(2). For the coal-tar oils, Eq. (3) gives an error that is acceptable for practical purposes (5.56%), while Eqs. (i) and (2) are unsuitable. Equation (2) gives unacceptably high results for the aromatic hydrocarbons and aromatic products from petroleum and coal processing. On the basis of our calculations, Eq. (3) is recommended for the determination of molecular weights of individual hydrocarbons, petroleum fractions, gas condensate fractions, coal-tar oils, and products of coal hydrogenation, with molecular weights from 62 to 460, boiling points 0-500~ density 0.62-1.17, and contents of sulfur, nitrogen, and oxygen up to 5% each or up to 10% total. For the fractions and oils, it is recommended that the basic parameter for correlation should be t b = tvol_av - 4 where tvol_av is the volume-average boiling point at normal pressure. LITERATURE CITED i. 2. 3. 4.

E. N. Sudakov (ed.), Calculations of Basic Processes and Equipment in Petroleum Refining [in Russian], Khimiya, Moscow (1979). M. L. Kreimer, R. N. Ilembitova, and E. A. Akhmadeeva, Tr. Bashk. Nauchno-Issled. Inst. Pererab. Nefti, No. 14, 160-168 (1975). Z. V. Driatskaya et al. (eds.), Crude Oils of the USSR [in Russian], Vol. 4, K h i m i y a , Moscow (1983). J. A. Gray, Ind. Eng. Chem. Process Des. Dev., 22, No. 3, 410-424 (1983).

SCREENING TEST METHODS FOR MOTOR OILS UDC 621.433:621.892

V. D. Reznikov

One of the conditions for successful development of motor oils in a short period of time with acceptabl e costs is the use of reliable screening test methods in the stage of laboratory studies and primary engine evaluations of the properties of experimental materials with the aim of optimizing the composition of the oil being developed [i]. The problem of reliability of these methods has become particularly acute because of the fast-increasing costs of engine tests for qualification and classification. According to [2], classification tests on a motor oil for conformance to the classes SG/CE (API), G3, PD-I, and D3 (CCMC) and qualification tests for conformance to the specifications of the leading firms MAN, VW, and Daimler-Benz now cost 1 million deutschemarks (with single tests by each of the prescribed methods). From 1970 to 1989, the cost of development and engine-test certification of an oil in the highest class for its time had increased from 148 to 1680 thousand deutschemarks, i.e., by a factor of more than ii [2]. This is due mainly to an increase in the number of engine tests on an oil for classification and the ever-increasing use of full-scale engines instead of single-cylinder units in evaluations of a number of the most important service properties of oils. A similar trend can be noted in the situation in our country. For example, tests in the UIM-6N-NATI unit have been replaced by tests in full-scale D-240 and D-245 engines [3]. Here we are presenting a review of screening tests methods for oils - methods that are used to evaluate the properties of experimental formulations in the stages of optimizing the composition before performing classification and qualification tests in engines. These same methods are often used to evaluate the properties of oil base stocks and individual functional additives for their combinations - the building blocks for complete additive packages. Principal Requirements Imposed on Screening Test Methods The results of screening tests must correlate satisfactorily with evaluations of the same properties of the oils by classification tests in engines; the screening tests must All-Russian Scientific-Research Institute for Petroleum Processing (VNII NP). Translated from Khimiya i Tekhnologiya Topliv i Masel, No. 9, pp. 28-33, September, 1992.

532

0009-3092/92/0910-0532512.50

9 1993 Plenum Publishing Corporation

TABLE 1 ....

Arbitrary d~signa~1on 1.1

Name of method

Sources

Test conditions

GOST 11063--77, [8] Oxidation by air in a thin

DK-NAMI

layer and in bulk (in rotating L-shaped flask) at 200~

Rating indexes Oxidation time up to the initial formation of insoluble sludge (induction period of sludge formation); viscosity increase during induction period Increases of viscosity and carbon residue; content of insoluble sludge; change of acid number during test period

1.2

Determination of oxidation stability in universal appare~s

GOST 18136--72, [6] Oxidation in bulk at temperatures u p t o 250~ by bubbling wihh air or oxygen with or without catalysts (strips, foil, coils, or soluble Metal compounds, with oxidation time and conditions depending on enduse and properties of oil)

