Chapter 5
Hydraulics Research Giant, 1949-1963, Part I: River Modeling, Potamology, and Hydraulic Structures   back arrow

Expanded Functions and Facilities

In the aftermath of war, hydraulics research at WES returned largely to traditional civil functions: flood control, river and harbor improvement and regulation, and design of hydraulic structures. Activities in these and other areas such as tidal estuary modeling and potamology increased far beyond prewar efforts. In addition, for the first time the Station assumed a program of applied research that, rather than concentrating on specific problems connected with individual projects or structures, appreciably broadened the scope of experimental work. Military endeavors, though greatly reduced, found an unexpected release in studies of the effects of nuclear explosions in water.

Aerial view of large metal hangars at WES 
to house models As its workload evolved, the Station went through a major physical renovation. In 1946 the acquisition of adjoining property increased usable acreage by more than half. Extensive clearing and grading and placing of compacted fill in gullies provided broad level areas for new model shelters and soils studies, while widened paved access roads improved internal communications. To serve Corps personnel in the Vicksburg area, where housing had been a critical problem in the prewar period, the MRC directed construction of a dozen single family homes, an apartment complex, and a small (and unpopular) trailer camp. Clearing underbrush from around the Station’s lake improved its appearance, and grounds keepers waged a constant battle to clear the lake of turtles. One weekly “turtle report” to the Director listed a three-foot alligator among its victims. Snakes occasionally found models to be convenient resting spots.1

More importantly, new facilities replaced the outdated wooden sheds that housed numerous hydraulic models. Station Directors found that portable sheetmetal hangars, with materials obtained from military surplus, were both ideal and available. Through the late 1940s WES constructed a number of huge hangars on the graded upper level of the reservation adjacent to the Mississippi River Flood Control Model. By 1947 the center of activity for model studies had been moved from the lower level in front of the WES main building to new facilities on high ground. While the Hydraulics Division relocated most of its operations there, the Soils Laboratory occupied the central portion of the old main building, which had housed flumes, small models, and other hydrodynamic experimental equipment.

More Growing Pains

Henry B. SimmonsGrowth inevitably altered relationships between WES employees. “Old hands” nostalgically recalled that the Station in its earlier years had benefitted from an extraordinary camaraderie. Its predominantly young cadre of engineers and technicians had been bound together by common experiences brought on by the Depression and World War II. Excited by the prospects of innovative scientific research, they maintained a level of interaction possible only in a limited, almost family-like atmosphere. Henry Simmons, a WES employee since early 1940, noted of the World War II era that “In those days everybody ate lunch together out on the lawn out of brown paper bags,” and that there was constant discussion about what was going on. Everybody knew everybody else — including the director — by first name. Simmons also represented a generation of WES employees who found it possible to rise to prominence despite a lack of formal education. Later an internationally known pioneer in estuary modeling and WES Hydraulics Laboratory Chief, Simmons had dropped out of Mississippi State College for financial reasons halfway through his senior year. He never received a college degree.2

A listing of WES hydraulics projects in progress in 1950, the 20th anniversary of experimental work at the Station, illustrates the remarkable growth of the Hydraulics Division’s activities during the previous decade. It also serves to indicate the difficulty of maintaining the close personal relationships that had benefitted the Station’s personnel during its early years. With growth came distance.

From 23 projects in progress in 1940, the Division’s workload had almost doubled to 42 in 1950. (A complete listing of projects in progress in 1950, with sponsors, is included in Appendix B.) Site-specific studies of dams and appurtenant structures included work for Belton Dam, Texas; Cheatham Dam, Tennessee; Folsom Dam, California; Fort Randall Dam, South Dakota; Garrison Dam, North Dakota; Genegantslet Reservoir, New York; Oahe Reservoir, South Dakota; and Philpott Dam, Virginia. River flood control and navigation efforts involved constructing and instrumenting the huge Mississippi Basin Model and model studies of Memphis Harbor, Tennessee; the Hoosic River, Massachusetts; and the Mississippi River near the Greenville, Mississippi, bridge. Another river project, begun for the Buffalo District, entailed model studies of the Niagara River and Niagara Falls necessary to design improvement and protective works. Large-scale models of Charleston Harbor, South Carolina; Delaware River Estuary; Grays Harbor, Washington; Raritan River Estuary, New Jersey; Savannah Harbor, Georgia; and Port Washington Harbor, Wisconsin, exemplified tidal and wave action research activities.

By 1950 OCE had for the first time also involved WES in a number of applied hydraulics research projects as a part of centralizing its research programs. Civil Works Investigations (CWI), unlike site-specific studies, involved broad research initiatives for general application. CWI projects in progress at the Station in 1950 represented a broad range of the Corps’ hydraulics engineering mission, including: General Spillway Model Tests, Conduit Intake Model Tests, Cavitation Research, Sluice Outlet Model Tests, Model Study of Sluice Coaster Gate, Slide Gate Model Tests, Use of Air Instead of Water in Model Testing, Scale Effects on Spillway Discharge Coefficients, Hydraulic Capacity of Meandering Channels in Straight Floodways, Study of Wave Force on Breakwaters, Stability of Rubble-Mound Breakwaters, Study of Harbor Design, Scale Effects in Harbor Models, Analysis of Hydraulic Experimental Data, Effects of Model Distortion on Hydraulic Elements, Simulation of Air Entrainment in Models Involving High Velocity Flow, Hydraulic Instrumentation, Development of Turbulence Meter, Prototype Analysis, and Roughness Standards for Hydraulic Models.3

Administrative Evolution

Eugene P. Fortson Despite its enlarged mission and the plethora of projects performed after World War II, for nearly two decades the administrative structure of the Hydraulics Division remained comparatively stable. Fortson served as Chief until 1970 with the exception of a stint in Korea from 1951 to 1952. From 1947 until 1951 the division consisted of only three branches:

John J. Franco Thomas E. Murphy Frank B. Campbell George B. Fenwick In 1951 the Hydraulics Analysis Branch under Frank B. Campbell joined the existing three, but the following year the Rivers and Harbors Branch absorbed the Mississippi Basin Model Branch, restoring a three-branch structure. This arrangement lasted until 1962 with Fenwick, Brown, and Campbell continuing as chiefs. In that year Fortson added the Nuclear Weapons Effects Branch under Guy L. Arbuthnot, Jr. The Nuclear Weapons Effects Branch in turn left the Hydraulics Division in 1963 to become a separate division with Brown as Chief. The Hydraulics Division then expanded to five branches:

This basic structure continued into the 1980s. (See Appendix A: Organization Charts.)

The Korean War

Peace in 1945 did not bring international stability. By 1948, as a reinvigorated WES faced its enlarged mission, the Grand Alliance between the Western democracies and Communist Russia had collapsed. The Soviet occupation of Eastern Europe and subsequent clumsy attempts to force the Western powers out of Berlin marked the onset of the Cold War. Within three years after the defeat of the Axis powers, the United States and the Soviet Union faced off as antagonists in a bipolar world. China’s fall to communism in 1949 further intensified the confrontation between East and West.

Over 100 WES employees saw service in 
Korea with the 434th Engineer Battalion The Cold War turned temporarily hot in June 1950 when Communist North Korea invaded South Korea. The United States responded immediately, sending regular Army troops to the Theater of Operations and calling up reserve units. One such reserve unit was the 434th Engineer Construction Battalion, established and sponsored by WES in 1948. Using WES facilities for training, the unit specialized in bridge and road building and repair. About seventy WES employees, many from the Hydraulics Division, reported for active duty in August 1950, then left for further training in Colorado before sailing for Korea in January 1951. Eventually over 100 Station personnel served with the 434th, while others joined different units. The drastic drop in manpower disrupted numerous WES functions, as did attendant sharp cuts in Federal expenditures for civil works.

