sign in sign up
<<
Home , Quantity, System of measurement

NAT'L INST. OF STAND & TECH NIST PUBLICATIONS AlllOb 722D0A NBSIR 75-928

The National for , and

Paul E. Pontius, James R. Whetstone, and John A. Simpson

National Engineering Laboratory National Bureau of Standards Washington, D.C. 20234

Final

May 1978

DEPARTMENT OF COMMERCE

)NAL BUREAU OF STANDARDS * 75-119

NBSIR 75-928

THE NATIONAL MEASUREMENT SYSTEM FOR MASS, VOLUME AND DENSITY

Paul E. Pontius, James R. Whetstone, and John A. Simpson

National Engineering Laboratory National Bureau of Standards Washington, D C. 20234

Final

May 1978

U.S. DEPARTMENT OF COMMERCE, Juanita M. Kreps, Secretary

Or. Sidney Harman, Under Secretary

Jordan J. Baruch, Assistant Secretary for and Technology NATIONAL BUREAU OF STANDARDS, Ernest Ambler. Director

CONTENTS

Page

TABLE OF CONTENTS i LIST OF FIGURES iii LIST OF TABLES iii

0.0 EXECUTIVE SUMMARY 1 1.0 INTRODUCTION 3 2.0 STRUCTURE OF THE MEASUREMENT SYSTEM 4-33 2.1 6 2.1.1 Mass 6 2.1.2 Volume 8 2.1.3 Density 9

'. 2.2 Basic Technical Infrastructure . . . 9 2.2.1 Documentary Specification System 12 2.2.1.1 Standardization Institutions 12 2.2.1.2 Survey of Documentary Standards 14 2.2.2 Instrumentation System 16 2.2.2.1 Measurement Tools and Techniques 17 2.2.2.2 The Instrumentation Industry 19 2.2.3 Reference Data 20 2.2.4 Reference Material 23 2.2.5 Science and People 23 2.3 Realized Measurement Capabilities 25 2.3.1 of Reference Standards 25 2.3.2 The Measurement Process 29 2.4 Dissemination and Enforcement Network 30 2.4.1 Central Standards Authorities 30

2.4.2 State and Local Offices of Weights and Measures. . . 30

2.4.3 Standards and Testing Laboratories and Services. . . 31 2.4.4 Regulatory Agencies 32 2.5 Organizational Input-Output Transactions Matrix 33 3.0 IMPACT, STATUS AND TRENDS OF MEASUREMENT SYSTEM 33-38 3.1 Impact of 33 3.1.1 Functional, Technological and Scientific Applications 33 3.1.2 Economic Impacts — Costs and Benefits 36 3.1.2.1 Processes Relating to Quality of Material 36 3.1.2.2 Processes Associated With Quality Control and Regulation 36 3.1.2.3 Processes Associated With Preparations of Objects or Materials 36 3.1.3 Social, Human, Man-on-the-Street Impacts 36 3.2 Status and Trends of the System 37 4.0 SURVEY OF NBS SERVICES 38-55 4.1 The Past 38 4.2 The Present—Scope of NBS Services 39 4.2.1 Description of NBS Services 39 4.2.1.1 Mass Measurement 39 4.2.1.2 Volumetric Measurements — Glassware 39 4.2.1.3 Metal Volumetric Standards 40 4.2.1.4 Density Measurements 40 4.2.1.5 Measurement Assurance Program 40 4.2.1.6 Office of Weights and Measures 40 4.2.2 Users of NBS Services 41 CONTENTS

Page

*. 4.2.3 Alternate Sources . . . 42 4.2.4 Funding Sources for NBS Services 45 4.2.5 Mechanism for Supplying Services 45 4.3 Impact of NBS Services 46 4.3.1 Economic Impact of Major User Classes 46 4.3.2 Technological Impact of Services 48 4.3.3 Pay-off from Changes in NBS Services 50 4.4 Evaluation of NBS Program 51 4.5 The Future 53 5.0 SUMMARY AND CONCLUSIONS 54 References 55-56 Appendix A. METHODOLOGY OF THE STUDY 57-59 Appendix B 60-65 References 66

i i LIST OF FIGURES

Page

Figure 1: Measurement Process Infrastructure 5

Figure 2: Infrastructure Associated with the Metric 5

Figure 3: Voluntary Standards Infrastructure 7

Figure 4: Practical Hydrometer Scales 13

Figure 5: Mass Customer Profile (Major SIC Groups) 43

Figure 6: Mass Customer Profile (Three Digit SIC Groups) 43

Figure 7: Volume Customer Profile (Two and Three Digit SIC Groups) ... 44

Figure 8: Density Customer Profile (Two and Three Digit SIC Groups) . . 44

LIST OF TABLES

Table 1: Voluntary Standards Organizations 8

Table 2: Industrial Effort Relating to Mass, Volume and Density .... 21

Table 3: Commercial Weighing Devices in Use 22

Table 4: Devices Used in States' Weights and Measures Activities. ... 22

Table 5: Tolerances for Commercial Test Weights 27

Table 6: Typical Commercial Scale Tolerances 28

Table 7: Typical Measurement Uncertainties 29

Table 8: Summary for Users of NBS Services 47

Table 9: Petroleum Measurement Commitment 49

i i i I

I THE NATIONAL MEASUREMENT SYSTEM FOR MASS, VOLUME AND DENSITY

Paul E. Pontius, James R. Whetstone, and John A. Simpson

National Bureau of Standards

Mass, a property of all matter, is to be a property of a material, however, unique in the sense that the concept is such extension requires an assumption on generally deduced from the motion of a the homogenity of the sample measured. body under the influence of an applied Considering that most of the measurement force. It is for this reason the mass processes within the mass, volume and unit for measurements over the range of density measurement are based on objects which can be physically handled economic compromise rather than technolog- is now, and for some to come will ical capabilities, it is presumptuous to continue to be, embodied in an artifact assume that NBS has any substantial impact standard. The mass of one object with on the total system. On the other hand, respect to another is deduced from a NBS, by providing a means for access to comparison of the gravitational forces the unit, is generally considered to be acting on the two objects. The additive a partner in all measurement activity. of mass makes it possible to con- Consistency of results commensurate with struct appropriate subdivisions and requirements is a fundamental to all of multiples of a single defining artifact. the activities the system supports. Reso- The present mass unit, the International lution of inconsistencies, by referral mea- Prototype , is one of the Base surements, by suggested process modifica- Units of the International System of tion, or by basic changes in measurement units. Access to the unit is provided by concept is the essential support to the Bureau International des Poids and Mesures system provided by NBS. (Sevres, France) through prototype kilogram In spite of the size of the system, it replicas of the unit. Prototype appears to be in almost complete control No. 20 and No. 4 provide such access for and operating to the general satisfaction the , and are in turn the of all concerned. There are several rea- basis for the U.S. practical mass scale as sons for this: (1) mass standards are established by the ni chrome kilograms Nl intrinsically stable so that the unit once and N2. injected into a subsystem remains reason- Volume is a quantity derived from the ably consistent; (2) the instruments used unit. It is inseparably coupled are reliable; and (3) the size of large with mass. The intent of the founders of subsystems concerned with voluntary stan- the was that one cubic deci- dards, weights and measures activities, meter of water at maximum density have a government requirements, etc., result in mass of one kilogram. The realization was inertia with respect to change. Through proven to be imperfect, thus introducing the mechanism of Measurement Assurance for a period of time the liter as a volume Programs, NBS or any other interested unit. (The current unit for volume is the party, can test the effectiveness of a cubic meter). Precise mass measurement given subsystem relative to the intended requires a knowledge of the displacement use of the measurement results. The volume of the objects being compared. same medium provides a means for a running Hydrostatic weighings are used to (a) audit of the subsystem to insure that determine the displacement volume ratio stability is not lost. relative to an object with displacement With the advent of the NBS-2 kilogram volume established by direct measurement comparator the precision of measurement and (b) displacement volume relative to has reached the where factors a known density of the fluid. affecting the short term stability of the Density is defined as mass per unit defining artifact can be studied. These volume. Density is a property of factors include cleaning procedures, mois- interest for a variety of fluid mixtures. ture films and environmental contamination. The density of a fluid in conjunction Problems associated with the air density with mass measurement is a means to algorithm, e.g., oversimplification and determine the containment volume of a non-uniformity, introduce small offsets variety of volumetric standards. Mass and in the practical mass scales of the vari- volume are both used to determine quantity ous countries. These offsets, too small of material. Density is usually considered to be of concern in most weighing

1 operations, do not affect the consistency of relates to specific problems, the practical measurements in a given country. resolution of which has no visibility. Resolution of the problem, however, is Generally, most of these problems relate to important in support of work on fundamental specifying what the measurement is supposed physical constants. to accomplish and there are those who There are certain identifiable expect NBS to address his particular constituencies within the system which rely problem, however ill defined. exclusively on NBS services. Support to On the other hand, there are pressures the State Weights and Measures activities, to work through an interface (identifiable coordinated by the NBS Office of Weights constituency) both from within NBS and and Measures, extends to about 84,000,000 from other sources. Working through an weighings in supermarket chains alone, interface tends to isolate NBS from prac- which in part supports a $145,100,000 a tical measurement problems. Interfacing year industry which provides the needed through a system of facilities which are weighing devices. The petroleum industry primarily regulatory raises the question has a worldwide investment of $1,800,000,000 as to whether such a system can provide in volume measurement facilities at 650,000 advice and guidance with regard to methods sites, manned by 50,000 full time person- and procedures without compromising its nel. Each "barrel" is measured on the regulatory function. The legalization of average of 12 between the well head all measurements, as proposed by the Inter- and the consumer, with either custody national Organization for Legal , transfer or accountability between each diverts attention from the measurement pair of measurements. All measurements on result to traceability of the unit. In local or imported crude and products are some instances, the consumer of measure- based on NBS calibration of metal volumet- ment data is relieved from the burden of ric provers. In the beverage industry, proving that the results are adequate for approximately $2,500,000 is expended the intended purpose. In other instances, annually for density measurements in the there is an added burden of not only prov- production of $15,400,000,000 of products. ing that the measurements are not only NBS is providing support in the develop- adequate but also legal. ment of procedures for both custody trans- The above listed problems not fer and accountability of nuclear material. withstanding; the discharge of NBS respon- The Mass Measurement Assurance Program pro- sibilities in the of mass, volume and vides direct support to the Department of density, is judged to be satisfactory. Defense measurement systems, measurement Except for the maintenance of certain staff systems within the nuclear and aerospace skills, the level of effort now displayed industries, and to manufacturers of results in slow but steady progress. The precise weighing equipment. nature of the services provided and the Problems in the interface between the role of the section will change as its Mass and Volume Section and the rest of aging staff moves into retirement. The the measurement system are, in general, course of activities for the future matters of judgment. The section, recog- includes support to the system as it exists nizing the service roll of measurement, has at the time, and a continuing effort to worked diligently to establish and maintain improve the efficiency of measurement. a line of communication with the individual users (a non-coherent constituency). Such an approach requires not only an under- standing of measurement process limitations but also a clear definition of what is to be expected from the measurement effort. The latter requirement is the most diffi- cult, and where it has been possible to define acceptable limits relative to a particular undertaking there has been visible evidence of success, e.g., savings to the Food and Drug Administration rela- tive to the calibration of glassware, the adoption of measurement assurance proce- dures within the nuclear weapons complex, and the efforts supporting accountability and safeguarding of nuclear materials. Assistance provided to large numbers of users through consultation generally THE NATIONAL MEASUREMENT SYSTEM FOR MASS, VOLUME AND DENSITY

Paul E. Pontius, James R. Whetstone, and John A. Simpson

National Bureau of Standards

1. INTRODUCTION incorporate knowledge of the volume or den- sity of the objects of interest as well as Weighings for purposes of trade were made the density of the surrounding environment. at least as far back in time as the Early The qualities of mass, volume, and density Bronze Age and a well developed system of are related by the following definition: standards and enforcement of standards was in effect in Egypt by the dynasty. MASS = VOLUME X DENSITY The use of the balance—as a symbol of fair- ness and justice — dates from the same mists One can determine one of these quantities if of antiquity [1]. While it was common know- the other two are known, e.g., the mass of ledge that early governments used one set of petroleum contained in a tanker can be de- weights in the collection of goods for taxes termined from measurements of tank and another to disperse goods for services, and petroleum density. by simply interchanging the loads on the Maintenance of the unit of mass is, by the balance pans, it was easy to demonstrate Treaty of the , the responsibility of that there was nothing wrong with the bal- the Bureau Internationale des Poids et ance. In normal commerce, other than the Mesures (BIPM). The responsibility for usual haggling over the details of the maintenance of the for mass and transaction, the measurement process worked the units for volume and density in the well. However, with the first technological United States is one of the oldest and most advance which made weighings rapid and con- basic missions of NBS. It remains a unique venient, the steelyard thought to be intro- responsibility since the mass unit is the duced by the Romans, the situation changed. only Systeme Internationale (SI) unit which The buyer was at the complete mercy of the in principle is not independently reproduc- seller who could alter the instrument at ible. Of the two Prototype Kilograms desig- will, changing either the sliding , or nated No. 4 and No. 20, received by the U.S. the lever ratio. The widespread usage of under the Treaty of the Metre, Prototype the steelyard in commerce was the genesis of Kilogram No. 20 is the de facto United regulatory activities relating to weights States' unit of mass [5]. In , K4 and weighing devices. It was the diversity and K20 are merely a mechanism for transfer- of local units, staunchly defended by the ring a defined unit from one mass measure- tradesmen, that ultimately led to the devel- ment system to another. opment of the metric system of measurement The derived unit of volume is now ex- uni ts. pressed as length cubed. High precision Mass, in the context of mass measurement, measurements of artifact displacement, or is based on physical laws formulated by containment, volumes are determined from [2], thus relative to the history of length measurements. Such measurements are weighing, and even the metric system, the complex and rarely done, e.g., the work of concept of mass is a recent one. Because of Cook [6,7] in measuring the density of mer- the size of the solar system, Newton could cury and that of Bowman, Schoonover and consider the planets and their satellites as Carroll in measuring the density of single points in the of space [3]. The con- crystal silicon [8]. Practical measure- cept of mass was essential to the formula- ments of displacement volumes use Archimedes tion of his laws of motion. As a consequence, Principle, e.g., hydrostatic weighing in a mass cannot be measured directly in the fluid of known density. normally encountered environment. Practical volumetric measures are vessels For example, a balance, contrary to popu- whose volume is defined by weighing the ves- lar opinion, is a force comparison device sel filled and empty, using a liquid of which cannot separate the gravitational known density. This operation inextricably force acting on the object from the buoyant ties volume measurements to mass and density. effects of the atmosphere [4]. While one Where weighing operations are impossible, can compare these forces to a very high pre- such as in determining the capacity of large cision, an algorithm must be employed to sep- volume tanks, direct dimensional measure- arate the two effects. This algorithm must ments ("strapping") are used [9]. In these —

cases, the errors normally encountered in the link is as short as is practicable. Each the length measurements are very small rela- route has its own unique characteristics. tive to the total volume of the tank. The ultimate goal in any case is to obtain The derived unit of density is currently a consistency of measurement compatible with expressed as kilograms per cubic meter [10]. the manner in which the measurement is to be (In this case the SI unit is so large that used [11]. it has little practical meaning.) By far, the majority of density measurements are 2. STRUCTURE OF THE MEASUREMENT SYSTEM made for determination of the characteris- tics of substances in the liquid phase. A The commonality of certain elements in variety of arbitrary units are used, each specific measurement processes provides a associated with a particular class of logical basis for ordering groups of measure- liquid. The simplicity and precision of ment processes into systems and subsystems. the hydrometer is a practical means of To illustrate the ways in which the system is quality control for any products produced or "tied" together, figure 1 is a diagram of the marketed in liquid form, ranging from milk factors which affect the design and operation to alcohol -water mixtures and from petro- of a specific measurement process. The task leum and its products to paints and acids. requirements, which arise from the desire for Derived units involving force, e.g., success, or the consequences of failure, are direct force measurements, pressure measure- the basis for the details of the process. ments, electric strength measurements, These requirements, as perceived at the mea- are directly tied to mass measurement since surement level, establish the specific process mass is one of the base SI units. Each of and procedural details. In operation, there these is covered in other surveys, is a value judgment for each measurement nonetheless, the consistency within accept, scrap or rework items, or remeasure. and between these seemingly diverse fields In the end, the item and result pass on to is directly related to the characterization others and in combination with many other and dissemination of the mass unit. elements, i.e. results from other measurement Measurements of mass, volume, and density processes, contractual requirements, entre- are undoubtedly the most pervasive measure- preneurial judgments, etc., constitute the ments performed in the world. Weighing oper- completion of the task. Economic factors, ations are involved in activities ranging success/failure ratio, changes in the con- from teaching addition in the elementary sequences of failure, and changing tech- grades to establishing the mass of a testing nology are all "feedback" factors contributing device for use in determining the mass and to refinement of process detail. center of gravity of space vehicles and mass The one element common to all mass, volume determination of objects used in current NBS and density measurement processes is the con- measurements of the universal gravitational ceptual basis for the measurement. This ele- constant, G. In present commerce, almost ment, however, is more devisive that adhesive all commodities are priced by weight, as are relative to the total system. For example, materials more precious than gold and cheaper as the task requirements, either real or than dirt. Most physical and chemical proces- perceived, approach the limit of measurement ses are mass based. Chemical reactions, as capabilities, each step in figure 1 must be well as the of nuclear reactions, treated explicitly—not only relative to the depend on mass ratios. Moreover, techniques specific process but also relative to the exist for conveniently performing weighings other. elements contributing to the accom- over a range of 10 , literally, almost plishment of the task. Those who work in this nothing is too large or too small to be area become, by necessity, "understanding" weighed or to be similarly described through oriented. In contrast, when the requirements the related properties, volume and density. are modest with respect to capabilities, many The almost infinite number of measure- of the steps can be treated implicitly and ments of mass, volume, and density in the those who work in this area become United States tie to the Mass and Volume "procedure" oriented. In the first instance, Section of NBS in many different ways. In the interest is in "how much," i.e. the mass some cases the relation is established by of an object, the volume of a tank, or the legal, or contractual, requirements. In density of a material. In the latter other cases, the relation is either a matter instance, the interest is in how much one is of necessity or of convenience. In some cases in error if he assumes that the mass of an the link between the U.S. Prototype Kilogram adjusted weight is 1 kilogram, or the and the point of judgment as to the accept- capacity of a "certified" prover is 1 , ability of a given measurement relative to a and so on. specified requirement is through a hierarchy The most visible common element is the need of calibration laboratories. In other cases for access to a consistent set of metrics. .

Another common element, which to date has no able bureaucratic infrastructures relate to identifiable bureaucratic structure, relates providing access to consistent metrics ade- to the availability of adequate instrumenta- quate for the task at hand, and to the value tion. Finally there is a commonality in the judgments for similar measurements (voluntary basis for the value judgment, particularly standards ) where there are many processes engaged in the The infrastructure related to the metric same or similar tasks, and where the measure- is shown in figure 2. All who have need have ment requirements are modest. The identifi-

Figure 1 INFRASTRUCTURE OF THE MEASUREMENT PROCESS CONCEPT SCIENTIFIC AND TECHNICAL BASE

MEASUREMENT PROCESS INSTRUMENT PROCEDURE METRIC

ERROR ITEM RESULT LIMIT

SCRAP AND VALUE PERCEIVED TASK DISPOSITION REWORK JUDGEMENT REQUIREMENTS REQUIREMENTS

ACCEPT

SUCCESS REFINEMENT OTHER TOTAL ELEMENTS TASK FAILURE CONSEQUENCE

Figure 2 INFRASTRUCTURE OF THE METRIC

DEFINITION OF UNITS

t — BIPM* (KILOGRAM)

NAT. LABS OF ^(REF^DATA) MASS OTHER | VOLUME NATIONS DENSITY J

PRECISION EQPT. USERS (A) MFR'S.

USERS (B) USERS (C)

BIPM = International Bureau of Weights and Measures, Sevres, France OWM = NBS Office of Weights and Measures .

access to adequate mass standards, either ill conceived, there is little to be gained from NBS and similar laboratories of other by imposing change. On the other hand, nations, and in some cases, directly from technological progress may eventually BIPM, or from other sources served by the dictate requirements beyond the capabilities echelon shown. The first level dissemination of currently accepted practices, at which from NBS provides access to those with real time the old processes must be discarded and needs (USERS (A)), or perceived needs (USERS a new and more appropriate one devised. Thus (B)), which require direct access to national because of the diversity of requirements at standards. Precision equipment manufacturers any point in history there are (a) subsystems are also served. These in turn provide in- which are stable with formal infrastructures; struments and standards which are themselves (b) existing processes which, for various adequate for USERS (C). The industrial or- reasons, are undergoing change with resulting ganizations served in turn provide services pressures on existing infrastructures; and for a restricted set of users (USERS (R)), (c) processes being developed for emerging re generally limited to groups within the quirements which are not as yet sufficiently particular organizations. Government labora- stable to permit the development of a formal tories, usually organized in hierarchial i nf rastructure. echelon systems, also provide access to a restricted set of users. Finally, through 2.1 Conceptual System the Office of Weights and Measures, the state laboratories provide services for their 2.1.1 Mass inspections and in most cases to unrestricted users, USERS (C). It should be noted that It is very difficult to construct a con- both USERS (B) and USERS (C) include equip- ceptual mass measurement system which would ment manufacturers serving measurement areas be acceptable at all levels of interest. in which the requirements are modest. From the beginning, and even now, most The infrastructure associated with value weighings are associated with determining judgments; i.e. voluntary standards, regula- quantity of material. Highly developed tions, etc., is considerably more complex. weighing devices were widely used long be- The task is to establish a consensus agree- fore the reasons why such instruments were ment between the user-supplier, or the successful were understood. The concept of regulator-regulated, as to what constitutes mass does not exist without motion. Few, an adequate measurement under a specific set however, would interpret the simple act of of conditions. The procedural steps leading placing a weight on a balance pan as motion. to a voluntary standard are shown in figure 3. Mass and force are inseparable, but an in- The time span from start to general acceptance tellectual separation has been maintained is often in excess of 10 years. If the stan- by definitions, e.g. mass is measured by a dard is in response to a regulatory require- balance; force is measured with a spring. ment, the regulator may reject the entire While it is only necessary for a conceptual effort at any time. Obviously, once a stan- system to be logically correct, there are dard becomes an accepted practice, it is ex- good reasons for it to also reflect reality. tremely difficult to make substantial changes Few are motivated to pursue measurement in spite of the fact that most standardizing detail beyond an understanding of some organizations require periodic review. The conceptual system. Decisions are then made standardizing organizations, with interest on the basis of this understanding. If the in mass, volume and density measurements concept as perceived is in error, then the are shown i n table 1 decision may also be in error. The structures described in figures 1,2, The mass of a body is revealed by a and 3 are frameworks within which one can body's resistance to change in state of discuss various aspects of mass, volume, motion, i.e. inertia, or by the fact that a and density measurement. The actual measure- force of attraction exists between any two ment system is far more complex. Unlike mea- bodies, that force being proportional to the surement systems which originated around product of their and inversely pro- the turn of this century or later, mea- portional to the square of the be- surements of mass, volume and density started tween their mass centers, i.e. gravitation. with the dawn of technology. These measure- For all practical purposes, the same pro- ments are necessary to the accomplishment of perty of a body is responsible for both ef- about every human endeavor, be it by an indi- fects, thus consistent measurements systems vidual, a profession, or by a society. Funda- can be constructed using either effect. mental to these accomplishments are measure- Mass measurements on, or near the surface ment processes that work well enough for the of, the earth are based on gravitational intended purposes. Where the resulting effects. Body mass changes of astronauts in processes are adequate, regardless of being orbit were measured with systems based on inertia! effects [12]. For a conceptual mass The traditional concept of mass equality, measurement system, it is sufficient to "Two bodies have equal mass if, when placed state simply that mass is one property of on the opposite pans of a "perfect" equal any object. arm balance, result in the balance arms being horizontal," or paraphrases thereof, has been a serious retardant to the under- standing of mass, and mass measurement, and the development of instrumentation. If one Figure 3 accepts that mass is a unique property of an INFRASTRUCTURE OF THE object, then, for the purpose of mass mea- VOLUNTARY STANDARD PROCESS surement, the object can be considered as a mass point located at the mass center of the object. For two such mass points, if the IDENTIFY attractive gravitational forces at the same REQUIREMENT location in the earth's gravity field are equal, then the two points have equal mass properties. If the response of the two DEVELOP points to the application of a given force TECHNOLOGY is the same, the points also have equal mass properties. For mass measurements in the earth's TENTATIVE gravity field, any force measurement device *] STANDARD COMMITTEE with sufficient resolution, be it a tradi- OF tional two pan balance, a nonsymmetric one INTERESTED pan balance, an elastic device, a pressure TEST AND PARTIES device, or whatever, is a candidate for a EVALUATE mass . Over the past ten years, the traditional equal arm balance has lost its position of prominence, even at DRAFT the highest level of mass measurement, that STANDARD of providing access to the International Prototype Kilogram [13]. There are now a variety of high technology weighing devices SUBMIT FOR which are convenient to use and are adequate REVIEW for practically all requirements. To complete the conceptual measurement system, one must be able to extend or sub- RESOLVE divide the unit, i.e. one needs a defini- COMMENTS tion for addition. For a collection of objects fixed relative to each other, there is a unique point at which the total mass of PUBLISH AND the collection can be considered to be con- DISSEMINATE centrated. In terms of measurement, this is just another mass point, differing only in quantity. It remains only to have the force TEST sensor respond to the gravitational force acting at this point. Simply stated, the mass of two objects together on a balance ALL pan is the sum of the masses of the individ- GENERAL INTERESTED ual objects. In reality, considerable care ACCEPTANCE is exercised in the design, construction, l PARTIES and operation of weighing devices to assure I that with a single object, or a collection REVIEW of objects on the "pan," so to speak, the effective mass center of the "pan" load has a fixed position relative to the force sen- sor and the direction of the gravitational field. Considering an object, or collection of objects, as a mass point in the concept in- troduces ambiguities in the realization of the measurement. When these discrepancies are important relative to the intended

7 .

