Downloaded from SAE International by Columbia Univ, Wednesday, April 15, 2020

Causes and Effects of Sludge

Formation in Motor Oils
By D. P. Barnard,l E. R. Barnard,l T. H. Rogers,l
B. H. Shoemaker land R. E. Wilkin 2

S INCE it appeared reasonably certain that the

key to the presence of sludge deposits, re~
from quite large proportions of asphaltic material which un·
doubtedl y did come from the oil to almost negligible pro-
portions of such substances, with the major part of the de-
gardless of form, lies in the oxidation of the oil
to give the small amounts of asphaltic substances posits consisting of "blow-by" carbon, inorganic material
which are inevitably present, the authors pre. and, occasionally, water. In appearance, aside from their in-
variably being black, these deposits differ quite widely, the
pared the following paper to summarize and range extending from a thin hard "lacquer" to a soft putty-
appraise the evidence on this point and to present like mass of considerable bulk. Frequently, deposits in the
a method for measuring oil stability. form of gelatinous masses are encountered and, in fact,
such accumulations are usually responsible for the clogging
After discussing sludge, its composition and of screens and filters. In such cases the non-fluid accumu-
effects, engine observations of oil oxidation are lation generally consists largely of oil in which carbon and
stated and the conclusions are reached: (a) that inorganic material are held in suspension by the presence of
appreciable oxidation of each of the oils occurred some small amount of asphaltic matter. The results aris-
in a 50·hr. run; that (b), variation in engine out. ing from the presence of these deposits are almost as varied
put markedly influenced oxidation of the oils as their consistencies and range from accumulations in oil
pans and other relatively stagnant locations which, when
while variation in sump temperatures did not; an engine is dismantled, appear quite objectionable but which
and (c), that even with the wide difference in the actually do no harm, to deposits in such vulnerable spots as
character of the oils used, chloroform insoluble ring grooves, oil screens and filters where actual interfer-
material ("carbon") did not vary greatly but ence with engine operation results. It appears reasonably
appeared to be determined mainly by engine certain that the key to the presence of sludge deposits, reo
condition. gardless of form, lies in the oxidation of the oil to give the
small amounts of asphaltic substances which are inevitably
A comparison of laboratory and engine tests is present. It is for the purpose of summarizing and apprais-
made and the Indiana Oxidation Test is described. ing the evidence on this point and presenting a method for
Service observations are stated and commented measuring oil stability that the following paper has been
prepared.
upon. In conclusion it is said that the indication
is that the formation of sludge in motor oils IS Sludge, Its Composition and Effects
due primarily to asphaltenes resulting from oxida- It is generally recognized, particularly by those actively
tion of the oil. engaged in engine maintenance, that sludge may appear in
a variety of forms ranging from deposits quite soft and
gelatinous in nature to those of the texture of coke. Regard-

T HE term "sludge" has been used for years to denote

virtually any deposit-except combustion-chamber car-
bon-found in an internal-combustion engine. While
it is generally understood that sludge deposits ongmate in
less of their physical characteristics the analyses. of sludge
deposits show in general very little material which can be
positively identified as having resulted from oil deterioration.
there always being some doubt as to whether the carbon
the deterioration of the lubricating oil there is no uniformity present results from "blow-by" of products of combustion.
in the constitution, or even in the appearance, of such ac- Even in the case of deposits occurring on the under side
cumulations. Analyses have shown compositions varying of piston heads, the method of their formation has not been
[This paper was presented at the International Automotive Engineering clearly understood. In spite of the fact that sludge deposits
Congress, Chicago, Sept. I, 1933.]
1 Research Laboratory, Standard Oil Co. (Ind.) may consist either almost entirely of carbon or, at the other
2 Technical Division, Standard Oil Co. (Ind.)
extreme, of unchanged oil, a certain consistency has b~ell

