Chapter 2
The Vogel Years, 1930-1934   back arrow

Facilities and Equipment

Lieutenant Herbert D. VogelIn building WES literally from the ground up, Vogel faced four challenges: constructing adequate facilities, acquiring proper equipment, hiring competent personnel, and attracting sponsors for projects. In meeting each of these, he was extraordinarily successful.

While outlining plans for the original proposed laboratory in Memphis, Vogel had been authorized by the Office of the Chief of Engineers to spend $50,000. This modest sum was to cover Vogel's salary in addition to the construction of a World War I-type building of "elephant iron." In Vicksburg, however, with the solid backing of General Jackson and the Mississippi River Commission, Vogel later estimated he had spent nearly $1 million in the Station's first year. Upon completion in November 1930, the brick main building alone cost $122,000, with Jackson providing the necessary approval. The building consisted of an open, high-ceiling experiment hall flanked on both ends by two-story wings. The main hall was long enough to house a 165-footlong flume, small movable models, and other laboratory equipment. Not partitioned from the main hall, the east wing served as a pump room with enough open space to allow the assembly and disassembly of movable models. Three offices, a calculating and drafting room, a carpenter shop, a darkroom, and a sediment-reduction room occupied the west wing.1

Completed main WES buildingConstruction of main WES 
building, 1930Because hydraulics experiments required a large and stable water supply, Vogel supervised construction of a dam immediately behind the headquarters building. Completed in mid-October 1930, the earthen structure soon held a 40-acre lake that Vogel named after Major General Brown, who gave the order to move the project from Memphis to Vicksburg. The lake provided water directly to the main building and to the large open area in front of the building through conduits. A pumping system insured that water could be transferred from the lake even during droughts, although this later proved inadequate in extreme conditions.2

Early indoor WES flumeBrown's Lake and Vogel's homeVogel acquired commercial equipment from various sources when available, including pumps, gages, and laboratory experimental equipment. However, much of the gear, especially larger apparatuses, such as holding tanks, flumes, weirs, and traps inside the main building, had to be designed and built on the premises. In cases where equipment was entirely lacking or ineffective, WES personnel quickly developed the expertise to devise and manufacture new types. Two factors made this possible: practical construction skills developed through experience, and the uninhibited ability of WES engineers to apply innovative ideas to distinctively American conditions. In the latter, WES from its birth was a pacesetting institution in hydraulics engineering. When European prototype conditions failed to match those in the United States—particularly the presence in North America of large, meandering, alluvial rivers with complex beds and basins—WES designed and built new structures on scales unheard of in Europe. In its first year of operation, the Station's Yazoo Basin backwater model was the largest in the world. WES also soon led the engineering world in the use of exaggerated model distortion.3

Recruiting Personnel

Joseph B. Tiffany surveys the 1937 
Mississippi River floodJames G. 
Jobes, one of first WES civilian engineersEarly bed load samplerRecruiting quality personnel proved arduous but surprisingly rewarding. Hydraulics specialists were inherently difficult to locate in the best of civil service appointments. The Vicksburg District assigned three junior engineers to WES—James G. Jobes, a graduate of the University of Michigan; Georgia Tech alumnus William Willingham Woods; and Isham H. Patty, who was actually a pharmacist by education. Resorting to unconventional methods, Vogel took advantage of Jackson's license to write his own travel orders and visited MIT, the University of Michigan, the University of Illinois, the University of Iowa, and other institutions engaged in hydraulics research. At each he attempted to recruit personnel for the Station, offering to hire their top graduates as "laborers" at $100 per month with 15 percent deducted. Since jobs were at a premium at any wage, a number of highly-qualified and capable young men who might not have been available in better times accepted Vogel's offer "with alacrity." University of Illinois graduates Joseph B. "Joe" Tiffany and Frederick R. "Fred" Brown, for example, came to WES in 1933 and 1934 respectively, then remained to achieve lengthy and distinguished careers as researchers and administrators. Tiffany had been valedictorian of the Illini class of 1932. Vicksburg native John J. Franco, an electrical engineering graduate of Mississippi State College (later Mississippi State University), began his stellar 40-year WES career in 1933 as a gage reader.4 Vogel also persuaded OCE to allow lieutenants pursuing postgraduate studies to "intern" at WES. Through this program he attracted Lieutenant Francis H. Falkner and Lieutenant Paul W. Thompson, a former Freeman Scholar, both of whom succeeded Vogel as WES Director. Vogel referred to his hand-picked, professional-grade cadre as "brilliant engineers" with no lifetime theories to uphold, any of whom "would have been glad to prove Sir Isaac Newton wrong."5 As an added attraction, most were single men, a fact that Fred Brown modestly claimed provided a "great boon to the young ladies of Vicksburg."6

Vogel’s successors, including Falkner, continued his hiring practices, in some cases with stunning results. Upon graduating with a degree in engineering from the University of California at Berkeley in May 1935, future hydraulics pioneer and Corps of Engineers administrator Jacob H. Douma found himself in the envious position of having two job offers, one from the fledgling Tennessee Valley Authority (TVA), the other from the almost equally nascent Waterways Experiment Station. The former position paid $105.00 per month, the latter $110.00. Douma chose WES for the five dollars more. Upon arriving in Vicksburg, Douma received an unexpected boost to $120.00 per month with the high-sounding grade of gauge reader pro-tem.7

Within a year, Vogel had assembled a civilian staff of about 20, including four professional engineers, eight sub-professional engineers, one clerk, two skilled workmen, and six laborers. By mid-1932 the total had increased to 34, then, as Station activities burgeoned, to 185 in 1933, 215 in 1934, and 401 in 1935. Of that latter number, 16 were professional engineers and 103 sub-professional engineers, while the number of laborers had surged to 236. As previously stated, some of the "laborers" were actually engineers by education. (Numbers for 1935 were inflated due to construction of the huge Mississippi River Flood Control Model, discussed in Chapter 3). Surveyors, draftsmen, photographers, and other trained specialists complemented the diverse work force.8