Table 1. Standardizing Organizations with Interest in Mass, Volume and Density

Organization Classification Area of Interest International Treaty of the Meter (BIPM) Treaty Units International Standards Organization (ISO) Suppliers Marketing International Organization for Legal Metrology (OIML) Treaty Legal Metrology Government and National Department of Defense (DoD) Government Procurement General Services Administration (GSA) Government Procurement Nuclear Regulatory Commission (NRC) Government Material Control American Society for Testing Materials (ASTM) (1) Procedures, Materials, American National Standards Institute (ANSI) (1) Instruments, Artifacts Professional National Conference of Weights and Artifacts, Instruments, Measures (NCWM) (2) Procedures National Conference of Standards Laboratories (NCSL) (2) Procedures, Administration American Petroleum Institute (API) Suppliers Petroleum Products Scientific Apparatus Manufacturers Association (SAMA) Suppliers Apparatus Institute for Nuclear Materials Nuclear Materials Management (INMM) Suppliers Control

(1) Requires equal participation between users and suppliers (2) Users and suppliers participate

usage of the result, a treatment of the ap- Displacement volumes can be computed from propriate ambiguities must be included in direct measures of the exterior dimensions the definition of equality. The major am- of solid objects. However, unless the ob- biguities are associated with the displace- jects are of very uniform , the re- ment volumes, air density, surface finish, sults are relatively crude. When this and the response of the surface of the ob- method is used, the geometry of the objects jects to changes in environment, i.e., cor- is generally limited (cube, cylinder or rosion, moisture layers, etc. In some sphere) with surface finish adequate for instances, the difference in the position length measurements by interf erometry. Vol- of the mass centers of the objects being ume determinations by length measurement are compared, relative to the mass center of the usually limited to two classes of measure- earth must also be considered. ment: those associated with determining the density of various reference liquids, e.g. 2.1.2 Volume water and mercury, or those associated with very large volumes, e.g. large objects and The concept of volume is the quantity of liquid storage tanks. three dimensional space enclosed by a closed Precise displacement volumes of a variety boundary surface. Therefore, its dimensions, of forms can be obtained by Archimedes' as a derived unit, are those of length raised Principle, e.g., hydrostatic weighing in a to the third . Volume is an additive fluid of known density. Volumetric coeffi- quantity. The major ambiguities of concern cients of expansion enter in two places: for in the realization of the measurement are the object and for the density of the fluid. associated with definition of the closed The surfaces of the object must be compati- boundary surface and . For each ble with the fluid so that the boundaries material, volumetric changes with tempera- remain clearly defined without causing the ture change are related to length changes mass of the object or the density of the with for that material (usually fluid to change. One of the largest ambigu- expressed as volumetric coefficient of ities in this method is associated with air expansion) bubbles which are entrapped on the surface of the submerged object which, in turn, regard to a reference material has an addi- cause apparent changes in the mass of the tional complexity over and above those asso- object. The method is also subject to the ciated with measurement: that of the normal ambiguities associated with mass mea- uniformity of the reference material over surement. Generally, the method is limited time and as obtained from various sources. by the availability of equipment and handling problems associated with the object. 2.2 Basic Technical Infrastructure The largest number of volume measurements are associated with determining the amount Mass of liquid which is "contained" in, or can be The accepted standard mass unit is the "delivered" from, a variety of appropriate mass of the International Prototype Kilo- vessels, i.e., pipets, burets, flasks, metal , a pi ati num-i ridium cylinder of height containers or capacity standards, process very nearly equal to its diameter. The mass tanks, etc. When the container design is of this object is exactly one kilogram by such that it can be weighed in the filled definition. There are no environmental res- condition, the containment volume is deter- trictions associated with the definition. mined by gravimetric methods. The container In the comparison of the International Pro- is first weighed in an empty condition, then totype Kilogram with its replicas, the am- filled to an appropriate index with a liquid biguities mentioned above are not large, of known density and reweighed. The volume provided that sufficient time is allowed of the liquid, determined from the defini- for the surfaces to come into equilibrium tion of density, is assumed to be the volume with the environment since the objects are of the container. The method is subject to made from the same material with nearly the all of the ambiguities associated with mass same surface finish. and volume measurements, as described above. The acceptance of the International For "delivered" volume, there is an ambigui- Prototype Kilogram as the standard of mass ty associated with the tendency of the liq- did not end the confusion associated with uid to cling to the walls of the container mass, weight and force. In 1901, the 3rd during drainage. In this case, the con- General Conference on Weights and Measures tainer is filled and drained for a specific (CGPM) declared: time interval before the first weighing. 1) "The kilogram is the unit of mass; With proper attention to the ambiguities, it is equal to the mass of the one can establish the volumes of large con- International Prototype Kilogram;" tainers by transfer methods, e.g., a repeti- 2) "The word weight denotes a quantity tive operation consisting of filling a con- of the same nature as force; the tainer of known volume to an appropriate weight of a body is the product of index, and emptying it into the larger its mass and the acceleration due container. to gravity." This second statement, a deviation from the 2.1.3 Density general policy of the CGPM of defining only units but not quantities, demonstrates the Density is defined as mass per unit vol- wisdom of this policy. ume. It is not a measurement unit in the The 11th CGPM, in 1960, established the same sense as mass and volume. It is a unit derived unit for force, the Newton, which is generally used to characterize a material, defined as the product of mass and accelera- be it gas, liquid or solid. In certain tion. This action, in essence, relegates cases density may be a measure of mass dis- the word "weight" to mean an object which tribution within the boundaries of the en- has a mass property of interest. In reality closed volume, e.g. certain gas and gas the words weight, mass, and force are still mixtures, certain pure liquids, and certain syntactically confused. The to which solids. In many cases density is merely a the populace, as well as those directly in- useful characteristic of an object. terested in mass and force measurements, Because of large numbers of ambiguities accept the Newton remains to be seen. which are involved in both mass measurement, In this country, the most common metrics and direct measurement of volume using a for mass measurement are the avoirdupois length metric, the most precise density work , with a growing use of decimal subdi- has been limited to the characterization of visions, and the kilogram, with the usual specific reference materials: water, because subdivisions. Traditional subdivisions of of its prominence in the establishment of the pound are required by legislation to be the metric unit of mass and its availability used in some areas, e.g., the Troy pound and as a reference material; mercury, for use in are used in precious metals trade; pressure measurement; and most recently, grains, in coinage. All now have been single crystal silicon. Density work with defined, directly or indirectly, relative to the kilogram. There are no inconsistencies called "apparent mass," and was assigned between the various metrics. on the basis of the observed differences. In determining the mass of an unknown Later, the precision of the "apparent object, the metric can be introduced in one mass" scale was preserved by normalizing of two ways; by direct comparison with an the data to obtain the value one would appropriate summation of mass standards, expect if the unknown had been compared relying on the instrument indicating system with a hypothetic standard of stated to subdivide the smallest practical stan- density in a specific environment. The dard, or by direct measurement using a de- oldest "apparent mass" scale used in this vice which has been designed to indicate country is based on a standard of either mass units directly or some mass re- 7.8 g/cnr at 0°C and an air density of lated quantity of interest, or both, e.g. 1.2 mg/cm . The newest scale, originating the grocery store pricing scale. The choice in Europe and Japan, is based on a stan- of method depends upon the end usage of the dard of density 8.0 g/cm^at 20°C and an 3 measurement results. air density of 1.2 mg/cm . The system- In the comparative mode, the constancy of atic difference between the two scales is the metric is determined by the available observable on many instruments. Finally, summation of mass standards. The uncertain- there is the "Newtonian" mass scale, e.g. ty of the result is a function of the uncer- "true mass" or "mass in vaccuo," for which tainty of the mass value assigned to the the International Prototype Kilogram is standards, the precision of the instrument, the defining artifact. the care exercised in making the measure- ment, and the attention given to significant Volume ambiguities. In the direct reading mode, The founders of the metric system, in an the constancy of the metric is assigned to attempt to establish a reproducible mass the instrument. The uncertainty of the re- unit, defined a mass unit, the gram, to be sult is then a function of both the long the mass of one cubic centimeter of pure term stability of the instrument and its water at the temperature of maximum density. precision, as well as the uncertainty of the On the basis of this definition, and the mass scale as embodied in the instrument. newly established length unit, the meter, The instrument design usually includes some the Kilogram of the Archives was construc- compensation for significant ambiguities. ted using arbitrary mass units as embodied It remains for the user, however, to ascer- in the Pile of Charlemagne [14]. Using tain that such treatment is indeed appropri- Archimedes' Principle, and objects with ate for his particular measurement. In measured displacement volumes in terms of practically all cases, commercially avail- the meter, the summation of arbitrary mass able instruments are designed to be used as units which would be equivalent to one direct reading devices. Most instruments, kilogram was established. It then remained however, can be used in either mode of to construct and adjust the newly estab- operation either as is, or with minor lished kilogram to be equal in mass to the alterations, if the instrument indi- appropriate summation of arbitrary mass cating system has not been arbitrarily units. This work was a substantial scien- rounded. tific effort, with careful attention to The most confusing technical problem the ambiguity associated with the buoyant concerns the basis for stating mass values. effect of the air. There was no way to In addition to confusion between mass, check the realization however, since the weight and force as mentioned above, there available platinum-i ridium alloy used in are at least three "mass scales" in common the construction of the Kilogram of the use. To understand the origin of these Archives was too porous to permit a scales, consider the case in which a stan- hydrostatic determination of the displace- dard of known mass is compared with two ment volume. It was generally accepted, "unknowns," one made from the same material however, that the kilogram was equivalent as the standard, one made from a different in mass to that of one cubic decimeter of material. In comparing the like materials water at its maximum density point [15]. of essentially equal mass, the A later group of pi ati num-i ridium kilo- buoyant effect becomes a second order , from which the International Proto- correction, and the value transferred to type Kilogram was selected on the basis of the unknown is on the same basis as the being most nearly identical in mass to the value assigned to the standards. Where Kilogram of the Archives, was of sufficient the materials are different, the buoyant quality to permit hydrostatic determination effect must be reconciled in some manner. of displacement volume. It was then evident The practice was started long ago where that the intended realization was not the value assigned to the unknown was achieved. In 1903, the unit of volume was defined by the third CGPM as follows: value. Density differences as great as 20 "The unit of volume, for determinations of ppm have been found [26] between the first high precision, is the volume occupied by a and last fractions of a distillation. mass of one kilogram of pure water, at its Uncertainties in the density of "pure" maximum density and under normal atmospheric or "normal" water, still under study, limit pressure; this volume is termed the . "[1 its use as a reference material in high ac- No reference is made to the . curacy measurements. Water, however, is an Careful measurements had shown that the mass adequate reference material for a wide vari- of one cubic decimeter of water at its maxi- ety of measurement activities. The most mum density point differs from the mass of recent fit of the original data has been the kilogram by 27 milligrams. As measure- done by Wagenbreth and Blanke [27]. In this ment precision has increased, the confusion work the data has been adjusted to the In- between the cubic centimeter and the ternational Practical Temperature Scale of milliliter has done likewise. 1968 (IPTS-68), and the density values, ex- The eleventh CGPM in 1960 requested a re- pressed in SI units (kilograms per cubic solution of this controversy which subse- meter) are tabulated for the temperature quently occurred at the twelfth CGPM in range 0° to 40°. Density values, as given which the definition given in 1901 was abro- in this work, are used at NBS for all volume gated, the recommendation made "that the measurements in which water is used. In some name 'liter' should not be employed to give cases, the values can be adjusted for use the results of high accuracy volume measure- with normal tap water. ments, but that high accuracy measurements The problem of uniformity of the refer- be expressed in the SI units, i.e., in cubic ence fluid was handled in a different way by meters or cubic decimeters." Cooke in his work with mercury. In essence, For liquid measure, the most common the final results apply to a large quantity metrics are cubic decimeter (usually called of mercury obtained by mixing samples from liter in spite of the above recommendation) a variety of sources. The mix was then sub- and the U.S. gallon, which has a defined re- divided, with smaller quantities being lationship to the SI length unit. furnished to national laboratories interested in high accuracy pressure measurements. Density Following the demonstrated ability to During the era encompassing the third compare of silicon crystals at the CGPM, about 1900, the thermal coefficient of 1 X 10" level [28], Bowman, Schoonover, and expansion of normal (to be defined) water Carroll [29] have established the density of was measured [17,18]; and in 1937, Tilton single crystal silicon artifacts with and Taylor at NBS [19] published tables (ex- measurement uncertainty on the order of 1 ppm. pressed in grams per milliliter) of the re- In this work, the volume of a group of steel lative densities of water as a function of spheres was determined by interferometric temperature from a formula they constructed measurement. Spheres were selected because to fit the early data. Around 1910, BIPM manufacturing processes have now been per- [20,21,22] had measured the true volume (in fected to the point that nearly perfect geo- cubic centimeters) of a milliliter by hydro- metry is readily obtained. Working with the statical ly weighing objects of directly buoyant forces in a fluid of arbitrary den- measured volumes in normal water. "Normal" sity, one can determine the volume of other water is now defined as that having a objects relative to the measured volume of "normal" or "average" isotopic abundance the spheres, the result being independent of ratio. the mass unit. This work suggests that the Modern workers [23,24] have shown that natural density of single crystal silicon the isotopic abundance ratio in samples of may be more constant than obtainable with water drawn from various locations on earth water in any form. Artifacts of known den- may vary enough to result in density differ- sity in combination with mass measurements ences of as much as 5 or more parts per and Archimedes' Principle, can be used to million. Without knowledge of the isotopic characterize either fluids of interest or abundance ratio, we are left with an uncer- other objects. Crystal samples have been tainty of this amount in the accepted furnished to other laboratories, including density of a particular sample of water. BIPM. Additionally, there is an even larger source In contrast to the high technology work of uncertainty. The vapor pressures of the briefly described above, most practical den- various isotopic forms of water are signifi- sity measurements are made to characterize a cantly different [25], so that distillation wide variety of fluids, again a problem of to remove impurities or boiling to eliminate interest since man discovered the pleasures dissolved gasses can further upset the iso- of alcohol. The resulting measurement sys- topic abundance ratio from its "normal" tem, revised and improved over the centu-

11 ries, is a simple, practical solution to a of the fluid are well known, and provided very complex problem. This complexity that the desired scale is defined relative arises from the fact that where solutions to the SI unit for density. [30] The most are involved, volumes are not additive. important development is the availability of Further, the changes in density of the solu- a new type of instrumentation for density tion with temperature cannot generally be measurement based on the inertial properties predicted from the characteristics of the of mass, described in Section 2.2.2.1. solution components. Based on Archimedes' Principle, a hydro- 2.2.1 Documentary Standards System meter floats in the solution at a level where the buoyant force acting on the sub- We define documentary standards to merged portion of the instrument balances include all documents which are formulated the gravitational force. The sensitivity for the purpose of providing solutions to of the instrument depends on the ratio of constantly reoccurring routine problems. This the stem cross section area to the submerged includes recommendations, test methods and volume. A reading scale in the stem can be procedures, and specifications, both graduated in any units which are related to hardware and performance. All standards are, the density of the fluid. Further, by in a sense, after-the-fact documents, and as selecting the density of some arbitrary such reflect an acceptable consensus decision fluid at a particular temperature as unity, with respect to some particular problem. the instrument is not dependent on any Standards provide a degree of order which in system of measurement units. Since distilled many cases produce economic savings to both water is a widely used dilutant, it is the the producer and the user. Perhaps more basis for a variety of "density" scales. important, standards provide a means to Figure 4 illustrates some of the more common address the legal problems associated metrics currently in use. with supply and demand, e.g. the The starting point for hydrometer calibra- specific desires of the customer, the tion is the selection and preparation of the limits of control by the regulator, reference fluids. In the case of simple the limits of guarantee or warrantee solutions, for example, a series of solu- by the producer, and some form of proof tions of known dilution are prepared from that "due care" was exercised to comply pure materials. For more complex fluids, with appropriate restraints. density is determined by weighing using weighted glass plummets of known displace- 2.2.1.1 Standardization Institutions ment volume, or picnometers of known con- tainment volume. These solutions are used The first of these is the apparatus set to establish appropriate scales on reference up by the Treaty of the Metre of 1875. This hydrometers. Corrections for temperature structure has two levels, diplomatic and effects are essential. Sets of reference technical. Nations who are signatories of hydrometers for the common metrics are the Convention of the Metre are represented maintained at the various national labora- on the General Conference for Weights and tories. The reference standards are used to Measures (CGPM). This body, composed of na- calibrate "master" hydrometers for manufac- tional delegates, elects the International turers and certain classes of users. The Committee for Weights and Measures (CIPM) "master" hydrometers are an essential part and formalizes the international adoption of of the manufacturing process. It is not recommendations. These recommendations in generally possible to control the physical essence promote uniformity in communications dimensions of the hydrometer to a degree by defining the various base units, in this commensurate with the sensitivity of the case the kilogram and the meter, and the instrument. For each precision instrument, various derived units. the scale must be established relative to an It should be emphasized that the organi- appropriate "master" hydrometer as one step zation established by the Treaty of the in the manufacturing process. The finished Metre is concerned only with the realiza- instrument may then be checked against tion of the units. The utilization of the another "master" hydrometer, either by the units, so defined, is left to others. manufacturer or the user, or it may be International Bureau of Weights and Mea- submitted to a national laboratory for cali- sures (BIPM) is the laboratory arm of CIPM. bration relative to the appropriate From its location just outside of Paris, it reference standard. serves as secretariat to the advisory Recent developments include a method to committees, maintains the prototype kilo- establish reference hydrometers without gram and periodically intercompares the using reference fluids, provided that the National Prototype Kilograms belonging to surface tensions and thermal characteristics the member nations.

12 .

Figure 4 PRACTICAL HYDROMETER SCALES

COMPARISON OF VALUES

— Hydrometer Scales. ->-

J3 | J J

m Si,

OH

^ If $

*C/OS 3t>£. L/7>OS

IP aire*?.

w v) 5 C L 0 J- Il

S T£Oi- 6^7C/V5 S'JT -SOZ.

13 A second international treaty organiza- united in a common endeavor, or the manufac- tion is the International Organization for turers and users of a product or technique. Legal Metrology (OIML). Its objectives are: In the latter case, it is generally true (1) to determine the general principles of that manufacturers' input is very strong un- legal metrology; less the area of interest concerns a large (2) to establish model draft laws and regu- number of users. lations for scientific and measuring The Federal Government within its pro- instruments and their use; and curement process generates specifications (3) to determine necessary and adequate which become, because of the size of the characteristics and standards to which government market, de facto national stan- measuring instruments must conform in dards. Other Federal Government agencies, order for them to be approved by member such as the Department of Defense and the states and for their use to be recom- various regulatory agencies produce stan- mended internationally. dards which impact all aspects of Actions of the Conference on Legal Metro- measurement. logy are in the form of recommendations and State and local governments also produce are not legally binding, but member nations standards of varying degrees of impact. By are morally obliged to implement decisions far the most important of these, within the of the Conference as far as possible [31]. context of this document, are the weights A nontreaty international standards organ- and measures laws which are currently issued ization of importance is the International by 775 separate jurisdictions. While the Organization for Standardization (ISO). The legal responsibility is at the state level, American National Standards Institute (ANSI) national coordination is obtained through is the United States member of the Interna- the National Conference of Weights and Mea- tional Organization for Standardization, sures, sponsored by NBS through the Office which now consists of the national standards of Weights and Measures. bodies of 52 countries. The United States' viewpoints to be presented in the technical 2.2.1.2 Survey of Documentary Standards work of the ISO are developed either through the interested sectional committee or CGPM resolutions define the metrics for through a competent committee of another or- mass and length, and the terminology for the ganization or, if there are none of these SI base and derived measurement units. Res- available, through a committee specially or- olutions are universally adopted by the 93 ganized as a USA Advisory Committee for the signatories of the Conventions. Efforts to ISO Technical Committee. provide access to these measurement units The work of the ISO technical committees generate a literature base associated with results eventually in ISO Recommendations the "how to" of measurement, e.g., instru- which may be embodied in the national stan- mentation design and development, constancy dards of the ISO's Member-Bodies. ISO re- of the unit, and the performance of the commendations, like all voluntary standards, measurement process over time and in the may be in the form of either detailed "how various environments in which measurements to" specifications or overall performance must be made. This literature base is a ma- specifications, e.g., "standard" test for jor resource for measurement application. determining the density of crude oil or a The community which is concerned with the listing of agreed upon ranges for hydrometers. "how to" of measurement is not generally re- ANSI and the American Society for Testing stricted by either proprietary or legal and Materials are the two largest of over limitations. 500 organizations in the United States who The application of the "how to" of consider the preparation of standards to be measurement to resolve specific measurement a major or important part of their reasons problems generates standards which can be for being [32]. The field of expertise categorized according to the nature of the found in both the ASTM and ANSI type of or- specific problem. For routine problems with ganization is best described by ASTM's wide application, e.g., instrument perform- managing director: "Actually, the Society ance specifications and the adjustment toler- has only one area of expertise and that is a ances for different classes of mass stan- full knowledge of management systems for the dards, or of interest to large sections of development of consensus standards" [33]. A the populace, e.g., weights and measures ac- great deal of care is taken to ensure user tivities, the resulting standards are read- participation in any committee to avoid ily accessible to all who have interest. antitrust action. Ease of access, in turn, results in such Voluntary standards organizations work standards inadvertently becoming a part of through a committee structure whose member- the "how to" literature base. The "how to" ship is composed of representatives of users

14 of measurement will be discussed further in the potential, and to a degree existing, use section 2.2.3. as nontariff barriers to trade. The largest number of standards relating Voluntary standards organizations, in es- to application are the "in house" standards sence, establish "due care" at the producer of the producers of goods and services. In or the producer/user level. By mutual con- order to limit liability with respect to sent, voluntary standards dealing with mate- contract compliance, warrantee or guarantee, rials are generated by the American Society such standards are not widely circulated. for Testing and Materials (ASTM), and those In many cases there is a concerted effort to dealing with devices by the American Nation- protect the proprietary nature of these al Standards Institute (ANSI). There is standards within a given organization, with close cooperation between the two institu- circulation based on the "need to know." tions and their standards are similar in While standards of this type are the basis nature. ANSI has an additional role re- for eventual consensus standards, the exis- lating to ISO. ANSI represents the United tence of a consensus standard does not pre- States' views in the international voluntary empt the use of the "in house" standard. standards effort. Both ANSI and ASTM stan- Product performance, established and main- dards are subject to periodic review. tained by "in house" standards, may be de- The role of voluntary standards in fended in a legal sense on the basis of American industry is complex and generally consensus standards. not well understood. Voluntary standards The basic differences between the OIML are "how to" documents for a variety of and the voluntary and government standards specific tasks. The standards are essen- are related to the means of deciding what tially defense mechanisms. In most cases, constitutes "due care" in a legal sense. the issuance of a standard is contingent OIML is an effort to define "due care" at a on a unanimous affirmative vote by the national level through the control of mea- reviewing body. The time interval be- surement detail. OIML recommendations per- tween concept and approval is on the order taining to artifacts and measuring devices of 5 years, with mandatory review inter- specify the placement of a metro! ogical con- vals of 3 to 5 years. After issuance there trol mark upon the article in question. is another time interval, also measured in That is every article of that type must years, during which the standard is imple- bear the "mark." Obtaining the "mark" for mented by the various users. As a conse- a product requires adherence to the specifi- quence, voluntary standards serve stability cations of the instrument or artifact engi- rather than viability. Nonetheless, the neering design contained in the pertinent collections of voluntary standards are used recommendations. by many as authoritative "how to" standards For example, a case in point is that of a for measurement. U.S. manufacturer of weights attempting to Professional societies, such as the So- market his product in a nation subscribing ciety of Automotive Engineers, the American to OIML recommendations [34]. A prototype Society of Mechanical Engineers, the Insti- approval was required before any weights tute for Nuclear Materials Management acting could be sold. Submission of a typical set through ANSI, develop and issue standards as of weights resulted in rejection based upon do some 360 trade associations. There are the dimensions of the weights rather than at least 20,000 such standards, accessible on the suitability of the weights for the via NBS SP 329, "An Index of U.S. Voluntary specific application. Engineering Standards." An ASTM subcommittee Further, the 59th National Conference of on balances has reviewed all of the ASTM Weights and Measures passed a resolution specifications to identify those in which pertaining to OIML in which it defined its mass measurement was limiting. As expected, view of the "moral obligation" involved in in an excess of 20 volumes of specifica- the treaty. That resolution specified tions, there were no mass measurement re- that the NCWM was not morally obligated to quirements which could not be satisfied with implement any OIML recommendation, i.e., a available instruments and procedures. In model weights and measures law, which had many cases, however, the specifications did not been approved by the NCWM after review not reflect a knowledge of the convenience and acceptance and only then if the U.S. had and economy associated with modern methods cast an affirmative vote for the model regu- and instruments. lation in the International Conference of Federal Standards, issued by the General Legal Metrology. Services Administration for use by the gov- The U.S. Advisory Panel who reviewed the ernment in the purchase of consumer goods, first 19 recommendations found 18 technical- represent a different approach to the "due ly unacceptable. The ultimate impact on care" problem. These standards represent an U.S. industry is a cause of concern due to arbitrated resolution of differences between 15 user requirements and vendor capabilities, the context of trade and the enforcement of with specific attention to the factors used regulatory requirements originating at the to differentiate between an acceptable and a state or local level. The principal univer- nonacceptable product, i.e., carefully sal documentary standard is Handbook 44, address the legal aspects of "due care." "Specifications, Tolerances and Other Tech- These standards are subject to periodic re- nical Requirements for Commercial Weighing view. There are approximately 28 such stan- and Measuring Devices." This loose leaf dards relating to mass, volume and density. style handbook, periodically revised, is In the case of volumetric glassware, the technically the output of the National Con- size of the government market relative to ference of Weights and Measures (NCWM). The the total market is such that these specifi- Office of Weights and Measures of NBS acts cations are essentially industry specifica- as the Secretariat of the NCWM, providing tions, with obvious benefits to the guidance and coordinating various activi- nongovernment user. ties. HB44 Has been adopted by all 50 The Department of Defense Standards rela- states and incorporated into their weights ting to the procurement of weapons systems and measures laws. In addition, 28 states have an important role in American technolo- have chosen to adopt uniform weights and gy. Weapons procurement involves high tech- measures laws following a model law prepared nology design, development and production, by NBS and approved by NCWM. usually where there is no precedent for de- sign, or where performance specifications 2.2.2 Instrumentation Systems are far more stringent than those associated with normal commercial products. Currently Broadly speaking, the instruments for such procurement is based on a vendor res- measurement of mass, volume and density con- ponsibility concept. In this framework, the sist of a wide variety of measuring devices, procurement procedures must address the artifacts, and procedures. Drastic changes problem of what constitutes "due care" on in all of these categories have occurred the part of the vendor in the course of over the past few decades. The precision carrying out the contract. balance industry has changed from the isola- The resulting system is set up under the tion of a tightly controlled craft to a high basic documents which affect all areas of production, high technology industry with measurement: worldwide sales and service. A few foreign firms satisfy essentially all of the market MIL-Q-9858A "Quality Program Require- demands. These same firms, concentrating on ments" the problem of making good measurements MIL-C-45662A "Calibration System Require- rapidly and conveniently, have had an impact, ments" on all types of weighing devices. The mar- MIL-I-45208A "Inspection System Require- ket for weight sets has essentially disap- ments" peared, with only one small company in this DSAM 451553 "Evaluation of a Contractor's country devoting part of its efforts to sup- Inspection System." plying conventional weight sets, and perhaps a few supplying special sets to weights and The implementation of these documents has measures officials. The use of volumetric resulted in the establishment of a system glassware, particularly in clinical labora- for measurement control, the operation of tories, has been replaced by a variety of which is a contractual requirement, and the completely automated test equipment. New inspection of which is carried out by the instrumentation has been developed for contract administrator. In excess of 65% of determining densities of fluids. Weighings the total calibration activity of NBS di- which were previously done in the market are rectly supports the resulting systems. It now being done with weighing devices incor- should be emphasized that the system is, in porated in complete packaging systems. Pro- view of the procuring agency, a minimal ef- cedures, which were carefully designed to fort necessary to production of acceptable maximize the performance of the older in- goods, i.e., a definition of "due care." No struments, are no longer applicable. attempt is made to resolve problems where The above changes have been made the required measurement effort has no rela- economically feasible by concentrating the tion to the quality of the product or where design and manufacturing effort on the areas the output of the system is inadequate for in which there is a large market, i.e., from the task at hand. a few grams to a few kilograms. There is Weights and measures activities, a state, only one firm in this country, and perhaps and to some degree a local, function, are in the world, that markets a 30kg capacity basically concerned with the legality of instrument of other than traditional design. mass, volume and density measurement within Only two companies in the world offer custom

