US1991715A - Packing set shape - Google Patents

Description

Feb. 19, 1935. H T WHEELER 1,991,715

PACKING SET SHAPE Filed July 29, 1952 2 Sheets-Sheet l f "Y @y Feb. 19, 1935. H. T. WHEELER PACKING SET SHAPE Filed July 29, 1952 2 Sheets-Sheet 2 IN VEN TOR.

Patented* Feb. 19, 1935 UNITED STATESz PATENT oFFicE PACKING sur snare Harley T. Wheeler, Dallas, Tex. Application July 29, 1932, serial No. 626,040 12 claim. (o1. 28s-26) This invention relates to the normal pressure applied by a porous elastic structure against contiguous surfaces. and its chief object is to control the amount of friction produced by the combination of thrust due to the fluid pressure and the pressure due to the expansion of the packing by saturation.

I desire to obtain, by a combination of shape and width of porous packing rings, a form of packing which will wear for long periods and pro- 'l duce a minimum of friction upon the moving party A further important object is to select the correct relations of ring shape, annular space width, the depth of the set assembly and the density of the packing rings, so that the lowest possible friction may be obtained for any given rod and box diameter.

IWith the foregoing objects in view, other objects and advantages in the matter of arrangement will be disclosed as the description proceeds, accompanied by the drawings, whereinz' Fig. 1 is the cross-section of a stumng-box fitted with a set of variable slope frusto-cones, a1'- v ranged according to this invention.

Fig. 2 is a set of stepped frusto-cones in a stuffing-box. Y

Fig. 3 is a set of variable arc concave-convex packing rings in a stuffing box.

Fig. 4 is a cross-section of a pump liner having a piston fitted with a set of variable slope frustocones made according to this invention.

Fig. 5 is a set of stepped frusto-conical rings on a piston similar to that of Fig. 4.

Fig. 6 is a 'set of variable arc concave-convex packing rings on` a piston similar to that of Fig. 4.

Fig. '7 represents the effect of fluid pressure thrust on the various ring shapes herein discussed.

line of support, showing alsovthe variation of 45 ring density. y A

Fig. 10 is a wide set of cones and the line of support. t

Fig. 11 is a medium width set of truste-cones and the line of support.

' Fig. 12 is a wide set of frusta-cones and the line of support.

of packing. A

.- Fig. l14 is a phase chart of various ring oon- Fig.I 8 represents the effect of saturation oliv Aof pressure Fig. 15 is a diagrammatic outline of the variable slope frusta-cone set.

Fig. 16 completes the set of Fig. 15.

Fig. 1'1 is a set of stepped frusto-cones.

Fig. 18 is a set of variable arc concave-convex 6 rings.

Referring now to Fig. 1, the machine frame is bored for a stumng-box, and the rod 41 extends through the latter. A packing gland 42 is held in place and adjusted by cap screws 43, 44, the 10 set of packing 45 arranged according to this invention sealing the joint between the stumng box wall 40 and the movable rod 41. While no exact explanation is offered at this point as to the action of the divergent slopes of the ring l5 faces it should be apparent that the reaction to pressure will cause the rings to move against the rod and seal the pressure impressed upon them.

Fig. 2 shows a stufilng-box bored in the frame 20 46, the rod 47 extending through the latter, the

stepped frusta-conical rings adjusted by the screws 49.

Fig. 3 shows a box bored in the frame 51, with extended rod 52, the variable arc set-of concave- 25 nut 60 threaded to the rod '58. A set of packing rings 62 made according to this invention seals the joint between the follower head 5'1 and the '35'"" liner 56, against pressure P as will be elsewhere explained.

Fig. 5 shows a partial cross-section of a piston head 63 holding a set of stepped frusta-conical rings adjusted by the follower head 64, the as- 40 sembly to work in a liner similar to that of 58.