Civilian Personnel Employed at WES, 1931-1935
Classification Number Employed at End of Fiscal Year
1931 1932 1933 1934 1935
Professional Engineers 4 6 7 9 16
Sub-Professional Engineers 8 10 52 66 103
Clerks 1 2 4 7 9
Tradesmen & Skilled Workmen 2 4 11 14 37
Laborers 6 12 111 119 236
Totals 21 34 185 215 401

Organizational Evolution

Vogel instituted the first simple laboratory organization in January 1931, shortly after the beginning of experimental work. This consisted of three laboratory groups with a single group coordinator in general charge of all activities. The Hydraulic group dealt with fixed-bed models, the Sediment group worked only with movable-bed models, and the Soils group performed supporting studies. (Fixed-bed and movable-bed models are discussed later in this chapter.) By October 1932 the volume and diversity of work had expanded beyond the capabilities of the group coordinator, leading Vogel to abolish the group structure and establish two independent hydraulic sections. These handled fixed- and movable-bed models, respectively. The leader of each section was responsible for all design, construction, and operation of each of his models, and at the same time carried on all correspondence and wrote reports. This arrangement lasted only until January 1933.9

The real functional subdivision of WES began in January 1933 when Vogel set up three sections with separate, though interrelated, functions:

The Research and Experimentation Section conducted technical research and gathered data. It in turn consisted of four groups, one each for fixed-bed models, movable-bed models, tidal models, and soils laboratory work. To free the Research and Experimentation Section from the burden of construction details, Vogel established a Construction Section as a service unit, while a new Administration and Reports Section provided clerical and drafting services.

Only a few months later, in September 1933, Vogel initiated yet another structural overhaul. Retaining a three-unit format, he entitled the new sections:

The Experiment Section, headed by Patty, had complete control over design and operation of models. Within it, three groups specialized in particular model types: Group 1 under Jobes and later James B. Leslie dealt with fixed-bed models; Group 2 under Robert B. Cochrane with movable- bed models; and Group 3 under Henry Sargent with tidal models. Each had from three to six subgroup leaders and was charged with from six to 12 projects at any given time. Tiffany headed the new Research and Publications Section, which conducted technical experimental research and edited reports that were to be published. All service functions fell to the Operations Section, including construction, administration, and the soils laboratory. However, the soils laboratory soon left the Operations Section to form Group 4 of the Experiment Section.10 This structure lasted until November 1935, more than one year after Vogel left WES.

First WES Projects: Sedimentation Studies

Cover of Paper H, first WES 
publicationResearch and publication began at WES even before completion of the dam and main building, although not with hydraulic models. The MRC had for decades gathered sedimentation data from the Mississippi River and its tributaries. On orders from the MRC president, Vogel compiled and examined these records. In July 1930 he completed the first WES paper, published by the MRC as Sediment Investigations on the Mississip pi River and its Tributaries Prior to 1930. Paper H of the U.S. Waterways Experiment Station. Vogel cleverly labelled this first WES effort as "Paper H." Subsequent publications were, in order, papers "Y," "D," "R," "A," "U," "L," "I," and "C."

Small current meterRiver sediment samplerShortly after the publication of Paper H, the MRC ordered Vogel to coordinate a new investigation to project the rate of silting in proposed flood control reservoirs and to add to the fund of data concerning the movement of sedimentary material through the entire Mississippi River system. Personnel from the four districts of the Lower Mississippi Valley Division took periodic sediment trap samples at three river depths—surface, mid-depth, and bottom. Samples came from 26 locations on the Mississippi River and its major tributaries and outlets, including the Missouri, Ohio, Old, Arkansas, Yazoo, Ouachita, Red, and Atchafalaya rivers. Preserved in special containers and mailed to WES, the first samples arrived in late August 1930. The WES sediment reduction laboratory, housed in the west wing of the still-unfinished main building, ran tests to analyze sediment compositions and volumes.

Upon completion of the project in September 1931, the influential Engineering News-Record called it "the most systematic and complete" study of its type ever performed.11

First WES Models: Ohio River Lock and Dam

First 
WES indoor model, Ohio River Lockand Dam No. 37, December 1930 WES model tests began at the end of 1930. (These were not the first conducted by the Corps. The St. Paul District's suboffice at Iowa University had conducted model tests for Hastings Dam, on the Upper Mississippi River, in 1929 and 1930.12) On 27 October 1930, only two weeks after completion of the WES dam, the Cincinnati District Engineer requested through the MRC that WES perform a study of a section of the Ohio River where the proposed Lock and Dam No. 37 complex was to be redesigned. Amid noticeable excitement, construction of the first two indoor WES models began on 3 December. Located in the main hall of the headquarters building, one represented a 4,000-foot section of the Ohio River with a 1:300 horizontal scale and a 1:60 vertical scale. This produced a 5:1 distortion, since the horizontal ratio was five times that of the vertical (300:60). The other model, which was undistorted, reproduced a smaller river reach. Following European techniques, workmen built both by cutting templates of galvanized iron to conform with cross-sections of soundings along the river reach, then spaced the templates less than one foot apart inside a simple lumber-framed box 30 feet long, 12 feet wide, and 2 feet deep. Crews molded sand into the spaces between the templates up to about an inch below the top edges of the templates and then carefully troweled a cement surface to the level of the templates. This yielded a fixed-bed solid contour of the river bottom.13

After completion of the model river sections, replicas of the existing dam and its appurtenant structures were added. These could be altered or remodeled in various fashions to determine the effects of revisions in the prototype. Project personnel then introduced water flow into the models and made adjustments until the flow patterns in the model corresponded to the empirically determined patterns of the prototype. The models were then considered verified. Experiments could be performed on a number of alternate dam and lock designs, each subjected to different river levels, including maximum flood stage. Following several weeks of tests—final observations were not made until 15 May 1931—a WES report furnished detailed recommendations for use by the Cincinnati District in its choice of dam alterations.14 However, before action could be taken, a directive from OCE called for the Ohio River Division to restudy its entire canalization program on the Ohio River. The Cincinnati District then suspended plans for improving existing dams.