Fortson, who had served with the 3rd Military Railway Service in Iran during World War II, commanded the 434th in the Theater of Operations from February through December 1951. As in the first conflict, Fenwick replaced Fortson as division chief until the latter’s return in 1952. John J. Franco, another longtime Hydraulics Division employee, World War II veteran, and future Waterways Branch Chief, succeeded Fortson as battalion commander in Korea until June 1952. He was one of the last of the original troops to return to Vicksburg. The 434th performed with distinction in its wartime role, receiving a Meritorious Unit Commendation from Eighth Army Commander General James A. Van Fleet.

MBM: Administration and Construction

Construction and operation of the Mississippi Basin Model (MBM) more closely represented a return to the Station’s original flood control mission than any other project of the immediate post-World War II era. To administer the “supermodel,” OCE in August 1945 established the Mississippi Basin Model Board to “determine policies and programs for the subsequent development and operation” of the giant facility. Members were the president of the MRC, the WES Director, the Missouri River, Ohio River, Southwestern, and Upper Mississippi Valley Division Engineers, and a representative of the Chief of Engineers. Functioning until 1970, the Board usually met on an annual basis, with subcommittees meeting more often during the interim. Subcommittees submitted numerous reports to the Board, which then produced detailed reports on its meetings and made recommendations pertaining to construction and use of the model.4

In 1946, within the WES administrative structure, Station Director Colonel Carroll T. Newton established a MBM Branch under Franco. The following year Dewey succeeded Franco. Dewey reported for active duty in August 1950, serving in Korea until late 1952. He then returned to WES and briefly assumed his former position as director before leaving the Station to head the Corps’ model of San Francisco Bay in Sausalito, California. McGee guided the MBM Branch during Dewey’s absence in Korea. On Dewey’s permanent departure from WES, Fenwick’s Rivers and Harbors Branch absorbed the MBM Branch, which was reduced to section status.

Margaret S. Petersen One of the Station’s most notable hires assigned to the MBM was Margaret S. Petersen. Shortly after graduating from the University of Iowa in civil engineering, Petersen and her best friend and fellow Iowa graduate Irene Miller responded positively to job offers from Tiffany. (The pair had gotten a total of two job offers – one from WES, the other from the library of a paper company in Wisconsin.) Although they “didn’t have the faintest idea where Vicksburg was,” and “couldn’t find it on the maps,” the two arrived at WES in August 1947. From August 1947 until November 1949 Petersen reviewed and analyzed hydraulic, topographic, and hydrographic data for designs of proposed testing programs and model-operating techniques. She also was in charge of several research projects while teaching courses in fluid mechanics for the WES staff. On the latter date Petersen became one of the first and only female administrators in the Corps, rising to the office of Chief, Research Sub-Section, Mississippi Basin Model Operation Section, WES.5

In 1952 Petersen and Miller applied for leaves of absence without pay to pursue graduate studies at Iowa. However, many of the Hydraulics Lab’s employees were expected to return from the Korean War in the near ivefuture and assume their prewar positions. Because there was not a lot of work available and funding was insecure, the WES personnel office refused to grant a leave of absence to Petersen and Miller. Both resigned from the Station in order to return to school. Petersen went on to a stellar career with the Corps, including a short stint back at WES in 1964 during which she served at Chief, Wave Dynamics Division, Water Waves Branch, Hydraulics Laboratory. After retiring from the Corps in 1977 she served as an associate professor in the Department of Civil Engineering and Engineering Mechanics at the University of Arizona. Her River Engineering, published in 1986 became a standard work in the field. In 1995, looking back on her years at WES, Petersen opined that her experiences at the Station had been invaluable due largely to the wide variety of hydraulic engineering work in progress and, more so, to the “very competent” Hydraulics Lab employees and consultants that served as role models.6

Map of area 
reproduced by Mississippi Basin Model and drainage basin of Mississippi RiverAfter prisoners of war prepared most of the model site by 1946, WES crews directed by Dewey began construction of the concrete portion of the model proper. The original WES Definite Project Report, submitted to OCE in April 1943, predicted model completion about one year after site preparation. This overly optimistic calculation may well have been based on previous WES experiences, such as construction of the Mississippi River Flood Control Model in 1935. In that instance WES had built the giant edifice — then the world’s largest hydraulic model — in a mere four months.

Mississippi Basin Model constructionMBM construction, in fact, continued at a varying pace until 1966. The long building process was due primarily to irregular funding and, less so, to unforseen technical problems. Funding went through an unusual and often disruptive sequence. WES in the beginning charged construction, operation, and maintenance costs to the divisions having sections to be reproduced in the model. The divisions in turn prorated costs to their districts through complicated formulas. By the mid-1950s this method had become so unreliable that Congress took over funding. Beginning in Fiscal Year 1957 direct Congressional appropriations covered construction and verification of the model. Amounts ranged from a low of $400,000 in that year to $810,000 in Fiscal Year 1958.

Completed Mississippi Basin ModelThe sheer size of the MBM presented new construction demands. Covering an extensive area, over 200 acres, it had to be built in sections with expansion joints between them to absorb the expansion and contraction of the concrete. Areas between concrete sections were sodded to prevent erosion. Until 1953 workers constructed individual section blocks directly in place on a carefully prepared subgrade using the traditional template method. This involved cutting sheetmetal templates to cross sections obtained from topographic maps and set about 2 ft apart to the correct elevation on the model site in positions located by a rectangular grid system. Crews then placed concrete between the templates and molded it to correct elevations. Highly expansive clay beneath the model caused undue shifting and by 1953 WES engineers developed a contour method for completion of the rest of the model. Using enlarged contour maps for construction plans of sections to be molded, technicians fabricated the sections on an assembly line. Section blocks then cured for seven days before being carried by truck to the model site and set on concrete piles. The piles extended to a depth of 10 feet, passing through the expansive clay stratum to where the soil moisture content was stable. Although slower and more expensive, this provided the necessary equilibrium.

Reproduction of model details also consumed substantially more time than previous efforts. The MBM eventually included all bridges, levees, highway and railroad fills, and other pertinent elements of the extensive prototype. Corps districts furnished information so Federal and many private levees could be modeled to location with proper heights, grades, and alignments, while highway departments and railroad companies provided data necessary to simulate their construction efforts. Channel and overbank roughness required further refinement. WES engineers working with a pilot model determined that carefully brushed and scored concrete, concrete ridges, and concrete and brass parallelepipeds (usually cubes), properly spaced, could accurately simulate channel roughness. To replicate overbank phenomena such as trees, workers installed folded screen wire cut to the scale of the average height of trees where aerial photographs showed trees in the prototype. Expanded metal fastened to the model on cleared areas provided adequate resistance to flow where the brushed concrete was not effective, with some areas requiring two or three metal layers.