167 May, 1934

 Downloaded from SAE International by Columbia Univ, Wednesday, April 15, 2020

168 S. A. E. J0 URN A L
(Transactions)
noted in that they are invariably found to contain some
asphaltenes which are materials insoluble in petroleum ether
(or "hexane") but soluble in chloroform, which material is
undoubtedly the product of oil deterioration. For example, .
the results of analyses of a large number of used crankcase
oil and sludge samples yielded the values given in Table I.
Data of the type given in Table I lead at least to the sus-
picion that, even though most of a sludge may consist of
substances other than those arising from oil deterioration,
neverthel.ess such deterioration products are necessary com-
ponents If only to serve as binders and peptising agents.
~a.ny of t~e visible effects of sludge formation are quite
famlhar, particularly rhe clogging of screens, passages and
filters. It is of course unnecessary to dwell upon these trou.
bles in detail, as it is well known that the two former effects period is plotted against the number of rings found to be
may be responsible for engine failure while the third pre- stuck and inoperable at the end of that period. Two sets
vents t~e filter from performing its duty of cleansing the oil of data are given in this chart covering two different oil-
of foreign matter. However, Figs. I and 2 may be referred sump temp~ratures, all other operating conditions being held
to as examples of deposits of this type. constant. Many other mechanical causes of high oil con-
The effects on engine performance and life due to stuck sumption in engines are well known, but it is reasonable to
piston-rings, however, do not seem to be quite as well known expect that rings sticking, either permanently or temporarily,
although, actually, it appears that ring sticking is one of may be responsible for many of the observations of high and
the most important phases of the sludge problem. A few erratic oil consumption which are frequently reported. Fur-
examples ~ay serve to illustrate this fact. In Fig. 3 is ther, not only do sticking rings impair the output of any
shown a piston from a test engine which failed at the end of particular engine, but susceptibility to this trouble very defi-
37 hr. of ~ projected 50-hr. run due to stuck rings and sub. nitely limits the permissible outputs of truly high-specific-
sequent. selzu:e of ~he overhe~ted pistons. It was necessary output engines such as are necessary in the aviation field.
to reb mid thiS engme followmg the failure, as the pistons The two-cycle engine also suffers much from ring trouble
were completely r.uined. This represents a rather exag- when serious attempts at high mean effective pressures are
gerated case, but It does show the same behavior that is made. In this particular type, of course, the problem is
freq.uently observed i? service over longer time periods. aggravated by the hot gases playing directly on the exposed
Fig. 4 shows an Oll-consumption curve for this same en- portions of the upper rings at the moment of exhaust-port
gine in which the average oil consumptiori for a 5o~hr. test- opening. That ring sticking can be such an important lim-
iting factor in engine performance is inevitable from the
• See the Journal of the Institute of Petroleum Technicia.ns, vol. 12,
p. 582, 1926, Moore and Barrett; see also the National Petroleum Ne'ws fact that, as rings become stuck, heat t~ansfer becomes im-
Aug. 13, 1930, p. 63, Lederer ~nd Zublin. . ,
• See th.e paper by Dietrich presented at the Symposium on Motor Lubri- paired with the result that piston temperatures increase.
cant~ dUring !he March 5, 1933, meeting of the American Society for In the aviation engine, stuck rings are liable to result in
Testmg Matenals.
piston failure and probable wrecking of the engine. Even if
actual failure does not occur, overheated pistons tend to in-
duce detonation and, to a certain extent, seizure and abrasion.
Occasionally a badly carboned and stuck set of rings will
result in greatly exaggerated cylinder wear, which is no
doubt accentuated by the fact that under such conditions the
filter will also in general be inoperative. No exact valua-
tion can be placed on these difficulties, but it is quite safe
to assume that any reasonable means of their elimination
would be well justified.
A number of miscellaneous examples of sludge are illus-
trated in Figs. 5 to 13 inclusive which are given for the
purpose of illustrating a part of the variety of forms in
which sludge deposits occur and, by way of contrast, the
results obtained by operation on an oil not susceptible to
sludge-forming oxidation.
The foregoing discussion may be summarized briefly by
stating that, although the bulk of most sludge deposits may
consist of "blow-by" carbon and other matter foreign to
the lubricating oil, the actual existence of a deposit is in
general due to the presence of products of deterioration and,
probably, oxidation of the oil.
Engine Observations of Oil Oxidation
While it has been recognized to a certain extent" that
oxidation is an important factor in the service of crankcase
oils, this is not always the conclusion 4 nor have the sources
of the insolublescommonly reported in used oils been defi-
Vol. 34, No. 5
 Downloaded from SAE International by Columbia Univ, Wednesday, April 15, 2020

SLUDGE FORMATION IN MOTOR OILS 169

oxidation took place primarily on the under side of the pis-

tons and at other places well above sump temperatures.
(3) Even with the wide difference in the character of oils
used, chloroform insoluble material ("carbon") did not vary
greatly, but appeared to be determined mainly by engine
condition. In the white-oil runs this material appeared in a
few minutes, without any accompanying evidence of oxida-
tion of the oil, and it was concluded that the source of such
material is combustion-chamber blow-by.

