Sie sind auf Seite 1von 8

ASL E TRAXS ACTIO X S

7, 24-3 1 (1964 )

Friction of Greases and Grease Components


during Boundary Lubrication
By DO UGLAS GOD F RE YI

Downloaded by [75.149.200.233] at 07:54 25 July 2012

.Ifeasurements have been made of th e coefficient of friction of a steel pin sliding on a steel ring
lub ricated w it h thr ee com me rcial greases, various ex periment al greases or pastes, and ot her com ponent s of greases. A fter break- in at room tem perature, heat was applied w hich redu ced t he friction
of t he greases to about half t heir room tem perature value. Th e friction increased on cooling. Th e
opposit e effect w as observed wit h base oil. Th e ph enom enon w as dem onstrated in a ball t hrust
bearing operating und er boundary condition s as a reduction in to rque w it h heatin g and an increase
in torque on cooling. Th e reduction in friction w it h heatin g was [ound t o be du e to th e soap
t hick ener. Pastes of CaC0 3 , fil me d silica, and clay in oil, t he base oil, and octadecane did not
calise a redu cti on in friction. It w as concluded t hat grease t hick eners infl uence th e friction of steel
sliding on steel because of th e for matio n of films on t he surjace of t he m etal .

placed on top of the upper race fixture. The torqu e was


plotte d during rotation at 10 rpm. The balls and raceways
were smeared with commercial sodium grease, and a dab
of grease plac ed in each concavity of the cage.

Intr od uction

A COM M ON concept is that the soap in a grease does not


affect boundary friction . Bondi et at. (1) present ed data
showing no difference in coefficient of friction betwee n a
soda base grease and its base oil. However, experience
and recent papers (2, 3) suggest that gelling agents and
fillers do produce an effect during sliding. The purpose of
the experiments described her e was to det ermine the effect
of heat , the most important operational variable, on the
fric tion of steel sliding on steel lubricated with greases.
The data will show that during heating under conditions
of boundary lubric ati on, soap thick eners reduce the
friction.

LOrD
RESTRAINING
FORCE - - - - - - .

Exp eri ment a l

Tribo meter. T he apparatus consists of a steel pin with


a hemispherical tip sliding again st the rim of a horizont ally mounted rot ating steel ring. The load of 1 kg
and sliding velocity of 1 cm per second were chosen to
minimize hydrodyn amic lubricati on and prom ote boundary conditions. A schematic of the elements of the
tr ibometer and a photograp h is shown in Fig. 1. Detail s
of specimens, specimen pr epar ati on, and procedure are
given in App endix r.
Thrust bearing. Seven run s were made in which the
torque of a ball thrust bearing was measured during
heating. A commercial bea ring with a 1/ 6-inch bore and
i -inch balls was mounted horizontally on the rotating
tabl e above th e heater. The upper race was a tt ached to
a fixture with a torque arm and restrained from turning
by th e strain gage. A dead weight load of 6600 gm was

F I G.

1.

Tri borneter-schernati c of sliding elements and photo-

graph.

Detail s of the lubricants used ar e given in Appendix

II.
P resented as a n America n Society of Lu bri ca tion Engineers
pap er at th e Lu br ication Conference held in Rochester , New York ,
October 15-1 7, 1963.
I Californ ia Resea rch Cor pora tion, Richm ond , California.

Results and discussion

R ecorded data . Figure 2 is a tracing of the actual dat a


recorded for sodium stear ate in 480 neutral run on th e

24

25

Friction of Greases and Grease Components durin g Boundary Lubrication

200

I-

a:

15 0

LL 0.3

HEATER

LL

ON

02

10 0

b 0IF...---.....----:~~&i.IlJ~~IIIlIJi_~
. . . oJII!AlII...~---~- ...-....,.....-:.~_;j
.- - - - - - --- -- - -~- - '~

oL--

.-

--;:';,---50

NO . Of"

Downloaded by [75.149.200.233] at 07:54 25 July 2012

F IG. 2.