16 b

built kilogram balances with performance in the instrument to changes in environment and the microgram range. One in this country the ambiguities associated with the concept. offers an instrument based on NBS design for The sophistication and, to a large degree, about $75 000, and one in Japan, for about the cost of the instrument depends on the $125 000. Equal arm balances with 5 000 1 degree to which the manufacturer has mini- capacity have been custom built; they are mized these effects by design. Depending on not available commercially. Most large the materials to be weighed, ambiguities capacity weighing devices are designed for associated with the concept must be consid- use as commercial weighing devices, i.e., in ered at the part in 5 000 to 10 000 level. accordance with Handbook 44, or other appro- Beyond this point, the instrument indication priate specification. The nature of the in- may or may not be an appropriate measure of dustry is such that a large amount of capi- the mass of the object or material being tal investment is required to design, de- weighed. Most weighing instruments are in- velop, market, and service a new instrument. tended to be used as direct reading de- As a consequence, manufacturers are very vices, and as such, are adequate for all reluctant to supply special equipment to requirements where the accepted uncertainty limited markets. Performance claims are of the result is determined by factors which generally limited to that obtainable by are not related to the measurement process, simplified procedure in relatively poor e.g., the grocery store pricing scales. environments. This creates a competitive market in which Two other recent developments have had a , convenience, and maintenance costs, large influence on the instrumentation sys- and perhaps certain regulatory requirements, tem: the availability of stainless steel for are important factors. use in mass standards, and the ability to As the limits for acceptable uncertainty characterize an individual measurement approach the of the ambiguities process, and groups of similar processes. The associated with the process, one must either stability of stainless steel weights, e.g., restrict the magnitude of the process ambi- constancy of mass values, is superior to guities, e.g., impose environmental controls brass in any form, and perhaps even superior and limit the class of objects to be to the platinum-iridium alloy of the proto- weighed, or one must adopt more sophisticat- type kilograms. Careful material control ed procedures and methods for data analysis. associated with the "built in" weights is Specifying one particular weight material, important to the success of modern instru- for example, extends the speed and conveni- ments. Computer analysis of large sets of ence of the pricing scale into the calibra- data provides quantitative parameters tion laboratory which is concerned only with descriptive of actual instrument performance the maintenance of the unit. Such action, which in turn guides the development of however, does not necessarily account for equipment and procedures. ambiguities in the assignment of mass values The most important development of the to other materials. With few exceptions, last decade, generally made possible by everyone must use commercially available electronic advances, is a change in the way equipment for mass measurement. Therefore, in which the instrument indication is deter- one must resort to procedures and techniques mined. Traditionally, changes in the angu- to make adequate measurements for require- lar position of the beam, as the balance is ments which are beyond the performance subjected to different "pan" loads, have claims of the manufacturer. been used to indicate differences in mass. The instrument indication is in essence In increasing numbers of instruments, the an amplification of the mass difference beam is driven to a prescribed "null" posi- between the object in question and a "tare" tion by applying an auxiliary force to the setting. The prerequisite for changing the beam through a servo system. The current in mode of operation of the instrument from a the "feedback" loop of the servo is then a direct reading mode to a comparative mode measure of the mass difference. There are a is that the indicating system not be number of advantages: the use of flexural intentionally rounded to obscure random- elements instead of knife edges, rapid oper- like variability as the sensitivity is ation, and a signal which can be read increased. di rectly by a computer. In a comparative mode of operation, with appropriate procedures and attention to 2.2.2.1 Measurement Tools and Techniques environmental conditions, one can achieve the maximum instrument performance. The Measurement tools and techniques are convenience of direct reading, however, is largely determined by the way in which the generally sacrificed. In a comparative result is to be used. Two sources of ambi- mode, the constancy of the unit is shifted guity must be considered, the response of from the instrument to a stable artifact.

17 The significant ambiguities are accounted For centuries, precision weighings were for by attention to environmental factors, made by comparing the mass of the object of and by treating the ambiguities associated interest to the mass of a summation of known with the concept in the data analysis. weights with an equal arm balance. The Where requirements are stringent, it may be necessary weight sets were provided through necessary to use supplementary measurement a hierarchy structure of mass standards. The data to determine the displacement volumes adjustment problem was solved by using two of the objects of interest, and to determine piece weights, the weight body with an ap- the density of the air in the vicinity of propriate cavity, and a cover or seal for the instrument. that cavity. In the adjustment process, Thje most readily identifiable artifacts small pieces of material were added to in mass measurement are the weights or achieve equality with the known standard weight sets which provide local access to according to prescribed limits of adjustment, the measurement unit. In those nations and the cavity was sealed. The mark on the which are members of the Treaty of the seal was, and in some cases still is, a Metre, a national laboratory has the task of symbol which verified the authenticity of transferring the unit as embodied in the the adjustments. Such weights were also, appropriate prototype kilogram to a more and still are, used to verify the perfor- practical weight material. This work, in mance of direct reading commercial weighing essence, substantially reduces the magnitude devices. The structure was simple, and of the ambiguities associated with the con- easily understood— to make a better measure- cept in providing the nation with access to ment all one needed was a set of weights the unit. Errors in the initial transfer, from a higher level in the hierarchy, e.g., however, become systematic offsets of the weights which had been adjusted closer to practical mass scales of the various nations. the prescribed nominal value. There are in effect two systems for the Modern weighing devices remove the choice dissemination of the unit with reference to of weight sets from the user. As before, the practical artifact standard. Starting the user assembles an appropriate summation with an artifact with defined value, an of weights to obtain an "on scale" condition equal arm balance, and a mass of material using the appropriate controls on the in- slightly in excess of that of the artifact, strument and then relies on the indication one can carefully remove material from the to subdivide the mass of the smallest inter- "unknown", until a comparison indicates a nal weight. The instrument manufacturer has "null" condition. At this point, one could taken the responsibility for the construc- declare the mass of the two objects to be tion and adjustment of the reference "identical." As one increases the sensitiv- standards. The weight materials are chosen ity of the balance, at some point it becomes on the basis of stability. Using techniques impractical to make further adjustments. such as electropolishing, weight adjustments The residual difference is then determined can be made on a production basis with devi- from the balance indicating scale. One can ations from nominal essentially limited only now announce the results of the work in one by the uncertainty of the values assigned to of two ways: the mass of the adjusted object appropriate reference standards. The in- does not differ from the defined mass of the strument case provides clean, dust-free, and standard in excess of some appropriate essentially nonaccessible storage. In at amount, e.g., difference expressed as a per- least one case, instrument assembly is done centage error; or, the mass of the object in a "clean room" environment. The design relative to the mass of the standard is a philosophy is in essence to remove the res- particular value, usually expressed as a ponsibility for the detail measurement pro- nominal value plus a correction, the sum of cedures from the user with a well engineered which is the assigned mass of the object. instrument. The instrument cost is consid- For example, the CGPM, in distributing the erably more than a traditional equal arm bal- prototype kilogram, announced that maximum ance and a set of weights, however, in terms difference in mass between the prototypes of cost per measurement, the traditional and the defining artifact did not exceed methods are obsolete. 1 mg. The reports of calibration from BIPM The change in emphasis from the closeness in contrast, state mass values relative to of adjustment to actual value is not without the defining artifact with uncertainties on problems. Everyone whose concept of mass the order of a few micrograms. The relative measurement was formulated before the last importance of these two points of view, one decade, including the scientific community, emphasizing the closeness to the nominal views mass measurement as a hierarchy of value and the other value relative to the de- mass standards. The most prevalent manner fined value, has changed drastically over the of expressing measurement error is relative last few decades. error, e.g. in error so many parts per

18 million or percentage. While this may be a analysis, one concentrates on character- convenient or valid way to express error in izing a particular measurement process. some cases, the error in mass measurement Performance indices are developed which is not quantity proportional. A part in a reflect the influence of major perturba- billion measurements can be made at the tions on the process, and which are valid kilogram level, but not at the gram or mil- prediction limits for uncertainty of mea- ligram level. Further, long standing regu- surement results over the range of vari- lations and specifications which are the ables for a given process. The programs, basis for a number of inspection systems once established, provide a continuous cannot cope with instruments which are monitoring of process performance. The "better" than the available hierarchy validity of each measurement is checked standards. relative to past process performance. With one exception, volume and density measurement is essentially mass measurement 2.2.2.2 The Instrumentation Industry instrumentation applied to a different set of artifacts, e.g. a wide variety of volu- Weighing instrumentation can be metric containers, hydrostatic weighings to considered in two classes; Laboratory determine displacement volumes or fluid den- Precision Balances and Commercial Scales sity, picnometers, etc. One characteristic and Balances. The nature of the two of interest is the coefficient of expansion industries is significantly different. of fluids over a rather wide range of tem- The laboratory precision balance industry peratures. Traditionally, a picnometer (a is primarily foreign, centered in Europe and special glass container of known contained Japan. The individual companies, while few volume) is filled and sealed in a tempera- in number, are large, some with worldwide ture controlled bath. [35] After sealing, sales and service organizations. Total the picnometer is cooled to a temperature sales figures are carefully concealed, compatible with the weighing device. From however, it is estimated that one European the weighing data and the characteristics firm has captured approximately 70% of the of the picnometer, one can determine the world market. Marketing in the United temperature-density characteristics of the States is through wholly-owned subsidiaries, fluid. The method is tedious and time con- or through agreements with American firms. suming. A very successful instrument is now Many of the balances in the NBS Mass Labora- available which reduces the task to simple tory, and all of the balances furnished to routine. The instrument, using the inertial the State Laboratories, with the exception properties of mass, relates the response of of the 30 Kg balances, are foreign made. a quartz tube, filled with fluid at a speci- The restrictive nature of the various states fied temperature, to an oscillating force weights and measures regulation has gener- field, to the density of the fluid. The in- ally blocked these firms from the Commercial strument, in essence a "density comparator," Scales and Balances field. can be used over a range of temperature. The American industry devoted to labora- When well characterized fluids are used for tory precision balances is a group of small calibration, near state-of-the-art precision companies. The equipment is usually mar- can be obtained and density determination keted through laboratory supply houses or can be made in a few . manufacturers representatives. Few have From the previous discussions in this nationwide sales and service organizations. section, it should be clear that a wide Most, however, operate short courses in variety of techniques for taking and service for journeymen who operate general analyzing the data are used. The degree of balance and scale repair organizations. The complexity of these techniques and proce- industry has been somewhat protected by dures is directly related to the manner various "Buy American" clauses which are in which the result is to be used. One used occasionally in certain government such technique, now called Measurement contracts. Assurance Programs, was developed in the In contrast, the Commercial Scales and mass measurement area. [36] This technique Balances industry in this country is large, addresses the problem of how one obtains with a number of firms having nationwide and the assurance that the measurement results international sales and service organiza- are valid when forced to utilize the maxi- tions. In general, the industry has been mum capabilities of a particular measure- conservative and very conscious of the ment process. Using redundancy of limitations associated with the various measurement, "check standards", i.e. weights and measures regulations. A wide objects similar in nature to those which variety of equipment is available for both are measured routinely but which remain single weighings and large complex systems, as a part of the process, and statistical e.g., packing systems, counting systems, 19 etc. Pressures from many sources, includ- sents only one hierarchy system - that ing marketing, prepackaging, and to a associated with equity in trade. The systems certain extent the foreign precision established within the Department of Defense balance industry, seem to have generated and by Department of Defense procurement some recent innovations evident in most procedure are orders of magnitude larger. grocery markets. Large mass standards, totaling in excess of Volumetric glassware is produced by large 3 million pounds, are installed in various firms who generally market by the case to deadweight machines used to calibrate large users and various scientific appara- force measuring devices. There is no tus supply houses. Special apparatus for estimate of the total number of artifacts which there is no large market is produced which are used in the industrial or scien- by small companies. The grade of the glass- tific community. ware is established by compliance with speci- fication requirements, some of which are 2.2.3 Reference Data performance specifications and some of which are detail specifications. The size of the Reference data are defined to include all market is established in part by the break- information or data which are taken from the age rate in use and by the cost of cleaning literature under the assumption that it is after use. Throw-away plastic apparatus is applicable to the task at hand. Such data, a significant competitor. in general, are presented in a format which Hydrometers are generally made by small is convenient to use. The format, however, companies, usually in conjunction with usually reflects judgments on the part of liquid-in-glass thermometers. Again the the author which are rarely discussed in grade is established by specification com- detail. The ability to judge the validity of pliance, and as with all glassware, the data relative to a particular usage is, breakage is a factor which tends to support in many cases, beyond the capabilities of the industry. Hydrometers are not usually the user either because of insufficient back- general purpose instruments, but related to ground on the part of the user, or lack of specific uses, e.g., the tax structure for detail discussion relative to the limitations alcoholic beverages, the petroleum industry, of the data, or lack of time. In many the sugar industry, etc. As a consequence, instances the validity of the data is deter- the market is generally governed by specif- mined by the reputation of the author. Dif- ic industry actions. ficulties arise, however, when the judgments Metal volumetric containers, such as used of the author are not appropriate to the task for testing liquid and gas metering devices at hand. and for calibrating process tanks, are pro- Perhaps the largest body of reference duced by one or two small companies. Most data are related to the "how to" of mass, of the market undoubtedly consists of re- volume and density measurements. Concepts placing damaged devices. Large companies of measurement are introduced in the ele- which produced such items in the past have mentary grades and amplified in the high now either sold that portion of the business school and early college . The con- or just stopped production. Again there is cepts are applied to practical measurement an interface with regulatory requirements tasks in text books and test procedures which to some extent dictates the equipment throughout science and industry. For exam- detai 1 s. ple, authoritative treatises on analytical Table 2 provides an estimate of the size chemistry contains chapters or sections de- of the American industrial effort relating voted to measurement detail [37, 38], weights to mass, volume, and density measurements as and measures officials receive formal in- compared to the larger category Engineering struction, defense department personnel are and Scientific Instruments (SIC 3811). formally trained in calibration procedures. Table 3 provides an estimate of the num- Many industries conduct formal training in ber of devices in use for commercial weighing the measurement techniques appropriate to and an approximate cost per device. Current- their products. The result is a great confi- ly available precision weighing devices with dence throughout the populace in the ability capacities from a few grams to a few kilo- to make mass, volume and density measure- grams cost from $3 000 to $10 000 and custom ments. In reality, however, this confidence built precision instruments from $20 000 to is limited by the level of understanding of $125 000. Weighing systems which are a part the person who undertakes the task of of a total system usually require a large instructing the "how to" of measurement. capital investment. Tabular data on the density of both dry Table 4 is an estimate of the number of and moist air are available in various met- devices which are used by state weights and rological literature, such as the Smithsonian measures officials, an activity which repre- Tables [39]. For requirements beyond the accuracy of these tables, air density is sources of reference data which are adequate usually computed using supplemental measure- for mass measurement. When such data are not ments of pressure, temperature, and relative adequate for the materials under study, ap- humidity, and appropriate formulas. There are propriate data are determined by separate

Table 2

V alue of Shipments (million of dollars ) 1970 1969 1967 1963 1958 Code Description 1972 1971

758.6 ] 3811— Engineering and scientific 1110.5 1215.9 1115.3 1049.0 722.1

i nstruments 6.5 6.5 3.7 3811371 Laboratory precision 6.8 bal ances 4.6 3829321 Hydrometers, glass, all 8.1 types 3229423 Scientific and Lab Glassware 92.5 64.1 3229425 Industrial and Technical 62.4 47.3 Glassware 169.4 131.4 90.02 70.99 3576-- Scales and balances, 138.3 152.7 except laboratory 18.8 12.2 6.5 3576011 Railroad track and motor 16.9 truck scales 8.9 6.3 7.0 3576021 Bench and portable industrial 10.8 scales 6.9 3.8 3.0 3576023 Floor scales (including 6.3 built-in) 9.9 8.7 3576025 Predetermined weighing and 11.9 check weighing scales 10.6 6.0 3576027 Automatic bulk weighing 10.2 18.7 10.2 7.2 3576029 Miscellaneous industrial 14.7 scales 12.1 11.6 3576031 Computing scales 19.9 2.5 1.3 3576035 Miscellaneous computing 3.3 scales 13.8 11.0 10.9 3576041 Bathroom scales 17.5 1.8 1.1 1.3 3576043 Person weighing scales 2.1 1.2 1.4 3576048 Miscellaneous kitchen, baby 2.3 scales 3.2 2.9 2.0 3576051 Mailing and parcel post 3.9 scales 2.3 2.6 3576082 Accessories and attachments 3.2 8.6 6.0 3576084 Parts for scales and balances 9.1 3.44 2.2 3576000 Miscellaneous household 6.1 scales and balances

a number of published formulas, all of which experiments which in turn become a part of are adequate for some . the reference data base- There is evidence that a fundamental lack of Water, and particularly distilled water, knowledge concerning the density of moist air is a reference material which is used for is one on the transfer of many purposes in addition to volume and mass value from an object made of a dense ma- density measurements. Tabular data regard- terial to an object made of a light material, ing the density of water at various temper- e.g., platinum-i ridium alloy to stainless atures are available from many sources, steel, or single crystal silicon. [40] practically all of which are based on mea- Coefficients of expansion of metals are of surements made in the early 1900's. Changes primary importance in length measurement and in the definition of the unit for volume and in many scientific and industrial applica- in the temperature scale are not contained tions As a consequence, there are many in the older references. As mentioned

21 b )

Table 3 Total Number of Number of Devices Retail Prices for New Type of Device Devices in Use Produced in 1969 Devices (Average) Drum computing scales 298,600 13,393 $275-525 Fan computing scales 74,800 3,073 $1 50-425 Projection scales 50,000 $800 Prepackaging computer scales 29,000 3,985 $3,200-5,600 Package checking 48,500 7,511 $225-243 Railroad track scales: 574 36 $16,000-30,000 Load Cell Railroad track scales: 7,000 31 $17,000-40,000 Mechanical Motor truck scales: 121 29 $3,000-7,000 Load Cell

Motor truck scales: 54,500 1 ,939 $3,000-10,000 Mechanical Other Platform Scales: 282,894 22,810 $225-20,000 Load Cell and Mechanical Beam Scales which have 4,720 capacities 100-1250 lbs. Drum type computing and 3,320 automatic meter scales Fan type computing scales 36,000 10 capacity vehicle 50 16 oz. beam scales 205,000 500 lb. parcel post scales 220

Table 4 Total Number Total Number Type of Equipment in Use Type of Equipment in Use Weight sets Test measures and provers

Cube weight sets (1/16 oz tc 2 lb) 841 1 gallon 1,312 Decimal pound weight kits 455 5 gallon 2,793 (0.001 lb to 0.3 lb) 50 gallon 251 Decimal ounce weight kits 127 100 gallon 380 (0.01 oz to 0.5 oz) 100 gallon (stainless 5 8 lb test weight kits 163 steel Small individual mass standards 500 gallon 58 50 lb mass standard 36,647 1000 gallon 24 25 lb mass standard 191 Mi seel laneous test 300 Miscellaneous mass standard 140 measures and provers Scales, balances, etc.

Over and under 5 1 Package checking 726 10 lb field balance 305 Miscellaneous scales and balances 272 Platform scales 97 Large mass standards 500 lb mass standard 1,143 1000 lb mass standard 2,628 2500 lb mass standard 122 5000 lb mass standard 37 Miscellaneous mass standard 7 2500 lb dollies calibrated 36 5000 lb dollies calibrated 4 Miscellaneous moving dollies cali- 16 brated

22 before, NBS uses and recommends the tables and the density of the fluid, was used to es- of Wagenbreth and Blanke [27]. tablish the displacement volumes of the Limitations on the density of water as a crystals. The uncertainty of the announced reference material are associated with the density for these artifacts is on the order isotopic content. Menache [41] and Emmet and of 1 ppm, the major components of which seem Miller [42] find that particular continental to be associated with phase shift at the sur- waters vary in density as much as 10 ppm face of the steel sphere in the interfero- because of isotopic composition. Information metric measurements of their diameters and on the variation at one place with time indi- perhaps the air density algorithm used to cates a change of less than 1 ppm. The dis- establish the mass of the silicon. All of tillation process, however, may alter the the characterized samples are on long term isotopic content of the waters. These limi- loan or have been requested for loan by var- tations were in essence the reason for devel- ious national laboratories. Since the mate- oping a solid object density reference rial is brittle it must be handled carefully. material (single crystal silicon) to be Care must also be exercised in the prepara- described in section 2.2.4. tion of the material for hydrostatic weigh- ing. There are plans to characterize a 2.2.4 Reference Materials larger quantity of this material and make it available through the Standard Reference Reference materials, that is, objects or Material program of NBS. materials which have been characterized by Oceanographers interested in relative careful measurement, are made available to changes have established a reference material others for use in local measurement process- designated Standard Mean Ocean Water (SMOW). es. In one sense calibrated mass and volume SMOW is taken from depths of 500 to 2,000 artifacts, considered to be one of the in- meters in open ocean areas where no direct struments for mass measurement, could be dilution by continental runoff or glacial classified as reference materials. The re- melt water can occur. Large bodies of data mainder of the reference materials is asso- are based on SMOW from a particular loca- ciated with density measurements, primarily tion, however, samples from other areas the density of certain fluids. taken at various times are adjusted to agree Cook of the National Physical Laboratory with the established characteristics of (NPL) studied the density of mercury. This SMOW. Density is only one of the character- work included mercuries from various sources istics of interest. Samples of SMOW are as well as a mixture. In one study density available to oceanographers and available was determined by the hydrostatic weighing from several sources. method using a tungsten carbide cube, with displacement volume computed from measured 2.2.5 Science and People dimensions. In the second study density was determined on the basis of the mass of mercu- Mass, volume and density measurements, to- ry contained in a cubic volume constructed of gether with length measurements, are some- optically flat quartz plates, the computed what unique in the history of civilization. volume being based on interferometric length Practical instruments for the measurement of measurements. The announced volume from the these quantities were essential to emergence density of mercury, uncertain by about of the arts and crafts from the earliest

1 ppm, applies to the mixture. Sufficient agrarian societies. Practical instruments quantities of the mixture were furnished to were essential to transition from barter to various national laboratories for use in commerce in the form that we know it today, precise pressure measurement. first intertribal, then between communities, Early work of Bowman, Schoonover and and later between nations. Finally, prac- Carroll using an instrument based on the tical instruments for the measurement of principles of the Cartesian Diver, demon- these quantities were for centuries the most strated the ability to compare the densities precise instruments available which could be of small single crystal silicon artifacts to used by philosophers and scientists for test- a precision of 0.1 ppm. [43]. This work ing theories relating to natural phenomena. suggested the possibility of using single The only common factor between these diverse crystal silicon as a density reference. areas is the instrument, a situation which

Bowman et al . [8] determined the density of resulted in growth of one of the most secre- selected large single crystal artifacts by tive crafts in the history of technology - first computing the volume of a pair of that devoted to the development and manufac- steel spheres from interferometric measure- ture of precision balances. ments of diameter. The volumetric ratio The characteristics of the three groups method, which is essentially a hydrostatic of users, broadly designated as (1) the pro- weighing but independent of the mass unit ducers of goods and services, (2) the regulators, and (3) the scientists, are such ways. In commercial transactions involving that there is little intragroup communication the exchange of certain materials, contracts of significance. Producers of goods and ser- can be challenged, in which case one defense vices are profit oriented, that is, primarily is evidence that the measurements were legal- concerned with establishing and maintaining ly correct. As a protection against fraudu- a market for a product or a service. Their lent and liability charges relating to the success depends upon ingenuity in the appli- goods and services produced, the producers of cation of science and technology in resolving such must conform to the legal interpretation practical problems relating to the particular of what constitutes "due care." In these product or product line. Competition, sup- matters if one has a choice between a defense ported by antitrust regulation, makes it nec- based on the justification of new procedures, essary to protect this ingenuity. or on the precedence of old established pro- Professional engineering society meeting cedures, the path of wisdom is usually the agendas rarely include plant visits and even latter. then competitors must register separately. These three groups have coexisted for the Plant access is always restricted, except in last few centuries, relying on the guilds to certain very large basic-type industries, produce the necessary balances and weighing e.g., steel plants, automotive assembly devices. Events in the early 1800's resulted plants, large power plants, etc., with fre- in an attempt to impose an additional common quent interdepartmental restrictions. Except factor - the adoption of a uniform set of for companies which conduct non-product re- measurement units. The units for mass and lated research, communication within the area volume were to be based on natural phenomena of producing goods and services and with the rather than replicas of existing artifacts. other two areas designated above, is gener- The scientists of the time could not agree on ally limited to elements which are common to the basis for the length unit, e.g., the the group, such as problems relating to the length of a pendulum of a specified period or definition of "due care," e.g., test proce- a fraction of circumference of the earth as dures, specifications, and "out of the measured along a particular meridian. The ordinary" contractual requirements. For the metric system, as developed in France, is average professional the peer group struc- based on the latter, with the mass unit ture, if such exists, is related to these based on the maximum density of pure water. common elements. As a result of a fire England was faced with In contrast, the scientific community is the problem of replacing the accepted mass entirely different. Generally, it is not unit. The resulting action, a reconstruc- profit oriented. It is more or less dedi- tion of the destroyed artifact, established cated to the advancement of knowledge rather the pound as the mass unit for the English than the application of knowledge to the speaking countries. This action favored the solution of practical problems. The estab- producers of goods and services. Both the lished review procedures associated with most construction of the metric units for mass and scientific literatures emphasize the peer length and the reconstruction of the pound groups. The individual scientist in turn were major scientific efforts. relies on the quality of the literature of With the completion of the artifacts the his field and concentrates on contribution dissemination mechanism in essence created to that literature base. In the design and a new discipline - metrology. The metrolo- construction of apparatus he concentrates on gist was concerned with the conservation and minimizing the influence of measurement dissemination of the unit. The metrologist variables which mask the property of inter- utilized material from many sources in the est. By exercising care in the definition accomplishment of his assigned tasks. Scien- of the problem he is usually able to achieve tists on occasion had found it necessary to his goal supported in part by open discus- improve mass measurements in order to ac- sion with interested colleagues, laboratory complish particular tasks. Using such lit- visits, and participation in one or more erature, together with some engineering and established peer group activities. a lot of good judgment, the metrologist was The member of the regulatory group is not able to make substantive contributions to the interested in either the advancement of literature with regard to improving balance knowledge or the application of that knowl- performance. He was, however, never really a edge. He is concerned with the actions part of any particular discipline. His which define a measurement relative to es- closest relationship was with the scientific tablished laws and, in certain areas, to as- field. Scientists quickly adopted the metric certain that measurements are made accord- units whereas the technologists were and ingly. He has essentially no interaction still are reluctant to change. The metrolo- with the scientific area. He does influence gist also realized that solution to many of the producers of goods and services in two his problems must come from the scientific

