THE AMERICAN MIXED-FLOW TURBINE AND ITS SETTING AMERICAN SOCIETY CIVIL ENGINEERS ARTHUR T. SAFFORD, M. Am. Soc. C. E. AND Edward pierce Hamilton, Esq. WITH DISCUSSION BY Messrs. B. F. GROAT, CHARLES W. SHERMAN, C. M. ALLEN, DANA M. WOOD, H. A. HAGEMAN, ROBERT E. HORTON, FORREST NAGLER, GEORGE A. ORROK, F. W. SCHEIDENHELM, JAY M. WHITHAM, GARDNER S. WILLIAMS, FLOYD A. NAGLEl HARVEY LINTON, and ARTHUR T. SAFFORD and EDYWRD PIERCE HAMILTON. m- r>„ ir ^65744 ■ AMEKIOAN SOCIETY OF CIVIL EN6INEEES INSTITUTED 1852 TRANSACTIONS This Soclsty IS not responsible for any statement made or opinion expressed in its publications. Paper No. 1503 THE AMERICAN MIXED-FLOW TURBINE AND ITS SETTING* By Arthur T. SAFFORD,t M. Am. Soc. C. E., , AND Edward Pierce Hamilton,:}; Esq. r '• With Discussion by Messrs. B. F. Groat, Charles W. Sherman, C. M. Allen, t Dana M . Wood, II. A. Hageman, Bobert E. Horton, Eorrest Nagler, George A. Orrok, F. W. Scheidenhelm, Jay M. Whitham, Gardner S. Williams, Floyd A. Hagler, Harvey Linton, and Arthur T. Safford and Edward Pierce Hamilton. 1 Synopsis The Proprietors of the Locks and Canals on Merrimack Eiver were incor- j,porated in 1792 for the purpose of making the stream navigable from tide- water to the New Hampshire line. The development of the water power avail- able at the Pawtucket F alls, in Lowell, Mass., was commenced by the Proprietors jn 1821, and since that time this power Das been utilized in increasing amounts |)y tbA almost continuous addition of new and improved water-wheels. Much ||vater-whee! hUtoty was enacted at Lowell during the Nineteenth Century. The late James B. Francis, Past-President, Am. Soc. C. E., was connected with the ' Proprietors of the Locks and Canals from November 22d, 1834, until his death, )n September 18th, 1892, and was Chief Engineer from 1845 to 1885 ; his son, he late James Francis, M. Am. Soc. C. E., followed him in that position, jtiitil 1893. Uriah A. Boyden was closely connected with the developments at |iOwell, and the first Boyden wheel was placed in the plant of the Appleton Com- /pny, in that city. Asa M. Swain was a pattern-maker in the Lowell Machine i Shop and, later, developed and built the Swain wheel, at North Chelmsford, • Presented at the meeting of May 3d, 1922. t Engr., Proprietors of Locks and Canals; Cons. Hydr. Engr., Lowell, Mass, t Milton, Mass 1238 THE am: bFLOW TUEBHSTE AND ITS SETTING Mass. The late Hiram F. Mills, Hon. M. Am. Soc. C. E. (successor to Col Francis as Chief Engineer of the Proprietors of the Locks and Canals), first came to Lowell to test wheels for the Swain Company, and James Emerson engaged to design a Prony brake and, subsequently, began his wheel testip there, Clemens Herschel, Past^President, Am. Soc. C. E., at various times worked under the late James B. Francis, at Lowell, and as Hydraulic Enginee: of the Holyoke Water Power Company, from 1879 to 1889, built the Holy ok Testing Flume in 1881. The library of the Locks and Canals contains a wealth of early water wheel history; and in view of the present interest in high-speed and high efficiency water-wheels and their settings, particularly for hydro-electric de velopments, it has seemed, to the present Chief Engineer, that a review of th development of the turbine runner and of water-wheel settings may be timely Many of the so-called modern features are from a quarter to a half centurj old, and were common knowledge to the men of that time. It is also hoped by this paper, to call particular attention to the splendid work of the earlj hydraulic engineers and millwrights. Such a review shows that the modern high-speed runner is the result a gradual development, brought about by ever-increasing demands for mor power, speed, and efficiency. A wheel of the high-speed propeller type wa patented and on the market fifty years ago, but the mechanical connection through crown gears and belts favored wheels of slower speed, and it was no until the development of a generator of the umbrella type that the high-spee wheel came into its own. Such a review also shows that the desirability of, and the reasons fot smooth and easy passages for water-wheel channels were fully appreciated Diffusers and draft-tubes of excellent design were in use, the scroll was well known, excellent settings were often used, and numerous examples of relative^: high efficiency are on record. During the latter part of the Nineteenth Century, quantity production wit “cut and dried” installations became all too common, and too frequently t value of the proper design of the waterways and draft-tubes was overlook The boiler-maker ruled instead of the hydraulic engineer. Hydraulic pract during the last few years represents some improvements and a return to m of the good features developed in the middle of the last ceni;ury- The writers have not had the benefit of the recent expeilm^tal work wheel design, made by the water-wheel builders and their engineers, exc the results of their Holyoke tests, which usually have been available. Tl paper does not attempt to discuss the low-speed, high-head runner, nor remarkable development of very large units, such as some of the recent stallations at Niagara Falls. In these installations, however, hydraulic con tions differ but little from those in smaller units, and the difficulties 6 usually mechanical and structural. It is hoped that the subject of this pa|; will be of sufficient interest to bring out so many additional data that t history of water-wheel design and practice in the United States may be bett known than it is at the present time. James Bicheno Francis. 1815 - 1892 . Asa Methajer Swain. 1830 - 1908 . Uriah Atherton Boyden. 1804 - 1879 . James Emerson. Hiram Francis Mills. 1836 - 1921 . Clemens Herschel. 1842 - THE AMERICAN MIXED-FLOW TURBINE AND ITS SETTING 1241 Early American Wheels The grist-mill and the saw-mill came to this country with the first settlers. One of the very earliest water-power developments was built by Israel Stoughton, in 1634, at the Lower Falls of the Neponset River, between Milton and Dorchester, Mass., where the head was about 8 ft. In succeeding years, it was used as a grist-mill, a saw-mill, and a powder-mill, and the power is now owned by the Walter Baker Chocolate Mills. On Mill Creek, on Boston Neck, there probably were tidal mills at a somewhat earlier date, but the Neponset development is interesting in that it has been in constant use for almost 300 years. The mill moved westward with the early farmer, hard on the trail of the frontiersman. The grist-mill was a most important com- munity center, and many a town grew up around its water power. All these early wheels must have been overshot, undershot, or breast-wheels, although there may have been flutter-wheels in later days. One may read of wheels, the buckets of which were made of ox-horns. All these old wheels were built on the spot to meet the conditions of the place. With the coming of the Nineteenth Century, the growth of the factory system began to call for power in quantities unthought of hitherto. At first, this was met by developing the large rivers of New England on a more or less co-operative basis. The breast-wheel was used almost entirely, and as its diameter was determined by the fall — mechanical limitations fixing the former — the rivers usually were developed in stages of 10 to 20 ft. To 1242 THE AMERICAN MIXED-FLOW TURBINE AND ITS SETTING increase their capacity, the wheels were lengthened axially and various me- chanical refinements were developed. By 1840, these wheels were the common prime movers in the big textile corporations of New England. (Fig. 1.) The example shown in Fig. 1 was one of several wheels installed in the Prescott Mills in 1844. It was constructed almost entirely of cast iron and had a diameter of 16 ft. These wheels gave good efficiencies, but were very slow, and were particularly liable to trouble from ice. They were basicly the same as those which had been used from time immemorial. As more power was required, attention was turned toward the turbine, then existing in a crude state. There had been in use in France from very early times — Buchetti gives the dimensions of one in the Departement du Gard in 1620* * * § — an early form of turbine known as “roue a cuve”, or “tub wheel”, shown in Fig. 2. This w’as a true reaction turbine. In 1804, Benjamin Tyler, of Lebanon, N. H., patented what was known as the “Wry Fly” wheel. The wording of the patent is obscure, but it would appear that the wheel was basicly the same as its continental ancestor. The remains of an old wheel in a saw-mill, at Bow, N. H. (Fig. 3), seem to comply closely with the description of the “Wry Fly”. Another early American wheel was the Parker, invented about 1828, f which was an adaptation of the Barker or Scotch Mill. It does not seem to have come into any very general use. In the same old mill at Bow there was a pair of spiral wheels, installed on a horizontal shaft. (Fig. 4). These were reaction wheels, and be- cause of the rather flat blade angle, gave comparatively high speed under a low head.:]: These wheels were replaced by a pair of Bose wheels (Fig. 5), which were of the impulse type, driven by a double jet. Bose wheels were common in the Northern States in the early half of the Nineteenth Century.§ In 1838, Samuel B. Howd, of Geneva, N. Y., patented an inward discharge turbine. It seems doubtful whether he appreciated the value of centripetal flow, since, in 1842, he patented an outward-flow wheel. Such were the early American wheels. The period of the spiral and Bose wheels cannot be fixed definitely, and they may not have been as old as the others mentioned, but it is probable that they existed previous to 1840. Transportation was very limited in the first half of the Nineteenth Century. By 1840, there were a few railroads and a considerable network of canals over a large part of the North Atlantic States. In many ways, communities, particularly those in the interior and the more inaccessible parts of the East, were largely self-sufficient, and many of the needs were met locally. The result was that the water-wheels of the period and for some time afterward were local in origin and more or less restricted to certain communities. This is well shown by the fact that Howd, instead of trying to manufacture his wheels, licensed builders in various localities to make them in their dis- tricts. His agents sold the rights for Middlesex County, Massachusetts, to the Proprietors of the Locks and Canals on Merrimack Biver for about $1 200, * J. Buchetti, “Les Moteurs Hydrauliques Actuels”, Paris, 1892. t W. C. Hughes, “The American Miller”, p. 48, Phila., 1856. t David Craik, “Practical American Millwright and Miller”, pp. 145-149, Phlla., 1870. § Loc. cit., p. 149. Fig. 3. — Tub Wheel. Digitized by the Internet Archive in 20i7 with funding from University of Illinois Urbana-Champaign Alternates t https://archlve.org/detalls/amerlcanmlxedfloOOamer Fig. 4. — Spiral Wheel. Fig. 5. — Flutter-Wheel and Pair of Rose Wheels. % n V / F‘ig. 6. — Flutteb Wheel. € \ > THE AMEKICAN MIXED-FLOW TUKBINE AND ITS SETTING 1249 and this corporation sold permission to use a single wheel to Solomon Dutton, of Sudbury, for $20. By 1840, we find a few large corporations using overshot and breast wheels of considerable capacity, and a great number of small mills all over the settled districts using undershot, overshot, and breast wheels and various forms of crude turbines. (Fig. 6.) The need for power was growing, and with the increasing development of many rivers, the question of wheel efficiency was beginning to attract attention. A country, formerly devoted to farming, and importing many of its neces- sities from abroad, was fast spreading westward, and manufactures were developing very rapidly to meet the growing domestic needs. As steam was still in its infancy, and water power was common throughout a large part of the country, the early factories were built where power was available. At the present time, it is interesting to study the location of the manufacturing towns of New England and to note how very many have grown up at the natural falls of the various rivers. On the large rivers, many of the devel- opments took place where crude dams and canals had been originally built for the purpose of inland navigation. 7 . — Fourneykon Turbine, 1827 . The Fourneykon and Jonval Turbines Fourneyron installed his first turbine at Pont sur TOgnon (Haute Saone), in France, in 1827. (Fig. 7.) It was an outward discharge wheel in which the water was admitted axially, turned outward through 90°, and discharged. This was the first modern turbine, and it was very successful. A number were 1250 THE AMERICAN MIXED-FLOW TURBINE AND ITS SETTING built, including a 14-in. wheel under a 354-ft. fall, installed at St. Blasier in the Black Forest of Baden in 1837.* * * § The Fountaine turbine of 1839t was improved by Jonval and constructed by Koechlin, in 1841. :{: It was generally known thereafter as the Jonval (Fig. 8), and was a true axial-flow turbine, with the guides in a plane parallel to the runner. It was equipped with a straight draft-tube and was well adapted to low heads. For many years these two turbines were the principal ones used in Europe. In the latter part of the Nineteenth Century, the inward flow turbine, as de- veloped by Francis and Swain, was designed and built on highly scientific lines in Germany and Switzerland. The Boyden and* Francis Wheels By about 1840, there were, in America, a number of embryonic turbines and a growing need for a prime mover of larger capacity and of better efficiency. In 1842, Ellwood Morris, in a paper published in the Journal of the Franklin Institute,§ described the Fourneyron turbine as then in use in France. Tur- bines had been mentioned in the Journal in 1839 and 1840, but the paper by Morris in 1842 seems to have been the first real announcement of the outward flow turbine in this century. * Glyniij “Power of Water”, p. 57, Lend., 1853. t J. Buchetti, “Les Moteurs Hydrauliques Actuels”, p. 80. t J. Buchetti, “Les Moteurs Hydrauliques Actuels.” § Vol. IV, Third Series (October, 1842), p. 217. THE AMERICAN MIXED-FLOW TURBINE AND ITS SETTING 1251 In 1844, Uriah A. Boyden designed, for the Appleton Company, of Lowell, a wheel along the lines of the Fourneyron turbine. (Fig. 9.) He improved and refined the foreign wheel both mechanically and hydraulically. His first wheel developed an efficiency of 78%,* and, before many years, this type of turbine was the favorite of many of the large corporations and was much used throughout Hew England. In the early part of his life, Boyden was at one time a leather splitter and had a shop in Cambridgeport, Mass. He was interested in science, and in 1826, the New Jersey Eagle published an article by him entitled “An Attempt to Explain the Cause of the Warmth at the Poles of the Earth”. In 1838, he was Engineer of the Hashua and Lowell Railroad, but after about 1840, his main work was in hydraulics. Boyden was the inventor of the hook- gauge, and he accumulated a considerable fortune from the sale of his patents on water-wheels, which, during the later years of his life, enabled him to devote his time to the study of pure science, without consideration of financial return. He was particularly interested in the velocity of light, the compres- sibility of water, and the study of “caloric”. When he died in 1879, his fortune of about $230 000 was left to Harvard College and was used for the founding of the Harvard Observatory in Peru. Concerning three Boyden wheels built for the Appleton Company two years later, Francisf states: “The wooden flume, conducting the water immediately to the turbine, is in the form of an inverted truncated cone, the water being introduced into the upper part of the cone, on one side of the axis of the cone (which coincides with the axis of the turbine) in such a manner that the water, as it descends in the cone, has a gradually increasing velocity, and a spiral motion; the horizontal component of the spiral motion being in the direction of the motion of the wheel. * * * The guides, or leading curves, are not perpendicular, but a little inclined backwards from the direction of the motion of the wheel, so that the water, descending with a spiral motion, meets only the edges of the guides.” The Appleton wheels were suspended from an overhead bearing. Mr. Francis states: “This had been previously attempted, but not with such success as to war- rant its general adoption. It has been accomplished with complete success by Mr. Boyden, whose mode is to cut the upper part of the shaft into a series of necks, and to rest the projecting parts upon corresponding parts of a box. * * * It will readily be seen that a great amount of bearing surface can be easily obtained by this mode, and also, what is of equal importance, it may be near the axis. * * * The cast-iron box is suspended on gimbals, * * *.” (Fig. 31.) For a description of the diffuser which Boyden fitted to most of his wheels, Francis may again be quoted: The object of this extremely interesting invention, is to render useful a part of the power otherwise entirely lost, in consequence of the water leaving the wheel with a considerable velocity. It consists, essentially, of two sta- tionary rings or discs [there was at least one example of a diffuser built integral with the runner] placed concentrically with .the wheel, having an * J. B. Francis, “Lowell Hydraulic Experiments”, p. 2, N. Y., 1883. t hoc. oit., p. 3. 1252 THE AMEEICAH MIXED-FLOW TURBIXE AND ITS SETTING interior diameter a very little larger than the exterior diameter of the wheel; and an exterior diameter equal to about twice that of the wheel; the height between the discs, at their interior circumference, is a very little greater than that of the orifices in the exterior circumference of the wheel, and at the exterior circumference of the discs, the height between them is about twice as great as at the interior circumference; the form of the surfaces [of the disks] * * * is gently rounded, * * *. There is, consequently, be- tween the two surfaces, an aperture gradually enlarging * * *. It is essential to the proper action of the diffuser, that it should be entirely under water; and the power rendered useful by it, is expended in diminishing the pressure against the water issuing from the exterior orifices of the wheel; and the effect produced, is the same as if the available fall under which the turbine is acting, is increased a certain amount. * * * ^‘The action of the diffuser depends upon similar principles to that of diverging conical tubes, which, when of certain proportions, it is well known, increase the discharge; * * *. “Experiments on the same turbine, with and without a diffuser, show a gain in the coefficient of effect, due to the latter, of about 3 per cent. By the principles of living forces, and assuming that the motion of the water is free from irregularity, the gain should be about 5 per cent.” In later years, the Boyden wheels were placed at the bottom of a heavy cast-iron quarter turn. These wheels were big and heavy, and were usually set in massive cut-stone pits. They were very expensive; a 9-ft. wheel built in 1861, for the Nashua Manufacturing Company, cost $19 375.32,* or about $26 per h. p. In 1851, Francis made a test of the power and efficiency of a Boyden wheel, built for the Tremont Mills of Lowell, which was the first scientific wheel test on a large scale. The wheel showed an efficiency of 79% ; an efficiency of 88% was claimed for two of the turbines built for the Appleton Company, but the test was not so accurate. However, it is certain that the Boyden wheel, in its best form, showed an efficiency of more than 80 per cent. It had a low capacity for its size, and was of very low speed. The Tremont turbine had a specific speed of 26. The part-gate efficiency of this type of wheel was very low. The one important factor in which the Boyden wheel 'differed from the Eourneyron, was that its design was scientific, and every effort was made to secure hydraulic and mechanical perfection. Boyden ap- preciated the fact that some energy remained in the water at the time of its discharge from the runner, and to remedy this he devised the well known Boyden diffuser. The guides and buckets of Boyden wheels were often made of composition bronze, which prolonged their lives considerably. At the plant of the Merri- mack Manufacturing Company, in Lowell, there is a Boyden wheel which was installed in 1853, and which is in excellent condition and still used at certain times of the year. Sixty-nine years is a long life for a prime mover, yet this wheel, except for its low speed, is much better than many wheels of much more modern design now in use in New England and elsewhere. From a consideration of the Boyden wheel and of the earlier Howd wheel, Francis became interested in the inward flow turbine. The Proprietors of Locks and Canals, of which Company he was Engineer, secur ed the patent • J. B. Francis, Papers, Vol. 18. Fig. 9. — Boyden’s First Turbine, 1844. Fio. 10. — Francis’ First Experimental Wheel, 1847. LiSRARY CrEL • ' ’is ■'.'.i ::ar4*H. !f.\ THE AMERICAN MIXED-FLOW TURBINE AND ITS SETTING 1255 rights for the Howd wheel for the locality in which Lowell was situated. Francis first built and experimented with a small model wheel which developed an efficiency of 71% and a specific speed of 19. (Fig. 10.) In 1849, he designed a center-vent wheel for the Boott Mills, which gave an efficiency of almost 80 per cent. A wheel of this type, installed in 1847, is still used to operate the head-gate hoist in the Northern Canal Gate-House in Lowell, and there is another (Fig. 11) in the grist-mill belonging to the Proprietors. This wheel, although theoretically better than the Boyden, was a true inward dis- charge wheel, and, as such, was restricted in power. In its original form, it was little used in the United States and soon disappeared, but it was the founda- tion on which the modern American water-wheel was built. These two American wheels — the Boyden and Francis — marked a great step forward. They were designed scientifically, full-sized drawings were made of the guides and buckets, and the path of a particle of water was studied for different relative velocities under a given head. Professor W. P. Trowbridge states that all the principles on which wheels had been designed, previous to this period, were wrong,* and that Boyden and Francis broke away from tradition and established entirely new methods. He Yolson Woodf made complete computations for the original Francis wheel (Boott) and came to the conclusion that the design was excellent. The computed efficiency was 79.31% and the observed efficiency, 79.37 per cent. The result of this period of scientific design has been the production of one thoroughly American wheel, the Howd wheel, as developed by Francis. This wheel was to be the forerunner of all modern reaction wheels, and the next step will be to trace its development into the mixed-flow turbine. The “Cut and Try” Period The period after 1860 was marked by the development of a great number of wheels of all types and combinations of types, good, bad, and indifferent, but mostly very bad. The sentiment of the period has been well expressed by “W. W. T.”,t as follows: “There is no use denying that our object was to make money. We had seen these parties build an ordinary casting weighing about 300 lb. and worth when finished $30, and charge for it $231. Such profits as that were well worth working for, so we made the experiment.” It happened that, obtaining an efficiency of only 51%, they were honest enough to give up the attempt, but other makers, turning out even worse wheels, were less scrupulous, and continued to foist their wheels on a mis- guided public. Another remark typical of the period is one attributed to Swain, inventor of the wheel of that name. Before a test of one of his wheels at Lowell, when some one said that the wheel bound in its bearings, he “guessed it would go, only put the water to it”. It did not take many years of testing at Holyoke to banish that idea, and nowadays some manufacturers even have special ball- bearing bridge trees made for their Holyoke tests. ♦ “Turbine Wheels”, N. Y., 1890. t “Turbines”, Transactions, Am. Soc. Mech. Engrs., Vol. XVI. Emerson’s Turbine Reporter, January. 1876. 