US3273390A - Venturi tube - Google Patents

Description

Sept. 20, 1966 w BROWN 3,273,390

VENTURI TUBE Filed Oct. 6, 1965 2 Sheets-Sheet l INVENTOR.

FIG -2 WILLIAM R. BROWN BY n ATTORNEY Sept. 20, 1966 w. R. BROWN 3,273,390

VENTURI TUBE Filed Oct. 6, 1965 2 Sheets-Sheet 2 2. l I II II I 3.127 MGD C 9.7 Hv

l3.8=l RANGE IO=| RANGE O 2 3 4 5 6 7 8 9 I0 I! I2 THROAT REYNOLDS NQ(X KY Flax-4 8 5 i I 7 I I I i I I E I I I I BETA RATIO f .5

O 2 4 6 8 IO l2 I4 I6 I8 20 2 2 24 NET LOSS- 70 OF DIFFERENTIAL.

INVENTOR FIG.- 5 WILLIAM R. BROWN BY M ATTORNEY United States Patent 3,273,390 VENTURI TUBE William R. Brown, St. Paul, Minn., assiguor to Pfaudier Pei'mutit lino, Rochester, N.Y., a corporation of New York Filed Oct. 6, 1965, tier. No. 505,089 Claims. (Cl. 73213) The present application is a continuation-in-part of my application Serial No. 276,480, filed April 29, 1963, now abandoned. This invention relates to an improved venturi tube which is used to produce a differential pressure for measurement :of fluid flow.

Generally, venturi tubes are recognized as providing optimum performance when their performance characteristics correspond to certain recognized advantageous conditions. It is generally accepted within the art that best performance is typified by a venturi tube exhibiting such characteristics as a high pressure differential, a low overall head loss and a stable co-efficient over a wide operating range.

The co-eflicient of a venturi tube is a factor that indicates the deviation of actual flow from a perfect frictionless flow. With varying flows, and more particularly decreasing flows and velocities, the corresponding pipe Reynolds number will decrease to a value at which there is a deviation from a normally straight line co-efficient curve. Where the coefficient curve deviates from a straight line condition, inaccuracy of flow measurement in a standard secondary instrument will occur, and the greater the deviation, the greater will be the inaccuracy in flow measurement. The deviation point is, therefore, a limiting factor which establishes the range over which a venturi tube can be used accurately with a standard secondary instrument. Therefore, to prevent the metering range of a venturi tube from being reduced to that value that includes the portion of the co-efficient curve encompassing the deviation point; and, for all tubes of current design, particularly the compact or high differential low loss type, it is necessary to move the metering range up to higher pipe Reynolds numbers on the straight line portion of the co-efficient curve. This can only be accomplished by decreasing the throat diameter of the venturi tube, which in turn decreases the beta ratio; the beta ratio being the ratio of throat diameter to the main pipe diameter. This reduction in throat diameter, or reduction in beta ratio, increases the velocity, providing a favorable increase in pressure differential, but an unfavorable increase in overall head loss.

For years, venturi tubes used commercially have contained inherent limitations which have imposed restrictions on their performance. For example, tests of over two hundred standard tubes, generally of the Herschel type, show that the discharge co-efficient deviates from a constant value at a pipe Reynolds number of approximately 200,000. Some variations of this conventional type of tube are able to provide a relatively constant co-efiicient down to pipe Reynolds numbers of approximately 110,000, but this appears to be the ultimate lower limit of stability of this type of tube. In the case of tubes of the high differential-low loss design, the lower limit of CO-Bl'IlClGIll stability is reached at a pipe Reynolds number of approximately 350,000. At pipe Reynolds numbers below these values, unstable co-efiicients are produced which cause inaccuracy of flow measurement. Similarly, venturi tubes in commercial use today, including both the conventional and high differential-low loss designs produce a comparatively high overall head loss, expressed as a percentage of the pressure differential as measured at the high and low pressure piezometer openings.

A further limitation of venturi tubes used commercially is that their accuracy is greatly reduced where high beta "ice ratios are utilized. This inaccuracy results from the fact that venturi tubes having high beta ratios produce coefficient instability at the normally used flow rates. Prior art designs, therefore, tend to utilize beta ratios of approximately 0.5, which of necessity, have been generally considered as being commercially acceptable.

Although contemporary design venturi tubes, are of necessity, being used commercially, their inherent limitations remain. My invention, however, greatly reduces these objectionable limitations. The limitations of prior art venturi tube designs which are overcome by my invention are broadly outlined in the following objects.

It is an object of this invention to provide an improved venturi tube that will provide a stable coefficient over an operating range that includes operation at velocities and pipe Reynolds numbers lower than obtainable with prior art Venturi designs.

