NASA TECHNICAL NOTE

z

TORQUE A N D LEAKAGE RATES OF VARIOUS LUBRICATED LIP SEAL DESIGNS AT A PERIPHERAL VELOCITY OF 120 FEET PER SECOND

by Luwrence P. Lndwig, Thomus N.Strom, dnd Robert L. Johnson

Lewis Research Center CZeueZand, Ohio

_+-N A S A -TN 0-541 1­

o.,/

t (74,

N A T I O N A L A E R O N A U T I C S A N D S P A C E A D M I N I S T R A T I O N W A S H I N G T O N , D. C. S E P T E M B E R 1 9 6 9

c

https://ntrs.nasa.gov/search.jsp?R=19690026391 2018-04-24T22:52:20+00:00Z

TECH LIBRARY KAFB, NM

Illllll11111lllllllllllllllIIH111111111Ill 0132202

TORQUE AND LEAKAGE RATES O F VARIOUS LUBRICATED LIP SEAL

DESIGNS AT A PEFUPHERAL VELOCITY

OF 120 FEET P E R SECOND

By Lawrence P. Ludwig, Thomas N. Strom, and Robert L. Johnson

Lewis Research Center Cleveland, Ohio

NATIONAL AERONAUTICS AND SPACE ADMINISTRATION .- -

For sole by the Cleoringhoure for Federal Scientific and Technical Information Springfield, Virginia 22151 - CFSTI price $3.00

ABSTRACT

Lip-type shaft seals were run at peripheral velocities representative of require­ments in mechanical components for rocket engines. The seals with butyl rubber Lips had prominent wear-in torque peaks that were attributed to lip temperature rise and sub­sequent lip shrinkage (Gow-Joule effect). This effect could cause sudden failure; grad­ual wear-in, however, reduced failure probability. The seals with glass-filled poly­tetrafluoroethylene lips also exhibited wear-in torque peaks. The seals with graphite-filled polytetrafluoroethylene lips exhibited low leakage and good wear characteristics. Runnifig torques of the various lip seal types were similar to that of a segmented-carbon shaft seal. For butyl and graphite-filled polytetrafluoroethylene lip seals, the dynamic gas leakage rates were similar to the static leakage rates.

ii

TORQUE AND LEAKAGE RATES OF VARIOUS LUBRICATED LIP SEAL DESIGNS

AT A PERIPHERAL VELOCITY OF 120 FEET PER SECOND

by Lawrence P. Ludwig, Thomas N. Strom, and Robert L. Johnson

Lewis Research Center

SUMMARY

Some mechanical components of rocket vehicles use lip-type shaft seals to prevent leakage of bearing and gear lubricants. Although required seal life is only several hours, the sliding velocity is high and data on seal life, leakage, and heat generation is of interest. Various lip seal designs were run at speeds (24 000 rpm) representative of rocket vehicle mechanical components. Running torques and lip temperatures were con­tinuously recorded in short-term runs to 600 seconds with zero pressure differential across the seal. Starting torques and leakage rates were checked at pressures to 14 psig (9.65 N/cm 2 gage). The lip orientation used produced a lip radial force that decreased as pressure increased. The seals with a butyl rubber lip exhibited prominent wear-in torque peaks that were attributed to lip temperature rise and subsequent lip shrinkage (Gow-Joule effect). These torque peaks could lead to sudden failure (burnout); gradual wear-in, however, reduced the probability of failure. Seals with a glass-filled polytet­rafluoroethylene lip also exhibited wear-in torque peaks; several sudden failures oc­curred. The seal with a graphite-filled polytetrafluoroethylene lip exhibited low leakage rates and good wear performance. Running torques of the various seals were s imilar to that of a segmented-carbon shaft-riding seal. For the butyl rubber and graphite-filled polytetrafluoroethylene lip seals, the dynamic gas leakage was s imilar to the static leak­age.

INTRODUCTION

Operational service time of the order of several hundred seconds at high shaft speeds (i.e. , 24 000 rpm) is typical of the requirements of some rotating mechanical components for rocket engines. With the addition of the ground checkout time, the total

Nitrogen gas at r Lubrication tube s u r e P3-,% ,-Sheet steel seal housing

,Seal l io !