1256 THE AMERICAN MIXED-FLOW TURBINE AND ITS SETTING Crude and. unscientific as this period was in many ways, it gave, as its result, the modem American mixed-fiow turbine, and it is well worth while to examine closely the various elements that went into the melting-pot. Since this period, almost every wheel has been a combination of the tub wheel in its many forms, the center- vent Howd-Francis, and the axial-flow Jonval. It would seem, perhaps, that a little too much credit was given to the Jonval, since almost every wheel, except the true Jonval, is merely a combination in some form or other of the first two named. All the scientific teachings of Boy den and Francis were thrown to the winds, and the great god, “Cut and Try”, came into his own. If a wheel did not come up to expectations, its buckets were chipped back, up, or down, or its blades pounded, until it gave something better. Such a period could hardly be avoided, since mathematical analysis and design of turbines were unknown to the majority of early wheel makers. The beginning of the testing system at Lowell, and, later, at Holyoke, did much to relieve the situation. Before many years, a manufacturer could not avoid a wheel test and sell wheels, with the result that a poor wheel was either improved or abandoned. For a long time, a few makers managed to avoid public tests, but gradually they were forced to make such tests, and by 1890 most of the wheels on the market were more or less satis- factory. During this period, combinations of all kinds were tried, and great ingenuity was shown, with the result that, by 1873, reported efficiencies of 90% had been reached. One of the early wheels was the Warren scroll, which dated from about 1853. It was a slightly modified Francis runner placed in a wooden scroll case without any guides. It was merely an offshoot of the Howd-Francis wheel and, except for its casing, was of little importance. In 1858, Swain made his first 6-in. model. This wheel was much like the Howd-Francis, except that its buckets were deeper, many in number, and curved outward from the inner discharge edge, so that the discharge was inward and downward. The diameter of the Swain wheel was still large for its capacity, and its buckets were rather shallow, but it was a really good wheel and was the direct predecessor of the modern low-speed reaction wheel. In 1870, on the advice of the late Mr. Mills, the bucket entrances were deepened and the capacity was somewhat increased. A test on a 72-in. Swain wheel by Francis, at the Boott Mills, in 1874, showed an efficiency of almost 84%, with a specific speed of *40. At this time, the wheel had no draft-tube. In 1909, C. M. Allen, M. Am. Soc. C. E., tested the power of this wheel, after a draft-tube and bevel crown gears had been added, and the efficiency, measured by an Alden dynamometer on the jack-shaft and by current meter, was 86.1 per cent. A power test made in July, 1921, in which a small generator was driven by a belt from the jack-shaft, showed an over-all efficiency, at the switch- board, of 73.9 per cent. After almost fifty years of use, this wheel seems still to be as good as ever; its bronze buckets are in almost perfect condition, and, to all appearances, it is capable of running for another fifty years. It is set in a wooden flume, and the water enters eccentrically, as in the Appleton wheel previously mentioned. Fig. 11. — Francis Wheel in Locks and Canals Grist-Mill. Fig. 12. — Swain Runnee. Ui*- i iJiJUuU ;• , Qr. I-:. \ THE AMERICAN MIXED-FLOW TURBINE AND ITS SETTING 1261 The Swain wheel, although one of the first of this period, was modern in every way as far as the runner was concerned. (Fig. 12.) The gate was cylinder and rather clumsy, but the part-gate efficiency was very high for that time, perhaps because the gate opened downward rather than upward. This wheel was a direct development of the Francis runner, probably by “cut and try” methods, as the inventor was a pattern-maker. However, it was really an excellent wheel and was the father of the modern wheel, just as the Howd- Francis was the grandfather. About 1871, a number of these wheels were installed in Lowell, replacing Boyden, center -vent, and breast-wheels, and several are still in use. In all the Swain wheels, great attention was given to the smoothness of the water passages, and hydraulically they were nearly perfect. The lower step bearing, shaped as an inverted cone, diverted the downward discharge to a horizontal direction with a smooth and easy transi- tion. The “American” water-wheel was patented in February, 1859. It was a Francis runner, modified to give an inward and downward discharge. Its buckets were shallow, and the size was still large for the capacity. (Fig. 13.) This wheel had attained a wide distribution over the United States by 1870, particularly in small saw-mills and grist-mills. It was the first stock wheel and marked the beginning of the period of quantity production of standard- ized wheels. As wheel capacities were increased, the “American” was de- veloped, through various stages, to the “Improved New American” wheel, as made in recent years by the Globe Iron Works, of Dayton, Ohio, successors to the original firm of Stout, Mills, and Temple. The first “American” wheel, as tested by Emerson in 1872, showed an efficiency of about 80% with a specific speed of 25. The firm of James Leffel and Company began building wheels at Spring- field, Ohio, in 1862, and has continued to do so to the present time. The first Leffel wheel was double; that is, it combined an inward discharge Francis runner with another of the inward and downward type, built together in one solid piece. (Fig. 14). This wheel attained even greater distribution than its contemporary, the “American”. Equipped with wicket gates, patented by Elijah Boberts, of Eochester, N. H., in 1854,* it gave an efficiency of about 74% and a specific speed of 30. Many manufacturers built the double wheel during this period, but all of them soon died out, except the Leffel. It is understood that the makers say that they do not quite know why the double wheel works, but it does, as shown by the fact that the double discharge Leffel Type F has recently shown an efficiency of 93% at Holyoke. It is rather interesting to note that the Leffel Company still sells a considerable number of its original model, although it makes several types of wheels, including one of a specific speed of 102. The Houston wheel had a rimner resembling somewhat a bevel gear, the inlet edges of the buckets being at an angle of about 45° with the shaft, the bottom diameter being the greater. This wheel was one of the better tur- bines of this period, showing an efficiency of as much as 88%, with a specific speed of 32. It did not have much to do with the development of the modern • Emeraon/a Turbine Reporter, January, 1876. 1262 THE AMERICAN MIXED-FLOW TURBINE AND ITS SETTING wheel and gradually disappeared; however, some modern high-speed wheels, especially a recent continental one, show a return to this shape of tilted bucket. The Risdon wheel (Fig. 15) which attained an efficiency of more than 90% in 1873, marked the high point of the slow-speed, low-capacity period. Despite the fact that it had a cylinder gate, it showed an efficiency of 73% at half load, the highest attained so far. It was an inward and downward flow wheel, and differed from the Swain wheel mainly in that its discharge was wholly axial. The Risdon Company was the first, since Boyden, to appreciate the diffuser or draft-tube, and to apply it to its wheels. The wheel tested in 1873 gave an efficiency of 90.5% with the diffuser, and 88.8% without it. The hydraulic design of the Risdon wheel was excellent, and it seems to have been satisfactory mechanically, although, at one time, some trouble was had with the gate mechanism. FIg. 15. — Risdon Wheel, 1873. Many pages would be required to describe some of the other wheels of the period such as the Angell, Barber, Blackstone, Bodine, Case, Curtis, Cook, Geyelin, Humming Bird, Humphrey, Luther Scroll, Tyler, Upham, Whitney, and others. However, they had little to do with the further development of the turbine, and, with the coming of the later wheels, soon disappeared. An especially interesting wheel of this time was the Wynkoop, a combination of an impulse wheel discharging into a reaction wheel. Its conservative makers claimed a mere efficiency of 175%, but on test it showed about 55 per cent. At this time, the idea was prevalent that a double wheel would give very high THE AMERICAN MIXED-FLOW TURBINE AND ITS SETTING 1263 efficiencies, and many of the ideas conceived bordered on perpetual motion. Of all the combination wheels, the Leffel wheel was the only one which con- tinued beyond the end of this period. The Propeller Type of Eunner Eecently, there has been placed on the market a propeller type of runner, for which many advantages are claimed, chief of which is' a high specific speed. It is proposed to demonstrate again the truth of the old adage, “there is nothing new under the sun”, and to show that wheels of almost exactly the same type were used many years ago, and that their high speed was appreciated. It is claimed that the modern propeller wheel is axial flow. Exception is taken to this, as it is difficult to understand how, in the matter of direction of flow, the propeller wheel can be different from any modern high-speed reaction wheel. The propeller wheel, as commercially installed, is set in a wheel case of the Francis type, the water being admitted through the usual wicket gates, in a direction approaching the tangential. True axial flow presupposes axial delivery of the water to the casing, as in the Jonval wheel. The propeller wheel must depend on the principle of the vortex, which is formed by the tangential admission of the rapidly moving water. It is only a step farther to imagine a wheel setting in which, guides being eliminated, a scroll may be substituted, of such dimensions as to give the same direction and velocity of flow to,, the wheel. It is granted that this will hold under only one condition of speed and discharge. The wheel is still a propeller wheel, and it is still operating on the same principle as it did when equipped with wicket guides. The old tub wheel consisted merely of a vertical shaft carrying a number of blades lying in planes inclined to the wffieel axis. The water was admitted to the wheel with a more or less downward direction, and eccentric to the axis. Under such conditions, somewhat of a vortex action must have resulted. This type of tub wheel was a true reaction turbine. How, then, but for the addi- tion of the draft-tube, is the modern propeller wheel different? In both cases, the water is admitted practically perpendicular to the shaft, and dis- charged axially, and the runner is of the same type. In July, 1860, J. W. Truax, of Kichford, Vt., patented a water-wffieel known as the “Green Mountain”, which, as manufactured in 1876, was as shown in Fig. 16. It was a four-bladed wheel on a vertical shaft, set eccen- trically in a wooden flume. The runner had a band around it, to which were attached a number of small, inclined plane saw-teeth, designed to make the leakage do work, but this does not change the fact that here is a runner identical in principle with the propeller wheel as made at the present time. Other than the maker’s circular, little information can be found regarding this wheel. Apparently, its use was largely local, most of the known installa- tions having been in Vermont. Truax, in his circular of March, 1876, states: “What has baffled the skill and ingenuity of inventors, has been to produce a wheel that would use a large quantity of water and obtain corresponding power for the water used, and, at the same time, attain sufflcient speed in the wheel itself to allow the power of the wheel to be transmitted with con- venience and small outlay on light falls. These points are obtained with this wheel.” 1264 THE AMERICAN MIXED-FLOW TURBINE AND ITS SETTING He further states that the speed of a wheel of this type depended on the incline of the blades. Not only did Truax appreciate what he had, but he also knew how he got it, which was unusual at this time. Unfortunately, no tests of this wheel are available, and the only record of it is the maker’s wheel-table, which is claimed to be the result of actual measurements. It gives to a 30-in. wheel a speed of 126 rev. per min. under a head of 1 ft. Assuming that the measurement of the revolutions per minute and of the water was correct, and, furthermore, assuming an efficiency of 60%, which seems fair and reasonable in the light of the performance of similar wheels of the same period, a specific speed of 125 is obtained. If the efficiency had been only 40%, the specific speed would still be more than 100. Fig. 16. — Green Mountain Propeller, 1876. In June, 1884, at the Holyoke Water Power Company’s flume, a test (No. 256) was made of a Chase Special wheel (Fig. 17), built by the Chase Tur- bine Manufacturing Company, of Orange, Mass. This wheel had an eight- bladed propeller runner, set, without any guides, in an iron scroll case. Here is another wheel, somewhat more modern, operating on the same principle. The buckets were less flat than in the ^^Green Mountain”, with a consequent reduction in speed. However, it showed a good efficiency, 78.9%, with a specific speed of about 50. In the Jonval, the blades were set radially around the periphery of a disk, which left a circle of dead area around the shaft. This meant that the effective lever arm of the runner was large, with a result that the speed was necessarily low. Both the Chase wheel and the “Green Mountain” differed from the Jonval, in that their blades were directly attached to a small hub, THE AMEEICAN MIXED-FLOW TURBINE AND ITS SETTING 1265 with almost no dead area. The blades were effectively acted on by the water to a point very close to the wheel shaft, which meant that, for a given diameter of wheel, a much higher speed could be attained. With regard to “spiral, or screw flood wheels”, David Craik,* in 1870, stated : “Their principle of action is the same as the screw propeller, which has, in a measure, superseded the paddle-wheel in steamboats — the difference being that the propeller is driven round * * * by the steam-engine, * * * while the screw water wheel * * * is driven or revolved by the force of the passing current against its^ oblique vanes * * *. To comprehend this similarity better, take a screw propeller, and place its axis upon suitable bear- ings, and parallel with the stream in a strong, uninterrupted current, and entirely submerged, and it will furnish a motive-power * * Fig. 17. — Chase Special Runner, 1880. He further states that, by this, he does not advocate using the common type of propeller for a wheel, as being driven rather than driving, “it re- quires a modification in structure and details”. This refers, of course, to a wheel with true axial flow, and one in which the water approaches without any whirling motion. R. E. Horton, M. Am. Soc. C. E., states if “A variation of the Jonval turbine, in which the number of buckets was reduced to two, was extensively used in saw-mills in northern Hew York. Owing to the large openings of the buckets, ice, drift, and other obstructions could pass through this wheel without injuring it. The vanes were nearly horizontal, giving a high speed of rotation. The efficiency was very low.” This was the Austin wheel, which was similar to the Truax wheel. The Modern Mixed-Flow Turbine In 1876, a number of 24-in. wheels, invented by John B. McCormick, of Brookville, Pa., were sent to Holyoke to be tested. They were the develop- ment of a type of large capacity wheel, invented by John and Matthew Obenchain, and were the first of the famous “Hercules” type. One of them, tested by Emerson, showed an efficiency of 89.2% with a specific speed of • "The Practical American Millwright and Miller", p. 150, Phila., 1870. t U. S. Geological Survey, Water Supply Paper No. 180, p. 13. 1266 THE AMEKICAN MIXED-FLOW TURBINE AND ITS SETTING 48, which is 13 more than the highest previous wheel, the Eisdon. A 36-in. Hercules”, tested at the present Holyoke flume in 1883, showed an efficiency of 87% in three different tests. The flow was inward, downward, and slightly outward. The buckets were much deeper than in any previous wheel and protruded below the band, thus allowing the outward discharge. (Fig. 18.) The wheel was of the cylinder-gate type and the blades had flns parallel to the line of flow, that were supposed to improve the part-gate efficiency which was 73% at half power. The production of the ^‘Hercules” ushered in a new period of wheel design. The inward flow principle of the Howd-Francis and the down- ward discharge of the tub and Jonval wheels had been combined in the correct proportions, and, aided by good mechanical construction, there had been evolved the American mixed-flow turbine. High speed was yet to be developed, but the be- ginning had been made, and, with few exceptions, all subsequent design was merely improvement pjg. is.— Hercules Runner, over the McCormick type. One must not forget the generations of millwrights who had worked to make this possible. It was not a sudden invention ; it was merely the crystallization and modiflcation of principles toward which all had been working. Many an old, self-trained mechanic contributed his mite to the development of this new type. This was a real American production, the result of evolution during a changing period in American history. The need arose, made itself felt, and eventually was met, not by the work of one great scientist, but by the multitudinous efforts of an army of old Yankee millwrights and machinists, many of the names of whom are either unknown or forgotten. The year after the introduction of the “Hercules”, Stilwell and Bierce, of Dayton, Ohio, placed the “Victor” turbine on the market. This wheel was much like the “Hercules” and had very deep buckets which hung even farther below the band than in the earlier wheel. Built with a register gate, it was of very simple and solid construction, and, although a little below the “Hercules” in efficiency, it also had a speciflc speed of 48. (Fig. 19.) Both the “Hercules” and the “Victor” were later modifled somewhat; the “McCormick” wheel was placed on the market by several makers, and the improved Leffel, double-discharge wheel, the “Samson”, appeared. Until about 1900, these were the usual types of wheel and practically every condition necessarily was met by the selection of some one of- the stock models of the several types then available. Such was the state of affairs at the beginning of the period of electriflcation, when higher speed and better regulation became of growing importance. During the last few years, all effort has been devoted toward increasing the speciflc speed of the mixed-flow runner, without suffering a loss in effi- ciency. The bottom, or discharge, diameter has been greatly increased, with % Fio. 20 . — Modern High-Speed Runner. ■'iW fr •■■ . A. li. • ‘ ■ ■p>i' ] .^' •^"'z l:« m 5 > ; Jz""-''-;. *C ~ '‘ V r*^ ■ <:■ ^ S%t^ I . '^'fv'f , /* ;K rX , zz ^ . • ' '' .- '''■■i?5!f ■• ■ \ ^ 'iM . . ' 'y' • •fi.- ■. y rtA!*-' • V>v ■:i- ' z.“ (/* • K’ T'!F * '>.r , .'i ^ t*fu > t r ' > i'tr .1 j| ■ • “.’ -■ ' '•>?! . ■• :fr ■ r-- f- 'ikc:- fc.:# ..Xk.. V ■■' . ; - -iZ •*'' , ,,-.r '• ■ 'n WhtM' >?uc;.;uf.!--.'J^ ''»Z»t ■5^ THE AMERICAN MIXED-FLOW TURBINE AND ITS SETTING 1269 the result that the modern high specific speed runner with a nominal diameter of 30 in., has a maximum diameter of more than 40 in. The buckets no longer protrude much below the band, and the inlet edges of the buckets usually slant, so that the clearance between the guides and the buckets is considerably more at the top than at the bottom. Fig. 20 shows one of the best of the modern high-speed runners, which gives an efficiency of more than 90%, and a specific speed of 102. In the United States, there seems to be a general belief, especially a'mong some of the large water-wheel manufacturers, that practically all the develop- ment of the modern water-wheel has taken place within the last twenty years and has been the result of the introduction of foreign practice in water-wheel design. Practically all development abroad has been in Switzerland and in Germany. In France, where the first turbines of record are found, compara- tively little has been done toward the development of the modern high-speed wheel. The Germans developed the mixed-flow turbine to some extent, but they did not create it — that honor belongs to the United States. The turbines of Fourneyron, Fountaine, and Jonval were all French in conception, although it has been said that Henschel, of Cassel, Germany, deserves the credit for the axial flow wheel, rather than Jonval, whom he may have preceded by a year or two. The Fourneyron and the Jonval were, for a long time, the only turbines used in Continental Europe. Meanwhile, the idea was exported to America, where Howd and Francis developed the forerunner of the mixed- flow turbine. This, in turn, was sent back to Germany and Europe took up the inward flow principle at about the time the development of the “Hercules” introduced a new period of wheel design in this country. A Swiss engineer, Mr. A. Streifl,* states: “It is a matter of fact that every marked improvement in the design of pressure-type wheels originated in America. The classic investigations by James B. Francis of his ‘center-vent’ water-wheel at the Boott Cotton Mills at Lowell in the year 1850, are the foundations of the modern Francis turbine. Professor F. Prasil, of Zurich, in his theoretical investigations published in the year 1905, did not hesitate to illustrate his conclusions with examples taken from the original Lowell experiments. It is, perhaps, not to be regretted that the Francis turbine was not reared in the same scientific atmosphere in which it was born, since this might have stifled the innu- merable original creations of the inventors who followed. Each small foundry, and machine-shop, so to -speak, manufactured water-wheels based on Francis’ principle, and hardly any new form of runner can be conceived that has not been made in the past, and is perhaps still running in some Hew England mill. It was not until the year 1875 that the J. M. Voith Company, of Heidenheim, Germany, took up the Francis wheel, and in 1876 that the Escher Wyss Company, of Zurich, Switzerland, followed suit.” Mr. Streiff further states that, in 1914, American practice in the con- struction of high-speed, low-head runners far surpassed the best that could be produced in Europe. Table 1 gives the characteristics of American water- wheels, for the period 1847-1900, on the basis of a 30-in. wheel. • “Electrical Engineering and Hydro-Electric Development”, Transaotiona. Inter. Eng. Cong., 1915, p. 498. 1270 THE AMERICAN MIXED-FLOW TURBINE AND ITS SETTING FAMILY TKEE OF MIXED- FLOW TUEBINE TUB WHEEL ^ 1 WRY FLY EARLY AMERICAN Spiral Howd Parker Rich Rose JONVAL - - * howd-francis MIXED-FLOW “Green Mountain” Chase I modern propeller “American” Leffel Swain — Houston Risdon “Hercules” “Victor” “McCormick” “Samson” j MODERN HIGH-SPEED FOURNEYRON ' 1 ? BOYDEN ? j ( i MODERN LOW-SPEED i THE AMERICAN MIXED-FLOW TURBINE AND ITS SETTING 1271 [M Francis 1847 HPi 0.2 Ns 17 American 1859 l-Pi 0.4 Ns 25 Swain 1858 Modem hPi 3.3 Ng 88 HPi 0.6 iSrs34 Risdon 1873 hPi 0.5 Ns 25 Hercules 1876 KPi 1.4 Ns 48 Modern hPi 2.8 Ns 10 Modem l-Pi 4.5 Green Mountain 1876 Ns abt.125 Modem Propeller FPi ±1.7 JVs±150 BUCKET OUTLINES OF RUNNERS OF APPROXIMATELY THE SAME RATED DIAMETER H.P. ON BASIS OF 30" Fig. 21. 1272 THE AMERICAN" MIXED-FLOW TURBINE AND ITS SETTING z-o: 'V) ' 7 r TO£,M2o2dSoOD ^ o; W."'.” opCC^W ^ o d‘0 £ o M »5 o2, 2, M d g » § B 1“ o «oosgo-^oo>uco^ic^booacaoooooooQOQo< > rf>. >-» p p O 00 _< bobwb^iobob rtj ^ ^ X ^ a: f^Q P'5, c g'isi'f ; I isS"3 S IfJ'S I - = ct> 2 si iisaliii -r Cx Hj 00 s g 2. 53 P g P d 0,0.“ ^P CD g-S aco CD P CD d w g,. CO iig^s rt-CD 2 ■ S' P'S 2.2? S‘p B ^ • • c cr S5s$ o-O OP. o^ B 3 - ^r..j,fDCD £-2 " ' M “ §g Hg. . . , , , ,pD, ' w a " ^ .1 pr o 2.0 ^.^dCP r? ^ ffc CD *«d 3 ^ -r I II ■ l| gs2 w a> 2^' ^ £ ^-:: “ 2 t=. -s o P Sp • ® 2 si CK » S i 5= o B H'cD ll- i" d"^ P P c+ ?fdx? WcD 8^ ^2 o d ap P's: o “ 2 .gS £. a=^ 2d O P' ® w 3 ; g‘2. p p22.b o = ^.crq on Pf - p P'5 £ 0^.2-"^® 2 ^.Pdo ri p‘ 5“ o S" t o-^ B tr <^5 >-J o g TABLE 1. — Charaoteristios of American Water-Wheels, 1847-1900, ON Basis of 30-Inch Wheel. THE AMERICAN MIXED-FLOW TURBINE AND ITS SETTING 1273 The Scroll Setting Many of the early turbines, such as the Boyden and the Swain, were built integral with their flumes. With the appearance of the small turbine, various settings were evolved. In the most common setting the wheel was placed directly in a hole in the floor of the flume which in some cases was widened at the end to form a box-like structure. By 1870, many of the small wheels were set in cast-iron globe-shaped cases and supplied by a pipe; this was particularly common with the higher heads. As the size of the wheel increased, the cost of cast-iron casings became, in most cases, prohibitive. Many of these early settings were excellent hydraulically, and often were quite as good as could be designed. In later years, however, as wheels increased in size, sheet-iron cases began to be used. All knowledge of hydraulics apparently was thrown to the winds, and structural conditions determined the shape of the casing. The boiler-maker ruled. Many of the early wheels were set in wooden scroll cases (Fig. 22) and had no guides. Fig. 23 shows a type of wheel quite common for many years. It was used much in saw-mills as an auxili- ary wheel for running back the log carriage. This diagram was copied from a drawing found among some of Boyden’s papers, in the flies of the Amoskeag Manufacturing Company, of Manchester, H. H. The construction of a similar wheel is described by Craik.* After 1850, many scroll wheels came into use. They usually were of iron and were entirely self- contained, that is, the scroll was built integral with Scroll, F^g, 23. — Wooden Center-Vent Wheel. “The Practical American Millwright and Miller”, p. 205, Phila., 1870. 1274 THE AMERICAN MIXED-FLOW TURBINE AND ITS SETTING the bearings which supported the runner. (Fig. 24.) The gate was usually a plain sliding rabbit-trap, at the entrance to the scroll, although a butterfly was also used. These wheels had no guides, and were very poor at part gate, although at full gate they gave an efficiency of about 70 per cent. Fig. 25 shows a model of a scroll wheel patented by Daniel T. Lakin, of Hancock, H. H., in October, 1863. Its runner was of the true Howd-Francis type, with flat blades, and was inverted so that it discharged upward. The interesting feature of this wheel was the regulating mechanism which was a register gate, fitted inside the runner and revolved with it. Closing the gate reduced the discharge area of the buckets and decreased the flow. Directly attached to the shaft are the flyballs which actuate the gate. The whole regu- lating mechanism is part of the runner and shaft, and the entire wheel is self-contained. There is no question but that such an arrangement would give poor part-gate efficiencies, due to contraction losses through the gate, but it is most interesting as an early example of the present practice of placing the flyballs directly on the shaft of a vertical unit. Fig. 24. — Scroll Wheel by Boyden, 1854. ^ Probably the best of the scroll wheels, and the one which survived the longest, was made by John Tyler, of Lebanon, H. H., grandson of Benjamin : Tylet who patented the “Wry Fly” in 1804. As made, in 1873, it gave an; efficiency of 81.6% and a specific speed of 29. The same runner placed in a' register-gate flume casing, gave a lower efficiency than when used in the scroll, j A scroll was sometimes divided into two compartments by a thin horizontal: I I i Fig. 25. — Lakin Wheel, 1863. Fig. 26. — Drouillaed Flume, 1880. Pig. 27. — Warren Scroll, 1860. / I /f .T'f” THE AMERICAN MIXED-FLOW TURBINE AND ITS SETTING 1279 diaphragm, so that, with the gate partly shut, the same velocity and direction of water might be retained in the lower half of the scroll. During the ‘^cut and try” period, the iron scroll casing was used by many makers, especially with smaller wheels, but, by about 1880, they all had disappeared with the exception of a few, such as the Chase which, however, was not a true scroll wheel. The Drouillard flume (Fig. 26) was an attempt to improve the open- flume type of setting. The time had now come in which, the increasing size of the wheels rendering the casf-iron case prohibitive in cost to most makers, and a setting more permanent Jh an the wooden flume being desired, wheel casings of sheet iron cani'i^Vo use . About this time, the use of the draft- tube enabled the wheels to be placed a^ove the tail-water, where they were more easily accessible for insp, action a nd repair. For many years, almost every wheel maker considered the Qiaft-t’^be merely as a device by which the wheel could be raised clear of the wa’iei^nd still retain the full head, and it appears that only one maker, Risdon and Company, appreciated its value in regaining velocity head. The development of the draft-tube will be discussed later. With the coming of the boiler-maker period, the scroll setting disappeared, and it seems to have bee*n almost entirely forgotten. Indeed one wheel maker of the present day is somewhat surprised that “examples have been found in what is left of some early mills in Connecticut of turbines of primitive con- struction, equipped with a volute form of water passage surrounding the runner.” Not only was the scroll used largely during the early years of the develop- ment of the American turbine, but its principle also was studied and under- stood. An article* by A. G. Hillberg, on scroll design, described a method of dividing the water used by the wheel, into a number of imaginary lines of flow, and of carrying these back through the scroll, thus studying the effects of shape. In the flies of the Proprietors of Locks and Canals, there is a study, presumably by Francis, dated 1857, of a scroll design for the grist-mill wheel in which this same method is used. The American Water Wheel Company, of Wareham, Mass., makers of the Warren wheel, were early users of the scroll case. (Fig. 27.) In a letter to the Committee of the Ninth Exhibition of the Massachusetts Charitable Mechanics Association, dated September l7th, 1860, that Company discusses its theory of scroll design. A few quotations are given, as follows : “Our improved, scroll is formed upon a plan which enables the water to pass, at the periphery of the wheel, at an uniform velocity at all parts of the wheel’s circumference. The water in the scroll passes (or should pass) towards the center of the wheel with an accelerated velocity in proportion to the diminu- tion of the area of the circles over which it passes.” The circles mentioned are a series of concentric circles, of area ratios of 1:2:3, etc., used in the construction of the scroll by graphic methods. This means that the water is given a uniform centripetal acceleration for each unit of angular distance traversed. • Engineering Record, October 3d, 1916, 1280 THE AMEBIC AN MIXED-FLOW' TUBBINE AND ITS SETTING “The principle of the centripetal action of the water in the scroll is the same as that of the vortex, or whirl-pool, accelerating the velocity of the water in proportion to the diminution of the area as it approaches the center.” After being abandoned in the United States, the scroll was taken up by the Germans, and largely used for high-head Francis runners. Pfarr credits Voith with being the first to build an iron, spiral-cased, wicket-gate turbine (1894).