. A single dam can have several roller gates, each of which can be operated independently. Each gate raises and lowers by means of a multiple, side-bar chain mechanism, similar to an enormous bicycle chain. The solid ends of each roller gate are fitted with sprockets, which engage inclined racks attached to the piers. As an electric motor housed within one of the piers hauls in the chain, the gate moves slowly up the racks. Only one end of the drum is driven, the other end merely rides up the racks. [5]

Construction of Dam No. 25, October 1938. The concrete baffles help dissipate the energy of the water passing over the spillway. (U. S. Army Corps of Engineers, St. Louis District)

A roller gate's trussed cylinder provides strength, and makes the gate very rigid along its longitudinal axis. An early improvement to the basic design was the attachment of a curved steel apron to the lower side of the gate, which reduced the required drum diameter and permitted use of a waterproof timber or rubber seal. The ends of a roller gate are equipped with flexible steel shields, the linings of which fit the slightly inclined planes of the faces of the piers. Water pressure keeps the end shields tight against the piers. [6]

The first roller gates in the United States were used on flood control and irrigation projects. In 1914, the Washington Water Power Company used three roller gates for spillway crest control at a dam located on Long Lake near Spokane, Washington. Each of these gates was 65 feet long. In 1916, the U.S. Bureau of Reclamation Service utilized seven roller gates, measuring up to 15 by 70 feet, in an irrigation dam located on the Grand River near Palisade, Colorado. [7]

An early improvement to roller gate design was the attachment of a curved steel apron to the side of the gate, which reduced the required drum diameter and formed a more reliable and waterproof seal. Dam No. 5A. (Clayton E. Fraser, Fraserdesign)

With the Grand Valley Diversion Dam near Palisade, Colorado, Reclamation Service engineers established the basic form for American roller gate dams. As a result of this project, Reclamation Service engineers also gained considerable experience in the fabrication and erection of roller gates. The outbreak of World War I in Europe prohibited the German patent holder, the MAN Company, from completing the design and fabrication of the gates. As a result, Frank Teichman, an engineer for the Reclamation Service, hurriedly reworked the German plans, and the Riter-Conley Manufacturing Company of Pittsburgh fabricated the gates. Elevations and sections of the Grand Valley Diversion Dam show the basic form of the piers and fixed service bridge to be the same as those in Dam No. 15 of the Upper Mississippi River 9-Foot Channel Project. The architectural style of the Grand River dam's piers are also clearly direct precursors of Dam No. 15's roller gate piers. [8]

Upper Mississippi Valley Division Head Engineer William McAlpine and the design team of the 9-Foot Channel Project chose to include roller gates in that project's dams for a variety of reasons. The Corps engineers had their choice of several types of movable crest gates: vertical lift gates, Tainter gates, sector gates, and roller gates. But roller gates had several advantages. The inherent strength of roller gate drums permitted construction of long gates with an economic use of metal. At the time, roller gates could also be built at lengths greater than any of the other types of movable gates suitable for pier dams. Corps engineers knew that these greater lengths would allow maximum clear openings through which to pass running ice, drift, and flood waters. The longer lengths would also reduce the number of dam piers, which also served as obstacles. The Corps wanted to achieve the freest flow possible on the Upper Mississippi River. Winter ice gorges posed a serious threat not only to the dams, but also to private property along the riverbank. In the spring, the river's flood waters carried heavy drift. The 9-Foot Channel Project engineers understood the necessity of passing the maximum amount of flood water quickly because the Upper Mississippi's banks were low, the bottoms densely farmed, and towns and railroads were located close to the river. [9]

The Corps of Engineers also selected roller gates because of the gates' massive construction, operating machinery, and their ability to be raised. The ruggedness of roller gates assured their positive operation under the most severe conditions. Corps engineers believed they would function better under freezing conditions than any other type of gate. The design of the seals, both side and sill, also meant that roller gates were less likely to leak. In addition, the engineers predicted that roller gates could be easily heated in freezing weather. (Unfortunately, this did not prove to be the case.) The Corps also preferred roller gates because of their ability to be raised above the water. The engineers were leery of submersible gates, such as some types of sector gates, because they were concerned about excavating the riverbed into which the gates would be lowered. They expected these excavations to fill up with sediment. [10]

By 1933, however, the 9-Foot Channel Project engineers had changed their minds about submersible gates. Construction of Dam No. 15 began in February 1932, and by 1933 enough gates were in operation for the engineers to learn that floating ice could not pass underneath the gates unless they were raised approximately half the distance from the sills to the surface of the upper pools. To open the gates this far during freezing temperatures, when the Upper Mississippi is usually low, lowered the upper pool below the desired level. This practice adversely affected the natural habitat of the area, and created scour problems below the dam apron and stilling basin. Opening the gates so wide also produced a very concentrated flow and caused serious erosion on dams with sand foundations. As a result, project engineers reconsidered the use of submersible roller gates that would allow ice, debris, and flood waters to pass over, rather than under, the gates.[11]

Almost simultaneously with its work on Dam No. 15, the Corps of Engineers was designing roller gate dams on the Kanawha River, a West Virginia tributary of the Ohio River. On the Kanawha River, Corps engineers specified that one of the dam gates be fully submersible. Dravo Contracting of Pittsburgh, the gate contractor, proposed instead to fabricate a gate with a movable flap hinged to the top of the drum. This flap could be lowered onto the face of the drum, providing a 5-foot overflow atop the gate. Although this innovation was not widely used in subsequent Corps of Engineers' roller gate dams, it pointed the way towards the use of submersible gates, the most significant innovation in roller gate design to emerge from the 9-Foot Channel Project. Corps engineers refined and improved the design and operation of roller gates throughout the course of the Upper Mississippi River project. These innovations resulted in the development of a decidedly American-style submersible roller gate. [12]

Roller Gate Pier and Operating Machinery House. Dam No. 25. (John P. Herr, John Herr Photography)

The Corps of Engineers first used submersible roller gates on the Upper Mississippi River at Dam No. 4, located in the St. Paul District. The Corps began constructing Dam No. 4 in November 1933. The submersible roller gates of Dam No. 4 submerge to a depth of 3 feet. At other dams in the St. Paul District, such as Nos. 3 and 9, Corps engineers installed gates that submerge approximately 5 feet, so that heavy loads of ice could easily pass over them.

Rock Island District staff made many of the design modifications relating to submersible roller gates. The construction of Dam No. 15 within that district had demonstrated the problems associated with non-submersible gates. Although Dams Nos. 20 and 16, which were designed in August 1933 and September 1934, were equipped with the original, German-designed, non-submersible roller gates, the district design team soon modified this basic design.

District engineers designed their first submersible roller gates in 1935 for Dams Nos. 11 and 18. A non-submersible roller gate had only one sill level; the gate lowered against the sill to form a bottom seal. But the new submersible roller gate had two sill levels: a higher upstream level and a lower downstream level, which were joined by a curved section of concrete. The Corps designed the submersible roller gate to either rest next to the higher level of the sill—to form a bottom seal—or slide along the curved section of concrete until it reached the lower level. At the lower level, the gates at Dams Nos. 11 and 18 are 8 feet below their normal closed position.