-Lip-to-shaft contact area (region of lubr icat ion f i lm)

L.UUIII. ' '-Shaft diameter,

(3. 175k0.Oo0 cm)0.002

CD-10364-15

Figure L - L i p seal apparatus (P2 > P1).

rotating operational time could be only several hours. For these conditions of high speed and short operating time, elastomeric lip seals have been used in turbopumps for rocket vehicles. Figure 1 shows a typical seal with an elastomeric lip. Usually the sea l con­sists of a sheet steel housing which is press fitted in the housing of the turbopump. This sheet steel housing carries the elastomeric lip which is loaded against the shaft by a garter-type spring and/or by the interference f i t between the lip and the shaft. In gen­eral, the seal surface speeds are high (> 100 ft/sec or 30.5 m/sec), and the amount of energy dissipated by the rubbing contact is necessary for calculation of a heat balance or thermal map of the mechanical component. For example, the heat generated by a seal next to a bearing could effect a change in the bearing internal clearance. Thus, some information on seal running torques is of interest. Further, during ground check out and before flight, the seal condition is checked by pressurizing the seal and measuring the gas leakage; thus some data on seal wear and static leakage ratio are of interest .

Manufactures of lip seals usually state seal operational capacity in te rms of limiting shaft surface velocity and pressure differential; these limiting figures usually apply to operational usage of hundreds or thousands of hours, and manufacturers give an approxi­mate maximum useful operation velocity of 50 feet per second (15. 2 m/sec) for a syn­thetic rubber l ip seal. No data are given by manufacturers on the limiting speeds or pressures for a useful life of several hours.

The useful life of a lip seal is dependent on the extent to which a lubrication film is formed between the lip and the shaft surface and on the seal lip temperature. Studies on

2

the lubrication of synthetic rubber shaft seals (ref. 1) indicate that full-film lubricxtion can exist and that the fi lm thickness was of the order of 0.0001 inch (0.00025 cm) thick. No mechanism for film formation was postulated, but the higher velocities (27 ft/sec or 8 . 2 m/sec) produced larger film thicknesses than the low-velocity (0.3 ft/sec or 0.09 m/sec) runs. This suggests that high peripheral velocities may not be detremental to wear life. The film formation is probably associated with small (micro) sl ider bearing geometries formed from the elastic deformation of the lip material in conjunc­tion with surface roughness and waviness. Similar microasperity lubrication of face seals is discussed in reference 2. However, as compared with face seal, the seal lip material has a much lower elastic modulus; therefore, local elastic deformation of the sliding interface can more readily occur under the action of hydrodynamic forces.

These elastic deformations are conducive to formation of a lubricant film between the lip and shaft; however, the data of reference 3 suggest considerable detrimental ve­locity effects. As velocities increase, out-of-round shaft geometry causes local gaps between the lip and the shaft surface because the lip cannot respond to the frequency of the shaft disturbances. In addition, seal interface temperature, a factor that affects film thickness, is dependent on the shaft surface velocity.

The objectives of this study are to (1) measure seal torque and gas leakage rates of various lip sea l designs and (2) compare the wear of the various seals.

The lip seals were operated at peripheral velocities of 120 feet per second (36.6 m/sec) in an oil-mist, nitrogen-gas atmosphere. Running torque, static torque, and gas leakage rates were measured at various pressures (from zero pressure differential to applied pressures that tended to reduce the lip contact force). The seal l ip materials were butyl rubber, glass-filled polytetrafluoroethylene, and graphite-filled polytetra­fluoroethylene. A segmented-carbon shaft-riding sea l was used as the basis of com­parison.

APPARATUS AND PROCEDURE

A schematic diagram of the apparatus used to evaluate the various lip seals is shown in figure 1. The seal lip r an against a 304 stainless-steel sleeve that was mounted on the stub end of a grinding spindle. The sleeve had a 1. 250-inch (3. 17-cm) nominal diam­eter and a 4- to 8-microinch (10. to 20. 3X10-6-cm) plunge-ground (no axial grind­ing motion) and polished surface finish. The maximum dynamic runout was 0.001 inch (0.0025 cm) as measured by a capacitance probe. A lubricating mist was produced by dripping the lubricant (a synthetic diester, MIL-L-7808) on the rotating shaft at the rate of 1 cubic centimeter per minute. The test sea l was supported by a housing that in turn was supported by a torque-meter with a 50-inch-ounce (35.3-cm-N) capacity. In some