* This type of wheel was much used in German and Swiss practice, and when high-head developments began in this country it was brought back from abroad. The first hydro-electric station at Niagara Falls was equipped with Fourneyron turbines made in the United States after design«^by Faesch and Piccard, of Geneva, Switzerland. It is interesting to know, hovv ever, that the firs^ turbines running under the full fall at Niagara were built b^ an American firm, James Leffel and Company, after the designs of A . F. £p-.rk8. In the earlier years of high-head development, the scroll was-' son)etimes used, but a great many installations were of the boiler-maker type. Py about 1904, however, the spiral casing was in general use for high-head wheels. In 1903, a vertical unit with a spiral, steel-plate scroll was placed in a plant on the River Glommen at Kykkelsrud Falls in Norway, under a head of about 60 ft.f This appears to have been the first example, on a large scale, of what is now the best modern practice. Until reinforced concrete made the modern setting possible, wood and ' cast or plate iron were all that could be used, aside from masonry. Metal cases were used, but, for a big unit, a plate scroll would have been considerable of an undertaking. It was much easier to build some kind of a plate casing, usually circular in form, around the wheel and to let it go at that. Efficiency ' was not sought as it is to-day, and in the days of direct drive and small units, ' a few per cent, did not mean very much. With the appearance of central | stations and large units, even 1% meant money, and over- all efficiency was ' considered as it never had been before. By about 1900, and for a considerable time thereafter, a majority of the plants had horizontal shaft units, which ‘j either drove generators or were direct-connected to machinery. Electrification called for a fairly high speed, and this was accomplished by using batteries of runners of small diameter. Until the high-speed wheel was developed further, no other alternative was possible, as long as high speed was required. The need for high speeds developed the runner and resulted in specific speeds of almost double those of the early wheels, such as the “Hercules” and “Victor”. Because of this, it soon became possible, even with low heads, to drive large hydro-electric units by a single runner. Three elements — the development of the runner, the need for high efficiency, and the growth of the use of reinforced concrete — mkde possible the modern vertical unit, which' appeared in this country about 1912. The first plans for Keokuk, Iowa, were for double-runner vertical units.:}: At present, the single vertical runner is used almost entirely for all low-head developments, and even high-head Francis runners have recently been placed in vertical cast-iron spiral settings, such * “Turblnen fiir Wasserkraftbetrieb,” Atlas, Plate XXVII, Berlin, 1907. t Koester, "Hydroelectric Developments and Engineering", p. 382, N. Y., 1909. ' t Engineermg News, September 28th, 1911. THE AMERICAN MIXED-FLOW TURBINE AND ITS SETTING 1281 as the Big Creek No. 8, 680-ft. development of the Southern California Edison Company. Fig. 28 shows an installation of 58-in. wheels operating under a head of about 37 ft., at the Gardner’s Falls plant of the Greenfield (Mass.) Electric Light and Power Company. This plant was built in 1913, and the acceptance test showed a wheel efficiency of more than 94 per cent. The Horizontal Wheel and Its Setting During the early days of the turbine, it was a natural step from the hori- zontal shaft of the breast-wheel to the placing of turbines in the same position. This setting was commonly used until about 1850. The wheels were usually of the outward discharge type, such as the spiral and Bose wheels, and often discharged into the open air. The double Parker wheel discharged outward into two wooden draft-boxes. (Fig. 29.) The efficient use of horizontal wheels required a draft-tube and until the draft-tube came into use this type of setting was not practical. About 1880, the horizontal setting began again to appear. Gates Curtis seems to have been the first to use it. In 1879, a pair of Curtis horizontal wheels were installed in the Hudson Piver Mill at Palmer’s Falls, N. Y. They were placed directly on the horizontal machine shaft, 16 ft. above the tail-race. A square wooden draft-tube was used and the head-water stood 12 ft. above the center line of the wheels. In a letter to the wheel maker, Mr. Curtis states : “We are satisfied by experiments made with pipes leading from the top and other parts of the draft-tube attached to glass and mercury gauges that we get equal results from the wheels, as we would were they placed at the 1282 THE AMEKICAN MIXED-FLOW TUKBINE AND ITS SETTING level of the tail-water. This arrangement of taking off power so far above the tail-water, thereby avoiding the annoyance and expense of bevel gearing, is proving most satisfactory in many ways * * A pair of horizontal Curtis wheels (Fig. 29) were tested at Holyoke in 1879, and found to give an efficiency of 72%, whereas, on a vertical shaft, one alone gave 83.6 per cent. About 1882, the Humphrey Machine Company, of Keene, N. H., began to build horizontal double units in cast-iron “camel- bacV’ cases of rather fair design, and, by 1890, this type of unit, constructed of boiler-plate, was in general use. Hydraulically, the majority of these settings were very poor. Some double horizontal settings were made of cast iron, such as the Risdon wheels in the Jefferson Mills of the Amoskeag Company of Manchester, N. H., J. B. Francis, Consulting Engineer, but they were much the exception to the rule. These wheels were interesting, because on one shaft were runners operat- ing under two different heads. There were other good horizontal settings. THE AMERICAN MIXED-FLOW TURBINE AND ITS SETTING 1283 such as that of the Lawrence Company of Lowell, placed in 1895 (Lig. 30), but they were not common. In many of the horizontal units built in later years, the casing was a con- tinuation of the penstock, and the water approached the wheels parallel to the shaft. Such a setting, .with a correctly designed casing, was a great improve- ment over the earlier ones in which the water entered a cramped casing with sharp corners, at right angles to the wheel shaft. Fortunately, however, increase in runner speeds and the use of reinforced concrete permitted the beginning of a new period in wheel settings and the rule of the boiler-maker is now a thing of the past. Probably the worst examples of horizontal wheels appeared during the Nineties. (Fig. 29.) The setting usually consisted of either a rectangu- lar box or a tube. The wheels were placed at each end, and discharged toward each other. A round metal draft-tube extended from the bottom of the draft-box into the tail-water. Every op- portunity for eddies and whirls was offered by the draft-box and tube, and sharp edges and corners abounded everywhere. The worst part of this type of setting was that the water was not conducted smoothly to the draft-tube. It was dis- charged at a fairly high velocity into a large box where it swirled around and was allowed to find its way out through the hole in the bottom of the casing. Sometimes each wheel had its own quarter turn and draft-tube and was set in a com- mon casing,' but, usually, both discharged into the same tube. One wheel catalogue shows an ordinary double horizontal setting, to which had been added a third unit with a vertical shaft. This unit discharged downward and was placed directly above and con- centric with the draft-tube opening. Perhaps the worst feature of the earlier horizontal wheels was the proximity of the runners, when both discharged inward against each other. In some cases, the distance from the bottom of one runner to the other was as small as one wheel diameter. In the early days of the boiler-maker period, this distance was usually from 1.5 to 2 runner diameters. In 1894, at the plant of the Appleton Company of Lowell, a test on Eodney Hunt wheels showed that insufficient distance between runners reduced the efficiency 10 per cent. A pair of 30-in. “Hercules” wheels were tested for the Middlesex Company of Lowell in 1896. As tried at first, the runners were 2.26 diameters apart. By increasing this distance, 16 in., or to 2.8 diameters, the maximum efficiency was increased 4.7 per cent. Modern runners, operating under medium to low heads, should be spaced from 3 to 4 diameters apart. The “camel-back” type of cast-iron draft-chest was gradually evolved, and . showed a great hydraulic improvement. The discharges of the two wheels, instead of being directly toward each other into an open case, were gradually 1284 THE AMERICAN MIXED-FLOW TURBINE AND ITS SETTING turned downward by a curving partition wall so that they were flowing in a more or less parallel direction when they met. Swain devised the idea of carrying this dividing partition all the way down, thus, in effect, giving each wheel a separate draft-tube. Fig. 29 shows the cone setting for a single horizontal wheel. This excellent design for small units under fairly high heads was used by several makers. The double horizontal wheel in a cast-iron, ‘^camel-back” draft-chest, when placed in a casing of good design and fitted with a properly flaring draft- tube, was an excellent prime mover. The best practice usually was to place this wheel in a casing that was a continuation of the penstock, so that the flow was parallel to the wheel-shaft. The approach to the wheel was eased by placing a cone-shaped casing over the up-stream end of the unit. In a setting of this kind, the water passage around the first wheel was sometimes cramped, and the flow to the down-stream wheel was .restricted. To obviate any pos- sibility of this, the casing was sometimes enlarged at the point opposite the gates of the up-stream wheel. (Fig. 29.) A, horizontal setting of Smith wheels in a casing of this kind was placed at the plant of the International Paper Company, at Glens Falls, H. Y. These wheels were tested in 1918 and V showed an efficiency of 88%, which is thought to be the highest authentic result ever obtained from a double horizontal setting. The best results that can be obtained from such a unit will always be from 2 to 4% less than the efficiency obtained from one of the runners tested, on a vertical shaft, under the best conditions. There are some places, however, where efficiency is not the primary requisite, and certain advantages of the horizontal setting make it suitable. The horizontal setting, to a very large extent, is vanishing,- but there are still certain cases to which it is pre-eminently fitted. The most important ^ of these is found in pulp mills, where the wheels are direct-coniiected to the ^ horizontal shaft of a battery of grinders. Conditions call for a fairly high speed, regardless of the available head and the multi-runner unit is often i the best adapted to the need. The Draft-Tube i In June, 1840, Zebulon and Austin Parker, of Licking County, Ohio, were I granted a patent for a “Draft-Tube for Water-Wheels”. Parker wheels, usually, if not always, were set in pairs on a horizontal shaft, fed by a crude scroll case, and each discharged outward into its own draft-tube which was merely a long rectangular wooden box. The object of this setting was not to gain efficiency, but to allow the wheels to be placed above the tail-water. The | French Jonval-Koechlin turbine, in 1841, was equipped with a draft-tube, I but the Parkers seem to have been the first to apply it. The first example of the use of a draft-tube to regain velocity head was the “diffuser” of Boyden (Fig. 31), which, by 1846 was being applied to most of the Boyden wheels built. It was shown to add about 3% to the wheel effi- ciency, by regaining the velocity head of the discharged water. Theoretically, i it should have saved almost 5%, so the diffuser efficiency was about 60 per THE AMERICAN MIXED-FLOW TURBINE AND ITS SETTING 1285 cent. The flare of the parallel disks was apparently too sudden, and the water failed to cling to the sides of the diffuser, with the result that the dis- charge was a succession of surges. In later years, Boyden wheels were not equipped with diffusers, probably be- cause their great size required unneces- sarily large wheel-pits which, in many cases, might have required the changing of the foundations of the mill. Only one diffuser is left in Lowell at the present time; it is on the 1853 Boyden wheel at the plant of the Merrimack Manufacturing Company. The dr aft- tube, as a diffuser, does not appear again for many years. By 1870, many wheels were set above tail-water on wooden stave or sheet metal tubes. Sometimes, the construction of the hooped wooden tubes required a slight flare, but usually they were straight. The design, if there was any, was usually based on the velocity with which a bubble of air would rise in water. In 1881, Gates Curtis, maker of the Curtis wheel, recommended the propor- tioning of draft-tubes for a velocity of from 5 to 8 ft. per sec. The general feeling among wheel makers was that a wheel .could be used with a draft- tube with little or no loss in efficiency, provided the tube was air-tight. About 1880, James Emerson made a series of experiments on draft-tubes, using a 15-in. Victor wheel. He began with a draft-tube the inside diameter of which was the same as the wheel skirt, and then, by inserting fillers, re- duced the area by degrees. The velocities in the tube varied from about 5.5 to 12 ft. per sec., with the best efficiency at the lower velocity, and this effi- ciency was considerably less than that obtained when the wheel was tested without a draft-tube. The draft-tubes used were all straight sided, and were from about 7 to 10 ft. long. At Holyoke, in 1873, the Eisdon Company tested a 36-in. wheel which gave an efficiency of 90.5% with a draft-tube diffuser and 88.8% without it. This trial must have convinced the Eisdon Company of the efficacy of the appli- ance, for it built draft-tubes of excellent design during the period when other makers were evolving weird affairs of boiler-plate. The Eisdon runner discharged axially over an annular section, and there was a dead area around the center of the wheel. To eliminate this, an inverted cone was fitted below the runner, so that the discharge passage was smoothed. The draft-tube, which was slightly flared and usually of cast iron, was excellent and far ahead of its contemporaries. It was claimed that the efficiency of the wheel was increased 2% by the use of the diffuser. However, during the period of the double horizontal unit, even the Eisdon draft-tubes fell down, and examples are found of two wheels discharging against each other in a cramped wooden draft-chest. 1286 THE AMERICAN MIXED-FLOW TURBINE AND ITS SETTING In the days of direct drive, a difference of 2 or 3% in wheel efficiency did not mean a great deal. The wheels were of such low specific speeds, that the total gain realized by the use of a draft-tube was only 2 or 3%, and the gain, in many cases, did not justify the expense of building a good draft-tube. It is interesting to note that “W. W. T.”, writing of the ^^Green Mountain” propeller wheel, realized that, with a wheel of this type, a draft-tube would be of greatly increased value. He says: “If Mr. Boyden would put his dif- fuser on one of these wheels, we would promise him a gain of more than 3 per cent. It would be more nearly 50”.* It would appear that during the period from 1870 to about 1900, or a little later, the use of the draft-tube, as a diffuser, was generally understood, but it was not generally used. The dollars and cents value of a draft-tube usually did not justify the necessary outlay. With the coming of electrification and higher specific speeds, the draft- tube became of greater potential value, and the use of reinforced concrete made it a simple matter to build a tube of any desired shape. The develop- ment of large central power stations meant that even 1%, when translated into kilowatt-hours per year, amounted to a considerable sum. The draft- tube was more than justified economically, and its use, as a diffuser, became universal. Apparently, the pendulum has swung to the other extreme, and there are now on the market various kinds of patent draft-tubes, each of which, like the old patent medicines, is promised to be the cure-all for every hydraulic trouble. The basis of one of these types is an inverted cone placed beneath the discharge open- ing of the draft-tube, with its axis co- incident with that of the runner. The object is to change smoothly and without shock the direction of the water from a vertical to a horizontal direction. The placing of a cone beneath the draft-tube is almost as old as the turbine itself. Some of the early Jonval- Koechlin turbines had it. Almost every German textbook on water-wheels from 1860 to Dr. Camerer’s “Wasserkraft- Fig. 32. — Swain Wheel, 1869. maschinen”, of 1914, shows one or more settings with cones below the draft- tube. English and French textbooks have the same type of setting, and Beardsley shows one.f The Swain wheel had a lower step bearing, and, by 1869, this had been smoothed off to form a cone (Fig. 32). Fig. 33 shows a Swain wheel, placed in 1875, at the Boott Cotton Mills, Lowell, Mass. To • Emerson’s Turbine Reporter, January, 1876. t “Hydro-Electric Plants”, p. 344. N. Y., 1907. Pig. 33. — Swain Wheel at Boott Cotton Mills. THE AMEKICAN MIXED-ELOW TURBINE AND ITS SETTING 1289 better conditions, the cone was continued, as shown, by wooden planks, and, in another instance, at the plant of the Hamilton Manufacturing Com- pany of the same city, it is made of concrete. The iron-cased Burnham wheels of about 1890, had a metal cone below the runner. Fig. 34 shows a wheel setting erected by the T. H. Eisdon Company at the Lawrence Woolen Company, Lawrence, Mass., in 1883. In the makeFs catalogue (undated), of about this period, it is stated: “This wheel and the two preceding it have diffusers. In the Lawrence Woolen Company, a very shallow pit was already constructed, in which the water only stood about 18 in. deep. If the end of the draft-tube had been set low and had discharged the water vertically, as in the preceding wheels, it could not have escaped freely, and there would have been a correspond- ing loss of head. Hence, the upper part of the tube was constructed as in the previous case, but the lower end has an outward discharge. We have carefully constructed diffusers, both with downward and outward dis- charges, and to obtain clear evidence of what the gain was, have tried the same wheel, both with and without a diffuser, and in every case have found a gain. “The increased economy of water obtained by a diffuser is 2 per cent. When the water leaves the water- wheel, it is traveling with a rapid velocity of about 3% of that due the whole fall. Of this loss, 2% may be saved by a diffuser. If the wheel has 200 h. p., 4 h. p. would be saved, and, if it must be made up by steam, the first year, which would more than pay for the diffuser. “Hence, where great economy of water is desired, a diffuser is a good investment. Every year we make more of them than in previous years. This arrangement here shown is the best possible for many locations, and, under such circumstances, we prefer a diffuser to any other arrangement. We always build the diffuser of cast iron, believing that the only way to obtain a satisfactory arrangement.” Water-Wheel Testing The question has arisen, at various times, as to the accuracy of results obtained at the present Holyoke Testing Flume. Before discussing this mat- ter, a rough survey will be made of all the large-scale turbine tests of record. Such tests as those made by Francis, on the Tremont turbine (Boy den), the Boott center-vent, the Booft Swain, and the Tremont and Suffolk Humphrey are above suspicion. They were carried out under scientific conditions, with practically no regard for expense, by the most eminent hydraulic engineer in the United States, if not in the world. The test of a 42-in. Swain wheel. Pig. 34 . — Risdon Diffusek. there would be a saving of about $200 1290 THE AMERICAN MIXED-FLOW TURBINE AND ITS SETTING at Lowell, by Mills, in 1869, has been questioned by James Emerson, but un- justly it would appear, and without good reason. In 1859-60, the City of Philadelphia conducted a series of tests at the Fairmount Water-Works. They are understood to have been rather crude, and the results are not considered trustworthy. Instead of using a Prony brake for the measurement of power, a method was used whereby the wheel hoisted a given weight. The wheels tested were Jonvals and early scroll^ wheels; a Jonval built by Stevenson gave the best results. The wheel tests at the Centennial Exhibition in 1876 were tested under the direction of Samuel Webber. The highest results were given by the Kis- don wheel. The following is quoted from Mr. Webber’s discussion of the paper by the late R. H. Thurston, M. Am. Soc. C. E., on “The Systematic Testing of Turbine Water-Wheels in the United States”, before the American Society of Mechanical Engineers:* “There have been some remarkable tests reported from the Old Flume [Emerson’s], which it has been impossible to repeat or duplicate at a later date with the same wheels, and the 90% test of the Risdon wheel, referred to ,, by Professor Thurston, is one of them. I have no doubt of the correctness •' of the Centennial test, which gave 87.68% net effect from this wheel, for ] other wheels tested by other engineers, in various places, corroborate it very closely, but I have never myself got so high a result from any other wheel; f| and it should also be noted that the very high efficiency reported from some | wheels has been usually found in the tests of very small wheels, of 15 or 20- ^ in. diameter, where a very considerable effect might be exercised upon the ; wheel by the man who handled the lever at the brake.” j In the same discussion, another view is given as to the accuracy of the 1876 tests, as stated by the late Charles E. Emery, M. Am. Soc. C. E. After describing how petty politics resulted in the appointment of Mr. Webber, he ' says : “As the only object was to obtain reliable and creditable results, the writer undertook to co-operate with Mr. Webber in the conduct of the work. This, ; however, did not prove an easy task, as Mr. Webber had made a great many tests, had acquired certain methods of his own, and did not care to go into many of the refinements which such an opportunity would have made of great scientific value. 'While he was pleased with the action of the Judges, he^ evi- dently felt, moreover, that his authority was from another source, and insisted ! practically on having his own way, and under the circumstances little else could be done. The original design was such that the water was necessarily admitted to the weir approach at the side, so that, although ample provision had been made in the length of the approach, the current on the side opposite the inlet was very much the stronger. The speaker suggested a series of baffling screens or racks, such as are described in Mr. Francis’ work, and, finally a single one was hastily applied, but with the spaces between the bars * so wide that there was still ample area for the water with greater velocity to pass along the side where it did before. Any further improvements were, however, considered unnecessary by Mr. Webber, and the experiments were conducted with the water approaching the weir at very different velocities on the two sides; and, moreover, on account of th e recoil of the current from • Transactions, Am. Soc. Mech. Engrs., Vol. VIII (1887), p. 359. THE AMERICAN MIXED-FLOW TURBINE AND ITS SETTING 1291 one side of the approach, floating bits of wood showed that all of the water did not approach the weir at a right angle, and the evidence of this deflected current was sufficiently marked to show a ridge in the crest of the fall at the weir itself/^ James Emerson was a most interesting and outstanding character during the closing years of the “cut and try” period. He had been a sailor at one time, but soon after the close of the Civil War, he became interested in the testing of water-wheels. He was associated with early wheel tests at Lowell in 1869-71, and afterward moved to Holyoke, where he opened a commercial flume for the testing of wheels. The business prospered and many wheels were tested during the Seventies. A number of years were required to bring all the makers to a belief in the flume test, but Emerson’s efforts were finally successful, as is shown by the prestige of the Holyoke test at the present day. James Emerson deserves the credit for the early development of the testing system, which was as necessary to the development of the reaction wheel as the “cut and try” period. One without the other would have been useless, but the combination of the two was what produced the “Hercules”, the “Victor”, the “McCormick”, and the “Samson” turbines. Emerson was a man of decided ideas and strong prejudices. His attitude, unfortunately, was that, if a maker did not have sufficient faith in his wheel to have it tested at Holyoke, which incidentally meant money in Emerson’s pocket, the wheel was beneath his notice, and, as far as he was concerned, did not exist. Most of the wheels of the period, however, did go to his flume, and thanks to him, there is a rather complete list of tests on early American wheels. The absolute accuracy of his tests might well be questioned, but, comparatively, they must have been correct within 2 or 3 per cent. Emerson was of extreme iconoclastic tendencies, and no institution in existence seemed to have escaped his abuse at some time or other. His most unusual book, “Hydraulics, Dynamics, etc”,* treats of everything from tech- nical matter concerning water-wheels to notes on the Bible, theology, law, woman suffrage, and spiritualism. He did much to educate the public by his annual reports and by his monthly magazine. The Turbine Reporter. What- ever were his bad points, his name is closely linked with the history of the development of the American mixed-flow turbine. Emerson’s testing flume was taken over by the Holyoke Water Power Company and, a little later, in 1881, the present testing flume was built under the direction of Clemens Herschel. The standards of the Holyoke test, from this time on, were the same as they are to-day. Concerning the accuracy of Emerson’s tests, Mr. Herschel states: “So far as I l^now, the Emerson tests are reliable, with the exception of some that must needs have failed, due to the routine, or mechanical, that is, hasty and perfunctory, manner in which they were in the course of time all conducted”. As the arrangement of the present Holyoke flume is well known and has been described by many writers, it will not be necessary to describe it in detail. The water is measured by a sharp crested weir with which little fault * Willimansett, Mass., 1894. 