The new submersible roller gates did not totally solve the scour or water elevation problems of the upper pool that had manifested themselves at Dam No. 15. As a result, Corps engineers attempted to mitigate these problems by reducing the depth to which the gates submerged. The roller gates in Dam No. 21 were designed to submerge to a depth of 4.67 feet, but engineers were still dissatisfied. In 1936, the Corps designed roller gates for Dam No. 22 that submerged 8 feet. In this design, the engineers also incorporated a Poiree dam trestle the same height as the downstream fender on the adjoining piers. The Corps set the dam trestles immediately downstream from the downstream level of the sills. As a result, the trestle units acted as weirs. In 1937, the Corps hired the Worden-Allen Company of Milwaukee to furnish enough Poiree dam trestles to retrofit the dams in all three districts built prior to Dam No. 22. Still, the Rock Island District was not happy enough with this solution to include it in its new designs. For its last three dams, Nos. 13, 14, and 17, the district engineers once again reduced the submersible depth to 4.67 feet. [13]

Corps officials were in constant communication with the companies that manufactured the roller gates. Companies such as the Allis-Chalmers Company of Milwaukee, the M.H. Treadwell Company of New York, and the McClintic Marshall Corporation of Chicago were included in the continuing dialogue involving every possible detail of roller gate design. The first indication of a problem triggered a flurry of correspondence between the Corps and private engineers to find solutions. Typical problems involved such items as discrepancies in interior bracing in the drum assemblies, end shield adjustments, and 10-degree changes in pressure angle designs for rack and spur rim teeth. The Corps of Engineers had to accurately resolve these and countless other details on a day-to-day basis in order for the leviathan project to proceed. [14]

A roller gate is raised and lowered by a multiple side-bar chain similar, on an enormous scale, to a bicycle chain. Dam No. 25, Roller Gate Operating Machinery Construction Drawing, January 1937. (U. S. Army Corps of Engineers, St. Louis District) (click on image for a PDF version)

Tainter Gates

While roller gates attracted the greatest attention in the national civil engineering journals, most of the gates used in the 9-Foot Channel Project were Tainter gates. By the end of the project, the U.S. Army Corps of Engineers had developed Tainter gates that were so advanced that they made roller gates obsolete. Indeed, Tainter gate design developed so rapidly that one Corps officer commented that "the first dam of a series may be partially out-of-fashion by the time the last one is built." [15]

Viewed from the side, a Tainter gate and its armature look like a pie-shaped wedge: a lengthwise segment of cylinder with triangular arms extending from each end. The cylindrical section of the gate forms the damming surface. While a roller gate is lifted up and down, the arms of a Tainter gate pivot on pins attached to the supporting piers. The Corps of Engineers opened and closed the first Tainter gates on the Upper Mississippi River by means of a cable or chain attached to the lower side of the gate shield and driven by machinery located above the gate on the dam's service bridge. The shape of the gate was such that the water pressure behind the gate had little effect, and the hoist machinery merely had to overcome the deadweight of the gate. By the end of the project, Corps engineers had eliminated the hoisting method in favor of a system operated by a line shaft and motors. [16]

In cross-section, a Tainter gate and its armature resemble a pie-shaped wedge. From the smaller end, triangular arms extend toward the wider, waterside face of the gate. Like this gate on Dam No. 8, the first Tainter gates used on the 9-Foot Channel were non-submersible. Dam No. 8 Construction Drawing, June 1935. (U.S. Army Corps of Engineers, St. Paul District) (click on image for a PDF version)

The Corps of Engineers continually refined Tainter gate design throughout the 9-Foot Channel Project. The submersible Tainter gates on Dam No. 24 are elliptical; steel skin plates totally surround a steel truss frame. The streamlined shell protects the gate's framework from ice and debris, and provides a smooth surface for the water. Dam No. 24 Construction Drawing, December, 1937. (U.S. Army Corps of Engineers, St. Louis District) (click on image for a PDF version)

The Corps also experimented with a truncated form of submersible elliptical gate. Dam No. 25 Construction Drawing, January 1937. (U.S. Army Corps of Engineers, St. Louis District) (click on image for a PDF version)

Arch-shaped, non-submersible gates were used at Dam No. 14, where the river bedrock made it very expensive to build submersible gates. Like the elliptical gates, arch-shaped gates are encased in steel siding. Dam No. 14 Construction Drawing, August 1936. (U.S. Army Corps of Engineers, Rock Island District) (click on image for a PDF version)

North American hydraulic engineers had been using radial gates based on the same principles as the Tainter gate for over 100 years by the time the Upper Mississippi design team began incorporating them into its designs. Tainter gates are of American origin. As early as 1827, when Captain Marshall Lewis applied for a patent on his semi-circular, cast-iron gate turning on pivots connected to the gate by arms, he admitted he had not designed a new gate type but merely made some important improvements in what was already known as the "common paddle gate."

Refinements in Tainter gate design continued throughout the nineteenth century. In 1840 and 1841, George W. Hildreth of Lockport, New York, and George Heath of Little Falls, New York, patented such similar refinements in the design that a series of court cases and state legislative actions ensued. By 1853, the noted French hydraulic engineer Poiree had adapted a similar segmental arc gate for use in movable dams. Wisconsin lumberman Theodore Parker made further refinements in the basic design. Parker sold his rights to Jeremiah Burnham Tainter who patented the gate system in 1886. In 1889, Major William L. Marshall became the first Corps officer to use Tainter gates, adopting a manually-operated version of them for use on a movable dam across the Rock River between Rock Falls and Sterling, Illinois. [17]

Tainter gates are economical and simple to fabricate, erect, and operate. But there was still room for improvement when the UMVD design team began assimilating them into its Upper Mississippi River designs in 1932 and 1933. At first, Corps engineers considered Tainter gates too small and too unreliable, in terms of their operation under adverse conditions, to be used in the principal spillway sections of the dams. In addition, most American Tainter gates averaged 30 to 35 feet in length. Given the Corps' requirements for a 100-foot gate that operated reliably under all kinds of conditions, Corps engineers initially designed the main section of the dam with roller gates, and completed the movable portion with a series of Tainter gates. The engineers had discovered that three or four 100-foot-long roller gates situated in the thread of the stream were all that were necessary to pass ice, drift, and flood waters satisfactorily. [18]

Initially, the Corps of Engineers believed the combination roller and Tainter gate dam was ideally suited to the particular needs of the 9-Foot Channel Project. However, many contemporary experts in the field of movable dam construction believed that dams should never employ more than one type of gate. In his 1937 work, Hydraulic Structures, Armin Schoklitsch argued that "water level[s] can be regulated and the debris deposits sluiced out more effectively if all the regulators [gates] are alike . . ." The decision to use more than one type of gate in any dam was, according to Scholditsch, "essentially a matter of cost." The U.S. Army Corps of Engineers, however, adopted the combination gate for the same reason that Schoklitsch advised against it—to accommodate the need for an exact regulation of flood stage variables. Again, the Upper Mississippi dictated a change in engineering philosophies. [19]

During the 9-Foot Channel Project, the Corps developed several innovations in Tainter gate design. Prior to the project, engineers believed that Tainter gates, like the original German-designed roller gates, should be raised above the water rather than submerged. However, the Corps' work in developing submersible roller gates eventually resulted in the design of a submersible Tainter gate. Corps engineers designed many of the 9-Foot Channel dams, such as Nos. 20 and 16, to include both submersible and non-submersible Tainter gates. The earliest form of a non-submersible Tainter gate was comprised of a drum-type gate with steel skin plates on the curved water-side face. The water-side face of the submersible gate was identical to that of the non-submersible gate. However, Corps engineers modified the submersible gate to include an additional overflow plate that arched back downstream from the top of the gate face.

In May 1935, Corps engineers initiated a new kind of submersible Tainter gate at Dam No. 18, located in the Rock Island District. On this new gate, steel skin plates totally surrounded a steel truss frame. The Corps designed the gate to be the shape of a three-quarter ellipse; the back of the gate face was convex rather than concave. The streamlined steel shell of this new gate protected the gate's steel framework from ice damage, and provided a smooth unobstructed surface for the water that passed over the gate in its submerged position.

The 9-Foot Channel Project engineers had conducted model studies that indicated that the earlier, drum-type gates created "a negative pressure on the crest that may cause vibration and excessive fatigue, or corrosion of the metal." Manufacturers, however, found the new, elliptical-shaped gates to be difficult and expensive to build, and the truss framing required the distortion of certain connection angles. As a result, Corps engineers revised the framing, substituting a girder frame for the truss frame. Dam No. 11 was the first representation of this more sophisticated, elliptical, Tainter gate design.