---,-:!-::-

100

150

~=__--~

200

a:

a.
~

i3

W-

a:

:>

I-

50

REVOLUTIONS

R ecorded data to show nature of frictio n sodium stea rate in 480 neutral (st eel vs. stee l; load , 1 kg ; speed, 1 em / sec: room air)

pin on ring tribometer. It is presented mainly to show


the nature of the tracin gs and that in most cases friction
was not steady . The fricti on force varied within each
revolution, numb er of revolution s, and temperature. An
analysis of the data in Fig. 2 showed that the coefficient
of friction on starting was 0.16, and durin g the first few
revolutions kinetic friction vari ed from 0.17 to 0.12.
After 50 revolution s this variation had reduced to values
between 0.11 and 0.12. Every lubricant-material combination tested showed these break-in characteristics .
R epeatabilit y . Repeatability of coefficient of friction
varied with the lubricant used and with the number of
revolutions: between a and 20 revolution s the repea tability was poor, 0.01 ; between 20 and 50 revolutions repeat ability was 0.005. During heatin g and
cooling repeat abilit y was 0.0 1 except for two dry soap
powders which decomposed.
Commercial greases

Tribom eter. All the data for the commercial grease are
plotted in Fig. 3. T he sodium soap grease, the EP grease,
as well as the calcium complex (salt and soap ) grease,
all showed a reduction in fricti on with heatin g and an
I ~ RUN

2 .!:!P RUN

increase in friction with cooling. Th e reason for this


change was not clear. The EP grease was liquid at about
100 C, but friction continued to decrease on up to 200 C.
The lower friction at higher temperatures is certainly a
desirable characteristic for grease. Another observation
from the curves was that the calcium complex grease and
the EP grease produce a faster and more well-defined
break-in than does the sodium grease. No correlation was
found for coefficient of friction and worked penetrations.
Thrust bearing. The sodium grease was used to
lubricate the ball thru st bearin g operating at 10 rpm and
6600 gm load. The torque decreased with heating from a
stable value after break-in of 1.5 gm at room temperature
to 0.2 gm at 200 C and increased on cooling to the stable
1.5 gm value at room temperature. Th e experiment was
repeated five times with new bearings and fresh grease
with identic al results. A 480 neutral base oil permitted
torque to rise with heating. It was concluded that the
reduction in friction with heat was not unique to the pin
on disk apparatus, but applicable to bearin gs operating
under boundary or partial boundary lubrication.
The following experiments conducted on the pin on
disk tribometer were designed to determine which part
of the grease was responsible for the reducti on and the
mechanism of action.

Q
to
ii

...
...o

Variation of cat ion of soap

.l 5 ~--_j--------+-----'------f

.10 f----t--....:::.-~~~~_+___a_~~:;.._~-_l

t-

'"

~ .051---

'"uo

---+- - - - - - - - + - - - --

BREAK - IN

-----l

100

ROOM

COOLING

HEATIN G

50

100

15 0

200

250

A series of metal steara te greases was used to dete rmine


if the fricti on reduction with heating correlated with
physical properties of the grease or with phase changes
reported by Hotten and Birdsall (4) . Figure 4 presents
curves consisting of the average lines only" for all the
metal stearate greases used and shows first of all that
the decrease of friction with heating and the increase with
cooling is common to all the metal steara tes.
Table 1 shows that no correlation exists with coefficient
of frict ion at room temperature and the penetration hard ness of the stearate grease. Also ther e is no correlation of

NO. Of REVOL UTIONS

F IG. 3. Fricti on of steel on stee l lubrica t ed wit h co mme rcia l


greases (medium gra de) (load , 1 kg ; speed , 1 em /sec : roo m air ) .

2 F rom this poin t on only average lin es will be p rese nted in


th e figures. A ll ru ns were repe a table within th e limits sta te d .

26

D OUGLAS G ODF REY


.2 0

.20

SODIUM

PQTASSIUM

.15

.15

"-

. 10

",--

<; ~

.0 5

.10

'----- .>

.2 0

.0 5

.20

MAGNESIUM

CALCIUM
.15

I\.

.........

.10

I-

E.0 5

.10

<, .>:

--

\
I

.0 5

I..

Downloaded by [75.149.200.233] at 07:54 25 July 2012

o
.2 0

I- .2 0

BAR IUM

I.J

ALUMINUM
.1 5

::: .15

1"-

I.J

.10

./\

<,

./

.10

>
I

.0 5

.0 5

B REAK
IN

roo

ROOM

COOLi N..

TEMPERATUA~.c

LITHIUM

.15

.0 5

HEATI NG

.2 0

.1 0

50

100

150

roo
200

250

NO. OF' REVOLUTIONS

\..

.:>
' -~

~:'==~:::~~==:'I:==::::::===~1
COOL IN..