24 field, but he was frustrated by an inabili- 2.3 Realized Measurement Capabilities ty to define his "real world" problems in a manner which would attract the necessary Any measurement result is the product talent. His problems could not compete with of a particular measurement process. the exploding new field of modern . The quality of the results is generally He was not a part of the regulatory system, expressed by a quantitative statement but, by his position, he was the highest au- regarding the area of doubt, or uncer- thority for decision. He was not a producer tainty, associated with the result. The of goods and services so that his judgments magnitude of the uncertainty depends on relative to dissemination were not always two factors; the quality of the reference consistent with users intentions. He was standards, and the details of the partic- almost never challenged, either because of ular measurement process. A necessary his authoritative position or because usage condition imposed on the measurement requirements had not caught up with measure- process is that the magnitude of the ment capabilities. uncertainty be acceptable relative to As with all other facets of the concept of the manner in which the result is to be mass, volume, and density measurements, the used. Satisfying only this condition is science and people aspects have also experi- not always sufficient. Other imposed con- enced substantial change over the last few ditions are related to compliance with decades. There are many factors which have contractual requirements and conformance contributed to the change. Engineering cur- with accepted practices, both of which ricula were changed drastically in the late may or may not be of importance relative 1940's and early 1950's. New disciplines to the intended usage. In most cases were established to bridge the gap between realized measurement capabilities are science and engineering, e.g., engineering not limited by either the available science and engineering physics. The con- standards or equipment but rather by stancy of artifact mass units, a problem insufficient knowledge concerning the which plagued the metrologist, essentially performance of the process and concern- disappeared with the adoption of stainless ing adequate acceptable limits relative steel. Computer technology took over much of to the end use of the result. the routine drudgery of lengthy hand computa- tions associated with precise measurement, 2.3.1 Quality of Reference Standards and for the first time it was possible for the people involved in measurement to actu- The reference standards for the mass ally think about what they were doing. unit are the artifacts, or mechanisms, Statistical techniques were developed to which give the local measurement processes characterize a total measurement process access to the unit and multiples or frac- in the environment in which it had to be tions thereof. In size, reference stan- used. At the same time emerging measurement dards range from fractions of a milligram requirements, first associated with the space (small pieces of quartz fiber which must age and later with environmental protection, be handled with a manipulator and a were beyond the capabilities of currently microscope) to large 50,000 lb stainless used measurement methods and established steel discs, as used in the NBS force mechanisms for inspection. As a consequence generators. The quality of this array the perceived role of the NBS mass and volume must be judged from two viewpoints, section changed. constancy relative to the precision of It was recognized that the data base and measurement at a particular level, and methods of analysis used in characterizing constancy relative to the intended usage. the processes used for maintaining and dis- In most cases, the latter viewpoint is seminating the mass unit could also be used the most important. in other ways, and further, the principles The precision of measurement at the used had wide applicability in many measure- prototype kilogram level (on the order ment areas. It became possible to formulate of .001 mg) is now such that changes in both the problems associated with precise mass can be observed which reflect the mass measurement and the suggestions for response of the material to the immediate those with practical mass measurement prob- environment. [44] Cyclic changes seem to lems in such a manner as to attract the in- correlate with changes in pressure and terest of the scientific community and relative humidity. Constant rate increases industry as well. The emerging synergistic are attributed to contaminants in the atmo- relationship is faced with many difficulties, sphere. Both of these are surface effects, mostly associated with past practices, perhaps related to the catalytic properties however, it seems to be growing steadily. of platinum. These changes are significant relative to the highest precision of 25 measurement (10 to 20 standard deviations). oz is 480 grains; 1 apothecaries' dram

Kilograms made from stainless steel do not is 60 grains; 1 apothecaries' scruple is appear to vary as much as the platinum 20 grains; 1 oz troy is 480 grains;

iridium kilograms. Such changes are 1 pennyweight is 24 grains; and one carot negligible relative to the precision of is 200 mg.) To complete the picture, the most kilogram capacity balances (on the class C weights are typical of those used order of .05 mg). to test commercial weighing devices to Before the general availability of verify compliance with performance speci- stainless steel, most weights used in pre- fications shown in Table 6. cise weighings were made of brass. Con- A different hierarchy of adjustment stancy of mass was then a continuing tolerances is used for laboratory mass problem. First with the quality of the measurements. The commercial tolerance brass, then with atmospheric corrosion, structure tacitly assumes that the ambi- and finally with coating materials and the guities associated with the concept and material used for adjustment. Lacquer the precision of the instruments are coatings were hygroscopic. Plating dis- negligible relative to the size of the continuities and plating solution residue tolerance. The tolerance structure caused variability. Oxidation of adjusting associated with precision weights is material caused problems. NIBS Circular 547 frequently so small that both of the delineated the detail requirements for above factors must be considered in detail weights which would be accepted for cali- to verify compliance. This leads to a bration. Strict inspection and rejection considerable degree of confusion in the at NBS resulted in improved stability. interpretation of calibration results. With the increasing use of stainless steel, One school of thought advocates strict and the recent changes in instrument control of weight materials and designs in design, constancy relative to most usage order to maintain the simplicity of hier- requirements is no longer a problem. archial structure relative to weight sets, Reference mass standards are available generally at the expense of the user. The in a variety of forms. Both regulatory and competing school of thought, believing that competitive pressures have resulted in a one should be able to assign a mass value hierarchial structure in which the quality to any object reasonably stable in mass of a mass standard is judged on the basis properties, advocates the abolishment of of its form and the difference between the the hierarchy beyond the requirements of mass of the weight and its assigned nominal commercial weighing. value--the adjustment tolerance. The high- In summary, mass standards are available est order weights have the smallest adjust- which are sufficiently stable in mass for ment tolerances, and in general, are more all known requirements. This fact is carefully constructed. The various levels generally down graded in the manner in are frequently designated as primary, which the mass value is assigned to the secondary, tertiary, quarternary and so on standards. The variety of adjustment down to the level of weights which are tolerance structures is confusing to those considered adequate for testing commercial who must obtain standards with reference weighing devices. The hierarchial structure to a particular usage. An ASTM subcommittee exists primarily to satisfy the legal has been working on the task of unifying requirements for the calibration of and simplifying the various specifications. weights, i.e. the comparison of weight While this document is now nearing with a similar weight from a higher completion, the growing acceptance of the echelon. direct reading instruments has greatly There are in excess of a dozen different reduced the priority initially attached classes of weights, ordered by construction to the task. The problem, however, detail and adjustment tolerance, the most remains in areas which utilize a hierarchy common being Class C- -Commercial Test of mass standards. For an increasing Weights. Table 5 shows the adjustment number of users, the reference mass tolerance limits originally established standards are the "built-in" weights in in 1918. This tolerance structure has been the modern balances. These weights are extended to include weights up to 10,000 of good quality and are carefully adjusted lbs. It is of interest to note the in- to well within the established tolerances. fluence of the crafts in the listing of Generally it is not possible to calibrate the customary units, many of which are such instruments in the traditional manner, still in use but now defined relative to i.e. the use of hierarchy standards. the kilogram. (1 lb Avoirdupois, the Reference standards for glass volumetric equivalent of 16 oz or 7,000 grains, is apparatus consists of volumetric and mea-

defined as 0.45359237 kg; 1 apothecaries' surement pipets, volumetric flasks and 26 —

TABLE 5 —Toleraaces for Class C—Commercial Test Weights

[The maximum error allowable on each weight is tiven in the co!:^n^s hcacl'-d " Tolerance"]

Customary

Carats 1 ' I

I Avoirdupois Grams 1 Apothecary troy ; PeunyveigilU

I I !

Te- De- De- De- De- De- De- Ttoler- Ttoler-1 Toler- Toler- toler- Ttoler- Tcler- nomi- nomi- nomi- r.omi- nomi- noml- coini- auce ance ance ance ance ance ance tslic-Q nation nation nation nariGn nation aiition

-

lb. gr. gr. gr. oi. ap. gr. 02. t. gr. dv.-t. gr. c mg 600 50 10 10 000 1 12 1 1000 10 10 000 10 2500 70 10 409 25 6 soeo 1 10 1 500 10 5000 5 2000 60 5 250 20 6 2000 0. 5 6 0. 5 400 5 4000 5 1000 40 2 150 10 4 1000 0. 3 5 0. 5 300 5 30C0 3 500 30 100 a 3 500 0. 2 4 0. 5 200 5 2000 3 200 20 5 3 200 0. 15 3 0. 4 100 3 1000 2 100 10 500 70 4 2 100 0. 10 2 0. 3 50 2 500 1. 5 50 7 200 40 3 2 50 0. 05 1 0. 2 40 1. 5 400 1 0 20 5 100 30 2 1 5 20 0. 03 r. ap 30 1. 5 300 1 0 10 3 50 20 1 1.0 10 0. 02 0.2 20 1.0 200 1.0 5 2 20 _ 10 oz 5 0. 015 5 0. 2 10 1.0 100 0. 5 2 10 7 10 1.0 2 0. 010 4 0. 2 5 as 50 0 t 1 0 7 5 3 8 0.5 1 0. 005 3 0. 1 4 0. 5 40 0.3 0.5 0. 5 2 3 S 0.5 0 5 0. 005 2 0.1 3 0.4 30 0.3 0. 2 0. 3 1 2 4 C. 5 0 2 0. 0C25 1 0.05 2 0.3 20 0 2 O 1 0. 2 2 0. 3 0. 1 0. 0020 =. ap. 1 0. 2 10 0. 2 0. OS 0.15 500 1. 5 I 0.2 0. 05 0. 0)15 2 0.05 0.5 0.2 5 0. 1 0. 02 0. 10 200 0.7 t 0.2 0.02 0. 0015 1 0. 03 0.4 0. 15 4 0. 1 0.01 O.OS 100 0.5 a l 0.01 0. 0015 0.3 0. 10 3 0 05 50 0. 35 0. 05 0 2 0. 10 2 0. 05 2il 0. 20 i 0.05 0 I 0.05 1 0.03 10 0. 15 a (is 0.05 0. (53 5 0. 10 0.02 0. 0» 0. 025 2 0. 05 0.03 0.025 1 0. 04 0.02 0. 020 0. 01 0. 015 0. 005 0. 010 0. 004 0 COS 0.003 0. C07 0.002 0. 005 0.001 0. OOS

straight precision burets. Perhaps the comply with specifications. This situation major factors differentiating the quality is usually resolved by a concentrated of such glassware are related to the index effort to change the specification. For mark which defines the contained volume example, changing from an etched and filled the quality of the mark, its location, index to a "fired on" index requires a and permanence. The user must set the level change in Handbook 44 if the ware is to be of the miniscus to coincide with the plane acceptable to the state weights and established by the index by adding or measures officials. removing small quantities of liquid. He Precision volumetric glassware is basic is interested in the quantity of liquid to the current methods for the preparation in the vessel when this condition has been of solutions in "wet" chemistry and clini- satisfied. In the case of precision glass- cal pathology. Testing of this apparatus ware, the location of the index is estab- for compliance with specifications or for lished by actual test. In the case of adequacy relative to a particular require- burets, there may be several test points ment is now within the capability of most which are subdivided linearly in the users with one exception. A legal measure- marking process. The reference standards ment as defined by the regulators requires are the basis of a variety of glassware an artifact which has been calibrated by which is graded according to the quality higher authority. The calibration efforts of the marking, and the difference between of NBS exist essentially to support the the actual volume contained relative to a states weights and measures operations particular index and the nominal volume. with only occasional requests for calibra- The useful life of all glassware, includ- tions relative to a particular end use. ing the reference standards depends on For example, some calibration work directly breakage and the cost of cleaning. supports manufacturers quality control The present manufacturing processes systems. produce glassware which is in accordance Metal capacity standards, e.g. "slicker with current specifications. These speci- plate" containers, cubic bottles and fications generally contain both perfor- provers, are used either singly or in mance and detail requirements. The sales combination to test flow metering devices, groups favor the detail requirements but or in the calibration of large tanks. The when basic manufacturing methods are devices are compromises between the proper- changed, the ware produced may no longer ties one would like to have in a standard,

27 TABLE 6—TYPICAL BASIC TOLERANCES FOR SCALES INDICATING OR RECORDING IN AVOIRDUPOIS UNITS (Edited, for complete table see NBS Handbook 44, pages 19, 20, 21)

Test load Mai ntenance tol erances Acceptance tolerances

From To*

Ounces avdp. Grains Ounces Pounds Grains Ounces Pounds

0 2 2 1 2 4 4 2 4 8 8 4 8 16 16 8 Pounds avdp.

1 2 25 1/16 0.004 13 1/32 0.002 2 4 1/8 .008 1/16 .004 4 7 3/16 .012 3/32 .006 7 10 1/4 .016 1/8 .008

50 75 1 .062 1/2 .031

75 100 1 1/2 .094 3/4 .047

100 150 2 .125 1 .062

800 1 ,000 14 .875 7 .438

1 ,000 and over 0.1 percent of test load 0.05 percent of test load

* but not including

BASIC TOLERANCES FOR SPECIAL SCALES INDICATING OR RECORDING IN AVOIRDUPOIS UNITS

Type scale Maintenance tolerances Acceptance tolerances

Prescri ption 0.1 percent of test load 0.1 percent of test load Jewelers 0.05 percent of test load 0.05 percent of test load Cream and moisture 0.5 0.3 grain Test (18 gm load) Animal, livestock, 0.2 percent of test load 0.1 percent of test load crane, axle-load, hopper (other than grain) and vehicle Railway track scales static 0.2 percent of test load 0.1 percent of test load uncoupled in-motion 0.2 percent of test load 0.2 percent of test load Wheel load weighers 3.0 percent of test load 2.0 percent of test load Railway coupled in- Acceptance and Maintenance Tolerances motion (100 car a) Difference between motion weight and static weight test train) of train shall not exceed 0.2 percent. b) Difference between motion weight and static weight on 100 car weights shall not exceed. 1) 0.2 percent on 30 cars 2) 0.5 percent on 5 cars 3) 1.0 percent on any car

28 and the ability to handle the standard in and procedures. This is not to suggest use- "Slicker plate" standards use a flat that all measurements should be made ac- glass plate on a flat surface of the con- cording to these published techniques but tainer to define the contained volume. rather to identify the magnitude of vari- Such artifacts are usually of heavy con- ability from various sources which the user struction and are limited in size to at must evaluate relative to the manner in most a few . Capacity standards which he intends to use his result. Ulti- are of lighter construction, the location mately, the uncertainty associated with of the liquid level in the neck being his result is based on three factors, the inferred from the liquid level in a paral- systematic error introduced into his lel glass tube with reference to an appro- process by NBS calibrated standards, and priate linear scale mounted on the vessel. the systematic error and random error The calibration may consist of determining associated with his process. If one or the volume relative to the scale reading, both of the latter factors are the predomi- or the adjustment of the scale to a posi- nant terms in his uncertainty, technically tion appropriate to a nominal volume and there is little need for direct contact then determine the "over" and "under" with NBS. For example, the Class C toler-

amounts relative to appropriate scale ance for 1 kg from Table 5 is 100 mg. graduations. There is no formal hierarchy of capacity standards. The manufacturers, of which there are only a few, can furnish TABLE 7—TYPICAL MEASUREMENT calibrated standards adequate for most UNCERTAINTIES requirements. In some specific uses, such as in the petroleum industry, the items are Mass routinely sent to NBS for calibration. Reference standards for density Level Uncertainty measurements include picnometers, reference 50 000 lb + .2 lb fluids, solid objects of measured density, 50 lb + .000 03 lb and calibrated hydrometers. Hydrometer 1 kg + .11 mg sensitivity to change in density is 1 g + .003 mg determined by the ratio of the cross 1 mg + .000 5 mg sectional area of the stem to the dis- placement volume of the bulb, and, Large Capacity Measures consequently, the instruments are fragile. Test Tank 3 000 liters* + 1 liter A high quality hydrometer has a hand 100 prover + .000 20 gal marked scale permanently enclosed and 30 gal prover + .000 33 gal anchored in the stem. The calibration is 5 gal "Slicker Plate" + .000 31 gal for a given liquid at a specified tem- perature. Corrections are necessary for Displacement Volume other liquids (surface tension correction) Alumi num Kg 360 + .000 5 cm and other temperatures. Volumetric Glassware

2.3.2 The Measurement Process 1 Liter Flask + .1 cm

A measurement process includes the 50 ml Buret + .01 cm instrument, and all of the measurement 10 ml Buret + .005 cm details including preparation of material, 25 ml to making the observations, supplementing 50 ml Transfer Pipet + .01 cm data, data analysis, and interpretation. The NBS measurement processes for mass, Hydrometers volume and density are sufficiently well API Scale Range 10 to characterized that the uncertainty state- 21 in .1 API + .92 API ments are based on quantitative values for random and systematic error which are Specific Gravity based on large collections of data Scale Division .0005 + .0001 reflecting both current and past perfor- mance. Typical measurement uncertainties Alcoholometer 56-72 are shown in Table 7. %vol. in A% Divisions + .02% It has been the intent to completely 3 describe the various measurement processes *The recommended SI unit for volume is m . in the literature to demonstrate the mea- In practice, however, the term "liter" surement capabilities which can be achieved is used for one cubic decimeter, and with currently available instrumentation milliliter (ml) for one cm^.

29 e

The uncertainty associated with a calibration any object sufficiently stable in mass to of the same weight from Table 7 is .11 mg. warrant the calibration effort. The cali- If compliance with Class C is adequate for brated objects, in turn, provide fixed his intended usage, the esthetics of an reference points on a hypothetical mass NBS calibration is a costly luxury. None- scale such that others can adjust working theless, for reasons associated with com- weights to whatever tolerance desired. pliance or conformance, it may be necessary A second class of de facto central for the user to have standards calibrated standard authorities have been established periodically at NBS. by various levels of the Federal Govern- ment. Of these, the two of largest impact 2.4 Dissemination and Enforcement Network are the Federal Procurement Specification, as issued by the General Services Adminis- There are a number of networks more or tration, and the regulations of the Depart- less loosely interlocked. By virtue of ment of Defense, previously mentioned. In the embodiment of the mass unit in arti- the former case, GSA acts as a mediator fact standards, NBS is the focal point of between the federal user and the supplier the network concerned with the dissemina- to obtain products which are adequate for tion of the unit. The dissemination of the intended use. Federal specifications, methods, or the "how to" of measurement unlike voluntary standards, address the has in the past been the literature of quality control aspects of procurement and the various disciplines, i.e. chemistry, clearly delineate the basis on which a pro- clinical pathology, etc. The most widely duct will be accepted or rejected. The recognized enforcement network is that DoD regulations with regard to measurements associated with equity in trade, a also stem from the generic problem of responsibility delegated to the individual quality control with respect to military states. Other enforcement networks have procurement. been established by various segments of the government such as the Department of 2.4.2 State and Local Offices of Weights Defense, the Internal Revenue Service, and Measures the Nuclear Regulatory Commission and others. A different type of enforcement One of the largest central authorities network is that associated with voluntary is the National Conference of Weights and or consensus standards, mostly concerned Measures, an organization of state and with the "how to" of measurement. With local weights and measures officials. The the emergence of the concept of measure- historical tie between NBS and NCWM is ment as a production process, a concept long standing. The Constitution of the which concentrates on the adequacy of field United States, Article I, Section 8, measurements relative to the intended empowers Congress ". . . . to fix the usage, the role of the Mass and Volume standards of weights and measures." Section relative to these various networks Congress has chosen to allow the individual is quite complex. states to perform this function. However, as early as 1836, Congress, in a joint 2.4.1 Central Standards Authorities resolution, directed the Treasury to fabricate standards for customhouses and The units, as defined by the Convention states and to establish an Office of of the Metre and the Committee General Weights. In 1901, this office was dis Poids and Mesures, were declared designated as the National Bureau of legally permissible in the U.S. by Congres- Standards. As the nation grew, a need was sional action in 1866. This action was felt for a closer coordination of the followed in 1893 by the Mendenhall Order state activities and the National Bureau in which the Secretary of the Treasury of Standards met with state officials to announced that the International proto- discuss "protecting the public against type standards for mass and length would short weight or measure." A dialogue be regarded as fundamental standards. The developed which is now carried out through practical mass unit in the U.S. is embodied the agency of the National Conference of in a pair of ni chrome kilograms. In the Weights and Measures. dissemination of the unit, the Mass and A typical state Office of Weights and Volume section is not concerned with the Measures is a unit in the state Depart- conformance of the reference mass standards ment of Agriculture although it may be to any specification requirement for under the Department of Law and Public construction or adjustment tolerance. The Safety (New Jersey), Department of Labor perceived role of the Mass and Volume (Rhode Island), Department of Commerce section is to establish mass values for (Arkansas), or elsewhere. It is headed by an appointed state official and has a 2.4.3 Standards and Testing Laboratories staff of inspectors which varies according and Services to the size of the state by at most a few dozen. Weight and balance manufacturers offer Within the state, various local calibration services, i.e. tests for com- jurisdictions may also have weights and pliance with or adjustment to comply with, measures authority. In all, there are a tolerance specifications for both weight total of 775 offices and a total of approx- sets and weights internal to certain imately 3000 officials. Besides control- balances, with reports which are acceptable ling the weights and measures activities of for most requirements. The world's largest retail trade, depending on the state's supplier of precision balances (by far in industry, anhydrous ammonia (fertilizer) or excess of 50% of the market) relies on NBS lumber or grain may be of official for the maintenance of their standards. interest. Recently, under both the Fair Maintenance and test of weighing equipment Packaging and Labeling Act of 1966 and is usually done by separate organizational the general use of consumerism, these units in large facilities, or by service offices are increasingly involved in con- contracts with reputable repair service sumer protection in areas remote from organizations, e.g., those made available classical weights and measures activities. by the instrument manufacturer. Metal In 1965 Congress appropriated funds to volumetric apparatus manufacturers, of equip state laboratories with standards which there are only a few, provide both whose values were known with uncertainties refurbishing services and calibration smaller than those needed to support the services. tolerances mentioned in section 2.3. This The production of precision glassware is program, managed by the NBS Office of controlled at the manufacturing level by Weights and Measures but calling on other the same methods as used in NBS calibra- NBS resources, has to date delivered 42 of tion. The manufacturer guarantees compli- the 53 sets of artifacts and instruments ance with federal specifications. These needed by U.S. states, territories, and specifications, in turn, can be used by the District of Columbia. anyone in the procurement of apparatus. Provisions of this law call for the GSA has established procedures to monitor states as a prerequisite for the award to the quality of the products. Considering supply a suitable laboratory and trained the normal breakage rate and the increasing metrologists. In the absence of educa- use of "throw away" apparatus, there is tional facilities noted before, a train- little motivation for testing services to ing program has been established to calibrate apparatus. provide initial and continuing education Attempts to interest independent testing for weights and measures officials and laboratories in providing calibration ser- technicians. Only the State of California vices, particularly with regard to hydro- has a comprehensive in-state program and meters, have not been successful. These less than 6 states have a full time train- laboratories are primarily concerned with ing officer, although some twenty states performing specification tests according to have a program of some sort. procedures prescribed by ASTM, The American The state weights and measures offices Concrete Institute, The American Welding provide a calibration service for mass Society, etc. Either the volume of work which is accessible to the general user is too small relative to the cost of estab- community in addition to performing such lishing a service, or the nature of the services for their own inspection opera- industry using the instruments is such that tions. The NBS interfaces with this the risk is too large with respect to the system only through contact with certain profits. Of approximately 130 members of users who have tried these services with the American Council of Independent inadequate results. The impression is Laboratories, less than ten list weighing that the service is seldom used where the as a particular specialty. uncertainty of the measurement must be There are a few companies, some known. The service is a convenient way worldwide, who provide calibration services to comply with certain imposed require- to the petroleum industry. The services, ments, however, with few exceptions, the consisting of meter and tank calibrations, facilities are not oriented to end use contract directly with the oil companies. probl ems. It is presumed that these companies exist