Despite scour problems, Rock Island District engineers incorporated the elliptical Tainter gate into the five dams founded on sand that they designed after Dam No. 18. The engineers also continued to test new designs. At Dam No. 22, district engineers added elliptical shields to both ends of that dam's one submersible Tainter gate. When the gate was submerged, the shields prevented water from seeping between the gate and the piers. The Corps followed up on this experiment. In 1939, the Corps' Hydraulic Laboratory at Iowa City made 19 models of Dam No. 22's submersible gate, and conducted 157 tests on them in order to develop a satisfactory stilling basin for submersible Tainters, as well as a design for improved Tainter gate hoisting machinery and operation. [20]

The first form of Tainter gates used on the 9-foot channel had steel plating only on the water-side face of the gate. A submersible gate, such as this one on Dam No. 8, can be distinguished by the additional overflow plate on the top of the gate, which provides a streamlined surface for breaking through the water. Dam No. 8, Submersible Tainter Gate. (Clayton B. Fraser, Fraserdesign)

At Dam No. 22, the Corps of Engineers also introduced a new type of non-submersible Tainter gate. Like the submersible elliptical gate, the new non-submersible gate was totally surrounded by steel siding. But, rather than being elliptical, the gate was arch-shaped, similar to the non-submersible Tainter gates that had metal sheathing only on one side. The Corps apparently wanted the improved longitudinal rigidity, increased strength, and ruggedness of the elliptical Tainter gate but did not need, at a dam founded on bedrock, the scour diminution offered by the submersible gate. Also, because of the bedrock foundation, it would have been very expensive for the Corps to have constructed the two-level sills necessary for submersible gates. The Rock Island District used these new, non-submersible, Tainter gates exclusively in Dam No. 14, which was also founded in bedrock.

Submersible elliptical Tainter gates and non-submersible arched Tainter gates supported the construction of gates at unprecedented lengths. Corps engineers were soon building Tainter gates at lengths of 60 feet. Eventually, Dam No. 24, located in the St. Louis District, employed Tainter gates that were 80 feet long.

Dam No. 24, Submersible Elliptical Tainter Gate. The shields on the ends of the gate prevent water from seeping between the gate and the piers. Typical of later gate designs, the gate armature on Dam No. 24 is solid. Earlier designs, such as Dam No. 8 pictured on the opposite page, featured open framework. (John P. Herr, John Herr Photography)

The Corps of Engineers designed Dam No. 24 in December 1937. With this dam, constructed in 1938-1939, the Corps attained its highest level of Tainter gate technology during the 9-Foot Channel Project. Corps engineers incorporated fifteen 80-foot-long Tainter gates into the 1,340-foot long movable portion of Dam No. 24. The large size of these gates, and the relatively ice-free conditions that characterize this stretch of the Upper Mississippi, convinced Corps engineers to entirely eliminate roller gates from this dam. At the time of construction, the Corps believed the Tainter gates at Dam No. 24 to be the largest ever constructed. [21]

The Corps designed the Tainter gates at Dam No. 24 to be fully submersible and elliptical in section. Project engineers selected the elliptical design because it permitted the shell of the gate to act as a beam between the end supports, eliminating the need for extensive internal bracing and framework. The design also reduced both the quantity of steel required to fabricate the gate and its operational weight. An additional reduction in weight, as well as an improvement in corrosion resistance, resulted from the use of high tensile, phosphorous chromium steel for most movable portions of the gates.

The construction of Dam No. 25, which has truncated elliptical Tainter gates, December 1937. (U. S. Army Corps of Engineers, St. Louis District)

Dam No. 24 represented a vast improvement in Tainter gate design. Within the space of just a few years, the Corps had improved the design of Tainter gates so dramatically that roller gates, the principal engineering feature discussed in early technical articles related to the 9-Foot Channel Project, were entirely superseded by a cheaper, simpler, and more reliable gate type. These developments made roller gate technology obsolete, effectively bringing to an end the short history of combination roller-Tainter gate dam construction in America.

The Corps' methods of operating Tainter gates also evolved significantly during the Upper Mississippi River 9-Foot Channel Project. Rock Island engineers originally designed all 40 Tainter gates in Dam No. 20 to be hoisted by locomotive cars running on the dam's service bridge. As an experiment, the district later modified two of the Tainter gates so that they were hoisted by individual electric motors, coupled to line shafts, housed in installations above each gate. The new design, which eliminated the need for the locomotive cars, was a success. After September 1934, the Corps ordered all new Tainter gates on the Upper Mississippi to be equipped with motors and line shafts. Although more expensive to construct and install, the new method permitted rapid, simultaneous operation of several gates and required fewer operators. As Corps engineers continued to modify Tainter gate design, they also continued to perfect the line shaft and motor assemblies. [22]

The 9-Foot Channel Project engineers also refined the auxiliary parts of the gate systems. After using the emergency bulkheads for the roller gates at Dam No. 15, the Corps realized that considerable silt accumulated inside these units, increasing their weight and making them more expensive to clean. Consequently, the engineers modified the bulkhead units to provide for end guide and reaction rollers, buffer blocks, molded rubber end seals, more efficient curb plate splices, a better pickup device, and projecting angles in the bulkhead recesses against which the end seals bear. The Corps also improved the Tainter gate emergency bulkheads. Engineers replaced the welded units with riveted units, reducing the number and cost. The new Tainter gate bulkhead units were also stiffer and more stable, and hazards due to defective welding were reduced. The Rock Island District first used these modified roller gate and Tainter gate bulkheads at Dams Nos. 11 and 18, and they soon became the standard units used in the rest of the district's dams. [23]

The U.S. Army Corps of Engineers continues to develop and improve Tainter gate technology on the Upper Mississippi River. The Corps recently completed Lock and Dam No. 26R (the Melvin Price Lock and Dam), which replaced the aging Lock and Dam No. 26. On the new dam, the Corps utilized state-of-the-art Tainter gate technology. Each of the 9 non-submersible gates of Dam No. 26R measures 110 feet across and 42 feet high. As such, the new gates are nearly 3 times the length of the gates of the original Dam No. 26, and nearly 40 percent larger than the 80-foot gates erected at Dam No. 24. However, the principal difference between the modern gates of Dam No. 26R and those constructed in the 1930s is their enormous size. The Corps built and operates Dam No. 26R in essentially the same fashion as the earlier 9-foot channel complexes. [24]

The abandonment of the roller gate dam system on the Upper Mississippi was a logical progression in river hydrotechnology. The Corps replaced the Mississippi River's roller gates with Tainter gates in the same way that it replaced the Ohio River's wicket gates with roller gates. Thus, many of the dams on the Upper Mississippi River have a particular significance because they represent a technology that is no longer designed or constructed, due to advances reflected in the system's own unique evolutionary designs.

The pre-1936 roller gate piers have buttress detailing, large industrial sash windows, rubbed concrete wall finishes, and hipped roofs. Dam No. 6. (Clayton B. Fraser, Fraserdesign)

Architecture of the 9-Foot Channel Project

As in the case of engineering, the architectural designs of the Upper Mississippi River 9-Foot Channel Project also evolved over the 10-year construction period. The Corps of Engineers employed two distinct architectural styles in the navigation structures of the 9-foot channel, which can generally be identified by their date of construction. Dams built before 1936 are relatively ungainly and utilitarian in design. By contrast, post-1936 structures reflect sophisticated, streamlined, Art Moderne styling. [25]

The Corps of Engineers constructed the 9-Foot Channel Project during the Depression of the 1930s, and the changes in architectural design owe much of their evolution to the same era. The 1933 Century of Progress Exposition held in Chicago focused on national and international advances in technology and engineering, including the work of German designers. It is very likely that these currents affected the 9-Foot Channel Project, especially when viewed in light of the comparisons that were already being made between German roller gate designs on the Rhine and the Corps' work on the Upper Mississippi River.