Comparison of Laboratory a n d Engine Tests

Comparative Results with Various Laboratory Tests.—Hav-
ing established that oxidation is an important factor in the
performance of motor oils and that oils differ markedly in
their behavior in this respect, the next consideration was the
development of a test method by which the oxidation be-
havior of oils in service m a y be predicted. A great deal of
attention has been paid to oxidation tests for light lubricating
oils such as transformer and turbine oils, all of which involve
the use of relatively low temperatures. A s it is apparent
from the data in the previous section that temperatures well
over 300 deg. fahr. are concerned in motor-oil oxidation,
none of these tests are suitable on account of the time re-
quired, if for no other reason. Tests which have been in-
vestigated are as follows:
(1) Fixed-Time Tests at High Temperatures.—These
were:
(A) Sligh Oxidation Test5.—In this test a 10-gram sample
sealed in a flask containing oxygen is heated for 2 % hr. at
200 deg. cent., after which the oil is analyzed for naphtha
insoluble.
(B) Oxidation in Air at 450 Deg. Fahr.—This is the test
nitely explored. T o establish some basic information on described by Davis and Blackwood 6 , in which the oil is
these two questions, a series of exploratory tests were carried tested for viscosity change, acidity and naphtha insoluble
out in a water-cooled engine on the following oils: after 12-hr. oxidation. This test has been used with vari-
(i)-A "white oil" which oxidizes autocatalytically at 210 able time making it more similar to Tests C and D .
deg. fahr. to form colorless acidic oxidation products be- (2) "Life" Tests at 300-3J5 Deg. Fahr.—It has been found
ginning after about 75 hr. of continuous oxidation. This
. reaction can be prevented virtually indefinitely by the use of
antioxidants. At higher temperatures, 300 deg. fahr. or
above, asphaltic materials are formed from the acids.
(2) T h e "white oil" of item (1) containing an anti-
oxidant not effective above 275 deg. fahr.
(3) A conventional motor oil which, upon oxidation, re-
quired a temperature of 330 to 340 deg. fahr. to form ap-
preciable quantities of asphaltic substances in 50 hr.
(4) A n unfinished distillate which upon oxidation formed
asphaltic materials several times as rapidly as did oil N o . 3.
This work m a y be summarized by the following state-
ments, which represent the conclusions which were drawn
at the time this stage of the work had been completed.
(1) Appreciable oxidation of each of the oils occurred in
a 50-hr. run. T h e white oils formed mainly acids, while
oils N o s . 3 and 4 formed acids and asphaltenes. Oil N o . 4
formed a m u c h greater amount of asphaltenes than did oil
N o . 3, indicating that sludging is definitely a function of
oxidation stability.
(2) Variation in engine output markedly influenced oxi-
dation of the oils, while variation in sump temperatures
from 130 to 210 deg. fahr. did not. Oil N o . 1 formed oxi-
dation products at a linear rate and oil N o . 2 gave identical
results with oil N o . 1. All of these facts tend to prove that
5
See the Proceedings of the American Society for Testing Materials
vol. 24, 1924, pp. 964-972.
8
See Industrial and Engineering Chemistry, vol. 23, 1931. p. 1454.

May, 1934
 Downloaded from SAE International by Columbia Univ, Wednesday, April 15, 2020

170 s. A. E. J 0 URN A L
(Transactions)

that the results obtained by continued oxidation of an oil

under fixed conditions of temperature and oxygen supply
afford a much more complete and informative picture of its
oxidation behavior. In early work some tests were made at
268 deg. fahr. but, as brought out in a previous section, this
caused much less asphaltene formation than is experienced
in engine operation for a similar time. Most of the work
has been carried out at temperatures from 305 to 375 deg.
fahr., using either oxygen or air, and two specific tests were
extensively used and compared with engine performance.
(C) 305-Deg. Fahr. Test with Oxygen.-In this test oxy-
gen at the rate of 10 liters per hr. was passed through 300 cc.
of oil maintained at a temperature of 305 deg. fahr. Sam-
ples taken periodically were examined for acidity, viscosity.
naphtha insoluble and sometimes other properties.
(D) 34I-Deg. Fahr. Test with Air.-This test was car-
ried out in the same general manner as (C) except for the
use of a different temperature and air instead of oxygen. A
detailed description of this test is included in a later section.
All of the above tests, (A) to (D) are carried out in glass IO-mg. value, which in a sense is an arbitrary choice, has
flasks without metal present. With the exception of the proved to be very satisfactory as an index because the
Sligh test, provision is made for a supply of oxygen in excess amounts of insolubles formed during the initial period are
of the rate consumed by the oil. A detailed study of the usually well below this, the rate generally increasing rapidly
Sligh test has shown that, with the higher-viscosity oils, which when the To-mg. concentration is reached.
may be characterized as having a higher oxygen absorption, The asphaltene development observed in the engine tests
sufficient oxygen is not present. is similar to that of the laboratory test (See Fig. 14), al-
Test (B) gives single values for change in viscosity and though, before asphaltene formation has become pronounced,
acidity, for evaporation loss and for formation of insolubles. there is a somewhat more gradual development. This may
Tests (C) and (D) give curves which show the development readily be accounted for by possible pickup of small amounts
of acidity, viscosity and insolubles with time. This is an of asphaltenes accumulated in inaccessible places which are
important feature, as will be seen from an inspection of the not thoroughly cleaned after each test, as well as to the fact
curves in Fig. 14, showing some typical tests. During the that only a part of the oil is undergoing oxidation at a given
initial period of oxidation the amount of insolubles is al- time. For this reason the time required to develop 20 mg.
most negligible. Once the formation of asphaltenes begins of sludge per 10 grams of oil in the engine has been arbi-
in appreciable amount, the rate becomes increasingly greater rrarily set as the engine sludging-time. After the oil has
until a maximum is reached; after this it becomes almost reached an asphaltene concentration of IS to 20 mg., the rate
linear. Similarly, viscosity does not change in simple fashion accelerates very rapidly.
with time. For these reasons fixed-time tests cannot furnish Comparison may now be made between the engine tests and
an entirely satisfactory comparison of oxidation stability of the oxidation-test results, by methods (A), (C) and (D), as
oils. shown in Table 2. It is obvious that the Sligh test reflects
As, from the practical side of the problem, one is chiefly in a measure the behavior of an oil in service, but it is also
concerned with the service which may be obtained from an equally obvious that the test is not suited for quantitativI';'
oil before sludge formation begIns, this initial period, the correlation. Oil A used in runs Nos. 20 and 25 had what is
length of which varies with the oil, is the most significant considered a low Sligh value-4.5 mg. per 10 grams of oil-
measure of the sludging behavior of the oil. Accordingly, but its service was immensely inferior to several other oils
sludging results obtained in Tests (C) and (D) are ex- giving Sligh values only a few milligrams less. Similarly, the
pressed in terms of the time required to form TO mg. of correlation for the oil G, run No. 26, is quantitatively far our
naphtha insolubles per TO grams of oil (O.T per cent). This of line. With these very definite indications, supplemented