1292 THE AMERICAN MIXED-FLOW TURBINE AND ITS SETTING can be found. The power is measured with a dynamometer of the Emerson type, consisting of a broad-faced cast-iron pulley, with a wood-lined metallic brake band, cooled by a constant flow of water. Several sizes of dynan^ometers are used, depending on the size of the wheel. With a heavy load, the brake seems likely to stick and work unevenly at the lower speeds. It is thought that a considerable improvement might result from the use of an Alden absorption dynamometer. The late Professor Thurston stated* that the limit of error at Emerson’s flume was from 3 to 5%, and that: “A careful examination * * * has led the writer to the conclusion that the Holyoke Testing Flume, and the methods of observation and calculation employed there, are capable of giving the- efficiencies of turbines tested cor- rectly within a limit of error of certainly less than one per cent, and prob- ably to within one-half of one per cent. For all practical purposes the results of the trials of turbine water-wheels, at the Holyoke flume, may be taken as exact, and absolutely trustworthy.” In conclusion, it is only fair to state that, although all available sources that gave promise of any useful information have been consulted, there may be a few omissions. An historical study is no better than the sources con- sulted, and, in the present case, the sources were, for the most part, scant and scattered. Many thanks are due to Mr. S. S. Kent, Assistant Engineer, The Pro- prietors of the Locks an d Canals, for his help in preparing this paper. • “Systematic Testing of Water-Wheels”, Transactions, Am. Soc. Mech. Engrs., Vol. VIII, p. 49. DISCUSSION ON THE AMERICAN MIXED-FLOW TURBINE 1293 ■Discussioisr B. F. Groat,* M. Am. Soc. C. E. (by letter). — There is no better means of becoming familiar with the fundamentals of any type of engineering devel- opment than by way of history. Development history is bound to be a chronicle of successes and failures, showing indisputably the effects of right and wrong efforts. The history of the turbine and its setting embraces the true science of its design. A little study of that science will convince any one that it is not essentially the result of individual effort, but the inevitable result of civilized effort under the influence of economic stress. These truths are made clear by the authors, who essay to demonstrate the truth of the old adage, “there is nothing new under the sun,” by showing that there is nothing new under the water. However, their paper discloses a record of advance that surely means positive improvement in things, which, although they may not be new, have been made to yield a new result, or record. Let credit then be given to the man who made the new record for that advance, whether it be large or small. There is nothing so discouraging to the scientific laborer as the failure of those about him to grasp his point of view. It may be vital to an intelligent appreciation of the value of his work. It is the cheapest remuneration he can be accorded. The authors are certainly correct in their view that progress has been made by a multitude of small steps rather than by large strides. They also exhibit, in a striking manner, that some of the steps have been backward, only to be retraced contrariwise. In particular cases, it is sometimes hard to say whether the success or the failure has led to the positive advance. Thou- sands of supposed inventions are merely the results of ignorance, and may easily be shown to be only the discarded products of richer experiences of the past. There is always the chance, however, that changed conditions may create a new demand for an old device to a new use. The draft-tube, as described by the authors, is a good example. At first, it was used apparently for the sole purpose of elevating the turbine above the tail-water. When scientific views broadened and economic conditions changed, the diffuser principle became better known, with the result that draft-tubes are now in general use. It is doubtful whether there has been any great increase in efficiency, as applied to turbine runners per se, since the time of Boy den and Francis. Hydrodynamically, the efficiency of a turbine depends almost entirely on the entrance and exit angles of the buckets, supposing that water passages are smooth and curves gentle. Nothing proves this truth of science more certainly than the astonishingly uniform results of tests on good designs of the greatest variety. The curvature of the bucket is of little consequence as long as it leads easily from entrance to exit. This is one of those fundamental condi- tions of Nature which makes successful designers of turbines, just as a rich soil makes good farmers. No disrespect is intended for engineer or farmer; both must be proficient to secure the best results. The idea is merely to remind ourselves of the opportunity that Nature offers, namely, 100 per cent. * Civ. and Hydr. Engr., Pittsburgh, Pa. 1294 DISCUSSION ON THE AMERICAN MIXED-FLOW TURBINE After the main features of turbine runners had been developed by the eminent engineers mentioned by the authors, and proper designs had been established, the succeeding considerable advances were to be looked for in the setting. Every form of runner requires its individual setting. It is in the setting that the greatest advances have been made during recent years. It is not improbable that any good runner of the older time, such as the Swain runner, shown on Eig. 12, with slight modification and a specially designed setting, could be made to yield an efiiciency about as high as that of the best modern runners in their settings. As to the classification of water-wheels, the list and method given in Table 2 were worked out by the writer at the University of Minnesota between 1901 and the time of Basshuus’ classification. Let A, P, and h, be the required speed, power, and given head for a water power development. Solve for I = P -f- fi^. Divide I by the index of any wheel and the quotient will be the necessary number of such wheels. Multiply _ the speed coefficient by N, and the result will be the required diameter of such wheels. Divide the total required power by the previously obtained num- ; ber of wheels and the result will be the power of each wheel. In Table 2, the ; columns headed ‘‘Range” show the probable errors occasioned by using the cor- responding coefficients, the various sizes of wheels of any given design varying ^ more or less as to relative performance among themselves according to the : design ; the foregoing data have been taken from the published catalogs of the various makers and probably refer to performance at best speeds in all cases. j The data for any individual wheel given in Table 2 were calculated by students in the School of Mines. The “index” was determined by the J equation : I N^P ' = I in which N is the speed of the wheel, in revolutions per minute, P, the num- . ber of horse-powers, and h, the head, in feet, acting on the wheel. It is ! apparent that this is the square of the specific speed as determined by Basshuus. Although the specific speed is a useful constant for the water-wheel | designer, its square is equally useful to the hydraulic engineer who wishes to ^ examine a number of types of wheel with reference to their applicability to a given water-power project. The use of this constant is explained in a preced- ing paragraph. D. W. Mead, M. Am. Soc. C. E., aptly calls it the “specific ; power” coefficient.* It is the power of a wheel of the same type that will run at unit speed under unit head. It is surprising to know that the beveled buckets of the Houston wheel, ; showing an efficiency of 88%, have almost disappeared. It would seem that : some such plan as this for receiving the water on the buckets would be appreciably better than the present plan, which involves a right-angle turn, although not so abrupt as a superficial view of an axial cross-section might: suggest. Probably questions of construction have led to the abandonment "Water Power Engineering. DISCUSSION ON THE AMERICAN MIXED-FLOW TURBINE »-»• to to io 10 CO to CCK-itOCOtO ktfc. CO 00 h-A CO CO to to 1-1 h-i to to >-iO:^»^tOOrf^C5COO-.?0~?0505 0Hf>-C0»0t0 — ggg§g-ggg5g§8ggggg^g§ggg8§g§8g§§§o8£5;g Index. 10 i-iioosoo^oo^ij^Oh^oowoj-i^tfs-Jw^wcceociocSi-iooo" co-cso!^ Range, percent- age. oot^ls^w^^t^tlooooooo* OO^oSo!^ODO^I^OOOt^^OH^^OOOOOO--* Qc OJ o ’to • H-t OS in io to ’>-1 — bi in o ’-j ’c» •" o htx L ^ on Range, percent- age. , I-.1-H-1. . • . toi-ii-i- h-.toto • • >-i tei-.i-.to • to- o “Cj Oi -1 00 • oso — os- • • ooioot-.a5- os* os h-. i-. to to o co Number of wheels. Victor High Pressure Water-Tight Syracuse High-Head Type Snoqualmie Type Little Giant Std. Flume Std. Cyl. Gate Standard High Duty Std Special High Head Little Giant Std Deep Flume National Poole-Leffel Special Water-Tight Deep Scroll L. C. Cyl. Gate Bradway Patent Flenniken New Pattern Hunt D. C. Cyl. and Reg. Gate Special High Duty Victor Reg. Gate “1900” Water Tight Crocker McCormick’s Holyoke McCormick New Am. Special .Samson (1st set) United States Victor Std. Cyl. Gate Standard Victor Inc., Cap. Cyl. Gate.... Samson (2d set; Leviathan Imp. New American Smith Wicket Gate Name of wheel. Platt Iron Wks. Co Wm. Bartley & Sons Alex. Bradley & Dun Dayton Globe Iron Wks. Co Platt Iron Wks. Co Munson Bros. Co S. Morgan Smith Co James Leffel & Co Risdon-Alcott Co Trump Mfg. Co Munson Bros. Co Munson Bros. Co Case Wheel & Mill Co Poole Eng. & Mach. Co Wm. Bartley & Sons Munson Bros. Co Risdon-Alcott Co C. P. Bradway Me. Wks Dubuque Tur. & Roller Mill Co Rodney Hunt Ma. Co Risdon-Alcott Tur, Co Risdon-Alcott Tur. Co Platt Iron Wks. Co Wm. Bartlev & Sons E. D. Jones & Sons Rodnej' Hunt Mch. Co Dubuque Tur. & Roller Mill Co S. Morgan Smith Co Dayton Globe Iron Wks. Co James Leffel & Co Camden Wt. M^h. Wks Platt Iron Wks. Co Trump Mfg. Co Platt Iron Wks. Co James Leffel & Co Risdon-Alcott Tur. Co Dayton Globe Iron Wks. Co S. Morgan Smith Co Manufacturer. Dayton, Ohio. Bartley, N. J. Syracuse, N. Y. Dayton, Ohio. Dayton, Ohio. Utica, N. Y. York, Pa. Springfield, Ohio. Mt. Holly, N. J. Springfield, Ohio. Utica, N. Y. Utica, N. Y. Bristol, Conn. Baltimore, Md. Bartley, N. J. Utica, N. Y. Mt. Holly, N. J. Camden, N. Y. Dubuque, Iowa. Orange, Mass. Mt. Holly, N. J. Mt. Holly, N. J. Dayton, Ohio. Bartley, N. J. Pittsfield, Mass. Orange, Mass. Dubuque, Iowa. York, Pa. Dayton, Ohio. Springfield, Mass. Camden, N. Y. Dayton, Ohio. Springfield, Ohio. Dayton, Ohio. Springfield, Ohio. Mt. Holly, N. J. Dayton, Ohio. York, Pa. Location. 1295 TABLE 2. — Partial List of American Water-Wheels with Average Coefficients for Each Type. 1296 DISCUSSION ON THE AMETIICAN MIXED-FLOW TUKBINE of this feature, at least temporarily. This is to be regarded as a setting feature rather than a runner feature. It is quite possible that there is room for an improvement here. The authors have stated that the best results to be obtained from a horizontal setting (double unit) will always be “from 2 to 4% less than the efficiency obtained from one of the runners tested, on a vertical shaft, under the best conditions.” With this statement the writer cannot agree, although he is aware of the prevalence of the opinion. He believes that by some modifi- cations of the White hydraucone, it will be possible to derive, from a given runner, the same efficiency in horizontal as in vertical settings. The value of a rotary discharge in some such device has not yet been fully appreciated, nor realized. A rotary discharge is all that is required to increase the efficiency of a horizontal unit. The paper names a very high efficiency obtained on a place test of the Gardner’s Falls plant. Such a high value, although possible, raises the ques- tion of the kind and precision of the test. It would seem desirable to have a detailed description of this test, especially as to the determinations of power, discharge, and head. A testing code for hydraulic turbines is now under discussion for adop- ! tion. Therefore, all matters throwing light on methods of testing would be particularly valuable at this time and the authors are certainly high in j authority in this branch of hydraulics. With reference to one of the earliest vacuum tests on a draft-tube, Mr. Gates Curtis is quoted as saying (page 1281) ; “We are satisfied by experiments made with pipes leading from the top and !l other parts of the draft-tube attached to glass and mercury gauges that we get • equal results from the wheels, as we would were they placed at the level of the ^ tail-water. This arrangement of taking ofi power so far above the tail-water, f thereby avoiding the annoyance and expense of bevel gearing, is proving most : satisfactory in many ways. * * *” q How can the majority of the framers of the turbine test code exclude the | vacuum test in the face of its obvious value? Is it because such tests fre- j quently locate losses of 3 and 4 ft. of head in the draft apparatus ? | The most pleasing and valuable parts of the authors’ paper are the refer- , ences to the leading men in the development of hydraulics in America and ; in the portraits given. j Charles W. Sherman,* M. Am. Soc. C. E. (by letter).— The part of this , paper relating to the development of the turbine and describing and illustrat- ing some of the early types of wheels, is of particular interest to the writer. Such information is difficult to obtain. Little or nothing beyond conjecture | is to be found relating to the efficiency of the earlier types of wheels, and : although information of this character is generally of slight significance at present, questions occasionally arise in which such data would be of service. A number of years ago, the writer had occasion to investigate the water- power privileges affected by the diversion of the water off a small drainage area, from the head-waters of one of the tributaries of the Nashua River, for * Cons. Engr. f Metcalf and Eddy), Boston, Mass. DISCUSSION ON THE AMEEICAN MIXED-FLOW TURBINE 1297 the water supply of Fitchburg, Mass. The fall in this stream was consider- able, and was utilized at a number of small privileges, in addition to which there were several abandoned privileges. A large percentage of the plants located on the smaller privileges were using old and comparatively inefficient types of wheels. Some of the wheels were inaccessible, and no record was available of the size, make, or other particulars. As it is generally accepted as an axiom, that undeveloped or abandoned water-power privileges are entitled to less damages for diversion of water than fully developed privileges, it follows that partly developed privileges, or those using inefficient types of wheel, are entitled to less damages than those fully developed, with efficient power plants. Information on which to base satisfactory estimates of the efficiency of the existing plants, therefore, would have been of real significance in this investigation, but practically nothing was available. A few old wheels of the types of some of those in use, were found in mill yards, in cases where they had been superseded by better wheels, and photographs and measurements of some of them were obtained. Poor reproductions of some of these photographs have been pub- lished by the writer.* They are, however, sufficient to show in general the type of runner used in the Pose and Blake wheels. The owner of one of the mills stated that a Pose wheel tested by his father showed an efficiency of 62% when new. No information was available as to the method of the test, and the accuracy of this determination is perhaps open to question. Among the other types of wheels found on this stream were Humphrey, Pider, and Stevens (made at Ayer, Mass.), as well as Hercules and Leffel turbines of very old types. C. M. ALLEN,t M. Am. Soc. C. E. (by letter). — The authors have mentioned one or two old installations which are still being used and which give approxi- mately the original efficiency. The writer has made a number of tests of efficiency on old installations and has found that they were still doing excellent work. For instance, two vertical 80-in. Boy den wheels in Holyoke were tested by mounting the brake on a horizontal jack-shaft. The bevel gears in both cases were of cast iron. The discharge from each wheel was carefully meas- ured over a standard weir by A. F. Sickman, Assoc. M. Am. Soc. C. E., Hydraulic Engineer for the Holyoke Water Power Company. The highest efficiency of the wheels at full gate, which in this case includes the losses from gears and bearings, was 83.5% and 83.7%, respectively. At the time of this test, these wheels had been in operation for more than thirty years. In another case, a Hercules wheel which had been installed in 1886, was tested thirty years later, both before and after cleaning. The brake was mounted on the horizontal jack-shaft, allowing a gear loss of 4% ; this test checked the original Holyoke test of the runner. The wheel had been idle for some time and was covered with tubercles. After it had been cleaned, there was an increase of about 40% in the power. * Journal, Boston Soc. of Civ. Engrs., Vol. Ill, p. 23. t Prof, of Hydr. Eng., Worcester Polytechnic Inst., Worcester, Mass. 1298 DISCUSSION ON THE AMERICAN MIXED-FLOW TURBINE Although, occasionally, old wheels are found that are still doing as good work as they ever did, the majority are doing otherwise; altogether too many old wheels are operating, the efficiencies of which are not more than 50 per cent. Dana M. Wood,* M. Am. Soc. C. E. (hy letter).— The authors have asked for additional data of historical interest regarding old water-wheels. Fig. 35 illustrates an interesting early design of wheel. The turbine apparently occupies its original position, but the mill with its equipment is gone, and the dam, seen in the background, is in the last stages of decay. The head was apparently about 8 ft. The turbine was set on the bed of the river and had no draft-tube, the water being discharged upward and outward. There were evidences of a 6-ft. power flume. The cast-iron gate shown, admitted the water to a spiral casing, and the shaft, with coupling on top, indicates that the turbine rests in its original position. The plate, visible on the gate-casing, bears the inscription, ‘‘Geo. H. Jones, Little Giant Turbine, manufactured by J. C. Wilson & Co., Picton, Ontario (Canada), 1870”. The power site is at Paquetteville, Que., Canada, on a small stream tributary to Hall Stream, in turn tributary to the Connecticut River. H. A. HAGEMAN,t M. Am. Soc. C. E. (by letter).— The authors have pre- sented an interesting paper in reviewing the history, in the United States, of the development of the pressure type of water-turbine runner. The mixed-flow turbine and its appurtenances have passed through trying periods in their development and have required the persistent efforts of many hydraulicians for more than flfty years to perfect the modern unit of the Francis type. In the early Nineties, turbines of the Francis type, having either vertical or horizontal shafts and- equipped with one or more runners, could be obtained only for medium or low heads, low powers, and medium speeds. Prior to that period, the power was transmitted from the turbines to the driven load by gears or pulleys, and turbines were seldom direct-con- nected to the machine to be operated. About that time The Cataract Construction Company, of Niagara Falls, N. Y., the predecessor of the Niagara Falls Power Company, was preparing to develop its first hydro-electric plant and was in the market for three 5 000- h. p. turbine units to operate at 250 rev. per min., under a head of 135 ft. The turbines were to be direct-connected by long shafts to alternating-current generators. The International Commission of Engineers, headed by the late Dr. Coleman Sellers, of Philadelphia, Pa., which the Company selected, invited bids from the leading manufacturers of water turbines throughout the world and finally chose a design by Faesch and Picard, of Geneva, Switzerland. The turbines selected were of the double-runner, Fourneyron type, without draft-tubes, and were built and installed by the I. P. Morris Company, of Philadelphia. The turbines and the generators of these units were un- precedented in capacity and were the admirati on of the Engineering Profes- * Engr.. Stone & Webster, Inc., Boston, Mass, t Hydr. Engr., Stone & Webster, Inc., Boston, Mass. Fig. 35. — Old Tukbine at Paquettevillb, Que., Canada. DISCUSSION ON THE AMERICAN MIXED-FLOW TURBINE 1301 sion. This installation was not only successful, but became the pioneer show plant in the central-station hydro-electric field, and, later, when all ten units were installed, it had a total capacity of 50 000 h. p. For many years, this plant was the largest hydro-electric station in the world. Keplacement of the original wheels by single-runner Francis turbines and draft-tubes was com- menced about 1910. The successful operation of this plant gave the central hydro-electric sta- tion the necessary opening for the future design and construction of large direct-connected water-wheel generator units. In 1899, when the present plant of the Utica Gas and Electric Company at Trenton Falls, FT. Y., was constructed, the Fourneyron type of turbine was again selected. The four units installed were of the single-runner, vertical- shaft type with draft-tubes, each with a capacity of 2 000 h. p., under a 264-ft. head, at 360 rev. per min. It seems to the writer that the two plants mentioned, containing turbines of the Fourneyron type, were largely responsible for interesting the designers of water-wheels in devoting greater effort to produce Francis turbines for much higher heads and power than could be obtained prior to 1890. This is particularly so on account of the higher part-gate efficiency of the Francis runner and also because of its better mechanical design and low maintenance cost. The result was that, about 1901, Escher, Wyss and Company, of Zurich, Switzerland, designed a turbine unit having a single runner of the Francis type, draft-tube, and vertical shaft for Plant FTo. 2 of the Niagara Falls Power Company. The capacity of the unit was 5 500 h. p., operating at 135-ft. head and 250 rev. per min. The completed plant contains eleven of these units and has been in successful operation for twenty years. From 18,90 to 1905, the foremost designs of water-turbine units were made in Europe, but about 1905, American designers began to produce large turbine units of the Francis type and take from the Europeans the leadership in the art. During the past fifteen years, turbine engineers in the United States have led the world in design, have produced Francis wheels of greater specific speed, head, power, and efficiency than ever before, and have developed this type of runner ahead of any other. The general design requirements of water-turbine units are the same to-day as they have always been, but with the perfection of the electric generator, the commercial possibilities of the hydro-electric unit has increased enormously, requiring the refinement of old designs and the development of new ones to meet the new conditions brought about by competition with other forms of prime movers. Development of water-turbine design has ever been progressive, more so during the Twentieth Century than before, and the general tendency has been to improve the efficiency of all the component parts of the unit, besides simpli- fying and bettering the mechanical details of construction. When the single- runner, 5 500-h. p., Francis turbines mentioned were installed at Niagara Falls, they were the largest units in the world. Since then the head, power, and .speed h.avq been greatly increased, and units for single runners with capacities 1302 DISCUSSIOlSr OlsT THE AMERICAN MIXED-FLOW TURBINE of 55 000 h. p. are in operation and similar units of 70 000 h. p. are being con- sidered. Thirty years ago there were few Francis turbine units operating under heads of more than 100 ft., whereas, to-day, units of this type are operating under heads as high as 800 ft.', and the head or power limitation has not been reached. Apparently, the authors are inclined to pass, almost unnoticed, the achieve- ments of hydraulic engineers and designers of water turbines of the past twenty- five years, as represented by a group of men, many of whom are, to-day, a part of the organizations of the leading water-wheel manufacturers in the United States. Although the writer would not detract in any way from the credit which is due to the designers and engineers mentioned in the paper, he would give much credit to those engineers who, in recent years, have added inventions and improvements to the water-wheel art, and developed the Francis turbine unit as a whole, making it an efficient machine i^f the highest type. Egbert E. Horton,* M. Am. Soc. C. E. (by letter).— Much of the history of the development of the American type of turbine is lost. The subject is complex, as inventors in different localities developed similar ideas more or less simultaneously. Some important features of this history, however, could be recovered from a thorough study of the records of the United States Patent Office. Mr. John B. McCormick has a full set of the water-wheel patents granted by the United States Government down to 1880, which the writer has been permitted to examine. From these and from an index of water- wheel patents. Table 3, which gives a list of a few patents that represent landmarks in turbine history, has been compiled. TABLE 3. — Important Early Water-Wheel Patents No. Date. To. 861 1 .S76 759 2 599 2 708 3 153 3 510 4 056 5 090 5 144 28 314 122 275 1804 Oct. 29, 1829 Oct. 22, 1830 July 26. 