A like arrangement is seen in Fig. 6, where the piston head 66 holds variable arc concave-convex rings 68, adjusted by a follower head 6'1. Referring to the movable surface and the direction 45 the packing ringA sets of Flgs. 4,

5,' and 6 are equivalents of Figs. 1, 2, and 3, the movable surface being presented outwardly.

The improvements comprised in this .invention hinge on balancing in some part the normallyV 50 1 applied pressure due to the thrust of the iluid Fig. 13 is a iluxion chart of the various widths pressure by' the normal pressure exerted by the saturation. There is always in these cases an excesa of normal uid pressure due to saturation of the packing rings by sai'uid which seals the joint, the force due to saturation being greater than that due to uid thrust. 'I'he sources of friction are various, and it is now desirable to combine the effects of friction to offset the factor of the curvature relation of the rod diameter and the stufling-box bore.

In Fig. 'l are shown ring shapes in outline, athe common square braid, b the standard cone, c and d the frusto-cones, and e and f the concave-convex rings. Considering first the thrust 'due' to the fluid pressure, which may be considered as a slow seepage flow away from the sdurce, the impressed pressure P is understood in all cases to act in a direction'parallel to the movable surface and toward the packing as shown by the arrows. In a square shape the uid tension, P, is resolved in any and all directions by reaction, for example the force W normal to the wall and R normal to the rod. In shapes b to f, inclusive, the wall reaction is that of the so-called annular wedge ring where the ring is inclined relative t'o the wall as explained in my application Serial No. 580,015, dated December 10, 1931. As this feature is common to all of the shapes of this specification, it is not here particularly discussed. In the shapes b to f, inclusive, the impressed pressure P is resolved by the parallelogram of forces into S normal to the ring face and R normal to the rod. In shape f, P is not resolved at the movable surface, as the ring is perpendicular to the rod surface.

Considering now the effects of saturation of the packing, Fig. 8 is the representation of saturation within the rings of Fig. '1. In a square shaped packing ring the pressure due to saturation is equal in any and all directions due to the fluid tension, as indicated by the small arrows within the contour of section a; and creates a nlm of pressure on the movable surface, as shown by the opposing small arrows. The shapes U' to f', inclusive, show longitudinal rea'ction and its direct effect against the movable and stationary surfaces. Shapes b', c' and d produce a fllm on the contiguous surfaces while e' and f due to their curved surface tend to reflect the force in the manner shown by the curved arrows. It may be noted that the necessary width corresponding to the shape is also shown relatively.

It is important to now consider the amount of friction developed by use of my packing. The depth of a set of packing is as important as any other factor. An eicient shape, even when used on the correct width will be very ineicient if the set is of incorrect depth. Another and controlling factor which has .not been considered is termed the overhang, and the coincident value of ring support.-

Good results are obtained by the graduation of the porosity of the packing rings, thereby causing a more uniform distribution of friction by controlling the drop of internal pressure. This effect is indicated by the varied shade lines of Fig. 9, wider spaces denoting high porosity, and the close ones indicating low porosity. The impressed pressure is permitted to penetrate into the p orous rings so that the reaction of the ring material against that pressure will produce a gradual drop of internal pressure along the length of the contact surface. In order that the subject maynot be confused with the distribution of friction it is here to be noted that the methods ered with the amount of friction created on any one increment of surface. We have to consider methods of obtaining the least possible amount of friction on any given rod and box diameter by obtaining the least sum of normal pressure due to the iluid pressure thrust, plus that due to saturation. It is to obtain minimum friction that my packing is devised. The theoretical consideration of thrust versus saturation in a single ring is different from that of a complete set, due to the factor of support. The matter of annular space width also has considerable effect on support as the narrower the width, the greater the depth must be t`o secure minimum friction. That is to say, a double width will require much less depth, if the rings are properly designed.

