Design and Test of a Model Siphon Spillway Citation Ramseyer, George Numa (1923) Design and Test of a Model Siphon Spillway. Bachelor's thesis, California Institute of Technology. doi:10.7907/E8MJ-QA38. https://resolver.caltech.edu/CaltechETD:etd-06232004-143720 Abstract No Abstract. Item Type: Thesis (Bachelor's thesis) Subject Keywords: Civil Engineering Degree Grantor: California Institute of Technology Division: Engineering and Applied Science Major Option: Civil Engineering Thesis Availability: Public (worldwide access) Research Advisor(s): Unknown, Unknown Thesis Committee: None, None Defense Date: 1 January 1923 Record Number: CaltechETD:etd-06232004-143720 Persistent URL: https://resolver.caltech.edu/CaltechETD:etd-06232004-143720 DOI: 10.7907/E8MJ-QA38 Default Usage Policy: No commercial reproduction, distribution, display or performance rights in this work are provided. ID Code: 2699 Collection: CaltechTHESIS Deposited By: Imported from ETD-db Deposited On: 24 Jun 2004 Last Modified: 03 Oct 2019 23:20 Thesis Files Preview PDF (Ramseyer_gn_1923.pdf) - Final Version See Usage Policy. 2MB Repository Staff Only: item control page

WHESIS REPORT Design and Test of a Model Siphon Spillway. by George IN. xamseyer Department of Civil Engineering Class of Nineteen Hundred ‘twenty whree. ‘Galifornia Institute of Technology Pasadena, California. 6. Spillways in General Siphon Spillways Advantages of Siphon Spillways Siphon Spillway Problems fests (a) Apparatus (b) Tests Proper (c) Conclusion Data Sheets page 16 17 21 22 23 25 1 Spillways in General. In its ordinary use the spillway is a devise ‘for removing excess or surplus water from a resev~ oir or canal, in order that the water level within the resevoir or canal may not rise above the point consigered sate or fixed uon as the maximum allow- able height. When used to discharge water from a canal it discharges the water when it becomes fill- ed above its normal capacity. Such a spillway by the very nature of conditions must be automatic in its action and absolutely reliable. It serves as a safety value to prevent the canal being overloaded and its banks tovped and washed away. ‘The s.ill- wey must be so located that the water may be divert- ed quickly and without damaging tne adjacent land. {It should be easily accessible and orovide for the greatest degzee of etfeciency »ossible. In the design of the ordinary spillway we must consider two factors; first, the maximum quant- ity of water to be discharged and, secondly, the max- imum rise above the normal water surface which is to be expected or ratner that the structure, whether it ke @ canal or a dam, will stand. Having established the discharge and nead of water available the length of svillway required tnay be computed by using tne standard weir fornula, Be 3 Q = KBu~ in which, Q = Quantity of water discharged in cubic feet per second. K= A constant or the efficienty. B = Width of spillway H =The effective head or the distance from the crest of the spillway to the water surface. In order to vrovide for a safe foundation the spillway should, in general, be located where the canal is well in cut. ‘he design of a spillway is largely controlled by surrounding conditions. In the case of a storage dam the spillway location is controlled by the natural slope of the land that leads away from the spillway. The chief criticisim in the past has been that spillways have not been built with sufficient cap- acity for the extraordinary floods. Provision has been made for high water in most cases but the fail- ure of many dams shows that the extraordinary flood of the century has not been provided for. It must be remembered that during the life of the structure there may occur some combination of rainfall or melting snow which will produce a flood exceeding any previously known maximum. It is this flood that the designer of a spillway must provide for. An example which will illustrate this point is & project on the Cache Creek, a tributary of the Sacramento Niver, California. Observations indicated @ maximum flood of about 20,000 sec. ft., but from various indications, one of the engineers figured out a possible, although improbable, flood during earlier years reacning 60,000 sec.ft. ‘This figure was regarded as absurd, but nevertheless the spillway was designed for a capacity of 60,000 sec.ft. It had hardly been finished when a flood occurred of approximately 65,000 sec. ft. Had not the spillway been designed for this capacity it would undoubtedly have been swept awaye In San Diego, California,y, they were not so fort- unate,for when the big flood of 1916 came the dams at Sweet water and at Otay, which had nowhere near sufficient Spillway capacity,were washed out causing tremendous loss to the city both in lives and in capital,to say nothing of prestige. The safest and only correct method of designing @ Spillway is to make a very thorough and complete study of all