APRIL 1975


Basic Definition:-

Vortex tube sand trap consists of a tube open along its soffit which is embedded across the channel bed of a stream. The tube is open at one end to facilitate free discharge of some flow entrapped from the canal bottom.

The flow which is permitted to escape moves through the tube in a spiral motion; hence the device is called "vortex" tube. Therefore, a vortex tube sand trap is a device for removing from a canal, the sediment which travels near the bed. This device, when properly set and dimensioned, has proved to be very effective for the removal of sand and gravel moving as bed load in canal streams.


Available Information:-

A properly designed diversion works can exclude a portion of the material before it enters the channel. However, many diversion works were constructed before much was known about proper design for excluding or by­passing sediment. In India sediment excluders and extractors have been developed. In U.S.A. vortex tubes have been developed but general design criteria to assist field engineers in designing the vortex tube has been lacking. The phenomena is when flow is passed over the opening, a spiral motion was set up within the tube. This device was observed to be very effective in removing large material even to size of cobblestones.

In general, for operation at ultimate capacity, the most effective diversion method developed is continuous sluicing adjacent to canal headworks in a diversion approximately 60' to curved guide walls leading to headworks and sluiceways greatly increase the efficiency of the general sediment exclusion plan. Each diversion dam with its individual design requirements and field conditions requires separate studies to give the most satisfactory solution. 

Martin and Carlson used the performance effectiveness as the ratio of the concentration of the sand in the water passing through the sluiceway to concentration of sand in water passing through the headworks, i.e. Cs /Ch

Intermittent sluicing was tried and this method proved to be quite efficient in removing sand. However, it was recognized that the fluctuation of the canal water level due to varying discharge through the headworks during the cycle would cause sloughing of canal banks constructed in sandy material. Intermittent sluicing has been used successfully where the canal banks can withstand comparatively rapid water level fluctuations with sloughing.

A curved guide wall extending upstream from a point between the sluiceway and the over flow weir, with a vortex tube extending across the face of the headworks immediately upstream from the headworks sill, were installed in some designs. This arrangement gave a satisfactory ratio of Cs /Ch 7.5.

The ratio of Cs /Ch was lower for small Discharge and small velocities.

G.L. Koonsman and Maurice L. Albertson have presented the efficiency of trapping E in relation to other parameters such as the tube diameter, the channel width, the sediment concentration and Froude number.

They concluded by stating that; when the concentration of sand or bed load in a canal was excessive, the vertical action in the vortex tube was destroyed and deposition in the tube resulted, leading to low efficiencies. The results have shown that slightly better efficiency was found for values of F greater than.1, than for values less than 1. The critical depth gave the highest efficiency with the opening lips were on the same elevation. As the down­stream lip was lowered, increased efficiency was attained with F less than 1. Efficiency increased with relative tube size.

When silt exclusions devices were designed from the results of model tests carried out in Irrigation Laboratory in Colombo, the devices worked very satisfactorily for materials such as sand and larger sizes, but were not suitable for silt in suspension. This was due to the fact that most of these devices were based on model tests and it was found to be difficult to reproduce silt in suspension in models. 

S.V.Rao ; stated that unless the experimental technique was developed such as to correctly simulate the efficiency factor in the model, the proto­type structures naturally could not be expected to behave as originally contemplated. A model could operate with an efficiency representative of the prototype only if the same sediment concentration, i.e. same type of distribution of sediment at different points in a cross-section, was simulated in the model as obtaining in the prototype. This could be achieved when the following factors were considered:-


  1. The precise laws governing sediment transportation.
  2. The equipment for determining the sediment distribution inside a section should be available.
  3. Having known 1 and 2, the prototype features should be established in the model preserving dynamic similarity.


Parshall recommended a vortex tube to be laid at angle of 45" with the axis of flow. Laboratory tests as well as field tests indicated that the optimum action of the vortex tube occurs when the water passing over the lip was moving at or near the critical velocity. That was when the velocity head was equal to one half the depth of water at the lip of the tube. No particular success was obtained in small scale model studies of vortex tube, therefore the laboratory tests were confined to tube sizes ranging from 4 inches to 12 inches in diameter which were assumed to be full scale dimensions.

Vortex tubes have been failure when the velocity in the canal was low or the tube was set below grade in the channel.