First Outdoor Model: Illinois River Backwater

In late December 1930 the Station began its first experiments with an outdoor model. The Chicago District Engineer requested a model study of the Illinois River to determine the limit of the river's backwater—the maximum distance that the river would back up from its mouth in times of flooding. Flowing across the fertile farmland of central Illinois and emptying into the Mississippi River just north of St. Louis, the Illinois presented the Corps with an important flood control challenge. By defining the limit of the river's backwater, the Corps could design a levee system to protect the entire area from inundation. Accurate calculation would not only insure that levees extended far enough upriver to check flooding, but would save a great deal of money by avoiding unnecessary construction above that point. The project had a degree of urgency because the Chicago District needed data within 30 days to make recommendations to Congress.15

In the Illinois River project, WES demonstrated that American researchers were capable of exploring new channels in large-scale modeling. This was necessary because river conditions in the United States often differed materially from those in Europe, both in size and complexity. No European river, for example, rivals the Mississippi in length, volume, or meandering tendencies. Even the Illinois River is large by Western European standards. With this in mind, Vogel and other Americans who had studied in Europe felt that models of rivers at home must have greater dimensions and more distortion than their European counterparts, a concept most European engineers questioned.

First WES outdoor 
model, Illinois River backwater, January 1931The Illinois River project posed an immediate challenge. Because a reliable model would be much too large for the WES building, Vogel ordered it constructed outdoors. European river modeling practices called for design and construction of a concrete fixed-bed model, but this was impossible in the short time allotted. According to Vogel, he and Clarence Bardsley resorted to innovative but simple techniques. Bardsley, a Freeman Scholar in 1928 and 1929, had taken a leave from the faculty of the Missouri School of Mines for a short-term job at WES. Acting on the assumption that a reliable model could be carved directly into the loess soil of the Station's grounds, WES workmen began digging into a flat area stretching from the headquarters building south toward a highway. Using topographical maps as a guide, Bardsley had templates cut from steel sheets and fitted to the ground to trace the river's channel. Crews then simply dug into the soil, carving out channels to depths determined by soundings of the prototype as marked by the template. After completion of the model's channels, miniature overbank structures were added—first the existing levees, then additional levees proposed by the Chicago District.16

The model deviated radically from its European progenitors. Neither European publications of the time nor Freeman's Hydraulic Laboratory Practice suggested that models could be dug into the ground. Also, the mere size, the ratios, and the distortion the model incorporated were revolutionary. At nearly 600 feet in length, Vogel claimed that his creation was the largest hydraulic model in the world. While European engineers considered a horizontal scale of 1:300 to be large, the Illinois River model’s scale was 1:1,200. Its vertical scale of 1:48 produced a 25:1 distortion, a ratio also unused in Europe but necessary to reproduce conditions in so large a prototype as the Illinois River.17

To calculate the Illinois' backwater limit, Vogel used the model to replicate several scenarios. The final and most crucial experiment involved simulating the maximum known historical flood flow of the Illinois with the maximum known flood stage of the Mississippi River at its confluence with the Illinois. This was accomplished by allowing a proportionate measured volume of water into the model at its source. Also, a weir was placed at the end of the model (which represented the confluence of the Illinois and Mississippi Rivers) and raised to represent flood levels on the Mississippi. Since the Mississippi at flood stage would cause the Illinois to back up in even normal conditions, the combination of simultaneous floods on both rivers would produce the maximum backwater influence on the Illinois.18

Results from the experiments indicated that the limit of backwater from the Illinois was about 120 miles up from the river's mouth. According to Vogel, Congress established the mark as the the Illinois' backwater limit, enabling the Corps to complete its levee program with a degree of confidence and at minimal cost. After the recording and reporting of data, the model was demolished to clear space for another.19

Yazoo Backwater Project

Yazoo River Basin backwater modelWES had scarcely finished the Illinois River backwater project when in March 1931 the MRC requested a similar investigation of the Yazoo River Basin. Forming the eastern border of the verdant Mississippi Delta, the Yazoo and its numerous tributaries form a large basin susceptible to flooding. Like the Illinois, the Yazoo flows into the Mississippi, which, when at flood stage, produces an extensive backwater up the Yazoo. In the great 1927 flood, still fresh in most memories, the Yazoo had at one time actually run backward up its channel due to the extreme stages on the Mississippi. By 1931, MRC plans called for extending levees on the Mississippi just north of Vicksburg. Since raising the level of the Mississippi at flood stage would affect the Yazoo River backwater limits and retain water in the Yazoo Basin, the MRC called on WES to determine the new Yazoo backwater limits under several scenarios.20