the rainfall data available and assume the very worst conditions possible then add a sufficient factor of safety and tese design the spill- way for this capacity. there are two general types of spillways, namely, overflow spillways and siphon spillways. ‘he dis- tinguishing features of the two types are that the capeity of the overflow spillway depends on the length of crest and the height of water above the crest and is inereased in no way by the distance through which the water falls below the crest; on the other nand the cap- acity of the siphon spillwey depends on the area of cross section of the smallest part of the siphon and the difference in elevéstion between head water and teilwater. In other words the sivhon utilizes the fall (below 33.9 feet) to increase the capacity while the overflow spillway s:akes no use of this fall. 2 Siphon Spillways. In Europe siphon svillways were used as early as 1870. ‘the first types required the use of an ejector to cause them to function properly. AS soon as the water rose above its normal level the ejector came into action and sucked the air from the siphon proper which consisted of @ large pipe. ‘the ejector made possible the functioning of the siphon long before the water had risen above the level of the bend. vYhe presence of the smaller pipe, which is nec- essary to set the Siphon in overation forms a weak point in the arrangement since it is liable to become choked with debris or its action impaired by frost. ‘the circular section is also undesirable from an, econ- omic standpoint and renders it impracticable except for small diameters. the credit for the design of the first siphon spillway without an ejector goes to Heyn, & German + s siphons. It > {> civil engineer. Fig.I shows one of h t; is built of riveted sheet iron and is simply fixed on a floodgate in an irrigation canal. Under test it showed an effeciency of about 50%. the action of this siphon depends on the fact that the water at the beginning in a small stream along the lower surface of the crest of the siphon draws with it more and more air from tne u,ver ert of the sivhon and this quickly brings the siphon into operation. Contemooraneously with Heyn, Gregotti of Italy constructed spillways without ejectors. He demon- strated their applicability under the most varied circumstances and perfected their construction by the use of reinforced concrete. Figure W. shows two type especially favored by Gregotti. In France a higher state of developement and more effecient functioning was obtained by using reinforced concrete and proportioning the different parts to effect this effeciency. fear of interference with operation during freezing weather was the chief reason for American engineers failing to adopt this type of spillway but a modificationof design led to the overcoming of this defect and to its subsquent adoption in this country. This menace was obviated by lower- ing the intake leg below the normal surface of the water and the draining of the sealing basin. This also prevented debris collecting or lodging at the throat. The use of siphons as spillways is much more recent in this country than in Europe. ‘The first to be built in this country was on the New York state barge canal. More recently they have been con- structed on several projects of the U.S. Reclamation Service and on many California irrigation projects. Some of the more notable exam-les are Orland, Salt River, Guma, and Sun River of the teclamation Service; the South San Joaquin Irrigation District, Sweetwater dam near San Diego,and the Southern California Edison Co. at Huntingto, Lake, California. As previously stated, the height utilized to produce flow in an overflow spillway is figured from the surface of the water above the spillway to its erest. ‘he available head is from the water surface above to the water surface below the Gam. in the siphon, however, flow is oroduced by a nead ecual to the difference in elevation of the water surface at the inlet of the siphon and the elevetion of the water at the outlet end. If the outiet is not submerged it is measured to the center of the outlet opening. the tendency has been, in most cases, to assume that heads in excess of 24 ft. were not adaptable to Siphon installation because it was accepted that the limit of vacuum draft had been reached at that point. Some siphons have been built with neads in excess of 24 ft. and in specific cases with heads as nigh as 52 ft. Kost engineers agree however that 64 ft. is the limit of vacuum draft. There are two general forms of Siphon tubes, the first hes a throat and outlet of equal cross section put not necessarily of uniform shape. ‘his type is rather common in Europe and was followed to some extent in the developement of the Ocoee River project in Tennessee. The second type has a contracted throat section and expands with a divergent tube as an outlet to overcome local losses of head and other