Another object of this invention is to provide an improved venturi tube that will produce an overall head loss lower than that obtained by any venturi tube presently in commercial use.

A further object of this invention is to provide an improved venturi tube that can incorporate higher beta ratios than those presently considered as being commercially acceptable, without affecting co-efficient stability and accuracy of flow measurement.

A still further object of this invention is to provide an improved venturi tube that will produce a greater pressure differential, for greater accuracy of flow measurement, without adversely affecting the overall head loss.

A venturi constructed in accordance with my invention generally comprises a converging section, a throat section, a recovery section, an upstream high pressure piezorneter tap, a low pressure piezometer throat tap and an element disposed in said throat immediately upstream from the throat tap, the element adapted to direct fluid flow over and about the throat tap to create a quiescent region immediately above and about the throat tap without materially interfering with the mainstream flow.

These and other characterizing features and advantages of my invention will become more apparent upon consideration of the following detailed description when taken in conjunction with the accompanying drawings depicting the same, in which:

FIGURE 1 is a sectional view of an apparatus according to my invention;

FIGURE 2 is an enlarged fragmentary, perspective view of a portion of FIGURE 1 showing the development of an encapsulated quiescent region over and about the throat low pressure piezometer tap;

FIGURE 3 is a view similar to FIGURE 2 but showing another embodiment for developing the quiescent region;

FIGURE 4 is a family of curves showing the coefficient C plotted against the throat Reynolds number of a commercially available high differential low loss venturi tube (1) and a typical improved venturi embodying my invention (11) taken at comparable operating ranges, and the coefficient curve of the most successful tube constructed and tested according to my invention representing the ultimate performance that can be obtained in venturi design (III); and

FIGURE 5 is a family of curves showing beta ratio plotted against the net loss of head of known commercially available conventional Herschel tubes (IV and V), a commercially available high differential low loss tube (VI) and an improved venturi tube embodying my invention (VIII).

Referring now to the drawings, FIGURE 1 shows a venturi tube generally designated 10, of a type that is inserted within a pipe. Venturi 10 includes an inlet section 12, a cylindrical throat section 1 4 and a recovery section 16. Located on the outer surface of throat section 14 is a holding flange 18. The venturi tube is inserted in a pipe line 20 having an upstream section 22, including a flange 2'4, and a downstream section 26 having a companion flange 28. The holding flange '18 is clamped between flange 24 and companion flange 28 to support venturi in pipe line 20. Any suitable fastening means may be used to hold flanges '24, 18 and 28 together, as for example a plurality of nuts 30 and bolts 32.

A high pressure piezometer conduit 34 is located within holding flange 18. Conduit 34 extends from a point on the outer surface 36 of holding flange 18 to a point on the upstream side surface 38 of the holding flange, communicating with an annular high pressure section 40 at high pressure piezometer opening 41.

A low pressure piezometer conduit 42 is also located within holding flange 18. Low pressure conduit 42 extends from a point on holding flange surface 36 to an annular low pressure chamber 44 and further continues from chamber 44 to communicate with the inner peripheral surface 46 of throat section 18 at low pressure piezometer opening 48.

Attached to inner surface 46 of throat section 14 and immediately upstream from low pressure opening 48 is a generally ramp-shaped element 50. One ramp-shaped element 50 is provided for each low pressure opening 48 communicating with inner surface 46. As shown in FIG- URE 2 each element 50 includes a base surface 52 attached to inner surface 46 in line with low pressure opening 48 and a downstream surface or riser 54 upstanding from inner surface 46 adjacent low pressure opening 48. A diagonal surface 56, which is a continuant of the converging sections inner surface 57, connects the upstream end of base 52 to the upper edge 58 of riser 54.

Riser 54 extends radially inward from inner surface 46 in a plane generally normal to the flow through venturi 10 to truncate element 50. The height of riser 54 should be suflicient to permit its upper edge 58 to pierce the boundary layer of fluid flowing through venturi 10.

In any flow along a surface there is a very thin layer of fluid which has a zero velocity due to adherence of fluid to that surface. At some distance from the surface the flow is at constant velocity. The boundary layer, then, is that area adjacent the surface in which the velocity of flow varies from zero to maximum. This boundary layer for a venturi tube is very thin because substantially all the flow through the throat section is at a constant velocity with very little loss. For this reason I have found that the height of riser 54 should be a maximum of 2% of the throat diameter. This height permits riser 54 to just pierce the boundary layer while not materially extending into the mainstream of flow. Allowing upper riser edge 58 to pierce the boundary layer produces a curving of the flow stream lines in the area near low pressure piezometer opening 48 which in turn generates a lower local pressure in this area than is experienced in the main flow. As shown in FIGURE 2, this area of lower local pressure defines a quiescent region 60. The height of riser portion 54 also determines the slope of diagonal surface 56 which is also a critical factor, as too abrupt a slope will cause the formation of eddy currents over upper riser edge 58 thereby destroying the quiescent region 60. F or this reason a shallow enough slope should be provided to produce a smooth, relatively eddy free flow over upper riser edge 58. I have found that for best results this slope should be between and 50.