31 because of the necessity of the oil com- of harmful substances, such as the Environ- panies to meet various regulatory require- mental Protection Agency and the Nuclear ments, in which case it is less expensive Regulatory Commission, are examples of the to contract for the services than to estab- latter viewpoint. The Nuclear Regulatory lish their own service organizations. By agencies have worked closely with the Mass far, most of the large oil and pipeline and Volume Section since special nuclear companies have their own service materials accounting procedures require and organizations. use both mass and volume measurements using The primary function of the echelon state-of-the-art techniques. calibration laboratories of the Department The Mass and Volume Section has been of Defense (DoD) is to provide calibration aggressive in encouraging the regulatory and testing facilities to the various agencies to change from a "traceability branches of the department. With few of the unit" concept of measurement to a exceptions, e.g. government owned equip- "traceability of the measurement" concept. ment and certain support to defense con- The following examples are illustrative of tractors, these services are not generally the nature of the transition. available to the user community at large. The recognition of adequacy relative to Cooperation with local weights and measures the manner in which the result is to be officials, and with local industry is used is illustrated by the work with the largely dependent upon the work load and Internal Revenue Service's Division of the attitude of the local command. Similar Alcohol, Tobacco and Firearms Control. The services are provided to the National Aero- function is the control of the use of ethyl nautics and Space Administration contrac- alcohol. All legal ethyl alcohol is manu- tors through a series of contract operated factured under bond. No matter what the facilities. All of the top echelon labora- use of this alcohol, be it beverage, tories in these two networks are partici- medicinal, cosmetics, or any other purpose, pants in the Mass Measurement Assurance it is taxed at the time of withdrawal from Program. bond on the basis of equivalent 100 proof gallons (i.e. a 100 proof mixture is a 2.4.4 Regulatory Agencies fifty percent mixture with water). The determination of proof at the time of with- Regulatory agencies are concerned with drawal is a consensus between the IRS, the the application of measurement technology supplier and the purchaser. The measure- to certain specific measurements. Gener- ments are made with "proof" hydrometers. ally, the uncertainty associated with the From the IRS viewpoint, the required pre- measurement result is only one factor in cision of measurement is established by the judging the adequacy of the measurement. proof increments in the tax tables. Modifi- Regulatory agencies view measurements in cation of these procedures are used in the one of two ways. One view typical of the case of wine and similar alcoholic bever- older agencies is defensive in nature. In ages. There is a mechanism for tax rebate this view, in the event of prosecution, the where ethyl alcohol is not used as a bever- use of NBS calibrated standards is consid- age. Over the years under the guidance of ered to be a means to divert attention from NBS, IRS has established a calibration the measurement detail, i.e. the adequacy service adequate to support these activi- of the measurement is established by "trace- ties. On occasion NBS acts as a consultant ability" of the unit. Requirements estab- in regard to techniques and basic data. lished by the newer regulatory agencies The producers and purchasers of the tend to emphasize the uncertainty of the material, however, rely on the NBS hydrom- measurement result relative to the manner eter calibration services. in which the result is to be used, i.e. the Finally, the work on voluntary standards adequacy of the measurement is established relative to the nuclear industry (Institute by "traceability" of the measurement. In for Nuclear Materials Management and the this case the adequacy of the measurement Nuclear Regulatory Commission) is an result is established by careful analysis example of an area where perceived measure- of the entire process including standards, ment requirements are beyond the capabili- methods, control of systematic errors, etc. ties of traditional weights and measures In these cases NBS involvement is primarily regulatory procedures. In this case NBS consultation. The standards may or may not has been involved in helping both the regu- be NBS calibrated. lated and the regulator to (1) quantify State weights and measures regulations the requirement and (2) understand the are typical of the former viewpoint men- concept of measurement as a production tioned above. Regulatory agencies which process and measurement assurance and are concerned with limiting concentrations (3) ways and means to improve measurement 32 techniques in both mass and volume on the basis of the manner in which the measurements. results are used, one relating to the determination of quantity of material; one 2.5 Organizational Input-Output Matrix relating to matters concerning regulation and quality control; and one relating to In an attempt to more clearly define the properties of objects or materials. While morphology of this measurement system, we these systems are not completely indepen- have created an Input-Output Measurements dent, the structures reflect practical Transactions Matrix. This matrix presents boundary conditions associated with the in compact form all we have been able to desired result. Within this structure, the deduce about the interconnections in our impact, status and trends of the subsystem system and it justifies careful study. can only be discussed relative to the Along the upper and left hand border of desired result. All mass, volume and the matrix you will find twenty-five organi- density measurements are merely one of many zations or groups which have been identified elements involved in a variety of specific as potential users or suppliers of measure- endeavors. In at least two of the above ment data in the field of mass, volume, and subsystems success is not limited by the density. Many of these are both users and precision of measurement, but rather by suppliers. In their role as suppliers, they social and economic judgments. appear on the left, as users along the top border. 3.1 Impact of Measurements By survey, by interview, by a retrospec- tive study of NBS calibration records, and 3.1.1 Applications by the experience of the authors, the inter- actions between these elements of the system Every legal exchange of material, be it were estimated and quantified according to at the local market or in the world trade the code at the bottom. The interactions arena, has two elements, the price and the were examined for magnitude, derivative of quantity of material. The same is also the magnitude, and importance to the sys- true in the exchange of products except tem's well-being. Particular care was exer- where the product is marketed as a unit. cised to maintain a uniform scale of values, As a consequence, there is at least one, particularly relative magnitudes. It is and in some cases many, determinations of realized that this scale reflects a degree quantity for each transaction, and the cost of subjective judgment even though utiliz- of measurement is a part of the cost of ing the individual opinions of several know- doing business. Order in the marketplace ledgeable people independently. is contingent on the acceptance of uniform First, as a supplier of measurement in- measurement procedures. The major con- formation, we note, not surprisingly since straints on the resulting measurement our charge is measurement, that NBS inter- processes are the cost of measurement acts strongly with most of the other ele- relative to the cost of the material, the ments. Only a few industries are blank. constancy of the mass, or volume, of the These are industries where mass, volume, and material over the interval between ship- density measurements are not frequently ment and receipt, and the ability to make made. Our interactions with the knowledge adequate measurements in the environments community, for example, is strong, stable, associated with custody exchange. In all important and moderately successful. cases the measurements must be within the NBS as a rule seems to listen fairly well capabilities of the available personnel and to what people have to say. Except for the be defendable in court, that is, in accor- knowledge community and the upper echelons dance with a consensus agreement and rela- of the standards bodies, these transactions tive to an acceptable artifact standard. are largely assessments of perceived needs. Direct NBS involvement with this subsystem is providing access to artifact standards 3. IMPACT, STATUS AND TRENDS OF which are accepted as authoritative. In- MEASUREMENT SYSTEM direct involvement by NBS is associated with the availability of instrumentation The national measurement system for and measurement methods. mass, volume and density, being defined as The subsystem concerned with quantity the totality of all such measurement measurements is very large and diverse. It processes from the defining standards to includes those activities in which the end the point of ultimate usage, is so vast product is formulated by recipe, including that further subdivision is necessary the preparation of solutions in the chem- before one can address impact, status and istry laboratory to the bulk production of trends. One can identify three subsystems paints, acids, beverages and other liquid

33 ) ) r r — ) -

1 27) (exc. DIRECT - (exc. FORESTRY Jj 35) 36) 37) & 38) TESTING cr h- CD CO 1, 80) Publ.) MEASUREMENTS NATIONAL COMMUNITY Education, WEIGHTS B) CD LU — CD CO — (OWM's) Req.Aq. 391) SVCS/PRINT TRANSACTIONS U Aq.) 5 EQPMT FEDERAL q; _j ro MINING ELECTRICAL Go A Gp LU LU FAB. Gp Gp UTILITIES bal Gp MATRIX FOR S & 6 - SERVICES PUBLIC LOCAL LOCAL A C) 1 IDE AND E) — -a AGENCIES t-i H- CO >- o PRODUCTS OWM's) o MASS. VOLUME, E OF AGENCIES Req. & TRADE/INS/FIN/REAL r- lO S \ 5 - '2 3 2 2 4 1 2 2 2 2 2 1 2 2 2 1 1 KNOWLEDGE COMMUNITY 3 2 2 (Science, Education, 2 2 1 2 3 1 1 3 1 1 1 1 2 N 2 2 N 2 Prpf. Soc . 4 Publ . 2 INTERNATIONAL r—3 3 3 2 4 3 2 2 4 2 4

METROLOGICAL 1 3 3 2 3 1 2 ORGANIZATIONS 2 4 4 4 4 2 ? 1 1 1 2 2 3 1 3 DOCUMENTARY — — 3 1 2 3 1 4 2 1 STANDARDIZATION 1 4 2 3 3 2 3 1 1 1 1 X 1 1 3 1 2 1 1 ORGANIZATIONS 2 2 2 3 1 2

1 2 2 2 1 1 2 2 3 3 4 INSTRUMENTATION 1 2 2 i— r 2 2 1 1 3 2 1 1 1 1 INDUSTRY 2 1 2 3 3 1 3 2 2 2 2 2 3 2 (SIC Maior Go 381 3 2 2 2 2 2 2 2 L, 1 2 2 2 2 1 1 5 4 1 4 1 2 V r 4 2 4 1 2 3 1 2 2 NBS 4 4 3 2 3 4 2 3 2 1 1 2 1 2 1 2 2 3 1 2 N 1 1 1 1 N 6 OTHER U.S. NATIONAL STANDARDS AUTHORITIES 1 1 4 1 7 5TATE & LOCAL 2 2 2 1 3 1 4 1 T" 2 1 3 1 3 3 4 OFFICES OF WEIGHTS 1 2 2 4 3 2 1 1 3 2 1 4 1 3 1 3 4 3

& MEASURES (OWM's) 1 1 2 R R 1 2 2 R 1 1

8 STANDARDS & TESTING 1 1 2 2 1 2

LABORATORIES 1 2 2 2 3 3 1 1 1 2 1 1 1 1 2 AND SERVICES R R R 1 9 f EGULATORY 3 1 3 1 3 1 3 1 3 1 1 / GENCIES 1 1 1 3 1 1 3 2 1 1 2 3 2 2

excl. OWM's) R R. • 1 1 1 n R R 1 R R R R R R , R R EPARTMfNT Of 1 1 2 1 I 2 1

t EFENSE 1 1 1 3 3 1 1 3 1 l l

excl. Stds. Labs) 1 N R R R R R N R R R if civilian federal 2 1 2 2 2 2 r t 2 1 GOV'T AGENCIES (excl. 1 2 1 1 2 3 3 1 1 1 1 1 1 1 1 1 l Stds. Labs & Reg. Aq. N 2 R R R R R R R R R R 12 STATE & LOCAL

GOVERNMENT AGENCIES 2 1 3 1 1 1 1 1 R R R (exc. OWM's & Req. Aq. R R R R 13 INDUSTRIAL 2 2 2 1 2 1 2 1 2 2 1

TRADE 1 3 2 2 3 2 2 3 3 1

2 1 R R 1 1 14 MluiRoRLslKV 2 3 2 2

FISHING; MINING 1 2 1 2 1 1 1 4 l 4 4 4 1 1 4 2 l (SIC Div. A & B) R R R R R R R 15 CONSTRUCTION 1 1 1 1 1 2 1 3 1 i ! i 2 (SIC Div. C) 1 R R R R 16 F00D/»TOB/TE)iTlLE/ 2 2 2 2

APPAREL/LBR/FURN/PAPER/ 1 1 4 1 1 1 1 1 3 4 2 1 3 1 3 LEATHER (SIC 20-26, 31) R R R R

2 ' 17 CHEM/PETROL/RUBBER/ 2 2 2 3 2 1 PLASTICS//CLAY/ 1 1 2 1 2 2 1 2 1 1 3 3 1 4 1 1 1 1 2 3 2 GLASS (SIC 28-30, 32) 2 R 1 R R R R R 18 PRIMARY & FAB. METAL PRODUCTS 1 1 1 1 2 1 1 1 1 1 1 3 1 1 2 1 (?|C 33-34, 3?1) R 19 MACHINERY,

' EXCEPT ELECTRICAL 1 2 3 3 3 2 2 1 2 2 1 1 2 2 2 1 3 1 2 4 4 2 4 (SIC Maior Go 35) R 20 ELECTRIC AND ELECTRONIC EQPMT 1 2 1 (SIC Ma.ior Go 36) R

21 1 RANSPORTATION 2 2 2 2 1 2 1

f QUIPMENT 2 1 2 1 1 1 2 1 1 1 1 2 2 1 2

SIC Major Gp 37) R 2 1 R R 1 1 2 22 1 RANSPORTATION S 2 2 1 F UBLIC UTILITIES 1 1 3 2 1 1 2 4 2 2 2 2 2 1 3 4 3 1 3

1 SIC Div. E) R R 23 TRADE/ INS/FIN/REAL 4 2 EST/PERS SVCS/PRINT-PUB 2 3 4 1 1 1 2 3 2 2 4 1 1 3 1 1 2 4 1 4

(SIC F-H, bal. I. 27) R R R R R R R 1

24 t 3 1 2 1 1 HEALTH SERVICES 4 1 2 2 1 1 1 2 4 2 (SIC Major Gp 80) 1 R 1 3" 1 25 3 1 T 1 T

GENERAL PUBLIC 1 3 2 1 1 1 1 1 1 1 2 4 1 4 R 1 R 1 R R R R R R R R R 1 1 —-s 1 KEY TO MATRIX ENTRIES

- C IMPORTANCE OF TRANSACTIONS D - ( IN) ADEQUACY OF SERVICES

1 = Purely convenience 0 = No improvements needed = 2 Strongly desirable 1 = Could be improved 3 = No real alternatives 2 = Marginal 4 = Essential 3 = Serious deficiencies 4 = Out of control B - RATE OF CHANGE A - MAGNITUDE OF TRANSACTIONS N = Declining Tri vial 0 = Stable Minor 2 = Growing Moderate 4 = Growing explosively Important R = Flow of requi rements info dominates Major

? = Unknown, X = Not studied, Blank = 0

34 products- These measurements are subject on manufacturer's specification which can to essentially the same constraints as be maintained with minimum training is those associated with the exchange of better than nothing. A more complex goods, except that the cost of measurement system, based on the premise that complete- is established relative to the quality and ly trained metrologists will be available, market value of the end product. The NBS would tend to collapse from the lack of role, however, is essentially the same. adequate manpower. There are other ramifi- The subsystem associated with quality cations, previously discussed relating to control and regulation is also a diverse "due care." The element is large, and subsystem with a different set of con- through inertia, the features have been straints. A characteristic of such mea- adopted in other disciplines, i.e. American surements is that the measurement effort is College of Clinical Pathology. defensive rather than productive. The con- The third element consists of the mea- cern is generally whether the measurements surement processes associated with regula- are adequate for the purpose, or whether tion other than weights and measures. Some the product does or will conform to the of these systems, those associated normally appropriate specification. Decisions con- with compliance, have features from both cerning the details of the measurement the weights and measures, and the military process are relative to risk. The genera- systems. All are concerned with legality tion of the elements of the subsystem arise of measurement. The one component with from social pressures, the need for product which the Mass and Volume Section has the acceptance, the need for national defense, most direct input is associated with the and the health and welfare of the nation. control of nuclear material. In this case, Three subelements, discussed below, illus- the source of detail specification is the trate the NBS role with regard to this regulated facility. The regulator, usually subsystem. the Nuclear Regulatory Commission, general- Perhaps the oldest of the elements is ly provides only guidelines and goals. In that associated with weights and measures-- the process of establishing a facility, the states weights and measures activities. detailed procedures are submitted for re- Constraints on the system are those associ- view and approval. The collection of ated with the legality of measurement, the approved procedures eventually becomes the need for viability, the need for simplic- operation manual. The role of the inspec- ity, i.e. parties involved must understand tor is that of enforcing compliance with the detail, and the resulting procedures the approved procedures. This approach is must be within the capabilities of the also being considered in regulations con- available personnel. The net results in a cerning the introduction of pollutants into labor intensive system which operates most the atmosphere. In this element, the effectively at the consumer level, and with people being regulated have considerably great difficulty in interactions with more interest in measurement detail. The capital intensive organizations. NBS has decisions, however, are theirs relative to supported this system by providing well the economics of measurement, including calibrated reference standards for mass the maintenance of the system, and the and volume, and good instrumentation. announced goals established by the regula- By and large, this equipment supports an tor. Generally, the desired precision of echelon of field standards. measurement is somewhat beyond that A second element associated with quality associated with either weights measures control is the system established by the and military procurement with respect to Department of Defense. One of the unique mass and volume. characteristics of this element is that the The last subsystem, those concerned established system at any given time must with the properties of objects or materi- be capable of immediate expansion, thus it als, is also a large diverse system. The is very similar to the weights and measures elements of this system are mostly con- elements, with echelons of calibration lab- cerned with adequacy of the result of oratories. Each laboratory maintains measurement relative to a particular standards of mass and a measurement capa- requirement, that is, the uncertainty bility appropriate to the echelon, partly associated with the measurement result as to support the various commands, and partly compared to functional limits related to in anticipation of immediate expansion. the task at hand. In essence these are Visibility is in the form of calibration the processes associated with the product intervals and inspection "marks." One can which in turn are monitored by the various argue the cost effectiveness of such a quality control and regulatory systems, if system, however, in the event of rapid such exist. These measurement processes expansion, a simple inspection system based range from those associated with the

35 constancy of the defining mass unit, to the ness of the effort relative to what the determination of the density of dislocation effort is expected to accomplish. In the free single crystal silicon, to the support interface between the producer and the con- of force and pressure measurements. Some sumer, and between the regulator and the of these processes are concerned with prop- regulated, the visibility of NBS seems to erties of special materials, such as the be highly desired, particularly in the form density of crude oil and refined petroleum of calibrated standards. On occasion, NBS products, the results being the basis of calibration is used to resolve disputes in measurement systems relating to both regu- lieu of other legal action. As before, the lation and commerce. Measurement processes cost of the NBS services provided is minus- in this element are usually free of all cule relative to the value of the goods constraints other than that of producing subject to quality control and regulation. adequate results in the most economical Again, this subsystem is not receptive to manner. efforts directed toward establishing self- sufficiency. Aggressive attempts to im- 3.1.2 Economic Impacts prove the utilization of both methods and instrumentation have not only been resisted 3.1.2.1 Processes Relating to Quantities of but met with animosity. In contrast to the Material previous subsystem, it is felt that this area would welcome NBS expansion of routine In a free economy one factor contributing calibrations, not from the standpoint of to the stability of the marketplace is the product improvement, but from the resulting ability to make adequate measurements of simplification of paperwork requirements. quantity. The system that has evolved places NBS in the role of higher authority 3.1.2.3 Processes Associated With with regard to measurement standards while Properties of Objects or Materials generally reserving the right for decisions with regard to measurement detail. Con- The processes in which mass, as an sidering the intricacies of the judgment intrinsic property of a material or an ob- factors this perhaps is as it should be. ject is of interest require the ultimate There are adequate techniques for the mea- in mass measurement capabilities, e.g., surements which have to be made. The dollar the constancy of the defining artifacts volume associated with exchange of material and the density of single crystal silicon. is a significant fraction of the gross These processes are without exception within national product. Yet direct NBS interac- the scientific community and in support of tion is minuscule. On the other hand, this specific scientific goals, e.g., the deter- subsystem is not receptive to attempts to mination of Avagadro's Number, the realiza- establish self-sufficiency. Change in the tion of certain electrical quantities, the cost of materials or changes in regulatory gas constant, the universal gravitational requirements are the major motivating constant and the like. While a substan- factors for improvement. It is clear that tial part of the total effort in mass, NBS visibility is desired, and on occasion volume and density measurements at NBS has aggressively pursued, yet direct economic been in support of these efforts, the di- benefits remain intangible. rect economic impact resulting from this effort cannot be measured at this point 3.1.2.2 Processes Associated With Quality in time. This does not mean that these Control and Regulation efforts are totally devoid of such benefit. The emergence of new procedures and refer- In both cases, the investment in ence materials as a result of this work measurement processes is associated with reduces the cost of certain types of mea- the entrepreneur's assessment of risk and surements, e.g., the "solid object" density not by the available precision of measure- standard. In the course of this work, the ment. This follows from the fact that efforts devoted to understanding the philo- neither function exists until such time sophies of measurement, in addition to de- that there are facilities capable of pro- velopment of procedures, have the greatest duction of parts or products more or less potential for economic impact on the entire adequate for the introduced usage. Pro- measurement community. fessional Quality Control people and inspectors are not generally trained 3.1.3 Social Impacts metrologists. In both cases the assessment of the effectiveness of the efforts are The greatest impact of the Mass, Volume, measured in terms of compliance with the and Density system to the average citizen established plan rather than the effective- lies in the area of retail trade. Thanks to

36 a functioning weights and measures enforce- problems. Such structures are generally ment system, the citizen in his daily life established on the basis of a set of can rest assured that he can base his pur- conditions at some point in time. Con- chasing decisions on posted price-per-um't tinually changing technology, marketing quantity. It is not necessary to search for procedures and the like may result in the gas station giving the largest gallon more effective ways to achieve the de- and prepackaged foods can be bought without sired result, or may make obsolete the qualm. Yet, this state of affairs is not problems which originally established the without cost. In 1972-73, the State of bureaucracy. In these cases, the advan- California reported litigation costs of over tages of changing the system must be com- $100,000 in a single "short weight" case. pared with the disadvantages associated In the same year, the State of New Jersey with disrupting the work force. Generally, reported 1,321 citizen complaints of which if the measurements are not vital and if 1,183 were successfully prosecuted with the costs are accepted as a part of busi- $54,750 in penalties assessed. The State of ness, the work force is preserved. Because West Virginia reports over $60,000 recovered mass, volume and density measurements are for consumers. Most of the populace have so widely used, and because the measure- some knowledge of the general concepts of ment techniques are not difficult, there these measurements, and a large percentage are a number of areas where present prac- are involved in some degree through such tices are continued as the result of a things as keeping track of body weight, social decision rather than a technological procuring supplies and materials, and form- one. ulating various recipes. Marketing prac- tices have changed substantially over the 3.2 Status and Trends of the System past two decades, however, and with the preponderance of prepared and prepacked The precision of measurement has now foods, pricing scales interfaced with reached a point where the ability to trans- computers, etc., the populace must rely fer the unit to other objects is limited by largely on the system to look after his the characteristics of the defining arti- interests. Systems, on the other hand, tend fact and the characteristics of the weigh- to be inflexible, bureaucratic and not ing environment. The longevity of the generally sensitive to individual problems. present defining artifact is contingent Achieving and maintaining order in the on the resolution of these problems. In marketplaces through a bureaucratic struc- terms of the national measurement system, ture is not without cost. The degree of the questions are largely academic. uniformity in available products and prices One cannot seriously doubt that the across the country is one of the charac- national measurement system is producing teristics noted by many foreign visitors. results adequate for the requirements as Sizable as these numbers are, they are perceived. Problems which are brought to a vanishingly small fraction of total NBS are seldom associated with the ability sales. It is impossible to overstate the to measure but with (A) inadequate defini- benign influence of the weights and mea- tion of the measurement requirement sures activity on social order. In an era relative to the task at hand and (B) in- where alienation from government is ram- adequate understanding of the measurement pant, the citizen can see, if only by the process details. These problems are equal- seals on the scales in the supermarket and ly prevalent in science, industry and com- on the gas pumps, that his local govern- merce. For modest measurement requirements, ment is protecting his interests. these factors do not necessarily affect the The rest of the system impacts only adequacy of the result relative to its lightly on the citizen. He does have an intended use. The efficiency of the mea- interest in the health services delivery surement effort, however, decreases sig- system and his spokesmen are expressing nificantly, e.g. more measurement effort a legitimate concern over the adequacy of is expended than the task requires, or the laboratory work. These tests, upon measurements are made which are unrelated which a great deal of diagnosis is based, to the basic task. depend on elaborate automatic volume mea- The parts of the system which are suring equipment whose accuracy is large- satisfied with the measurement results ly unproved. Efforts are now under way currently obtained are essentially static. to examine this part of the system. In many cases, there is little motivation A second social impact which influences for change. Where measurements are reason- measurements relates to inflexibility of ably important relative to the task at hand large bureaucratic structures which are the trend is to eliminate or at least designed to address certain specific reduce operator judgment, e.g. digital 37 indications, sometimes with computer inter- wrong, error estimates were conservative. facing, and completely automated processes This was also a period in which the charges such as in the packaging industry. for calibration were not retained by NBS, One factor which will influence the but returned to the Treasury. As a con- direction of the system is the political- sequence, calibration was routinized with economic pressures for increased govern- frequent waiver of fees on the basis of mental intervention in the area of need. There was essentially no communica- standardization and enforcement of stan- tion between NBS and the ultimate user of dards. These forces, typified by the the calibrated items. International Organization for Legal The latter era started with the general Metrology, would bring a very large frac- acceptance of the then new constant load tion of the measurement system under single pan balance. This gave the user a surveillance. The legalization of precision of measurement equal to, and in measurement has two basic limitations. In some cases better than, that available at those situations where adequate results NBS. Conservative individual estimates of cannot be obtained by "legal" means, the the area of doubt associated with calibra- system will have to be circumvented. The tion results were no longer acceptable. laboratories supporting legal metrology This in essence, forced a re-evaluation of (e.g. state weights and measures offices) the services provided by NBS. cannot provide guidance without getting Changes in services evolved around the into a conflict-of-interest situation. idea that in a given facility, a measure- That is, the regulator cannot provide ment process is essentially a production measurement guidance without assuming process. While the objects after responsibility for the adequacy of the measurement and the results pass on to result if, at some later time, he must others, the measurement process remains and pass judgment on the legality of the in a sense is a capital investment. One resulting measurement process. can utilize statistical techniques to evalu- ate the process, and to establish quantita- 4. SURVEY OF NIBS SERVICES tive parameters which describe its performance. These parameters are the basis 4.1 The Past for realistic uncertainty statements and in turn are the basis for judgments concerning A brief history of the Mass and Volume the adequacy of the results relative to the Section is presented in Appendix B. For a intended usage, the need for process im- discussion of the past NBS services in the provement, and perhaps most important, pro- area of mass, volume and density measure- vide a means to monitor the performance of ments, there are two time periods of impor- the process independent of individual tance, one extending from the establishment judgments. of NBS to approximately 1960, and one from In the 1960's methods and procedures were 1960 to the present. In the earlier period, developed in accordance with the "new services provided were essentially calibra- approach" for all of the calibration tion of weights for manufacturers, and for services provided. The most completely the states' weights and measures organiza- developed system is in the area of mass mea- tions. A considerable effort was made in surement with extension into a number of the formulation of specifications for mass measurement facilities through the Mass Mea- standards. A quasi -regulatory function was surement Assurance Program. The next most generated by limiting items acceptable for developed system is in the area of capacity calibration to those which complied with standards, primarily to bring the NBS pro- the specification. These specifications cesses into a state of control and to extend reflected the metrologist view of "what the the techniques into the area of application. user should have" rather than a careful In the area of hydrometer calibration and assessment of what was actually needed. the calibration of volumetric glassware, Work in the latter part of the period, measurement methods were developed incorpo- i.e. studies of the response of a balance to rating the desired features but little was thermal air currents, the development of a done to incorporate the methods into the 50 lb balance for the states' laboratories, calibration procedures. In these areas the and the initial work on high precision kilo- calibration work load is small, and the new gram balances, was instrumental in causing procedures were substantially different from the change in service provided. Typical current practices. It was felt that the new of this era was the acceptance of the procedures would not really benefit the validity of a measurement based on the users of hydrometers or volumetric glassware. reputation of the person who made the mea- surement and, in the interest of not being 38 4.2 The Present--Scope of NBS Services e.g., kilograms or pounds, with uncertainties commensurate with the requirement, and lim- The variety of services available from ited only by the amount of effort and the NBS which support the National Measurement uncertainty of the reference kilograms, Nl System range from providing access to the or N2. mass unit and the calibration of appropri- (2) For those who do not have the facilities ate artifacts, to providing guidance to or for some other reason do not elect to do that section of the system associated with their own "work downs" or "build ups," NBS "equity in the marketplace" and to provid- will calibrate ordered sets of weights from ing consultation to all who ask for assis- 30 kg or 50 1 b to 1 mg or the appropriate tance. The calibration services of the equivalent. Weights less than 1 mg require Mass and Volume Section, and the relation special balances (quartz fiber) and handling between the Office of Weights and Measures techniques (working in the field of a micro- and the National Conference of Weights and scope) and a continuous maintenance of both Measures are the services most widely skill and equipment which cannot be economi- recognized and easiest to characterize in cally done at NBS. Where such weights terms of services provided and user are required, e.g., for work with small clientele. The consultation services, quantities of short life isotopes, the NBS supported in part by NBS funding, and by will calibrate suitable "starting standards" "other agency" funding, are the most as in (1) above, and suggest suitable difficult to characterize. Consultations procedures for "working down" at the user's range from informal conversations relative faci ty. to selection of equipment, procedural dif- (3) NBS will assign mass values and in some ficulties, and problem definition, process cases adjust to specified nominal values, to design, development of procedures and the large objects from 50 1 b to 50 000 lb. (The refinement of reference data. The follow- adjustment to nominal value is with refer- ing sections cover the direct services ence to force machines, and the requirement provided by NBS. The indirect, or con- that the suspended weight exert a specified sultation services, are described in force in a given gravimetric field.) The 4.3.2 adjustment of large weights is a matter of convenience, considering the handling prob- 4.2.1 Description of NBS Services lems and shipping costs. NBS will not nor- mally adjust weights below 50 lb. There are two organizations within NBS (4) The section will assign mass values and which contribute to the measurement system. associated uncertainties to any object from