Vol. 34, No. 5

 Downloaded from SAE International by Columbia Univ, Wednesday, April 15, 2020

SLUDGE FORMATION IN MOTOR OILS 171

by a large amount of laboratory data indicating its lack of although the viscosity reached 390, and, by an extension of
suitability, no further attention was given to this test. the test, the oil reached the sludging point at 13 hr., with a
Test (B) was studied on a group of four oils of S.A.E.-60 viscosity of about 500. Further, test (B) did not rate the
grade, the results of the engine tests being collected in Table four oils, as to viscosity increase, in the same order as found
3 (Table 4 gives operating temperatures for these runs). in the engine test. On the other hand it will be observed
Table 4 also gives laboratory-test data by Methods (B) and from Table 3 that test (D) did rate the oils in the proper
(D). It is indicated that, because of the high temperature order, with the exception of the minor variation between
involved, test (B) emphasizes viscosity increase but does not oils Nand C, and also predicted the sludging behavior of
reliably predict sludging stability. This is shown particularly oil N.
by the results on oil N. In the engine, this oil reached the An inspection of Fig. IS and, particularly of Table 2,
20-mg. sludge-value at 42 hr. and, at 60 hf. contained 4I.2 shows that, in general, the laboratory sludging-time accord-
mg. of asphaltenes and had a viscosity of 172. By test (B) ing to test (C) (305 deg. fahr. using oxygen) correlates with
no significant amount of insoluble was formed in 12 hr., the corresponding engine sludging-time. There is, however,

May, 1934
 Downloaded from SAE International by Columbia Univ, Wednesday, April 15, 2020

172 s. A. E. J 0 URN A L
(Transactions)

As opposed to the above, test (D), using air at 341 deg.

fahr., showed that the relative sludging-time of the two oils
A and F differed markedly from that indicated by test (C).
The 341-deg. fahr. air-test, although rating many of the oils at
approximately the same stability as test (C), rates oil F in
an entirely different class, and in line with the engine test.
Extensive laboratory experiments have shown that oil F is not
an anomaly. Due to the fact that the test using oxygen
"punishes" an important class of oils unduly, it became obvi-
ous that the use of such a test could not afford a valid corre-
lation with performance.
As shown by comparison of the sludging data in Table 2,
the results of test (D) correlate well with engine performance.
Many of the oils did not form significant amounts of sludge
in the 50-hr. test, and a more elaborate comparison includ-
ing viscosity change is given in a later section. Because test
(D) correlated so reliably with available engine and service
data, it has been adopted as a research tool by the authors'
Laboratory and will be referred to as the "Indiana Oxidation
Test." It is described in detail as follows:

Indiana Oxidation Test

As shown in Fig. 16, the oil under test is placed in a glass
tube which is held in an oil bath regulated at constant tem-
perature. The tubes are made of regular Pyrex tubing, 20 in.

one marked discrepancy in the case of oil F. The two oils, long and 1Y4 in. internal diameter. A flowmeter is provided
A and F of engine tests Nos. 24 and 25, show the same to measure the stream of air which is delivered into the
laboratory stability, 31 to 32 hr., whereas the engine results oxidation tube by means of a glass tube (3/16-in. internal-dia-
are markedly different. The negligible asphaltene formation meter), supported oy a cork and reaching to within ~ in.
in run No. 24 indicates that the laboratory sludging-time of of the bottom. A "Bright Stock" of high flash-point and
this oil should be over ISO hr. as was the case with tests good oxidation stability is used for the bath oil.
Nos. 14, 16 and 19, which also showed negligible asphaltene The test is started with 300 cc. of oil, the level of which in
formation in the engine. the tube should be well below the bath level. The test oil