1838 Oct. 18, 1839 May 30, 1838 Apr. 30, 1842 July 8. 1842 July 3, 1843 Mar. 26, 1844 May 21, 1845 May 1, 1847 June 5, 1847 May 15,1860 Dec. 26, 1871 Benj. Tyler. New Hanapshire, “Wry Fly turbine”. Zebulon and Amasa Parker. Calvin Wing, Gardiner, Me. Samuel B. Howd, Geneva, N. Y. T. Rose, Windsor, N. Y. N. Johnson, Triangle, N. Y. Samuel B. Howd, Arcadia, N. Y. R. Rich. Wbitelaw Stirratt, Paisley, England. N. Johnson. J. Leffel, Springfield, Ohio. U. A. Boyden. (Diffuser.) U. A. Boyden. A. M. Swain, Lowell, Mass. M. W. and J. T. Obenchain. The so-called “American type”, large capacity, high-speed, mixed inward and axial flow turbine is undoubtedly almost wholly an American invention, but it is not the result of the labors of any one man or generation. As it stands to-day, it embodies several distinct and probably essential features, (a) inward flow of the water to a vortex chamber surrounding the entire runner; (h) a volute or scroll case, designed to give the inflowing water the proper rotary motion in the vortex chamber; (c) pivot guide-gates; and (d) a draft- tube to recover the energy remaining in the water discharged from the runner. ♦ Cons. Hydr. Engr., Albany, N. Y. DISCUSSION ON THE AMERICAN MIXED-FLOW TURBINE 1303 In view of the manner in which the water operates and the feature which most distinguishes this from all other types of turbines and water-wheels, namely, the inward flow to a vortex chamber surrounding the runner, it is suggested that the designation, “volute case vortex turbine”, would better describe this machine hydraulically than the somewhat indefinite designation “American type of turbine” now in common use. It would also afford a basis of getting away from the designation “Francis turbine”, a term of European origin which, for reasons hereafter stated, appears to the writer to be objection- able. If, as the writer believes, the most essential feature of this turbine was the invention of Samuel B. Howd, then following the precedent used in the case of the Fourneyron turbine, the American type should properly be called the Howd turbine. There is, however, a distinction in that in all its essential features the Fourneyron turbine stands to-day practically in the same form in which it left the hands of its inventor. The volute case vortex turbine, however, involves at least three other features generally deemed essential, which were not contributed to it by Howd. These are the volute case, pivot guide-gates, and draft-tube. The authors seem inclined to ascribe to Howd a moiety of the credit and to give Francis a large modicum. The writer has no desire to detract from the just reputation of one of America’s greatest hydraulicians, the late James B. Francis, Past-President, Am. Soc. C. E. In a paper covering much the same ground as that of the authors, but in somewhat different form, the writer stated*: “The first suggestion of a wheel of this construction appears to have been made by Poncelet in 1826. The early American inward flow wheels were, how- ever, but a modified form of the central discharge wheel which, in turn, was evolved from the rouet. In 1838, a patent was granted to Samuel B. Howd, of Geneva, N. Y., for a simple form of inward flow wheel. Wheels made by Howd were extensively used, and were known as “United States” turbines. “Following the general lines of the Howd wheel a turbine was designed by James B. Francis in 1849 for the Boott Cotton Mills. This wheel, constructed at the Lowell Machine Shops, was experimented upon by Francis and many similar wheels were afterward constructed on the same pattern. As the Fourneyron wheel differed from its predecessors mainly as a result of high- grade mechanical construction, so also the Francis center- vent wheel differed from the Howd and other earlier inward flow turbines. Regarding this wheel, Francis says in ‘Lowell Hydraulic Experiments’ : ‘In the design for the Boott wheel the writer has so modified the form and arrangement of the whole as to produce a wheel essentially different from the Howd wheel as above de- scribed, although it may possibly be technically covered by the patent for that wheel’. As above stated, impartial comparison would hardly leave any thing to the credit of Francis in the invention of this wheel, although the construction was so much improved as to make it, as he says, ‘essentially a different wheel’. In fact, the progress of the last century in industry has been very largely the result of improved mechanical construction of machines, the principles of operation of which have been known from an earlier period. The center-vent wheel was quickly displaced in America by the inward and downward flow American stock pattern type of turbine, and it is probable there was not a single wheel of the Francis type in use in this country for nearly half a century. In the meantime, the simple inward flow turbine was extensively used on the European continent under the name of the Francis * “The Turbine Water Wheel as a Prime Mover”, Bulletin, Clarkson Inst, of Technology, Potsdam, N. Y., Vol. 7, No. 1 (January, 1910). 1304 DISCUSSION ON THE AMEKICAN MIXED-FLOW TURBINE turbine. To meet the demand for a compact, large capacity, high-head water- wheel, it is being reintroduced into the United States, following the European designs, and it is now known here as the Trancis turbine’, . although it has the volute or spiral case developed abroad by Schiele and commonly used in America with crude central discharge wheels. It has the pivot gates of Eincke and probably resembles the original Boott turbine of Erancis less than that resembled the cruder Howd wheel. It is quite certain that this type of wheel is to be the hydraulic prime mover of the future.” - The writer has not experienced any important subsequent change of view in this matter, except to come to the conclusion that the work of Howd was of relatively greater, and that of Erancis of less, importance in this connec- tion than the foregoing quotation indicates. Bef erring to the Howd wheel patented July 26th, 1838, Erancis says, a wheel similar in its essential features was proposed in France in 1826 by Poncelet”.* The Poncelet turbine referred to is illustrated and fully de- scribed by Weisbach.f He says : ‘‘The first idea of an inward fiow tangential wheel is due to Poncelet”. This wheel, as shown by the illustration and as j intimated by the designation, was fed by a tangential spout, the water enter- j ing on a small arc of the circumference only. It did not embody the prin- -I ciple of the vortex chamber which was that the water surrounding the wheel j should rotate at such velocity that it could enter the runner without shock, j That is the essence of the Howd type of turbine and of the modern scroll case ' vortex turbine. The writer has been unable to find any precedent in the historical or patent literature of hydraulic motors for an inward flow turbine | fed around the entire circumference, antecedent to Howd’s patent of 1838. In order to get a better perspective of the situation attending Howd’s invention, it is worth .while to consider the opinions and practices of the times in water- j wheel matters. q At that date, there were no great firms engaged in the manufacture of 1 water-wheels. It was the usual practice for the millwright to build the water- ^ wheels for the mill at which he was employed, and as a rule the old-time mill- ! wrights were exceedingly staunch in their opinions and jealous of their i prerogatives. On the one hand, the dominant water motors were of the hori- • zontal, overshot, and breast types; and on the other, wooden spout-fed flutter, j tub, and scroll case, central discharge wheels. Eourneyron had invented the , outward flow turbine in Europe a few years previously, had brought it to a ' high state of mechanical excellence and efficiency, and its fame had spread the i world over. Simultaneously with Howd’s invention, Morin’s classical experi- j ments on the Eourneyron turbine were being conducted, the results of which , were shortly translated by Ellwood Morris and published in the Journal of | the Franklin Institute. Almost immediately following Fourneyron’s invent | tion, crude outward flow turbines had been brought into use in the United i States to meet the demand for a relatively high-speed water-wheel to drive the vertical English gate and muley saws in the mills of the pioneers.^ Calvin^ Wing had patented such an outward flow turbine without guides in 1830.' * “Lowell Hydraulic Experiments”, p. 61. t Du Bois, “Mechanics of Engineering”. Vol. II. Hydraulics and Hydraulic Motors, p. 364. DISCUSSIOTT ON THE AMERICAN MIXED-FLOW TURBINE 1305 The use of such wheels spread with great rapidity, despite the limited means for disseminating information in those days, and during the period from 1838 to 1850, when Howd’s work was performed, numerous patents were issued on cast-iron, outward flow turbines identical in principle with that of Calvin Wing and including the well known Rose, Johnson, and Rich turbines. The tendency of the times was strongly toward the use of outward flow turbine wheels. The average millwright had a sort of superstitious belief in the magical efficacy of centrifugal force, which it was well known augmented the discharge and increased the power of outward flow wheels. In the light of these cir- cumstances, the writer is inspired with a feeling of great respect for the moral courage and creative imagination of Howd who undertook, practically without precedent, and in the face of such well established opposition, to develop and commercialize a turbine water-wheel acting on directly the opposite principle by having the water flow inward instead of outward. Unfortunately, little is known or can apparently be learned about the life and work of Howd. When his flrst patent was granted he was a resi- dent of Geneva, N. Y. It so happened that some of the writer’s ancestors were pioneer millwrights and water-wheel builders in the same locality and at the same date. Diligent inquiry, however, has failed to reveal anything regarding Howd. It is entirely true, as the authors state, that Howd did not himself manufacture turbines, but sold patent rights with the expectation that the wheels would be constructed by mill- wrights, at least the writer has not been able to find any evidence to the contrary. In the ‘‘American Miller and Millwright’s Assistant”, by William Carter Hughes, 1851, appears a diagrammatic outline entitled “Howd’s Patent Direct Action Water Wheel”, which is here reproduced as Pig. 36. In the same book (page 57), is an article purporting to have been written by Howd, which contains the following: ^ “This is a wheel which, when properly located, is admirably adapted for mills of all kinds, working the water on the tourbillon principle, being the whirling vortex or better known as reaction principle. Its superiority over the old reaction wheel consists in applying the water on the wheel at the verge and discharging it at the center.” This quotation is offered in view of the author’s statement that “it seems doubtful whether he appreciated the value of centripetal flow.” Hughes’ book also gives “directions for making the several parts of Howd’s latest improved water-wheel and setting it up.” The text indicates that it was presumed that the runner crown would be made of plank, and directions were included for fitting the runner to a wooden shaft, octagonal wooden shafts being the usual type of construction in those days. In the few succeeding years, to the time when James B. Francis, with keen insight, recognized the merits of the Howd wheel, there were great improvements in facilities for mechanical operations. 1306 DISCUSSION ON THE AMEEICAN MIXED-FLOW TUKBINE It is not surprising that, with his own great ingenuity and the mechanical resources of the Locks and Canals at his disposal, Francis should have been able to have built for him by skilled mechanics, wheels of this type of better construction than those of the ordinary millwright, but a comparison of the diagram of the Howd wheel as constrlicted by Francis for the Boott Cotton Mills (Fig. 37) shows the layout of the guides and runner closely resembling Howd’s diagram. Fig. 36. There was no change or improvement whatever in the principle of operation. The writer has a copy of Howd’s Patent Mo. 861, dated July 26th, 1838. The patent drawings include a plan of the layout of the runner and guides practically identical with that shown in Fig. 36. The claim is of particular interest. It is as follows : claim as my invention — the application of the water upon the outside of the wheel and operating upon the principle of reaction by dis^arging inwardly on a wheel constructed and combined so as to operate as above de- scribed with the spouts or shoots giving the water a direction with the motion of the wheel applied to a reacting wheel as aforesaid.” Inasmuch as the writer has not found any precedent for an inward flow' reaction turbine supplied with water around the entire circumference ante- ^ cedent to Howd’s patent, it appears that credit for the invention of this? important type of prime mover clearly belongs to Howd, and that the runner as he originally conceived it, with the possible exception of the use of curved instead of straight guide vanes, was identical with the runner from which Francis and others designed or constructed similar wheels. The writer has felt impelled to take up this matter in view of the fact that the merit of Howd’s invention has been in the past, and appears likely^ to be in the future, greatly dimmed by the reputation of James B. Francis! In view of the fact that the turbine, next to if not indeed transcending the ^ steam engine, is one of the greatest of human inventions and that the high-1 head turbine of to-day embodies the principle of vortex circumferential feed, to the runner precisely as it left the hands of Howd, it has seemed proper to suggest that Samuel B. Howd is entitled to a place in the Hall of Fame as one of America’s foremost inventors. This discussion is given solely in the spirit of ‘'rendering unto Caesar the things that are Caesar’s”, and as stated, not with any desire to belittle the reputation or standing of James B. Francis, In view of the preceding and for other reasons, the writer does not feel willing to accept the “Family Tree” of the mixed-flow turbine as given by th( authors. Under early American turbines are included the spiral. Rich, anc Rose wheels. These, it is true, were early American turbines, but they were not in any sense antecedents of the volute vortex turbine. Again, the Ameri can, Hercules, Victor, and Sampson turbines were offshoots of the Obenchaii and McCormick wheels, which, in turn, were direct predecessors of the moderi: high-speed runner. As a matter of fact, the essential features of the volute vortex turbine of to-day were more or less separately, but simultaneous!:^ developed and cannot be properly ascribed to a common ancestral tree. bub 3 ec| to revision, the direct antecedents of the principal features of the modenj DISCUSSION ON THE AMERICAN MIXED-FLOW TURBINE 1307 volute-vortex turbine appear to the writer to have been substantially as set forth in Table 4. The draft-tube has not been included, as it appears to have been the independent creation of Jonval and has received little subsequent change or improvement, except the outlet cone, undoubtedly suggested by Boyden’s diffuser, and built at an early date by Swain and others, and more recently by Moody. TABLE 4. — Prototypes of Modern Volute-Yortex Turbines, Showing THE Parallel Development of Various Features. Development of inward and mixed inward downward flow principle. Development of the scroll- vortex chamber. Development of pivot guide gates. (1) Scroll or Spout Feed; no Guides: (a) No Reaction: Primitive wooden scroll central discharge water-wheels. ICirc. 1820) (b) With Reaction: Poncelet Turbine (1825) (2) Full Circumferential Feed with Reaction and Guides: Samuel B. Howd, 1838. (3) Increased Capacity and Speed, Howd Type: Francis-Howd 1847 Swain 1860 (4) High-Speed Type: Obenchain— Little Giant 1873 John B. McCormick Circ. 1870- 1885. (5) Propeller Type: Austin wheel. Screw current meter. Marine propeller. Nagler high-speed runner. (1) Vertical Spout-Fed Flutter Wheel (Medieval). (2) Primitive Wooden Scroll Case Central Discharge (Cite. 1820). (3) Parker Scroll Case (Pat. 1829). (4) Iron Scroll Case Reaction Turbines without Guides: Tyler, Reynolds etc., etc. (Circ. 1850) (5) Scroll Case Reaction Turbine with Guides: Warren (Pat. 1860) Early Risdon. Schiele Turbine. (6) Partial Vortex Scroll Case with Pivot Guides: Thomson Vortex Turbine (Circ. 1870) (1) Early Leffel Turbine 1845- 1860. (2) Thomson Vortex Turbine (Circ. 1870). (3) The Fincke Pivot Gate, developed recently in Europe. The authors, quoting Pfarr, credit J. M. Yoith, of Heidenheim, with being the first to build an iron, spiral-cased, wicket-gate turbine, thus incorporating in one machine all the essential features of the latest volute-vortex turbine. Wooden, scroll-case, central-discharge wheels with flat vanes were in common use in Central and Northern New York in the early part of the Nineteenth Century, quite certainly as early as 1820. It is probable that they were origi- nated simply as an improvement or evolution of the vertical spout-fed flutter wheel, which has been traced back to Medieval times. From about 1850 to 1880, scroll case wheels without guides, including the Reynolds, Tyler, and Carley, were in common use. As far as the writer can determine, the Warren wheel, referred to by the authors, was the first scroll case reaction turbine with guides, which appears to have been patented in 1860. It was probably some- what later that James B. Thomson invented his vortex turbine (Fig. 38). This wheel had a volute case, the manner of feed resembling somewhat more closely that of the modern turbine in a concrete volute chamber than that of the high-head volute-vortex turbine in an iron scroll case. It had pivot guides, however, and appears clearly to antedate Yoith’s claim of being the first turbine which incorporated all the essential features of the modern volute-vortex turbine. 1308 DISCUSSION ON THE AMERICAN MIXED-FLOW TURBINE The earliest use of pivot guide-gates is somewhat in doubt, and possibly , originated with the early Leffel wheel. The modern type of pivot guide-gate ; is a recent European invention, commonly ascribed to Fincke. •b Fig. 38. ji With reference to the propeller type of runner as developed by Mr. Forrest ; Nagler, it may be noted that this runner differs from that of the volute-^; vortex turbine in several particulars: (1) . — It is more essentially a downward than an inward flow runner. j (2) . — The crown plate and rim band are dispensed with and the runner^ vanes project radially and are otherwise unsupported from the hub as in a screw.'; propeller. • (3) . — The plane of each runner vane is nearly at right angles to the axis,^; thus giving a runner of high speed of rotation. (4) . — The number of vanes is much more limited than in the case of the vortex type of runner. The first of these features, namely, axial flow, has been common since the days of Jonval. The omission of crown and rim bands, although, perhaps, not, very essential, has been not uncommon in other water-wheels, beginning with' the vertical flutter, tub, and scroll central discharge types. : TABLE 5. Name. Page. Crown plate. Rim band. pif»<*Wst'.nnft 188 Yes No 1 Economy 194 No No JYT. nnmpoiirid 194 No No I Mullikin 200 Yes No ii Coleman 216 Yes No Walsh 335 Yes No Meropr, 355 Yes No ' ♦ Copied from originals on file with the Oswego Canal Company. DISCUSSION ON THE AMERICAN MIXED-FLOW TURBINE Number of experiment. ! men o ooooeooocooooi— -o Duration, in minutes. WW CC WW vt 0»t^05CC0cOGC00;coOh{^>^c;T Total head and fall, in feet and inches. ^ CO ^ M ^ CO 09 M .tcco OdaoacdooA Effective head when wheel out of w'ater from head to center inlet, in feet and inches. wco CO CO CO CO CO CO CO CO CO CO CO CO 4i. hp. >e. err ' 4 ^ 4^ 4^4^^0^4^»^4^4^4^C^O>-th-L^ COh-*- 4i^ — 0 ^ -a — * CO CO •Q 4*^ 4^ 05 Oi Cn 1 + -H+ Effective head, in feet and decimals. ►--00 00 OOCr4^C;TcoboOOQciDtOtoii.^Cn 00 0 0 0 0 0 0 0 0 0 0 Oi 05 CO CO 0 Velocity due to head, in feet and decimals. h-^ *-* 1-* to M W to • 1-^ to to CO to CO to OCO--1-5COCOOOO- 4^QOtOC;TKfiOt--cO V\ OOC5CO^}(Xi>yCX* 1— t— .-*OTOIO»OCO 00 (OIOCOCJIOOCOO* COiOOOOOCOOOO Revolutions of wheel. CO ►-- I-- 1-1 to to to to • ►-* to to CO to to to ^ OC^CDCOOOCO* hP^OOtOC5o Revolutions per minute. a^o 00 oootDcoo* oooco^-^oii-^-toco 00 050509 09* 09b90 4»^bo4^05H-- 05 Velocity of circumference diameter = IOJ 4 in. Bo S Ratio. >-» to CO CO CO CO CO CO 4^ CO CO CO to CO 4^ CO coo OD Oca»Oi05 4*->-*-0'OC0004^000CO 0^1^ 0^ 0^— '^O^GDSOOSC^'J-^nH 0 05 05 Depth of w'ater in reservoir, in inches. Size = 641/6 sq. f t. to to ^ to to to to to to to 0 0 ?D 0 CD 00 to 1-^ 0 0 CO 0 ►-►•o 05 ototo^coao-*^ boo Vi o^cDoocobtwboboboboOQo Quantity of water, in cubic feet. CO CO CO CO CO CO CO CO CC CO CO CO 4>^ 4k. 4^ CD 00 --1 -> -si C5 -> 4 ;*. ^ ^ 00 0 0 H-- UT CO C0:D 00 OOODOOOtOOCODXOOOQOCOOD>-i hx 01 Quantity if flowing through inlet without con- traction or obstruction. Size of inlet = 6.127 sq. in. CCUT oi vxvivxvxvxvxvxvx:jf0x0zvn^^ ^00 4^ 4^CtOCOCOO-:i05C;iCO«p^04^'— • •^4^-q^ -^4k05 Useful effect in pounds, raised 1 ft. per min. CO 4k 4k CO 4k 4k 45 . 4k 4k to C7T ot 4^ tOC5 4k 4k'--C0O4k00C5CJ^C;KJO^t0^1 ►--•kl 0 0 tD«-^ C» to --"X -) O ’ 0 X 0 to C0C5 •-l 05 ment, on page 1238, to the effect that “a wheel of the high-speed propeller type ’ was patented and on the market fifty years ago”. It is granted that the wheels j to which they refer had radial blades few in number, but the high-speed feature ; is questioned. They required the addition of a suitable guide case, the addi-^| tion of suitable diffusers or draft-tubes, the elimination of the rim, approxi- mately 50% reduction of blade area, and a considerable modification of the i blade angles from the substantially helical form originally used, in order to . permit of the possibility of securing high speed as it is understood at present.. It may be of interest to note that all the early designs of wheels of the propeller type discussed by the authors have been considered by the officials of the U. S. Patent Office in their investigations of the Nagler design. These officials have expressed the opinion that these modern designs of high-speed runners and turbine settings, although they include certain old elements, do produce new and useful results and decided advances in the field of hydraulic engineering. The horse-power of the last runner, designated as the modern propeller type. Fig. 21, has been increased about 50% above the figure given. ^ The specific speeds have been increased more than 50 per cent. ^ Table 1 is in- teresting, the most significant features being the high efficiencies and increases of power made by the Swain and Kisdon turbines. In connection with the section on draft-tubes, no distinction is made be- tween the form of modern concentric tube, the measure of the efficiency of which is in the closeness with which it approaches the form of natural hydraucone-, developed and defined by Mr. W. M. White, and the old form. DISCUSSION ON THE AMERICAN MIXED-FLOW TURBINE 1317 which merely provided for transition in a chamber from axial to radial flow. It is the development of the proper form of chamber, that has permitted the attainment ot results, impossible with any of the chambers of which Figs. 32 and 33 are typical. George A. Orrok,* M. Am. Soc. C. E. — The speaker is particularly inter- ested in Table 1 which shows that forty-five years ago runner efficiencies, as tested at Holyoke, had reached a figure in excess of 90 per cent. In 1920, efficiencies from 92 to 93% were common, and runners built in small country shops had given these test results. The speaker believes that the published figures of more than 95% for one of the largest runners and draft-tubes tested in the field, is the present-day limit of attainment, and he is inclined to attribute this more to the cleverness of the runner designer than to the saving due to a perfectly designed draft-tube. F. W. ScnEiDENHELM,t M. Am. Soc. C. E. (by letter). — Some of the efficiencies recorded in Table 1 seem truly remarkable. Great credit is due such men as the designers and makers of the Risdon wheels of 1873 and the “Hercules” wheel of 1876. It is more than likely that the tests of the present Holyoke flume are somewhat more severe in results than the earlier tests made there, thus accounting for the fact that to the better known wheels of the period extending from the construction of the present flume in 1881 to the end of the century are imputed lower efficiencies than the premier wheels of 1873-76. Even if this is true, the results shown in Table 1 for the period prior to 1900 would indicate on the face retrogression in the early years of the present century. However, such is probably not the case. Undoubtedly, the most reliable index of performance of the water-wheels of a given type is the extent to which the manufacturer is willing to guarantee that performance. Still the progress of guaranties of efficiency has not been compatible with so small an increase in attained efficiency as one might infer from Table 1 to have taken place since the dates there shown. Nowadays, a test efficiency approximating 94% is fully as rare as one of 90% is reported to have been in the Seventies; nevertheless, the difference between present- day guaranties and those of, say, 1900, are considerably greater than 4 per cent. At present, there is such a thing as a guaranteed efficiency of 90%; whereas, as late as 1905, guaranties rarely exceeded 80 per cent. A possible explanation is that the highest efficiencies shown in Table 1 represent exceptional, not average, wheels; that the then existing state of the art did not afford types the individual wheels of which were consistently of high efficiency. However, one of the outstanding characteristics of the present state of the art is the consistency which exists in this very respect. If this was not the case, the manufacturer would not be so ready to make guaranties of high efficiency and the purchaser would have to pay a considerable insurance premium to compensate the manufacturer for failures to attain the guaranteed mark. As a matter of fact, failures are now rare. ♦ Cons. Engr., New York City. t Cons. Engr. (Mead & Scheidenhelm) , New York City. 1318 DISCUSSIOJT ON THE AMERICAN MIXED-FLOW TURBINE Perhaps a further explanation lies in the extensive use prior to 1900, and even later, of “stock” wheels, whereas plants of larger capacity justify special runner development wherever the controlling conditions so indicate. In short, much as credit may he due to the pioneers for the attainment of certain exceptional efficiencies, to the present generation belongs the credit of making efficiencies around 90% reliably available. Again, the authors apparently are of the opinion that the results of hydraulic turbine tests at the Holyoke flume are accurate and reliable. It seems, how- ever, that of even greater significance from the practical standpoint is the fact that the field tests of modern reaction turbines of great capacity generally yield efficiencies higher than those obtained in tests of the smaller homologous runners at Holyoke or elsewhere. This is likely the rule only for cases where the head under actual operating conditions is materially greater than the head under test conditions and where, therefore, a constant friction loss is a larger proportion of the power under the lower test head than of the power under the higher operating head. Nevertheless, the writer must confess that, until about ten years ago, he, like many others, was inclined to be skeptical of laboratory tests as a direct measure of the efficiency to be expected from the full-size runner in place. On the contrary, the fact can no longer be success- fully controverted that, unlike laboratory results in general, laboratory tests of turbine runners show at least as low efficiencies as field tests. A difference of 2% is by no means unusual. The development of the hydraulic turbine has been largely empirical. Time and again theory has not stepped in or explained progress until after experimentation, such as marked the “cut-and-try” period described by the authors, had made that progress a reality. The present-day development seems to be no exception. j Thus, the increase in unit or specific speed, under the incentive of lower generator costs, has been taking place gradually. The step to a still higher specific speed for application in a given problem of the moment seems to be hindered, not by theoretical considerations, but only by lack of precedent. The writer believes that the development of reaction wheels has not yet reached the point where one can determine the maximum specific speed which will ulti- mately be practicable for a given head. Similarly, the reaction turbine has been applied successively and success- fully to higher and higher heads. About 1912, the feasibility of application to heads of approximately 600 ft. was a matter awaiting confirmation by practical operation. At present, the reaction turbine has been used with heads of more than 800 ft. The ultimate upper limit seems to await determination by practical test rather than theory. Recently, the writer has had occasion to refer to Weisbach’s “Mechanics of Machinery and Engineering”, 1846-47, as translated from the German and published in Philadelphia, 1848-49. The two volumes of this work had recently been given the writer from the library of the late Frederic P. Stearns, Past- President, Am. Soc. C. E., and, as now appears by signatures in the volumes. DISCUSSION ON THE AMERICAN MIXED-FLOW TURBINE 1319 must originally have belonged to “U. A. Boyden”. This was undoubtedly Uriah A. Boyden, the “father” of American mixed-flow hydraulic turbine design. By reference to the prefatory matter in the second volume of Weisbach, which deals at length with water-wheels and turbines, it appears evident that in 1849 both the English translator and the American editor considered that this was “the only theoretical treatise on water power of the least practical value hitherto printed in the English language”. Inasmuch as Boyden designed his first water-wheels five years or more before he could have read what Weisbach had to say on the subject, it would seem that in his case, also, experimentation and common-sense design preceded theoretical analyses. More- over, it is possible that his subsequent scientific analyses, especially of the flow of water through the turbine, found inspiration, if not basis, in WeisbacVs detailed treatment of the “theory of reaction turbines.”