HEATING

ROOM

100

I
100

200

TEMPERATURE. C

50

100

150

200

250

NO. OF' REVOLUTIONS

F IG. 4. Fri ction of steel on steel lubricat ed with meta l steara tes in 480 neutral (load , 1 kg ; speed, 1 em/ sec ; room air )

T AB L E 2

Comparison of Coefficient of Friction and Solubilit y of


M etal St earates in Oil

T ABL E 1

Comparison of Coefficient of Friction and Penetration of


M etal Stearate Greases

Cation

Coef. of friet.
afte r room temp .
break-in

Penetr ati on
P 60

Stear at e

Temp.
C

Coef.
of frict.

Sodium
Calcium
Lithium
P ot assium
Barium
Ma gnesium
Aluminu m

0.12
0.11
0.11
0.11
0.10
0.10
0.09

443
509
313
361
509
284
509

Sodium
Lithium
Lithium
Sodium
Aluminum
Aluminum
Calcium

100
100
200
200
100
200
200

0.13
0.11
0.09
0.09
0.08
0.07
0.07

Solubility
of steara tes
in oil ;
from Ref. (4)
% by wt.
0
0
6
10
25-50

In finite
Infinit e

27

Friction of Greases and Grease Components durin g Boundary Lubric ation


frict ion at 100 or 200 C with known melting (dropping)
point s, or free acid content or the presence of excess base.
In the case of magnesium and aluminum stearate, the
greases were completely liquid at 130 C and above, but
there was no cha nge in the friction curve at the melting
point. F riction continued to decrease when the lubricant
became liquid .
Reference (4) shows a great difference in the solubility
of different metal stearates in oil at various temperatures.
Perhaps the reduction in fricti on of greases with heatin g
is a result of increasing quantity of dissolved soap in the
oil. Ta ble 2 suggests that there is a correlation between
friction and soap solubility .

.20

...:
z

4 8 0 NEUTRAL" ,

52
....

.15

1\

...

ii:

...o

.10

....

Downloaded by [75.149.200.233] at 07:54 25 July 2012

...
...o

ii:

r-.

51

52
....

8%

II

.15

./

~----,

....z

- ----

W .IO

--

/:==;;:--

30 %BENTONITE )

......

' , ~ 35% C .

co,

" ::'~----':::'-:'?

<,

./

\r... 35

% C. (OH)2

8 .0 5
COOLING

HEATING

ROOM

100

200

TE M P ERAT URE .

50

~~

"

.I'

/~ECOMP05ED

SOAP

SODIUM STEARATE../

.0 5

BREAK -IN

HEATING

COOLING

II

ROOM

100

200

10 0

T EM PE R ATURE . C
50

100

150

200

250

100

150

100

20 0

FIG. 6. F riction of steel on steel lubricat ed wit h base oil and


with dry soap powders (load, 1 kg ; speed, 1 em/sec; room air).

same as the whole greases. Th erefore, the reduction in


friction with whole grease is attributable to the soap alone.
Octadecanoic (stear ic) acid produced a similar and nearly
reversible reducti on in friction with heating. (T he curve
is presented lat er. )
Investigation of mechanism

BREAK-IN

/'

'\

NO. OF REVOLUTIONS

.25

--

F umed silica, bentonit e, calcium carb onat e, and calcium hydroxide were each stirred into 480 neutral to
make pastes. Th e results of fricti on measurements are

~'::~-----

./

CA L C IUM 5TEARATE-U-.....
:

Inorganic th ickeners and fillers

....,: .2 0

...!.

II

2 50

NO. OF REVOLUTIONS

Since octadecan oic acid was the simplest mat erial


which produced reduced friction on heatin g, run s were
made with octadecane and octadecane derivatives with
various degrees of surface activity . Th e friction-t emperature curves for octadecane, octadecanol, octadecylamine, and octadecanoic acid are presented in Fig. 7. All
these mat erials were liquid above 75 C. Th e results show
that the least surface activ e mat erial , octadecane, failed
to lubricate when heat was applied. The increa se in
friction with heat is attributed to greater amounts of
metal to metal contact resultin g from a lower energy of
adsorption than that at lower temperatures. Th e increase

FIG. 5. Fr iction of steel on steel lubricat ed with pastes of


inorganic thicke ners in 480 neut ral (load, 1 kg; speed, 1 em/ sec;
roo m air).

presented in Fig. 5. Th e graphs show that durin g heating,


fumed silica and CaC0 3 increased friction , Ca( OHh
reduced friction , and bentonite had only a small and
irr egular effect.
These results show that some inorganic components
used in greases do not reduce fricti on durin g heating and
so the phenomenon is not common to all combinations
of oil and solids.
Friction of grease components

The base oil and two dry soap powders were tested
separately using the same procedure as that used for the
greases. The result s are presented in Fig. 6, which shows
that fricti on of the base oil increased with heatin g and
decreased with cooling, which is the opposite effect of
the whole grease. Figure 6 also shows that dry soap
powders caused a reducti on in friction with heat , the

.25r-- - ---,r-- - - - - - --,- - - - - - - -,

rtr:\
-~~
v ----:::::;:::::=:::::::::-t---,
\

~~

.15 t -..;::.".-

- ir --

\
.10

:5

"\l--:::-.