In Fig. 9 is shown a single width, cone-shaped packing, the packing gland 69 taking the thrust due to the impressed pressure. As a hypothesis theline my is denoted the line of support, all lthrust above the line being transferred to `the gland 69, all thrust below being resolved by the parallelogram of forces into force S normal to the ring face and R normal to the rod. That portion of the rings under the line .ry is termed the overhang, or unsupported packing. It is this portion which causes the friction of contact due to thrust.

Referring now to Fig. 10, cone-shaped packing rings of relatively greater width are shown. Owing to the necessity of using less packing depth on greater widths the line 'cy obviously lies at a greater angle with the rod and approaches mz as a limit, the latter being perpendicular to the tip of the conically faced gland 70. In comparison with the volume of the entire set, the overhang increases in proportion to the width, which accounts for the increase of friction with width enlargement, up to a certain maximum.

'Ihe substitution of my frusto-conical rings for Y the cones on a narrow width as in Fig. 4, produces more friction as the overhang is increased thereby. But in Fig. 12, a wider width, it should be apparent that the frusta-cone becomes efficient as the overhung portions are perpendicular to the rod and the pressure P is not resolved. The packing above the line :ry is supported by the conically shaped gland face so that less fluid pressure thrust is directed toward the rod.

The effect of uid thrust which is caused by the drop of the internal pressure, may be partly balanced by saturation effects by the use of the correct ring shapes. vI will no w tabulate the features which produce my correct design of packing set, free of hysteresis, which transfers most of the thrust due to pressure drop to the packing gland and which creates the minimum possible friction for the curvature relation:

v1. The annular wedge ring reaction, which is defined as the reaction of a conical wedge shaped ring against the stationary surface, is correctly accomplished when the ring causes the seepage flow to be ata minimum through the wedge shaped portion of the ring, being approximately 45 degrees for a narrow width, and changing to as much as degrees for an 'extreme width, for the same material. l y

2. 'Ihat diameter of the stuffing-box as compared to therod diameter, which causes the saturation effects to build up a film under pressure on the movable and stationary surfaces, with a minimum of created friction at the movable surface. l

3. That shape of the ring which causes a minimum of thrust due to the drop of internal-pressure to be resolved normally to'the movable surface.

4. That combination vof annular ring vwidth and ring shape which will make the least overhang, or volume under the line of support.

5. That arrangement of porosities which will cause a uniform drop of pressure and thereby uniformly distribute the friction over the surface of movable contact. i

Referring now to my Fluxion Chart, Fig. 13, as developed in application, Serial No. 533,430, filed April 28, 1931, the graphs illustrate pulsations of pressure, the abscisses are in impressed pressure and the ordinates in measured friction, the arrows on the curves indicating the friction d ue to the rise and fall of pressure. Curve a is that of a 45 degree cone, the hysteresis value fk being due to excessive volume increase which cannot be controlled. Curve b shows the pressures acting on my frusto-cone, for example that of Fig. 11, the proportion of friction rise being less, as is shown in the hysteresis value ih. Curves c, d, e and j are the results of similar tests in this development and will be discussed elsewhere.

All packings so far invented, which include the conical shape, my frusto-cones, braided types, plastic and metallic packings and the like have the common characteristic of creating friction in some proportion to the increase of impressed pressure. I have worked out `the elments of design which will eventually lead to the relation of constant friction with pressure increase, as indicated by curve e of Fig. 13.

With reference to said lFig. 13, Oa: is the initialset of a packing, the curve e being for the present the action of a hypothetical type of packing whose measurable friction. neither increases or decreases with either .rise or fall of impressed pressure, and a defining term of flat-compounded is applied to denote such a lack of change. All curves which increase about the initial-set, such as c 'or d are termed over-compounded. A curve such as f, is termed undercompounded.