1.3

Evaluation of period of sludge formation for additive oils

19}

Oxidation by air in bulk (20 ml) at 230~ with vigorous mixing by a rotating rod

[I01

Oxidation by bubbling with ~zr in the bulk (300 g) at 165~ for 64 h in the presence of a soluble catalyst (organic compound or iron) Oxidation bv bubblinR with Viscosity increase; IR aboxygen (5 llters/h) in bulk sorption at wavelength (300 ml) at 149~ for ~8 h 5.8 ~m; acid number in%n the presence of catalysts crease; content of insoluble (Fe, Cu, and Pb strips) sludge and lead in oxidized oil; change in weight of strips

1,4

ERCO--'rEST

1.5

ARCO Railroad Oil [11] Oxidation tes~ ( modi-

fied Sinclair ~ethoa) 1,6

AMUPOT

Amoco

[12, 13]

Conditions the same as in method 1.5, but test duration 72 h; catalysts Fe and Cu strips

[14--19]

~

(Modified Union Pacific~x~dation Test)

1.7

TFOUT

(Thin Film

OxygenUptake ..... Test)

1.8

PSU -- microoxidation test

[16, 20--23]

Changes in optical density during each hour of test; time to reach an optical density of 0.5 (decrease in light transmlssion relative to initial value) Change of viscosity in each 8 h of test

Viscosity increase; IR absorption (1710 cm -I band); changes in base and acid numbers; weight loss from copper strips; change of pH; overall inffex calculated by an empirical formula xidation in a thin layer Induction period, defined 1.5 g) by oxygen in a as the time to the start closed vessel at 160~ in the presence of soluble cat- of a rapid drop of presalysts, water, and oxidation sure in the closed v~ sel (bomb) inltiators (nitrated hydrocarbons); initial pressure 620 kPa, test time from tens to hundreds of minutes Oxidation in a thin la~er by Induction period, defined a flow of air at one of sev- as the time to t6e start eral temperatures in the of severe deposit forma210-300~ interval or higher, tion at the given temtest time from tens to hunperature; temperature dreds of minutes; variants ~ependence of the inducare possible: oxidation at tion period; chromatographic 150-~75~ with a test time and IR spectroscopic data of 10-15 h, in the presence on oxidation products, of catalysts, HzO , and N20 in particular the change in molecular weight

d i f f e r e n t i a t e c l e a r l y among oils w i t h respect to the level of the p r o p e r t y being evaluated; results must be o b t a i n e d rapidly, w i t h good r e p e a t a b i l i t y and r e p r o d u c i b i l i t y ; the e q u i p m e n t must be r e a d i l y a v a i l a b l e and low in cost; labor costs must be low, and only small amounts of test sample must be required. The first two of these r e q u i r e m e n t s are the most important. If these two r e q u i r e m e n t s are met, the s c r e e n i n g test m e t h o d can be used to p r e d i c t the results of e n g i n e tests. Therefore, any s c r e e n i n g method, before it is i n t r o d u c e d into the p r a c t i c e of lube oil development, should (without exception) be checked for c o r r e l a t i o n and a b i l i t y to differentiate, these checks being based on a s t a t i s t i c a l l y s i g n i f i c a n t n u m b e r of samples w i t h different compositions. U n f o r t u n a t e l y , this r e q u i r e m e n t is not always met. An e x p e r i m e n t a l check has shown that some of the s t a n d a r d l a b o r a t o r y m e t h o d s do not give a s a t i s f a c t o r y p r e d i c t i o n of engine test m e t h o d s [4, 5]. A m o r e r e l i a b l e p r e d i c t i o n of these results can be o b t a i n e d on the basis of a c o m b i n a t i o n of l a b o r a t o r y e v a l u a t i o n s of those p r o p e r t i e s of the 0il w h i c h together i n f l u e n c e (for example) the f o r m a t i o n of v a r n i s h and carbon deposits on pistons [6].