The Yazoo study led to construction of a model even bigger than its Illinois River predecessor. Stretching over a 13,000-square-foot area in front of the headquarters building, the new project miniaturized a 125-mile stretch of the Mississippi River adjacent to the Yazoo Basin along with the entire Yazoo backwater area. Unlike the Illinois model, when construction time was severely limited, the Yazoo model was a fixed-bed concrete structure. This made long-term, multiple-experiment use possible. It also eliminated excessive seepage and drainage, phenomena that experience with the Illinois River model indicated were out of proportion to similar effects occurring in nature. As with the indoor Ohio River models, skilled technicians cut sheet metal templates to reproduce the rivers' contours, set the templates on a prepared base, filled spaces with sand, and covered the structure with concrete. The concrete surface, left purposefully unfinished to produce overbank roughness, was quite similar in net effect to the natural surface within the area considered.21

The Cutoff Controversy

While performing studies of the Ohio River dam site and the Illinois and Yazoo backwaters, WES became embroiled in a long-term investigation of an old and controversial issue: the effects and desirability of cutoffs on the Mississippi River. As an alluvial, meandering stream, the Lower Mississippi forges a serpentine course from the mouth of the Ohio to the Gulf of Mexico. Historically, this course changed continuously, sometimes quickly and dramatically, as the river cut new channels and abandoned old ones. Natural changes occurred most commonly at bends where the river tended to widen a bend further and further until it formed a loop with a narrow neck. Eventually the river cut a shorter channel across the neck, usually at very high water stages, abandoning the former channel, leaving it to form an oxbow lake or fill with silt. The process then began anew.

Corps studies indicated that cutoffs occurred at the rate of about 13 to 15 per century, each shortening the river's length from approximately six to over 20 miles. But since the river habitually began to lengthen its channel, repeating the cycle after any natural cutoff, the total length of the river from Cairo to Baton Rouge, Louisiana, remained almost unchanged from the early 1800s to the 1930s. Cutoffs did not occur below Baton Rouge.

Some cutoffs were manmade. In 1831, Henry Shreve, river boat captain and founder of Shreveport, Louisiana, ordered a channel dug across the narrow neck of Turnbull's Bend, about 80 miles above Baton Rouge.22 Shortening the river notably pleased river boatmen and other commercial interests, but the long-term repercussions of tampering with the river's natural course remained unknown.

From the mid-1800s until the 1930s the Corps and the MRC adamantly opposed further cutoffs, either natural or artificial. Corps attitudes were shaped largely by Charles Ellet, Jr., an influential engineer under contract to the Federal government, who in 1851 warned that river cutoffs were detrimental and actually presented increased dangers of flooding. Ellet's beliefs were echoed 10 years later by Captain Andrew A. Humphreys and Lieutenant Henry B. Abbot in their report for the Corps, Physics of the Mississippi River . Humphreys and Abbot specified, mistakenly, that although cutoffs lowered water stages upriver, they increased river stages below them by half as much. Later events appeared to support their argument. Three natural cutoffs above Memphis and one at Vicksburg in the 1870s and another at Waterproof, Louisiana, in 1884 produced drastic changes in the Mississippi River's alignment, wiped large tracts of agricultural land literally off the map, and interfered seriously with navigation. Corps and MRC activities thereafter concentrated on preventing rather than encouraging cutoffs.

Greenville Bends exemplified the meandering regime of the Mississippi RiverAs always, disasters stimulated reanalysis. Following the great Mississippi River flood of 1897, longtime river student James B. Miles recommended cutoff construction to Congress. He uncannily predicted the exact number of cutoffs and exact mileage reduction in the length of the river as that of the plan adopted and executed 40 years later. Miles and others argued that shortening and straightening the river would lower its bed, lower the level of flood stages, and hasten the flow of floodwaters to the Gulf. In the aftermath of the 1927 superflood, William E. Elam, Chief Engineer of the Mississippi Levee District, presented a paper to the ASCE in which he attempted to show the benefits of cutting off the river's Greenville Bends, a notorious labyrinth of loops near Greenville, MS.23 Elam and other engineers favoring cutoffs had no clear plans of how to execute and control such operations in a river as large as the Mississippi.

Corps policy even after the disaster of 1927 remained staunchly opposed to cutoffs. The "Jadwin Plan," which Congress had accepted in 1928 as the Corps' master design for flood control on the Lower Mississippi, pointedly did not include cutoffs, as Jadwin was a vocal opponent. Yet only four years later, the Corps was committed to a massive cutoff program with WES playing a major role.

The Corps Opts for Cutoffs

Yucatan Cutoff provided a natural clinic 
for cutoff studies Cutoff advocates received a boost from a timely natural occurrence. In the fall of 1929, shortly before Vogel relocated from Memphis, about 40 miles downriver from Vicksburg the Mississippi River completed a natural cutoff. Called the Yucatan Cutoff, it was especially unusual because it came at low water. Jackson and Vogel visited the site in December. The new channel, which took another two flood seasons to capture the majority of the main stem’s flow, was also atypical in that it was not across the narrowest part of the neck of the bend, but rather passed through a slightly curving channel nearly two miles in length. Since it did not upset the river either upstream or downstream in a detrimental way, some observers rightly concluded that a narrow channel a mile or two in length and gradually developed — like the Yucatan — was superior as a cutoff route to the typical short, wider cuts across narrow necks of land. Mild curvature of the channel also seemed advantageous in preserving a deep navigation channel. Learning from nature, cutoff advocates then called for "shortening but not straightening" the Mississippi.