Various widths of packing create widely different amounts of friction. On what I call the single width, or single standard, I nd that the 45 degree cone, the square braid, the V-shape and similar types of this width, show theirv lowest friction without leaking at the box wall, as indicated by curve a of Fig. 13. A greater width using my frusto-cone gives a curve b. As theY width increases it becomes more difficult to seal the wall surfacefdue to the buckling of the laminations making up the ring, or of the ring itself. To realize the ideal curve e the synthesis of normal pressure indicates that new combinations of the ve tabulated elements of design are necessary. Y

Referring now to Fig. 14 which is a representation of the flux field in the annular space. By the use of this graph the phasing of the seepage ow preparatory to designing the packing for low friction may be determined. The stationary surface plane is on the side s, the movable surface is r. The small arrows indicate seepage ow direction. By the relative spacing of the lines the amount of flow is indicated,:the greater seepage being where the lines are closely spaced. The curve a is that of a packing 4having excessive seepage flow along the box wall and a tight joint at the rod, this being the usual condition of wide rings at this time. Curve b is the reverse of a and shows tightness at the wall and leakage at the rod surface.- It should be obvious that neither a tight joint. This phasing is obtained by taking internal pressure readings at various points in the stuiling box.

The curve c shows excessive seepage at both wall and rod surfaces. But the serial curves d represent retarded flow at the wall and rod surfaces and fast movement in the center of the annular space. That is, there is a high saturation at the contiguous surfaces and the seepage flow which is necessary to expend the energy to cause reaction takes place in the center region of the packing ring and without disturbing the films at the contiguous surfaces. Relations at the wall and movable surfaces vary simulta-J neously as the pressure changes.

I have discovered that, while acertain width of packing will reduce the normal pressures due to saturation, the proper angle may be given to the ring at the movable surface so that there will be produced only enough pressure to balance the force normally received from the thrust due to internal pressure drop.

To consider the proper widths to produce a balanced packing, reference is made to Fig. 15, in which the packing gland 73 is the end-wise support of the packing rings. A hypothetical line of support my originates at the tip of the conical gland face '73. On a wide strip the 45 degree cone tends to buckle in a long box and to draw away from the wall thereby permitting leakage. However, the overhang of the ring a below the line of support is acted upon by some thrust. The inner ring b is my frusto-cone, contacting at the wall at about 45 degrees and normal to the rod, or 90 degrees. Thus pressure P normal t0 the lower part of the frusto-cone and below the line of support is not resolved. The only reaction which 'can occur in the ring is the longitudinal reaction du'e to saturation as shown by the small arrows. The structure of the frustofcone is such that the increase of volume due to saturation is sufficient to create a film at both rod and wall surfaces. The rings a and b are the outer and inner rings respectively of the set to be formed, where minimum friction isdesired.

It may be observed that the foregoing is the nucleus of a packing set, that is, an inner ring operated solely by saturation, and an outer one .the wall, but contacts with the rod at an acute angle thus resolving pressure from thrust normally to the rod. The rings d, e, and f respectively are weaker in saturation effects but stronger in resolving thrust. The thrust T at any point in rings c to f inclusive may be resolved in the manner previously noted.

The foregoing combination of saturated rings and rings responding to fluid pressure thrust will give on one width, for a given rod diameter, the low friction curve c of Fig. 13, for example, but no other width can be found of this shape which will show as low a friction. The tendency of the ring a to buckle due to its length is partly counteracted by receiving the end thrust of the superimposed rings, f, e, d, c, and b, and is forced to elongate to a degree due to the compression.

For the next greater width it may be observed that a greater saturation pressure is desired and less thrust resolution. Referring now to Fig. 17, the packing gland is a conical supporting face, the line of support being my. The ring ais a combination of a frusta-cone on the pressure side and a cone on the supporting side. Ring b is a frusto-cone contacting normally with the movable surface and with the wall at approximately 45 degrees. The point of intersection of the conical surfaces of this and adjacent rings is on the line of support my. Thus this packing set is a succession of frusto-cones stepped together. The combination gives a higher saturation pressure and the pressure P is not resolved at any point.