Vol. 34, No. 5

 Downloaded from SAE International by Columbia Univ, Wednesday, April 15, 2020

SLUDGE FORMATION IN MOTOR OILS 173

is kept at a temperature of 341 deg. fahr. (corrected), which

requires a temperature of approximately 342 deg. fahr. in the
bath. Air is passed through the oil at a rate of 10 liters
per hr. (measured under laboratory conditions). Periodically,
depending on the oxidation characteristics of tqe oil, 25 cc.
of oil are removed, 10 grams of which are immediately
weighed into a tared Erlenmeyer flask. This portion is
diluted with 100 cc. A.S.T.M.-precipitation naphtha and
allowed to stand 3 hr. before filtering through a prepared
Gooch crucible. The crucible containing the insolubles is
washed with approximately 100 cc. of naphtha, after which
it is dried Yz hr. at 300 deg. fahr. and weighed. The amount
of insoluble is expressed in milligrams per 10 grams of oil,
and a sufficient number of samples are taken for test to deter-
mine accurately (a) sludging time, that is, the time required
to form 10 mg. of naphtha insoluble per 10 grams of oil and
(b), the loa-mg. point, that is, the time to form 100 mg.
of insolubles. These results are conveniently obtained by
plotting insoluble against time of oxidation on a log.-log.
chart. .
The determination of viscosity increase, which becomes the
more important criterion when sludging time is much ovel
50 hr., is made by taking a IOO-CC. sample every 50 hr. This
run in an air-cooled engine. The technique of making these
sample is promptly run for viscosity and put back in the tests is described as follows:
oxidation tube. Oxidation tests are ordinarily continued Engine-Test Methods.-In making engine tests for sludge
either to the loa-mg. point or for ISO to 200 hr. for "sludge- development, a definite procedure, based on a number of
less" oils. trial runs, was worked out. The method adopted, largely
Experience with this test by several laboratories has shown arbitrary, was adhered to as rigidly as possible in the making
that the sludging time may be checked within about 5 per of all t e s t s . '
cent by different operators on different apparatus. Accurate The engine used in these tests had the following specifica-
tions:
temperature control, fairly exact control of air rate, change
of the oil in the bath before marked thickening takes place
and careful cleaning of the oxidation tubes, are essential. As
it has been found that the sludging time varies inversely with
the partial pressure of oxygen (within a limited range),
correction must be made for tests at high altitudes. The tubes
are cleaned by washing with naphtha, followed by soaking
first with alcoholic KOH and then with chromic acid clean-
ing solution. Two engines were actually used. They had different cool-
ing-fin arrangements but otherwise were identical. The data
Correlation of the Indiana Oxidation given in Table 2 were obtained on one engine while those
Test with Performance, (C) in Table 4 were observed on the other. In each case the new
In the comprehensive study of the correlation of the engine was "run in" at light load and moderate speed, after
Indiana test with performance, a second series of oils was which the load and speed were increased until the cylinder-
head temperatures averaged about 450 deg. fahr. The engine
was then dismantled and thoroughly cleaned, inspected and
measured. After reassembly, several preliminary tests were
made. During these tests temperatures in virtually every part
of the engine were measured by means of thermocouples.
Tests were run for 50 hr. with the engine delivering 24 b.hp.
at 1500 r.p.m. Seventeen thermocouples were placed through-
out the engine in each test. These, in conjunction with the
cradle dynamometer, afforded a reliable and convenient
method of checking and controlling test conditions.
The routine for each test was as follows: The engine was
dismantled, all parts being removed except the crankshaft.
camshaft and timing train. The individual parts were cleaned,
and the carbon from the ring grooves and combustion cham-
ber weighed. The engine was then reassembled, using new
gaskets throughout, after which it was flushed with a mixt~re
of 50 per cent benzol and 50 per cent alcohol by motonng
with the dynamometer. Further flushing with oil from the
supply to be tested was next in order. The crankcase was
then filled with a weighed charge and the engine started.
After 15 min. of operation at light load, the engine was

May, 1934
 Downloaded from SAE International by Columbia Univ, Wednesday, April 15, 2020

174 S. A. E. J0 URN A L
(Transactions)

stopped for final valve adjustments. The test was then started, It was found during the tests that, as the engine became
the engine being run continuously for 50 hr. except for a worn from running and frequent cleaning, the oil consump-
brief stop in the thirtieth hour to check the crankcase level, tion increased. When the consumption became too high to
for which a calibrated gage-stick was used. No oil was added permit 50 hr. of operation without refilling, the engine was
to the crankcase during the test. Samples were taken at fitted with new rings and pistons. In one instance, regrind-
intervals of 10 hr. and examined for acidity, total A.S.T.M.- ing of the cylinders was necessary to keep the oil consump-
precipitation of naphtha insoluble and the proportion of this tion between the limits set. It would have been desirable,
which is chloroform soluble, and viscosity increase. At the of course, to be able to make all tests at a fixed oil-consump-
end of a test, the remaining oil was drained and weighed. tion. This being impossible, particularly as the engine was a
The consumption for the 5o-hr. period was considered to be multi-cylindered one and also because of the rather wide
the difference between the original weight and that of the range of oil viscosities encountered, it was found necessary
drainings, samples being regarded as consumed oil. to work as far as possible between certain limits which were
After the test, the engine was dismantled and the appear- taken roughly as 0.r6 and 0.26 lb. per hr. (total consump-
ance of the parts noted. The condition of pistons and rings tion). Only a few tests were made outside this range; also,
was recorded and carbon scraped from the various parts was it will be noted that a few tests were made in which the
collected and weighed. New parts were supplied to replace 50-hr. period was not strictly adhered to. Not all of the engine
those broken or unfit for further service. Photographs were tests that have been made are recorded in Tables 2 and if-
taken of various parts as a record of the appearance. Parts Many of those tests not recorded in this paper were simpie
coated with the "lacquer-like" deposit previously de~'cribed duplicate or check tests. Others were those into which such
were buffed clean and bright. variations had been introduced as to render the data not

VoL 34. No.5

 Downloaded from SAE International by Columbia Univ, Wednesday, April 15, 2020

SLUDGE FORMATION IN MOTOR OILS 175

pertinent to the present discussion.