533

TABLE 2 Arbitraryl designa-- Name of method tion

Sources

2.1

Papokm~thod

JOST 23175--78

2.2

M'ODGOST

[io]

2.3

DSC (differen tial calorimetry

[24--27]

TGA (thermogravimetric analysis)

[27, 28]

2.5

HTT (Hot Tube

[29]

2.6

Panel Coker~Test

A I M standard D 3462; GOST~ 12337--84 (determination of coking tendency on panel), [7]

2.4

Test conditions Isothermal oxidation in an air atmosphere at a given temperature (most often 250~ Modification of method 2.1; isothermal oxidation at a constant temperature inthe 200-350~ interval, with air fed at 1 liter]min to the vaporizers Oxidation in a thin film (isothermal, or with a programmed temperature rise) under excess oxygen or air pressure in the presence of soluble catalysts (metal naphthenates, oxidized hyarocarbons) . Heating with a preassigned program in an oxidizing andTor inert medium (nitrogen, argon) Passage of oil and air through a glass tube placed in a ~urnace at 320~ for 16 h (several versions of this method) Periodic spraying of oil onto a hot panel surface; test duration, oil spray and drain schedule, and panel temperature depend on the end-use of the oil

Rating indexes Time until the formation of equal quantities of varnish and working fraction on the vaporizer eights of oil and varnish expressed as percentage of resh oil weight) in relation to time or temperature

i

Induction period of oxidation up to the start of an exothermic effect; activation energy of oxidation and deposit formation; ratio of areas on thermograms by the "two-peak" method Oil loss from vaporizer; dependence of thermooxidative or thermal decomposition of oil on temperature; activation energy of oxidation Rating of deposit quantity in glass tube on a special scale; critical temperature for deposit formation Weight of deposits on panel, or visual rating (in comparison with a standard scale)

It should be kept in view that the correlation of results of laboratory and engine tests is not always described by a proportional relationship, and that the results of tests are often expressed in arbitrary, semiquantitative units (for example, in merit or demerit ratings). Hence, a screening method should be considered acceptable in terms of reliability of prediction if it will rank a large number of samples with different compositions in exactly the same order. Consequently, with variation of the level of the rated property of the oil over a broad range (from unsatisfactory to very high), rank correlation is adequate. Developers and consumers of lube oils, additive manufacturers, and engine manufacturers use a number of diverse methods for screening tests (laboratory and engine). We will examine only certain typical methods that are used in practice. Since the assignment of one method or another to the evaluation of a specific property of an oil is largely arbitrary (in many cases, the test result reflects several properties of the oil at the same time), we will be guided by a classification of methods that is generally accepted [7]. Methods for Evaluation of Antioxidant Properties of Motor Oils A summary of these methods for different types of oils is given in Table i. Method i.i is most often used in our country. The induction period of sludge formation, which is taken in this method as the main rating criterion, depends not only on the high-temperature oxidizability of the oil, but also on its dispersancy. In [8] and in a number of other publications, the induction period of sludge formation has been shown to correlate satisfactoirly with the tendency of oils to form deposits in engines under test-stand conditions. Undesirable features of the method are the long test time (up to 60 h), and the low level of repeatability and particularly reproducibility of the viscosity increase during the oxidation period. The test time is shorter in method 1.3; however, in order to confirm its suitability as a screening test, a large volume of data must be accumulated and systematized in order to determine whether the results correlate with engine test evaluations. Methods 1.2, 1.4, 1.5, and 1.6 are similar in many respects: The oil is oxidized in the bulk at a moderately high temperature in the presence of catalysts; the oxidizing gas is bubbled through the oil sample at a rather high rate. In most cases, method 1.4 gives results that correlate well with results from engine tests on oils by Sequence III-C and III-D methods [i0].

534

~0

~8

56

G~

~, h

Fig. i. Correlation of induction period zi of oxidation of seven oils by the modernized TFOUT method with the time 9 in Sequence III-E engine tests up to a 375% viscosity increase

[19].