Tests carried out by Rohwer, et al on vortex tubes installed in channels 8 feet and 14 feet wide. The tubes used were 4 inches and 6 inches diameter set at various angles to the flow. Conclusions from these tests were as follows;-


  1. The tubes were most active when the depth of water in the channel was slightly less than critical.
  2. Straight or taper tubes were equally efficient in removing sand.
  3. Angle of tube for angles less than 90° to the direction of flow had little effect on efficiency.
  4. Efficiencies of trapping were conspicuously better when elevations of the upper and lower lips were the same.
  5. The tubes would remove from 70% to 90% of bed load carried by the flame.
  6. Tubes in a channel that was 8 feet wide seemed to be more efficient in sand removal than ones installed in a channel 14 feet wide.
  7. When the Froude number of the flow immediately upstream from the tube exceeded 1.3 a considerable amount of sand and gravel was thrown out of the tube and re-entered the channel.


To aid in selecting and justifying a location for the diversion intake structure a hydraulic model is often constructed and operated over a wide range of conditions for both present and future. The model study is made to investigate a specific objective and may require in installation of a sediment ejector or excluder device. Following the specific use, the model is often used to extend the range of data sufficiently to generalize the design of a particular structure for future application to similar problems.

The first step towards economical handling of sediment in irrigation system is to prevent as much sediment as practicable from entering the system at the headworks. Usually there are several requirements considered in selecting a type of sediment removal device. Six of these requirements are the following:-

  1. A sufficient quantity of water must be available for sluicing-type devices.
  2. A sufficient head should be available to produce adequate sluicing velocities.
  3. An ample water supply should be available in the downstream channel to carry the sediment away where sluicing devices are used.
  4. The minimum size of sediment particle to be removed should be known.
  5. The amount of sediment to be removed and gradation of particles should be known.
  6. There should be adequate space available for sediment excavated from a settling basin.


It has been stated that the simplest and probably the least effective sediment excluder is the gated sluiceway placed adjacent to the canal head gate.The effectiveness of this type of excluder depends on the volume of the pocket that can be created by raising the sill of the canal head gate above the sluiceway floor. In many cases curved guide walls are provided at the head gate. These are curved training walls to form an approach channel to the head gate and sluiceway of an intake structure have been found to be efficient in excluding sediment from canals. These walls artificially create secondary spiral currents similar to those formed in a natural curved channel.The radius of curvature and the position of the guide walls with respect to the inlet and the channel currents are usually developed from model studies because the relationships of the several factors involved have not been determined in terms of hydraulic or sediment parameters.

The canal guide wall type of intake operates more efficiently when water is available for continuous sluicing. However, when the amount of sluicing water is limited, operators have sluiced intermittently with satisfactory results.

On vortex tube, Dominy stated that, tests indicated that narrow sluice gate which produced higher velocities in the sluice way approach channel combined with the vortex tube greatly improved the exclusion of sand from the canal. Vortex tubes have been built in several locations and have proven to be efficient devices for sediment ejection.

In certain installation, the vortex tubes perform excellently. However, its general use is limited in canals and canal structures because it cannot function properly under the conditions involving a wide variation in flow or when velocities are low. To achieve maximum efficiency, the flow across the tube should be near critical depth. If a particular section is designed to produce critical flow, operation at partial discharge greatly diminishes the efficiency of the tube. The tube is useful only in the removal of bedload and is practically ineffective in the removal of suspended sediment.



Review of past studies indicates that the vortex tube type of sand trap has been found to be superior to other types of sediment ejectors.The following design features are indicated based on findings of previous investigators.

  1. The Froude number of the flow in the section containing the vortex tube should be near 1.0.
  2. Amount of flow removed by the tube depends on slot opening as well as depth and velocity of flow. An average extractor ratio of about 10% was indicated.
  3. The shape of tube was not particularly important as long as area was sufficient and shape such that sediment would not escape from the tube once it had entered.
  4. Efficiency of   trapping increased   as   size   of   material increases.  
  5. Straight tubes   performed equally   as   well   as   tapered ones.  
  6. There seems to   be   a limiting length   of tube   for   optimum operation about 15 ft.                
  7. The angle of tube should be in the range of 45°- 65° from the direction of flow.