By taking samples of the oil at 10-hr. periods, curves
showing the rates of acidity formation, viscosity increase and
asphaltene formation are obtained. The results of acidity
determination were rather erratic and, in view of lack of
evidence of any undesirable practical effects of acidity forma-
tion in itself, very little weight has been placed on these
results, the usual observation being that the laboratory-test
and engine-test results followed the same general trend. Vis.
cosity increase becomes quite an important factor in somt
classes of oils, particularly those of the heavier grades which
contain residual stocks. This class of oils is subject to oxida-
tion at a rate even greater than light stocks; but no insoluble
material is formed for a very long time, during which period
the viscosity may. increase to such an extent that the oil is
unsatisfactory for further use.
Table 4 shows data on the engine tests and sludge and
viscosity results on the corresponding oils when oxidized by
the Indiana test. Inspection data on the oils used are

1HZ7H[W
in Table 5· The temperatures of NO.3 cylinder head and
the oil sump are given as significant indexes of general
temperatures. Oil consumption is indicated because it is
important consideration, the concentration of oxidized prod-
ucts being inversely proportional to the amount of the 1HZ7H[W
in the sump. The analysis and viscosity of thl" final (gen-
erally 5o-hr.) oil sample are given, as well as the engine necessarily limited to about this length of time due to the
sludging-time as previously defined. It will be noted that oil consumption. Inasmuch as the operating conditions on
the control oil A was run in tests Nos. 1, 9 and 12, to deter- this test are about as severe as can be imposed in practice,
mine uniformity of engine conditions. The engine sludging- approaching the temperature limits at which the aluminum
times of 22, 18 and 20 hr., respectively, give assurance that alloys involved can be safely used, it appears that the oils
the results are reliable. having engine sludging-times over 50 hr., corresponding to
Fig. 17 shows a plot of the engine against the laboratory a laboratory sludging-time of more than 70 hr., will be sub-
sludging-time. Widely varying types of oil-Coastal, Mid- stantially sludge-free in practice. With such oils, the addition
Continent, Pennsylvania, as well as some whose method of of make-up oil becomes a governing factor in preventing
refining makes them practically independent of crude source insolubles from building up.
-are included, and there are no exceptions from a reasonably A rather interesting comparison can be made with runs
good correlation between oxidation-test results and perform- Nos. 13 and 16, two oils which have the same laboratory
ance. There are several oils which had not formed 20 mg. stability. The major difference of the oils is their viscosity,
of asphaltenes after 50 hr. of engine service. The test is 80 and 122, which resulted in a lower consumption as is

May, 1934
 Downloaded from SAE International by Columbia Univ, Wednesday, April 15, 2020

176 s. A. E. J 0 URN A L
(Transactions)
shown in Table 4. According to the indicated mechanism of
oxidation in the engine it follows that the smaller residual
amount of the lighter oil from run No. 13 should show a
higher concentration of asphaltenes than does the more viscous
oil. The 50-hr. asphaltene-values of IL7 and 34 mg. thereby
confirm this theory.
In the laboratory test, considerable attention has also been
given to the rate of asphaltene formation after the sludging
point is passed. This is obviously of considerable practical
importance, although it is a difficult matter to get reliable
engine data pertaining thereto, because of the effects on
consumption produced by the insolubles previously formed.
For this reason it has been feasible only to develop a corre-
lation of initial sludging-times in Jaboratory test and engine
test, which permits the assumption that laboratory sludging-
rates are similarly valid.
Due to the fact that a fixed amount, representing a small asphaltenes found near the end of the "dropping-level 16-lb.
proportion of the oil, undergoes oxidation in the engine at run," the "constant-level 16-lb. run" shows a flattening of the
a given moment, it follows that the concentration of oxida- curve after long operation, due to the diluting effect of fresh
tion products obtained will vary with the amount of the oil oil. It can be shown mathematically that this curve approaches
in the sump. This is shown in Fig. 18, which gives results a maximum which is a function of the sludging time of the
with a 34-lb. charge of oil compared with the normal I6-lb. oil and the rate of addition. Thus, in practice, the consump·
charge.· Similarly, the addition of make-up oil to keep a con- tion becomes an important factor governing the degree to
stant sump-level, which is ordinary service practice, has an which oil deterioration can go, and it is not to be anticipated
effect upon asphaltene concentration which becomes quite that the very high asphaltene values found in some of these
significant after 25 hr. of operation. This is also shown in test runs would be obtained in actual operation.
Fig. 18, where it is seen that, opposed to the steep rise in From the foregoing discussion it is obvious that laboratory

