60
It I A I.
,< 1 1)
fi
x
(;
I 3 I’: b> It 1 N G
pruteeted in its eiijoyrnent fay 14. i v j i .
EFOltE tahng up a discustiioii rnoiitlis and sonietiniea years, or1 prubof this subject as expressed iri lems confronting them. This is the distlie title, t h e r e a r e c e r t a i n tinction between a commercial laborafundamental principles with reference tory and a research lahoratory. The to industrial research o r g a n i z a t i o n s pressure of commercial demands upon wliich first should be discussed, and the former do not permit much time to there arc certain assumptions to he bo given to research study. Quick restated. To justify the formulation of sults niust be obtained from them. In any principles of management in a rea research laboratory, alrile there must search laboratory, we must assunic tliiit be a certain willingness to be driven the laboratory is an established affair arid to report promptly knowledge alor, a t least, is accepted as becoming an ready acquired, there must also hr a n established affair. There is no need to unwillingness to he driven unduly on discuss principles of nranagenient in a problems requiring study and careful researcli latioratory that is organized investigation. for oiie or a few specific objectives; An industrial research laboratory is nor is it worth while to discuss principresumed to exist to satisfy coniinerples of management unless they can be cia1 needs. It must contribute someapplied on a long-time basis. Furtherthing to industry, and especially to the more, there can he no principles of particular industry supportiirg it, to ~nanagenicntreally worth stating- unless .iustifv-its existence. Tliere is this disthe laboratory is a fairly large one and employs sucli a nuiiiber tinotion, therefore, between an industrial research laboratory of meii that some subdivision of work and real organization are and those laboratories engaged in pure research. The latter required. There is one other very broad essential~~knamely, niay contribute to industrial needs, but that is not their that a research laboratory be recognized as an important main purpose. An industrial research laboratory must conpart of the business organization by which it is supported. tribute to tlie commercial needs of tlie industry. UndoubtA research laboratory search- Sor knowledge, particularly edly, in many industrial research laboratories true pure in its own field. Two things are important in this connec- scientific research is carried on. It is, howevor, usually tion: one is an insatiable thirst fur knowledge; the other limited to the field of the industry. Otherwise i t is questionis ample library and record facilities wliere this knowledge able whether it can be justified. This close relation existcan be sought, orderly assernbled, and stored away. Tbere ing between the connnercial needs OS ari industry and its zre continuous demands made npon tlie researcli lahoratory research laboratory may at times present dificulties. Wliat for information in regard to tlie product.^ inanufactured the selling or producing parts of an industry may demand and their uses. I’urely commcr~:ialmen can hardly bc cx- of a research laboratory are possibly unreasonable both as to pected to have this knowledge in accurate form, arid t h y substance and form. Tlic question may be r a i d as to must rcly upon the research laboratory to give it to them whether expenditures for research arc justified unless rein tliat forin arid to correct from time to time tlie crroneous sults are obtained u,!iicli are of immediate commercial adopinions mhicli are apt to be developed wlierc lack of ac- vantage. Sonietiines, although fortunately less now than curate knowledge exists. A thorough search and orderly in days gone by, demands have been iriade on the research assembly of available knowledge is also a neccssary require- laboratory by those in charge OS commercial problems for ment in all research projects. reports iii fornis that result in obscuring the truth. S a t u I t cannot be expected, of course, that every inenilm of a re- rally, tho rescareti laboratory cannot conforui to such desearcii laboratory will liave, to the same high degree, a thirst mands, for it is essential in research work that there be a for knowledge, but this sliould he the mainspring back of devotion to truth as i t is seen. Tlic old question as to what the workers in a good rosearcli laboratory. A thirst for constitutes truth is unsolved as much liere as elsewhere. knowledge is essential, as otherwise there will not be h i l t In view of these difliculties n e must presume that tlic reu p that power of anticipating answers to inquiries which search laboratory has corifiriried itself, in the niinds of those ufteri conre, as it were, froni a clear sky. A thirst fur hiowl- in control OS cornniercial operations, as being an important edge, resulting in its acquisition and assembly, will always part of the organization, and that i t is depended upon for tend towards keeping a research laboratory ahead of the help, truthful information, and an earnest endeavor to supply game. It nil1 be secii froit! this that it slrould be niadc up the organiaatioii’s needs, patiently, arduously, and devotedly. largely of men who are students, who like study, and who With this picture (not by any means ideal) of the kind of are not afraid to put many weary honrs, extending thmngh mseareh laboratory of which the principles of management I
INDUSTRIAL AND ENGINEERING CHEMISTRY
70
relating to the industry, with few exceptions, should be available. Some of these should be bound a t the end of the year and the others should be stored, a t least For a period of time. Trade journals of course ought not to be kept more than a short time. ORGANIZ.4TION WITH
REFERENCE TO FIELD COTERED
Obviously, if the field is large enough, it must be subdivided into several divisions, and the work under each head still further subdivided. The director of a laboratory cannot be expected to have as much knowledge in each of the fields comprised by the industry as those heads under him should possess. He may have some form of knowledge based, perhaps upon long experience, which is invaluable when passing upon the work and recommendations of heads under him. He must have a knowledge sufficient, in reference to each of these heads, to enable him t o give some guidance to the work they are doing or projecting and to be a t all times ready to discuss their problems with them. Sometimes it may be desirable for the head of the laboratory to have men working directly under him doing work of which he sees the importance and which, because of his personal interest, he wishes to handle directly. This is perfectly proper, providing the men working for him come to him frequently and fully understand his objectives. Sometimes it will be found, however, that employees so engaged might work far better under the direction of a lower head. The question of how many heads should exist under the manager depends upon the laboratory and the capacity of the heads under him. It may be better to have only two or three heads and let them subdivide the work. There is no hard and fast rule relating to subdivision. It ought not, however, to be carried so far as to destroy the unity of the whole. Certain policies must be determined by the manager of the laboratory or must have been determined through long association with the work without formal statement on his part, Variations of these policies should and will, in every properly organized laboratory, come in for review.