contributing defects. This is the type most generally used in siphon design in this country and also in Europe at present. y ¥ 4s ea f: rgihs yi si 7 F = = + ae i, FIGS, I ce i P LSS we ., Fp ees Oe F a rng Ps if . Par iad yer we os ya of “ FTE Ee, “here is hardly any available data to aid the designer in the determination of either the size or shape of the vital parts of the siphon to realize maximum effeciency and proper functioning. whe great cost involved and the greater risk of failure discourage most designers from departing to any marked degree from the prevailing practice especially as they have no assurance of increased effeciency or ‘accomplishment of the desired purpose. No experiments have been made to determine other then the overall effeciency. Practically nothing is known of the relative losses in the various parts of the siphon. It is readily seen that a slight change in some part of the siphon might greatly effect the overall effeciency. This is indeed a field that -IMight well be explored. It is true, no doubt, that the overall effeciency is whet is desired but to in- sure economic design the effeciencies of the various parts would be very desirable. the formula used for calculating the effeciency a= xa [2 gh is-~ in which Q = Quantity of water discharged expressed in cu. ft. 1 sec. K= A constant or tne effeciency of the Siphon | -A = The cross sectional area of the throat or smallest section expressed in Square Ft. H = The fell expressed in ft. The siphon Spillway was first introduced in this country by GF. Stickney on N.Y. State Barge Canal. Since then many experiments have been conducted and the Siphon has been improved and developed so that now it is used to discharge far greater volume than was originally contemplated. The discharge through a closed conduct, as previously stated makes it poss- ible to utilize the fall at a dam to increase the velocity and consequently the quantity discharged. The Siphon is put into operation by a slight rise of the water. The fluctuation of the water above the normal stage and ig restricted within narrow limits. The rate of discharge over a dam depends on the depth of water on the crest, increasing as the depth increases and hence an overflow spillway may not have much discharge capacity until the water surface rises to @ considerable height. Such a spillway is not immediately responsive to variations in the stream flow and does not closely control the water level. Although siphons were known to the ancients and have been utilized for nundreds of years, their use . was confined to small tubes, of an inverted U-forn, for conveying liquid over the edge of a vessel and delivering it at a lower level; it is only within recent times that the siphonic principle has been applied to conduits of large size and capacity, for discharging at a dam. The action depends on the difference in pressure at the extremities of the conduit, the flow being to- ward the lower level and ceasing when the levels 10. coincide or when air collects at the highest part of the conduit. In order to pass water over the summit of the Siphon higher than the upper water surface, it is necessary to utilize the air pressure on the water surface. A clear idea of siphonic action may be had if it is understood that the flow is due to the push of the air on the upper water surface. This proves ‘that the maximum head available is 33.9 ft. which is equivalent to a pressure of 14.7 per Sq. In. A siphon in operation, and acting under the siphonic head or less, will continue in operation as long as the inlet .is submerged sufficient to prevent the admission of air, whether or not the outlet be submerged. This is due to the fact that the water is flowing at such a great velocity that air bubbles cannot rise in the lower leg against the outflow. The problem is thus resolved into the devising of methods for priming the siphon. A siphon that is sealed against the entrance of air, will be vrimed when there is sufficient | flow thru the conduit to absorb and carry out the air entrapped therein. In some siphons it is nec- essary that the lower leg be sealed by discharging below the lower water surface or by a Sealing basin at the end of the leg, whereas others discharge freely and are sealed by a jet of water in the siphon. In order that the inflow velocity and consequently the entrance loss be small, the inlet to the sipnon should be large. ll. whe inlet should also be placed well below the water surface in order to exclude floating matter which might clog up the conduit. Whe upper leg Should taver gradually from the inlet to the mini- mum section, called the throat, at the top of the lower leg. ‘he upper bend, called the crown, should be curved and of moderate radii to reduce friction loss. ‘the lower leg may be either inclined or vertical. whe outlet should be flared upward so that air bubbles, rising thru the water while the. Siphon is priming, may