On a still wider width, if the friction is to be maintained at a low value, the saturation must be increased. So in Fig. 18 the packing'gland 76 has a conical supporting face, the line of support being my, the ring a is a combination having a concave face exposed to the pressure side and a conical depression fitting the gland face 76. Due to thev curved face, ring a has a reflex action which may be indicated by arrows z and w. Ring a is my concave-convex type contacting at the wall at approximately 45 degrees for narrow widths and at the rod normally. Both ends of this ring have a strong reflex action; pressure P cannot be resolved at the point of contact with the rod but is opposed by the reflex action, pressure P in the center of the ring can be resolved and P" can be resolved and is opposed by the reflex action z. 'I'he balance of the rings are concave-convex, the arcs converging with successively greater radii, approaching the outer cone face as a limit.

It should be obvious that if friction is to be decreased some form of a film must be substituted for actual contact between the packing'set and the harder movable surface. This has been attempted yet accomplished only in a slight degree. My research and investigations during the analysis of those packings for which claims have been advanced in this line of endeavor show that they are all deficient in performance. By synthesis of normal pressure I found that a. relation thus obtained may be stated by the expression:

Friction of contact for a given rod diameter is inversely proportional to the annular ring width.

Interpreting this expression: if a narrow width on any rod diameter is investigated it will be found that the friction decreases as the annular space is increased and that there is'a certain Width for every diameter which obtains the lowest value. If a 45 degree cone is considered as a standard packing, then, for example, on a fiveeighths inch rod, the best width is found to be one-quarter inch, and at three inches it will be nine-sixteenths of an inch, increasing gradually between the given diameters. But, if it is desired to secure the lowest possible friction from ve-eighths to three inches diameterl of rod by using all available shapes, it will be found that at ve-eighths inch diameter rod the width of packing is one-half inch and that in rods up to one and one-half inch it is still One-half inch, increasing slightly from the latter diameter to three inches.

It should be apparent that the combinations of curvature possible is unlimited. There are in commercial practice over 2000 stuffing-'box sizes show that any one type of packing -ring made on a given slope can properly meet the conditions of but a fraction oi' the total. The correction of this incongruous condition lies in properly proportioning a stumng-box for each size rod based on the minimum possible friction oi' contact, which reduces the number of boxes necessary to onetenth of those now in use. The application of the shapes, combinations of shapes and conical ring slopes as are disclosed by the foregoing schedules to the proper sized boxes have never been used in practice.

Returning now to Fig. 13, the iluxion chart, showing the characteristics of different packings. For example, on a. given small rod diameter, the lowest possible friction is obtained by design of the stufng-box, selecting the correct ring shape and using the correct ring depth, resulting in the curve b. The preferred arrangement of ring p0- rosities is shown in Fig. 9, and is now used to regulate the internal pressure drop, resulting ina uniformly distributed friction and giving a still lower friction as curve c. The latter curve shows no excessive saturation and no hysteresis as the fluid has instant entrance and egress to the packing structure due to the porosity arrangement. It is therefore to be observed that the packings herein described cause a proportion of the thrust due to the fluid under pressure to be transferred to the movable surface andthe friction increases in some proportion to the pressure increase.

Referring now to Figs. 16 and 18, the fluid pressure Q is shown by the arrows in direction opposite to that of the pressure P. It is the direction of pressure Q to the ring and the shape of the set which is the key to flat-compounding.

The relation of the direction of impressed pressure application to the shape of the packing ring has become important since my discovery of the effects-of saturation. There is now a. basis for the comparison of the results of any design, and it may be noted that similar shapes now in existance are used in a manner which is the reverse of my designs due to basing those designs on the thrustl of the fluid under presure and to the present accepted law that friction is independent of the area of contact. I have gone into some detail to show that a porous structure in contact with a. solid body creates friction in proportion to the area of contact and is also influenced by other factors of seepage flow, the rate of pressure application, the annular Width space, and particu-V larly the relation of curvature before deilned.