The pier houses of the Moderne, post-1936 structures are incorporated into the overall pier design of the dam structure. The roofs are flat, and the buildings have slit, three-pane windows. Dam No. 5A. (Clayton B. Fraser, Fraserdesign)

Passages from the Century of Progress Exposition's guidebook reflect the post 1936, dam pier design of the 9-Foot Channel Project: "Consider the architecture of the buildings. Wonder, perhaps, that in most of them there are no windows. Note curiously that these structures are for the most part unbroken planes and surfaces of asbestos and gypsum board . . ." Architecture and planning were seen as elements of a "huge experimental laboratory" designed to further modern concepts in both fields. As the Corps of Engineers was completing the Upper Mississippi 9-Foot Channel Project, the 1939 World's Fair in New York again focused on technology, but this time without German participation. Nevertheless, the Depression-era "message of the modern" had not been lost on America's engineers. The Corps of Engineers had manifested that message into the concrete and steel structures of the 9-Foot Channel Project. [26]

Buildings and Esplanades The Corps of Engineers also designed the lockmasters' and assistant lockmasters' residences, control stations, and other ancillary buildings at each 9-foot channel site. Like the navigation structures, these buildings are simple and utilitarian. In these cases, however, the Corps borrowed from Colonial and Georgian models, rather than modern. Lock and Dam No. 10, Esplanade and Landscaping Drawing, February 1938. (U.S. Army Corps of Engineers, St. Paul District) (click on image for a PDF version) Living Room of Lockmaster's House, Lock and Dam No. 10, February 1938. (U.S. Army Corps of Engineers, Rock Island District) Lockmaster's House, Lock and Dam No. 8. (Clayton B. Fraser, Fraserdesign)

Locks

The Corps of Engineers modeled the general lock design for the Upper Mississippi River after the Ohio River canalization. As on the Ohio River, the standard Upper Mississippi River lock measures 110 by 600 feet, and is equipped with miter gates. The Corps of Engineers had originally used roller-type gates on the Ohio River locks, but malfunctions caused by ice and sediment encouraged the development of a new gate system. By 1913, the Corps had perfected a miter gate design for the Ohio River locks, a mechanism consisting of two hinged panels that, when closed, forms a miter point or "V" configuration pointing upriver. Miter gates are located at both ends of a lock, opening and closing to allow river traffic to pass through the chamber.

Although the Corps refined the design, miter gates were the most traditional type of lock gate and had been used for hundreds of years on canal locks prior to their use on the Ohio and Mississippi Rivers. There were few problems involved in designing miter gates when the locks they enclosed were 18 feet wide, an average mid-nineteenth century size. By the 1870s, miter gates reached widths of 80 feet, which were still workable. But when Corps engineers began designing 110-foot-wide miter gates for the Davis Island Lock on the Ohio River, they concluded that such large miter gates were not feasible with available materials. This ushered in a 30-year period in which the Corps either restricted miter gates to an obsolete 80-foot width, as in the case of the Rock Island District's Moline and Le Claire locks, or experimented with other types of lock gates. [27]

In 1913, the Corps' Louisville District engineering staff under William H. McAlpine, who later played such a significant role on the Upper Mississippi River 9-Foot Channel design team, finally solved the engineering problems inherent in developing a 110-foot-wide miter gate. The Louisville engineers combined an interior bracing frame, surrounded by steel-skin plates, with the traditional flat-leaf form of a miter gate. The resulting gate was strong enough to do its job well, yet light enough to be easily operated. This 1913 miter gate design was used on the Upper Mississippi River, with Corps engineers continuing to refine the design throughout the course of the 9-Foot Channel Project. [28]

Corps architects designed the 9-Foot Channel Project control stations with neoclassical arched windows and simple detailing. In all cases, simplicity was the hallmark of design. Central Control Station, Lock and Dam No. 7. (Clayton B. Fraser, Fraserdesign)

The most significant alteration to the Ohio River design involved the plan of the culverts and ports used to flood and empty the lock chambers. To fill and empty the chamber, the Ohio River locks utilized a series of small culverts passing directly through the lock's river wall above and below the dam. Each culvert was individually controlled by a valve. On the Upper Mississippi, Corps engineers replaced this complex arrangement with a system of large longitudinal culverts, controlled by four valves located in the base of the lock walls. This system provided greater dependability and required less maintenance. The Corps located the intake and discharge openings in the lock walls, respectively above the upper gates and below the lower gates. A series of small ports branched off the main culverts and flooded or emptied the lock chamber. [29]

The Upper Mississippi River engineers decided to use Tainter valves to control the flow of water into and out of the lock chamber. This design feature also represented a significant departure from the Ohio River project, where a roller valve design had been developed. The special Board of Engineers, in their final survey report of December 1930, had not dictated the design of the valves and operating machinery for the locks. However, the report noted that three types of valves were suitable: Stoney roller valves similar to those used in the Panama Canal locks; butterfly valves, such as those used at the Emsworth and Dashields Dams on the Ohio River; and Tainter gates, as used on the Welland Canal. The Tainter valves that were ultimately incorporated into the Upper Mississippi River locks function in the same way as the Tainter gates in dams. The Tainter valves are raised and lowered by electrically-driven cable hoists. In the lowered position, the seals on the sides and bottom of the valves close off the culverts, preventing water from either entering or leaving the lock chamber. [30]

Corps engineers perfected other aspects of lock design during the 9-Foot Channel Project. Rock Island engineers experimented with rubber gate seals rather than the conventional wood-on-concrete seals. The Corps also corrected problems in lock gate operating machinery. At Lock No. 20, the original specifications called for single-speed motors that allowed 1 minute for opening and closing the gates. Before these machines were installed, however, the Corps' experience with 1-minute cycles at Lock No. 15 showed this speed to be too great for safe operation. As a result, engineers reworked the motors for the lock gates at No. 20 for a two-speed operation. The high speed winding developed 25 horsepower (hp.) at 1200 revolutions per minute (r.p.m.), resulting in a closing time of 1 minute. The low speed winding developed 5 hp. at 300 r.p.m. for a closing time of 4 minutes. These re-worked motors served the lock effectively for over 50 years. [31]

Lock No. 24 Construction Drawing, January 1936. (U.S. Army Corps of Engineers, St. Louis District) (click on image for a PDF version)

Miter gates are electrically operated. Motor assemblies housed in machine pits in the lock walls power the arms and gears which open and shut the gates. The lockmaster controls each pair of gates independently by switches located in weather-proof cabinets on the lock walls. Miter Gate on Auxiliary Lock No. 26. (John P. Herr, John Herr Photography)

The Corps also had to contend with outdrafts, or varying water currents, in the design of the locks. Upon completion of Lock No. 20 in 1935, the engineers discovered that strong outdrafts were making navigation in and out of the upstream end of the lock very difficult. In an effort to cut down on the outdraft, the Corps ordered the lockmaster's staff to keep the Tainter gate closest to the lock closed as much as possible. The engineers also encouraged tow operators to hire an extra tow boat to help hold the head of their tow in place.