MBM Instrumentation

MBM Instrumentation MBM designers from the earliest stages of its development considered automatic instrumentation desirable and necessary. Total manual operation of the model would require a full-time staff of about 600, which would be unduly expensive and difficult to train and maintain. Available manually controlled instruments also could not accurately reproduce and chart some complex Mississippi Basin phenomena. WES devoted approximately four years, from 1943 to 1947, to the study of automatic instrumentation and to the testing of commercial and pilot instruments embodying various design principles. After investigating products of about 125 manufacturers throughout the United States, model engineers determined that available instruments did not have the accuracy of measurement or the range required for use on the model. Consequently, WES developed specifications for new instruments and invited manufacturing companies to bid or to submit designs for alternatives that would accomplish the desired results. In 1948 contracts awarded to Infilco, Inc., Chicago, and Leupold and Stevens Instruments, Portland, led to production of the necessary automatic devices.7

MBM automated control centerThe MBM incorporated three types of automatic instruments: inflow, stage, and outflow. A single master timing device synchronized the operations of all. Centrally located instrument houses on the major streams provided control centers. Programmers located in the control houses could regulate the introduction of water into the model through the inflow controllers, monitor water levels with stage devices, and measure discharges at selected points on a stream with outflow instruments. Automatic recorders, also in the control houses, made permanent records of data received electrically through transmitters, in addition to registering the month, day, and hour of model flood periods as determined by the master timer.8

WES engineers continued to make improvements in instrumentation while the model was being constructed. Use of automatic instruments was a complete success. Hundreds of tests indicated that automated experiments were typically more accurate than manually conducted ones, especially when identical tests were repeated. Rather than the 600 personnel needed to manually run the MBM as originally projected, the new instruments required only 60. Savings were about 50 percent, as the cost of the instruments in use on the model was offset by the savings in salaries required for manual operation.9

The 1952 Flood

Piecemeal construction of the MBM had certain benefits. Partly at the insistence of Station Director Linder, WES sectionalized the model so various parts could be used for separate studies without involving the entire facility. Portions of rivers such as the Missouri, Tennessee, or Arkansas could be tested independently as they were completed to study local problems. Thus the model was “many models in one.” As individual sections were completed, engineers installed instruments, added roughness and other features, and verified them for local testing. Verification usually involved reproduction of the maximum flood of record for each reach of the model. After introduction of model equivalents of the flows for that flood, crews adjusted channel and overbank roughness to make model stages agree with those of the prototype. With the model thus adjusted to reproduce past occurrences, it was expected to produce occurrences that might be experienced in the future. Construction and verification generally proceeded geographically from north to south, since the Mississippi Flood Control Model at the Station already covered most of the river reach below Memphis.10

MBM directors put some individual sections into operation as early as 1949. Fortuitously, by 1952 the Missouri River segment was fully operational. In April of that year a great flood threatened to reach a crest discharge of almost twice the maximum flood of record at Sioux City, Iowa. The Missouri River Division (MRD) requested WES to operate the MBM on a 24hours-a-day basis to assist in predicting crest stages and discharges. During a critical 16-day period, WES and MRD personnel maintained almost constant contact by telephone. Model tests first indicated that levees at Council Bluffs and Omaha were not high enough to contain the crest of the flood. The MRD subsequently evacuated the affected areas and raised the levees five days after the WES prediction. Flood crests exceeded the original grades of the levees but were contained by the recent additions, saving the two cities from flooding. Further tests indicated that certain levees should be raised but that others would be overtopped before efforts would be effective. The MRD then concentrated on building up levees where there was a chance to prevent overtopping and evacuating people from other locations. This spectacular use of the model, according to the MRD, was a prominent factor in the success of the flood fight that prevented damages of an estimated $65 million.11

Basin-Wide Testing Program

By 1959 the model had been completed downriver as far as Memphis. MBM directors then began a comprehensive testing program that coordinated the entire model structure as it expanded. Further construction by 1966 added the Mississippi River from Memphis to Baton Rouge, the Arkansas River, and the Atchafalaya River basin. Basin-wide tests, which continued through 1969, concentrated on analyzing the effectiveness of reservoirs in controlling floods and in developing procedures to obtain the greatest overall flood protection. Further tests were to determine the efficiency of Corps plans for operation of floodways and to check the adequacy of project levees in the Lower Mississippi River Valley.12

The lengthy test series reproduced four historic floods, 1937, 1943, 1945, and 1952, for which adequate data were available. Each had different characteristics and represented various ways in which floods occurred on the Lower Mississippi. The 1937 flood, for example, was the maximum flood of record on the Ohio River and portions of the Lower Mississippi, while the 1943 and 1945 events saw flooding of the Missouri River as well as the Ohio and Mississippi. Tests also reproduced three hypothetical floods representing early spring, late spring, and winter phenomena. Procedures involved introducing model equivalents of the flood flows at 114 model inflow points on the major tributaries and routing flows through the model to the downstream end at Baton Rouge or through the Atchafalaya Basin. All seven of the model floods were reproduced six times, each time with a different set of variables, most of which pertained to operation of existing or planned reservoirs.

MBM Sightseeing Facilities

Mississippi Basin Model visitorsDuring its period of basin-wide testing the MBM gained international renown as a tourist attraction. Beginning in 1964 visitor facilities provided self-guided tours on a seven-day-a-week basis. Facilities included a visitor assembly center, a 40-foot-high observation tower, an operation observation room near the center of the model, and elevated platforms, walks, and sidewalks at selected locations throughout the area. Maps, pictures, other visual aids, and recorded lectures provided visitors with information about the model and other work done by WES and the Corps of Engineers. Through the remainder of the decade the model drew about 5,000 visitors a year, including domestic and foreign engineers, Corps officials, and sightseers in general. Perhaps more than any single construction project completed by the Corps of Engineers — and certainly by WES — the MBM brought public attention to the development and possibilities of hydraulic modeling.

MBM Retired

With its basin-wide testing program completed in 1969, the MBM no longer had a clear mission. Tests on individual problems continued into 1971, but high operating costs and declining demand for conventional model studies, largely due to the use of computers to replace or complement model investigations, led Corps leaders to put the MBM on a standby basis.

MBM revived during 1973 Mississippi River floodThe Corps found one more use for the mighty model in 1973. In the fall of 1972 heavy rains in the Mississippi River basin saturated the ground and filled flood control reservoirs. By the following April the Lower Mississippi River experienced its largest flood in decades. A potentially disastrous situation arose at the Old River control complex when a wing wall failed and a large scour hole developed in front of the structure. Failure of the entire structure would have resulted in the Mississippi taking a new main channel down the Atchafalaya basin, bypassing Baton Rouge and New Orleans.13 MRC President Major General Charles C. Noble considered opening the Morganza Floodway, which had been completed just downriver from the Old River complex in 1953. However, the floodway had never been operated and serious questions arose concerning the impact of its use on the Atchafalaya basin and whether it would divert polluted water through Baton Rouge and New Orleans. Noble consequently requested reactivation of part of the MBM for tests. Despite having only two remaining full-time staff, the MBM was operational within 48 hours. Tests performed on an around-the-clock basis over a three-week period indicated that opening the Morganza Floodway would improve conditions at the Old River complex without endangering water supplies at Baton Rouge or New Orleans. Also, the veteran model again showed the ability to pinpoint levees that were in danger of overtopping.14 (The emergence of numerical modeling and the expense of maintaining the MBM, led the Corps to finally relinquish control over the facility. In 1993, the city of Jackson took custody of the MBM.)

Niagara Falls and St. Lawrence Seaway Projects

With the major exceptions of the MBM and the Mississippi River Flood Control Model, WES activities involving river models declined precipitously in the years following World War II. This indicated the Corps’ shifting priorities from flood control, which had progressed geometrically since the early 1930s, to other engineering areas. Two notable WES projects used river models in attempts to preserve and enhance the beauty of Niagara Falls and to facilitate construction of the St. Lawrence Seaway. Rather than flood control, both centered on aesthetic, industrial, and navigational factors, and both were fraught with political as well as technical difficulties.