Colonel Harley B. Ferguson Although engineers positively influenced toward cutoffs by the Yucatan event still formed a minority in the engineering community at large, and certainly within the Corps, their numbers included Colonel Harley B. Ferguson, then South Atlantic Division Engineer. By the late 1920s Ferguson had become the Corps' most outspoken proponent of cutoffs, and on 22 November 1930, he submitted a report recommending cutoffs in the 370-mile stretch of the Mississippi between White River and Old River. Instead of adopting the European technique of making dry cutoffs to the full dimensions of a river's channel, then diverting the river into the cut, Ferguson promoted a pilot-cut plan that permitted a more leisurely approach. Integral to Ferguson's thesis was allowing the river to gradually do the major part of excavating a new channel. This would avoid high velocities at the time of diversion and prevent raised flood stages downstream, both invariable results of the European method. Bold in concept and without precedent, Ferguson's ideas soon gained the confidence of Chief of Engineers Brown, who became a staunch patron. Brown later stated that

Ferguson was the first and only responsible man who ever brought to the Chief of Engineers the serious proposition to make artificial cutoffs on the Mississippi River. Whatever credit is due for a courageous effort to lower the height of floods on the confined waters of the Mississippi is due to ... Harley B. Ferguson.24

Reflecting his faith in Ferguson and his new-found advocacy of cutoffs, Brown appointed Ferguson to replace Jackson as president of the MRC in July 1932.

First WES Cutoff Model

Indoor Greenville Bends model pioneered use of models for cutoff studies, 1931

Indoor Greenville Bends model pioneered use of models for cutoff studies, 1931 Vogel had begun model studies of cutoffs long before the arrival of Ferguson. In November 1930 the Office of the Chief of Engineers had ordered MRC President Jackson to begin an investigation of the effects of cutoffs in the Greenville Bends. On 18 November Jackson instructed the Vicksburg District to initiate a full field study. On the same day he directed Vogel to perform a model study of the effects of cutoffs at each of the four necks in the Greenville Bends.25

In December 1930, at the same time that its Ohio River dam models were in use, WES began construction of an indoor model of the Greenville Bends reach. The 80-foot-long structure represented 98 miles of the Mississippi River, stretching from immediately below the mouth of the Arkansas River to four miles below Lake Lee, south of Greenville (River Mile 401 to River Mile 499). Initially built as a movable-bed model with a sand bed and gravel added to simulate rough overbank conditions, the model in its first two series of tests showed "no substantial agreement" with readings taken directly from the river. Consequently, technicians concreted the channel in place and made other adjustments. On the third test series, measurements on the model agreed closely with those observed in nature.

After this verification, over a period of months experiments enabled project engineers to predict stages in the Mississippi River at different points in the Greenville Bends for a variety of cutoff scenarios. Calculation of river stages was complemented by other projections. After exhaustive trials WES personnel found that well-soaked, creosoted sawdust ideally simulated natural detritus movement in the model. This material was used to determine what sedimentary deposits might be formed as a result of cutoffs. Threads placed in the model indicated the direction of currents, and dyes injected into the upstream reaches revealed eddies and other current phenomena.

Test results, published in April and August 1931, pointed to a revolutionary conclusion: cutoffs did not raise river levels below them. All studies, in fact, indicated a general lowering of flow lines above cutoffs — 2.2 feet in the case of Tarpley Neck — and that any rising of flow lines below cutoffs were short term. This further contradicted the theories of Humphreys and Abbot. The model also gave no indication of detrimental effects due to cutoffs, but showed a slight tendency toward improved conditions in some cases.26

Outdoor Cutoff Model

Outdoor Greenville 
Bends model 1932In the spring of 1932 WES constructed a larger, outdoor model for cutoff studies. Taking advantage of the existing Yazoo Basin model that included the adjoining stretch of the Mississippi River, project engineers extended the Mississippi River part of the model to simulate another 155 miles downstream to Old River. The entire complex represented about 280 river miles and covered 17,000 square feet, including the Yazoo Basin. The upper portion of the outdoor model overlapped 16 miles of the lower portion of the indoor Greenville Bends model.

Outdoor model tests concentrated on the effects of cutoffs at seven locations, ranging from just downstream of the Greenville Bends at Sarah Island to Esperance Point below Natchez, MS. Conclusions, derived from both the indoor and outdoor models and published in April 1932, were mixed. The WES report, for instance, indicated that a cutoff at Diamond Point, about 15 miles downstream from Vicksburg, would have a number of good effects and no bad ones. Cutoffs near Natchez were seen as having "dubious" value, and a cutoff considered at Ashbrook Neck in the Greenville Bends, according to the model study, should be "studiously avoided."27

Ferguson and the Cutoff Program

Observers at Diamond Point 
Cutoff, 1933Ferguson succeeded Jackson as MRC president in June 1932. In the administrative transition, he brought cutoff proponent Gerard H. Matthes from the Corps’ Norfolk (Virginia) District to take over as the MRC’s Chief Engineer. Both Ferguson and Matthes took an immediate interest in Diamond Point as the potential site of a manmade cutoff. According to Vogel, Jackson had already begun field work for a cutoff at Diamond Point before his departure from the MRC, but this was vehemently denied later by Matthes.28 In either case, Ferguson ordered cutoff work at Diamond Point to proceed posthaste.

As the first manmade cutoff in nearly a century, the Diamond Point project served as the initial test for both Ferguson's master plan and the WES model study. In September 1932, two hydraulic dredges began excavating channels on opposite sides of the bend's neck, working toward each other, and by January 1933 only a 50-footwide plug separated the two. Amid substantial fanfare, Ferguson on 8 January 1933 departed from Vicksburg on the steamer Control, accompanied by a quarter boat and party that included Vogel. Standing on the banks of the cut, the group watched as four dynamite blasts removed the final barrier. Because the river level on the upstream side was almost 5 feet higher than the downstream, water rushed through a shallow trench, quickly causing the banks to cave in and clearing a 60-foot-wide channel. Thereafter, as intended, the channel expanded gradually without disrupting the river's normal levels, capturing only 10 percent of the flow in the next two weeks. The new channel did not capture the majority of flow until the spring of 1937, and by the end of that year had become the main stem.29

Interpreting the Diamond Cutoff as a total success and vindication, Ferguson proceeded vigorously in implementing the remainder of his cutoff program. By 1939, when Ferguson left the MRC, 12 manmade cutoffs had been completed that, when combined with the Yucatan Cutoff, shortened the Mississippi by about 115 miles between Memphis and Baton Rouge.30 Three later cutoffs increased the total shortening to 170 miles by 1942.