Locks Nos. 11, 16, and 18, completed in 1937, had equally severe outdrafts. Locks Nos. 21 and 22, completed in 1938, also had outdraft problems. Rock Island engineers now realized that they had made inadequate engineering assessments. As a result, the district staff began experimenting with designs that could retrofit all six locks. The engineers developed a 500-foot, concrete, cell foundation extension to the upstream end of the riverward wall of the auxiliary lock. The foundation cells extended almost to the surface of the upper pool. Only a small portion of the wall was underwater, offering virtually no obstruction to the flow of water. By May 1940, the Rock Island District had adapted this structure for all the individual lock sites. [32]

Locks Nos. 13 and 14, which went into operation in 1939, also had strong outdrafts. Here, however, the Corps engineered different solutions. At Lock No. 14, the Corps adapted the riverwall extension directly to the riverward wall of the main lock. At Lock No. 13, Corps engineers added a simple 1,064-foot-long mooring levee extension to the upstream guidewall. With the upstream guidewall extended, boats could be physically held against the guidewall. The Corps added a similar mooring levee extension to the upstream guidewall at Lock and Dam No. 15. [33]

Tainter valves, which control the flow of water in and out of the lock chamber, work on the same principle as the Tainter gates in the dams. Construction Drawing for Valve Operating Machinery for Lock No. 24, October 1935. (U.S. Army Corps of Engineers, St. Louis District) (click on image for a PDF version)

The extensions of the upstream guidewalls at Locks Nos. 13 and 15 proved such a success that the district installed similar, but smaller, additions to Locks Nos. 11, 21, and 22. District engineers added approach flow deflecting dikes at Locks Nos. 11 and 22, where the outdrafts were more severe. They also added extensions to the upstream landwalls at Nos. 11 and 22, and a transitional section of approach dike and a mooring cell to No. 13. However, engineers still had to deal with outdrafts at the other nine complexes in the district. [34]

The Corps of Engineers designed all of the 9-foot channel installations with main and auxiliary locks but, at most of the complexes, the auxiliary locks were never completed. The Corps constructed auxiliary lock foundations, and equipped the incomplete locks with emergency gates. These gates open when the pool is drawn, allowing river traffic to pass. At some sites—Nos. 1, 15, 19, 26, 26R, and 27—the Corps completed the auxiliary locks because the dams are so high that there is no other way to provide for passing navigation except through the locks. The additional lock can also be used in case of accident or repair to the main lock, or as an auxiliary should increased river traffic require it. [35]

Foundations

In some river locations, Corps engineers situated the 9-foot channel locks and dams on bedrock, which provided a firm foundation. At others sites, however, the Corps located the locks and dams on silty, sandy riverbed. Here, the Corps had to support the concrete foundations of the locks and dams on a multitude of pilings driven into deep layers of riverbed sand—challenging conditions that called for specific engineering solutions.

The disparate foundation conditions that confronted the Corps can be seen within the Rock Island District. Foundation construction was not a particular problem for the four Rock Island District lock and dam systems sitting on bedrock: Nos. 14, 15, 20, and 22. But seven other complexes had to be built on sand: Nos. 11, 12, 13, 16, 17, 18, and 21. At these locations, the district built the structures upon concrete foundations set on sealed timber pile configurations.

By 1934, however, Corps engineers had learned that the land walls of the locks resting on the timber pilings did not have adequate stability. The lock land walls were incapable of resisting the horizontal thrust imposed by the back filling necessary to create the lock esplanades. To remedy this situation, the engineers added a series of reinforced concrete struts to the foundations of the lock chambers. The top of these struts were level with the lock floor. Thus, a portion of the horizontal thrust imposed on the landwall by the esplanade fill carried to the riverward wall of the lock. The engineers also required struts between the land walls and the riverward wall of the lock upstream from the upper miter sill and downstream from the lower miter sill. [36]

Similarly, Rock Island engineers redesigned the foundations of the guidewalls at several locks. The UMVD had departed radically from standard practice in designing the guidewall foundations. Traditionally, guidewalls were located on rock foundations, where such foundations existed. During the 9-Foot Channel Project, however, the Corps constructed guidewalls on timber cribbing partially filled with rip-rap. Concrete was supported on timber stub pilings either placed directly on rock or driven to refusal. The substitution eliminated the necessity of expensive cofferdams for this portion of the work.

The placement of the foundation pilings determined the final shape of the lock and dam structure. Pile Spacing Plan for Lock No. 25, August 1935. (U.S. Army Corps of Engineers, St. Louis District) (click on image for a PDF version)

But Corps engineers discovered that the guidewalls at several of the locks resting on pilings had the same weakness to back fill as the land walls. To solve this problem, the Corps held an inter-district conference. Struts could not be used, since no riverwalls existed to absorb horizontal stress. Therefore, the conferees decided that battered piles with a wider crib and some additional vertical piles would give the necessary support. [37]

The St. Paul District also arrived at unprecedented solutions for some of its lock and dam foundations. Unlike the sites at St. Anthony Falls, Minnesota, where the substrata of sandstone and limestone enabled unshored methods of construction, Locks and Dams Nos. 3-10 required concrete foundations set on pilings. At Lock and Dam No. 3, district engineers decided to entirely replace the dam substrata with a more stable type of river sand to accommodate the driving of the necessary underpinnings. Such a departure from standard practice in piling configurations required exact calculations. Corps experts were immediately brought in from a variety of offices for consultation, including William McAlpine of the UMVD.

The Corps of Engineers tested such experimental procedures at various testing centers, including the University of Iowa, the Soils Laboratory at the District Engineer's office at Zanesville, Ohio, and the Fountain City, Wisconsin, laboratory of the St. Paul District. The staff of the District Engineer's office in Zanesville built "photoelastic" models that simulated the unstable strata with a gelatin element to pinpoint the difficulties related to stable foundation design. [38]

The St. Paul District implemented other innovative technologies for foundation construction. For some 9-foot channel foundations, St. Paul engineers used "Z" piling, a type of steel piling first rolled only 2 years before the beginning of the 9-Foot Channel Project. "Z" piling was useful in situations calling for maximum resistance to buckling or bending. The name "Z" comes from the shape of the piling; when one pile interlocks with another or a group of such materials, the result places the maximum amount of metal away from the neutral axis. The "Z" piling configuration gives great strength to the construction. The design of the curved portion of the abutment was felt to be unique at the time of its construction. In order to prevent "bulging" (radial pressure at the fill of the curves), three tension bands were used at varying elevations. These heavy plates were augmented by three lighter plates to ensure additional strength. One end of the outer plate was anchored in concrete. The plate then followed around the exterior of the abutment. Connection to the other end of the plate was followed along the straight section of the abutment and welded to the "Z" piles. The tendency for outward movement at the corner was deflected by the "wrapping action" (circumferential tension) in the steel plating. This method of engineering and construction was an unusual application of "Z" piling and was observed with great interest by contemporary engineers. [39]

As in the case of the St. Paul District, the Corps of Engineers made specialized testing a critical component of the 9-Foot Channel Project. Corps engineers tested the project's technology on a daily basis, with every conceivable element of the project receiving expert scrutiny. Such tests provided valuable research that formed the basis for future improvements. When tests of critical importance were not available in contemporary literature, Corps personnel developed their own. In some cases, as in the testing of substrata stability for the Lock and Dam No. 3 foundations, the ultimate decision to remove the unstable silt layer corrected the problem and rendered the testing results moot. Other experiments failed entirely, as in the design of an auger with specially designed flaps for the retrieval of soil samples from beneath the riverbed. Nevertheless, the Corps of Engineers' research contributions to the body of engineering knowledge remained important.

With conditions ranging from bedrock to sand, the Upper Mississippi riverbed provided a challenge for the U.S. Army Corps of Engineers. At Lock and Dam No. 3, the Corps entirely replaced the original substrata with a more stable type of river sand, which provided better support for the pilings. Timber piles being delivered to the Lock No. 3 work site, May 1936. (U.S. Army Corps of Engineers, St. Paul District)

Positive results quickly evolved from the research and testing process. For example, experimentation in cofferdam design in relation to seepage at Dam No. 6 enabled Corps engineers to improve cell design and configuration, as well as pumping methodologies. Because of its special substrata conditions, Lock and Dam No. 3 provided the impetus for testing everything from trial concrete mixes to probe resistance and soil density. Corps engineers also conducted load, pile, and paint tests, as well as tests for concrete temperature. The 9-Foot Channel Project engineers also drew upon earlier Corps projects for information. Corps engineers at the Huntington District's London Lock and Dam on the Kanawha River in West Virginia had done concrete testing that helped the Upper Mississippi River engineers in their pursuit of elusive data and computation of probable effects of construction. [40]

Contractors also entered the testing realm. The New York firm of Spencer, White & Prentis, the general contractor for Lock and Dam No. 6, developed a new type of skid pile driver that was built on the job site. Vulcan Number One steam hammers equipped the pile drivers; steam power was provided by the Chicago, Burlington and Quincy Railroad. The contractors also developed a new method of keeping pile hammers level, based on a circular steel slot bolted and welded to the hammer's s side. The improved pile hammers drove the piles and sheets plumb, eliminating the need for guide lines.