METHODS OF HANDLING RESEARCH PROBLEMS Sometimes one man may have assigned to him a great many problems, or he may have one problem only, which may be large or small. Every research problem has its objective, and very often, if the objective is properly stated, the problem is more than half solved. There may be, however, a series of objectives evolved in the course of the study of the main objective. We therefore see that problems subdivide themselves and represent a process of evolution. First a project is laid out and assigned to one man, with or without assistants. His study should first involve a careful review of the literature and work already done in the laboratory. Then work is started. After a certain amount has been finished, a preliminary report is presented which indicates the next step or steps to be taken. I n fact, it is quite usual that a number of separate consecutive steps are found to be involved. Each of these is an outgrowth of the original project. It is doubtful whether, in the development of the study of a research problem, any logical order prevails. If logic does prevail, it is probably noted after the fact. The important thing, however, whether logical or not, is the proper laying out of research projects. Each one should be carried to such a point that i t would indicate some other research project, unless indeed, as is unusual, the matter is completely and satisfactorily solved. Let me emphasize the thought that is here outlined: every research problem passes through a state of evolution, and each step becomes an individual offspring of the original project and deserves a certain amount of identification by
Vol. 24, No. 1
a separate report. What its progeny may be should also be indicated.
LABORATORY PLANNING The laboratory is the housing and may be the home of research operations. It must be laid out with regard to the field covered by the research. Just how elaborately it shall be built and equipped depends upon the money available and the purpose which those authorizing the expenditure of money may have in mind. All of us are familiar with the imperfect and inexpensive equipment used by the early chemists. We do not want a research laboratory to be dependent upon such facilities, and yet the question may properly arise as to how far research laboratory expenditures should go in providing expensive edifice and expensive equipment. I really have no opinions on this subject; “handsome is as handsome does,” and I am sure that all real research workers are better pleased by proper room and facilities than they are by fanciful decorations. The building should house a suitable library. It should provide general office space and telephone connection between departments. It should have a room for the receiving of supplies and preparation of samples; and refuse should be delivered daily to this room. The question of subdivision is a debatable one. Much space is often lost by such subdivision, Those rooms assigned for certain men to Tvork in may be unavailable for other use. For this reason, although I may be wrong, I have resisted the idea of subdivision. A certain number of private offices are necessary. One of these private offices, perhaps, can be used for relatively large gatherings, if desirable. But to divide the main floors where research is being done, into separate rooms simply for the purpose of segregating the workers and their work does not appeal t o me. Certain subdivisions must be made where offensive gases or vapors are given off, and where such cannot be properly drawn away, but beyond this the utility of space is often impaired by walls and corridors.
ORGANIZATION OF FINANCES An industrial research laboratory may work under the budget system or without it. KO matter which system is followed, occasional expenditures must be covered by special authorization. Both systems have their disadvantages. Budget systems imply previous knowledge of work to be done during a coming year which is hard to foresee. Without the budget system there is an implication of confidence in the economical operation of the laboratory. T h a t ever system is adopted, there should be a careful record made of the expenditures for salaries, supplies, rent, insurance, etc., and these expenditures should be classified with as much detail as can be reasonably expected. By this classification one is able to compare expenses from year to year, and, if increases are noticeable, a check can be made t o see whether such increases are justifiable. A complicated system of accounting is unnecessary. The purchase of apparatus and supplies should not be made without the approval of the manager or others in authority, when expenditures exceed a certain stated amount for any one item or group of closely related items. Economy must be exercised. The man in charge of purchases must be continually on the alert to determine whether apparatus ordered cannot be supplied by a transfer rather than by purchase. Where a given research problem has been studied arduously for a sufficient length of time without practical results, consideration should be given in conference as to whether further expenditures are justified, for how long, and for how much.