Since 1877 both the United States and Canada had diverted water from the Niagara River above Niagara Falls to produce electric power. Treaties in 1909 and 1910 limited diversions to daily quotas for each country to preserve the scenic beauty of the falls. In order to meet wartime power needs during World War II, a further series of agreements increased diversions. Fearful that continuing diversions would diminish the attractiveness of the falls, yet highly dependent on diversions for power, the United States and Canada in 1950 signed a comprehensive new treaty to regulate use of the Niagara River. Although power requirements were an important consideration, the treaty specified that the primary obligation of the two governments was to “preserve and enhance the scenic beauty of the Niagara Falls and River,” and stated that the two countries would complete any remedial works necessary to distribute water so as to produce an unbroken crest line around the falls for all flows.15

Administration of the treaty fell to an International Joint Commission, which in turn established the International Niagara Falls Engineering Board. The Board, drawn from technical agencies of the two countries involved, was to undertake an engineering investigation of the Niagara River and Falls and to make recommendations to the parent organization. The Buffalo District and the Federal Power Commission furnished experts for the United States. In addition to beginning onsite investigations in 1950, the Board called for model studies. Since the project was international and inherently sensitive, both countries constructed models, ostensibly to complement one another. Sponsored by the Buffalo District, in late 1950, Fenwick, Earnest B. Lipscomb, Robert G. Cox, and Cody D. McKellar of the Rivers and Harbors Branch began design and construction of a model at WES. At the same time, the Hydro-Electric Power Commission of Ontario built a second model at Islington, Ontario, although substantially smaller and with different scales than its WES counterpart.16

Reproduction of the prototype area involved a set of problems not encountered in earlier river models. Unusually swift currents, eddies, the dramatic drop in elevation from Lake Erie to the stretch below the falls, the large volume of water cascading over the falls, and the presence of water intakes for power plants were alien to more traditional studies. In addition, the lack of hard data concerning channel depths, current speeds, and other phenomena that the model had to incorporate forced project field engineers to invent new investigative methods. These included suspending weights on steel cables from helicopters in order to determine water depths and tracing current directions by studying aerial photographs of ice flows. Model verification further required uncommon adjustments for roughness — added with wire screening, stucco, sheet metal, and small rocks — to reproduce turbulence in the vicinity of the falls.17

The Niagra Falls model Upon completion and verification, the indoor WES model covered an area nearly the size of a football field, representing part of Lake Erie, 26 miles of the Niagara River, Horseshoe and American Falls, and the scenic stretch approximately one mile below the falls. Even bridges and the proposed and existing power intakes were precisely incorporated. In 1951 a first test series determined that increased water diversions for power, as allowed by the 1950 treaty, would result in “intolerable” effects on the falls if no remedial works were constructed. This led to evaluation of several proposed plans of remediation. WES tests indicated that the key element in maintaining adequate flows would be a 1,705-ft-long gated control structure built into the river from the Canadian shore above the falls. Based on these model studies, in 1953 the International Joint Commission adopted a scheme calling for construction of the control structure, but reduced its length to 1,550 feet. However, the structure was to be designed so additions could be made if needed. Other hydraulic adjustments included extensive excavations and fills on both flanks of Horseshoe Falls. Model tests also showed that a proposed 450-foot-long gated structure near the U.S. shore was not necessary.18

By 1957 the International Commission had supervised construction of the massive control structure. Although results were generally quite good, flow levels at the falls were at times somewhat less than anticipated. Proposed design revisions for power intakes above the falls were also a source of some concern. Consequently, in 1959 WES conducted a follow-up study to evaluate the effectiveness of the prototype gated control structure. Tests reinforced the WES recommendation that the edifice extend 1,705 feet into the channel rather than the 1,550 feet originally accepted by the commission.19 Upon completion, the addition allowed adequate flow diversions for hydroelectric power while maintaining the falls as one of the great scenic wonders of the world.

Barnhart Island - Lake St. Francis Reach, St. 
Lawrence Seaway model outh Cornwall Channel, St. Lawrence Seaway 
indoor navigation channel Construction of the St. Lawrence Seaway, another joint effort by the U.S. and Canadian governments, benefitted from WES model studies.20 Plagued by controversy and opposed by various commercial elements such as railroads and East Coast ports, technical planning for the massive project began in 1940. In that year the Corps of Engineers established the St. Lawrence River District solely to carry out survey work and to submit plans for features to be included in the vital International Rapids Section. A 46-mile-long stretch forming part of the border between New York State and Ontario Province presented by far the most challenging engineering and political problems. In 1942 the St. Lawrence River District submitted an extensive design memorandum for the river reach that the Corps accepted as the basis for any subsequent action. OCE then dissolved the St. Lawrence River District and delegated its responsibilities to the New York District.21

Both American and Canadian engineers concluded that model studies would be necessary to determine optimum designs in several areas. In the Galop Rapids Reach, a notoriously treacherous stretch near the head of the International Rapids Section, the St. Lawrence District had actually submitted two designs that differed substantially. Both involved extensive excavation of a navigation channel, relief cuts, addition of structures to reduce river velocities, and removal of various existing structures such as dikes and locks. The New York District asked for model studies to test the effectiveness of both plans, a project that OCE assigned to WES in early 1943. In June of that year Fenwick spent two weeks at the St. Lawrence site conferring with project experts and studying surface currents. He discovered several discrepancies between Corps drawings and actual prototype behavior and also noted that river maps did not show numerous eddies and other local flow characteristics. In a relatively rare occurrence, human experience and observation substituted for technical expertise: Fenwick and the New York District Engineer relied heavily on a local commercial fisherman, Ed LaFlair, to furnish details concerning the river bottom and surface flow patterns. At Fenwick’s insistence, LaFlair spent several weeks at the Station as a consultant.22

On his return to WES, Fenwick supervised construction of a fixed-bed model of the Galop Rapids Reach with Shields E. Clark, Jr., as project engineer. Part of the model consisted of removable and interchangeable concrete blocks so different river conditions and alterations could be tested without breaking and remolding sections of the model. Changes were so extensive that the concrete blocks were soon abandoned in favor of a soil-cement mixture soft enough to be carved to desired configurations but hard enough to resist erosion or deformation. Early results of the two-year testing series, completed in October 1945, indicated that the first St. Lawrence District design was unworkable because it would not effectively reduce river velocities. The District’s alternate plan also had serious deficiencies. WES studies then went far beyond their original intent of testing the two proposals, resulting in numerous changes. This led to the Corps’ acceptance of a WES-developed revised alternate plan ultimately implemented by the St. Lawrence Seaway Development Corporation in the Galop area.23

Lipscomb supervised further St. Lawrence tests from 1955 to 1958, requested by the Buffalo District while the Seaway was under construction. A 1956 series involved reproduction of a 4-mile section of the Long Sault Canal, including the massive Eisenhower Lock and Grass River Lock, to determine the effects of surges in the canal between locks.24 Other investigations used two models of the highly complex Cornwall Island Reach near the lower end of the International Rapids Section. A small, detailed model dealt only with cross-currents between Barnhart Island and Cornwall Island, while its larger outdoor counterpart miniaturized several river miles above and below Cornwall Island. Tests in both utilized a remote controlled replica of a large ore-carrier-type ship such as those used on the Great Lakes. As a result of WES efforts, the Corps recommended adjustments to the river channel that would divert more water to and require more dredging on the Canadian side of Cornwall Island.25 In one of the most controversial issues of the entire Seaway project, the Canadians rejected U.S. recommendations, largely for political reasons. The Corps eventually accepted the Canadian position, a decision that led to substantial difficulties in implementation.26