/

30

" 10

I

o / 20

60

I00

1~o

~o

2~0

T, min Fig. 2. Quantity of deposits Q as a function of oxidation time 9 at 230~ for Class CD SAE 30 oil [22]. Methods 1.5 and 1.6 were developed specially for oxidation stability evaluations of oils used in locomotive diesels. They are modifications of a well-known method - the Sinclair Railroad Oil Oxidation Test. By making the oxidation conditions much more severe (by a factor of 2-4), the test time has been curtailed. Improvements have been made in the procedure for rating the test results, and their correlation with engine test results has been verified. In particular, good correlation was shown in [ii] between the results of oxidation tests by method 1.5 and the results of engine tests on oils by the EMD 2-567 method. The coefficients of correlation were above 0.9 for all of the rating indexes with the exception of the base number decrease. Method 1.6 is characterized by completely satisfactory metrologica! indexes [12, 13]. It has been proposed for standardization. Oils that have been used in locomotive diesels and oils that have been oxidized by method 1.6 contain similar products of aging [12]. Methods 1.2, 1.4, 1.5, and 1.6 have common disadvantages: the long test time and the laborious analyses of the oxidized oil. Methods 1.7 and 1.8 differ significantly from those examined above in that they require only a short test period (up to some hundreds of minutes) and very small oil sample requirements; they also differ in the conditions of oxidation, which takes place in a thin film in these methods. They are widely used in plant laboratories and research laboratories in many foreign countries. In terms of cost, time expenditure, metrological characteristics, and correlation with the results of engine tests, they generally meet the requirements for screening test methods. Method 1.7 has been standardized in the United States. According to ASTM D 4742, it is intended for the evaluation of oxidizability of oils used in gasoline engines, but the method is often used in studying the influence of individual or combined additives on the oxidizability of oil base stocks with various compositions. Good correlation has been demonstrated repeatedly between the induction period of oil oxidation by method 1.7 and the engine-test time up to the moment of limiting (375%) viscosity increase in the Sequence 535

III-D method. In connection with the replacement of Sequence III-D by the more severe method Sequence III-E in testing API Class SG oils, the TFOUT method has been modernized [19]. By partial replacement of catalysts and changes in the catalyst contents in the test oil, it was possible to change the oxidation conditions without harming the correlation between the induction period in laboratory oxidation and the viscosity increase in engine tests by the Sequence III-E method. From the mixture of soluble catalysts (metal naphthenates), the copper naphthenate was eliminated, which was related to the use of low-solubility copper compounds as antioxidants added to certain modern motor oils. The correlation of the oxidation induction period in the modernized TFOUT method with the test time in the Sequence III-E method corresponding to a 375% increase in oil viscosity at 40~ is shown in Fig. I. In [19], good repeatability was noted for results obtained by the modernized TFOUT method. Method 1.8 has many modifications and applications. It is used to evaluate the oxidation resistance of synthetic and mineral oils, to investigate the mechanism and kinetics of oxidation processes, and to investigate the effects of various additives and catalysts on these processes. In combination with IR spectroscopy and chromatography, it offers a means for studying the oxidation products. In [16], from a comparison of data obtained by several laboratory methods, including the TFOUT method, it was concluded that one of the versions of method 1.8 is to be preferred in evaluating the oxidizability of API Class SG oils. Since method 1.8 is universal and can be used in many different versions, and since the oxidizability of oils used in gasoline engines is rated by method 1.7, it is advisable to review experience in the use of method 1.7 to evaluate the antioxidant properties of diesel oils. Studies performed by Caterpillar [20, 22] have demonstrated good repeatability of results from laboratory evaluations of oils by method 1.8, and have shown that the results are consistent with full-scale engine test results; these studies have provided a sound basis for considering the use of method 1.8 in screening tests. The version of method 1.8 in question consists essentially of isothermal oxidation of a thin film of the oil by a directed flow of air on the surface of a cup made of the same material as the engine piston. On a plot of the quantity of deposits that are formed as a function of oxidation time, there are four characteristic sections (Fig. 2). As long as the antioxidant in the oil has not been consumed, no deposits are formed, or the quantity of deposits increases very slowly. After exhaustion of the antioxidants, the quantity of deposits starts to increase rapidly; then stabilization sets in, and, finally, a decrease in the quantity of deposits as a consequence of decomposition and the formation of volatile substances. The induction period of oxidation is determined from the point at which a tangent to the section of the curve characterizing rapid growth of deposits intersects the time axis. It was shown in [22] that the induction period of isothermal oxidation decreases as the temperature is increased from 210 ~ to 260 ~ but the dependence of the induction period on the reciprocal of the absolute temperature is individual for each oil and is practically linear when plotted on semilogarithmic coordinates. Hence it is sufficient to perform an experiment at a single temperature, for example 230~ that is characteristic for the conditions of oil oxidation in the piston ring zone; thus it is possible to limit the test time to something reasonable (the induction period of oxidation is 2-2.5 h for oils with a high level of antioxidant properties). Oils that have passed engine tests and those that have failed show at least a twofold difference in the induction period of oxidation. This difference is many times greater than the probable error in determining the induction period at a single, constant, accurately maintained temperature. Method 1.8 can also be used to evaluate the thermooxidative stability of motor oils and their tendency to form high-temperature deposits [24]. In this case, the rating index is the quantity of deposits formed during isothermal oxidation over the course of the selected time. The temperature and oxidation time are chosen on the basis of the thermal stability and oxidizability of the products being tested. Methods for Evaluating Thermooxidative Stability of Motor Oils A summary of these methods is presented in Table 2. They are based on high-temperature oxidation and thermal decomposition of oils on the surface of a metal (steel or aluminum