As shown in FIGURE 2 the width of riser portion 54 is greater than the diameter of low pressure opening 42. This permits the quiescent region 60 to extend about as well as over low pressure piezometer opening 48. In this manner element 50 causes the fluid flowing through venturi 10 to pass over or bridge low pressure opening 52, thereby creating a quiescent region 60 directly above and about the low pressure opening without substantially altering or affecting the main flow through the venturi. Preferably the width of riser 54 should be approximately the square root of t ree times the diameter of piezometer opening 48.

This size permits centering the low pressure piezometer opening within a quiescent region that is generally the shape of an equilateral triangle. As shown in FIGURE 2, riser 54 forms the base of the triangle while the mainstream flow past element 50 forms the other two sides of the triangle about the low pressure opening 48. The width of riser 54 may be made wider to insure the development of a quiescent region completely about the low pressure opening 48. However, the width should be kept to a minimum to prevent unnecessary disruption of the mainstream tlow. Keeping the width of riser 54 to a minimum also insures that the total frontal area of element 50 is kept to a minimum. The frontal area of element 50 should obviously be minimized to avoid excessive interference with the flow of fluid through the venturi. I have found that for best results element 50 should have a frontal area of approximately 2.8% of the total throat area. Any greater frontal area will result in excessive head loss which in turn impairs the accuracy of high and low piezometer readings.

FIGURE 3 shows another embodiment of the generally ramp-shaped element 50 of FIGURE 2. Element 50a of FIGURE 3 is a right conical section having the area of its riser 54a equal to the area of riser '52 of FIGURE 2. Element 50a also operates to form a quiescent area 60a over and about a low pressure piezometer opening 42a. Element 50a however has the additional advantage of a semi circular cross section which allows fiuid to roll smoothly oif the elements curved surface 5 6a, thereby decreasing the possibility of eddy current formation.

The improved venturi tube according to my invention provides co-eflicient stability at pipe Reynolds numbers as low as 15,000, which permits accurate flow measurement over a range of pipe Reynolds numbers previously unattainable without compensation in the secondary instrumentation. When compared with venturi tubes currently considered as being commercially acceptable, this is extraordinary; conventional designs can produce co-efiicient stability at pipe Reynolds numbers down to only approximately 110,000 to 200,000. My invention further provides co-eflicient stability at very low throat velocities. One tube showed no appreciable deviation from stability at the lowest flow rate that could be measured; another was stable from a flow rate of 0.8 feet per second to the highest velocities that could be attained. Moreover, the data was reproducible even where deviation did occur; at no flow rate was the co-eflicient erratic. This is markedly different from the experience with all known types of venturi tubes at very low rates of flow.

FIGURE 4 delineates a family of curves showing the co-eflicient C plotted against throat Reynolds number. Curve I shows the co-eflicient of a commercially available high differential low loss venturi tube of the heretofore generally accepted design having a beta ratio of 0.776. As can be seen, this co-eflicient which is relatively stable at .73 at high flow rates becomes unstable at Reynolds number of 500,000 or less. At Reynolds number of 200,000 or less the co-eflicient rapidly diverges from the .73 value, and the accuracy of measurement will be completely lost without extensive secondary instrument correction. In contrast to this, the venturi tube made according to my invention, as shown in curve II of FIGURE 4 has a co-eflicient which remains stable even at low rates of flow. The co-eflicient remains constant at 0.683 down to Reynolds numbers of 150,000 and only thereafter does the co-eflicient deviate from this constant value. The deviation does not exceed 1% (the normal standard ap plied to commercial tubes) until the Reynolds number drops well below 100,000. Curve III shows a co-eflicient of the best tube so far tested. This co-eflicient actually showed no measurable deviation with flows down to a Reynolds number of 50,000. However, it is believed that curve II will be more typical of the performance of tubes embodying my invention in commercial practice. This allows tubes of my improved design to operate at rela,

tively low flow rates without any correction of the secondary instrument.