In order to determine pumping needs, Lazarus White, president of Spencer, White & Prentis, also constructed a model of the Lock No. 6 cofferdam at a scale of 1:24. According to the project report, "The effect of berm and ditches and the relation between depths of sump and elevation of water were clearly demonstrated by the model and the results and information obtained many times justified its cost." [41]

The St. Louis District also made major advances in the analysis of the load bearing capacity of pile-founded structures. In St. Louis, Principal Engineer Lawrence B. Feagin recognized that little field data existed for determining the resistance of piles subject to lateral loads, such as those resulting from backfill against the locks land walls, or water pressure behind lock walls and dams. Lock No. 26 was designed to be founded upon 14,200 timber piles, including 5,000 concrete piles placed in those areas that supported the greatest loads. Feagin resolved to determine the degree of safety afforded by this type of construction for Lock and Dam No. 26 and future projects within the district.

Feagin conducted a series of tests upon groups of vertical piles subjected to static and cyclical loadings. These tests provided information on the behavior of the foundation piles under constant and changing loads, and led to a modification of the original lock design that introduced a series of pile-founded concrete struts extending, at floor level, between the land wall and the intermediate wall, and between the intermediate wall and the river wall. These struts distributed the horizontal loads among the three locks walls, reducing the load on any single wall. Feagin conducted similar load tests upon groups of batter piles at Lock and Dam No. 25, and incorporated batter piles, as well as lock wall struts, into the design of this structure. Lock No. 24, which rested upon a bedrock foundation, did not require the extra structural support afforded by the struts. [42]

At Lock and Dam No. 26, the movement experienced by both structures greatly exceeded the limits projected by Feagin's testing program, indicating that the pile foundations were "grossly underdesigned." Nevertheless, Feagin's pile load tests represented a pioneering effort in the scientific analysis of pile-founded structures. Prior to Feagin's efforts, engineers had little field data regarding the way large pile-founded structures behaved under loads. Feagin's testing program, though inadequate by modern standards, provided a wealth of information on this vital subject. [43]

The Corps conducted numerous tests to determine the load-bearing capacity of pile-founded structures. The results of those tests often resulted in the redesign of the lock and dam structures, as was the case at Lock and Dam No. 25. Driven Pilings at Dam No. 25, May 1938. (U.S. Army Corps of Engineers, St. Louis District)

During the 9-Foot Channel Project, the U.S. Army Corps of Engineers also devoted a great deal of time to the design of the dam sill and apron to assure that water passing over the spillway did not carry off foundation sands. In timber-driven foundation pilings, the structure gains significant structural stability from the sands that surround the pilings. Loss of sand through erosion, frequently the result of scour, can significantly reduce the structure's stability.

In the 9-Foot Channel Project, Corps engineers designed the movable gates on each dam to include an elaborate stilling basin. These stilling basins served to control the water's hydraulic jump and dissipate its energy so that it flowed placidly downstream. The first element of the stilling basin consisted of a reinforced concrete apron located directly below the dam sills and extending 36 to 52 feet downstream. Corps engineers varied the length of the apron according to gate type and hydrological conditions. A series of concrete baffles, positioned at the upstream edge and midpoint of the apron's dissipated the energy of the water passing over the spillway. Engineers placed timber mattresses laden with derrick stone against the edge of the apron, as well as an additional 48 feet downstream, to protect the pile foundations from scour. Timber and stone mattresses protected the upstream face of the dam. The Corps also drove steel sheet pile cutoff walls above and below the dam to prevent the passage of water directly beneath the structure."

CHAPTER SEVEN NOTES

1. Tweet, Transportation on the Mississippi and Illinois Rivers, 82; Old Man River 4, No. 1 (January 1937): 15-20; 4, No. 9 (September 1937): 12-15; 2, No. 5 (December 1935): 8-13; 3, No. 8 (October 1936): 19-22; 4, No. 5, (May 1937): 16-21; 5, No. 3 (March 1938): 12-18; 5, No. 2 (February 1938): 2-21, Old Man River Safety Bulletins 1938-1940, Box 2, Entry 1626, NAKCB; "Project Information: Upper Mississippi 9-Foot Channel Project prepared by U.S. Corps of Engineers Office, Corps of Engineers, St. Paul District, St. Paul, Minnesota (n.d., n.p.) (photographs); Leonard H. Dicke, "The Upper Mississippi River Waterway," 1-5, RG77, subgroup: St. Paul District, General Records 1934-1943, 9-Foot Channel Project, Box 39, Entry 1629, File 4013.1/66 to 4013.1/86, NAKCB; Gjerde, "St. Paul Locks and Dams," 128-161 (site map and section elevation reproductions); "The Upper Mississippi River Canalization Improvement" (U.S. Army Corps of Engineers, U.S. Division Engineer, UMVD, St. Louis, Missouri, February 1938, revised May 1939, revised May 1940) 1-17; and Elliot, "Movable Gates," passim.

2. "Roller-Gate Dam Erection," 410.

3. In 1902, a Krupp roller dam was built on the Main River near Schweinfurt, Germany. Soon roller gates predominated on both the Main and the Neckar Rivers. The roller dam at Kibling, Germany, which was built before 1915, involved a particularly remarkable use of the technology. This dam had one 28-foot high roller providing a 46-foot clear span. Another interesting pre-1915 example is a dam near Stuttgart, Germany, which had two spans, 92-foot by 12-foot. McAlpine, "Roller Gates in Navigation Dams," 420; Roberts, "Kanawha River," 338; 'Building the Rolling-Crest Dam Across Grand River," Engineering News 76, No. 2 (July 13, 1916): 60; Johnson, Davis Island Lock and Dam. 162; and F. Teichman, "Large Roller-Crest Dam, Grand Valley Project, Colorado," Engineering News 76, No. 1 (July 6, 1916): 4. Although individual roller gates had been built in Europe that were both larger and longer, Dam No. 15 was the first to incorporate so many gates of such an aggregate length.

4. P.S. Reinecke, "The Rhine and the Upper Mississippi," The Military Engineer 30 (May-June, 1938): 167-171.

5. McAlpine, "Roller Gates in Navigation Dams," 420-421; and Roberts, "Kanawha River," 338, 340-342.

6. "U.S. Engineer Office, Improvement of Mississippi River, Development Near Rock Island, Illinois, Hearing on December 22, 1930," 67; Leland R. Johnson, Men, Mountains and Rivers: An Illustrated History of the Huntington District, U.S. Army Corps of Engineers (Huntington: U.S. Army Engineer District, 1977), 137-138; and Johnson, Davis Island Lock and Dam, 162.

7. Teichman, "Large Roller-Crest Dam," 4.

8. Ibid., 2-4; and "Building Rolling-Crest Dam Grand River," 60-61.

9. Roberts, "Kanawha River," 339. Vertical lift gates's name explains their technology well. The rectangular sliding gates at Dam No. 19 on the Upper Mississippi are a form of vertical lift gate. Sector gates are roller gates in which the roller is a sector of a circle instead of a cylinder. John S. Scott, A Dictionary of Civil Engineering, 2nd ed., (Baltimore: Penguin Books, 1965), 270.

10. Roberts, "Kanawha River," 340; and H. Doc 137.

11. The original German-designed gates that the Corps started with at Rock Island were those covered by the MAN Company patent as fabricated by the S. Morgan Smith Company. They were almost identical to those used at the New England Power Association's Bellows Falls Dam. H. Doc 137, 97. Discussions relevant to reasons for the modifications in these original designs are in Gross and McCormick, "Upper Mississippi River Project," 315-316; and Roberts, "Kanawha River," 340. Standard roller gate construction is covered in "Building Roller-Crest Dam Grand River," 61-63. Construction dates on Dam No. 15 are covered in Final Report Lock and Dam 15, 60. Completed on March 31, 1934, the dam was halfway done in 1933.