The Mass and Volume Section is particu- 1 mg to 1 kg, provided that the object is larly careful about the interface with the sufficiently stable in mass to warrant the users of the calibration services. In most measurement, and provided that the displace- cases and particularly in the case of new ment volume is such that the object can be users, each user is contacted by telephone handled with existing equipment. to determine what is needed and when. The (5) The section will make weighings of conversation is acknowledged by a formal convenience on any object up to 50 lb or document stating the estimated delivery 30 kg at no charge provided that there is date, the approximate cost, and special no reasonable alternate source. requirements, if any. (6) Where repetitive measurements are There are no clearly identifiable required by other staff members, the section organizational units within the section. will instruct, and supervise measurements While some staff members are more knowledge- made by others on section equipment. able in some areas than others, most can do any of the tasks that the section performs, 4.2.1.2 Volumetric Measurements — Glassware subject to limitations due to the physical strength necessary to handle some of the (1) The section will calibrate the standard larger equipment. items of volumetric glassware, made in ac- The calibration activities of the Mass cordance with the GSA specifications, under and Volume Section offer the following the provision that the requirements of the services essentially "at cost" to the user: specific work demands such a calibration. The user is advised that the calibration 4.2.1.1 Mass Measurement applies only to fluids similar in character to water. (1) For those who desire to do the calibra- (2) The section will, on occasion, assist tion of subdivision or multiples of the unit large agencies, e.g., Food and Drug Admin- themselves, NBS will establish mass values istration, in monitoring the quality of the for suitable pairs of starting standards, products purchased by testing samples chosen ) .

at random from a given lot. The agency dards, or directly by the user. The section makes the decision regarding the disposition will not accept certain types of hydrometers of the lot in accordance with the rules in where the hydrometer scale is not directly the GSA specifications. related, or relatable, to density, e.g., (3) The section will accept lots of pipets lactometers. which are used by the manufacturer in the (2) The section will calibrate solid density maintenance of a quality control system. standards for use in specific measurements. (One method of testing such a system is to (3) The section will determine, or consult put a "known" pipet in the inspection line with, or assist, in the construction of with the "unknowns.") apparatus for determining the density of a wide variety of small objects, e.g., 4.2.1.3 Metal Volumetric Standards gradient columns, special floats and the like. (1) The section will determine the contained (4) The section will determine the density or delivered volume, and calibrate the indi- of fluids within the capabilities of the cating scale on all standard size provers. existing equipment, and provided that the The section will not normally adjust, or fluids do not require special handling pro- seal, the location of the reading scale with cedures beyond that available in the normal respect to the prover body. These items go 1 aboratory directly to the user, with no intermediate steps. The user in turn makes an effort to 4.2.1.5 Mass Measurement Assurance Program monitor the consistency of the section's work. The Measurement Assurance Program, as

(2) The section will calibrate .1 and 1 cu- operated by the section, has not been sold

' 1 bic foot bottles, which also go dir.. to in an aggressive manner. Participation is the user. voluntary, and prior to making the decision, (3) The section will calibrate slicke ju the prospective participant is told quite standards. Adjustment, if required, must be clearly that it requires a lot of work. If done by the manufacturer. on the basis of his analysis he requires (4) The section will conduct a standardizing either (1) the maximum performance of his test, and adjust as necessary, Stillman measurement processes to satisfy the re- standards, e.g., bell provers of 1 cubic quirements of his specific task, or (2) the foot capacity. This is currently done with benefits in the form of reduced measurement a single purpose test apparatus, and gradual effort by virtue of the quantitative know- decrease of requests does not warrant fur- ledge of descriptors of his measurement pro- ther refinement. (Most items of this type cess, he is then encouraged to participate. are calibrated by "strapping," e.g., deter- Approximately six months is required to make mining the displacement volume from direct the MAP operational in the normal mass mea- linear measurement.) surement facility. Beyond that, the fre- (5) The section will calibrate either to quency of usage is up to him. If he has the contain or to deliver, any metal container necessary supporting facilities, he may provided that it is sufficiently rigid to elect to run the program himself with only be handled in the filled condition, that it minimal connection with the section. In can be leveled, and that the end point is other cases, the section may provide all of sufficiently well defined to warrant the the supporting effort. Periodically, the calibration; and further, provided that the knowledge of his process performance param- effort is necessary to the manner in which eters is updated. (At this time all of the the vessel will be used. (This is of course present participants have been updated at limited by the capacity of the existing least once.) Currently some of the users equipment. are hierarchy laboratories, others are in (6) The section will accept tests of various support of certain direct measurement apparatus provided that the results of the requirements. Some do almost everything test are not proprietary, and subject to the themselves, and others rely on the section condition that the results will benefit a for continuous support. large number of users. 4.2.1.6 Office of Weights and Measures 4.2.1.4 Density Measurements The second operational unit is the Office (1) The section will accept hydrometers for of Weights and Measures. The role of OWM is calibration which are of such quality, both to provide leadership and those technical in manufacture and precision, to warrant the resources that will assure accuracy of the calibration effort, and provided that the quantities and quantity representations in items will be used as manufacturers' stan- all commercial transactions for all buyers

40 .

and sellers in the United States, and to which commodities are packaged for retail promote a uniform national weights and sale, and the extent to which voluntary measures system. product standards are being adhered to by In fulfilling its mission, the Office of industry. Weights and Measures engages in a wide range Determining the relationship between num- of activities all of which are essentially bers of package quantities available at funded by NBS. Foremost is the assistance retail and its effect on consumer ability offered to the States in the following to make a value comparison. areas: Distributing published standards and reg- ulations resulting from the Fair Pack-

Laws and Regulations : aging and Labeling Act, providing infor- The development of model weights and mation and assistance to State officials, measures laws and technical regulations and generally promoting uniformity in for the States and local jurisdictions. Federal and State regulation of the la- The evaluation and updating of laws and beling of packaged consumer commodities. regulations at the request of the States. Providing advice and counsel to the pack-

Standards and Laboratories : aging industry on means and methods for The development and dissemination of achieving the aims of the Fair Packaging

design and performance specifications for and Label i ng Act. capacity for use as State and local ref- Technical Training erence, laboratory, and field standards, The conduct of formal and informal tech- and the encouragement of manufacturers nical training sessions for weights and to make available standards that conform measures officials, laboratory technolo- to such specifications. The conduct of gists, and industry representatives.

the New State Standards Project, which Information Dissemination : provides for the issuance of new labora- The preparation and dissemination, in tory standards and instruments and conveniently usable form, of data on modernization of State weights and weights and measures units, systems, and measures laboratories. Consultation and equivalents, to satisfy the particular recommendations on laboratory facilities, needs of the Federal Government, State organization, instruments, and technical and local governments, educational in- procedures. The calibration of secondary stitutions, business and industry, and standards for the States and industry. the general public, including the devel-

Testing and Equipment and Procedures : opment and maintenance of archival and The design of testing equipment and the reference collections of published development of testing procedures for material weighing and measuring devices. The development, the preparation, the The conduct of the master railway track publication, and the dissemination of ap- scale test project and the testing of propriate information on standards, test- commercial track scales operated by the ing equipment, technical procedures, railroads and other industries located technical investigations, and standard throughout the nation. practices.

Conferences and Meetings :

Measurement Studies : The plan and conduct of an annual Nation- The identification, analysis, and al Conference on Weights and Measures, solution of technical problems in the including all secretariat functions to measurement area of commerce. the Conference and its standing The study, including field investiga- committees. tions, of weighing and measuring equip- ment, the preparation of specifications 4.2.2 Users of NBS Services and performance tolerances, and the es- tablishment of standard practices involv- The principal user community of the Of- ing the use of such equipment. fice of Weights and Measures is the State The examination of prototype commercial and local weights and measures officials. weighing and measuring devices and equip- A small, but important, additional service ment submitted by manufacturers for con- of this NBS group is their prototype formance with National Bureau of approval program. This program allows Standards requirements. manufacturers of measuring devices used in trade to obtain approval certifying that

Packaging Practices : the device satisfies the requirements of Market surveying to determine the Handbook 44. Such certification is ac- weights, measures, and quantities in cepted by many jurisdictions in lieu of

41 their own inspection. The program is small from other national laboratories using local (about 40 inspections per year) but growing. organizations as "fronts." In the area of The Mass and Volume Section, on the other mass measurements the results are used as hand, has a much more diffuse clientele. The "informal international audits" rather than most serious customers are those who sub- sources of calibration. scribe to our Measurement Assurance Pro- The higher order calibrations provided by gram (MAP). There are currently 18 facili- NBS reflect the uncertainties from three ties participating in the MAP program. separate measurement processes, the defining The nonclientele group contains all of artifact to the prototype kilograms, those whose measurements are now adequate the prototype kilograms to the ni chrome for the intended purpose, or at least kilograms, and the ni chrome kilograms to the thought to be by the user. artifacts being calibrated. Of these three In order to establish a profile for the processes only the latter has been com- users of NBS services, the customers over a pletely characterized. With the excep- four-year period were categorized according tion of a small group of measurements to nature of their business using the Stan- which were made when NBS-2 was still at NBS, dard Industrial Classification (SIC) codes. essentially no calibrations have been made While it was not possible to categorize all relative to the prototype kilograms. This of the customers according to this method has primarily been due to a lack of equip- of classification (87% for Mass, 67% for ment and to uncertainties associated with Volume, and 61% for density), Figures 5 the air density algorithm. Present calibra- through 8 show the frequency of re- tions are on the basis that the historical quests from various categories over the values assigned to the nichrome kilograms period of the study. are exact. The uncertainties assigned to the various 4.2.3 Alternate Sources pairs of kilograms of the Mass Measurement Assurance Program participants are on Technically the user's decision to accept this basis, thus these laboratories are or reject an alternate source for calibra- capable of providing an alternate source tion should be based on the uncertainty of for most of their requirements which for- the calibration relative to the task at merly were sent to NBS, as well as for hand. Generally, if the user's process others if they desire. (An "informal cannot detect the small differences between international audit" shows that the like objects which are clearly identifiable practical mass scale as disseminated by in the calibration process, the user cannot NPL and by NBS may be offset by one part utilize the full benefits of the calibra- in ten million. This is attributed to tion. Practically, many decisions with re- problems associated with the air density gard to source of calibration are based on algorithm.) Since the consistency of the political or economic factors rather than work of each of these laboratories has technical requirements. been evaluated both relative to NBS and With access to the defining unit, and relative to other participating labora- with the world's most precise balance, tories, one or more could function as an NBS-2, BIPM provides the highest level cal- interim calibration source in the event ibrations. The uncertainty of the results of disaster at NBS. The maintenance of are limited by the instability of the such service is important, not from the defining artifact. BIPM has provided cali- standpoint of providing calibrated weights brations for selected artifacts to be used which are in turn used to calibrate other in redefining certain electrical units, weights, but because wel 1 -cal ibrated arti- i.e. the and the faraday. BIPM has facts are needed to characterize modern provided calibration services at the kilo- weighing devices and measurement processes. gram level for at least one foreign balance With the completion of current manufacturer. While BIPM is not in a posi- activities directed toward the development tion to provide general access to the de- of equipment and procedures and the evalu- fining artifact, the precedent has been ation of the air density algorithms, it is established with regard to providing support planned to re-calibrate these pairs of for certain types of activities. kilograms relative to the prototype

While all national laboratories are ki 1 ograms. acceptable alternate sources for calibra- Perhaps the most widely available cali- tions with regard to military specifica- bration procedure is that associated with tions, most national laboratories limit establishing compliance with adjustment their services to organizations within the tolerances. Here the user's choice of territorial boundaries of their governments. source may be either practical or economi- On occasion users may obtain calibrations cal or both. The state laboratories and 42 I

3 CU s E o 3 H o M 4-1 3 CU cfl T3 > U CO O o H OJ i-t u 3 cj 3 •rl 3 CU 3 l-l T3 E i O O P-t cfl E •H CO cu •rl o o cfl CO to 3 Pn 3 o u 4-1 cu cu -a CT 1-1 3 cu •H o 3 W U >-i o CO 4-1 4-1 4J < CO C CU CO •V CO C u i— cu o 4-1 C c o 3 CU E5 T3 S CO o to 4J T3 E w C 4-1 u OJ C 3 s=> CO cu t3 tfl C/3 H u o- c CU 4J M cu w cu CO 4-1 o cu 0) > Pi CJ CO O 00 o. o O o s u (J a Cfl c (0 e •H to CU CO •rl u c CO > a E M 3 cu c CO s 1—1 ,n C3 T3 •rl CU t-H t-l 3 cO rl C/3

a Inrinrin

rO

Figure 5 Mass Customer Profile (Major SIC Groups)

CO 3 4-1 uCfl

cfl rl a. 0i 4-J

cfl 13 •rl H cO rl O O CO CU •rl E cu c .cO 3

3 -a (3

CN m 00 r-l CN

43 i

CO cO

CT) •H )-l CO C 01 4-1 cO c O 4-1 CJ 3 •H CO 3 3 T3 1 4J a T3 0) 4J O Cfl CO O c CO CO M •H 4J CJ M rl a Z M 3 (X 01 •H 60 Cm S3 OJ c C er o e C CJ c •H n •H W a OJ CO o M C a) CO CO S a CO cu rl cj cj e c 4J O OJ B CO PM cu o IW (1J x cfl o cd C X* oo o X 2 cu s w £ •H u 3 B c_> rH -a W pi 4J H CO H •rl o B & -a ai rH CO •rl cO to CU ft) 4-J CU X) S-i oo o_ c a i— $ 03 M cj U c C o b cu a rj cu CU i—O, cd o > •rl CJ rH c u & S-i !-i i-> 4-1 t-i H O n! —4. CO c CO H 3 B u B & CJ fl 0) 01 3 00 »rl U CJ o CO CU 4J tu u •rl u 1) ct) •H ft 3 £> a) cO a. X- t-l ca H CJ H C M 00 CO r-H o 01 •H u (X Pm W H C/3 Q < S P-t

CN 00 CTi o CO oo oo oo a\ 00 CN C I CO CO Figure 7 Volume Customer Profile (Two and Three Digit SIC Groups)

CJ 3 XI 3 O X! u O CJ s 3 cu X) B O 1^ ex U cu >-i •H (3 PH -Si CU 3 CO E cu c cr o 0) C X) ex rl W •rl XI o Qj OJ XI B O a CJ C cu B X) cfl o x: CO o c a a •H CJ 3 o 4J CI) o •rl CO cu S _±£ a V-J U CO a 0) o cd s CJ 4-1 •H •H co C S U CO c T3 B C 01 •rl cfl 3 0 u cu cfl •rl CO oo x) CO •rl o CO C CU u CJ (-J 3 rl 2 U H H 00 CO a n •<) vo 00 CTi O o o o 00 CN CM CN CN

Figure 8 Density Customer Profile (Two and Three Digit SIC Groups,

44 . " . e

the weight and balance manufacturers pro- Staff Funding ($1000) vide calibration services complete with Reimbursabl proper evidence for compliance with Yr Total Prof NBS Work Total

"traceabi 1 ity" requirements. Many former 1970 21 6 300 160 460 users of the NBS services now utilize these 1971 17 6 325 165 490 sources with substantial economic savings. 1972 17 7 380 no 490 (Some states provide such services at no 1973 9 2 210 60 270 cost.) In some cases NIBS calibration 1974 10 3 205 35 240 services are used for this purpose, i.e. the NBS assigned values are compared with The large change between 1972 and 1973 was the adjustment tolerance limits to verify the result of the establishment of the compliance, and the limitation on the use Dimensional Technology Section. The work of is then set by the tolerance class rather maintaining and characterizing the measure- than the uncertainty of the value. In ment processes of NBS, and general support •these cases the decision is political to the appropriate segments of the national rather than economic, i.e. use the NBS measurement system, accounts for most of the calibration services in lieu of justifying work and is funded directly by NBS. Mea- an alternate source. surements which are done for others are, in There are no technical reasons to most cases, charged on the basis of "at support the NBS calibrations in the areas of cost. volume and density measurements. The The staff profile of the section does not methods used at NBS are well documented and reflect the utilization of the services widely disseminated. Adequate equipment is provided by both the Office of Measurement generally available in the user's facility. Services and by the Statistical Engineering However, with the exception of IRS (hydrom- Laboratory. In both cases, many of the eters) and FDA (glassware) previously men- respective staff members are considered as tioned, attempts to develop alternate "ex officio" members of the Mass and Volume sources have not been successful. The rea- Section. sons for this are somewhat complex. In the The Office of Weights and Measures case of the small user, i.e. a hospital lab- operates under a total budget of just under oratory, the practice of self-calibration $450,000 with a staff of 18. diverts resources from the primary mission of the laboratory. Further, such practice 4.2.5 Mechanisms for Supplying Services raises the question of defense in the case of challenge by those who make judgments Access to the unit is provided through based on the measurement output of the the calibration services. The transfer laboratory. mechanism is the Report of Calibration which For large uses, i.e. the petroleum indus- is returned to the submitter along with the try (volumetric process), the alcohol and calibrated items. In the case of mass cal- sugar industry (hydrometers) there appear to ibrations the report is essentially a com- be two reasons, both associated with the puter prepared "laboratory notebook" which interface between the producers and users of not only gives the results of the calibra- measurement data. One is the need for a tions in several different forms, but also consistent defensible reference. The second completely delineates the details of the is the widespread belief that where large measurements which were made and the status amounts of are involved, measurement of the performance of the measurement pro- data will always be biased in the favor of cesses used. All calibration reports show the organization which has produced the the conditions under which the results data . In such an environment the need apply, as well as realistic certainty state- for unbiased "third party" calibrations is ments which permit the user to evaluate the essential results relative to his requirements. In some instances, where there is a history of 4.2.4 Funding Sources for NBS Services NBS calibration of certain items, the pre- vious history is taken into account in the The two NBS organizations which are assignment of value, and associated administratively separate have different uncertai nty funding patterns. TT5 There is a published strategy for The Mass and Volume Section profile, both biasing measurement data where dis- for funding and staff, is shown below. agreements are resolved by NBS mea- surement with the party having the greatest discrepancy with the NBS paying the cost for the referee measurement.

45 The documentation associated with the The Office of Weights and Measures, Mass Measurement Assurance Program includes acting as the secretariat of the National the calibration of the participant's stan- Conference of Weights and Measures, provides dards, the characterization of the partici- both administrative and technological ser- pant's process based on measurement data vices to that organization and its various from his facility, a verification of the study committees. OWM provides a calibra- consistency of the results through indepen- tion service to the state laboratories for dent calibrations of an arbitrarily selected items which are not designated "primary" set of weights by both the participant and standards. The items include trailer NBS. In addition, the participant is pro- mounted provers used to test a variety of vided with control chart data which will tanks and metering devices, large weights permit a continuous monitoring of his pro- used for testing large capacity scales, cess performance. Follow-up documentation miscellaneous glassware items used for may include recalibration of his standards testing liquid packaging containers, and at NBS, recalibration of his standards in the like. OWM conducts extensive training his facility relative to furnished NBS operations for all levels of personnel standards, a re-evaluation of the consis- within the state weights and measures tency of measurement between his facility organizations. OWM also operates a pro- and NBS, and finally, an occasional "update" gram designed to verify the consistency of of pertinent parameters for his process. mass measurement in the various state labor- Detailed access to measurement methods atories. OWM operates, on an intermittent and procedures is provided through the gen- basis, the railway track scale test cars eration of a literature base, and through which provide a means for calibrating rail- consultation. The literature base consists way master scales, and such other private- of a series of publications ranging from ly owned scales as arranged for in the "OVERLAP," an irregular news letter to MAP course of planning the itinerary for the participants, to NBS reports, Technical two test cars. Notes, Monographs and formal publications in a. journals such as "Metrologi " For the most 4.3 Impact of NBS Services part method and procedures are covered in detail in NBS reports. The intent, in gen- 4.3.1 Economic Impact of Major User eral, is to provide a depth of detail suf- Classes ficient to not only permit the interested user to make the measurement, but also to The user classes, discussed in permit him to assess the uncertainty of his Section 4.2.2, show a degree of commonality result. The methods presented can be sim- between the three areas of measurement. plified as appropriate to his requirements, This in turn suggests that there may also or to the requirements addressed by be a commonality between all of the users voluntary standards groups. of the NBS calibration services in all mea- Tutorial assistance has been provided surement areas. In considering impact, it through participation in seminars conducted seems important to identify this common- by the Mass and Volume Section, and by the ality in order to differentiate between the Office of Measurement Services. On occasion users who are in essence "told" to use NBS invited talks have been presented at various (e.g., by contractual requirements) and society conferences, including the National those who use NBS calibration services as Conference of Standards Laboratories, the a matter of need or convenience. This is IEEE, the American Defense Preparedness not to imply that those who are "told" to Association (formerly the American Ordnance use NBS would not normally come to NBS Association), the Aerospace Industries anyway, but rather to differentiate between Association, and others. The most effective realistic calibration needs and the artifi- access to methods and procedure, however, is cial needs generated within the

direct consultation relative to a particular i nf rastructure. measurement requirement. These consultations Table 8 is a summary of all of the users range from inquiries concerning the avail- of all NBS calibration services, by dollar ability of appropriate standards to system volume, and by number of facilities served problems concerning the health and welfare for fiscal year 1975. The military con- of the nation (e.g. determining the amount tractor category of the private sector in- of coal dust in mines, and the costing of cludes 70 corporations which account for in measurement in laboratories of the Food and excess of 87% of military research, devel- Drug Administration) and large segments of opment, test and evaluation contracts the national economy (e.g. bulk measurements ($5,474 billion); 46 of these 70 corpora- of petroleum and nuclear materials). tions, which account for 50.9% of the RDT&E contracts, are large industrials

46 1

(Fortune's "Top 500"). The total industrial generally end-use oriented, accounts for sales of these 46 (1974) is $129,463 bil- 65% of the total NBS calibration effort. lion, thus the military business is only Non-NCSL members with accounts over about 2.5% of their total effort. Further, $5000, of which approximately 50% of the total sales of these 46 represent only the $225,910 is associated with the about 15.5% of the total sales of the "Top calibration of products from less than 10 500" ($833,956 billion or 59.3% of the instrument manufacturers, accounts for 1974 GNP, $1,406.9 billion). 12.7% of the NBS effort. For the remainder, 22.3%, the average cost per facility, for 678 facilities, is slightly Table 9 Survey of Users of All NBS less than $600. An unidentified number Calibration Services (1975) of the facilities in this group are sub- contractors which are required to No. of "maintain" certain measurement capabili- Cost of Facilities ties, and, on occasion, items submitted

Calib Served for periodic recal i bration have been received in the same unopened package Total Govern- which was used to ship the items after the ment $ 542,812 63 previous calibration. The above summary of the users of NBS Dept of calibrations is descriptive of the users of Defense $ 486,085 (89.5%) (35) mass, volume and density calibrations. The impact of the NBS services is the degree to OTHER 56,727 (10.5%) (28) which these services benefit the user in attaining measurements adequate for his needs, either realistic or imagined (e.g., established by a bureaucratic infrastruc- Total Private $1 ,234,844 871 ture). The degree to which direct services are required is the ratio of an acceptable Military Con- uncertainty of the results from the user's tractors $ 615,050 (49.8%) (224) process, under actual conditions of use, to the uncertainty of the NBS calibration mea- OTHER $ 619,794 (50.8%) (647) surements. For example, the ratio of the uncertainty of the internal balance weights of one balance manufacturer who is a parti- cipant in Mass MAP, to the uncertainty of Overal $1 ,777,656 934 the NBS calibration is on the order of 1.5 to 2, obviously requiring direct contact. NCSL Member- As a result, the dissemination of the mass

ship $1 ,157,862 ( 65%) (238) unit as embodied in stainless steel is widely available with little degradation. Non-NCSL $ 225,910 (12.7%) (18) On the other hand, the same ratio for (accounts acceptable commercial weights ranges from over 100 to 1000, thus adequate services are $5,000) available from many sources and those who insist on NBS calibration are merely buying OTHER $ 393,884 (22.3%) (678) a name. Generally speaking, there are very few uses of mass, volume, and density (Average cost/user $580) calibrations with requirements commensurate with the uncertainty of the calibration The National Conference of Standards resul ts. Laboratories is a non-profit, laboratory- In contrast, the consultative activities, oriented organization which promotes co- from short telephone discussions to operative efforts toward solving common long term continuing studies are directly problems faced by standards laboratories in related to specific measurement require- their organization and operation. With few ments or measurement assurance require- exceptions the membership is comprised of ments. Two such long-term activities are government laboratories supporting hier- related to the nuclear power industry and archy chains of calibration laboratories, the petroleum industry. and organizations which must interface with In the nuclear power industry, the first contractual requirements imposed by govern- quantitative measure of the amount of mate- mental agencies, primarily the Department rial in a "spent" reactor core occurs after of Defense. The membership, which is not the core material has been dissolved from