Packing constructed in accordance with the showings set out in the drawings will have a minimum of friction and will be self adjustable through saturation, and will tend to be lubricated by the maintenance of a film of iluid along the moving surface.

What I claim as new is:

1. A stuffing box in a machine frame having a rod therein and subject to a fluid medium under pressure, a set of packing rings including an inner packing ring having two conical intersecting surfaces exposed to said medium and an outer ring having a cone-shaped depression in contact with said frame.

2. In a stuiiing box having a rod therein and subject to a fluid medium under pressure, a set of packing rings including an inner packing ring, exposed to said medium, said inner ring having opposed faces, each face being inclined from the outer margin of the ring inwardly toward the source of said pressure medium at a plurality of angles thus providing intersecting conical surfaces, and

an outer packing ring having non-parallel conical faces.

3. A stuiiing box in a machine frame having a rodtherein and 'subject to a fiuid medium under pressure, an inner packing ring therein having a convex face including an outer conical portion and an inner portion at an angle to said conical portion, said face being exposed to said medium and an outer ring having a cone-shaped depression in contact with said frame.

4. A stuing box having a rod therein and subject to a fluid medium under pressure, an inne!` packing ring having parallel concave andfco'nvex surfaces and an outer packing ring having opposite faces, one of which is inclined at a plurality of angles relative to said rod and the smaller end of the conical surface being presented inwardly of said box.

5. In a stuiiing boxhaving a rod therein and subject to a uid medium under pressure, an inner packing section therein having parallel faces including an outer conical zone and an inner zone radial of said rod, and an outer packing ring having a cone-shaped depression on its outer facev and the opposite face comprised of intersecting conical surfaces, each inclined at a different angle relative to said rod, and a plurality of contiguous frust-conical packing sections having parallel faces positioned between said innerL and outer sections.

6. In a stuing box having a rod therein and subject to a fluid medium under pressure, an inner packing section comprised of intersecting surfaces, one of which is conical, each surface being inclined at a different angle to said rod from the adjacent surface, an outer packing section having a cone-shaped depression on one face and the opposite face comprised of intersecting conical surfaces, and a plurality of intermediate contiguous frusto-conical packing sections having non parallel faces positioned between said inner and outer sections.

' 7. In a stufling box in a machine frame having a rod therein and subject to a fluid medium under pressure, an inner frusta-conical packing section having its inner face comprised of intersecting surfaces presented at different angles of inclination relative to said rod and exposed to said medium, an outer frusto-conical packing section contacting with said frame, and a plurality of contiguous frusto-conical packing sections positioned between said inner and outer sections and completely filling said box, all of said rings being positioned with their smaller ends presented inwardly toward said fluid medium.

8. In a stufiing box in a machine frame having.

adjacent the box wall .than on the margin adja- 4 cent the said rod.

9. In combination with a frame, an outer conical support member, a set of packing rings the outer of which is shaped to t against said support member, and a plurality of adjacent rings each having one area parallel with said conical support member, and another area extending inwardly atan angle different from that of said first mentioned surface.

10. In a stuiiing box subject to uid pressure at one end, and having a rod therein, a set of porous packing rings therein, said rings being set. at an angle relative to said rod inclined from the outer margins of said rings forwardly in the direction of said fluid pressure, said set of rings being of greater thickness on the outer side away from the moving rod.

11. In a stuffing box having a, rod therein and of packing rings therein including an inner packing ring exposed to said medium said inner ring having intersecting conical and lradial surfaces on its opposite end surfaces said radial surface being presented toward said rod to contact therewith, and an outer packing ring having opposed corneal faces inclined from its outer edge inwardly toward the source` of movement of the fluid medium, and means in said box supporting the inclined outer faces of said set of rings.

12. In a stuillng box subjected to a fluid medium under pressure, a rod therein, a set of packing rings including an inner packing ring. exposed to said medium, said inner ring having opposed intersecting conical surfaces, and an outer packing ring having nonparallel conical faces, all of said HARLEIT. WHEELER.

Images