The Old River Dilemma

While the MBM took shape and Niagara Falls studies were completed, WES hydraulics engineers and geologists became acutely concerned with conditions that could potentially alter the entire regimen of the Lower Mississippi River. Harold N. Fisk’s geological reports on the Lower Mississippi River Valley, conducted under WES auspices in the 1940s (discussed in Chapter 4), brought intensified attention to an old problem. In the early 1800s the Mississippi River channel followed a large meander to the west near Angola, Louisiana, called Turnbull’s Bend. It was located about 300 river miles from the mouth of the Mississippi at Head of Passes and 80 miles upriver from Baton Rouge. The Red River flowed as a tributary into the Mississippi channel at the upper west end of the bend.27 At the lower west end of the bend the upper end of the Atchafalaya River trickled into the Mississippi. The Atchafalaya channel stretched lazily south to the Gulf of Mexico near Morgan City, Louisiana. During high water periods the Mississippi reversed the flow of the upper Atchafalaya, turning the smaller stream into a distributary.28

Old River, after Shreve's Cutoff, became a 
distributary channel of the Mississippi River as the Atchafalaya River and threatened to capture the 
Mississippi's flow In 1831, Shreve ordered a channel dug across the narrow neck of Turnbull’s Bend, eliminating the meander. The Mississippi immediately took the new shortcut and the upper channel of the bend dried up. However, the lower channel of the bend, which connected to the Atchafalaya, continued to flow. This vestigial link became known as Old River. By the late 1800s water from the Mississippi flowed more regularly through the Old River channel into the Atchafalaya, even at normal river stages, eventually converting the smaller stream into a permanent distributary.29

The MRC expressed strong concerns about the Mississippi-Atchafalaya connection in the late 1880s. Even earlier, some river observers had speculated that the Atchafalaya might eventually capture the main channel of the Mississippi.30 Flow of the Mississippi into the Atchafalaya continued to increase into the 20th century. In 1932 one of the first WES hydraulics reports warned of the inherent danger of diversion.31 Still, it was Fisk’s 1944 report that brought the problem into focus. For the first time, Fisk interpreted Mississippi River diversions in detail based on geological studies of floodplain features. He concluded that major changes in the river’s channel had occurred when an actively meandering loop of the Mississippi, such as Turnbull’s Bend, reached an adjacent flood basin, such as that of the Atchafalaya, and intersected a tributary stream that had a channel capable of carrying low-stage flow from the big river. Fisk’s data also indicated that, historically, the flow of the Mississippi had gradually shifted from an existing main channel to such a new course over a span of no more than 100 years. A critical period seemed to be reached when a distributary captured about 40 percent of the Mississippi’s flow, after which diversion was inevitable. Thus if geological analyses of previous diversions were accurate, the Atchafalaya was an ideal prospect for a new main stream in the relatively near future.

Old River Geological Investigations

By the mid-1940s the MRC, the Corps, and others had reached the point of alarm. Readings indicated that discharge from the Mississippi into the Atchafalaya was swelling at a disturbing rate. In the meantime, WES geological operations had moved from their original base in Baton Rouge to the Station, where in 1948 Station administrators established a Geology Branch as part of the Soils Division. Most of Fisk’s WES contingent in Baton Rouge then relocated to the Station.32

With Fisk as a consultant, WES conducted more detailed investigations of the Old River phenomenon that confirmed that diversion was highly probable.33 Pessimistic projections indicated that if diversion continued at the prevailing rate, the major part of the flow of the Mississippi would be captured by the Atchafalaya by 1968. More conservative opinions held that the main river would not change course until about 1985. Corps leaders accepted 1975 as a reasonable compromise. Sentiment was unanimous that a major change in the river’s course would spell drastic ecological alterations and economic disaster.

Old River Model Studies

Action was imperative. Corps leaders began to consider several proposals for dealing with the threat of river diversion. Ultimately they opted for a complex arrangement of edifices involving an Old River closure dam, a new inflow channel into Old River with a navigation lock, an upriver low-sill control structure and channel into the Red-Atchafalaya system, and a lengthy overbank control structure for use during flooding. Never had engineers attempted such a project to manage the channel of a major alluvial river. The low-sill structure was of utmost importance, as it would regulate flow of the Mississippi into the Atchafalaya basin even at low water periods. It also presented the most challenging technical problems.

In 1947 the MRC commissioned WES model investigations to determine the effects of the proposed control structures on stages and flow conditions in the Mississippi River, the Red River, and the Atchafalaya River and basin. Tests employed the venerable Mississippi River Flood Control Model under the supervision of Lipscomb, assisted by Joseph W. McGee of the Rivers and Harbors Branch. Model reproductions of three floods, the 1927 and 1945 prototypes and the synthetic project flood, provided data. These indicated that the control complex would effectively prevent river diversion and also enabled the Corps to plan coordinated use of the complex during floods with the Morganza Floodway, only three miles downstream.34

Portion of Mississippi River 
flood control model showing Old River control structureIn 1953, in a more specific study, Fenwick and Franco supervised construction and operation of a smaller model that replicated only the 11 mile river reach in the immediate vicinity of Old River. Experiments concentrated on determining the effects of the proposed control complex on sedimentation in the Mississippi River and on the flow from the Mississippi into the Atchafalaya River. As a result of the study, the Corps enacted a major revision in its plans, reversing the relative positions of the overbank control structure and the low-sill structure. Whereas the original construction proposal placed the low-sill structure upriver from the overbank control structure, model tests showed that this would produce undesirable sedimentation phenomena.35

Further Geological Investigations

While hydraulics specialists analyzed the riverine aspects of the Old River project, construction engineers recognized that before any major edifices could be designed in detail, further geological evaluations of the entire Old River area were necessary. The massive facilities — especially the low-sill structure — required foundations built on complex subsurface strata. Excavations at the low-sill site would extend approximately 65 feet below the surface of the ground. Consequently, engineers insisted that foundation conditions at each construction site be adequate to assure that settling of the structures would be uniform and not excessive. Also, the new intake channel and lock channel demanded the location of erosion-resistant materials to support them.

In 1949 the MRC, in one of its final acts as administrative agency of WES, authorized a major Old River geological investigative project. Station personnel took more deep borings in the area, scrutinized aerial photos, and addressed a thorough review of geological composition and chronology.36 A second WES investigation, ordered by OCE and completed in 1953, involved further borings and provided an inclusive subsurface review.37 Fisk, who left LSU in 1950 for a position with Humble Oil, served as consultant on both studies. The WES reports led to the selection of the most advantageous locations for erection of the control complex in terms of soil and strata composition.

Old River Hydraulic Structures

After the Corps accepted the control complex plan and specific site selection, Brown’s Hydraulics Structures Branch performed a series of studies aimed at evaluation and design improvement of a number of Old River mechanical elements. Experiments involved replication of such components as the vertical-lift gates of the low-sill structure, panel-gates of the overbank structure, and filling and emptying systems of the Old River Lock.38

While the Corps refined plans for the Old River project, implementation of the plans would require Congressional and Executive approval and a massive appropriation of funds. The Eisenhower administration looked skeptically at any large water project. The Old River problem involved a particular sense of urgency; thus, Congress in 1954 approved with the support of the Administration the Corps’ entire Old River design plan and authorized $47 million to start construction. Work began on the low-sill structure in late 1955. Further appropriations followed, allowing completion of the project by 1965. The complex stands as one of the world’s great engineering feats, although damages resulting from the 1973 flood caused grave concerns.