3

cases, the runs were made with zero pressure differential; in other cases, the seal was pressurized (nitrogen gas) to reduce the lip contact force (see fig. 1). This procedure is the opposite of the conventional mode of operation in which the seal assembly is such that the sealed pressure tends to increase the coiitact force. However, in some short-t e rm applications, such as in components for rocket vehicles, applying the seal in the pressure unloading configuration is sometimes used so that the seal prevents gross lu­bricant leakage and allows the case to vent as the altitude of the vehicle increases. Measurement of starting torque and leakage provided a relative gage of lip contact force. Shaft speed (24 000 rpm) was monitored by a magnetic pickup. Leakage was measured by a wet-test gas flowmeter, and the torque was continuously recorded on magnetic tape. In some runs, a thermocouple was attached to o r embedded in the seal lip (see fig. 1).

A schematic diagram of the test seals, typical commercial seals, is shown in fig­ure 2. The approximate cross sections of the seals with butyl rubber lips of 80 duro­

,Sheet steel housing7, Sheet steel housing, I I I,’ ,,-Loading spring 7, \\, I

I Springs 1;

I Lip LCarbonI

I segments

Shaft diameter, 1.250 in. 13.175 cm)

Nominal outside (a) Butyl rubber lip. (b) Glass-filled (c) Graphite-filled (d) Segmented-diameter, polytetrafluoro- polytetraf luoro- c a r b n shaft­2.00 in. (5.08cm) ethylene lip. ethylene lip. r id ing seal.

CD-10365-15

Figure 2 - Seal types evaluated.

meter hardness and the seals with glass-filled polytetrafluoroethylene (80 percent poly­tetrafluoroethylene) lips a r e shown in figures 2(a) and (b), respectively. These seals are loaded by a garter spring. Figure 2(c) shows the approximate cross section of the seal with a graphite-filled polytetrafluoroethylene lip. Mechanical springs for lip load­ing were not used in this design. The approximate crQss section of a segmented-carbon shaft-riding seal that was run for comparative purposes is shown in figure 2(d). No ex­ternal heat was used; thus, the lip temperature rise was attributed to frictional heat generated at the sliding interface.

4

RESULTS AND DISCUSSION

Seal R u n n i n g Torque and Temperature

Seal with butyl rubber lip. - The torque and temperature variation during running t ime for butyl rubber l ip seals operating at zero pressure differential and a velocity of 120 feet per second (36.6m/sec) is shown in figures 3 and 4. The traces of figure 3 are typical of seals that failed suddenly after several minutes of operation. The sudden fa i l ­ure was postulated to be a result of seal l ip shrinkage. Shrinkage is known to occur in heated rubber materials under tensile stress (Gow-Joule effect). When assembled on the shaft, the lip is in tension and exerts a radial force against the shaft. Shaft rotation causes frictional heat generation at the sliding interface, and this heat can increase the lip tension (and radial force). More tension leads to more heat, and sudden failure can be induced. In contrast, the torque and temperature traces in figure 4 are for a seal that did not fail, probably because the initial lip inside diameter was larger and the ra­dial lip force lower than that of the run shown in figure 3. The s teep rise in torque (at

-I350

200

350 1 5 0 1

325

100 300

2 2 . 5 r20.0t3L­28

17*5115.0

1 2 . 5 1 ._ a? 3

10.0 g i­

5. 0

2.57-iO L

0 60 120 180 240 Time, sec

Figure 3. - Seal r u n n i n g torque and l ip tem­perature as func t ion of operating time. Lipmaterial, butyl rubber; shaft surface ve­locity, 120 feet per second (36.6 mlsec); MIL-L-7808D lubr icant rate, 1cubic cent i ­meter per minute; in i t ia l l ip inside diam­eter, 1.170 inches (2.972 cm); pressure differential, zero.

5

-20 425

Temperature ­400 /---

TorqueA v u n v 0 - 0

300"'[ 0 60 120 180 240 300 360 420 480

Time, sec

Figure 4. - Seal r u n n i n g torque and l i p temperature as funct ion of op­erating time. Lip material, butyl rubber; shaft surface velocity, 120 feet per second (36.6 mlsec); MIL-L-7808D lubr icant rate, 1cubic centimeter per minute; i n i t i a l l ip inside diameter, 1. 180 inches (2.997 cm); pressure differential, zero.