FIGURE 4 also illustrates the striking difference in operating ranges between commercially available high differential low loss venturi tubes and tubes of comparable size and beta ratios embodying this invention, for given differential and flow rate. For example, using a standard commercially available secondary instrument capable of measuring a maximum pressure differential of 9.7 feet of water and the generally accepted commercial standard for venturi tube accuracy (plus or minus 1% the range of contemporary commercially available low loss tubes is approximately 4.9: 1. In contrast to this, the range of my tube of comparable size is 13.8:1, or almost three times as great. These ranges are measured between the point where the curve deviates 1% to the flow rate where the tube produces the maximum pressure differential in the range of the secondary instrument. In this case, the upper limit is a flow rate of 3.127 million gallons per day. Using half of the generally accepted tolerance as a criteria (plus or minus 0.5%), the range of accuracy of the commercially available tube is approximately 2.8:1 while that of my tube is approximately 1.

My improved venturi tube provides a substantial increase of the differential pressure over that provided by conventional tubes of equivalent size at given flow rates without a proportionate increase in the overall loss of head. This increase in differential pressure varies with the beta ratio of the tubes. For example, with an 8-inch tube embodying my invention, the increase in pressure differential ranged from 1.6 times the pressure differential created by a conventional type tube of .41 beta ratio to 2.16 times the pressure differential in the case of a .80 beta ratio tube. This increase in the differential produced at a given beta ratio permits the use of a higher beta ratio to produce any given desired pressure differential for actuation of a secondary instrument. This results in a substantial reduction of the overall pressure loss which must be incurred to produce the necessary differential pressures. This reduction is very large in comparison with conventional venturi tubes and is also substantial in comparison with contemporary high differential low loss venturi tubes.

FIGURE 5 shows the net loss vs. beta ratio relationship for an improved venturi tube of my invention and a series of commercially available modified high differentiallow loss and conventional Herschel type tubes. Curve VII shows the net loss of a tube made according to my invention and curves IV, V and VI show the losses of the modified and conventional commercially available tubes. It will be observed that at all beta ratios considered satisfactory for commercial use, the venturi tube produced according to this invention shows the lowest overall loss of pressure yet obtained by any known venturi tube, ranging as low as 2% of the pressure difierential.

Upstream conditions, such as an elbow relatively close to the converging sections of conventional and contemporary high differential-low loss venturi tubes cause instability of the co-efficient in terms of deviations from a flat characteristic and changes in the numerical value of the co-efiicient. -In the case of these known venturi tubes, it is usually necessary to provide a considerable length of straight pipe between an elbow and the inlet into the tube in order to attain reliable operation. However, in the case of tubes embodying this invention, upstream conditions have much less effect on the reliability of the tube. While an elbow or a similar flow disturbing element upstream of such a tube will change the numerical value of the co-eflicient, it does not affect the flat characteristic of the co-efficient. For example, an elbow as close as one pipe diameter upstream of this tube does not substantially affect the stability of the co-efficient. For this reason, it is possible to obtain the ultimate performance of the characteristics even in instruments where it is not feasible to utilize excessive length of straight pipe upstream of the measuring venturi.

A further and very important advantage of my invention is that the obstruction acts as a protecting device for the low pressure piezometer opening. As is well-known in the art, all types of venturi tubes are extremely sensitive to the characteristics of the piezometer openings in the throat. In order to function properly, this opening must be round, clean, sharp and completely free from any minute burrs or other obstructions which could disturb the flow characteristics. Any deviation from a clean, sharp opening and the measuring area produces completely inaccurate and erratic readings in all known types of venturi tubes. For this reason, the production of throat piezometer openings is a costly and tedious job and adds considerably to the overall cost of the tube. However, because of the protective characteristics of the obstructions, tubes embodying my invention are insensitive to minor defects in the throat piezometer openings. In fact, in a series of tests, the low pressure piezometer openings were intentionally deformed and roughened and no change in the stability of the tube or accuracy of the readings was detected.

Thus it will be appreciated that my invention fulfil-ls all of its intended objects, providing a venturi which has a stable co-efficient over an operating range of velocities and pipe Reynolds numbers lower than ever obtained with prior art venturi designs, as well as producing an over all head loss lower than that obtained by any venturi tube presently in commercial use. My invention also provides co-efficient stability and accuracy of flow measurement at beta ratios higher than those presently condsidered as being commercially acceptable, and produces a greater pressure differential, for greater accuracy of flow measurement without adversely affecting over all head loss.

Heretofore in the manufacture of venturi tubes great care was taken to insure a smooth and :as obstruction free inner surface as possible because any obstruction or irregularity of the inner surface greatly altered the flow characteristics and accuracy of the tube. However, with an element 50 designed and located in a venturi tube as taught by the present invention it was found that a ramp having a frontal area of only a few percent of the total flow area will generate a pressure differential much greater than the equivalent throat area change, without substantially altering or affecting the mainstream flow.