Potamology Investigations

Concurrent with construction of the Mississippi Basin Model and design of the Old River Control Complex, WES engineers began the most extensive investigations ever conducted of the fundamental nature of the Mississippi River. These were, to a degree, an enlargement and continuation of WES directive energy, bed materials, and sedimentation studies in the 1930s and of Friedkin’s experiments on the meandering of alluvial rivers during World War II. Encompassing all aspects of the river’s constitution and behavior, this potamology or river science program involved personnel of both the Station’s Hydraulics and Soils Divisions in addition to outside consultants.

Early interwoven willow mattress revetments Failure stimulated the potamology program. By the 1940s the Corps had spent tens of millions of dollars in attempts to stabilize the channel of the Mississippi. The greatest single expenditure was on revetments — covering the river’s banks with materials to prevent cave-ins, scour, or other changes caused by the action of water on the soil. Revetments commonly covered not only the above-water banks, but extended substantial distances underwater along the river bottom into the channel. Since the late 19th century, revetments on the Mississippi had evolved from the use of crude interwoven willow mattresses, held in place by rocks, to articulated concrete slabs fastened together with steel cables. River engineers faced a constant battle in keeping revetments in place, as failures were common and expensive.

By the 1940s the revetment program took on added importance. Since the Corps’ overall plan for flood control in the Lower Mississippi Valley called for massive new levees, many of which were under construction, it was necessary to keep the river’s channel between levee lines. Continued meandering would threaten the levee system and necessitate costly setbacks. To compound the problem, the straightening of the river channel by the Corps’ cutoff program had increased current velocities, accelerating bank caving and meandering.

Events of 1946 were particularly distressing. During the low water season of that year the Vicksburg District placed new revetments on the actively caving bank of Reid-Bedford Bend, about five miles downriver from Vicksburg. That fall a major failure removed several hundred thousand cubic yards of bank material and several hundred feet of the new revetment in a matter of hours. Further massive failures at the site in late January and early February 1947 duly alarmed Corps planners. Shortly thereafter the MRC commissioned WES to perform a major study of river meandering and bank stability. Station Director Colonel John R. Hardin then called a conference of WES, MRC, and Vicksburg District engineers with the grandiloquently stated purpose of “Finding out Why Mississippi River Revetments Fail so Rapidly and What Can Be Done About It.” Hardin expressed his obvious chagrin by stating that “The condition of the river today indicates that no ground whatever is being gained.”39

Within a matter of months WES proposed a program of study accepted by the MRC. Objectives included:

The Hydraulics Division was to conduct several large-scale laboratory projects and coordinate activities of the Soils Division, its Geology Branch, and the Instrumentation Branch. The Memphis, Vicksburg, and New Orleans Districts were to provide personnel and equipment for field observations.40

Hydrodynamic pulsimeter WES potamology 
conference attendees In April 1948 WES hosted the first of a series of potamology conferences. Hardin encouraged representatives of the MRC, WES, and the three districts involved to give intensive thought to means for determining bank and revetments failures and developing methods of preventing them. Later that year a second conference included outside hydraulics consultants for the first time. Boris A. Bakhmeteff of Columbia University and Lorenz G. Straub of the St. Anthony Falls Hydraulic Laboratory of the University of Minnesota provided expertise on certain river phenomena, especially the influence of turbulence. Hunter Rouse of Iowa University joined Bakhmeteff and Straub the following year, giving WES invaluable ties to the most advanced academic institutions engaged in hydraulic research in the United States.41 Since soils studies were an integral part of the potamology investigation, soil mechanics pioneers Arthur Casagrande of Harvard University and Donald W. Taylor of MIT also participated in future conferences, either in conjunction with the hydraulics consultants or in some cases only with WES soils specialists.42

Potamology Field Investigations

Hydrodynamic pulsimeter Both the Hydraulics and Soils Divisions conducted field investigations. Largely due to the influence of Bakhmeteff, efforts of the former concentrated on measurement of river turbulence and on attempts to determine the effects of turbulence on underwater revetments. Almost no empirical data existed for turbulence phenomena. Early attempts in 1948 and 1949 were unsuccessful, largely because adequate instruments were not available. Project directors determined that experimental equipment must be designed that could accurately measure turbulence in deep, swift water while suspended at any desired location from a boat or barge. By 1950, WES engineers had developed a hydrodynamic pulsimeter that met the desired requirements. The full-scale instrument consisted of a 2,700-pound cast-iron disk, 5 feet in diameter and 4 inches thick, that could be suspended from cables. A rudder along with horizontal and vertical stabilizers provided equilibrium in even very swift currents and kept the device in proper orientation with water flow. In the center of the disk a pressure cell measured pressure fluctuations, while a current meter attached to and suspended immediately above the disk measured velocity fluctuations. An oscillograph recorded all data.43

Use of the apparatus in 1950 enabled field crews to obtain accurate measurements of pressure and velocity variations in the Mississippi River for the first time. Computations made by Bakhmeteff, reinforced by field measurements, indicated that slabs of concrete revetment even as thick as 6 inches or greater could be lifted from the bottom of the river by turbulent forces. However, before more data could be obtained, wartime funding cuts resulted in suspension of most hydraulic investigations in the potamology program.44

Soils studies were more lengthy. In a first-phase project, directors specified particular areas for investigation, where revetment had failed or where bank slides had otherwise been troublesome. These included such obscure river reaches such as Point Menoir, Hardscrabble Bend, and Wilkinson Point. The last provided a spectacular, though distressing opportunity for analysis in 1950 when more than 4 million cubic yards of bank slid into the river, destroying and completely removing a considerable length of revetment. The slide extended far enough laterally into the bank to crevasse the main-line levee 800 feet inland, starting overbank flooding behind the levee line. Only an emergency construction effort by the New Orleans District averted a catastrophe. At all locations soils personnel first collected and reviewed existing data, including geological studies and existing borings. An extensive boring program, further geological analyses, and other tests produced detailed reports on soil types, subsurface strata, permeability, and other factors likely involved in bank failures.45

Accepted soil mechanics theories held that riverbank slides, including those under revetments, could result from three different types of failures: shear failure; failure by scour; or flow failure. Shear failure occurred when the forces acting on a soil exceeded the strength of the soil. Failure by scour followed when a sufficient quantity of sand was scoured by river action at the toe of a bank of revetment to permit the top stratum to slide into the river. In what WES engineers and consultants considered the most likely scenario, flow failure occurred as a result of the sand substratum of a bank becoming saturated with water, decreasing the shearing resistance of the sand and leading to instability of the slope. Observations and a process of elimination led the potamology team to conclude that the failures under consideration resulted from flow failure, especially the liquefaction of fine sands in point bar deposits.46

Rotary cone penetrometer designed by Hvorslev M Jull Hvorslev Soils studies also led to improved instrumentation and soils sampling methods, prescribed goals of the potamology program. Advances were largely the responsibility of M. Juul Hvorslev, a native of Denmark whom WES hired as a special technical consultant in 1946. In potamology investigations, field crews found locating and sampling of fine-grained sands, materials especially susceptible to liquefaction, to be very time consuming, expensive, and sometimes impossible with existing equipment and methods. Under Hvorslev’s guidance, WES designers developed a rotary cone penetrometer that, when attached to a truck-mounted drilling assembly, had the capacity to measure strengths of fine sands at depths up to 200 feet. Extensive lab tests, correlated with field experiences, proved its relative accuracy.47