47 . F

its cladding. The amount of material is reduction table accepted as an industry determined from assaying samples drawn from standard. Discrepancies in the table are a tank of "known" volume. The result not amplified when large temperature and density only establishes the monetary value of the changes occur in transit, e.g. on-loading material, but also establishes the base for in the middle East; and off-loading at a material accountability under the nuclear North Atlantic port. safeguards program. The Mass and Volume The current volume reduction table is Section has developed and documented mea- based on thermal expansion coefficients and surement techniques for determining the densities of North American crude oils and contained volume of large tanks. Working products prepared by NBS around 1915. In with voluntary standards groups, the tech- 1973 production from North American fields niques are being incorporated in standards had dropped to 25% of the total production. which will become industry standards. This It has been known for some time that the type of measurement support activities at 3 table does not reflect the characteristics U.S. government reprocessing facilities which of presently marketed crudes, however, the handle material used by the military. Should error relative to the cost of the material commercial reprocessing activities be au- was considered acceptable. At the current thorized for the U.S. the section's activ- price for imported crude oil, a .16% error ities in this area would support that over a temperature differential of approxi- effort. The projected value of the re- mately 40°F represents an inventory loss in sultant electric generating capacity for transit of $5 million per month at the cur- 1985 from reprocessed fuel is in excess of rent import rate of 8 million barrels/day. $5 billion. Should the plutonium fuel cycle The joint NBS/API activity is to character- be used then tank volume measurements be- ize both crude and products on a worldwide come critical for accountability and safe- basis (67% of 1975's proven reserves). This guards purposes. data will be the basis for a consumer's The Mass and Volume Section supports the decision relative to both the method for petroleum industry in three ways: the cal- reducing values to a common basis and the ibration of large volumetric provers which measurement detail associated with deter- are used worldwide to test metering devices; mining the appropriate parameters. the calibration of hydrometers, density being an important parameter in processing 4.3.2 Technological Impact of Services and application; and currently the American Petroleum Institute/NBS Physical Properties Historically, NBS impacted the weight Project. An estimate of the worldwide com- manufacturing industry through its technical mitment to petroleum measurement is shown publications such as Circular #3 and 547. in table 9. The tolerance classes set up in these The average "barrel" is measured twelve documents are widely accepted. times with numerous custody changes between The impact on the precision balance the well head and the ultimate user of the industry has been less profound due perhaps finished product. Shipper/receiver differ- in large part to the fact that this ences in terms of dollars are based on bulk industry is largely European based. How- measurements of quantity, which in turn is ever, at the 2500 lb level the "Russell" dependent on the properties of the material balance, named after a Bureau employee, is being transferred. Measurement detail is a the standard of the nation. More recently, consensus decision establishing the basis the NBS-2 type of kilogram comparator, man- (e.g., NBS calibrated provers) and the ex- ufactured by a U.S. concern, has found a tent to which systematic errors are mini- market among other national laboratories. mized with respect to the cost of the Commercial weighing devices manufactured material and the cost and practicality of to conform to Handbook 44 also reflect NBS measurement procedure. Attention to detail thinking. is directly proportional to the cost of the With respect to volume standards, the material NBS role has been less innovative, but the The worldwide pricing structure is based methods of calibration throughout the upon the bulk volume and density at 60°F. industry owe a great debt to NBS. The Since custody transfer takes place at tem- continuing calibration effort, although peratures differing from the standard tem- small, provides the unifying tie point perature, computation of the volume at 60° for a sizable, but diffuse, industry. is done using the measured temperature and The impact on the makers of density the density at the time of transfer (e.g., measuring equipment has been of yet another as measured with a hydrometer), the kind. In this case, the detailed techno- measured value as determined by a logical base of density measurement was calibrated metering device, and a volume developed at NBS. The proper configuration, 48 Table 9 PETROLEUM MEASUREMENT COMMITMENT

Crude Finished Comroi tment Petrol eum Products Total

Measurement $800,000,000 $2,000,000,000 $2,800,000,000 Facili ty Investment

Equivalent 22,500 27,500 50,000 Ful 1 -Time Personnel

Measurement 255,000 395,000 650,000 Poi nts*

Frequency "Average" Barrel Measured 12

Cost of 0.5% Measurement Error Per 1000 Barrels Petroleum at $13.00 Per Barrel $65.00

Monitoring discrepancy per day at cur- rent import rate $500,000.00

*Tanks, pipelines (meters), tankers, barges, tank cars and tank trucks

ESTIMATED WORLD-WIDE OIL INDUSTRY INVESTMENT IN BULK PETROLEUM LIQUID MEASUREMENT

methods of use and necessary precautions The section staff members are consultants are all spelled out in NBS technical to the Institute of Nuclear Materials Manage- papers. The industry is based on these ment (INMM) Writing Group 8 and subgroups principles. thereof which have the task of preparing At a higher level of accuracy, the NBS specifications or guides for the measurement techniques of hydrostatic weighing repre- of nuclear materials in cooperation with sent the "state-of-the-art" and are uni- ANSI. The specifications in preparation are: versally recognized as such. This work (1) ANSI Nl 5.18 "Mass Calibration Techniques continues with the silicon density work for Nuclear Materials Control" previously mentioned. (2) ANSI NT 5.1 9 "Volume Calibration for NBS has a technological input to other Nuclear Materials Control" specification-preparing groups, both The above standards, when accepted and government and private. For example, the formally approved in accordance with ANSI appropriate section of the NASA Mass Prop- procedures, will no doubt be made mandatory erties Manual were prepared by the section. by NRC (Nuclear Regulatory Commission) in Procedures and statistical methods prepared the nuclear industry. by the section have been incorporated in The staff has participated in the work of various published API procedures. Where ASTM subcommittee El 8, Hydrometer Subcommit- possible the section has taken the initia- tee. In early 1974, the parent committee, tive to see that the procedures are appro- El, on Methods of Testing, elected to dis- priate for the task, that the user band itself. The subcommittee El 8 has been understands what he is supposed to do, and redesignated El .05, under a different parent that the equipment furnished will indeed committee. The section will continue to be do the job. a part of this committee because of the

49 strong reliance of the industry on artifact parameters. For the first time realistic standards. uncertainty statements, based on demonstra- The National Conference on Weights and tive evidence, permitted unbiased comparison Measures had its beginning in 1905, when of measurement results from different mea- the Director of the National Bureau of surement processes. Realistic uncertainty Standards called a meeting of representa- statements, in turn, permitted an assessment tives from the several States of measurement process detail relative to "to bring about uniformity in the State end use requirements. laws referring to weights and measures, The techniques were extended to the field and also to effect a close cooperation by changes in reporting format, tutorial between the State inspection services training at NBS, and technical publications. and the National Bureau of Standards." This effort permitted the development and From an initial attendance of 11 persons, acceptance of alternate sources for routine the Conference has grown until now it calibrations for both weight sets from 100 g brings together at its meetings a total of down, and for volumetric glassware. Weight approximately 500 persons comprising manufacturers, utilizing adaptations of the weights and measures officers, other offi- techniques used at NBS, were able to market cials of the Federal, State, and local calibration services acceptable to a large governments, and representatives of equip- portion of the user community at a cost to ment manufacturers, industry, business, and the user substantively less than the cost consumers. The Conference develops many for NBS calibration. In the area of volu- technical and general recommendations in metric glassware, the Food and Drug Adminis- the field of weights and measures adminis- tration was the primary user of the NBS tration, and its programs explore the calibration services, primarily to verify entire area of this economically important compliance with specification tolerances. segment of governmental regulatory service. NBS efforts devoted to the development and The participation is in national and acceptance of suitable specifications, to international activities in the fields of the verification that manufacturers' pro- weights and measures standards and prac- ducts did indeed meet these specifications tical metrology. Such participation resulted in a change in the method of pro- includes: cessing glassware. The agency now procures (a) the development of and negotiation glassware through the established procedures toward the adoption of both national and of the General Services Administration with international specifications in the a direct saving in excess of $150,000 per assigned area; year, the amount previously spent annually (b) correspondence and other communication for NBS calibrations. with experts of the United States and of The reduction in the number of calibra- other nations; tions for manufacturers and suppliers (c) technical committee activities; opened a direct channel of communication (d) attendance at and participation in with people involved in operating measure- appropriate meetings and conferences of ment processes. As the characterization of the Federal and State Governments and of the NBS processes progressed responses to regional groups, business, industry, and inquiries became objective rather than educational institutions. subjective. Ultimately, the routine "fee schedule" calibrations were replaced by "at 4.3.3 Pay-off from Changes in NBS Services cost" calibrations in which the calibration effort was tailored to the particular cus- Substantive changes resulted from two tomer's needs. The measurement assurance factors; one, a concentrated effort in the program concept was developed and augmented. early 1960's to completely characterize the This concept has now been accepted in many mass measurement processes at NBS, and two, other areas of measurement. The concept is the recognition of measurement as a produc- used in the nuclear weapons industry to tion process whose product (measurement monitor the performance of the facilities results) should be evaluated relative to the involved. As a direct result of this pro- end use rather than relative to state-of- gram, the section was asked to make an ob- the-art measurement capabilities, e.g. the jective survey of the nuclear reprocessing "best" measurement process is one which industry relative to mass and volume produces adequate results in the most measurements. economical manner. As a result of the first Continued efforts devoted to the charac- effort, the basis for judgment concerning terization of measurement processes and the the adequacy of a measurement was changed development of measurement methods provided from operator opinion to quantitative tests direct support to accomplishments ranging relative to pertinent process performance from the redetermination of Avagadro's

50 -

Number, to the resolution of litigation con- time, the consumer is assured that the cerning the performance of a group of 30 kg requirement is satisfied. balances procurred by OWM for the new state Perhaps the largest benefit from change laboratories. Studies on the density of is the progress which has been made in single crystal silicon resulted in a "solid relating the ideas of Mach, Carnap, object" density standard overcoming the lim- Eisenhart, and others to modern measurement itations associated with water. Procedures processes. It is now possible to describe were developed and documented which would most measurement processes in generalized provide a base for determining the volume of terms independent of process detail. This large tanks in the nuclear industry. permits a logical delineation of require- Detailed analysis of the calibration user ments relative to a particular task, which clientele and the way in which the users in turn provides a list of the elements utilized the calibrated items were made. In which must be tested to establish that the many instances it was questionable as to performance of that process is adequate for whether the NBS calibration efforts had any the intended usage. As this work is refined real impact on the quality of the work done and disseminated, it will impact the design by the user. It was concluded that the and development of future measurement "catalog-fee schedule" invited commitments systems. One of the encouraging factors of NBS facilities without prior discussions relating to this work is the ever increasing concerning the adequacy of the effort rela- number of people who are now aware that tive to the particular task. Many requests there is something more to measurement than for calibrations were, and still are, possession of adequately blessed standards required for compliance with various stan- and the blind following of a specified dards. It was decided that participation in procedure — people who can discuss in some the preparation of voluntary standards detail measurement processes, process should be limited to consulting rather than characteristics, and measurement require- direct participation. It was felt that as a ments relative to an identified need of consultant, NBS could aggressively increase importance in accomplishing a specific the span of input to the voluntary standards task. procedures (see figure 3). In the case of old, established standards, this has not 4.4 Evaluation of NBS Program always been successful. However in the de- velopment of voluntary standards for the The two groups within NBS which support nuclear industry, NBS, acting as a consul the mass, volume and density system are ant, has an effective input over the whole organizationally separated with essen- span of the process, from the identification tially unrelated missions. One group is and quantification of the requirements by concerned with the legality of measurement the regulation, to the development of and equity in the marketplace, particularly methods and techniques, the preparation of in regard to consumer products. This group the appropriate voluntary standard, the is, in a sense, semi -regulatory . That is, demonstration of the adequacy of the stan- having created a system for control largely dard, the acceptance of the standard by based on visibility which has become a con- the regulator, and the manner in which com- sensus system for the various states, this pliance with the standard will be judged. group is mostly concerned with implementa- The increased span mentioned above has tion of the system. The system controls a resulted in a fundamental change in the large segment of the instrumentation indus- scope of certain new voluntary standards in try as well as the control measurements at the nuclear industry. Traditional older the consumer level. Typical of all large standards are essentially detail standards consensus systems, it cannot respond rapid- specifying itemized equipment lists and ly to changes in technology. The ritualis- procedural steps; such standards, essen- tic aspects of the effort tend to take tially without exception, ignore discussing precedence over the efficiency of the ef- in detail the uncertainty of the result. fort and, in like manner, administrative (It is presumed that adequacy relative to activities predominate. While the techno- a particular requirement has been con- logical requirements of the system are not sidered by the committee which prepared stringent, the lack of technological the standard.) The new standards are resources seriously limits the scope of performance standards, concentrating on the programs. the methods for assessing the performance The second group is concerned with the of existing processes relative to speci- adequacy of measurement throughout the fied requirements. In a producer-consumer system. The constituency of this group is relationship, this action gives the pro- voluntary with one exception. As "keepers ducer a freedom for choice and at the same of the kilogram" the group is a "servant "

to all who desire, or are directed to associated with measurements and less establish, a direct traceability to the emphasis on attempting to transfer pre- unit." This group, in the early 1960's, cise absolute measurements. Greater aggressively undertook a long-range plan confidence can be placed on the use of to change from merely being "keepers of the the measurements and standards at lower kilogram" to becoming a center for measure- costs, both to The National Bureau of ment excellence in the areas of assigned Standards and to commerce and industry. responsibilities. One of several goals was The panel believes there would be bene- that no current or future task should be fits in extending the assurance program limited by the inability to make adequate in the future to all dimensional measurements. Through the mechanism of measurements. characterizing the calibration processes, Thus, over a six year time span, the the basis for measurement assurance was intent of the program was finally under- established, e.g. operationally verifiable stood, at least with regard to the trans- evidence that process performance is as fer of the unit. It would seem that more expected. The efficiency of the measure- attention to the educational aspects of ment effort was addressed by emphasizing the program would have considerably reduced that the "best" process was that which the friction and would have shortened the produced adequate results with minimum time span between introduction and effort. Cooperative programs were es- acceptance. tablished with other facilities to study With one exception, the measurement factors affecting the consistency of assurance approach is not reflected in measurement over time and over the voluntary standards. The exception is environmental conditions in which with members of the Institute for Nuclear measurements must be made. Material Management who are incorporating By and large, the transition has been the approach in some of the voluntary successful. Measurement assurance rela- standards being prepared for the nuclear tive to the transfer of the unit has been industry. There may be several reasons for adopted in several measurement areas. this, one of which is the lack of aggres- Even this, however, has not been without sive or appropriate selling on the part of difficulty, as illustrated by the various NBS. Traditionally, voluntary standards National Academy of Sciences Evaluation are consensus standards in which the pre- Panels. paring committee assumes the responsibility In 1969, the evaluation panel for the adequacy of the specified procedure

recommended: . . . extension of the Pilot and interpretation relative to a specific Program (now called the Measurement task. Few if any such standards include Assurance Program) into the volume error analysis or any discussion of methods calibration activity, specifically, and to assess the uncertainty of the result. into other NBS calibration activities, Individuals who constitute these commit- generally . . . tees, however, recognize that while such As the Measurement Assurance Program standards may be necessary to order in the was developing in the area of length marketplace, or in the simplification of measurement, in 1970 the evaluation panel the exchange of data, the use of these was divided in opinion to the point that standards is not always sufficient rela- a minority evaluation was submitted. It tive to the task at hand. The resulting was quite clear that NIBS had failed standards tend to represent a minimal con- miserably in presenting the intent of the sensus relative to an interfacing problem program. rather than realistic efforts at the point In 1972, the panel "endorses" NBS of decision. policies pointing to expansion of the MAP The above implies that if the benefits program, and in 1975, the panel report of measurement assurance are to be realized states: in industry, there must be a one-on-one "The panel is extremely pleased to see the interaction at the point where the benefits progress which has been made in the Mea- are real. With regard to NBS as a whole, surement Assurance Program in the Mass this is a null effort for two reasons; one and Volume and the Dimensional Technol- obviously the futility of a one-on-one con- ogy Sections. We believe the measure- tact with everybody, and the second, the ment assurance approach will provide a absence of an identifiable measure of more viable method of transferring stan- success. If a company should adopt mea- dards from the National Bureau of Stan- surement assurance procedures with any dards to commerce and industry by measure of success, and there have been placing a greater emphasis on measure- several which have seen fit to do so, the ment assurance and the error bands fact is immediately privileged information. 52 It would seem inevitable that as the kilograms and the transfer of the mass unit constituency grows, the philosophy would as embodied in these artifacts to artifacts become identifiable in future voluntary made of other materials such as stainless standards. Such an effort is a long, steel, or single crystal silicon. The slow process, as is any worthwhile reform, technology is available and it is only a with periods far longer than the normal matter of concentration of effort. While budgetary cycle. NBS will have a role in this work, relative With regard to other aspects of the to practical mass measurement, these prob- program of the latter group, the range of lems are largely academic, the resolution activities has increased greatly over the of which will be beneficial to a relatively past 15 years. "Customized" at-cost cali- small segment of the scientific community. bration services are furnished in all Undoubtedly, isolated efforts will be made measurement areas. There are active to construct a mass unit based on a con- efforts with tangible results ranging stant of nature. Should these efforts be from the development of precise research successful, the results will be limited by weighing devices and the study of air the complexity of the measurement technique density as it relates to mass measurement to definition of the unit and is not likely to development and tutorial efforts con- to be competitive with the artifact-based cerning bulk mass and volume. system for some time. The program details reflect a With regard to dissemination, the consideration of the question, "What is current practices originated at the turn required?", "When is it needed?", "Will of the century. Today, there is no tech- large segments of the user community nical reason why BIPM cannot be given the benefit in any way?", "Will the tech- responsibility for the transfer of the nical competence of the section improve by mass unit from the defining artifact, what- undertaking the task?", and "Are the neces- ever it may be, to practical artifacts sary supporting services available?" With for all who have need for direct access to regard to the latter question, this group the unit. In fact, such is being recom- has been particularly successful in utiliz- mended informally to emerging nations by ing NBS resources outside of the group, and both BIPM and others who are knowledgeable other facilities in efforts supporting the concerning mass measurement. A decision various tasks. to centralize first level dissemination of the unit would be advantageous from 4.5 The Future both an economic and a scientific point of view. It would, however, effect the cur- The most striking characteristic of the rent roles of some of the various national

Mass, Volume and Density Measurement System 1 aboratori es. is its apparent stability. For hundreds of One factor which will set the level of years, it has evolved by a slow linear ex- the routine access to the unit as current- trapolation, each change firmly based on ly provided by NBS is dependent upon the past experience. This apparent stability degree to which those who provide goods comes in part from inertia of all parties and services are required by law or regu- involved; the scientists and educators, the lation, to defend their measurement suppliers of equipment and the users, and practices or results. Most of the current in part from the fact that the needs which requests for calibration in mass, volume were the genesis of a formalized system and density could be technically satis- have not changed in centuries. The impact fied by referral to others, e.g. state of near revolutionary changes over the past laboratories, manufacturers, commercial 25 years has influenced not only the entire laboratories, or by the users themselves. system, but the current role of NBS and Such facilities, however, do not general- other national laboratories and will indeed ly have sufficient stature to alleviate influence the future as well. questions which could arise in a defense Considering the system as two elements, situation, e.g. questions concerning one including activities associated with adequacy of method and results, questions the definition of the unit, and the other concerning "unbiased" results, questions including activities associated with the concerning the adequacy of alternate dissemination of the unit, the future of methods and procedures relative to those the former element is by far the easiest which are specified in contracts or reg- to discuss. ulations. The prudent entrepreneur is With regard to the defining artifact, the now reluctant to utilize services other national laboratories collectively will re- than those of NBS and if increased pres- solve the problems associated with the vari- sure is placed on measurement defense, ability of the platinum-iridium alloy will be more so. Increased pressure on 53 the defense of measurement results will cannot be resolved by NBS, e.g. social bring an increased calibration load, or implications of change and a large scale will require a mechanism to extend the education program relative to the funda- present NBS "disinterested third party mentals of measurement. While NBS is a blanket" to cover other facilities. partner in the literally millions of mea- In terms of the immediate future, NBS surements which are made each day, by must continue to maintain and provide providing access directly or indirectly access to a practical mass scale of suf- to the unit as requested and by acting as ficient quality to serve all of the needs an unbiased reference in the case of of the country. The mechanisms for such dispute, it is presumptuous to assume access, however, are influenced by both NBS to have much, if any, real leverage the quality of the instrumentation general- relative to the activities of the system. ly available to the user and the level of Over the course of history the develop- adequacy required for his particular task. ment of commerce and industry has been Transfer of the unit to certain equipment first the recognition of need and then manufacturers with minimum degradation the application of ingenuity to satisfy will be a continuing need. With regard that need. This has not changed in prin- to the general user, however, the quality ciple, only in complexity. Scientific data of the instruments has reached the point and measurement fundamentals are partners where sophisticated time-consuming efforts in such a system, while standards serve are necessary to characterize their perfor- only to smooth the interactions between mance. Simple procedures to verify ade- various elements of the system. Very quate performance relative to end use few tasks require the ultimate measure- requirements will not require traditional ment capabilities, e.g., studies of access to the mass scale. This in turn fundamental constants and certain will further erode the calibration work classified requirements. For the load. remainder, there are numerous solutions There are requirements for mass, volume which will produce adequate results, the and density measurement instrumentation final form of which depends on many (or systems) which are beyond the capa- factors. bilities of commercially available equip- In one sense, NBS has utilized its basic ment. For economic reasons, commercial assets (motivation, resources and time) to equipment features rapid, convenient mea- develop a competence in determining real- surement at a precision level adequate for istic requirements and in provided guidance the largest majority of users. It is not in the development of processes which will practical for most manufacturers to meet such requirements in the most economi- consider modifications to production cal manner. There are, however, economic- units to improve performance beyond an political -social forces, both with NBS and order of magnitude or to undertake the within the system, which strive for a more development of precise instruments for a tightly governmental ly controlled system limited market. An increasing role of NBS resulting in new responsibilities for NBS in areas where the requirements are beyond which may or may not be within the present the capabilities of commercial equipment capabilities of the organization. In the is consultation relative to increasing the absence of major change, as might be im- precision of measurement and, in situations posed on the structure by these forces, where the need is relative to the health the current level of effort at NBS is con- and welfare of the nation, the development sidered sufficient to maintain a minimal of adequate prototype measurement systems. level of competence, to maintain an ade- For such services to be effective, however, quate access to the unit, to maintain requires that "hands on" measurement reasonable progress in the advancement of activities, e.g. the providing of routine measurement application and capability. calibration, must remain above some Decreases in staff size or level of support "critical level," even if artificially would result in essentially zero progress generated. in one or more of these areas.

5. SUMMARY AND CONCLUSION

The national measurement system for mass, volume and density is stable and ade- quate relative to the needs as currently perceived. There are some questions con- cerning the efficiency of the current mea- surement efforts within the system which

54 , .

[12] Belew, L. F., Stuhlinger E., Skylab--A REh ERENCES Guidebook, National Aeronautics and Space Administration (EP-107), pp. 161, [1] Childe, Gordon, V., The Prehistory of 179, (1973). Science: Archaeological Documents, Part 1, The Evolution of Science, Readings [13] The International Bureau of Weights and from the History of Mankind, edited by Measures, 1875-1975; NBS Special Metraux, Guy S. and Crouzet, Francois, Publication 420. pp. 72-75 (New American Library (Mentor Books), New York, 1963). [14] Moreau, , The Genesis of the Metric System and the Work of the International [2] Cajori, Florian, Translation and Bureau of Weights and Measures, J. Chem. explanatory appendix, Sir 's Educ. 30, p. 3 (Jan. 1953). Mathematical Principles of Natural Phil-

osophy and His System of the World, pp. [15] Glazerbrook, . ., Dictionary of Applied 38-639 (University of California Press, Physics, p. 777. Berkeley, 1947). [16] The International System of Units (SI), [3] Rutherford, D. E., Classical Mechanics, NBS Special Publication 330.

Ch. 2, ( Interscience Publishers, Inc.,

New York, 1951 ). [17] Thiessen, M. , Schell, K . , Disselhorst, H., Wiss. Abh. Phys.-techn. Reichsanst 3, [4J Potius, P. E., Mass and Mass Values, 1-70 (1900). Nat. Bur. Stand. (U.S.), Monogr. 133, p. 12, (1974). [18] Chappuis, P., Trav. mem. BIPM 13, Dl- D40 (1907). [5] Tittman, 0. H, The National Prototypes

of the Standard Metre and Kilogramme; [19] Tilton, L. W., Taylor, J. K. , J. Res. Nat. U.S. Coast and Geodetic Survey, Appendix Bur. Stand. (U.S.) 18, 205 (1937) R.P. No. 18— Report for 1890. 971

[6] Cook, A. H. and Stone, N. W. §., Precise [20] Guillaume, C. E., Trav. mem. BIPM 14, Measurements of Mercury at 20 C- I, A1-A276 (1910). Phil. Trans. Roy. Soc. 250A (1957). [21] Chappuis, P., Trav. mem. BIPM 14, Bl- [7] Cook, A. H., Precise Measurement of the B163 (1910). Density of Mercury at 20 C- II, Phil. Trans. Roy. Soc. 254 (1061-2). [22] deLepinay, J. M., Buisson, H., Benoit, J.R., Trav. mem. BIPM 14, CI -CI 27 (1910).

[8] Bowman, H. A., Schoonover, R. M. , and 0 Carroll, C. L., A Density Scale Based [23] Friedman, I., Redfield, A. C, Schoen, B., on Solid Objects, J. Res. Nat. Bur. Harris, J., Rev. Geo. 2, 177-224 (1964).

Stand. (U.S.), 78A (Phys, and Chem. ) No. 1 pp 13-40 (Jan. -Feb. 1974). [24] Redfield, A. C, private communication (H.A. Bowman). [9] API Standard 2550 "Calibration of Upright Cylindrical Tanks", Amer. Pet. [25] Woh, M. and Urey, H. C, J. Chem. Phys., Inst. 2101 L Street, N.W., Washington, 3, 411-414 (1935). D.C. [26] Christiansen, Crabtree, Loby, Nature [10] Page, C. H. and Vigoureux, P., The In- 135, 870 (1935). ternational System of Units (SI), Nat. Bur. Stand. (U.S.), Spec. Publ. 330 [27] Wagenbreth, H. and Blanke, W., Official (Apr. 1972). Journal of the Physikal ische-Technische Bundesanstahlt, pp. 415-415 (June 1971). [11] Pontius, P. E. Notes on the Fundamen- tals of Measurement and Measurement as [28] Bowman, H. A. and Schoonover, R. M., J. a Production Process, Nat. Bur. Stand. Res. Nat. Bur. Stand. (U.S.), 69C (Eng. NBSIR 74-545 (1974). and Instr.), No. 3, pp. 217-223 (1965).

55 .

[29] Bowman, et al . , see [8]. [42] Emmet, R.T. and Miller, F.J., The Density of Oceanograph Standards, in

[30] Bowman, H. A. and Gallagher, W. H. , An preparation (1973). Improved High Precision Calibration Procedure for Reference Hydrometers, [43] Bowman, H.A., Schoonover, R.M., J. Res. Nat. Bur. Stand. C, Engrg. Cartesian Diver as a Density Com- and Instr., Vol. 73C, July-Dec. 1969. parator, J. Res. Nat. Bur. Stand. C, Engrg. and Instr., July-Sept. 1965. [31] Bulletin de L' Organization Interna- tionale de Metrologie Legale, No. 56, [44] Private communication, G. Girard, pp. 28-29 (Sept. 1974). BIPM, Sevres, France.

[32] Hartman, J., Dictionary of the United States Standardization Activities, Nat.