be deflected outward. Fig. 1 shows a siphon with vertical lower leg. the water rises above the air vent, sealing it agsin- st the admission of air, and spills thru the crown of the siphon, forming & jet at the throat thet jumps across the lower leg. This jet makes a dia- phragm of water which seals the upper part of the. Siphon against the entrance of sir from below. A certain quantity of sir is entrapped above the jet in the crown of the siphon ena this air is absorbed and carried out by the flowing weter. It is not necessary to eliminate all the air to induce -viming; the removel of a sméll part is sufficient to start action. A Siphon with slightly inclined lower leg and a sealing basin at its bottom is shown in Pig. @. A rise in the water level above the dam seals the upper end of the Siphon and produces a flow of water thru the conduit which fills the sealing eo SS 12. basin and thus seals the lower end. A jet of water is formed at the crown of the siphon which plunges into the water of tne sealing basin, carrying air apsorbed from above, and this air is carried outside the conduit, before bubbles can rise taru tne water. 4s in the previous case priming is a progressive process and is quickly accomplished. Fig. 3 shows a siphon with the lower leg on a flatter slope and terminating in a sealing basin. In this case, the inclination of the lower leg does not admit of a jet being formed directly. Recourse is had to an auxillary channel where jets are formed, that seal the upper part of the siphon. This causes the unper part of the siphon to prime and produce flow in the lower leg which carries out the air and primes the lower portion. The sealing basins are provided with small pipes to drain them of water when. the siphons are not in operation, so that they Will not become clogged with ice in cold weather. Another tyve of Sipnon with an inclined lower ‘ leg, which doen not require a sealing basin is shown in Fig. 4. In this siphon a curve is formed in the plane of the back wall of the lower leg, near the bottom, such that the water flowing thru the conduit will be deflected, and s& jet will be formed that will jump across the leg. whis jet makes a diaphragm of water which seals the leg against the entrance of air from below and causes the siphon to prime. 13. All the previously described siphons prime when there is a slight rise of the water level above the dam, the rise varying from 3 in. to 1; ft. depending on the size and the type of the siphon. Some types of siphons require submerged outlets, but sealing basins as a rule are objectionable and should be dispensed with if possible. When the water rises in the sealing basin, the air entrapped in the siphon is compressed, which may, unless relieved, retard or completely prevent priming. Also when priming starts there is a pulsation in the flow, due to air rising inside the siphon from the water of the sealing basin and restoring a part of the volume previously taken away. The sealing basin reduces the effective head somewhat and requires considerable additional masonry,thus increasing the cost of the structure. - During the operation, the shell of a siphon is under pressure, the intensity depending on the head. The air pressure tends to collapse the walls on the conduit, this, however is offset to some extent, by the fluid pressure within the siphon. The pressure at the inlet is equal to the pressure of the atmos- phere( 32.9 Ft.) plus the head of water above the inlet. At the crown the internal pressure is practically zero, if the siphon is acting under the maximum siphonic head, and in such case the unbalanced atmospheric pressure will amount to e117- per Sq Ft. The effective outside pressure decreases almost to zero st the inlet and at the outlet. 14. It has not been possible to derive a satis- -factory formula for the flow thru a siphon that will take all the factors into consideration for the simple reason that no formula or expermental data for deter- mining the loss af head at the bends etc. is available. The Sweetwater Dam, near San Diego California, has an overflow spillway at one end and a siphon spillway at the other, ‘where are six siphons, each six ft. high by twelve ft. wide, acting under a head of 223.5 ft. ‘he inlets are twelve ft. high by 8.5 ft. wide. The lower legs, which are inclined, are of practically uniform area throughout, varying from six by twelve ft at the throat to 8.5 by 8.5 ft. at the bottom and terminate in a sealing basin. ‘he discharge capacity of this spillway is estimated to be about 1200 cu. ft. per seg. some of the uses which a siphon spillway may be adopted with advaunt- age are eS follows:- 1. Yo regulate the water surface in a stream “we above & dam at the nighest permissable level, giving the maximum head for power developement, the max- imum storage, the greatest navigable depth, and re- duce the fluctuation of the water surface to a minimum. &. 