OCTADECYLA~~E

- t\ --

II .0 5 t - --

- ir --

BREAK-IN

--j

yl1\~.--/'

-7

\ '--

- - ~ - - - -----

- --

HEATING

-y{:::::,

OCTADECAN OIC ACID


- t-

--j

COOLING

ROOM

OCTADECANO L "
'"..----

~-;r-

.) OCT ADECANE

10 0

50

100

ISO

200

250

NO . OF REVOL UT IONS

FIG. 7. F riction of steel on steel lubricat ed with octadeca ne and


octadeca ne derivat ives (load, 1 kg ; speed, 1 em/sec; room air) .

28

D OUGL AS G ODFR EY

Downloaded by [75.149.200.233] at 07:54 25 July 2012

in friction of the 480 neutral oil with increa sing tempera ture is also a tt ributed to a decreasing ener gy of adsorption. Fi gure 7 shows that the friction of octadecane
decreases above 175 C and durin g cooling . This reduction
is attributed to the formation of oxida tion products (5) .
Octadecanol and octadecylamine form adsorbed films
which consist of closely packed vertically orient ed molecules. The results show that both of th ese materials
pro duced a decrease in friction up to about 100 C and an
incre ase in friction between 100 C and 200 C. This result
is consistent with th e th eory that it takes considerable
energy or hea t to desorb these polar molecules (6) .
Oct adecanoic acid provides the lowest friction at 200 C.
The momentary reduction in friction near th e melting
point is th e sa me as reported in Ref. (7) . T his
material is expected to react with the metal surface to
form an iron stearate film which would require more heat
.2 0

z
0

i=

.15

u,

-t-----\

u,
0 .10

t-

Z
III

u, .0 5

Jr/

-:

~A55 V5 GL A SS

sr>:
y
-----

"-

~:;~:

---" r--~

V5

ST EEL

LIQUID

III

BREA K- IN

HEAT ING

I
ROOM

10 0

I
200

T E M P E RAT UR E,

50

I
100

COOLI N G

100

15 0

200

2 50

N O. OF REVOLU TIONS

F IG. 8. Fricti on of mat erial s lu bricat ed with octadeca noic


(stea ric ) aci d (load , 1 kg ; speed, 1 em / sec ; roo m a ir) .

than any of the others to be desorbed . This theory is


ba sed on the assumption that the films are not durable
and need constant replacement on the sliding surface.
With glass sliding on glass, no reaction would be expected. The curves are shown in Fig. 8, where it can be
seen that liquid octa deca noic acid did not lubricate the
relatively inert glass . The reason for th e low fricti on
during break-in is not clear. The results with glass are
consistent with the theo ry that chemical reaction occurred between octadeca noic acid and steel to form a
chemi sorb ed film.
Tabor (8) ha s shown that under boundary lubricating
conditions that the coefficient of friction is the ra tio of
critical shear stress of the int erface material and th e
plas tic yield pressur e of th e underlyin g metal. Th e shear
st ress of th e film on th e surfa ce, th erefore, determines the
coefficient of friction. The films may be considered to be
solid soap when th e temperature is below the meltin g
point and adsorb ed films above th e melting point.
R eference (9) showed th at the shear strength of vari ous
solid soaps between anvil s was less at 100 C than it was
a t room temperature. These two references together sug-

gest that the reduced fricti on reported here for greases


and soap below their melting point was due to reduced
shear strength of the soap film with heat. The experimental work reported here in shows that above the melti ng
point friction of steel on steel decreases due to forma tion of chemisorbed surface films. In addition lower
friction was observed for soaps with grea ter solubility
in the base oil.
Conclusions