Vol. 34, No. 5

 Downloaded from SAE International by Columbia Univ, Wednesday, April 15, 2020

SLUDGE FORMATION IN MOTOR OILS 177

sludging-times well above 70 hr. have little significance as

far as comparative values are concerned and, hence, viscosity
change becomes the primary consideration. In this respect,
as shown in Figs. 19 and 20, the laboratory test is an adequate
measure of performance. Fig. 20 shows viscosity after
50 hr. in the engine, the value being determined by extrapola~
tion or interpolation in runs. Nos. 7 and 9, compared to vis.
cosity after 50 hr. in the laboratory test. In a correlation of
this type it should be remembered that variations in oil con-
sumption as well as in engine temperatures will influence the
viscosity increase of the oil. The major discrepancies are run

Lendency to ring sticking is a function of the sludging time

found by the laboratory test. Those oils having stabilities
less than 75 hr. averaged eight stuck rings, whereas the more
stable oils averaged about five. As may be expected from
the nature of the effect, the results as to actual number of
rings stuck do not show the same quantitative correlation as
in the case of asphaltene formation. One commercial, but
generally unorthodox, oil has been found exceptional in that
its ring-sticking behavior appears to be far worse than is indi-
cated by laboratory sludging-time. Another specially pro-
cessed oil is better than would be expected from the laboratory
test. These latter facts indicate that tendency toward ring
sticking may, under certain circumstances, be dependent on
some other properties of the oil in addition to its sludging-
time.
Service Observations
As the ultimate value of the laboratory work is deter-
mined by the results reflected in actual service, a series of
observations was made on vehicles in fleet operation in which
two oils which had shown laboratory engine-test behavior con-
sistent with quite different oxidation-test values were em-
ployed. The tests were carried out without altering the operat-
ing schedules of the vehicles involved. The test procedure
was quite simple, essentially being as follows:
The engines were thoroughly flushed with the oil to be
tested and new filters were installed. After being placed in

No. 7, in which the engine oil-viscosity increase was unusu-

ally high, and run No.8, in which it was extremely low. In
general, considering the wide ranges of viscosity increase with
the various oils in the engine and in the Indiana Oxidation
Test, it can be said that the correlation is at least fair if not
strictly quantitative. The increased oxidation of the oil in the
engine toward the end of the run, when the oil-sump level
is lower, explains this higher rate of viscosity increase. Dur-
ing the first 20 or 30 hr., the engine rate is only slightly
higher than that of the laboratory.
Table 4 also shows the observation regarding. ring stick-
ing in each of the tests. It will be seen that, in general,

May, 1934
 Downloaded from SAE International by Columbia Univ, Wednesday, April 15, 2020

178 s. A. E. J 0 URN A L
(Transactions)
The detailed data also indicated that oil A deteriorated
more rapidly during the hot summer months than in the
spring. A number of minor points were observed which, while
in general qualitative in nature, tend to confirm the labora-
tory work. For example, where laboratory engine-tests showed
that oil A resulted in filter deposits accumulating at about
three times the rate for oil VJ the average of a number of filter
cartridges taken from the coaches used in city service showed
total deposits of 50.6 and 20.2 grams on oils A and V respec-
tively. The asphaltene concentrations found in these deposits

service, samples of oil were secured at each drain period for

analysis, and deposits were taken from the oil filters and
certain other selected areas. Observations on each oil were
made for about six weeks, the procedure being repeated as
often as possible for the period March I to Sept. I, 1932.
The crankcase samples were analyzed by the same procedure
employed in the oxidation test. The deposit samples were
made oil free by extraction with naphtha and the asphaltenes
determined on the oil-free residue. Tables 6 and 7 give,
respectively, the laboratory and the service data obtained on
the two oils employed. Average values have been used in
presenting the service data, as, of course, variations are much
greater between individual observations than in the rigorously
controlled laboratory tests.
In spite of the fact that the field-test data are not quite so
clear cut as those obtained in the laboratory, it was con-
sistently observed that the proportions of asphaltenes found
in crankcase drainings and in engine and filter deposits
were much less in the case of the stable oil V as compared
with oil A, indicating that the former oil was oxidized to the
lesser degree. These differences were somewhat masked, of
course, by the fact that always some material from the previ-
ous observation period remained in the engine each time a are given in Table 7. Further, even mechanics engaged in
new oil was introduced. With such an allowance it would overhaul work reported visibly cleaner engines in the case
appear that oil V is actually non-sludging, as the asphaltene of the stable oil.
concentrations, noted in Table 7, were quite small indeed. In following the service tests, the attempt was made to
The tractor trucks undoubtedly represented the most severe determine whether the stable oil would lead to improved
type of service because of very high engine operating- oil economy as predicted by laboratory data. The average
temperatures, although the oil from motorcoaches in inter- values for consumption as given in Table 7 do show a trend
urban service shows equal deterioration, probably due to in this direction. As to individual observations, however, it
greater mileage between drains. (Continued on page 18I)