85 sec) was typical of the seals with butyl rubber lips. The steep r i s e and subsequent decrease were postulated to be the result of lip shrinkage followed by a small amount of wear-in (corrective wear), after which the seal operated satisfactorily (as evidenced by the small amount of lip wear measured at the end of the test). The lip wear was deter­mined by the width of the lip contacting the shaft; in this case, the lip was 0.025 inch (0.0635 cm) wide after operation. It should be noted that the lip temperature, measured 0.03 inch (0.076 cm) from the contact, is high (280' F or 410.9 K) under this type of operation.

An additional run in which the l ip contact force was decreased by an increase in pressure differential provided added evidence of the postulated corrective-wear theory.

2The run (fig. 5) was star ted with a pressure differential of 8 psig (5.5 N/cm gage), and considerable gas leakage (5200 standard cu cm/min) occurred because of the re­duced lip contact force. The im,wrtant point noted was the sudden increase in torque and subsequent decay each time the pressure differential was decreased. Each reduction in pressure differential was accompained by a corresponding increase in torque because of the increase in lip contact force.

Further evidence of the corrective-wear process, for the seals with butyl rubber lips, is provided in figure 6, which shows the torque variation with time for a sequence of runs. In the first run of 60 seconds, the peak torque during acceleration to 120 feet per second (36.6 m/sec) shaft surface velocity was 30 inch-ounces (2.12 cm-N). A torque r i s e near the end of run 1 is similar to that shown in figure 4 and suggests that the corrective wear process had been induced by l ip shrinkage that accompanied the lip tem­

6

7 +- E, a- 3 1[1 10 k- -- 3

Pressure differential,

8 6 (5.5) I (4. 1)

;as leakage rate, c u c m l m i n

5200

O L 60

4 2 0 (2. 8) (1.4) (0)

100 <50 0

-4

0 120 Time, sec

Figure 5. - Seal r u n n i n g torque as funct ion of operating t ime and pressure differential. Lip material, butyl rubber; shaft surface velocity, 120 feet per second (36.6 mlsec); MIL-L-7808-D lubr i ­cant rate, 1cubic centimeter per minute.

Run

20

0 2 .1e 5 P ;L,I 1

4

1 - u0 0 GO 120 180 240 Time, sec

Figure 6. - Seal r u n n i n g torque as funct ion of operating time and number of runs. Lip material, butyl rubber; shaft surface ve­locity, 120 feet per second (36.6 mlsec); MIL-L-7808-D lubr icant rate, 1cubic centimeter per minute; pressure differential, zero.

0

7

0

al

0

perature rise. After cool down, the second run of 60 seconds duration exhibited lower peak torque peak torque than the first run; no corrective wear occurred near the end of the second run as with the first run, and the running torque was lower than that for the first run. The third and fourth runs showed continued evidence of wear, in that both had lower peak torques and running torques than runs 1and 2.

Seal with glass -filled polytetrafluoroethylene lip. - The torque and temperature variation for a glass -filled polytetrafluoroethylene seal operating at zero pressure dif­ferential at 120 feet per second ( 3 6 . 6 m/sec) is shown in figure 7. An er ra t ic torque

16­

y 3 7 5 t 0 F Torque

al-L

CL aa­3

LE 150 e b­0 L

I- 325

3w "

0 60 120 180 240 300 360 420 480 540 600 Time, sec

Figure 7. - Seal r u n n i n g torque and l ip temperature as funct ion of operating time. Lipmaterial, glass-filled polytetrafluoroethylene; shaft surface velocity, 120 feet per sec­ond (36.6 mlsec); MIL-L-7808D lubr icant rate, 1cubic centimeter per minute; in i t ia l l ip inside diameter, 1.230 inches (3.124 cm); pressure differential, zero.

surge occurred near the end of 540 seconds of operation; this torque is unexplained. In additional runs (fig. 8) in which the pressure differential was decreased in steps from an initial value of 14 psig ( 9 . 6 N/cm 2 gage), the torque indicated a small wear-in effect. Each time the pressure differential was reduced (thus increasing the lip contact force), the torque increased slightly and then decayed slowly with increased running time Com­pared with the rubber butyl lip sea l design, glass-filled polytetrafluoroethylene lip ex­hibited a running torque that was less sensitive to changes in pressure differential. This relative insensitivity is probably the result of the greater stiffness of the filled polytetra­fluoroethylene and the lip geometry differences.