Measur ements of th e frict ion of steel sliding on steel


lubricated with greases under conditions of boundary
lubrication have shown :
1. The kind of thickener in a 480 neutral base oil
determines the level of friction .
2. Friction decreases during heating and incr eases
during cooling with soap powders, soap thickened oils, a
paste of Ca( OHh and oil, octadecanol, octadecylamine,
and octadecan oic acid.
3. The torque of a ball thrust bearing opera ted a t 10
rpm and 6600 gm load and lubricated with a commercial
sodium grease decr eased with heating and increased on
cooling. With a base oil only, the oppo site effect was
observed.
4. Friction increases on heating and decr eases on cooling with pa stes of fumed silica and of CaC 0 3 in oil, the
base oil alone, and octadecane.
S. The literature suggested that the friction of greases
decr eases with heat ing and increa ses on cooling becau se
the adsorbed soap decreases in shear strength with heating and incr eases with cooling. The experimen ts reported
her ein with materials of varyin g surface activity suggested that above th e meltin g point of the grease, friction
decrea ses with heating due to formation of sur face films.
The experiments also showed a correlation of coefficient
of friction and the solubility of soap in the base oil.
REFERENCES
1. BONDI, A., et al. " Develo pme nts in th e Fi eld of Soda Base
Gr eases," Proc. 3rd W orld Petrol. Congr. Sectio n VII , 1951,
pp . 3 73.
2. M ILNE, A. A., a nd COOKE, W . L ., " On t he In fluen ce of Grease
St ructu re on Bounda ry L ub rica ti on ," Proc. Conj. Lub rication
W ear, I nst . M ech. E ngs. 355-358 (1957) .
3. VOLO, M . J ., a nd VOLO, R . D ., " T he Nature of Soa p- T hic ke ned
Lubrica ting Gr eases," J . I nst . Petrol. 38, 155- 163 (1952) .
4. H OTTEN, B. W ., a nd BIRDSALL, D. H. , " P hase Ch an ges in
Soap Oil Disp ersion s," Ind . En g. Chern. 47, 447-45 1 (1955) .
5. BOWDEN, F . P ., L EBEN, L ., and TABOR, D ., "The Influence
of T emperature on th e Sta bility of a Mineral Oil," Tra ns.
Faraday Soc. 35, 900-905 (1939) .

6 . ZISMAN, W. A., a nd M URPHY, C . M ., " Trends in Lu b ricants,


Ad van ces in P et rol eum Che mistry and R efinery " ( K . A .
K obe and J . J . M cK ett a , Jr., eds .) , Vol. 2, pp . 53-57 . Wiley
(I n terscience), New York , 1959.
7. BARWELL, F . T ., a nd M ILNE, A. A., " T he Lu br ica ti on of
R ou gh Stee l Su rfa ces by a Ser ies of M et alli c Soa ps, " B rit . J.
A ppl. Ph ys. Su ppl. No . 1, Phys. Lubr. 44-48 ( 1951).
8. T ABOR; D ., " J u nct io n Growth in M et alli c F ricti on : The

Friction of Greases and Grease Component s durin g Boundary Lubrication


Role of Combined Stresses and Surface Contamination ,"
Pr oc. R oy . S oc. (L ondon ) Ser. A 251, 378-393 (1959) .
9. WHITE, J. R ., "Friction al Properti es of Soaps at Hi gh P ressures ," Lubr. E ngr . 10, 340-344 ( 1954).
10. GOODRICH, J. E ., a nd BURKE, J. J., "Oxida tion of Lubricating
Gr eases," N LG I Sp ok esman 24, 397-404 (1961) .

The rotating specimens were commercial test cups used


in several lubricant performance testin g machine s. The
cups are 0.5 inch high, with an outside diameter of
1.9375 inches. The narrow edge, approximately 0.09
inch wide, was used as the sliding surface. The material
is SAE 4615 steel, case hardened to approximately 60
Rockwell C.
For the investi gation of mechani sm, some of th e
rider s were boro-silicate glass of the same general shape
as the metal rider s. The rotating specimens were borosilicat e glass disks 2 inches in diameter and inch thick .
Cleaning. Th e riders were washed ultrasonically with
hexanes and rinsed with pure anhydrous ether, then
quickly dried with a blast of clean warm air.
The rings were washed with hexanes ; then the narrow
edge was abraded on 240-grit emery paper, using figure
eight motions, with frequen t rotation, to give a finish
with random abrasion mark s. The center line average
roughness of this sliding surface is 8.5- 10 flinches measured with a 0.03-inch wavelength cutoff. The rings were
then wiped with tissue wet with hexanes until the tissue
was no longer soiled. This was followed by an ultr asonic
wash in hexanes, then the rings were rinsed with CP
anhydrous eth er and quickly dried with a blast of clean
warm air. This cleanin g procedure produced a coefficient
of friction of 0.6 for unlubricated specimens.
The glass specimens were scru bbed in sulfuric-chromic
acid cleaning solution, thor oughly rinsed with distilled
water, and dried.
All cleaned specimens were stored in a desiccator and
subsequently handled with grease-free tongs or tweezers.