In some researches, it was found that the percentage of flow removed was a function of tube geometry and angle, as well as depth and velocity of flow across the section. From the tests carried out in these researches on vortex tube sand trap, the following design criteria are necessary for the successful operation of the device.

  1. The velocity and depth of flow across the section containing the tube should be such that the Froude number approximates O.B.
  2. The percentage of flow removed by the tube is a function of the depth and velocity of flow in the channel as well as width of opening area, angle, and length of tube. The flow removed usually ranges from 5% to 15% of the total.
  3. The width of opening should usually be in the range of 0.5 ft. to 1.0 ft.
  4. The ratio of length of tube to width of opening 4D should not exceed 20 with the maximum length of tube being approximately 15 ft.

Where L = length of tube   and   D = width of opening.

  1. The tube angle should be 45°.
  2. Straight tubes operate as well as tapered ones.
  3. The elevation of the upstream and downstream lips of the tube can be the same rather than having the downstream one lower.
  4. The shape of the tube does not seem to be particularly important as long as this shape is such that material entering the tube is not allowed to escape back into the channel. A pipe with a portion of the circumference removed seems to operate as well as other prefabricated shapes.
  5. The required area of the tube can be approximated by the relationship,       AT = 0.06 DL Sin Q,

Where AT = cross-sectional area   and   Q = angle of tube to direction of flow.

  1. With the foregoing design specification, the tube can be expected to remove approximately 80% of the sediment with sizes greater than 0.50 mm. The trapping efficiency of smaller sizes will be considerably lower.

Researches have shown that most of the previous works on the maximum efficiency of vortex tube appeared to occur when the Froude number in the approach channel was about unity or slightly less. Recent research has pointed out that what is important is not Froude number but the absolute velocity the approached channel which affects the overall efficiency of the installation in two conflicting directions.

Meanwhile the design of an efficient vortex tube sand trap ideally requires model tests as recommended by many researchers. Recent theory developed for the design and interpretation of the flow behavior in a vortex tube is thought to be of some value.






The following list of reference sources

  1. REHBOCK, T   "Distribution of Transported Sedimentary

:Material in River Diversions".

Hydraulic Laboratory Practice, ed. J.R. Freeman

(pub. Am. Soc. Mech Eng. 1929), pp. 178-179.


  1. MATTHES, G.B. "Diversion of sediment at Branching Channels". Trans. Am. Geophys. Union, Vol.14,

1933, p.506.


  1. RAUDKIVI, A.J.   "Loose Boundary Hyd'rau.Id.ca'", Pergamon, 1967.


  1. UPPAL, H.L.       "Sediment Excluders and Extractors"


Int. Assoc. for Hydr. Res. Report of 4th Congress, Bombay, 1951, pp 261-316.


  1. AHMAD, MUSTAO (1963) "Design of Silt Excluders and Silt Ejectors". Proc. W. Pakistan Eng. Cong., Oct.1963,
  2. 85-119


  1. NEEDHAM, J.   "Science and Civilisation in China" Vol. IV, Part 3. Camb. Univ Press, 1971, pp 211-378.


  1. KING, H.W.       "Exclusion of heavy silt from channels by Vanned Pitching ". Proc. Punjab Engineering Congress, Vol. VI, 1918.


  1. THOlv1AS, A. R. Discussion on "Diversions from Alluvial Streams ".C.P. LINDNER. Trans. Amer. Soc. Civ. Eng.,

Vol. 18, 1953, p.276.

  1. ROUVE & INDLEKOFER "The Influence of Secondary Motion on the Sediment Distribution in Channel Branches of Mild Slope Flow". Int. Assoc. Hyd. Res., Istanbul, 1973, Vol.l,pp 517-27.


  1. BLENCH. T   ,Discussion on paper by Parshall (Ref 12, below) Trans. Amer. Soc. Civ. Eng., Vol. 117, 1952, p 213.


  1. PARSHALL, R. L. (1931) "Sand Trap for Canals Works on Vortex Principle” , Civil Engineering (A.S.C.E., New York), Feb. 1931, p.419.


  1. PARSHALL. R.L. (1952) "Model and Prototype Studies of Sand Traps" Trans. Amer. Soc. Civ. Eng. Vol 117, 1952



  1. AIDw), MUSTAQ (1961)     Discussion of "Vortex Tube Sand Trap" Robinson (15) Proc. Amer. Soc. Civ. Eng. IR4t Vol. 87, Dec. 1961.