January, 1932
ISDUSTRIAL AND ENGIKEERIKG
CONCLUSION A research laboratory is a n organization made up of human beings with all the advantages and limitations which that involves. I n my opinion, the more really human are the members of a research organization, the more successful it will be. Or, to put it another way, the lese; it operates along the lines of discipline, such as we think of in connection
CHEMISTRY
71
with the army, the more successful it will be. The enthusiasm of a worker-must never be dampened, although properly it 111a) directed. He should always be encouraged when he shows a thirst for knowledge. All that may Le necessary is tactfully to guide his thirst for knowledge to the particular problems on which he should be working. RECEIVED .4ugust 29, 1931.
Flow of Petroleum Lubricating Greases4 34. H. ARVESON, Standard Oil Company of Indiana, Whiting, Ind.
I
of rate of shear is involved in the T HAS long been recognized PETROLEU?II lubricating greases' are disuse of lubricants. I n dispensing that the actual lubrication pensed and used as lubricants under a variety and in feeding from cups and of rubbing surfaces must be of conditions in which the factors determining the boxes the rate of shear involved accomplished by a fluid film if flow characteristics are of primary importance. is usually in the comparatively satisfactory results are to be Based on a norel principle, a aiscometer which low range from 0-1000 seconds- ' obtained. Xot so generally recand in the lubricating film from ognized is the fact that, a-hen predetermines the rate of shear, and which is 1000 seconds-' to indefinitely greases are employed, the luespecially designed f o r the purpose of measuring large values. brication must be accomplished the $0:. characteristics of lubricants, is described. The foregoing shows the imby a fluid film, rather than by The data at 77" F. (25" C.) on several worked portance of providing an instruthe soap structure itself. As a cup greases and a p u l p oil are presented in ment capable of producing varirule, greases are used as a matter ous rates of shear over a very of convenience in handling and graphical form. The large range of rafe qf shear large range. The numerous conapplication, rather than for any reported (0.08 to 132,000 reciprocal seconds) couers trol instruments, though very peculiar lubricating properties. the complete practical range of use. Among useful in their fields, are not hlany m e c h a n i s m s are conothers, the following conclusion is drawn: The suitable for this research because structed so that oil lubrication apparent aiscosity of greases decreases uith inof their restriction t o a narrow would be either impractical or range of rate of shear, usually extravagant. For this reason creasing rates qf shear in a manner characteristic very low in magnitude. I n most the lubricant used must possess of the particular soap used, approaching in the cases it is impossible, in the abthose characteristics which perlimit a value higher than, but of the same order sence of other data, to translate mit it t o be held in containers of magnitude as, the oil in the grease. the data from such instruments which are not liquid-tight, and to absolute units because of the to feed to the surfaces only when actual motion occurs. TLe consistency requirernents of differ- complex function of rates of shear involved. Particularly illustrative are the Saybolt tube, with its variable head, and ent mechanisms are almost as varied as the mechanisms themselves. I n the past no basic data of sufficient completeness the A. S. T. M. cone penetrometer, with its double-taper have been available which mould permit accurate e-t' q imacone. The hfachfichael and Stormer instruments come a little nearer to being applicable to the problem but are only tion of the needs of any particular case. suitable for use on thin greases. and cover only a comparaIMPORTANCE OF RATEOF SHEAR tively narrow range in rate of shear. The use of applied pressure on liquids and plastics in capilThe viscosity of a lubricating oil is its most important property, being fixed if the temperature and total pressure lary-tube viscometers to effect various rates of flow is well are fixed. (Variation with total pressure is ~ e r ysmall in known ( 2 , 6). Pressure-riscosity data have been obtained the normal range.) The analogous property of a grease, on a variety of compositions, including paints, lacqucrs, termed the "apparent viscosity," is not fiied under the above emulsions, and gels of many kinds ( I , 2, 6). I t is the comconditions but varies with the rate of shear. Table I shows mon experience of workers in this field that, when these the rate of shear in a simplified case of a film between a 2- methods are applied to lubricating greases, certain difficulinch (5-cm.) journal and concentric with its bearing when ties are frequently encountered. Grease lumps and air pockets alter the rate of flow, and blow holes occur in the the speed and clearances are as specified. charge. These occurrences are evident only from the fact TABLE I. RATEOF S H E 4 R IN CONCENTRIC I3EARING that the data do not check a previously determined value. (Speed, 1800 r. p m.) Bulkley and Bitner ( 5 ) managed to circumvent the difficulty CLE4RAXCE RATEOF SHEAR of irregular flow by attaching a perfectly horizontal caliInch (cm) Seconds-' 18,800 0 . 0 1 (0.025) brated capillary through a trap to the metering capillary 188,000 0.001 (0.0025) and noting the rate of movement of a mercury plug therein. 1,880,000 0.0001 (0.00025) Their instrument has the further advantage that the same I n the simple case of a material flowing through a tube of sample can be used repeatedly. l/a-inch (0.32-cm.) bore at a rate of 3 cc. per second (0.4 CONSTANT-SHEAR VISCOMETER pound per minute), the rate of shear is approximately 1000 seconds-'. It appears then that an extremely large range To avoid the above difficulties and to effect greater precision and ease of operation, a method, believed to be novel, 1 The term grease is used throughout in the restricted eense of a soaphas been developed, consisting simply in fixing the rate of thickened mineral lubricating oil.