Greenville Bends aerial 
view after cutoffsThroughout the Ferguson era WES continued to perform cutoff model studies, but their influence was, and is, debatable. Matthes in 1948 stated that the early WES reports were important in showing that river stages did not rise below cutoffs, but that, with the exception of the cutoffs proposed at Greenville Bends, WES findings were "far from encouraging," were "adverse to undertaking most of the cutoffs proposed," and were "of no help to General Ferguson." The fixed-bed models, Matthes continued, were of solid concrete and were incapable of simulating bank and bed erosion, factors basic to Ferguson's plan for channel rectification and bank stabilization that would accompany cutoffs. Cutoff model studies improved beginning in the summer of 1932, according to Matthes, when Ferguson ordered that experiments be conducted with erodible beds. Even then, data indicated that cutoffs at several projected locations would be ineffective.31

Convinced as to the validity of his theories, Ferguson was not likely to be strongly influenced by laboratory data, discouraging or not. Paul W. Thompson, Vogel's Assistant Engineer in 1932 and 1933 and third WES Director from July 1937 to September 1939, described Ferguson as "impatient of experimental results that failed to fit his own instinctive conclusions," but also as a man whose "instinctive conclusions were...often and uncannily right." Thompson believed that, in spite of denials by Matthes, the WES studies still "played an important part — more important than [Ferguson] ever admitted or perhaps ever realized."32 In any case, the cutoff program went forward unabated.

New Madrid Floodway

Part of the Corps' general plan for flood control on the Lower Mississippi River involved the creation of floodways: areas into which floodwaters could be diverted until water levels receded on the main stem, or could be routed to the Gulf of Mexico through alternate paths. The Birds Point-New Madrid Floodway was one of three constructed beginning in 1929. Located in Missouri and starting just below the confluence of the Mississippi and Ohio Rivers at Cairo, IL, it covered about 206 square miles. Engineers created it by lowering existing levees at selected points near the Mississippi by 5 feet, then building new levees about 5 miles farther west, away from the river. Conceptually, during very high flood stages, floodwaters would crevasse the shortened levees, diverting a portion of the river’s flow through the leveed floodway and lowering flood stages at Cairo.

Lacking precise data, the MRC in 1932 ordered WES to perform a model study to determine the effects of operating the floodway on the lands lying within it and to predict the draw-down on the Mississippi River with the floodway in use. With more than 100 miles of river to simulate, WES built an 80-foot-long outdoor concrete model of the river channel, the overbank between levees, backwater areas, and the floodway. Designers took special care to correctly place drainage ditches, levee borrow pits, and other details that would affect water levels, and raised miniature levees with soil taken from actual on-site levee borings. After comparing water levels and flows in the model with gage readings from the Mississippi, project engineers made the usual necessary adjustments (such as adding gravel to overbank areas to simulate roughness) until model and prototype readings agreed. Gage readings from the six highest floods in the vicinity since 1882 provided data for water levels and flows in the experiments. Model tests indicated that the new levees were high enough to contain any projected flood, that levels in the Mississippi would be lowered substantially during use of the floodway, and that the lands of the floodway would suffer almost no permanent damage from inundation.33

Other Model Investigations

Bonnet 
Carre Floodway model While most early activities at WES were centered around flood control projects on the Mississippi River and its tributaries, a number of experiments reflected a broader range of river engineering concerns. These included erosion control in floodways and along riverbanks, channel improvements for navigational purposes, and improved design of appurtenances for hydraulic structures. In January 1931 the MRC directed Vogel to determine if extensive erosion could be expected at dredged borrow pits in the Bonnet Carré Floodway, just north of New Orleans. Railroad companies were concerned that erosion at the borrow pits would undermine trestles and threaten elevated railway structures. WES designed a small outdoor model — only 30 feet long by 12 feet wide — that represented the floodway from the Bonnet Carré spillway almost to Lake Pontchartrain. Again demonstrating the ability to devise models for unique situations, personnel surfaced the cleared land areas in the floodway with cement mortar, but left ditches and borrow pits hollowed out from erodible natural soil. Spanish moss yielded a realistic covering for heavily wooded swamp areas, and other finishing touches included placing miniature railroads, complete with scale-model trestles, across the model. Results from experiments indicated that, although erosion could be expected in some places, it posed no threat to the trestles.34

full-size railroad embankmentEarly in 1931, a related study attempted to quantify the erosive actions and general destructive effects of floodwaters on low railroad embankments. This led to construction of a full-size railroad embankment, replete with crossties and track, in the 20-foot-wide canal leading from the spillway of the WES lake. In a pilot test, flood-level waters released from the lake cascaded over the embankment for over two hours while WES engineers took gage readings, motion pictures, and still photographs. Eventually the embankment withstood over 200 hours of flood-level inundation. To complement the outdoor tests, WES personnel designed and constructed scale-model embankments in an indoor flume, then duplicated outdoor tests with different grades of rock and riprap reinforcement. Both test series provided guidelines both for predicting flood damages to railroad embankments and for more efficient construction or remediation.35

