342 



National Resources Planning Board 



work who are attemptiug to keep abreast of the times. 

 It is felt that tliis poHcy promotes the understanding 

 and use of our products, and gives us our proportionate 

 share of the increased business." 



Types of Research Organization 

 in Manufacturing 



Industrial research is conducted, according to the 

 letters received, under a wide variety of organizational 

 set-ups. In the simplest cases, common in small organiza- 

 tions, such research as is done is instigated and carried 

 through by some of the same men who are doing the 

 production work itself. In companies large enough to 

 have a separately organized engineering department, 

 research is often a function of that department. The 

 next step is ordinarily the organization of a separate 

 research department, often with materials testing and 

 routine inspection as the backlog of its work. 



In still larger companies one begins to find decentral- 

 ization into more or less autonomous branch plants on 

 a product or a geographical basis, or both. In such 

 cases one often finds an engmeering department asso- 

 ciated with each branch performing research along with 

 other functions for the particular product or area in- 

 volved. In still larger decentralized organizations, 

 there will be a separate research department as well as 

 an engineering department for each branch. 



The ne.xt step in complexity of research organization 

 is the establishment of a central research laboratory' to 

 supplement and unify the work of the separate branch 

 engineering or research departments. In such cases 

 there usually is, at least on paper, a definite basis for 

 an appropriate division of labor between the central 

 research laboratory and the branch laboratories or 

 engineering departments. Either the central laboratory 

 undertakes such work as is of interest to several or all 

 of the producing branches, leaving to the branch labora- 

 tories the work germane only to their own branches; or 

 the central laboratory undertakes the "fundamental" 

 research, namely that not immediately applicable to 

 some current production problem, leavmg the "applied" 

 or "practical" research to those more directly concerned. 



Finally there is often some delegation of research 

 activity outside the industrial organization altogether. 

 This maj' take the form of joint or cooperative 

 research by a group of organizations within an industry, 

 through an association or institute. In such a case 

 the association or institute is likely to concentrate 

 on research intended to improve the competitive posi- 

 tion of the industry as a whole against other industries, 

 though some fundamental research of interest to the 

 industry as a whole may also be undertaken. 



Another form of delegated research, now becoming 

 more common than formerly, is that undertaken by an 

 independent research institute, such as the Mellon 



Institute in Pittsburgh which has served as a model 

 for several others, or by a university or college, under 

 a contract of some sort with an industrial client who 

 bears a large part or all of the research expense. 



An admirable example of cooperative delegated 

 research was the Steam Research Program sponsored 

 and directed by a special research committee of The 

 American Society of Mechanical Engineers, financed 

 by a considerable number of industrial firms, and car- 

 ried on at two universities and at the National Bureau 

 of Standards, which has led to three international 

 steam conferences, to an internationally agreed-upon 

 "skeleton steam table," to revised and much more 

 reliable working steam tables in three different countries, 

 and to greatty reduced uncertainties whenever the 

 performance of steam-driven machinery is discussed 

 across international boundaries. 



Other examples with which the authors happen to 

 be familiar are: An extensive program of research on 

 the art of cutting metals recently concluded, one on 

 so-called caustic embrittlement in boilers and on other 

 aspects of feed-water composition and treatment, one 

 on the characteristics and operation of super-pressure 

 boilers, one on fatigue and one on creep of metals, and 

 one on various aspects of the fluid-meter problem. 

 Still other examples of admirable cooperative dele- 

 gated industrial research can be found in the records of 

 the Engineering Foundation and of several of the major 

 national engineering societies. 



Research in Operation-Type Industries 



Because their product differs from that of manu- 

 facturing industries with a corresponding difference in 

 organization, it is necessary to give different treatment 

 in this report to tlie public utilities, electric, gas, rail- 

 road, telephone, and telegraph. In these industries, 

 as in manufacturing, research permeates every phase of 

 operation, though in varying degrees of formal organ- 

 ization, and the mechanical engineering-research 

 responsibilities are substantial. 



In describing the research functions performed by 

 mechanical engineers in the public utilities, it is nec- 

 essary to modify somewhat the classification used for 

 the manufacturing industries. In the paragraphs that 

 follow all the utilities will be discussed together under 

 the following headings: Materials, operation, new 

 devices and apparatus, and management and pro- 

 motion. 



Materials 



Among the critical problems of the electrical industry 

 the fuel problem looms largest because fuel is the 

 largest single item of material cost in the generation of 

 electricity. The sampling and testing of incoming 

 shipments of coal borders on the routine, but the inter- 