Further evidence of the wear-in process is provided in figure 9, which shows the rruzning torque as a function of time for a sequence of four runs. The first run exhibited a torque peak(50 in. -02 or 3 5 . 3 cm-N) that is probably the result of a wear-in process. Succeeding runs (2, 3, and 4) with the same seal had lower start ing torques than the first run and had no torque peaks.

8

N

Pressure differential.

leakage rate, c u cmlmin

110 = -Negligible

120 180 Time, sec

Figure 8. - Seal runn ing torque and l ip temperature as function of operating time and pressure differential. Lip material, glass-filled polytetrafluoroethylene; shaft surface velocity, 120 feet per second (36.6 mlsec); MIL-L-7808D lubricant rate, 1cubic centimeter per minute.

0

c'

T ._

+ 10

0t 0 60 120 180 240

Time, sec

Figure 9. - Seal r u n n i n g torque as func t ion of t ime and number of runs. Lip material, glass-filled polytetrafluoroethylene; shaft surface velocity, 120 feet per second (36.6 mlsec); MIL-L-7808D lubr icant rate, 1cubic centimeter per minute; pressure dif fer­ential, zero.

9

2.

Seal with graphite-filled polytetrafluoroethylene lip. - The torque and temperature variation for a seal design with a graphite-filled polytetrafluoroethylene lip at zero pres­su re differential and 120 feet per second (36.6 m/sec) is shown in figure 10. The start­ing torque of approximately 24 inch-ounces (16 .9 cm-N) was attributed to a lip contact forced judged to be high. During the run, the torque was er ra t ic and high initially; how­ever, the torque leveled off to a constant value of 10 inch-ounces (7.l cm-N) after 350 seconds of operation. The thermocouple cemented to the side of the lip registered a temperature of 275' F (408. 2 K) at the end of the run.

mist 20425b3 0 0 L

N

10.0 7 16re 'a,._ a- 12/- ,/­

7.5 $ 0l2*j1I-

5+ 5 4f-300 01- I I I I I I I I

0 60 120 180 240 300 360 420 480 540 600 Time, sec

Figure 10. - Seal r u n n i n g torque and l ip temperature as funct ion of operating time. Lipmaterial, graphite-filled polytetrafluoroethylene; shaft surface velocity, 120 feet per second (36.6 mlsec); MIL-L-7808D lubr icant rate, 1 cubic centimeter per minute; ini-t ia l l ip inside diameter, 1. 173 inches (2.979 cm); pressure differential, zero.

Segmented carbon seal. - A s a basis of comparison, the running torque of a segmented-carbon shaft-riding seal wa,s measured (fig. 11) I Each decrease in pressure resulted in a corresponding decrease in running torque. Since the pressure differential acts to increase the contact force between the carbon and the shaft, the relation of torque as a function of pressure shown in figure 11 should be anticipated. The torque t race shows that the segmented carbon sea l does not have the marked fluctuations in torque that occurred for the seal design with butyl rubber and filled polytetrafluoroethylene lips. However, the running torque at zero pressure differential is not significantly different froin that for the lip seals .

10

N

0 0

Pressure differential, p2 -.p1,

Leakage rate, cu c m l m i n

:11w 9 0 1 700 1 500 I 240 20

0

g l o ge e t- t­

.L

I5-

.-e

0 120 240 360 480 J

600 Time, sec

Figure 11. - Seal r u n n i n g torque and gas leakage rate as func t ion of t ime and pressure dif­ferential, Seal material, carbon-graphite; shaft surface velocity, 120 feet per second (36.6 mlsec); MIL-L-7808D lubr icant rate, 1cubic centimeter per minute; sealed gas, nitrogen (room temperature).

Start ing Torque, Leakage, and Wear

Seal with butyl rubber- -.__-lip. - Starting torque for hand-turned startup is plotted in figure 12(a)for various sealed gas pressures . Also plotted in figure 12(a) is the ruiiiiing torque at hand-turned speeds. As would be expected, this slow-speed running torque is somewhat less than the breakaway torque.