Append ix I
D ETAILS OF EXP ERIMENTS ON

PIN

ON RING TRIBOMETER

Op erational Featur es

Downloaded by [75.149.200.233] at 07:54 25 July 2012

29

The load was applied by dead weights directly above


the rider. Rider and weight s were held at the end of a
balanced arm mounted on two pairs of pivot bearings.
Friction was sensed by the strain gage as a restraining
force. The output of the gage was converted electronically
to coefficient of friction and recorded. The rotating specimen is fixed in a shallow cup which is mounted on the end
of a vertical shaft passing through the base plate . The
rot ating parts ju st clear a stationa ry flat ring heater. A
geared fractional horsepower motor with electronic speed
control provided the rotation of 4.12 revolutions per
minute .
Temperature was measured with iron-con stantan thermocouples which were mounted so as to cont act lightly
th e ring near the sliding surfa ce. The thermocouples are
not shown in Fig. 1.
S pecimens. The rider s for the pin on ring run s were
rollers manufactured commercially for needle bearings.
Each rid er was 0.5 inch long, 0.078 inch in diameter, and
the radiu s of the end was equal to the diameter of the
rider. The rider s were made of SAE 50100 steel, through
hardened to 60-62 Rockwell C. The center line averag e
roughness of the riders is 1-2 flinches measured with a
0.03-inch wavelength cutoff.

Experimental Procedure

The rotating specimens were fixed to the rotating cup ,


and the lubricant was smeared over the surface approxi-

TABLE 3
Properti es of Greases and Pastes
Base oil
vis. SSU

Thickener
Classificat ion
Co mmercial ball
bearing grease
Commercial EP grease
Commercial multi-duty

Co mposition
Sodium soa p
Calcium-lead soap
Calcium complex
(soap and salt
compl ex)

0/0

(weight )
25
11

100F
500
1066

210 F
61
78

Penetr ation
Dropping Unwo rked
pt. (F)
Po

Work ed
P GO

Steari c acid
content,
0/0 (weight )

450
235

232
235

280
285

o(excess base)

324

333

o(excess base)

411
509+
509+
246
223
227
409

443
509+
509+
313
361
284
509+

0.25
0.43
0.27
0.15
2.8

28

500

61

535

Experim ent al
Experimental
Experimenta l
E xperim ent al
Experimental
Experim ent al
Experimental

Sodium stearate
Calcium stea ra te
Barium stea ra te
Lithium stea rate
Pot assium steara te
Magnesium stea ra te
Aluminum stea ra te

15.9
15.6
17.9
15.2
16.6
15.4
15.9

512
512
512
512
512
512
513

62
62
62
62
62
62
62

374
406
< 104
419
446
< 104
< 104

Past e
P ast e
P ast e
Past e

Microni c CaC Oa
Ca( OH)2
Fumed silica
Bentonite

35
35
8
30

513
513
513
513

62
62
62
62

1.2

0
O(excess base)
0
0

30

DOUGLAS G ODFREY

Downloaded by [75.149.200.233] at 07:54 25 July 2012

mately 1 mm deep. The rider was placed in the collet


at the end of the arm , and the 1 kg weight was added
above the rider.
After calibration of the strain gage and thermocouples ,
rotation was started at room temperature. After 50
revolutions the heater was turned on. On reaching 200 C,
the heat was turned off, and the system was cooled with
a gentle blast of air directed toward the base to cause
the cooling rate to be near the heating rate. The coefficient of friction was recorded continuously.
Coefficient of friction was recorded on a chart. In the
case of variable friction within one revolution , the friction
was read at a point one-third the distance down from
maximum value. All experiments were run at least twice
and the value reported as an arithmetic average. The
coefficient of friction was plotted as a function of number
of revolutions of the ring and of temperature on the same
graph.
Appendix II
D ETAILS OF L UBRICANTS U SED

a metal cation. The hydroxide was used for sodium,


potassium , magnesium, barium, calcium, and lithium and
the isopropoxide (in water ) was used for aluminum. The
mixtures were heated sufficiently to complete the formation of the soap, then heated to 400 F to dehydrate the
grease. The grease was then cooled and milled.
The pastes were made by simply stirring powders into
the base oil at room temperature.
The base oil for the metal stearate greases and the
paste s was 480 neutral , a solvent-refined material from
paraffinic base crude with a viscosity of 513 SSU at
100 F.