536

alloy) or glass, with the formation of insoluble products that are deposited in the form of varnish or hard carbon (cokelike) deposits. Method 2.1 was initially used to evaluate the thermooxidative stability of oils without additives. Now that increasing quantities of additives are used in motor oils and the compositions of the additive packages are becoming more complicated, it has been found that the results obtained by method 2.1 are not consistent with engine test evaluations. In certain standards, however, this method is recommended as a means of monitoring the stability of quality and uniformity of batches of oil. Metrological certification of method 2.1 has not provided any grounds for its recommendation as a test method for evaluating oil quality. We must assume that the inadequate reproducibility and repeatability of results are due mainly to the lack of any controlled flow of air at the surface of the oil during oxidation. This problem has been rectified in method 2.2, which is a modification of 2.1. The The ring indexes have also been changed in method 2.2. With this method, it is possible to obtain an adequately accurate evaluation of (for example) the influence of exhaustion of an antioxidant additive on the tendency of oils to form deposits, the influence of oil-soluble catalysts on this process, and the critical temperature at which deposit formation reaches a maximum. No information is presented in [i0] to demonstrate any correlation with engine test results. Methods 2.3 and 2.4 are used separately or together as screening methods. By varying the test conditions (temperature, catalyst composition and quantity), consistency is achieved in the results of laboratory and engine tests [25, 26]. By comparing TGA results obtained in air and in argon [28], separate rating of thermal and thermooxidative decomposition of oils could be obtained, and the role of oxidation processes in the formation of deposits and volatile products could be clarified. DSC has been used in predicting the tendency of motor oils to form sludge [30]. The laboratory predictions were confirmed by engine test results. Method 2.5 is used extensively in evaluating the effectiveness of detergent and antioxidant additives, lube base stocks, and complete formulations of oils with additives. Depending on the purpose of the investigation, the test conditions may vary over a wide range. The capillary diameter, the temperature, the rates of oil and air feed, and the test time may all be varied. In [29] it was shown that Class CD diesel oils could be differentiated on the basis of the critical temperature of deposit formation. In test-stand evaluations and service tests, unsatisfactory conditions were noted for engines that had been operated on oils with a critical deposit formation temperature below 280~ There are may versions of method 2.6. Only six of these modifications are described in [7]. In our country, this method is used to evaluate the tendency of Group E oils to form high-temperature deposits, in the course of a 24-h test with an aluminum panel temperature of 315~ the oil is sprayed onto the panel for 15 sec out of each minute. Shorter tests at a different temperature are performed to evaluate the effectivenes of additives, singly and in combination. By selecting the proper test conditions for method 2.6, oils can be ranked in the same order as that based on piston cleanliness in engine tests. In [31], a simple and convenient method designated ELF/ECL is proposed for the evaluation of oils with respect to high-temperature deposition tendencies. It has the advantages of simple apparatus, short test time, and good correlation with the results of engine tests in the Renault 30TD turbosupercharged engine. For the evaluation of marine diesel lubes with respect to high temperature deposit formation and for the selection of additives to be used in such oils, the standard method DIN 51392 (Wolf-Streifentest) is used [32]. Here the laboratory evaluation serves as a means for preliminary screening; it does not in itself replace the engine tests. The rating index is the quantity of deposits formed in 12 h on panels heated to 250~ According to [5], the DIN 51392 method does not give a satisfactory prediction of oil behavior in engine tests. Thus, as a result of many years of study, various laboratory screening methods have been developed (TFOUT, PSU, DSC, TGA), tests with minimum requirements on test time and sample size, that will give a satisfactory prediction of motor oil behavior in engine tests; these screening tests can also be used to select prospective lube base stocks, synthetic components, and additives.

537

LITERATURE CITED i. 2. 3. 4.