Static leakage ra tes measured at the time the start ing torque values were deter­mined (before running at 120 ft/sec or 36.6 m/sec) are shown in figure 12(b). The curve for static leakage indicates a gradual increase in leakage as pressure differential increases to 6 psig inch (4 .1 N/cm 2 gage); be.tween 6 and 8 psig (4. 1 and 5. 5 N/cm 2

gage), the marked increase in leakage rate indicates that the sea l leakage gap has signi­ficantly increased. This point was termed the lip-opening pressure. After running at 120 feet per second (36.6 m/sec), the lip-opening pressure occurred at 5 .6 psig ( 3 . 9 N/cm 2 gage) compared with 6 psig (4 .1 N/cm 2 gage) before running. This lower

lip-opening pressure was attributed to wear that was measured as an 0.021-inch (0.032-cm) increase in diameter. Also plotted in figure 12(b) is the curve for the gas leakage rate at 120 feet per second (36.6 m/sec). The comparison of leakage rates

11

.a

m

x

1560 standardl4Y c u c m l m i n d c._E....::I,-Starting torque for E,

accelei-ation to "3 l2N

E z P

6 10 : Torque (hand- m -Before... turned speed1 r u n

w- .a-

P ,--Starting +

L

+ .a 0 7

x 4 .a

2

~0

- / i0

I - L l 2

I 4I I

6I

8I 1

1I I

0

2

0 2 4 6 8 I

Pressure differential, P2 - P1, psig

I I L I I I I I u . 1 I I I 0 1 2 3 4 5 6 7 0 1 2 3 4 5 6

Pressure differential. P2 - P1, Nlcm'gage

la) Torque. lb l Leakage rate.

Figure 12 - Seal starting torques and gas leakage rate as funct ion of pressure differential. Lip material. butyl rubber; in i t ia l lip inside diameter, 1. 180 inches (2.997 i m l ; sealed gas, ni t rqen; new seal.

330 standard c u cmlmin

"3 After r u n (see fig, 71E 80 -3

Starting torque Torque lhand // for acceleration turned Speed'

16 f to 120 ftlsec ;Running /'- 136.-6 misec)

r\ ,/,AStarting - -h A , - " "

8 ~ I I I I ~ ? ? I 7 4 6 8 10 12 14 16 16

Pressure differential, P2 - PI, psig

u - I I I I 1 I I I I I 1 I I I I I I I I I 0 1 2 3 4 5 6 7 8 9 1 0 1 1 0 1 2 3 4 5 6 7 8 9 IO 11

Pressure differential, Pz - PI, NIcm2 gage

la1 Torque. lb l Static leakage rate,

Figure 13. - Seal starting torques and gas leakage rate as function of pressure differential. Lip material. glass-filled poly tetrafluoroeiliylenc; sealed gas. nitrcqen; new seal.

: K t a r t i n g torquefor acceleration 1560 standard,,:;_----*­

c 1200 c u cmlmin --to 120 ftlsec ._ --. pol$etrafluoro­136.6 mlsec) p; E

6 ethylene lip seals z 20 Torque lhand-

3" 6 turned soeedl E

-Before r u n

3 Starting ~ 0

#-

P Running >', Im* ? - D u r i m r u n w eL ,f' at 120 ftlsec

i '36 "" .aD m A ,/ After

200

I I 0 2 4 6 8 1 0 1 2 1 4

Pressure differential, P2 - P1, psig

~ l l l l l l l I 1 I I I I I I 0 1 2 3 4 5 6 7 6 9 0 1 2 3 4 5 6 7 8 9

Pressure diflerential. P2 - P1, Nkm2 gage

la1 Torque. lb l Leakage rate.

Figure 14. - Seal starting torquer and gas leakage rate as function of pressure differential. Lip material. graphite-filled polytetrafluoroethylene; sealed gas, nitrogen; new seal.

12

shown in figure 12(b) shows that the dynamic leakage is s imilar to the static leakage at seal pressures to 6 psig (4. 1N/cm 2 gage). Dynamic leakage was not determined for

2 pressures greater than 6 psig (4.1 N/cm gage). Seal with glass-filled polytetrafluoroethylene lip. - The start ing torque values of the

glass-filled polytetrafluoroethylene seals are shown in figure 13(a). In comparison with the butyl rubber lip seal, the glass-filled polytetrafluoroethylene seal does not exhibit a lip-opening pressure (before running) in the differential p ressure range of 0 to 14 psig (0 to 9.6 N/cm 2 gage). Also, for any given pressure, the leakage rates are lower (fig. 13(b)) than that for the seal with a butyl rubber lip. These lower rates may be at­tributed to the greater bending stiffness of the filled polytetrafluoroethylene lip. Af t e r running, the leakage rate was significantly greater than that before running; this in­crease was attributed to the 0.014-inch (0.035-cm) diametral wear. Two other runs with new seals ended in failure after 15 seconds of operation; thus this seal design had high failure probability under these conditions of operation.