0 provide an outlet from a flume or forebay when a contrest flow is mainteined, which will operete with a slight raise in the water surface and vill dis- cherge the entire flow, in case of sudden shut down of the power plant or 4 sudden obstruction of the flow oe 0 use in place of gates valves to discharge water from a reservoir. whe crown of the sipaon may be placed entirely above the water level, so tnat no s leakage. is possible, and the Siphon pe arranges to be operated by the manipolation of small hand valves, ‘thus avoiding the use of low and cumbersome gete cists. 1o) 16. 3 Adventages of Siphon Spillways The greatest advantage of the Siphon spillway is the economy of ground space and construction cost% due to the greater quantity of weter which can be controlled in the same space and with the same variation in head, To bring out this point more clearly let us sight an example. In the Turbigo hydroclutin plant, near liilan, an ordinary overflow spillway is used. Near Verona & siphon spillway is in use. ‘hey both have practically the same discharge, 3000 and 2800 cu. ft. per sec. respectively. ‘he Veron spillway works from a face 59 ft. in leneth and 3" is the greatest variation in water level. On the other. hana the spillway at Turbigo is 300 ft long and a rise of 2 ft is necessary before the quantity of water required flows over. | Rapidity of control is guother advantage of the siphon spillway. 4 variation head of only a few inches causes the spillway to at full capacity. whereas the overflow spillway does not become fully effective until the danger point is reached. The siphon spillway is purely automatic and does away with the human element,which is often times an advantage. Some engineers were afraid that the siphon spill- way would not function properly in cold weather. These fears were largely exeggerated and by lowering the lip of the spillway the danger of being clogged by ice o# or debris had been largely, if not totally,done away with. 4. Siphon Spillway Problems. Altnough the siphon spillway has been used in Kurope and America for many years , nothing of a definite nature is known of the exact condition governing the design of the siphons. Practically no | data is available to aid the designer in departing | from the present practice to increase effeciency. Siphons in the past have been designed using the general principles of hydraulics, but the data avail- able is vague and indefinite. ‘there is almost no information regarding the relative losses in various parts of the siphon; the intake opening, the entrance leg, the throat, the outlet leg, the discharge loss, and the loss due to shocks and bends. It is entirely possible thet & very minor change in any of the deteiled parts would materially change the working of the siphon and increase the effeciency. | But the project of building a siphon is too large and costly an undertaking to experiment with unlessit is absolutely certain that the change will be for the better. In other words, the men investing fortunes in the build- ing of a siphon spillway are unwilling to depart from the general practice unless they can be assured tnat the new design is a better one, and is certain of properly functioning. the following is ean outline of invertigetion of siphon discharge as outlined by Mr. AsL. Sonderegger, consulting Engineer. 17. 18. Outline of Investigations of Siphon Discharge. A perusal of the current literature shows that in the past Siphons have been applied to regulate the ele- vation of a water surface, whether it be in canals or reservoirs, the idea being to secure a steady water level. The problem which should be investigated, is the construction of a siphon which would produce a constant discharge under varying water levels. The flood control problems in southern celifornia are complicated by the fact that storage reservoir sites of sufficient capscity to store the entire flood flow, are not available, and that recourse must be had to “detention reservoirs", which regulate the peak flow to such a quantity, as can be safely carried by the flood channel. At the same time it is desirable to use such reser- voirs, at least in part, for the storage of flood waters, for irrigation or domestic supply, as the combination of purposes will facilitate the financing of projects. A study of the run-off of the various streams shows that the peaks are of comparatively short duration but very erratic; that in many instances it might be possible to utilize comparatively poor sites, if it were possible to produce an instantaneous maximum discherge through 4 sopill-way located in the upper levels of the dam. 19° If the spillway is of the overflow type, its dis- ‘charge will depend upon the heed above the lip of the Weir, so that a considerable portion of the available on storage is wasted to acauire the head necessary for a maximum flow, the discharge increasing with the square root of the Srd power of the head (Yeir formula). On the other hana, if a siphon spillway is chosen, the initial head will be possible 30 feet and an increesing head over the lip of the signon will inereese the flow. in proportion to the square root of the head. ghile tne ro vortionate increase is thus much less than with an cc overflow, it is still considerable. It is believed thet a study of the shape of the outlet leg of siphons, poss- ibly a diverging outlet, might lead to a type thet would automatically stop sivhon action if the head increased beyond a certein point, and beyond such point would operete like a@ submerged opening, thet is minus the Sipnonic head. a It is assumed that with o siphon operating under & Sivhonie action of theoretically 32 ft., any head over that of the sipnon lip would contribute to the discharge in proportion of the square root of the total head. (This might be proven by experiment). 2. On account of the high velocity which the water poss- 1 esses when leaving e siphon, and the necessity of dischar i large volumes of water down e steep spillway, it would be desirable to have sipnons with outlet legs of flat grad- ients, so that they may be economically constructea in tne naturel rock, and with the point of discharge et some 20. distance from the base of the dam. This is particularly desirable in Southern California, where much of the rock is shattered and must be protected from washing by water under high velocities. See Siphon at Seon- Switzerland. whe problem is to regulate the admission of air and to produce @ vacuum on a flat grade. Possibly a flat gradient might also help to solve problems under item 32. Se on account of the shock and vibration experience in the starting and stopping of large siphons it is desirable that interruptions should not be too frequent. it is con- ceiveble that the fluctuations in the waterlevel of a res- ervoir, due to a varying influx, might cause a starting and stopoing of siphon action at snort intervals and for a considerable period of time, thereby endangering the struct- ure. “xperiments should be made with various types of siphon inlets and air ducts to cause a delay in the siphonic action until the waterlevel stands 2 or 3 feet above the lip, and to produce a sipnon discharge which would continue until the weterlevel has dropped to the elevation of the lip. In this way a reservoir storage of a few feet depth would limit interruptions in sipnonic action to long intervals. It is possible that a study of different types of inlet openings and of different siphon crests will lead to a sol- ution, combined with varying gradients of the outlet leg, Ww with and without weterlock at the outlet. /44 MODEL TF ‘ PLITE, EF SCALE = Gg my? ee acer 16 la Bote Kw ' Pe a nee Ld Siphon Spilway Model - Scale f"=7" PLATE L£ ale 5. Tests (a) Apperatus:- A siphon was designed similar to the one used some years ago by lir. a.f. Allen except that the inlet was so designed that wooden blocks could be inserted and thus vary the opening. . Plate 1. Snows the Siphon spillwa model and Plate a. Shows the two wooden blocks that were used in the tests. It was first thought that the sides of the siphon would be made of plate glass. Such a design by the use of a lamp on the outside would make possible the visual- zation of exactly what happened in the various parts of js the siphon during its operation. It is neecless to say that this would prove invaluable for the proper design of the siphon. However complication in design, cost, and limited time, outweighed its advantages and this design was eliminated. the plant available was two large cisterns, located in the hydraulics laboratory. hese two cisterns were separated by a 12 " concrete wall in which was cut a hole large enough to place the siphon. A siphon wes put in place. s#orms were then built around it and it was then cemented with a 1:1 mortar. “his held it in place very rigidly. whe water was pumped from one cistern into the other by means of a centrifugal pump driven by 4 D.C. motor with an adjustable field, which made possible a wide var-~ iation in speed, and consequently the quantity of water pumped. Zee Water in an overhead tank was used to prime the pump. The quantity of water pumped could be directly measured by means of a venturi meter. The two cisterns were connected by two valves which made possible pump- ing from either one into the other. _(b) Tests proper:- Tests were made on the two models; first, of the height at which siphonic action began and at the height at which siphonic action ceased, second; the time it ' took to prime after the height at which siphonic action began had been reached, and third; the effeciency for various values of Q. For the first test the pump was started and water pumped into the inlet cistern until siphonic action be- gan, the height of the water was then noted. ‘The pump nen was tees shut down and the height at which the action ceased was noted. It was