12. Roberts, "Kanawha River," 340-341.

13. Gross and McCormick, "Upper Mississippi River Project," 316; E.E. Gesler to Div., October 9, 1937, and E.E. Gesler to Chief, November 12, 1937, and May 28, 1938, RG77, Entry 111, Box 998, File 3524-Part 2; and RG77, Entry 111, Box 197, Envelope 7425.

14. L. Ylvisaker to Dravo Contracting Company, November 16, 1932, 1, RG77, subgroup: St. Paul District, Operations and Maintenance Files, 1931-1943, Box 395861, Entry 1626a, File 413b.3/05, NACB; Contracting Engineer, M.H. Treadwell Company to United Construction Company, January 12, 1934, 1-3, March 3, 1934, 1-4; United Construction Company to District Engineer, St. Paul, Minnesota, February 26, 1934, 1-2; RG77, subgroup: St. Paul District, Box 395861, File 413b.3, NACB; Old Man River 2, No. 5 (December 1935): 25-28, Old Man River Safety Bulletins 1938-1940, Box 2, Entry 1626, NAKCB; Lieutenant Colonel E.L. Daley to District Engineer, U.S. Engineer Office, St. Paul, Minnesota, February 20, 1934; Major Dwight Johns to United Construction Company, February 20,1934; Lieutenant Colonel E.L. Daley to M.H. Treadwell Company, Inc., February 17, 1934; United Construction Company, Winona, Minnesota, to District Engineer Office, St. Paul, Minnesota, February 17, 1934; M.H. Treadwell Company, New York, to United Construction Company, February 15, 1934; Captain Homer Pettit to United Construction Company, February 6, 1934; Lieutenant Colonel E.L. Daley to M.H. Treadwell Company, January 31, 1934; Major Dwight F. Johns to United Construction Company, February 13, 1934; Lieutenant E.L. Daley to M.H. Treadwell Company, January 24, 1934; M.H. Treadwell Company to United Construction Company, January 23, 1934; and United Construction Company to District Engineer, St. Paul, Minnesota, RG77, subgroup: St. Paul District, Box 395861, File 413.b3, NACB.

15. Elliot, "Movable Gates," 3.

16. Ibid., 8; William P. Creager, Joel D. Justin, and Julian Hinds, Engineering for Dams, 3 vols., (New York: John Wiley & Sons Inc., 1945), 3: 893; and Armin Schoklitsch, Hydraulic Structures: A Text and Handbook (New York: American Society of Mechanical Engineers, 1937), 648-657.

17. "Patent Lock Gate," The Farmer's Journal, Welland Canal Intelligence 39 (October 17, 1827); Deposition of William A. Gooding, October 19, 1850, in George Heath v. George W. Hildreth, Civil Action a-1363, Fifth Judicial District of the Supreme Court of the State of New York, Herkimer, NY; G.W. Hildreth, Canal-Lock Gate, Patent No. 1517, Patented March 19, 1840; G. Heath, Hydraulic Canal Gate, Patent No. 2393, Patented December 14, 1841; Heath v. Hildreth, Case No. 6,309, October 15, 1841, Circuit Court, District of Columbia; New York State Assembly Doc. No. 201, April 21, 1846; New York State Assembly Doc. No. 18, January 15, 1844, New York State Assembly Doc. No. 91, February 22, 1944; New York State Assembly Doc. No. 216, May 5, 1846; Johnson, Davis Island Lock and Dam, 135, 162; Mary Yeater, "Hennepin Canal Historic District," National Register of Historic Places Inventory-Nomination Form, Section 7: 2-3; and Gjerde, "St. Paul Locks and Dams," 125-128.

18. Daley, "Canalization of Upper Mississippi," 106; and Elliot, "Movable Gates," 9.

19. Schoklitsch, Hydraulic Structures, 638-656, 679-681; Old Man River 2, No. 5 (December 1935): 27; and Creager, et al, Engineering for Dams, 1: 43.

20. RA. Wheeler to Div. Engineer, August 16, 1935, RG77, Entry 111, Box 990, File 3524-part 2, WNRC; Drawings, No. M-L 18 10/39A; Annual Report 1940, 1160; Annual Report 1942, 1028; and Annual Report 1951, 1237.

21. Mississippi River Lock & Dam No. 24, Dam—General Arrangement—Tainter Gate, Drawing No. M-L 24 40/2 (December 1937); Elliot, "Movable Gates," 14; and Gross and McCormick, "Upper Mississippi River Project," 314.

22. "Final Report Laboratory Tests on Hydraulic Model of Lock and Dam No. 22, Mississippi River, Hannibal, Mo.," RG77, Entry 111, Box 179, Envelope 7245, WNRC; RA. Wheeler to Chief of Engineers, October 13, 1934, RG77, Entry 111, Box 993, File 3524-part 2, WNRC; "Mississippi River Lock and Dam No. 11, Final Report—Construction" (Rock Island: U.S. Army Corps of Engineers, Rock Island District, 1938), (hereafter referred to as "Final Report, Dam 11") 10, 53, RG77, NACB; E.E. Gesler to Chief of Engineers, January 13, 1937, E.E. Gesler to R.W. Kaltenbach Corp., April 16, 1937, and E.E. Gesler to Div. Engineer, June 26, 1937, RG77, Entry 111, Box 975, File 3524, WNRC.

23. "Final Report, Dam 11," 51, 56; and Drawings, Nos. M-L 11 58/1, 58/1A, 58/1A.1, and 58/1A.2, and M-L 18 58/2A.

24. Thorwald R. Peterson, "Replacement for Locks and Dam No. 26," The Military Engineer, 66 (1974): 287.

25. The architectural designs can also be differentiated by those structures for which the designs were completed between July 1931 and June 1934, and those completed between August 1934 and September 1936. William McAlpine and Lenvik Ylvisaker designed the first group; Edwin Abbott designed the second. C.W. Short and R. Stanley-Brown, Public Buildings: A Survey of Architecture of Projects Constructed by Federal and Other Governmental Bodies Between the Years 1933 and 1939 with the Assistance of the Public Works Administration (Washington D.C.: Government Printing Office, 1939), 509; U.S. Army Corps of Engineers, Drawings: River and Harbor Project—Mississippi River—Missouri River to Minneapolis Locks and Dams 3 through 10 (in two sheets for each complex) September 30, 1978; Jon Gjerde, HABS/HAER Inventory Cards—Locks and Dams 3 though 10, February 26, 1983; U.S. Army Corps of Engineers, "History of Construction—Dam No. 3, Red Wing, Minnesota," (n.d.); "Dam No. 4, Alma, Wisconsin," (E.J. Christenson, ed., n.d.); "History of the Construction of Dam No. 5, Mississippi River 4.8 Miles Downstream from Minneiska, Minnesota," (submitted by V.C. Funk, January 31, 1936), 77; "History of the Construction of Dam No. 8 Mississippi River—Genoa, Wisconsin—September 14, 1935—April 30, 1937," (submitted by Frank E. Rutt, Jr. Engineer, Acting Resident Engineer, n.d.), passim, St. Paul District Office Records; Wood, "A Nine Foot Depth in the St. Paul District," 6; Old Man River, January 1938, 14-16, RG77, subgroup: St. Paul District, Box 2, Entry 1626, NAKCB; and William Patrick O'Brien, National Park Service, Rocky Mountain Regional Office, field inspection, May 7-12, 1987.