Vol. 34, No. 5

 Downloaded from SAE International by Columbia Univ, Wednesday, April 15, 2020

INTAKE-SILENCER PROBLEMS 181

difficulties, and it had a somewhat broader silencing band fold, before a silencer is designed. Another common ex-
than the Quincke type; but still the range of frequencies cov- perience is to attempt to build a satisfactory silencer which
ered by a single Helmholtz chamber was not sufficiently is much too small for the volume of noise and range of fre-
broad. Furthermore, it was noted that, in addition to the quencies to be suppressed. It must be realized that a silencer
intake-noise period occurring in the speed range, say, from has a definite quantitative function to perform and must be
35 to 45 m.p.h., an additional intake period was found to dimensioned to its application as much as a propeller shaft
come in at another higher speed range with still other fre- or an axle shaft must be dimensioned to carry the loads im-
quency characteristics. This required that at least two cham- posed. The same may be said as well for the design of
bers would be necessary, and perhaps more, if the silencing resonance exhaust-silencers.
range was to be sufficiently broad. To meet these require- It should be mentioned that though the silencer has the
ments, the compound or series-type silencer, consisting of two outward appearance of a tin can, it follows in its action very
chambers in series off the main sound-channel, was worked definite and well-known physical laws. These same silencers
out. It was found that, for the same overall size of silencer on intake or exhaust can exert very definite influences on
two ranges of frequencies could be covered with about th~ engine performance, and often a poorly designed silencer will
same silencing effectiveness as a single chamber covering only more than offset gains anticipated from an expensive change
one of the frequency ranges. in engine constructiOll. When these possibilities are com-
Thus far we have considered only those noises issuing pletely appreciated by the designing engineers, the optimum
from the intake which were on the order of firing frequency. in engine silencing, performance and economy may be fully
Frequencies which are some of the higher harmonics of the realized. The intake silencer has permitted a gain to be
firing frequency likewise are present and must be silenced in made in specific power per cubic inch of cylinder displace-
order that a completely satisfactory job can be done. Some ment as, by its use, the silencer has made possible a wider
of these higher frequencies may be silenced by means of choice of valve timing without restriction by noise limi-
chambers or by means of sound absorbing material. The tations.
amount of silencing of the higher-pitched noises will depend
on the balance between desired degree of attenuation on the
one hand and cost and space limitations on the other. Or-
dinarily, the most practical silencer consists of a combination Causes and Effects of Sludge
of chambers and sound-absorbing material, the latter being
used primarily to silence high-frequency noises, such as hisses,
Formation in Motor Oils
originating in the carburetor. Other low-pitched noises is- (Continued trom page 178)
suing from the intake are most satisfactorily silenced by the
chamber-type silencers. must be said that they fluctuated quite widely indeed in com-
Paralleling the development of the intake silencer, a sat- mon with general experience. Both consumption and asphal-
isfactory exhaust silencer, also utilizing the principle of reso- tene formation, as observed in service vehicles, might rea-
nance silencing, has been developed. One important result sonably be expected to deviate quite widely from the actual
of these studies has been to show definitely that a design laboratory values if for no other reason than that, whereas
problem is presented in silencing either the exhaust or in- "make-up oil" is regularly added to the service engines, the
take, which is as tangible as the design problem in, for ex- laboratory work here recorded was all done without the addi-
ample, a crankshaft. Furthermore, the results to be expected tion of make-up oil and therefore with constantly dropping
from a given design can be predicted by paper analysis. oil level.

Cooperative Design-Problem Studies Conclusions

After the fundamental design-problems had been worked To summanze briefly the work covered III this report, it
out for the resonance type of intake silencer, the commercial is indicated that the formation of sludge in motor oils is
design-problems were worked out in cooperation with the due primarily to asphaltenes resulting from oxidation of the
AC Spark Plug Co. and with the Buick Motor Co. This oil. Engine experiments indicated that the type of oxidation
cooperative study is being continued between the Research involved occurred at temperatures well above those normally
Laboratories and the manufacturing companies already men- existing in the oil reservoirs and that, in general, good agree-
tioned, and most of the early difficulties are rapidly being ment could be expected between representative engine per-
overcome. formance and laboratory oxidation tests.
The chief difficulty still remaining is due to the lack of Of several laboratory tests investigated, the most promising
understanding on the part of the engine designer that the consisted of a continuous oxidation test in air at 341 deg.
intake silencer is a part of a tuned system which must be fahr. The results of this test correlate well with carefully
as harmoniously tuned as a piano. For example, too many controlled heavy-duty-engine tests with respect to asphaltene
times a silencer is designed for a given engine but, when formation and viscosity increase, and generally give sound
the silencer is tested, the expected results are not obtained indications of tendency to ring sticking. Extensive service
simply because in the meantime a change has been made in, tests with two oils differing widely in oxidation tests have
say, the valve timing, which to the engine designer seems served to prove the importance of oxidation stability, to con··
quite unimportant but which has a very decided effect on firm the conclusions of the laboratory work and to furnish
the intake-noise characteristics. Therefore, the engineer in further evidence of the effect of stability on oil consumption.
charge of engine development must realize the importance The authors wish to express their appreciation for the
of seemingly minor changes, thereby saving much testing valuable contributory work of Kenneth Taylor, J. O. Eisinger,
time. ' It is necessary in most cases that the engine should M. H. Arveson, M. L. Mack of the Research Department
be set as to displacement, valve timing, carquretor and mani- and H. R. Mathias of the Technical Division.

May, 1934