a very simple matter to deter- mine the exact time of the start of siphonic action as a roaring noise was very prominate. For test number 2, the water was pumped in, as before, and the time it took to obtain a prime after the water had reached the level at which siphonic action began was taken. The time was taken by means of a stop watch. For Test number 3, the pump was run at various speedo. The quantity of water discharged was measured by the venturi meter previously refered to. The head under which the siphon was operating was obtained by measuring the distance of the tailweter and headwater below the reference floor and subtracting the two. Knowing the cross sectional area at the throat the effeciency, c, was determined by the 4 RO6 equation | Q=cA JY 2 gh where Q —Quantity of water in cu. ft. per sec. eo = Effeciency A= Cross sectional area in sq.ft. g= Acceleration of gravity ~— 52.2 ft per sec h= Effective head in ft. The tailwater was measured by a wooden float which ‘was attached to a string and registered on a gage. The gage was calibrated and the depth of tailwater below the floor thus obtained. (c) Conclusion:- The results of the tests are shown in the data sheets. In the first test it is shown that by using Model 1. the siphon operated on a variation of 5 ", while by using Model 2 the siphon operated on @ variation of over 7 It is concievable that a siphon might, if operating on too small a range, be brought into operation to frequently end the vibration thus instigated might seriously weaken the structure. My experiments showed that siphonic action did not “begin until the water reached level (a) and ceased only when the water reached level (b). It is concievable, then, ' that by proper design the siphon could be made to operate with practically any veriation thought necessary. Test no. 2 brought out the fact that siphonic action is produced. by the exhaustion of the air within the siphon. “ith the greater curve of Model & prime occured much more rapidly because by the time the water had reached the height at which prime occured Model & had less air in the throat and inlet to dissipate and consequently primed mucn more quickly. Test no. 3 showed that the effeciency varied but slightly for either model. Tflate 3 shows the results of this test. Acknowledgements: - the writer wishes to acknowledge the value of the suggestions given by Prof. Franklin Thomas, of the " Institute, and lr. A.L. Sunderegger of Los angeles. Also the very velueble information he derived from reeding the revorts of Mr. 3.0. Allen, ir. Gl. Stickney, and Mr. A.v. sichelson. OL o9 “= A*INI/DISAT os yn — o7 og o£ Started ee ee es ee we me ne oe oe oe —— re ome me oe me ee me owe me o_o me ee we ee me mn ee oe On ae oe ome oe ee ew ee oe -— + 2 OF oo on om oe oe ee em oe - - { « al 3} Oe — oy ee ee ee ee « ao = 22-2 ------3' BY +--+ ----- Le Ave aiff. or say 25. red " 6% a 4 4 « P=) « : 26. DATA SHLET 3 2 Model 1 Time to prime after water reacned Vrial | ime seconds Lo we+---- ee 26 " Qo nee nen een en n-ne eee 2-2 --- 16.5 " ae z5 " ee 18 " ae 22 " QO sett crte ahehateietnnnnneneianenenninnenel aL ‘t Yo we een nnn ene nen 24.5 " Qo wae ee -------------------- 2S " Qo wane nn nnn enn en nnn tnt 25.5 " LO wenn nnn renee rrr ze.5 0 VL se-ee------------- w------- 25 " Avg. time 2226 SEG. Correction Yead water rdg 38 a" 3" 2," 3° DATA SHEET Model 1 effeciency eo factor 5 2 to be added to tailwater rdg. Tailwater rag 4'O° 4° 0" 4° 0" ae 2 a 1 " 42 4'0" 4'3" 4' 0" { a 43 Cor. dead 5 6. 56. 56. 518. 57" 515" 5 8. 57. ( 7] 5 10 Q cu. ft./sec. 1.65 200 2.20 2.50 Led 2.55 2.40 2.60 2.70 61.5 61 66.8 67.5 27. Started "28. DATA SHEET # 4 Model 2 Prime Finish Diff. ~-------- B’l]p -------------- 00 7" ~wn------ B'11g -------------- 7g ~-------- 4'Q" aaa------------ 3°" wane nnn SB 10k --------------- 7" --------- B']] --------------- 7" a BLL" --------------- 7" ~-------- 3 ‘11d --------------- "' a B'11p --------------- 73" ~---+----- 3 11p' --------------- 7" ~--------- B10} --------------- 7" ~--------- 3'10$ --------------- 6" Avg. diff. 7.2 Time to prime after water reached DATA SHEET Model r= 5 (eae ee cae os ee te tam ee som Oa eee in oe ey ne oe a a ee rn ee eer ee ee mm ne tw on oe tae ae me in ee Os OS a oe Ew en ye ne oe ee ne See oe A YO a ee ee ee ee (ome om nO oe tee ee ee ee eo Oe oe ew oe ee ee et eer a ae et rer rr me ts ene tae a ne ee Oe Os De OD ee a ee ne Ae SN kal ee ee ee tr ee ee ed Avg. time seconds ve y VN, 10.5 Sec. DATA SHEED # 6 liodel ze effeciency 1 + Correction factor 5 2 to be added to tailwater. Headwater Tailwater Cor. Head Q Eff. 8 6h 39° gil 54% 1.67 43 3) 6k 3105 9 OF 5'6" 1.88 48.5 3 7 3°10" 90" 5°42 2.16 56.5 t atl ry “A \ MN ry «.' Bs e'10 =o 0 p'42" 2.52 60.5 af 4 uw ‘ ' “ 3! 72 2°10 go’ = 548" 2.40 3») 625 Bt 3'10" g'o" 648" 2.46 64 3 9" 4° 0" 92" 5'5" 2.62 68 ' at ce" Poe 8) eo 39 4°1 9°3 5'6 8.65 ~— 68 B ' a 1 oN . eos 310 4°3 9°5 5'7 8.80 70