In 1954 potamology investigators initiated a second-phase soils study aimed at locating sites where revetments were planned susceptible to flow slides. Each year WES personnel performed penetrometer tests and borings and evaluated all borings from revetment sites made by the Memphis and Vicksburg Districts. By 1962 WES had compiled reports on 78 revetment sites. Of 30 locations where flow failures had occurred, WES studies predicted that 24 were unstable. Of the remaining six, five occurred at boring locations for which no prediction could be made due to lack of data. One failure took place near a location predicted to be stable, but the failure was more than 800 feet from the nearest boring and the boring data may not have been representative of soil conditions at the failure site.48

Potamology Laboratory Investigations

Laboratory investigations performed at the Station first concentrated on channel stabilization by means other than revetment. Model studies using dikes and baffles indicated that such structures could provide substantial protection to riverbanks in some circumstances.49 Efforts to provide alternatives to conventional concrete-slab revetment, such as the use of sand-asphalt revetment, were disappointing.50 Further attempts were aimed at developing a movable-bed model and operating techniques that could be used to predict future changes within a specified reach of the Mississippi River. Project directors elected to reproduce the Concordia-Scrubgrass Bend reach because it had experienced considerable bank recession and channel changes, but was not complicated by any man-made structures such as revetments or dikes. Lack of a material in the model that could both simulate the varying cohesiveness of prototype caving banks and serve as a true bed-load material after caving into the stream posed a particular problem. After numerous trials, engineers developed a crushed bituminous coal for use in the model bed, but mixed a binding agent with it to simulate cohesive properties of the river’s banks. Although the model appeared to reproduce conditions in the prototype, as called for in the original potamology investigational plan, the Corps did not call for further specific studies.51

Tests supervised by Straub at the University of Minnesota produced controversial results. Using a full-size concrete revetment mattress in a large flume, researchers attempted to reproduce turbulence in the Mississippi River that might lift or depress underwater revetment blocks. Since none of the revetment blocks were ever lifted off the bottom in model tests, project personnel concluded that turbulence was not the cause of such phenomena in the prototype. This conflicted with turbulence theories advanced by Bakhmeteff and Rouse and furthered by Tiffany that appeared to be supported by empirical evidence. Tiffany, in fact, completely discounted the Minnesota revetment tests.52

Hydraulic Structures Development

Continuing its traditional function of evaluating designs for hydraulic structures, Brown’s Hydrodynamics Branch performed a lengthy succession of investigations through the late 1940s and 1950s. In some areas these for the first time involved WES directly in a program of applied research. Whereas the Hydraulics Division had previously been concerned, at least officially, only with problems connected with the design of individual structures or with plans for specific flood control or navigation projects, WES engineers had long seen the need for a broader experimental program. Coincidentally, proposals for the establishment of a WES applied research program originated simultaneously at OCE and at the Station in February 1947, with letters from both offices proposing such a program passing each other in the mail. As a result of that correspondence, OCE designated WES as its primary research facility in a number of CWIs. The first were entitled “Wave Force on Breakwaters,” “Stability of Rubble-Mound Breakwaters,” “Study of Harbor Design,” “Scale Effects on Harbor Models,” and “Cavitation.” By 1949 OCE had assigned further CWI projects to the Station including “Effect of Model Distortion on Hydraulic Elements,” “Roughness Standards for Hydraulic Models,” and “Opening Forces on Miter-Type Lock Gates.” Work on lock gates, although authorized in 1949, did not begin until 1958.

WES hydraulics consultants 
conference, 1948 OCE guidelines for its CWI programs allowed WES to contract outside consultants on a continuing basis. The Station quickly moved to establish ties with the most renowned hydraulics experts literally from coast to coast. In November 1948 the Hydraulics Division hosted its first conference for consultants, with Robert T. Knapp of the California Institute of Technology, Morrough P. O’Brien of the University of California at Berkeley, Arthur T. Ippen of MIT, and Rouse of the University of Iowa attending. The consultants familiarized themselves with the Station’s facilities and staff, then discussed directions, designs, and techniques they felt WES research should take.53 Bakhmeteff and Straub joined Rouse and O’Brian as consultants for the first two WES conferences on cavitation and model distortion.54

Breakwater Design

Early WES CWI projects concentrated on breakwater design. As structures employed to reflect and/or dissipate the energy of water waves, thus preventing or reducing wave action in an protected area, breakwaters had been in common use since early Roman times. Breakwaters for navigation purposes are constructed to create sufficiently calm waters in a harbor area, thereby providing protection for the safe mooring, operating, and handling of ships and protection of shipping facilities. Sometimes breakwaters are constructed within large, established harbors to protect shipping and small craft in an area that would be exposed to excessive wave action. Offshore breakwaters serve as aids to navigation or shore protection or both, and differ from other breakwaters in that they are generally parallel to and not connected with the shore. By the mid-20th century, the Corps of Engineers was responsible for over 600 breakwaters of various sizes and designs.55

Generalized harbor breakwater model Rubble-mound breakwaters are the largest and most substantial of various breakwater types and are used almost exclusively in offshore and major coastal harbor protection schemes. They are typically constructed with a core of quarry-run stone, sand, or slag, and protected from wave action by one or more stone underlayers and a cover layer composed of stone or specially shaped concrete armor units. The structures are suitable for nearly all types of foundations and any economically acceptable water depth. They can be designed for either nonbreaking or breaking waves, depending upon positioning of the breakwater and the severity of anticipated wave action during the economic life of the structure.

Small-scale tests of rubble-mound breakwaters had been in progress at the Station since 1942, mainly for the Navy’s Bureau of Yards and Docks. These first determined whether the breakwater proposed for construction at Roosevelt Roads, Puerto Rico, would be adequate to withstand the attack of the largest waves occurring at the site. Shortly after the model investigation began, the Roosevelt Roads project declined in military importance and tests on the original problem were discontinued. However, because of the lack of knowledge concerning the phenomena of waves attacking rubble mounds, the Bureau of Yards and Docks authorized WES to broaden the scope of the investigation to include a study of problems of a general nature. Directed by Hudson and Jackson, the test sequence continued intermittently until 1950.56

GLarge flume for breakwater testsEarly investigations led to two important advances: development of model designs, appurtenances, and techniques necessary for further research; and an appraisal of the accuracy of existing formulas for design of rubble breakwaters. For tests, WES designed a large concrete indoor flume 18 ft wide, 5 ft deep, and 119 ft long, with the depth representing water 58 ft deep in a prototype. It could be partitioned longitudinally by the insertion of a dividing wall, creating narrower test lanes. A mechanical generator produced waves of desired heights and characteristics, ranging up to prototype waves 21 feet high by 300 feet long. These were measured by electrical gages designed specifically for that purpose, while an oscillograph recorded gage data. The wave and measuring mechanisms were used not only for the Roosevelt Roads breakwater project, but for other investigations such as the study of wave and surge action at the Terminal Island Naval Operating Base at San Pedro Bay, California.57 The size of the flume made possible the hand construction of model breakwaters of various compositions, materials, designs, and slopes which could be subjected to a wide variety of wave attacks.