* * * § Certainly, after reading Weisbach,f Boyden would have had no excuse for not appreciating the fact that in the language of the authors (page 1254), “some energy re- mained in the water at the time of its discharge from the runner.” Jay M. Whitham,:}: Esq. (by letter). — This paper is a noteworthy and interesting contribution on the subject, and preserves, in compact form, much of the early history of turbine development. The paper shows, and deservedly so, great admiration for the notable work performed by the late James B. Francis. Although the Howd wheel is mentioned, the credit for its develop- ment is given to Mr. Francis. This, the writer believes, is not fully war- ranted. It is to be regretted that so little is known of the achievements of Howd. To the writer, the modern mixed-flow turbine is Ilowd’s “grown-up child”, and it possesses his characteristics. Gardner S. Williams,§ M. Am. Soc. C. E. (by letter). — The authors have presented a very interesting history of the development of the American turbine, a subject which previously has not been adequately treated. Several years ago, the writer made investigations along the same line as those of the authors, and recalls a few points that are not included in the paper. As to the so-called wicket gate, it appears that a patent was issued for this device to John Temple, February 8th, 1859, and by him assigned to Stout, Mills and Temple, the predecessors of the Dayton Globe Iron Works, who manufac- tured it with the American wheel at Middletown, Ohio, until 1863, when the works were removed to Dayton. It has always been a characteristic of the “American Turbine.” How it happened that Leffel was permitted to use it, had been a mystery to the writer until he saw the authors’ statement that the gate had also been patented by Elijah Koberts five years earlier. The Leffel Wheel . — The development of the Lefiel wheel is particularly interesting in view of the fact that although originally the two parts were apparently expected to deliver approximately equal quantities of water, the capacity of the upper one has been gradually reduced until in the Samson * “Mechanics of Machinery and Engineering,” Vol. II, p. 254ff. t Loc. cit., p. 256. t Philadelphia, Pa. § Cons. Engr., Ann Arbor. Mich. 1320 DISCUSSION ON THE AMEKICAN MIXED-FLOW TURBINE wheel, as built in 1905, it embraced less than 10% of the inlet area. At one time, the writer was interested in gaining an idea of the relative effect of the two parts. The upper inlets were closed with blocks, and the wheel was run in that condition until a fairly complete and accurate record of power output had been obtained. The blocks were then removed, in which condition the wheel has been operating for several years. It appears from the record of the plant that, reduced to a uniform head of 17 ft., closing the inlets to the upper section of the wheel reduced the power about 12i% at full gate, about 9% at seven-eighths gate and about 5^% at three-fourths gate. The area of the inlets to the upper section was approximately one-eleventh of the whole inlet area. As no modification of the gates or guides was made, it is apparent that the lower section of the runner would give a greater power if the gates were designed for supplying that part alone, and the increase would be relatively greatest at full gate. Therefore, it seems quite reason- able to state that the upper section at full gate probably contributes its full share of the output, but that its relative influence on the power rapidly decreases with decreasing gate opening. TABLE 9. — Performances of Holyoke Turbines in Various Settings. Wheels. Draft chest. Center bearing. Distance apart clear. Horse-power at 16-ft. head, max- imum. Percentage of efficiency.* 2 Original. Plain Full 5 ft. 2 in. 268.50 76.02 2 Chipped. Plain Small 5 ft. 2 in. 272.50 77.04 2 W ith vertical diagonal plate None 5 ft. 2 in. 265.75 75.93 2 ” With 2-11-in. interior cones. Small 5 ft. 614 in. 276.40 77.60 2 “ 2-15^-in. “ ‘‘ . “ 5 ft. 614 in. 271. 00(?) 76.99(?) 2 “ 1-15-iD. exterior “ . 6 ft. 5 in. 278. 00(?) 78.16 2 “ 1-15-in. “ . 6 ft. 5 in. 284.00 76.04 2 “ 2-10-in. exterior and .5-in. interior cones Full 6 ft. 10 in. 276.30 78.20 2 With 2-10-in. exterior and 5-in. interior cones Small 6 ft. 10 in. 286.75 79.43 2 With 2-15-in. exterior cones. 7 ft. Sin. 289.75 77.29t 2 “ 2-15-in. “ “ . 7 ft. 8 in. 284.50 81.03 2 New “ 2-15-in. “ “ . 7 ft. 8 in. 289.60 81.38 2 “ 2-15-in. “ “ . 7 ft. Sin. 289.20 81.60 2 “ “ 2-18-in. “ “ . New 8 ft. 2 in. 288.00 81.50 1 Original. L. H. vertical cylinder gates 139.95 1 R. H. vertical cylinder ga.fes 142.30 Combined ... . . . 282.25 1 Np.w L. H. vertical wicket gates. 147.10 80.69 1 R. H. “ “ “ . 149.10 81.14 Combined 296.20 80.92 * The efficiencies given are based on the corrected discharge of the weir and are between 2 and 3% lower than those reported by the Holyoke testing flume, t Wheels over-gated. Michigan Lake Superior Power Company Tests . — Erom 1899 to 1901, the writer was engaged on an extended series of tests at the Holyoke testing flume for the Michigan Lake Superior Power Company. This was the third series of tests of pairs of horizontal wheels made at Holyoke. The wheels were designed in one of the early attempts to increase the speed of the then recog- nized type of wheel. As the solution of the problem was then stated, it was getting the power of a 36-in. and the speed of a 30-in. wheel, which was DISCUSSION ON THE AMERICAN MIXED-ELOW TURBINE 1321 accomplished with a 33-in. McCormick having a specific speed of 68, whereas that of the usual type was about 50. The wheels were to be used in .units of four and were tested in pairs. The requirements were 285 h. p. at 180 rev. per min,, with a head of 16 ft. and 80% efficiency for each pair of runners. The wheels were equipped with wicket gates. The first tests showed the wheels to be seriously handicapped by what the authors have designated as a “boiler-maker” setting. In justice to the wheel designer, the late James F. Jolly of Holyoke, it should be stated that the draft chest was designed without his connivance, and he realized at once what measures should be taken to improve conditions. Numerous expedients were tried, however, and an enumeration of them with the results obtained is presented in Table 9. The first improvement was to enlarge the outlets of the wheels by chipping ofi the edges of the buckets, thereby increasing the dis- charge and power. Wheels so treated are indicated as “Chipped” in Tables 9 and 10. Later, a new wheel was designed of larger capacity, which is desig- nated as “Hew.” The so-called full center bearing originally designed, consisted of a cast- iron girder, 12^ in. deep at the center and 6 in. at the ends, extending across the draft chest, the lower half of the bearing extending 10 in. below it with a diameter of 16| in. and a length of 15f in. The draft chest was 54 in. in diameter and only 7.0 ft. long with the runners projecting into it about 10 in. at each end, as in the “Boiler Maker Setting 1890” of Fig. 29. The small center bearing consisted of a ring about 2 in. wide and 2 in. deep sus- pended by three |-in. rods attached to the walls of the chest. Although made of cast iron, the original draft chest was a “boiler-maker’s setting” with a vengeance, and by the time the tests were concluded it was impossible to find any one who would admit having been responsible for its design. The units were finally equipped with a new chest having conical extensions 18 in. long at each end, and a new center bearing was designed, that occupied about one-fourth the space of the original one. By separating the wheels an addi- tional 6 in., it was possible to get substantially the same results with the new bearing as had been obtained with the small one, with 15-in. extensions (see ! Table 9), and to more than meet the specifications. Temperature Effects on Turbine Efficiency . — In the tests of these wheels the effect of cold weather on apparent efficiencies was clearly established, as is shown by Table 10. In the warm weather tests, the performances of similar pairs of wheels i. agree within less than 0.5%, both in horse-power and in efficiency, and in the cold weather tests these variations are about twice as great, a difference easily accounted for by the presence of variable quantities of floating ice. Wheels Hos. 86 and 87 were those tested separately on a vertical setting, with results indicated in Table 9. A Modern Horizontal Setting . — In 1916, the writer undertook to re-design the hydraulic plant of a large paper mill in Minnesota. The original in- stallation consisted of three pairs of wheels per unit in “boiler-makers” set- tings, which gave, for a 1-ft. head, 6.3 h. p. at the switchboaid, with an efficiency of 48%, equivalent to about 53% for the wheels. 1322 DISCUSSION ON THE AMEEICAN MIXED-FLOW TUKBINE TABLE 10. — Performances of Pairs of Holyoke Turbines, on Horizontal Shaft, Full Gate, 180 Kevolutions, 16-Foot Head. Warm Weather Tests Chipped wheels. New pattern wheels. ^nrn 2 and 3 10 and 27 62 and 63 108 and 128 Caianfpd bv Makers Writer Makers Writer Aug. 1900 Aug. 1901 Aug. 1900 Aug. 1901 Aug. 1901 T^£kmT%AT5it.nrP of wa,t,©r 76° Fahr. 74° Fahr. 76° Fahr. 65° Fahr. 74° Fahr. ‘nra'fr No. 1* No. 1* No 1* No 1* No. 1 1 Small New Small Small New 289.60 288.00 292.90 291.00 291.00 J5I Uvi oC* 1-^'-' ” •••••••• in fPPt 196.50 195.10 198.70 198.40 198.50 81.38 81.51 81.40 81.00 80.95 Cold Weather Tests New pattern wheels. New pattern wheels. 62 and 63 Makers Jan. 1901 Jan. 1901 33° Fahr. 33° Fahr. No. 1* No. 1* Small Small 288.40 285.50 199.40 200.08 79.67 78.30 86 aRd 87 Writer Dec. 1900 37° Fahr. No. 2 * No. 2 * Small New 285.50 283.60 199.40 199.30 78.85 78.46 lNU.IllU“Lo Ul * With 15-in. extensions, t “ 18-in. These units were replaced by two pairs, as shown on Fig. 42, having ; the same speed (200 rev. per min.), in settings designed jointly by the ; writer and the wheel makers, and gave, on test for 1-ft. head, 9.85 h. p. ( at the switchboard, with an efficiency of 74.6%, equivalent to about 80% ; for the wheels. In these tests, it was found that by introducing a plate . in the draft chest, extending across it and vertically down, the power was decreased to 9.83 h. p. for 1-ft. head and the efficiency at the switchboard was increased to 75.1%, equivalent to about 81% for the wheels. The in- crease in power of the new setting appears to have been about 56% and of wheel efficiency about 53 per cent. As in these tests the water was measured by current meters, the writer would suggest a possible error of 10% in the test on the original wheels and possibly 5% on those of the new installation. Scroll Settings— In February, 1905, the writer designed for the Edison Sault Electric Company of Sault Ste. Marie, Mich., an hydro-electric plant* in which the first low-head, direct-connected, vertical units were installed. The design embraced concrete, open-flume scroll wheel-pits and concrete draft- tubes, with roller bearings supporting the weight of the turbine and the gen- erator, and with conical covers Over the turbines. Each unit consisted of a 71-in. vertical Samson turbine making 100 rev. per min., directly connected to a 450-kw. alternator of the umbrella type, designed by the General Electric Company for this plant, on specifications drawn up by Alex Dow, M. Am. Soc. * Engineering Record, November 2d, 1907, p. 483. DISCUSSION ON THE AMERICAN MIXED-FLOW TURBINE 1323 C. E. This was the first hydro-electric plant, as far as the writer knows, wherein the electrical equipment was made to conform to the economical hydraulic requirements. In previous plants, the hydraulic machinery was subordinated to the electrical, and the inefficient horizontal installation of multi pie- wheel units resulted. Requirement of Constant Speed . — The requirements as to the turbines were complicated by the fact that although the ultimate head to be utilized was expected to be 16 ft., only 12 ft. of this head could be obtained at the start, and in consequence the wheel makers were asked to state what power could be expected from a turbine designed to give 600 h. p. with a 16-ft. head at 100 rev. per min., when it was run at the same speed under a 12-ft. head. It is an interesting criterion of the condition of the turbine industry in the United States at that time, that only one of those to whom the inquiry was sent, would even guess what power could be expected from the wheel at the lower head, and that one gave 385 h. p. at full gate, but a day later he reduced it to 285 h. p. The correct estimate was about 360 h. p. In July, 1905, the builder, referring to the matter of perfomance under the 12-ft. head at 100 rev. per min., proposed to test at a 12-ft. head and 100 rev. per min. a 56-in. wheel designed for a 16-ft. head and 100 rev. per min. He wrote: “No other wheel has any advantage over us in this matter. We are all in the same condition. We have had no occasion heretofore to make any such test.” Inasmuch as the writer had covered this point quite thoroughly during the tests previously discussed and had prepared a table of “Performances of McCormick Turbines at Over and Under Speeds”,^ he took the responsibility on himself for the performance at the 12-ft. head and authorized the construct- tion of the wheels. When tested in place, measuring the head from still water ^ American Civil Engineers Handbook (4th Edition), p, 1139, 1324 DISCUSSION ON THE AMERICAN MIXED-FLOW TURBINE in the forebay to water in the tail-race, and charging against the unit the water used by the exciter, which was independent and water-wheel driven, an efficiency of 68i% was obtained for the power on the switchboard, and cor- recting for the excitation, the efficiency of the main unit was 71%, which means about 78% wheel efficiency at a 16-ft. head, charging all flume and rack losses against the wheel. In commenting editorially on other plants. Engineering Record"^ stated: ‘^In none of these plants, fortunately, was there need of going to ver- tical shaft construction of which several examples have recently been seen. The vertical shaft does indeed enable one to keep the generating room above high water mark, but is liable to cause extra cost for attendance on account of step bearings.” Nevertheless, ten years later, the standard practice for hydro-electric plant design required the vertical unit. Draft-Tubes . — In the plant previously discussed, the concrete expanding draft-tube with a right-angle turn, was introduced, and one of the bidders for the turbines expressed himself as willing to guarantee 1% higher efficiency thereon than for an ordinary setting. The draft-tube was modifled in a plant built on the Huron Kiver, near Ypsilanti, and in one on the St. Joseph Kiver for the City of Sturgis, both in Michigan. However, no comparative tests were made, as far as the writer knows, until 1914, when the Argo Plant of the Eastern Michigan Edison Company, at Ann Arbor, Mich., was tested. This plant which had been constructed after the writer’s plans and under his super- vision, embraced two similar units. The contract for the machinery included a bonus and forfeit clause, and when the writer’s design for the draft-tube was submitted, the maker objected to it, declining to guarantee the performance of his machinery on it, and submitted a design of his own which was the recog- nized standard form for such constructions then and now generally adopted. As in other respects the two settings were practically identical, it was decided to equip one unit with the maker’s draft-tube. No. 2, Eig. 43, and one with the writer’s design, No. 1, Eig. 43. The plant was equipped with a weir for * Engineering Record, November 2d, 1907. DISCUSSION ON THE AMERICAN MIXED-FLOW TURBINE 1325 measuring the water and was tested by Charles M. Allen, M. Am. Soc. C. E., of Worcester, Mass., using an Alden dynamometer. The results ob- tained are shown in Table 11. TABLE 11. Maximum efficiency. Maximum Horse-powkr. percentage. 14-ft. head. 1-ft. head. Maker's draft-tube, No. 2 83^ 88 578 11.2 Writer’s “ “ No. 1 618 11.8 The maker^s guaranties were just met on their draft-tube, but he would have won a respectable bonus had he not objected to the writer’s design. The maker was not satisfied with the results of the test, and it was repeated when the generators were installed, with similar results, and draft-gauges. FTg. 44. 1326 DISCUSSION' ON THE AMEEICAN MIXED-FLOW TUKBINE attached just below each wheel, showed a considerably higher vacuum for the writer’s tube. This removed the last doubt as to the reliability of the result. Meanwhile, a series of tests had been made by the Allis-Chalmers Manufac- turing Company of Milwaukee, Wis., on models of draft-tubes, including the Argo draft-tubes, from which it appears that the so called “hydraucone” gave efficiencies between 1 and 2% higher than the writer’s Argo draft-tube. The opportunity for further tests was afforded by the construction of the Geddes Plant of the Detroit Edison Company, in which another pair of duplicate units were to be installed. A “hydraucdne” was constructed under one unit and a stream-line draft-tube of the writer’s design under the other. (Fig. 44.) The wheels were four-bladed ISTagler (propeller) runners built by the Allis-Chalmers Manufacturing Company and were identical, as were also all other conditions except the draft-tubes. A weir was constructed across the tail-race for measuring the water. Tests were also made on the “hydrau- cone” with a three-bladed Magler runner. The results obtained at 200 rev. per min. are shown in Table 12. TABj^E 3^2. \ Head, in feet. Brake- horse- power. Percentage efficiency. Corresponding power at 16-ft. head. Four-blade runner: On draft -tube 14.11 507.5 83.7 614 11.87 390 81.2 612 On hydraucone 14.59 514.5 81.5 591 11.77 375.5 78.8 595 Three-blade runner: On hydraucone 13.26 524.5 80.85 694 The results of these tests on draft-tubes show clearly that the theory on which their design has heretofore been based is largely in error. Combining the gain of 4i% shown in the Argo test for the writer’s draft- tube over the standard type, and the gain of 2% for the stream-line draft- tube at the Geddes Plant over the hydraucone, and the gain of 1% obtained for the hydraucone by its makers over the Argo draft-tube, it appears that the improvements in draft-tubes due to the writer’s designs have amounted to at least 7J% in output over the standard design. An increase of 7i% in output means a like increase of income and has been quite a satisfactory return to the owners for the small expense involved in the constructions and tests. If engineers more frequently took advantage of opportunities for comparative tests, technical knowledge would be much more rapidly advanced. Under-Sluices . — In several of the plants designed by the writer, provision for disposing of low-water flow during construction and for drawing the pond afterward has been made by means of under-sluices connecting with the pond at the inlet of the power house and discharging at the base of the draft-tube. (Fig. 44.) By proportioning the outlet of the under-sluice so as to cause it to discharge the water as a flattened jet along the bottom of the dr aft- tube, an action somewhat similar to that of an ejector takes place and the flow through DISCUSSION ON THE AMERICAN MIXED-FLOW TURBINE 1327 the turbine, and consequently the power, is increased. The device then becomes serviceable to increase the output of the plant at times of high water and, in a measure, to overcome the rise of tail-water. Tests at the Barton Plant of the Detroit Edison Company, have shown an increase of power at the switch- board of as much as 9% of that delivered with the sluice closed at the same head. Over-Fall as Head Increaser . — Another method of compensating for the loss of head due to high water is that indicated in the Edison Sault Electric Company Plant, wherein the stop-logs above the weir* may be raised by the crane and the surplus water be allowed to fall into the tail-water at the draft- tube outlet and by the resulting depression at the foot of the sheet cause an increase of effective head. A somewhat similar plan was used for a plant designed to be built on the St. Joseph River in Michigan before the World War. The flood water was to be taken through the power house in channels discharging at the sides of each draft-tube in a nearly horizontal direction, thus obtaining a similarly produced increase of head of greater magnitude than in the more nearly vertical drop at the “Soo”. Accuracy of Holyoke Tests . — The authors quote the statement of the late Robert H. Thurston, M. Am. Soc. C. E., as to the accuracy of the Holyoke tests. With the information Professor Thurston had when he made the state- ment in 1887, it was justifiable, but such is no longer the case. Investigations made by the writer at the Cornell Hydraulic Laboratory in 1899 brought to light the variations in apparent head that would be indicated by different devices for transmitting the level of the water flowing over the weir to the hook-gauge pail or the gauge-tube. A comparison of the heads indicated by the device in use at Holyoke and that used by Mr. Francis in determining his weir formula, to communicate the head of water to' the measuring instrument, has been already presented by the writer.f The original method used at Holyoke for communicating the head in the weir bay to the hook-gauge pail was through a pipe projecting a few inches beyond the wall of the channel, at right angles to the current, the end being cut off square and fully open to the cur- rent. The experiments of the late Hiram F. Mills, Hon. M. Am. Soc. C. E.:}: having already shown the erroneous indication of such a device, it was used only temporarily, and was replaced as soon as possible by the transverse per- forated pipe now, and since, in use. If the device used by Mr. Francis gives correct results to be used with his formula and those derived from his experiments, then the Holyoke device does not. For heads on the weir in excess of about 0.8 ft., the error introduced leads to an under-measurement of the water. This causes the reported efficiency to be higher than the true efficiency by amounts varying from zero at 0.8-ft. head to 2.23% at a head of 2.10 ft. on the full length weir, corresponding to variations of discharge between 50 and 205 cu. ft. per sec., as reported in the Holyoke tests. For discharges less than 50 cu. ft. per sec., the Holyoke indica- * Engineering Reced’d, November 2(1, 1907, p. 484. ^Transactions, Am. Soc. C. E., Vol. LXXVII (1914), p. 1311. t Proceedings, Am. Academy of Arts and Sciences, Vol. VI, New Series, Dost., 1879. 