Seal with graphite-filled polytetrafluoroethylene lip. - The start ing torques of the graphite-filled polytetrafluoroethylene seals are shown in figure 14(a). Comparison with the torques for the butyl rubber seal (fig. 12(a)) shows that the pressure does not unload the lip contact pressure as readily, and, therefore, the leakage is less (fig. 14(b)); the leakage curve of the butyl rubber seal (fig. 12(b)) is included for comparison in figure 14(b). The seal lip-opening pressure occurred between 10 and 12 psig (6.9 to 8 . 3 N/cm 2

gage) before running. After the run, the seal did not open in the range of 0 to 14 psig (9.6 N/cm 2 gage), and the leakage rates were the lowest of the three seal designs eval­uated. The change in diameter (permanent set plus wear) was 0.047 inch (0. 119 cm). In two other runs with new seals, the leakage and torque results were s imilar to that shown in figure 14. One of these runs was made without lubrication. In general, the graphite-filled polytetrafluoroethylene seal exhibited good wear and the lowest leakage rate of the three types of lip seals evaluated. (This evaluation is based on the total de­signs and not materials alone. )

The data suggest that the seal design with a graphite-filled polytetrafluoroethylene lip may be useful for very low-leakage sealing against a small pressure differential (to 8 psig or 5 .5 N/cm 2 gage) applied so as to reduce the lip contact force. However, long-term storage and running, in which cold flow of the lip material is important, has not been evaluated in this study.

SUMMARY OF RESULTS

Lip seals of various types were run to peripheral velocities of 120 feet per second (36.6 m/sec). Torque and lip temperature were continuously recorded during short­

13

t e rm (to 600 sec) operation with zero pressure differential across the sea Starting torques and leakage rates were checked at pressure differentials to 14 psig (9.6 N/cm 2

gage). The pertinent resul ts apply to the seal design as a whole and not solely to lip materials.

The following results were obtained with respect to the sea l running torque: 1. The seal with the butyl rubber lip exhibited a prominent wear-in torque peak.

This wear-in torque was postulated to be associated with l ip shrinkage due to heating (Gow-Joule effect). This lip shrinkage can lead to sudden failure at the shaft velocities employed, but gradual wear -in decreased the probability of failure.

2. Seals with glass -filled polytetrafluoroethylene lips also exhibited a wear -in process and high failure probability under some of the test conditions.

3 , Seals with graphite -filled polytetrafluoroethylene lips had er ra t ic and high wear-in torques that were attributed to the high initial l ip contact force characteristic of this sea l type.

4. In general, lip-seal running torques (after the wear-in process) were s imilar to that for segmented-carbon shaft-riding seals.

The following results were obtained with respect to leakage and wear: 5. The seal with a graphite-filled polytetrafluoroethylene lip exhibited the lowest

leakage rate and good wear characteristics. 6. Dynamic gas leakage was s imilar to the static leakage for the seal designs with

butyl rubber lips and graphite-filled polytetrafluoroethylene lips.

Lewis Research Center, National Aeronautics and Space Administration,

Cleveland, Ohio, March 5, 1969, 120-27-04-21-22.

REFERENCES

1. Jagger, E. T. : Study of .the Lubrication of Synthetic Rubber Rotary Shaft Seals. Conference on Lubrication and Wear, Inst. Mech. Eng., London, Oct. 1957, pp. 409-415.

2. Hamilton, D. B. ; Walowit, J. A. ; and Allen, C. M. : A Theory of Lubrication by Microirregularities. J. Basic Eng., vol. 88, no. 1, Mar. 1966, pp. 177-185.

3. Dega, R. L. : Zero Leakage: Results of an Advanced Lip Seal Technology. Paper 67-WA/L~b-l1, ASME, NOV. 1967.

14 NASA-Langley, 1969 - 15 E-4885

.- ............ . .