Soap Powders
The calcium stearate used as a dry powder was a
commercial material of technical grade. The sodium
stearate was a laboratory preparation made by reacting
stearic acid and sodium methylate in the presence of
methanol under conditions which excluded moisture.
Excess stearic acid was removed by ether extraction.

Greases and Pastes

Octadecane and Derivates

The properties of the greases and pastes used are given


in Table 3. The experimental greases were the same as
those described in Ref. (10 ). The greases were made by
melting stearic acid in the oil, addin g one equivalent of

The octadecane, mp 28 C; the octadecanol , mp 59 C;


the octadecylamine, mp 69 C; and the octadecanic acid
(stearic acid ) , mp 69.7 C, were commercially pure
materials used as received.

DISCUSSION
L.

C. BRu NsTRuM ( R esearch and Dev elopm ent


A merican Oil Compan y, Whitin g, In diana ) :

Department ,

In "A Sur vey of Journal Bearin g Lit erature" (A l) Full er makes


several conclusions which ar e pertin ent to this pap er. His suggestions for needed research includ e th e followin g sta tement :

" It a ppears that grease lubricated bearin gs have been only


explored rather perfun ctorily, and that it would be very useful to
determine th e limit s of load- carr yin g and oth er per formance factors that could be a tta ined by the use of grease both in hydr ody namic and hydrostati c applications."
Becau se the same sta tement may be mad e ab out the case of
boundary lubric ati on, it is gra tifying that the author has tak en
the pains to resolve the moot question of wheth er the soap in a
grease contributes to lubric ati on . Under the conditions selected by
th e author, the soa p obviously contributes; the difference in
friction is tw ofold . Furthermore, the author has shown that friction is relat ed to soap solubility, as influen ce by the cation. Thi s
is indeed substantial proo f that the chemistr y of soap in grease
contributes to lubric ation.
Full er also comments as follows on the contribution of grease
viscosity to load -carryin g ability of a bearing:
" Grease is one of the more viscous of the common lubric ants
that is used commercially. Frequentl y grease is used to provide
good boundary lubrication . However, it has been shown that
when properly supplied to a journal bearin g, a true hydr odynami c
action may be developed. Now, all other thin gs being equal, the
load- carryin g capacity of a bearin g depends directly upon the
viscosity of the lub ricant. .. . Thus, it might prop erly be expected

that a journ al bearin g running hyd rodynamically on a grease


would have a very great unit load-c arr yin g capacity ."
Although quantitati ve relati onships are still inad equate, viscometric considera tions ( A2) and bearing tests ( A3, A 4 ) tend to
confirm the relation between load-carryin g ability of a bearin g
and grease viscosity. Thus it app ears worthwhile to explore fur ther the author's finding that heating reduced the torqu e in a
thru st bearin g by 7.S-fold. Certainly the chemistr y of the grease
was effective here, just as it was on the tribometer , but does
chemistry alone account for the reducti on observed?
Several conditions menti oned in the pap er bear on this point.
For the tribometer experiments, the author, properl y for his purpose, selected conditions that would minimize if not eliminate
grease viscosity : bound ar y lubrication condit ions and geometry
that minimizes the flow of grease. Thu s in Table 1 we find that
grease viscosity, as jud ged by penetration, has been successfully
cancelled out of the experiment; there was no correlation between friction and penet ration.
On the oth er hand, conditions in the thru st-bearin g test are
quit e different. Th e load is low , there is rollin g ra ther than
sliding, and both the geometry and mor e abundant grease charge
favor grease flow. In this case the 7.S-fold reduction in torque on
heatin g might be attributed lar gely to the concommita nt reduction in grease viscosity. Some recent work ( A5) would predict
viscosity reductions of S- to lO-fold, dependin g on shear rat e and
temperature.
Very likely, most commercial applications of grease operate
under conditions of mixed boun dar y and hydrodynami c lubrica cation and both the ch"emical and physical properties of lubricant