V. D. Reznikov, Khim. Tekhnol. Topl. Masel, No. 7, 17-24 (1990). W. Grossmann, C. Eberan-Eberhorst, and J. Raddaz, Mineraloeltechnik, No. 8, 25 (1989). A. G. Nikitin, G. P. Belyanchikov, S. C. Arabyan, et al., in: Chemmotology [in Russian], MDNTP Obshchestva Znanie, Moscow (1991), pp. 88-94. P. A. Asseff, U. S. Dep. Commer. Nat. Bur. Stand. Spec. Publ., No. 488, 77, 78, 95-97

(1977). 5. 6. 7. 8. 9. i0. ii. 12. 13. 14. 15. 16.

N. Heckotter and W. J. Bartz, Motortech. Z., No. 4, 161-166 (1979). G. Jager et al., Schmierungstechnik, No. ii, 329-331 (1981). S. M. Hsu, Lbur. Eng., No. 12, 722-731 (1981). M. S. Borovaya, K. S. Ramaiya, N. A. Okinshevich, et., 21., Tr (Vses. Nauchno-Issled. Inst. Pererab. Nefti, No. 5, 3-10 (1977). Methods of Analysis, Investigation, and Testing of Petroleum Crudes and Products (Nonstandard Methods), VNII NP, Moscow (1986), pp. 108-142. F. C. A. Killer, Mineraloeltechnik, No. 5, 17 (1981). T. C. Chao et al., Lubr. Eng., No. i, 15-23 (1986). J. L. Thompson, R. L. Anderson, and D. A. Hutchinson, Lubr. Eng., No. 9, 768-774 (1988). R. D. Staufer and J. L. Thompson, Lubr. Eng., No. 5, 416-423 (1988). Chia-Soon Ku and S. M. Hsu, U. S. Dep. Commer. Nat. Bur. Stand. Spec. Publ., No. 671, 288-296 (1984). Chia-Soon Ku and S. M. Hsu, Lubr. Eng., No. 2, 75-83 (1984). S. Cunsel, F. E. Lockwood, and T. D. Westmorland, SAE Tech. Pap. Ser., No. 892164

(1989). 17. 18.

19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30.

31.

32.

538

H. Clemencon, Tribol. Schmierungstech., No. 5, 262-264 (1988). P. C. Hamblin, U. Kristen, and D. Chasan, in: Additive fur Schmierstoffe und Arbeitsflussigkeiten. 5. INt. Kolloq., Esslingen, Jan. 14-16, 1986; published in Esslingen (1986), Vol. 2, 7.3-1-7.3-2.7. Chia-Soon Ku, P. Pei, and S. M. Hsu, SAE Tech. Pap. Ser., No. 902121 (1990). J. M. Perez, F. A. Kelley, E. E. Klaus, et al., SAE Tech. Pap. Ser., No. 872028 (1987). T. H. Dincher, E. E. Klaus, and J. L. Duda, SAE Tech. Pap. Ser., No. 881617 (1988). F. N. Zerla and R. A. Moor, SAE Tech. Pap. Set., No. 890239 (1989). E. E. Klaus, J. L. Duda, and P. Wu. Proc. Jpn. Int. Tribol. Conf., Nagoya (1990); published in Tokyo (1990), Vol. i, pp. 409-414. J. M. Perez, P. Pei, Y. Zhang, et al., SAE Tech. Pap. Ser., No. 910750 (1991). E. Gegner and A. Born, Mineraloeltechnik, No. 8, 30 (1987). S. M. Hsu, A. L. Cummings, and D. B. Clark, U. S. Dep. Commer. Nat. Bur. Stand. Spec. Publ. No. 674, 196-208 (1984). J. R. Barnes and J. C. Bell, Lubr. Eng., No. 9, 549-555 (1989). S. M. Hsu and A. L. Cummings, SAE Tech. Pap. Ser., No. 831682 (1983), pp. 51-60. S. Ohkawa et al., SAE Tech. Pap. Ser., No. 840262 (1984). U. Kristen, M. Hutchings, and R. Schumacher, in: 7th Int. Kolloq.KraftfahrzeugSchmierung, Esslingen, Jan. 16-18, 1990; published in Esslingen (1990), Vol. 2, pp. 12.1-112.2-11. J. M. Georges, J. L. Loubet, N. Alberola, et al., in: 7 Int. Kolloq. KraftfahrzeugSchmierung, Esslingen, Jan 16-18, 1990; published in Esslingen (1990), Vol. 2, pp. 12.14-1-12.14-5. J. Hengeveld and T. Wajer, Mineraloeltechnik, No. 9, 20 (1981).