1328 DISCUSSION ON THE AMERICAN MIXED-FLOW TURBINE tion of efficiency is slightly low, the error amounting to 0.8% at a discharge of 25 cu. ft. per sec. over the full length weir. The first intimation received by the authorities of the Holyoke flume that the device was in error, was in March, 1900, when the writer made the experimental comparison referred to, and, after a thorough investigation, it was decided that the device should not be changed nor the reports corrected, on account of the large number of tests that had been reported embodying the error. In this decision, the writer concurred at the time, and he still concurs. The majority of tests made at Holyoke are primarily for determining the relative rather than the absolute merits of the devices tested, and, in the few cases where absolute results are desired, the quantity of water used is generally such that the error is not more than 1 per cent. It is, however, in the writer’s opinion, desirable that the facts be known, and in the tests herein reported by him the efficiencies have been corrected to those indicated by the discharge measured according to the Francis method. From his own observations cover- ing many weeks at the flume, the writer feels the utmost confidence in the Holyoke testing flume reports, and in calling attention to the facts, offers no criticism of either its designer or the operators, for it was not until many years after it was put into service that any one appears to have suggested the ■ possibility of a transverse perforated pipe in a flowing stream communicating anything but the true head of water above it. i Floyd A. Hagler,* Assoc. M. Am. Soc. C. E. (by letter). — In assembling ■ the fragmentary records and co-ordinating the steps in the development of the American type of mixed-flow turbine, the authors have made a great contribution to water turbine history. The writer has always been of the opinion that it is erroneous to designate the American mixed-flow turbine ; as being of the ^Trancis” type. It might better be called the “Howd” type or <: equally as well, the Swain or McCormick type, as the last two men are responsible for the most distinctive advancements in the present charac- ; teristics of the wheel. As the modern wheel is the product of so many men, all of them Americans, it would be well to drop all personalities in referring to this wheel and call it “the American mixed-flow turbine”, as the authors have done. The paper emphasizes the “scientific designs” of the late James B. Francis, Past-President, Am. Soc. C. E., and gives lesser weight to the advancements made by Messrs. A. M. Swain and John B. McCormick. It is true that Francis seemed to appreciate certain principles of relative velocity and entrance without shock, but he did not correctly apprehend the “centrifugal effect” of water passing through rotating channelsf and other principles which render some of his methods of design rather unscientific as compared with those used by the best modern designers. The “cut and try” methods used by Swain and McCormick are not to be regarded as entirely “unscientific”, when each new “cut” or “try” is based on previous experience. Much scientific experimentation proceeds along just such lines. Even Francis was guided m * Asst. Prof. Mechanics and Hydraulics, State Univ. of Iowa, Iowa City, Iowa, t “Lowell Hydraulic Experiments”, p. 42. Fig. 46. — Four-Blade Runner of the Austin Type. f >iTr;':u7> HlilT stiC HtiA .^li - .: ^v>'l —.f ^ .01" Fig. 47. — Whitelaw Runner of Barker’s Mill Type. Fig. 48. — The Centrifugal Wheel. DISCUSSION ON THE AMERICAN MIXED-FLOW TURBINE 1333 his designs by “but little theoretical information’’ and “principally by a comparison of the most successful designs”* * * § and developed his rides for proportioning turbines “by the aid of the experiments upon the Tremont turbine”. t Too much credit cannot be given McCormick for his contribu- tion in the development of the American mixed-flow turbine, even though his results were accomplished with a great deal of labor and cost. With regard to McCormick’s efforts, E,. E. Horton, M. Am. Soc. C. E., states “Probably the greatest achievement of any one man in advancing the development of the hydraulic turbine was that of John B. McCormick of Indiana County, Pa. About 1870, he found that by extending the bucket vanes of an inward flow turbine downward and outward, making them ladle or spoon shaped, he was able to greatly increase the outlet openings of a turbine of a given diameter * * *.” “The form of buckets or water passages through these latest and largest turbine runners differs but little from that developed by John B. McCormick, following earlier attempts at the production of an efficient inward flow turbine runner by Howd, Francis, and Swain * * *.” It would not have been inappropriate for a history of the development of the American type of mixed-flow turbine written in 1922, to have mentioned the name of the most successful designer, Mr. S. J. Zowski (Zwierzchowski), particularly since the authors mention his wheel as having the highest “specific speed” of any of its kind, namely, 102 (pages 1261 and 1269) and illustrate it in Figs. 20 and 21. The fact that Mr. Zowski is of foreign birth and that he applied his theories, which were learned in European schools, to the perfection of the American type of turbine is not sufficient reason for the failure to mention his name in the paper. Mr. Zowski is an American citizen and has been a thorough student of the American type of water turbine§ as he found it when he came to the United States, but he went further. The authors closely follow the development of the turbine until maximum values for specific speeds of 68 and 69 were attained in 1899 and 1897 by the Smith-McCormick and Samson runners, respectively (Table 1) ; then it is stated that a present-day model has a “specific speed” of 102. Certainly, an increase in the specific speed of a runner from 69 to 102 indicates that some radical changes must have taken place in its structure, and represents a greater betterment of the turbine speed and capacity characteristics than those obtained by Francis, Swain, or McCormick. There can be little doubt but that Mr. Zowski was the leader in this latter achievement, and his wheels to-day represent the runners of highest efficiency and “type characteristic” manufactured by three of the largest producers of the “American” type of water turbine. The Zowski models numbered, I, II, III, lY, Y, and YI, had “specific speeds” of 87.4, 92.8, 78.0, 90.0, 91.0, and 102.0, respectively, with * “Lowell Hydraulic Experiments”, p. 7. t Loo. cit., p. 44. t The Encyclopedia Americana, “Water Power”, p. 29, 1920 Edition, and Engineerina News-Record, October 7th, 1920, p. 685. § “The American High Speed Runners for Water Turbines’', S. J. Zowski, Michigan Technic, June, 1908 ; “A Rational Method of Determining the Principal Dimensions of Water-Turbine Runners”, S. J. Zowski, Michigan Technic, 1909 ; “Some Recent Tests of High- Power, High-Speed Water Turbines”, Engineering Record, November 28th, and December 26th, 1914. 1334 DISCUSSION ON THE AMEKICAN MIXED-FLOW TURBINE maximum efficiencies of 90.0, 87.2, 83.2, 89.2, 89.3, 90.1, and 90.7, respectively. In addition, his runners showed a flexibility hitherto unrealized.* On page 1262, it is stated that the makers of the ‘‘Wynkoop” wheel claimed an efficiency of 175% for their runner. Mr. Emerson statesf that an efficiency of only 135% was claimed, based on figures quoted from the maker’s table. Exaggerated as this last figure may be, it is a more accurate statement of the maker’s contentions. Several years ago, the writer had occasion to study the history and char- acteristics of the Austin wheel, mentioned on page 1265. The first wheel produced by Mr. Austin, shown in Eig. 45, was only 14.5 in. in diameter, and was patented by him in 1878. This runner, together with another 4.0 ft. in diameter, are the property of Mr. C. E. Kinne, of Watertown, M. Y. It was reported to the writer by Mr. Kinne that, on test, this 4.0-ft. Austin runner, “under 9-ft. head, made 150 rev. per min. under no load, giving a rim speed 27% faster than the spouting velocity of the water driving it; when loaded, it ran at 125 rev. per min., with a rim speed 6% faster than the spout- ing velocity of the water”. Based on these figures and on measurements of the openings through the runner, and using the most favorable coefficients in computation, the writer has never been able to ascribe a specific speed greater than 60 to this runner. This type of wheel does not have a large capacity for water, even though its speed is very high. In the attic of the Leete Iron Foundry, at Potsdam, K. Y., the writer photographed patterns of what appeared to be a four-blade wheel of this same type shown in Fig. 46. A runner made from this pattern, which was 4.0 ft. in diameter, should have more than twice the capacity of the 4.0-ft. Austin wheel, but no doubt would have less speed. If the Truax Green Mountain wheel performed in a similar manner, the specific speed of 125 ascribed to it in Table 1, based on claims made by the makers, is far too high. A Whitelaw runner of the Barker’s Mill type, similar to that shown in Eig. 4, is given in Fig. 47, and was photographed by the writer in the collec- tion of old water-wheels of the Bagley and Sewall Company, at Watertown, K. Y. Eig. 48 shows a runner which is an excellent example of the “cut and try” period. This wheel, commonly called the “Centrifugal Wheel” was first man- ufactured in the C. W. Leete Iron Foundry, at Potsdam, N. Y., in 1870. The wheel was photographed in the scrap heap of the Leete Foundry in 1919. Harvey Linton,:}: M. Am. Soc. C. E. (by letter). — Believing it to be suf- ficiently worthy of notice to be given a place in the history of water-wheel design and practice in the United States, so ably presented by the authors, the writer submits, from notes made at the time, the performance of a water- wheel of the Seventies. The water from the “Spring”, 788 cu. ft. per min., flowed quietly away. A few inches, under the surface was a 31-in. wooden turbine, making * Engineering Record, November 28th and December 26th, 1914. t “Emerson’s Hydraulics, Dynamics, etc”, 6th Edition, p, 193. t Philadelphia, Pa, DISCUSSION ON THE AMERICAN MIXED-FLOW TURBINE 1335 157 rev. per min., that furnished the power to drive a 4-ft. mill-stone and the necessary elevators, bolting machinery, etc. Except for the grist-mill and penstock, this noiseless little water-wheel, which had displaced an iron turbine, might easily have escaped notice. It worked under a head of 7 ft., and was estimated to develop h. p. That was in 1877. In 1880, the owner^s testimony was as follows : ^Tt continues to give the best satisfaction, even seems to work better than it did at first, * * * and to be as sound and good as when new. It does more work than was expected of it. The 36-in. Franklin turbine in the same penstock, will not compare with it in power and economy of water.” The miller stated that: ‘Tt grinds more than 5 bushels of wheat per hour, running one smutter, three bolts, elevators, one middlings purifier, and other machinery. It has fully 50% more power than the Franklin wheel and gives no trouble by clogging with drift, which is a serious objection to the iron wheel.” The appearance of that “Spring” indicated that all the force of the 7-ft. head of water was expended in useful work. This result was effected without the use of the numerous small openings and guides usually seen in a turbine wheel-case. At another grist-mill, a few miles distant from the pioneer installation, an 11-in. cast-iron turbine, constructed on the same general plan as the wheel previously described, displaced a 174-ft. over-shot wheel. This little turbine was rated at 5 h. p., under a 20-ft, head of water. In 1880 the owner said of it: “It does one-third more work in the same time than the I7i-ft. over-shot wheel which it displaced, using about the same quantity of water. * * * My old over-shot was frequently frozen up a great part of every winter.” The testimony of all users of this form of turbine was of similar character. The inventor was Eobert Wilson, of McLeansville, N. C. (Patent No. 168 202 dated September 28th, 1875.) He found such demand for his turbine (named by him, “The Inclined Plane Water-Wheel”) that, with the assistance of a number of millwrights, he abandoned his milling business and gave all his time to making and installing his wooden turbines. His methods do not appear to have been of the “cut-and-try” character, but were based on something more like close observation and a regard for the principles of hydraulics, that always brought good results. The apparently better class of turbines (constructed of cast iron), in use at that time, were more extensively advertised than the Wilson wheel. They were often failures; as much, perhaps, because they were not suited to the work put on them, nor to the sites for which they were chosen, as on account of their being inferior to this wooden wheel. The Wilson turbine was always built for the particular site and the work required of it. Its working speed, in revolutions per minute, was found to be the theoretical velocity of water, in feet per minute, discharged under the actual working head, divided by the i336 DISCUSSION ON THE AMERICAN MIXED-FLOW TURBINE circumference of the wheel, in feet. This working speed was said to be higher than that of any other turbine of its time. Figs. 49 to 53, inclusive, show the construction of a 28-in., upward dis- charge, wooden turbine and “chest”, as prescribed by Robert Wilson. Fig. 49 is the plan of the wheel; Fig. 50 shows the interior of the wheel with half its staves removed, the “head block”, and the depth of the “chest”; Fig. 51 shows the plane projection of the outer ends of the “inclines,” having a radius of 14| in.; Fig. 52 is a horizontal section of the chest and penstock; and Fig. 53 is a side view of the wheel, chest, and penstock. The level of the tail-water is a little Fibove the level of the top of the wheel when at work. Details, such as cast-iron flanges for the wheel shaft, bearings, etc., are not shown. Robert Wilson’s rules for constructing his turbine were, briefly, as follows: Knowing the flow of the stream, in cubic feet per minute, and the head of water over the turbine at work, he assumed that = - — , in which A is the area, in square inches, of the sum of the areas of the openings in his turbine that will discharge a volume of water yielding 1 actual h. p., assuming its efficiency to be 80% ; Q is the volume of water, in cubic feet per minute, discharged by the turbine to develop 1 h. p.; and q is the theoretical dis- charge, in cubic feet per minute, of water under the given head, through an aperture 1 in. square. In this 28-in. wheel, there are four oriflces, each 6 by 7 in., totaling 168 sq. in.; the head is 84 ft.; and A is found to be 17.9026 sq. in.; then. 168 17.9026 9.384 h. p. The stream flow being 728.634 cu. ft. per min.. 728.634 stream flow 77.647’ Q 9.384 h. p. Q z= 77.647 cu. ft. per min.; q = 9.7587 cu. ft. per min. The working speed is : 1 405.249 7.3304 77.647 j^ = l-.9026 sq.m. 191.5 rev. per min. The “draft” for wooden turbines is 14, that is, 14 in. measured on the circumference of the hub, to a 1-in. drop of the “inclines”. For iron turbines, the draft recommended is 1 to 1. The inventor claimed 80% efficiency for his wheel, judging entirely by results: he had no faith in the Emerson tests. In one instance, a 52-in. downward-discharge wooden turbine, with four oriflces in the chest, displaced an iron turbine of the same size, taking its place on the same shaft, to drive a 52-in. circular saw. A draft-tube was used in another location. DISCUSSION ON THE AMEKICAN MIXED-FLOW TUKBINE 1337 1338 DISCUSSION ON THE AMERICAN MIXED-FLOW TURBINE These wooden wheels could easily he rebuilt if the millwright had the orig- inal plans, and they could be used to advantage by many owners of small water-powers to-day. They were always constructed of green timber, yellow pine being preferred. Reinforced concrete could be used to advantage in the construction of the wheel-chest. Arthur T. Safford,* M. Am. Soc. C. E., and Edward Pierce Hamilton, f Esq. (by letter). — A somewhat more extended, although still far from com- plete, study of the early American water-wheels disclosed several interesting examples. Mr. C. E. Kinne, of Watertown, IST. Y., who is much interested in water-wheel history, has been of great assistance to the writers. It was from him that the photographs of the early reaction wheel, the Scotch Mill, Austin, and Truax wheels were obtained. Isaac Sanderson, of Watertown, Mass., built a paper mill at the lower falls of the Heponset River, at Milton, Mass., in 1817, and installed a wrought- iron tub wheel.:}: This is the earliest record the writers have found of an iron turbine. It is especially interesting that this wheel should have been in- stalled at the place where water power was first developed in the United States. The lower Heponset Valley was one of the earliest industrial centers of the country, and a variety of manufactures were in operation there previous to 1700. The 1840 reaction wheel (Fig. 54) was set at the bottom of an open flume without any guides, and discharged outward. Wheels of this model were common in Oneida County, Hew York, during the middle of the Nineteenth Century. The “Scotch Mill”, a modification of Barker^s Mill, was invented by Whitelaw and Starrett, of Paisley, Scotland, and patented in the United States in 1843. It is said that millers whose mills contained these wheels often had to suspend operations and devote an afternoon to chiseling a muskrat out of one of the tapering arms. The example shown (Fig. 55), was running in a grist- mill as late as May, 1919. The Austin wheel is mentioned on page 1265. Fig. 56 shows what is claimed to be the first Austin wheel made. Probably the first of the reaction wheels to come into use in New England was that of Calvin Wing, of Maine, which appears to have been invented in 1833. It was somewhat like a Boyden runner, consisting of top and bottom circular plates with vanes extending between them. The upper plate was pierced by a number of holes through which the water was fed to the interior of the wheel, whence it discharged outward. These holes were of less area than the bucket orifices and, consequently, the flow was restricted and the efficiency poor. Some of these wheels were installed at the Boston and Rox- bury Mill Dam about 1835. Mr. Forrest Nagler feels that too much credit cannot be given to Swain for his mixed-flow turbine. A search through the personal and technical papers of U. A. Boyden has revealed the following facts regarding the inven- tion of the mixed-flow turbine. In May, 1860, Swain was granted a patent on * Engr., Proprietors of Locks and Canals; Cons. Hydr. Engr., Lowell, Mass. t Milton, Mass. t Teele’s “History of Milton", p 372, Boston, 1887. Fig. 54. — Reaction Wheel, 1835. Fig. 55. — Scotch Mill, 1843, \ Fig. 56. — Austin Wheel. Fig. 57. — Boyden^s Mixed-Flow Runner, 1849. DISCUSSION" ON" THE AMEKICAN MIXED-FLOW TUEBINE 1343 a water-wheel covering three claims for the arrangement of the cylinder gate and the adjustable step-bearing. Nothing was claimed for the runner. In November, 1872, a substitute patent was issued with several additional claims, among which was one covering an inward-flow water-wheel with the discharge edge extending from the crown to the bottom edge of the band. This claim was broad enough to cover almost every mixed-flow wheel made. The Swain Turbine and Manufacturing Company tried to restrain James Ladd from in- fringing this re-issued patent. The bill was dismissed by the United States Circuit Court for the Massachusetts District, in January, 1877, and on appeal to the Supreme Court, this decision was upheld. The truth appears to be that Swain did not even attempt to patent the mixed-flow turbine at the time of the invention of his first wheel. He assigned the re-issued patent, which was extended to cover such a wheel, to the Swain Turbine and Manufacturing Company, and this Company attempted to secure a monopoly on all mixed- flow turbines. Justice Bradley delivered the opinion of the Supreme Court (102 U. S. 408). After covering the legal points of the case, he stated that the Stowe wheel, built in 1837, 1841, and 1850, although an impulse wheel, had buckets of essentially the same shape as Swain’s wheel. He also cited three wheels, each of which he said was the same as Swain’s and that if any of the three was of a date previous to 1860, it clearly showed that the idea of the mixed-flow prin- ciple was not new. These wheels were the Temple (1859), the Whitney (1864), and the Greenleaf (1854). Swain tried to extend his first patent by a re-issue to cover a principle which he may or may not have invented, but which he did not claim in his first patent, and this principle had come into common use since that time. It was held that a re-issued patent could not be extended to cover claims not in the first patent and that, although Swain may have been the inventor of the mixed- flow principle, he could not at this late period secure a monopoly of it. On June 13th, 1849, Uriah Boy den applied for a patent on an inward-flow turbine. Fig. 57 shows a photograph of one of his original models now in the possession of John B. Freeman, President, Am. Soc. C. E. Notice the hori- zontal dividing diaphragms which were to appear later as the fins on the “Her- cules”. They were designed to lessen the part-gate contraction of the cylinder gate. The following part of a letter tells the story in Boyden’s own words: “Boston, July 26, in 1876. “To the Judges of the Circuit Court of The United States for the District of Massachusetts : “I do not presume to know whether any general rule against noticing informal writing should prevent you from noticing this letter. I will deliver a duplicate of it to the complainant or its counsel, and I will also deliver a duplicate of it to the defendant or to his counsel. “Data indicate that the suit of the Swain Turbine and Manufacturing Com- pany against James E. Ladd in the Circuit Court of The United States for Massachusetts includes sham which you do not know. I tried to prevent any- thing from connecting me with said Company and their associates manufactur- ing water-wheels ; and I do not intend to include in this letter anything which causes do not urge me to mention. 1344 DISCUSSION ON THE AMERICAN MIXED-FLOW TURBINE “Hydraulics motors partly like Asa M. Swain’s turbines have been called Poncelet turbines, Howd turbines, inward flow turbines, center vent turbines, and converging turbines. I applied for a patent for improvements chiefly in said kind of motors; and some years after I applied, the United States issued a patent to me for them, and numbered it 10,027 and dated it September 20 in 1853 ; and the next two quotations are paragraphs from it. “ ‘In wheels of this kind, with which the water is directed obliquely to the wheels by the guides * * *, or by other things, the velocity of the water diminishes, or should, from the time it strikes the floats till it leaves them or till near the time it leaves them; and the change of the directions of its motion which should be produced during the same time is not sufficient to compensate the effects of the diminution of its velocity and the less space the water has in consequence of its approaching the axes of the wheels, so that the passages between the parts of the floats nearest the axes of the wheels are choked some. To produce the greatest possible effect of the water with this kind of motors, it is necessary that the resistance to the moving of the water through the wheels should be less than it is in them as usually made, so that the velocity of the water on entering these wheels should be greater than it usually is. There is no such disadvantage of partial choking in mere re-act- ing wheels, in which the velocity of the water increases as it passes through the wheels, and in which much contraction of the passages is needful. And this partial choking is also different from anything which happens in the common or Fourneyron turbines, in which the water diverges as it passes between the floats and out of the wheels at their peripheries instead of converging as in the Poncelet turbines. This disadvantage cannot be obviated by placing the inner parts of the floats farther apart, without their being directed so nearly toward the axis of the wheel as to cause a part of the force of the water to be lost. The fourth branch of my invention consists in making the distances between the rims of the wheels at the ends of the floats next the axes of the wheels greater than at their outer ends, * * *. This difference in dis- tances may be made by making so much of the upper part of the wheel as the upper edges of the floats touch plane or flat; and the upper surface of the lower rim next its periphery, including about one-eighth of the width of this rim, also plane; and from this place gradually curving down- ward, as by an arc of a circle; the curvature of this part being such that it will join the plane part without any sensible angle at the place of joining, and the distance of that part of the lower rim at the inner ends of the floats from the upper rim will be about 25% greater than at their peripheries. It is well to have the upper edges of the floats rather longer than their lower edges, so that the tops of their inner ends will be rather nearer the axes of the wheels than the bottoms of their inner ends ;***.’ ” The next following paragraph includes the claim of the improvement above described. “ ‘Fourthly, the shape of the spaces between the rims of water-wheels which the floats are fastened to, in which they flare toward the axes of the wheels; as above described; though I do not limit my claim to exactly the placing above described, but extend it to all placing which will essentially answer the same purpose. The flrst, third, and fourth branches of my claim apply only to such hydraulic motors as have guides or other things which cause the water to move obliquely toward the wheels in the way in which the wheels turn, and pass into the wheels at their circumferential parts, and after acting on the floats discharge from the floats inward. I do not extend these divisions of my claim to the class of tub wheels and undershot wheels in which the water generally flows into the wheels in streams with spaces between the streams, at which spaces the water does not flow into the wheels. Though I have described these water-wheels as being horizontal and the gates as being opened by raising, it is obvious that all these four branches of my claim are quite applicable to DISCUSSION ON THE AMERICAN MIXED-FLOW TURBINE 1345 wheels in other positions, and to cases in which the gate is opened by lower- ing, and I do not limit either branch of my claim to cases in which the wheels are horizontal or to cases in which the gates are opened by raising/ “Gay, Silver and Company made excellent turbines according to my design and under my supervision, and subsequently, in a. d. 1854, I made an elaborate design for a converging turbine with such flaring of the spaces between its rims as is claimed in said patent, for actuating their machinery at North Chelmsford in Massachusetts, and they made a turbine according to that design exclusively of my supervision, which they did not need, and actuated 4heir machinery. [This is the wheel shown in Fig. 24. It ran for more than fifty years before it was taken out about fifteen years ago.] I never removed that design from their shop. Harvey Silver of said firm, and other men, told me that said firm made spaces between its floats flare toward its axis and data indicate that he saw a copy of my said patent, but page 215 of the printed evi- dence shows that he testified or implied that the spaces between the floats of my converging turbines did not flare toward the axes of their wheels, so as by their flaring to produce any substantial effect on water’s motions in them; which is an error. “When such a converging turbine as I claimed works with its gate fully open, the water which flows through the upper parts of the spaces between its floats flows toward or nearly toward its axis when it leaves them, and it then moves horizontally or nearly so; and the water which flows through the lower parts of the spaces inclines downward when it leaves them, so that the inclina- tions downward at different distances below the tops of the floats are somewhat proportional to these distances. “Asa M. Swain’s patent dated May 15 in 1860, for alleged improvements in converging turbines, and the substitute for it dated November" 19 in 1872, represent a cylindrical piece around the lower parts of the floats, which has its diameter as large or nigh as large as the upper rim or part which joins the upper edges of the floats, and this piece is partly a substitute for the lower rim of a common converging turbine, so that it partly answers one of the purposes of such a rim. The upper outer part of each of his floats is nearly as de- scribed in my said patent, and from their upper parts his floats extend down- ward by double curvature to said cylindrical piece, so that the length of the discharging orifice between any two floats is greater than the height of the receiving orifice of the space between the same two floats; and it is greater than I showed in said patent, and his form involves wasting more of the power of water which actuates a turbine than the form which I showed by my said patent. The printed bill of complaint with the printed evidence represents that this suit is to check or stop alleged infringing a patent right to this one of Swain’s alleged improvements. “Waters implied that he did not know or believe that previous to Swain’s alleged inventing, any converging turbine was known in which the spaces between its floats had their discharging orifices longer than the height of their receiving orifices, which I had claimed in said patent; and much of his testi- mony in the printed evidence is on some advantage of such excess of length, and he testified or implied that because Swain’s turbines had such excess they were better than other turbines; yet the printed evidence shows that Waters and other witnesses neglected or blinked elements in deducing that inference and centrifugal force is one of those elements, and I next quote, for writing on some of its effects. * * This wheel of 1849 was a mixed-flow turbine. Let us now find the source from which Boyden derived this idea, for it seems that it did not originate with him. Amoug the papers of the late James B. Francis, there is a memorandum, a part of which is quoted herewith: 1346 DISCUSSION ON THE AMERICAN MIXED-FLOW TURBINE a* * * I jy-j. ^}ig street. I asked him if he did not think that the making the distance between the crowns of the wheel greater on the inside periphery than on the outside was new, also the mode of placing the gate between the guides and the wheel. He replied that he had application then pending for patents for those two things. I was much surprised — the more so as I had explained to him some months ago that I intended to make the wheels for the Boott Mill with a greater height between the crowns on the inside periphery than on the outside — and the impression I got at the time was that he thought it would not answer and he gave me not the slightest inti- mation that he claimed it as his invention. * * * “I do not know the date of Mr. Boyden’s application to the patent office, but I know that the improvements as applied to center vent wheels above men- tioned were made by myself and that I did not derive them or any hint of them as applied to this kind of wheel from Mr. Boy den or any one else.” Signed: James B. Francis." “Lowell, April 15, 1848. The model Lakin wheel shown in Fig. 25 was tested on February 16th, 1922. The power was measured on the jack-shaft by a small Prony brake, the water by a calibrated 3-in. meter. A maximum efficiency of 58.7% was obtained, despite the fact that the runner is only 4^ in. in diameter. It was also found that, at the lower speeds, the governor functioned smoothly and held the wheel to its speed. Fig. 58 shows a Curtis wheel, which is described on page 1282. A 32-in. Truax runner (Fig. 59), kindly lent by Mr. Kinne, was tested on July 6th, 7th and 8th, 1922, in the flume of the Proprietors of the Locks and Canals. The runner was first fitted with a wooden draft-tube 5 ft. long and a discharge diameter of 4 ft. A wooden casing, built according to the directions furnished by the maker for this size of wheel, was constructed around the runner for the first test. The runner was then tested with the casing, but without the draft- tube, and, finally, without either. Table 13 and Fig. 60 show the results of these tests. The results clearly prove the contention of the writers that a high-speed propeller wheel was on the market fifty years ago. Moreover, the characteristic curve of the Truax wheel closely resembles that of the modern propeller type (Fig. 61), and, although low in efficiency, has nearly the same speed. The results of this test are very different from those obtained by Mr. Forrest Nagler. The conditions, except for the addition of the draft- tube, were identical with those advised by the maker, and the use of the draft-tube on this particular wheel had been suggested. (See page 1286.) Mr. Nagler’s conditions were not fair to the wheel. The casing of the wheel is an essential part of the whole, particularly under a comparatively high head, such as Mr. Nagler used. Under very low heads, the omission of the casing is not so vital to the speed, as a vortex forms, which partly takes its place. Under higher heads, there would be no vortex, and the result would be the same as putting the wheel in a jet of water which had no whirling com- ponent. This is clearly shown by the very low speed obtained with the 10-in. model. Mr. Nagler’s model under 13-ft. head ran on the same principle as Barker’s Mill and not as a propeller wheel, to the successful operation of which Fig. 59. — Truax Runner. V DISCUSSION ON THE AMERICAN MIXED-FLOW TURBINE 1349 0 20 40 60 80 100 120 140 Revolutions per minute under 1 ft. head Fig. 61. 1350 DISCUSSION ON THE AMERICAN MIXED-FLOW TURBINE TABLE 13. — Tests on 32-Incii Truax Wheel. Date. Revolutions per minute. Head. Q. Horsepower. Efidciency. Ns WlT] a Casing a> :d Draft-Tube July 6, 1922*... 5.50 Runaway 198.5 5.71 33.35 12.30 57.0 78 ’.7 197.7 5.68 33.31 12.24 57.2 79.1 204.7 5.66 33 66 12.18 56.2 81.8 209 3 5.64 33.98 11.92 54.8 83.0 182.7 5.66 32.14 12.22 59.3 73.1 188.3 5.67 32.46 12.14 58.1 75.0 168.5 5 71 31.68 12.12 59.2 66.4 158.0 5.79 30.88 12.16 eo.o 61.4 147.7 5.85 30.26 12.10 60.3 56.4 136.7 5.89 30.00 11.88 59.3 5U4 228.7 5.73 36.34 11.90 50.4 89.0 2.55.8 5.61 36.82 10.75 45.9 97.1 278.3 5.52 38.34 8.91 37.1 98.0 294.5 5.51 39.68 6.48 26.2 88.9 148.3 5.78 30.34 12.16 61.2 57.6 With Gas iiNG, BUT Without Draft-’ Fube July 7, I922t... 132.3 4.88 28.66 7.94 50.1 51.4 164.3 4.88 30.82 8.22 48.1 65.1 202.7 4.87 34.14 8.11 43.0 80.0 239.5 4.89 35.90 7.18 36.1 88.4 No C ASING AND I 'lo Draft- Tube 203.5 4.85 39.87 8.14 37.1 80.8 242.5 4.91 41.00 7.28 31.9 89.6 348 4.95 runaway July 8, 1922t... 150 4.91 33.59 8.25 44.1 ^’.9 137.3 4.91 33.13 8.24 44.7 54.0 * Temperature of water = 66.5° Fahr. t Temperature of water = 67° Fahr. a whirling component in the water is essential, A modern Nagler runner would not have given a high-speed performance under the conditions imposed on the Truax by Mr. Nagler. The writers do not wish to detract from the very valuable development work done by Mr. Forrest Nagler in recent years on the propeller wheel. They feel, however, that the propeller type of wheel is not a new development, but has existed for many years in a state which, although crude and inefficient, had the same general characteristics as the present-day models. It must be remembered that the tests of fifty years ago, which showed efficiencies of 90%, are not absolutely substantiated. Emerson was a self -trained man with a great scorn for technical training of any kind other than “cut and try”. All his tests previous to the establishment of the Holyoke Flume are doubtful, and the results must be considered as comparative only. What Mr. Scheidenhelm says about these high efficiencies being reached by exceptional wheels is well taken. Emerson says of the Kisdon wheel, * * quite a num- ber of them have ranged along in the seventies in percentage, but, through some slight change after a first trial, every wheel tested (except two or three of the 20-in. size), has been made to return a useful effect of over eighty per cent, before delivery to purchaser, quite a number from eighty-five to ninety, and few even higher than ninety.”* * “Hydraulics, Dynamics, etc.’’, p. 211, James Emerson, Williamsett, 1894. DISCUSSION ON THE AMERICAN MIXED-FLOW TURBINE 1351 Mr. Groat has asked for more definite information regarding the efficiency test of the Gardner's Fall plant. One cannot do better than quote the descrip- tion of it which appears on page 45 of Bulletin 109 of the S. Morgan Smith Company. • Measurement of the effective head on the wheels were made by means of piezometers at different points. One of these at about the middle of each scroll was agreed upon to represent fairly the elevation of the water above the wheels. observed by a float gauge below the power house where the draft-tubes discharge into a short tail-race. The output in the res. agreement that the results indicated would be accepted by both parties above Greenfield Company in the canal above the power house and by the S. Morgan Smith Company in the tail-race “ measurements by the former in the tail-race. Ihe actual losses of head from the canal to the river at the end of the tail- Lheei'^®’’® " reasonable and less than expected, but indicated that the levds expected, practically the gross head between the two It, tho'^KrTl? ®ble condition was the strong wind blowing down the canal in the direction of the power house at the time of the best results at part gate. ‘‘A.. T. Safford^ Engineer"*. The head was taken as the difference between head and tail-water readings and no correction was made for the velocity head of the draft-tube discharge Fteley-Stearns current meters were used. Although the results of this test are high, the most searching criticism has failed to find reason for doubting them. Mr. Williams states that, due to his draft-tube designs, a gain of 7^% in efficiency has resulted over that obtained with wheels equipped with the ‘‘stand- ard type” of draft-tube. If by “standard type”, Mr. Williams means a long straight tube, or even a curved tube with a long straight section, the writers cannot agree. Can there be a better draft-tube than the long flaring four- diameter tube generally used at Holyoke ? Are not all the patent draft-tubes of to-day merely modifications designed to secure nearly as good results in a shorter over-all distance? The writers agree with Mr. Williams regard- ing the accuracy of the Holyoke Flume. An in- vestigation made in 1900 by one of them led to the conclusion that for wheels discharging more than 200 sec-ft., the error in favor of the wheel was somewhat more than 2 per cent. There is no question regarding the comparative accuracy of the testing flume, but the results are not absolute. It would seem that this fact should be more thoroughly appreciated at the present time when so much attention is being paid to tests in place. .u Nagler quotes Robert E. Horton, M. Am. Soe. C. E., who says that McCormick made probably the greatest single achievement when he extended the buckets downward and outward. Although McCormick was responsible for much valuable development work along these lines, he merely Obenehain “Little Giant” (Fig. 62), which pre- ceded the Hercules ' by about five years. Fig. 62. — Obenchain Turbine 1871. 1352 DISCUSSION ON THE AMERICAN MIXED-FLOW TURBINE Eegarding the ridiculous efficiency claims of some of the early water-wheel makers, Mr. Nagler states that the writers’ figure for the efficiency of the Wynkoop wheel should be 135% rather than the 175% given. The following is a quotation from one of L. D. Wynkoop’s circulars (undated), which is also reproduced in Emerson’s 1894 edition, on the page facing that referred to by Mr. Nagler: “Certificate. “From the experiments i)erformed with the Wynkoop wheel in the foundry of Messrs. Clapp and Hamblin of this city, I find it to utilize more than 175 per cent, of the absolute weight of the water used; probably nearer 200. This I regard as no violation of the principle laid down in our natural philosophy, viz., that no wheel can be invented which will utilize 100 per cent., as the wheel in question is not a single one, but such a combination of wheels, as cannot fail to give a vast increase of power. “I. C. Cochran, Principal of Owosso Union School. “Henry Gould, Millwright, Owosso City.” The “Little Giant” turbine mentioned by Mr. Wood consisted of two mixed- fiow runners, cast back to back, one discharging upward and the other down- ward. Mr. Wood’s photograph (Fig. 35) is of the Canadian model of this wheel in which the lower runner was omitted and only the upward discharging runner was used. The Reynolds wheel, made in Oswego, H. Y., was of the same type? except that instead of having two separate runners fastened together, it had a single runner with both upward and downward discharge. A wheel made by Cushman of Hartford, Conn., was also constructed on the same principle. Mr. Horton disagrees with the writers on several points. Perhaps it will be best to treat these in the same order in which they appear. The writers have made as thorough an investigation of the records of the Patent Office as was possible. Unfortunately, this could not be done until after the paper was prepared, but the results are in part incorporated in the discus- sion. Nothing was found which in any way changed the views of the writers with regard to the development of the water-wheel. In fact, this study was hardly necessary, for nothing of importance was found that was not already on file in the library of the Proprietors of the Locks and Canals. At the time that Mr. Francis became interested in the principle of inward fiow, he made a thorough investigation of Howd’s inventions and secured copies of the patents. At about this same time, Mr. Boyden was collecting information regarding the wheels of the period, and thanks to these two men, the writers found them- selves provided with a wealth of contemporary information. Mr. Horton will notice that, with the exception of Johnson’s wheel, every one that he mentions in Table 3 has been discussed by the writers. Leffel’s wheel of 1845, as men- tioned in Table 3, was a mere modification of a Jonval runner and was of no importance in the history of water-wheel development. This wheel did not have guide vanes, but was set in a scroll case. Mr. Horton is apparently under the impression that the modern wheel (“large capacity, high-speed, mixed inward, and axial flow turbine”) depends for its action on a volute case or scroll. This is far from the truth. Under the lower heads, a roomy open-flume setting is often much to be preferred, and DISCUSSION ON THE AMERICAN MIXED-FLOW TURBINE 1353 Holyoke tests are made with such a setting. It is granted that for structural reasons a scroll is of great advantage under high heads, but under these condi- tions the runner becomes low speed. He also states that “pivot guide gates” are one of the essential features of this type of wheel. This excludes all cylinder- gate wheels, such as the excellent “Hercules” Type D, which, although not of extremely high speed, is probably the most thoroughly American wheel made to-day. Mr. Horton feels that sufficient appreciation of Howd is not included in the paper. On a closer reading he will find that Howd is given credit for the inven- tion, but not for its development. The following is a quotation of the inventor^s own words from an advertisement* in which, decrying the principle of inward flow, he states that outward flow is the panacea, thus reversing his former claims : “A Kecent Important Hydraulic Discovery! Highly valuable to millers, manufacturers and millwrights. Howd’s latest improved Water Wheel! Com- bining all the excellent properties of his former wheel, the Reaction, Gilbert’s and the Turbine Wheels, and embracing in addition thereto several new and efficacious principles in its construction and operation, rendering it far superior to any of them in respect to durability, expense and efficiency. ******** “At the time the undersigned began his researches upon the subject, a Wheel well and generally known as the ‘Reacting Wheel’ was in most general use upon such streams ; and having often witnessed the operation of this Wheel, it appeared to the undersigned, from a comparison of the head and quantity of water applied to it, with the power obtained, that it was susceptible of much improvement, or that a Wheel upon some entirely new principle or combination of principles, might be invented which, under the same circumstances, would be more efficient. Prompted by this impression, after much close investiga- tion, and many trials and experiments, the undersigned in the year 1836, introduced to the public a Wheel now extensively known as ‘Howd’s Improved Water Wheel’, of which the following is a brief description: “The buckets, the impinging face of which are bounded by arcs of small circles, are placed between two flat rims, the upper of which is attached by arms to a perpendicular shaft. The water is admitted from without inwardly through shutes at right angles to the radius at the point of impingement, and the motion of the wheel is the same as the direction of the water. “This improved Wheel, upon being tested, was found to excel in every respect, the ‘Reacting Wheel’, insomuch that it has been extensively adopted, and has until recently been superseding all others used under low or ordinary heads of water. “But notwithstanding so much was gained by the Improved Wheel, the undersigned by a long series of scrutinizing observations, discovered several imperfections about it which, if removed or neutralized, would, in all probabil- ity, greatly enhance its utility. “The imperfections are these: First, the water being admitted from with- out on all sides through shutes, the outward apertures of which are open to the volume of water in the flume, and directed at the same angle on the periphery of the Wheel, an artificial whirling or circumgyration of the water in volume in flume is produced, which tends to remove it by centrifugal action, from the entrance of the shutes, and consequently diminishes the pressure to a con- * Published in the Wayne Standard (Wayne County, New York), of June 18th, 1842. 1354 DISCUSSION ON THE AMERICAN MIXED-FLOW TURBINE siderable extent, making a difference in the motion of the Wheel, that may be easily ascertained, other things being equal, to be proportionate to such diminution. “Secondly. — The water is unphilosophically introduced upon the Wheel by being admitted from without, or upon the periphery. For a rotary motion of the Wheel tends to throw it back, so that in causing the water to perform such an unnatural function there is a conflict between the centrifugal force and the force produced by the pressure of the water ; the result of which is that the power of the Wheel is less, by a quantity just equal to such centrifugal force, which is very considerable, operating as it does at every point of the periphery. ******** “Sixthly. — The water being discharged inwardly it is prevented, in cases of low heads, from flowing off readily, and in consequence thereof accumulates and rises above the wheel, and around the shaft several inches, the effect of which is equivalent to a reduction of the head an equal number of inches. * * * * * * * * “Being fully convinced of the existence of the above mentioned imperfec- tions, and incited by the hope of being successful in his attempts to remedy them, the undersigned carried out another protracted course of arduous experi- ments, testing every principle both by theory and practice, and finally in the month of September last, came before the public with another Wheel, desig- nated and hereafter to be known as ‘Howd’s Latest Improved Water- Wheel’, differing from his first Wheel mainly in these particulars: “First. — The water being admitted from within, the artificial whirling or circumgyration before alluded to, tends to move it centrifugally toward the periphery ; or in other words, increase the pressure at the point of impingement. “Secondly. — In consequence of the admission of the water from within, it is permitted to pursue its natural centrifugal force acting at the periphery. ******** “Lastly. — This Wheel when running to the right, has all the advantage of that natural and peculiar cirmumgyratory motion of the water, resulting as it is supposed from the diurnal revolution of the earth, and which never assists the old Wheel, but always retards it. “The undersigned has obtained Letters Patent for his improvement, and is now ready to dispose of Individual and Territorial Rights, to use the same. Numerous certificates, recommendatory of his latest Improved Wheel might be published, but rather than rely upon them for a proper public appreciation of its intrinsic merits, he has chosen to present for consideration the scientific principles involved in its construction, and to rely upon the opinion which candid, skillful and intelligent Mill- Wrights will pronounce, after such prip- ciples shall have been by them fully and carefully examined. If its inherent merits ‘unpuffed’ will not recommend and sustain it; it must sink ‘unpuffed’ into oblivion. “The public are cautioned to bear in mind the distinction between ‘Howd’s Improved Wheel’ and ‘Howd’s Latest Improved Wheel’. “Postage must be paid on all communications by mail, or they will not be taken from the Post-Office. “Samuel B. Howd.” “Newark, Wayne Co., N. Y., May 12, 1842. The advertisement shows a wood cut of a crude Fourneyron turbine. Mr. Horton’s Fig. 36 shows the Howd wheel of 1851. This was a later model, and might well have been modified in accordance with Francis’ improve- ments. Fig. 63 shows the Howd wheel as it was made in 1847. The diagram is an exact copy drawn from measurements made by Mr. Francis on one of DISCUSSION ON THE AMERICAN MIXED-FLOW TURBINE 1355 these wheels in June, 1847. The bucket outline is copied from a full-sized sketch accompanying these measurements. This wheel is basicly different from the one shown by Mr. Horton. For a 5-ft. wheel, the total guide area of discharge at the narrowest point was 1,51 sq. ft. and the same area for the buckets was 5.00 sq. ft. Advertisements of that time show the wheel set above tail-water and dis- charging into the air. Later, it was submerged, but apparently it was not in the early models. It is evident from these two facts that this wheel must have been primarily an impulse wheel. It was supposed to operate at a relative velocity of from about 0.33 to 0.50. Howd was the inventor of the wheel which inspired the Francis turbine, but did he appre- ciate it? Francis evolved, from Howd’s discarded invention, the genesis of the modern turbine. To which man should the greater credit be given ? Mr. Horton does not feel willing to accept the “Family Tree” of the modern turbine. Apparently, he feels that the “volute vortex” turbine suddenly sprang into being. Yet he admits earlier in his discussion that the turbine was the result of gradual evolution. He says, “Again, the American, Hercules, Victor, and Sampson turbines were off-shoots of the Obenchain and McCormick wheels * * *.” The first “American” turbine was patented in 1859, twelve years before the Obenchain “Little Giant”. Otherwise, Mr. Horton’s statement is quite correct, and a closer examination of the paper will show that this is precisely what the “Family Tree” and text demonstrate. The Parkers patented the draft-tube in 1840 and this preceded Jonval, as his wheels were not built until 1841.* Granting that the modern flaring draft-tube was suggested by Boyden’s diffuser, it is difficult to conceive how the diffuser could have suggested the “outlet cone”. . In his Table 4, Mr. Horton shows the marine propeller as a development of the Austin wheel by way of the current meter. The date of the Austin wheel is not known, but it probably was not earlier than 1860. The records of the Patent Office show no record- of any water-wheel patent issued to any one by the name of Austin up to 1873. The first screw current meter probably was Woltmann’s “Moulinet”. The earliest record of this found by the writers is a paper written in 1847 by Baumgarten,t in which it is implied that the meter is about ten years old. John Stevens’ steamboat of 1804 had a propeller, and the use of a screw propeller was suggested in 1680 by Hooke.:}: Mr. Horton states that the modem wicket-gate is a recent European inven- tion. It is merely the Leffel gate of 1862 made stream line. Can this be called an invention ? He does not feel that the Truax wheel is a good illustration of the early propeller wheels, and thinks the Austin wheel more suitable. The writers have * J. Buehetti, “Les Moteurs Hydrauliques Actuels”, p. IX. t Annales des Fonts et Chaussees, 1847. t Appleton’s “Cyclopedia of Applied Mechanics’’, p. 719, New York, 1880. PTg. 63. — Howard Wheel, 1847. 1356 DISCUSSION ON THE AMERICAN MIXED-FLOW TURBINE a four-bladed Truax wheel and a two-bladed Austin wheel, both loaned by- Mr. Kinne. Except for the difference in the number of the blades, the wheels are practically- identical. The Austin wheel was also made with three blades, and the Truax with two, three, four, and five blades. Some of the Truax wheels had the saw-teeth buckets around the rim, that were omitted in others. In preparing this paper, the writers have particularly had in mind small and medium-sized developments under comparatively low heads, in which field their experience has largely been. They have not attempted to trace the truly remarkable increase in size of modern units and the increasing use of the higher heads. This subject might well form a chapter by itself. It is realized that this increase in size has been an essential step in the development of the modem hydro-electric system. The writers feel that the advance has been basicly stmctural and mechanical rather than hydraulic. Full credit, however, is due to those designers who, in recent years, have done so much to increase the speed and power of the mixed-flow turbine. It is believed that the successful operation of a imit in place depends on the correct design of the wheel and its setting, based on proper velocity and velocity distribution and on sound hydraulic principles, rather than on any particular shape, patent, or popularity. Photomount Pamphlet Binder Gaylord Bros. Makers Syracuse, N. Y. PAT. JAN 21. 1908