26. U.S. Army Corps of Engineers, Drawing: "Mississippi River Lock and Dam Number 10—Esplanade Landscaping, February 1938, Rock Island District", No. ML10-38/3, St. Paul District Office Records, Map Division, St. Paul; O'Brien, National Park Service Field Inspection, May 7-12, 1987; Calvin D. Linton, ed., The Bicentennial Almanac: 200 Years of America (New York: Thomas Nelson, Inc., 1975), 332,347; Henry B. Ward, ed., "The American Association for the Advancement of Science: Preliminary Announcement of the Summer Meeting To Be Held in Chicago From June 19 to 30, In Connection With The Century of Progress Exposition," Science 77, No. 2003 (May 19, 1933): 463, 474; Reinecke, "The Rhine and the Upper Mississippi," 161-171; The Symbol of Arcturus official guidebook of the Century of Progress International Exposition (Chicago: A Century of Progress, Inc., 1933), passim; and Lois Craig, et al., 398-399.

27. The Illinois and Michigan Canal, completed in 1848, was a typical mid-nineteenth century canal; its locks were 18 feet wide. Mary Yeater Rathbun, The Illinois and Michigan Canal (Springfield, Ill.: Illinois Department of Conservation, 1980), 23; Ben Hur Wilson, "The Des Moines Rapids Canal," Palimpsest, April 1924, 120-130; Tweet, Rock Island District, 93-103; and Johnson, Davis Island Lock and Dam, 53-54.

28. Johnson, Louisville District, 176.

29. H. Doc. 137, 98.

30. Ibid.

31."Mississippi Lock and Dam 20; Final Report Construction," Vol. I: "Introduction and Lock" (Rock Island: U.S. Army Corps of Engineers, Rock Island District, March 1935) (hereafter referred to as "Final Report Lock 20"), 24, 48; "Final Cost Report Dam 20 Mississippi River" (Rock Island: U.S. Army Corps of Engineers, Rock Island District, June 1936) (hereafter referred to as "Final Cost Dam 20"), 1-2, RG77, Entry 81, Box 666, NACB; RA. Wheeler to Chief of Engineers, November 16, 1933, and July 23, 1934, RG77, Entry 111, Box 993, File 3524-Part I, WNRC; and Drawings, Nos. M-L 2020/1 and 40/1.

32. Mary Rathbun interview with G. Ron Clark and James Wright, U.S. Army Corps of Engineers, Lockmaster and Former Lockmaster (respectively) of Lock and Dam 20, Canton, Missouri, July 16, 1984 (hereafter referred to as Clark-Wright Interview); Inter-office Memo, May 21, 1940, RG77, Entry 111, Box 795, File 3524, WNRC; Annual Report 1940, 1160; Annual Report 1942, 1025, 1028; and Annual Report 1943, 943.

33. Annual Report 1942, 1028; and Annual Report 1951, 1237.

34. By 1951, engineers had developed a uniform remodeling program to handle the outdraft problems at all the lock systems in the district. Inter-office Memo, April 27, 1942, RG77, Entry 111, Box 975, File 3524, WNRC; Annual Report 1942, 1028; and Annual Report 1951, 1237.

35. H. Doc 137, 6, 21.

36. The final construction reports mention these being added at Locks 16, 18, 11, 21, 22, and 12—that is, all those in the district built on piles except Lock 17. It is not clear if the struts were not used at Lock 17 or were simply not added in. Lock 17 was the last lock constructed in the district. By then the struts may have been a standard feature drawn in at the initial design state and not meriting special mention. For a good description of the struts and their function see "Mississippi River Lock and Dam 21 Final Report-Construction," Vol. I: "Introduction and Lock" (Rock Island: U.S. Army Corps of Engineers, Rock Island District, August 1939), (hereafter referred to as "Final Report Lock 21"), 8-9, RG77, NACB.

37. R.A. Wheeler to Chief of Engineers, May 3, 1934, RG77, Entry 111, Box 997, File 3424, WNRC.

38. L.E. Wood, "Historical Sketch: Construction of L/D No. 3," Old Man River 5, No. 3 (March 1938): 12-18; and Old Man River 2, No. 5 (December 1935): 27, Old Man River Safety Bulletins 1938-1940, Box 2, Entry 1626, NAKCB.

39. William Z. Lidicker, "Mississippi River: Lock and Dam No. 3," Old Man River 4, No. 9 (September 1937): 12-15, Old Man River Safety Bulletins 1938-1940, Box 2, Entry 1626, NAKCB; and Lidicker, "Unusual Timber and Steel Bearing Pile Load Tests at Mississippi River Dam No. 3" (July 26, 1939), 1-13, RG77, subgroup: St. Paul District, General Records 1934-1943, Documents and Publications: Authorities for Publications of Articles, Box 22, Entry 1629, File 54.7, NAKCB.

40. Herbert G. McCormick and John W. Dixon, "Mississippi River Cofferdams," The Military Engineer 28, No. 158 (March-April 1936), 105-08; Harry Carlson, Report on Test of Relation Between Probe Resistance and Soil Density" (U.S. Army Corps of Engineers, St. Paul District, 1938), 1-5; Carlson, "Report on Trial Mixes for Dam Number 3" (U.S. Engineering Laboratory, Fountain City, Wisconsin, n.d.) 1-6, "Report on Investigations of Foundation Soil Densities at Dam No. 3" (U.S. Army Corps of Engineers, St. Paul District, 1937), 1-6, and E.J. Christenson, three memoranda: "Preliminary Report(s) on Load Test Nos. 3, 4, 5, Dam No. 3," June 23, 1937, 1-2, 1-2, 1-3, RG77, subgroup: St. Paul District, Operations and Maintenance Files, 1931-1943, Box 395851, Entry 1626a, File 313b.3, NACB; R.R. Philippe, memo: "Foundation-Lock No. 3," June 12, 1936, 1-6, RG77, subgroup: St. Paul District, Operations and Maintenance Files, 1931-1943, Box 395848, Entry 1626a, File 313a.3, NACB; "Discussion of Trial Mixes," (typewritten, n.d.), 1-42, RG77, subgroup: St. Paul District, Operations and Maintenance Files, 1931-1943, Box 395871, Entry 1626a, File 505.1, NACB; and Major Dwight F. Johns to District Engineer, Huntington, West Virginia, June 27, 1934, Lieutenant C.T. Hunt to District Engineer, St. Paul, Minnesota, June 14, 1934, and "Report on Tests on Temperature Rise In Concrete Conducted At London Lock and Dam, Kanawha River, Huntington District," (n.d., typewritten) 1-9, RG77, subgroup: St. Paul District, Operations and Maintenance Files, 1931-1943, Box 395857, Entry 1626a, File 405.1/11-4, 4/1, NACB.

41. Francis A. Landrieu, Ralph D. Salisbury, Mark Haima, George E. Oliver, and Lee H. Baron, "History of the Construction of Lock 6—Mississippi River—Trempealeau, Wisconsin—P.W.A. Project 11," (U.S. Army Corps of Engineers, April 1935, typewritten) 16-17, St. Paul District Office Records, Saint Paul, Minnesota.

42. Thomas J. Mudd, "Locks and Dam No. 26, Mississippi River—Alton, Illinois" (St. Louis: U.S. Army Corps of Engineers, St. Louis District, 1975), 3; and "History and Cost Report Lock 24, Mississippi River" (St. Louis: U.S. Army Corps of Engineers, St. Louis District, 1938), 1, typescript draft on file in St. Louis District Office.

43. Mudd, "Locks and Dam No. 26," 3-4; and "Final Report—Lock and Dam No. 26, Part I—Locks" (St. Louis: U.S. Army Corps of Engineers, St. Louis District Office) (hereafter referred to as "Final Report—Locks and Dam No. 26, Part I"), 16, 32, 34, typescript draft on file at St. Louis District Office.

44. Mississippi River Lock & Dam No.26, Dam—General Plan, Drawing No. M-L 2640/1 (December 1934); Mudd, "Locks and Dam No. 26," 3; Mississippi River Lock & Dam No. 25, Dam—General Plan, Drawing No. M-L 25 40/1 (January 1937); and Mississippi River Lock & Dam No. 24, Dam—General Plan, Drawing No. M-L 2440/1 (December 1937).