Using the flume and its appurtenances for a broadened scope of investigations in addition to site-specific studies, WES researchers evaluated breakwater design formulas already in use. Data appeared to indicate that one formula, first published by Ramon Iribarren Cavanilles in 1938, was sufficiently accurate for design of rubble breakwaters, but only if used in conjunction with coefficients developed during model tests. An alternative Epstein-Tyrrell formula was found to be of no greater accuracy.58

Succeeding the early test program for the Bureau of Yards and Docks, in 1951 Hudson’s Wave Action Section of the Hydrodynamics Branch began a long-term CWI program intended to integrate all important variables affecting the stability of rubble-mound breakwaters. More intensive investigations indicated that the Iribarren formula was less reliable than had been previously thought. WES then discontinued its use in correlating test data and Hudson developed a new, but similar, formula derived both from theory and from the results of model tests.59 This formula was eventually adopted by the Corps of Engineers and has been used worldwide.60

A number of general and site-specific investigations in the late 1950s and early 1960s dealt with recently-developed alternatives to conventional stone or concrete slab breakwater armor layers. Use of molded tetrapods, tetrahedrons, modified cubes, tribars, hexapods, and other special shapes, for instance, could be beneficial where adequate stone resources were not available or the molded shapes could be more damage resistant. Tetrapods, the first in the new generation of armor units, were developed at Danel’s Laboratoire Dauphinois d’Hydraulique at Grenoble in 1950. Extolling their merits, Danel claimed that tetrapods were much superior to either concrete blocks or quarried stone and could reduce construction costs in some cases by as much as thirty percent.61

In 1953, at the behest of the South Pacific Division, WES conducted a study that for the first time involved evaluation of tetrapods for use by the Corps. Although the program was site-specific, it fell under the Corps’ CWI initiative for general research, as the distinction between specific studies and their potential for broader application had always been blurred. Substantial damage had occurred to the breakwater at Crescent City Harbor, California, caused by storm waves and by waves that overtopped the breakwater even in normal circumstances. Because quarried rock of sufficient size to insure stability of the breakwater was not available locally, the Division Engineer requested tests of tetrapods as an alternative. WES efforts aimed at determining the size tetrapod required to insure the stability of the breakwater for different slopes and design-wave heights, and the optimum number of tetrapod layers that a protective cover would need. Conclusions were that 35 ton tetrapod units would be sufficient, if slight damage could be tolerated, and that two tetrapod layers provided optimum stability.62

Despite this endorsement of tetrapods and encouraging results in subsequent CWI tests,63 by the early 1960s a number of other designs produced superior results. Water Waves Section tests in 1958 indicated that tribars, an armor unit developed by Robert Q. Palmer of the Corps’ Honolulu District, were more economical than tetrapods.64 Quadripods, another American design, were also shown to perform as well as tetrapods.65 Further efforts in the late 1960s and 1970s dealt largely with dolos armor units developed in 1966 by E.M. Merrifield and J.A. Zwamborn.

The Corps of Engineers incorporated WES research in breakwater design, including use of both quarrystone and manufactured armor covers, in its authoritative Engineer Manual 1110-2-2904, Engineering and Design: Design of Breakwaters and Jetties in 1963. It and individual WES publications influenced breakwater design on an international level. Within the continental United States, construction or repair projects at Harbor of Refuge, Barcelona, New York; Port Washington, Wisconsin; Burns Harbor, Indiana; Monterey Harbor, California, and many other locations relied on WES recommendations.66 Further afield, WES site-specific studies guided efforts at Nassau Harbor, Bahamas; Tsoying Harbor, Taiwan; and at Nawiliwili Harbor, Lahaina Harbor, Kahului Harbor, and Hilo Harbor, Hawaii.67

Hydraulic Design Criteria

WES solidified its position as the Corps’ clearinghouse for hydraulics-related information through the establishment of a Hydraulic Analysis Branch in 1951. OCE transferred Frank B. Campbell from the Omaha District to head the new organization, which was responsible for the digest of thousands of technical papers, reports, graduate theses, and other publications and the dissemination of up-to-date design criteria.68

In 1952 the branch published the first edition of Hydraulic Design Criteria, a loose-leaf design manual for ready use by field engineers. The addition of new materials as data became available kept the original version updated on a continuous basis. Concentrating on spillway, outlet work, and gate and valve design, the first issue consisted of only 11 charts and five explanatory sheets, with 250 copies distributed. By the mid-1960s this had expanded to over 230 charts and 140 pages of text, and circulation surpassed 2,500 copies per year. Of that number, over 500 were distributed within the Corps, while other Federal agencies, consultants, universities, private individuals, and engineering firms on a national and international level purchased the remainder.69

Lock Design

As in breakwater and hydraulic structure design, WES involvement in lock design began with site-specific studies before evolving into a general research program. A first WES modeling effort, directed by Brown and Thomas E. Murphy in 1946 and 1947, dealt with the filling characteristics of Algiers Lock, Louisiana. In the New Orleans District, the lock was to connect the Mississippi River and the Intracoastal Waterway.70 Murphy was yet another Depression-era engineer who had first been hired in 1935 as a temporary gage reader.71 Further WES studies in the 1950s analyzed filling and emptying characteristics of the Calumet River Lock, Illinois, and the Barge Canal Lock on the Sacramento River in California.72 A project from 1960 to 1962 furnished filling and emptying design specifications for Holt Lock and Dam on the Warrior River in Alabama.73

Humbolt Bay, CA, model during large breakwater model testsBy the 1960s the Corps had conducted about 30 lock model studies on individual projects, WES efforts included, while the TVA had completed only six. These had kept pace with construction to that date. In spite of knowledge gained through research and construction, there were many gaps and serious questions left unanswered. For example, no general relationships had been developed between lift, filling time, depth of water in the lock chamber, and other factors. Nor were there reliable guides on the combination of lock size, lift, desired filling time, and potential filling system designs. On any lock where lift was about 20 feet, planners usually considered a model study necessary.74

By the early 1960s, however, the Corps had approximately 60 locks either under construction, planned, or authorized. Most were intended for large-scale, long-term canalization and navigation projects on the Arkansas, Ohio, Alabama, and other river systems. Because there was not time for site-specific investigations, the Corps needed standardized designs and procedures for a specific range of lock lifts, depths, and sizes.75 Turning to WES, OCE consolidated its lock design program at the Station in late 1961 as Engineer Study 820, Lock Filling and Emptying Systems. Lock design then occupied an entire section of Brown’s Hydrodynamics Branch. In 1963 this section split away to become part of a separate Structures Branch with Murphy as Chief. Branch efforts concentrated on developing standardized criteria for three lock sizes: 600 ft by 84 ft, 600 ft by 110 ft, and 1,200 by 110 ft.76 WES-developed criteria soon found use in Corps locks on a national basis.

Francis Escoffier of the Mobile District helped point the WES lock program in a new direction. During the 1950s he had become intrigued with a “longitudinal floor culvert system” used for lock filling and emptying designs on the Rhone River in France. By the early 1960s Escoffier’s interest in the French design intensified when the Mobile District became heavily involved in the design and construction of high-lift locks on the Alabama River. In visits to WES over a period of years, Escoffier had promoted the European method. WES engineers tended to favor the design in principle, but did not have time for extended studies and feared it would be excessively expensive. Undaunted, during one of Escoffier’s visits to the Station, he and Murphy sketched out a modified version of the French-inspired lock system on the hood of a car. A.M. Cronenberg of the Mobile District then prepared detailed drawings for a cost estimate. The end product was a less elaborate and cheaper version of the French system feasible for the Alabama River’s Millers Ferry Lock.77 A WES model test series supervised by Murphy from 1962 to 1964 indicated that the floor culvert system was superior to the Corps’ standard sidewall filling and emptying setups, especially for high-lift locks. Murphy, Jackson H. Ables, Jr., and Marden B. Boyd made further refinements that the Corps incorporated into the Millers Ferry and Jones Bluff Locks on the Alabama River and the Dardanelle Lock on the Arkansas River.78 This scheme ultimately became the standard for highlift locks on other projects on the Columbia River and on the Tennessee-Tombigbee Waterway.