Observers at railroad embankment test, 1931; note WES building in backgroundOther early experiments evaluated existing or proposed hydraulic structures. Models of nine locations on the Mississippi River dealt with proper placement and structure of dikes. For example, at Point Pleasant, Missouri, about 80 miles downriver from Cairo, local authorities had called for the removal of an existing dike system and its replacement with another. WES model experiments evaluated several plans, including leaving the existing dike system intact, removal of all dikes, and replacing or supplementing the existing system with alternative systems. Because the existing system functioned as well as any of the proposed alternates in the model, engineers took no action. This prevented unnecessary construction and avoided major costs.36 A similar study performed for the Jacksonville District showed the need for modification of spillway designs on the St. Lucie Canal. Model experiments covered six weeks and cost only $500, but resulted in a net savings of $25,000 in concrete use alone.37

First WES Tidal Model

Prior to 1933, the degree of diversity reflected by experimental activities at WES was moderate. In its first two years the Station concentrated almost exclusively on the engineering problems of inland waterways — flood control and river regulation/improvement for navigational purposes. This began to change as WES broadened its areas of expertise to include the engineering challenges of coastal inlets, harbors, and tidal estuaries. These prototypes presented a highly complex set of interrelated phenomena such as astral tides, littoral currents, wave action, wind action, salt water intrusion, and other factors not encountered in river engineering.

St. 
Andrews Bay Panama City, FL; the first WES harbor modelWES work in harbor engineering began in December 1932 when the Gulf of Mexico Division requested a model study to determine the more efficient of two proposed routes for a ship channel between St. Andrews Bay, Florida (the location of Panama City), and the Gulf of Mexico. In a demonstration of speed and skill developed over its two years of operation, WES designed, built, and verified a harbor model in only two weeks. The indoor structure replicated about 100 square miles of St. Andrews Bay and the surrounding area. Made of concrete, the model exactly reproduced the topography of the mainland and the bed of the Gulf of Mexico, but the bed of the bay in the model was concreted lower than the bed of the prototype. Project designers then covered the bed of the lowered model bay with a 2-inch layer of fine sand. This, they hoped, would provide the model with both fixed-bed and movable-bed characteristics where needed.38

Borrowing largely from European methods, WES operators attempted to reproduce the intricate hydraulic functions of the bay in several ways. Water flowing into the model from the proper direction simulated littoral currents, while raising or lowering the tailgate of the model reproduced tides. In a simple but effective procedure, workers used a gate extending the length of the Gulf of Mexico, inclined away from the model and hinged at the bottom, to simulate waves. A trained worker raised the gate with a hand crank, then allowed it to fall back to its position of rest. This generated a wave parallel to the shoreline. Through experiment and practice, engineers standardized the frequency and intensity of the waves until satisfactory results were obtained. Finally, 10 electric fans mounted on the Gulf of Mexico side of the model simulated winds from various angles to the shore and the surface of the water. WES experiments led the Gulf of Mexico Division to selection of a plan of improvement, but later developments in the prototype were highly disappointing (discussed in Chapter 3).39 The Station's first attempt at harbor modeling was not a success.

Theoretical Research

Movement of bed materials 
model Although WES was created as a practical institution intended to help engineers with problems in the field, part of its work turned to more theoretical considerations. The physics of water flow at river bends, for instance, had perplexed engineers for centuries. In 1876 James Thompson had published an interpretation in the Proceedings of the Royal Society of London that gained general acceptance into the 20th century. Based on the concept that water formed a helix or spiral as it flowed around a bend, Thompson's "heliocoidal theory" explained how materials from the concave (outer) bend of a river were transported by currents to the convex (inner) bend to create deposits or bars.40

In October 1932 Vogel, at MRC direction, initiated a series of experiments to study the movement of bed-load materials around bends. Particular attention was to be paid to the possibility of removing materials from the bed of a main stream by means of diversion channels. Once removed from the main channel, materials could be deposited as fill in low areas or passed along floodways, improving navigation and possibly helping reclaim swamp lands.41

Since WES already had outdoor models of several Mississippi River bends for its cutoff and channel improvement investigations, Vogel used them rather than engaging in new construction. The model used for an Island No. 9 dike study, for example, was adapted to the new project by cutting seven smaller channels leading out of the main stream to represent diversion channels, each of which could be easily opened or closed off. Observers could trace surface water movements simply by watching loose floats, while dyes released into the current indicated general flow directions. Still, neither floats or dyes accurately displayed current directions at the bottom of the stream, where most bed-load was carried.42

Oat grains provided a convenient material to simulate bed materialsVogel stumbled upon a simple material, derived from nature, that served as a reliable indicator of bed-load. Supposedly experimenting at his home on the WES reservation, he noticed that ordinary oat grains sank to the bottom of moving water with the heavier head resting on the bottom and the lighter chaff end pointing in the direction of flow, somewhat like a wind vane. Model operators further observed that oat grains drifted down channels to the concave side of the bends, then crossed to the convex side. The movement was not continuous or uniform, but was "jumpy, rolling, and sporadic." Of primary importance, grains invariably tended to move from regions of high velocity toward regions of low velocity, such as in the convex side of a river bend. Vogel deduced that bed materials were not swept across riverbeds by currents but were drawn to regions of low velocity by other forces, and that the heliocoidal theory of bed movements did not apply to broad rivers such as the Mississippi. As a practical resultt, model tests indicated that substantial amounts of bed-load material could be diverted by natural processes from main channels into secondary channels with lower velocities.43