  1. ROHVIER, CARL "Effect of Shape of Tube on Efficiency of Vortex Tube Sand Traps for various types of Sand". (In files of U.S. Conservation Services Office).


  1. ROBINS ON , A.R. "Vortex Tube and Sand Trap" Proc. Amer. Soc. Civ. Eng. IR4, Vo1.86, Dec.1960


  1. BROWN, n.R.M. (1962) "A Study of the Fluid Motion in Vortex Tube Sand Traps and the Factors influencing their efficiency". Ph. D. Thesis, University of London, 1962.


  1. KOONSMAN & ALBERTSON "Design Characteristics of the Vortex Tube Sand Trap". Int. Assoc. Hydr. Res. Report of the 4th Congress, Bombay, 1951, pp. 317-323.


  1. BROWN, D.R.M. (1964) "A Study of the Factors influencing the efficiency of Vortex Tube Sand Traps" Paper No.6715, Inst. C.E. London; (Abst+act in Proc. I.C.E. Vol.28, June 1964, pp 197-200) Taken largely from the Author's Ph.D. Thesis (16)


  1. FRANC IS , J.R.D. "Experiments on the motion of solitary grain along the bed of a water stream". Proc , Roy. Soc. London. Series A. Vo1.332. April 1973) pp 443-471


  1. BAGNOLD, R.A. "The Nature of Saltation and of 'bed load' transport in Water". Proc. Roy. Soc. London, Series., Vol.332 (April 1973) pp 473-504.


  1. RYAN, A.J. "Modern Control Works Solve Diversion Problems". Civil Engine{::ring, (A.S.C.E. New York) Vol.l1, Dec. 1947, p.30


  1. RINGLEB, F.O. "Separa t i on Control by Trapped Vortices" Chapter in 'Boundary Layer and Flow Control'; ed. G.V. Lachmann, Vol.1 pp. 265-293. Pergamon, 1961.


  1. MILNE-THOMSON, L.M. "Theoretical Hydrodynamics" Maemillan & Co., 5th ed. 1968


  1. BATCHELOR, G.K. (1967) "Fluid Dynamics" Cambridge Univ. Press. 1967


  1. BATCHELOR, G.K. (1964) "Axial flow in trailing line vortices". J. Fluid Meeh. Vol.20, Part 4 (Dee 1964) pp 645-58


  1. GARTSHORE, 1.S. "Swirling flow in a straight pipe" M. Se. Thesis, University of Lendon, Jan. 1960.


  1. DUNCAN, TROM & YOUNG "The l':echanics of Fluids". Edward Arnold Ltd. 1960.


  1. HENDERSON, F.M. "Open Channel Flow" Macmillan, New York, 1966 p.16.


Hydraulic Research International Association for Hydraulic Research, 1958, page 304.


International Association for Hydraulic Research, Report of the 4th meeting, Bombay, India, 1951, page 317-323 •


International Association for Hydraulic Research, Report of the 4th meeting, Bombay, India, 1951, page 261-316.


Dominy, Floyd E., Design of Desilting Works for Irrigation systems. Journal of the Irrigation and Drainage Division, ASCE, Vol. 92, No. IRA, Proc. Paper 4988, December, 1966, pp. 1-26.


Marton, Rarold H. and Carlson, Model studies of Sediment Control Structures on Diversion Dams, Bureau of Reclamation, United States.


Campbell, P.A., Vortex Tubes Sand Traps: A Study of the Motion, M.Sc. Thesis 1973, Imperial College.


Hydraulic Laboratory Report No. Hyd 384. Bur. of Reclam. U.S. Dept. of the Interior, Washington, D.C., Feb. 1954.


Hydraulic Laboratory Report No. Hyd 385. Bur. of Reclam. U.S. Dept. of the Interior, Washington, D.C., April 1954.


Hydraulic Laboratory Report No. Hyd 479. Bur. of Reclam. U.S. Dept. of the Interior, Washington, D.C., March 1962.


Hydraulic Laboratory Report No. Hyd 499. Bur. of Reclam. U.S. Dept. of the Interior, Washington, ·D.C., August 1962.