While I have shown and described the preferred form of my invention it will be apparent that various modifications and changes may be made therein, particularly in the form and relation of parts, without departing from the spirit and scope of the invention as claimed.

I claim:

1. A venturi tube adapted for fluid flow, said tube comprising:

(a) a converging section;

(b) a throat section;

(0) at least one low pressure piezometer opening in said throat; and

(d) an element attached to said throat section immedi ately upstream of each of said low pressure piezometer openings for creating a quiescent region over and about said low pressure openings, only, said elements having a width about the square root of three time the diameter of said low pressure openings and a height sufficient to pierce the boundary layer of fluid flowing through said throat section; the remainder of said tube being free of obstructions.

2. A venturi tube adapted for fluid flow having a substantially constant co-efficient for Reynolds numbers ranging from relatively high values to values below 100,000, said tube comprising:

(a) a converging section;

(b) a cylindrical throat section;

(c) a low pressure piezometer opening in said throat section;

(d) an element disposed in said throat section upstream of said low pressure piezometer opening for creating a quiescent region over and about said opening without substantially affecting the flow characteristics of the main stream;

(c) said element having its upstream end lying beneath the boundary layer of said mainstream; and

(f) said element having its downstream end truncated and extending through said boundary layer.

3. A venturi tube as set forth in claim 2 in which the width of said element is approximately the square root of three times the diameter of said low pressure opening and the height of said truncated downstream end is not greater than 2% of the throat diameter.

4. A venturi tube as set forth in claim 2 in which the frontal area of said element is approximately 2.8% of the total throat area, wherein the remainder of said throat area is obstruction free.

5. A venturi tube as set forth in claim 2 in which said truncated downstream end is generally of semi circular shape.

6. A venturi tube as set forth in claim 5 in which said element is a right conical section.

7. A venturi tube as set forth in claim 2 in which said element is generally ramp-shaped and includes:

(a) a base surface disposed upstream of and in line with said low pressure piezometer opening;

(b) a downstream surface upstanding from said base surface adjacent said low pressure opening; and

(c) a diagonal surface extending between the upstream end of said base surface and the topmost edge of said downstream surface.

8. A venturi tube as set forth in claim 7 in which the slope of said ramp-shaped element is between 15 and 50 degrees from the horizontal.

9. A venturi tube as set forth in claim 7 in which said diagonal surface is a continuant of the inner surface of said converging section.

10. A venturi tube adapted for fluid flow, said tube comprising:

(a) a converging inlet section;

(b) a substantially cylindrical throat section;

(c) a high pressure piezometer opening located upstream of said throat section;

(d) at least one low pressure piezometer opening located in said throat section;

(e) an element located in said throat upstream of and in line with each of said low pressure piezometer openings;

(f) said elements having their downstream ends truncated to form surfaces substantially normal to said throat section;

(g) said elements having a width slightly greater than said low pressure piezometer openings; and

(h) the remainder of said tube being free of obstructions.

References Cited by the Examiner UNITED STATES PATENTS 1,706,145 3/1929 Collins. 3,196,680 7/1965 Curran 73-2l3 FOREIGN PATENTS 841,541 2/ 1939 France.

RICHARD C. QUEISSER, Primary Examiner.

E. D. GILHOOLY, Assistant Examiner.

UNITED STATES PATENT OFFICE Certificate of Correction Patent No. 3,273,390 September 20, 1966 \Villiam R. Brown It is hereby certified that error appears in the above numbered patent requiring correction and that the said Letters Patent should read as corrected below.

In the drawings, Sheet 1, Figs. 1 and 2 should appear as shown below instead of as in the patent:

Signed and sealed this 5th day of September 1967.

[SEAL] Attest: ERNEST W. SVVIDER, Attesting Oficer.

EDWARD J. BRENNER,

Commissioner of Patents.

Claims (1)

1. A VENTURI TUBE ADAPTED FOR FLUID FLOW, SAID TUBE COMPRISING: (A) A CONVERGING SECTION; (B) A THROAT SECTION; (C) AT LEAST ONE LOW PRESSURE PIEZOMETER OPENING IN SAID THROAT; AND (D) AN ELEMENT ATTACHED TO SAID THROAT SECTION IMMEDIATELY UPSTREAM OF EACH OF SAID LOW PRESSURE PIEZOMETER OPENINGS FOR CREATING A QUIESCENT REGION OVER AND ABOUT SAID LOW PRESSURE OPENINGS, ONLY, SAID

Images