In a related study, Vogel supervised experiments to calculate the amount of bed-load diverted into a side channel of a straight flume rather than a river bend model. Noting that prior studies in Germany, performed primarily at Karlsruhe and sponsored by Rehbock, had limited applications, Vogel designed a larger and more practical apparatus than anything used in Europe. The WES flume was over 30 feet long with a 2-foot-wide cemented main channel. About 11 feet from the head of the main channel a 1-foot-wide side channel angled 30 degrees to the right. The proportional widths and angle of diversion represented the most commonly found conditions in nature, especially in the Mississippi Valley. Although similar devices were generically known as forked flumes, the Station christened its creation the more distinctive bifurcated flume. Tests conducted by Lieutenant Kenneth D. Nichols and C.D. Curran, under Vogel's supervision, involved introducing bed-load materials into the flume, then observing and carefully measuring the amounts carried by and deposited in the main channel and the diversion channel. Improving on German methods, the WES experiments used a variety of bed-load materials, usually sands, that could differ substantially in behavior, and also allowed exact measurement of materials carried completely through the model. As in the outdoor river bend model tests, results indicated that bed-load materials tended to move toward lower water velocities and that disproportionate percentages of bed-load materials moved to the smaller channel.44

C.E. Bentzel with 
tube in actionA third study related to bed load movement involved a lengthy series of indoor flume tests to determine the force of flowing water required to move the bed materials of the Lower Mississippi River. In 1932 Thompson designed a flume used throughout the testing sequence. Tiffany and then C.E. Bentzel succeeded him as project engineer. Because the study concentrated on the bed materials of the Lower Mississippi, the MRC acquired about 750 large samples taken directly from the river bottom. Workers molded the materials in the flume to simulate the river bed, adjusted the flume to a desired slope, then slowly flooded it from the lower end to avoid disturbing the bed. For the first time, tests provided a mass of data concerning specific bed materials, their location in the Lower Mississippi, the force of flow required to move them, their settlement tendencies, and other factors.45

WES bifurcated flume 
looking upstream WES bifurcated flume 
looking downstream Bifurcated flume tests also led to improved instrumentation. When work necessitated accurate determinations of water velocity and discharge distributions in the channels, standard velocity measuring devices proved too slow or imprecise. Bentzel then devised a velocity tube based on principles he had conceived while designing a flow meter for the gasoline line of his automobile. He secured the first WES-related patent on the instrument, with the right of manufacture retained by the U.S. Government.46

Expanded Mission: Soil Mechanics

While the Station emerged as the Corps' premier hydraulics research center, its mission expanded to incorporate other engineering fields. By the early 1930s, several American institutions, notably MIT and Harvard, began to offer courses and perform research in the new field of soil mechanics, later known as geotechnical engineering. Since many areas of hydraulics engineering such as sedimentation analysis, levee design, and underseepage of earthen structures, involved soils-related studies, WES incorporated soils testing into its activities at an early date. In 1931, just as the first WES hydraulics models were built, a small group of technicians began conducting mechanical analyses of bedload samples and sediment from the Mississippi River on a part-time basis. Housed in the west wing of the main WES building, this informally named Soils Section had by the late-1930s expanded its activities far beyond the support of hydraulics engineering at WES.

Spencer J. BuchananSoil mechanics at the Station received an enormous boost in 1933 when Vogel hired Spencer J. Buchanan, a Texas native and recent MIT graduate, to head soils-related work. Buchanan subsequently built the soils engineering program at WES into the most important in the Corps of Engineers before his departure in 1940. In 1939 WES established a Soil Mechanics Division on an administrative par with the Hydraulics Division. This set a precedent followed later in a number of cases: units originating in the Hydraulics Division to support the Station's hydraulics mission split away to form separate entities. As in the case of the Hydraulics Division, these became national and even world leaders in their respective fields.47

Hydraulic Modeling Ascendant

Vogel's tenure as WES Director ended in August 1934 upon his transfer to Command General Staff School. Less than five years had passed since his arrival in Vicksburg at the end of 1929, and less than four since the first WES experiments began in December 1930. Yet he had supervised a remarkable, and largely unanticipated, growth and transition. Carved from an overgrown creek bottom at the outskirts of a sleepy Southern river town, by Vogel's departure the Station had become the primary hydraulics research institution not only for the Corps of Engineers, but arguably for the entire nation. The increasing volume and diversity of work reflected the Station's prominence, rising from 13 projects in progress in Fiscal Year 1931 to 54 in Fiscal Year 1934. Vogel, in a 1934 article for The Military Engineer, boasted that WES models "in both number and size surpass those of any similar institution in the world." These served not only the needs of the MRC and Lower Mississippi Valley Division, but were used to perform experiments for districts representing every Corps division in the United States except two.48

The accomplishments of WES in the Vogel years are even more impressive upon consideration of the limiting factors present at its establishment:

By 1934 the situation had changed fundamentally in regard to all, and Vogel was largely responsible. With modest financial resources and in a time of national crisis, he had molded WES into a viable institution that was beginning to place the Corps at the leading edge of hydraulic modeling research. Numerous publications in the foremost professional journals of the time indicated the acceptance, both within and outside the Corps, of hydraulic modeling and of the Station's prominent role. Vogel, for example, defined the state of the art in river hydraulics in an article for the ASCE Proceedings of November 1933, an effort upgraded to the ASCE Transactions of 1935, with commentary.49

Perhaps the most striking statement in support of the success of Vogel, of WES, and of the American engineering community, was derived from a tour Vogel made of German laboratories in the summer of 1934, just prior to his leaving WES. This was his first return to Germany since receiving a Ph.D. from the Berlin Technische Hochschule in 1929. Whereas Germany had been the unchallenged leader in hydraulic modeling at the time of Vogel's graduate studies, he now sensed a remarkable change. In comparing German and American advances in the interim, he noted that since 1929 the Germans had made "considerable progress...but the advancement has been not nearly as rapid, or upon as broad a front, as in the United States." International leadership in hydraulics engineering was shifting across the Atlantic. Still, the Station had only begun to realize its potential.