Fnction 01 Greases and Grease Component s during Boundary Lubrication


affect tor qu e. Th e pr esent pa per is a step in th e direction of
separa ti ng th ese effects.
REFERENCES
F ULLER , D UDLEY D ., "A Survey of Jo urnal Beari ng Lit era ture," pp. 92 and 4 7. Th e American Society of Lubrication
En gineers, Chicago, Illin ois, 1958.
A2. SISKO, A. W., AND BRUNSTRU1\I , L . C., "Applica tion of a New
Grease E qua tion to Velocity P rofiles a nd Lu bricati on P rob lems," L ubrication E ng. 18, 307 (1962).
A3. BRADFORD, L. J ., BARBER, E . M. , AND M UENGER, J . R., " Grease
Lu bri ca tion with Plain J ournal Bearings," pr esented a t
ASME -ASL E Lubrication Confe rence, Bost on , October, 1960.
A-/. CHAKRABARTI, R. K., AND H ARK ER, R. J ., " Fr ictiona l Resista nce of a Radi ally Load ed J ournal Bearing wit h Grease
Lubricati on ," Lubricat ion En g. 16, 274 (196 0) .
A5 . BRUNSTRU1\I, L. C., BORG, A. C., AND SISKO, A. W., "Master
Cur ves for Grease Flow ," N LG I Sp ok esman XXVI , 6
(1962).

Downloaded by [75.149.200.233] at 07:54 25 July 2012

A1 .

J . PANZER ( Ess o R esearch & Engineering Co ., Li nden, New


J ersey ) :
M r. God frey has done an excellent bit of wo rk in studyi ng the
effect of grease components on friction during boun dary lubrica tio n. His caref ully perform ed exper iments provide some valua ble
insight s int o the beha vio r of greases. It was suggested by M r.
Godfrey th at th e reduction in frictio n with soa p-thickened greases
on heating is du e to the form at ion and behav ior of soa p films
whose shea r strength and solubility change with temp erature. Thi s
is based on th e cha nge in fricti on as th e temp erature is va ried.
H owever , whi le th e temp erature is cha nging, so is the number of
revo lutio ns th rou gh whi ch the slider is mov ing. Isn 't it possible
tha t the observed frictio na l cha nges could occur at a consta nt
tempera ture merely beca use the repeated shea ring of th e grease
causes chemical and physical cha nges in the soap? Cha nges in
thickener pa rti cle size, for exa mp le, could easily accou nt for the

31

behavi or repor ted by Mr. God frey . It would be of interest to


determine th e va riation in friction a t several constant temperatures
while th e slider mo ves th ro ugh 200 or mor e revo lutio ns. T he results of such exper iments wo uld clarify the mecha nism of the
above phenom ena still fur ther .
AUTHORS' CLOSURE:
Th e autho r wishes to tha nk Mr. Brunstrum and M r. Pan zer for
th eir discerni ng comments.
Mr. Bru nstru m's theo ry tha t the thickener could pro mote hy drodyna mic lubri ca tion seems very plau sible. E xperim ent s were
conducted on the tribometer in an effor t to evalua te thi s th eor y,
but, und er the conditio ns used, th e effect turned out to be ra ther
sma ll. All th e conditio ns were the sa me as th ose describ ed in th e
pap er except th at the speed was increased increm entally to 220 cm
per second ( 1000 rpm ) at room temperature . T he lubric a nts used
were the experimental sodium stea rate grease and its base oil, 480
Neut ral. One wou ld expect a greate r redu ction in th e coefficient
of friction with increasing speed with th e grease th an with the
base oil alone. At the highest speed th e friction of 480 Neutral
was 0 .070 0 .005 and the frictio n of the grease was 0 .055 0.00 5.
Perh ap s this sma ll difference was beca use th e a ppa rent area of
contact (the wear scar) was a sma ll circle only 0.3 mm in diameter , which suggests tha t th e sliding was in th e mixed film regio n
of lu bricati on .
In answer to Mr. Panzer 's questio n, it was shown th at pro longed run s of 2000 revol utions at room temp erature did not
change th e coefficient of friction significantly . Also, in most of
th e runs friction increased during cooling, th e final value approachin g th at measur ed befor e heating was star ted . Th us, the
mechani sm influ encin g th e fricti on durin g heating was either a
tempora ry one or tend ed to be reversible. It appears th erefor e
th at under the conditions used th e physical cha nges in th e grease
due to the num ber of revolut ions can be ruled out. Reference (2)
is a study of the effect on friction resulting fro m cha nges in a
grease caused by sta tic heatin g for 2 hour s.

Das könnte Ihnen auch gefallen