Bur. Stand. (U.S. ) Misc. Publ. 288 (1967).

[33] Cavanaugh, W. T., Measures of Success, ASTM Standardization News, Vol. 2, No. 3, p. 27 (March 1974).

[34] Landvater, J., private communication, Ohaus Scale Corp., Florham Park, NJ

[35] Johnson, A., Two Picrometers of Increased Convenience and Precision, J. Res. Nat. Bur. Stand. C, Engrg. and Instr., Vol. 69C, No. 1, Jan.- March 1965.

[36] Pontius, P. E., Measurement Philosophy of the Pilot Program for Mass Calibra- tion, NBS Tech. Note 288, May 6, 1966.

[37] Corwin, A. H., Weighing: Chapter III Techniques of Organic Chemistry,

Vol. I, Part I (Physical Methods of Organic Chemistry), Interscience Publishers, Inc., New York, 1959.

[38] Macurdy, L. B., Measurement of Mass, Treatise on Analytical Chemistry, Edited by Kolthoff, Elving and Sandell, Part I, Vol. 7, Inter- science Publishers, New York, 1967.

[39] Smithsonian Physical Tables, Smith- sonian Miscellaneous Collections, Volume 120, The Smithsonian Institu- tion, Washington, DC, 1954.

[40] Pontius, P. E., Mass Measurement: A Study of Anomalies, Science, Vol. 190, pp. 379-380, 24 Oct. 1975.

[41] Menache, M. and Girard, G., Concerning the Different Tables of the Thermal Ex- pansion of Water Betweem 0° and 40°C, published jointly by the Institute Oceanographique and the B.I. P.M. (1972).

56 APPENDIX A. METHODOLOGY OF THE STUDY 1. EXPLAIN THE NATURE OF YOUR BUSINESS IN TERMS OF SERVICES AND PRODUCTS RENDERED. The material in this study was based on a PLEASE LIST AS MANY USES AS YOU ARE AWARE number of published and unpublished docu- OF FOR MASS DETERMINATIONS. BE SPECIFIC. ments. Extensive use was made of the files of the NIBS Mass and Volume Section. Material 2. HOW MANY DOLLARS AND/OR MAN YEARS DOES on the program of the NBS Office of Weights YOUR COMPANY EXPEND ANNUALLY IN SOME and Measures and its relations with the FACET OF MASS MEASUREMENT? IS THERE SOME National Conference on Weights and Measures COST-EFFECTIVE FIGURE THAT HAS BEEN COM- was obtained from the annual reports of the PUTED IN TERMS OF THE INVESTMENT IN MASS latter body. Additional information was DETERMINATIONS VERSUS PRODUCTIVITY? obtained from the staff of the NBS Office of Weights and Measures. With respect to 3. IF YOU ARE A BALANCE AND/OR WEIGHT MANU- the International Organizations, use was FACTURER, PLEASE GIVE THE NUMBER AND made of material prepared for the U.S. SALES OF BALANCES ANNUALLY AND ALSO THE Advisory Committee on the International NUMBER AND SALES OF WEIGHTS. Organization for Legal Metrology by the NBS Engineering and Products Standards Division. 4. WHO ARE YOUR MAJOR CUSTOMERS AND THEIR Data on NBS customers was developed from the USERS (INTERNAL AND EXTERNAL) WITH files of the Office of Measurement Services. RESPECT TO MASS DETERMINATIONS? PLEASE A survey was run of selected customers, SPECIFY IN DETAIL THE ACCURACY AND PRE- using economic data from Standard and Poor's CISION YOUR USERS NOW REQUIRE AT EACH Index of Corporations, Dunn & Bradstreet, STEP IN THE TRANSFER. IS THERE A FORE- and Moody's Handbook of Common Stocks. For SEEABLE NEED FOR GREATER ACCURACY AND the purposes of our study, we have sought to PRECISION? IN OTHER WORDS, ARE YOUR obtain trends in the interpretation of our CUSTOMERS REALLY USING ALL THE ACCURACY customers' Standard Industrial Classifica- YOU NOW SUPPLY THEM? IF NOT, WHY NOT? tion codes instead of attaching significant (WE NEED AN EXHAUSTIVE ANSWER TO ALL OF weight to the individual numbers since many NO. 4). of our customers are conglomerates. A canvass of industry was made to estab- 5. WHAT TYPES OF INSTRUMENTATION FOR MASS lish a preliminary data base to confirm a DETERMINATIONS ARE USED IN YOUR few of our basic ideas concerning the struc- FACILITY? HOW DO YOU PROCEED IN THE ture of the mass measurement systems. The CALIBRATION OF THESE DEVICES? canvass consisted of two parts: an informal question letter and/or telephone interview 6. TO WHAT EXTENT DO REGULATORY AGENCIES with responsible individuals within an or- (LOCAL, STATE, FEDERAL) GOVERN THE PRO- ganization. In all cases no formal ques- DUCTION AND/OR SALES OF BALANCES AND tionnaire was sent "blindly" into an organi- WEIGHTS? zation. A high-ranking contact within a company was first established within our 7. DO YOU KNOW OF ANY PROBLEMS IN MANUFAC- area of interest; a preliminary telephone TURING PROCESSES WHERE AN IMPROVEMENT interview was conducted; and if the need IN THE ACCURACY OR PRECISION OR METHOD existed, a letter outlining other detailed OF MASS MEASUREMENT MIGHT HELP RAISE questions was subsequently sent to the in- PRODUCTIVITY? terviewee for his comments. An effort was made to look objectively at our NBS Customer 8. WHAT TYPE OF PERSONNEL DO YOU HAVE IN- Profile and select the SIC categories which VOLVED IN MASS MEASUREMENTS? IS THERE could be interpreted to imply "first-order ANY TYPE OF ADDITIONAL TRAINING YOU measurement-intensive" industries. By WOULD LIKE THESE PERSONNEL TO RECEIVE? selecting companies from our customer list IN WHAT AREAS? which possessed the majority of these codes within their products or services rendered, 9. DO YOU FEEL THAT IF YOU HAD TO "PROVE" we were able to develop a systematic THE ACCURACY OF YOUR MEASUREMENTS IN A "sample" of the users. The canvass COURT OF LAW, YOU WOULD HAVE SUFFICIENT questionnaire is reproduced below: EVIDENCE TO CONVINCE THE JURY? WHAT TYPE OF EVIDENCE WOULD YOU PRESENT FOR PLEASE COMMENT ON AS MANY QUESTIONS AS POS- YOUR DEFENSE? SIBLE. USE ADDITIONAL SHEETS OF PAPER TO PRESENT YOUR VIEWS. ALL ANSWERS WILL BE 10. IF YOU USED THE SERVICES OF NBS, PLEASE STRICTLY CONFIDENTIAL. COMPLETION ON OR STATE IN WHAT WAY. IF YOU ARE INVOLVED BEFORE NOVEMBER 5 WOULD BE APPRECIATED. IN THE MASS MEASUREMENT ASSURANCE PRO- GRAM (MMAP) PLEASE SPECIFY IN DETAIL .

YOUR LIKES AND DISLIKES. HOW CAN ANY, OR applying corrections to micro work ALL, OF THESE SERVICES BE IMPROVED? necessary. Production requirements are constantly changing and are generally 11. WHAT TYPE OF TRAINING SEMINARS AND/OR more difficult to meet each year. PUBLICATIONS WOULD YOU LIKE TO SEE EMA- Improvements are needed in weighing NATE FROM NBS? IN WHAT AREAS OF INTER- equipment available from the manufac- EST? TOWARD WHAT LEVEL OF PERSONNEL turer and not from the laboratory or SHOULD THESE BE GEARED? NBS calibrating organization. In most cases, the customer is using the accuracy supplied. A typical response to the above question- naire is this: " 5. Our laboratory has over 600 mass mea- suring instruments, representing most " 1. Our laboratory is under contract with manufacturers, ranging from platform the Atomic Energy Commission. Nature scales to quartz fibre microbalances. of work is classified and relative to Most instruments are commercially the National Weapons Defense Complex. acquired, although some balances are Mass determinations are used in: customized or adapted to our particu- 1. Research and development of lar needs. Majority of equipment is explosive and nuclear items. substitution type analytical balances. 2. Mass Standards Laboratory - Others include: Calibration of weights and Torsion Type Prescription balances. Top Loaders 3. Production of miniature de- Two Pan Free Swinging tonators and associated Two Pan Electronic devices. Torsion Roller Smith Type Quartz Fibre " 2. Seven man years are expended annually for calibration and maintenance of " Procedures for calibrating vary according balances and weights. In addition, to balance design and intended use. an undetermined number of scientists, Internal weights are calibrated for technicians and workers use balances bias. Precision is determined by in their daily work assignments. repeated weighings at designated loads. " 3. Does not apply. " Standard balances are calibrated via the " 4. Users of mass measurement equipment National Bureau of Standards Mass cover a wide spectrum as stated in Measurement Assurance Program. Question No. 1. Accuracy and pre- cision required vary according to " 6. Local, State or Federal agencies do use and range from being able to use not govern the production and/or sales a substitution balance without of balances and weights. There is no where + .010 milligrams is required at noticeable effect on weighing equip- the first transfer level (weight ment from these agencies. calibration) " 7. None " For example: Products being developed in the R&D areas may require a + .OlOmg " 8. Personnel involved in mass measure- uncertainty. Calibration of the mea- ments are hired as unskilled and suring equipment requires a minimum untrained. Our Mass Standards Lab- uncertainty in the mass standards, as oratory provides a training class to the balance used in the areas consume provide the necessary training to most of the allowable error. Mini- operate the equipment proficiently. ature detonators require weighings No additional training is required. less than 50mg to an uncertainty of + .2mg. Check standards used with " 9. Sufficient evidence is maintained to this production equipment must be held prove the stated uncertainties in our to less than + .015mg if product is to Standards Laboratory mass measurement be within the accepted tolerance. process. Statistical data and control charts furnished by the National " Accuracy and precision requirements in the Bureau of Standards Mass Measurement Mass Standards Laboratory are ade- Assurance Program are maintained for quate, but can be reduced further if this purpose. Similar information is 58 "

available for production balances when Mr. David M. Jeffers weighing to tolerance approaches the Administrative Technologist capability of the equipment used. Department of Pathology York Hospital "10. Our laboratory is an active partici- York, Pennsylvania 17405 pant in the National Bureau of Stan- dard's Mass Measurement Assurance Mr. Robert Schoeneman Program. There is very little criti- Internal Revenue Service cism of the program; the calibration Alcohol, Tobacco and Firearms Division data received has been extremely bene- Washington, D.C. 20226 ficial to our organization and the assistance and cooperation from NBS Mr. Charles Muirhead personnel is above our expectations. National Ocean Survey The turn-around time on some jobs is Oceanographic Division less than desirable at times, and 6001 Executive Blvd. more frequent issue of up-to-date Rock vi lie, Maryland 20852 control charts would be advantageous. Dr. James D. MacLowry

"11. I would personally like to see more Clinical Pathology Department detailed procedures written in Clinical Center "users'" language on calibration pro- National Institutes of Health cedures relative to mass measurements, Bethesda, Maryland 20014 other than weight calibrations. For example, procedures for calibrating small glass volumes and density mea- surements. Bowman and Schoonover's paper "Procedure for High Precision Determinations by Hydrostatic Weigh- ing" is an excellent document, but unsuitable for routine density mea- surements. A less complicated proven procedure is desirable in many laboratories.

We have not limited our contacts to the industrial community; indeed, we have estab- lished contacts within other governmental agencies, hospital and laboratory personnel, and also medical societies. Specific contacts include:

Mr. Charles Stockman Maryland Office of Weights and Measures Department of Agriculture College Park, Maryland

Mrs. Martha Winstead Quality Control Coordinator Southbend Medical Foundation Southbend, Indiana (219)234-4176

American Society for Medical Technology Ste. 200 5555 West Loop South Houston, Texas 77401 (713) 664-8121

59 "

APPENDIX B measurement started in this country with Hassl er. The past of the Mass and Volume Section Hassler was a competent metrologist, com- is inseparably intertwined with the history parable with the builders of the kilogram in of Weights and Measures and NBS. As one re- France, and the reconstructors of the pound views this history to identify strengths and in England. Fischer brought to the newly weaknesses, one cannot help but note the formed NBS the disciplines of Hassler, as he progress of people through the organization had learned them from those who had been in- over the years. As people move through the structed by Hassler himself [3]. He brought organization, so move philosophies and the with him the weighing designs of Hayford which way things are done, and to some extent, the predate those of Benoit of BIPM [4,5]. way others view NBS. One can see, through Hayford's work has never been given the at- the anecdotes of the past, only a generation tention it deserves, perhaps because of the removed, and through reviews of the histori- prestigious position of Benoit, and the fact cal records, a pattern which not only re- that Benoit' s presentation was somewhat more flects the development of the section, but general. Fischer, together with his knowledge also of the NBS as a whole. Historically, and the standards from the Coast Survey, be- it doesn't really matter why certain things came a part of the Weights and Measures "Sec- were done; it is certain that the reasons tion" of Division 1, under the direct super- appeared to be valid to those faced with the vision of Stratton. decisions which had to be made, but what was Among the earliest circulars of announce- done is important to an understanding of why ments of the services that the new NBS could things are the way they are today. For the provide were those in the area of mass mea- purpose of this discussion, the history of surement and in the testing of glass volume- the section, and NBS, has been arbitrarily tric apparatus. By 1910, Fischer was Chief divided into "eras," which, in addition, of the Weights and Measures Division, with have been arbitrarily named "Getting 3 sections which were the genesis of the Started," "Confusion," "Transition," and "A present Mass and Volume Section, and 2 sec- New Start. tions which would eventually become the Of- fice of Weights and Measures. Pienkowsky "Getting Started" was the Chief of the Mass Section through 1940. Circular 3, Design and Test of Stan- The basic tools of the trade are not new, dards of Mass, (Rev. 3), prepared by Fischer the balance originating some time in the and Pienkowsky, is a classic in the area of

Bronze Age, and with "traceabi 1 i ty " of the mass measurement. unit over at least 4500 years going back to Circular 3 treated both the problem of Nebuchadnezzar. The mathematical tools also design and adjustment of weights, and the originated far back in time, with averaging scientific use of mass measurements. For in the days of the Romans . Weighing by sub- commercial use, three sets of adjustment stitution, normally called "by the method of tolerances were established: Class A, for Borda," was used by Amiot in 1776. Transposi- state standards; Class B, an intermediate tion weighing, "the method of ," and class which saw little use; and Class C, for "least squares," are attributed to Gauss in commercial weights. This tolerance struc- the early 1800' s. It is doubtful, however, ture evolved from an end use requirement, that Gauss invented transposition weighing that of weighing large quantities of mater- since it is the basic method of testing an ials with multiple lever scales. The prob- equal -arm balance. He may, however, have been lem was to establish a commercial tolerance the first to treat the method in a mathemati- for "tip" weights which, when multiplied by cal analysis. Weighing designs started with the lever ratio of the available scales, Tartaglia in 1556, and Bachet in 1612, with later contributions by Hassler in 1817 and Broch in 1886. The science (or art) of mass The first five circulars are:

NBS Circular 1 The National Bureau of Standards 1 This, and the balance of this paragraph is 2 Measurements of Length and an extremely brief summary of the history Area Including Thermal Expan- presented by C. Eisenhart in references sion [1] and [2]. 3 Design and Test of Mass Standards 4 Verification of Standards of Capaci ty 5 Test of Clinical Thermometers 60 "

would not introduce an unreasonable system- operated out of Washington. For remote atic error in the weight of the goods. Hav- scales, such as mine tipple scales, a large ing arrived at these limits, Class B and number of commercial 50 lb weights had to be Class A tolerance limits were established "manhandled" to the scale location. With such that compliance with Class C tolerances practice, one could pick up two in each could be established with a minimum of mea- hand, an amazing feat which no doubt supple- surement effort. mented the paycheck through side bets with A story relating to the accomplishment of the miners. the above task has been passed down from With the organization of the National those who were present at the time. Fischer Conference of Weights and Measures in 1912, was in the habit of keeping notes on his additional sections were added to the Me- shirt cuff. The task of establishing the trology Division to support that organiza- tolerance structure had been under discus- tion. This action initiated a rift which, sion for some time, and at such a noontime up to now, is only partially repaired. Pi- discussion, Fischer worked out the toler- enkowsky, in charge of mass measurements, ances on his shirt cuff. He had them typed was concerned with the scientific aspects of immediately, but neglected to record the mass measurement up to 50 pounds, yet had to logic behind their origin. The shirt was interface with both the Scale Unit and the laundered and the detailed logic upon which legal aspects of weighing. The Scale Unit, the structure was based was gone forever. because of the handling difficulties with Pienkowsky was a tough taskmaster. Two large weights, was primarily engineering, types of calibrations were offered, one re- and, to some extent, art oriented. The sulting in a "test report," and the other in Scale Unit was concerned with the correct- a "certificate." The issuance of a "test ness of the result relative to a particular report" usually meant that something was in- usage. The Weights and Measures Unit was ferior, that the item did not warrant the concerned primarily with the legal aspects best effort. Years of experience were re- of metrology, e.g., right by definition. quired before one was qualified to work with The Mass Unit insisted that the balance was items which warranted a "certificate." a mass measurement device. The Scale Unit Weighing designs were used in both cases, was more oriented in the practical usage of and the quality of the measurement was es- large weights; the Weights and Measures Unit tablished by the use of "check equations," was only interested in establishing systems e.g., if A-B=X B-C=Xp and A-C=X then, if which could not be attacked legally. Two 1 , 3 , all is in order (X1+X2T-X3, the "check equa- anecdotes clearly illustrate the rift. tion" should be within some established limit. While the Weights and Measures Unit was While Pienkowsky was the only one who knew the stressing the importance of standards, one basis for the established limits, everyone in member of the Scale Unit would on occasion, the group knew what they were and strived and much to the dismay of the Weights and mightily to have their work be acceptable. Measures Unit, calibrate a standard with a In this era the thermal problems of the collection of flat irons, frying pans, stove equal arm balance were not as yet under- lids and the like, all of which had been stood. The best work was done in isolation. carefully calibrated in advance. State Computations were done by hand with the use weights and measures people were taught to of slide rules and logarithmic tables. The (1) pick up the weight in both hands (so as work was checked and rechecked. Work on not to drop it), (2) approach the balance large capacity scales, and the testing of with care (so as not to fall), (3) bow railway track scales started in about the slightly (so as to clearly identify the middle of the "era." A large effort was de- weight before it was placed on the balance voted to the testing of all kinds of large pan), (4) place the weight gently on the scales in addition to the railway track balance pan, and do likewise for the second scales. With the track scale test cars weight. Without the reasons which, of operating out of a field station on the out- course, were a matter of routine within the skirts of Chicago, a truck equipped with Mass and Scale units this was interpreted as large weights and weight handling equipment (1) approach the balance with awe, (2) genu- flect, (3) put the weight on the pan, and do it all over again. Needless to say, there were very strong feelings between the various sections. 1 Each operator was assigned a specific The close of the "era" was largely a mat- colored pencil, thus from the array of ter of coasting on reputation, or position. colored check marks on the computation Much time was devoted to detail of compli- sheet, one could identify the various ance with specifications, and to the exten- "checkers. sion of the tolerance limits to the variety

61 of weights in common use and some of which done on the drafts of the documents to were not so common, the onza and adarme. follow 547, but none were completed and The program was in essence a pub! i shed. semi regul atory action e.g., differ- Simplified procedures were published in entiating what would receive a "test report" the area of volume and density measurements. and what would receive a "certificate" in- A method for testing the new quick weighing creased. Aside from Circular 3, little was balance was published. Under the urging of done to advance the understanding of mass the Office of Weights and Measures, a 1000 measurement, and to provide a direct service lb equal arm balance was developed [6]. to the user. The Director signed each re- By and large other publications of the "era" port, which in turn was embossed with the followed a rather fixed pattern: general official Bureau seal. Since weights could classes of equipment were described together not be "marked" without destroying the cali- with brief outlines of test procedures used bration, the "mark" was engraved on the by NBS; some form of tolerance structure was case, which in turn was wrapped, tied with presented; detailed discussion of what was red ribbon and "sealed" with sealing wax and necessary to receive a "certificate," and an appropriate signet. what would get only a "test report;" and finally a description of the mechanism " Confusion" whereby one submitted items to NBS for test. Toward the end of the "era," the Mass and Reorganization early in the "era" changed Scale Section was once again returned to the the Weights and Measures Division to the Me- Metrology Division. At approximately the trology Division and established the Office same time, the Office of Basic Instrumenta- of Weights and Measures under the direct tion was established under the supervision supervision of the Director. This action es- of the Director and the position of Assistant sentially severed the Office of Weights and Director for Testing was established. Measures from the support of the technical With regard to the section, in spite of in- divisions, including editorial review. ternal conflicts between the Mass Unit and Authority to sign "Test Reports" and "Certi- the Scale Unit, this marked the start of ficates" was delegated to the section. The real progress, particularly with respect to ritualistic procedures were, however, main- the studies of the balance. The influence tained. of the Statistical Engineering Laboratory, About midway in the "era," the Mass Sec- established in the late 1940's was beginning tion and the Scale Section were combined and to be evident in the calibration programs. transferred to the Mechanics Division. This Higher management decisions, however, en- placed the new section in the area of ap- couraged dissemination through equipment plied science. Circular 547, the first of a manufactures and networks of calibration planned series, updated the design and tol- laboratories. The success of such an action erance structure established in Circular 3. presumes that the second level laboratories The philosophy of Circular 547 was oriented assume the role of interfacing with the more toward "what they should have" rather user. As it turned out, the second level than toward "what do they need." Manufac- laboratories elected to become "carbon cop- turers of weights were consulted and the ies" of NBS. Somewhere along the line, a tolerance structure was based on their es- "hierarchy" became firmly entrenched, with timates of the ability to adjust weights. mission oriented facilities located several These estimates turned out to be optimistic steps removed from NBS, and with no direct and, as a result, there were many rejections communication. for "out of tolerance." The situation was With respect to manufacturers, with tacit eventually resolved by revising the specific- NBS approval, calibration "certificates" and ations. the like were a part of doing business with This presentation e.g. is another example the "hierarchy," with NBS frequently being a of numerous attempts of NBS personnel to de facto extension of the manufacturer's achieve a semi regul atory action, first the quality control organization. It was a com- preparation of the specification, then the mon practice for manufacturers and distribu- rejection for noncompliance. Much work was tors to stock NBS calibrated items, with an add on "handling" charge far in excess of the NBS fee for the work. The practice of doing calibration work for manufacturers and distributors was encouraged by the use of "number of items calibrated" as a measure of 1 The onza is the ounce of Spain and Spanish the effectiveness of the section. America. The adarme is 1/16 of an ounce. By the end of the "era," several basic papers had been published with regard to de- 62 "

tailed studies of balance characteristics Director for Testing, however, was slow in [7, 8, 9, 10, 11]. The development of a pre- recognizing this need, with the result that cision kilogram balance had been started. frequently NBS was presenting two "faces" to Methods for calibrating weights in the the users. With the press of cost-effec- direct reading balances were developed and tiveness measures, by the end of the era it published. The two-knife constant load 50 was recognized in part within NCSL that the lb balance was at the state of evaluation of role of the calibration laboratory was the development model [12]. The program for 1 imi ted. new state standards had been prepared. The Through this era, the Mass and Volume "era" closed with the introduction of "trace- Section completely revamped both the program ability" through U. S. Air Force Bulletin No. structure of the section, and the services 520. The introduction of the concept without provided. Each action, starting with ob- definition created an endless scene of con- taining a commitment from the Statistical fusion. Engineering Laboratory to support the evalu- ation of the measurement processes, was the " Transition" result of micro-issue and micro-measurement system studies completed prior to starting The "transition" era included both the the action. The capabilities of the section change from the monolithic structure of NBS were augmented with the increasing use of to the present Institute structure, and the computers, both time sharing and otherwise. move from Washington to Gai thersburg. Hall- The section staff, relieved from the tedious marks of the era were the technological ex- detail of checking and rechecking, could plosion of the space age and the start of finally begin to think about what they were an extensive re-evaluation of the role of really doing. NBS. The "up through the ranks" method of The first computer prepared report sub- filling vacancies was wanning, with new faces stantiating a measurement process operating and new ideas appearing at all levels of in a state of statistical control was issued management. The National Conference of in late 1963. A large educational effort Standards Laboratories was formed, thus uni- in mass measurement techniques, and in the ting those who were faced with carrying out characterization of measurement processes, the quality control aspects of government was a part of the work with weights for procurement in the area of space exploration force machines. Studies were made concern- and defense. MIL-C-45662A, Calibration ing the performance characteristics of mass System Requirements, provided in part the measurement processes in other facilities. missing definition of "traceabi 1 i ty. In late 1965, the Director of IBS gave verb- Eisenhart's classic paper, "Realistic Eval- al approval to proceed with the Mass Mea- uation of Precision and Accuracy," was pre- surement Assurance Program (called at the sented in 1962 [13]. In 1963, by decree, time the Pilot Program in Mass Measurement). uncertainty statements were required on all By late 1967, the program was operational, reports. Later that same year "certi- and tests were underway to evaluate, for ficates" were abolished, a positive move the first time in a controlled manner, the away from the semi regul atory action associ- measurement capabilities of one segment of ated with the previous "report" and "certi- the national measurement system, that asso- ficate." Measurement results were reported ciated with mass measurement. on a "Report of Calibration." Where a Throughout this era there was a steady "certificate" was mandatory, the institute flow of papers, both tutorial and technical. Director would "certify" only that the Training conferences were held at NBS, and measurements were made. the section staff was invited to describe Early in the "era" there was a growing the programs at various meetings of profes- divergence of opinion concerning the inter- sional societies. By virtue of a knowledge facing of NBS with the users of the calibra- of the measurement process parameters, the tion services. At the section level, it was "lost time" on the move to Gai thersburg was recognized that various end use requirements less than one week. One particularly dif- could not afford the degradation of the ficult task associated with the move con- units through a hierarchy of calibration cerned the facilities which had been designed laboratories, and that there was a real need for the dissemination of measurement proce- dures. Most echelon laboratories were more concerned with emulating NBS than with 1 With the introduction of computer prepared being centers of measurement excellence. reports, many of the rituals, including Lower level laboratories were faced with tying packages with red ribbon, sealing real measurement problems. The Assistant with wax and an official seal, were di scarded.

63 and constructed on the basis of the Early in the era, the section accepted anticipated role of the sections in the task of the application of measurement the mid 1950' s and which, by now, had con- process analysis techniques to the problem siderably changed. of reprocessing spent reactor fuel elements [14, 15]. This initial work was supported " " A New Start by other agency funding. A continuing activity over the era has been in consulta- This is the current era, and from the tion with appropriate committees working on standpoint of calibration services, it is specifications and guidelines for both mass the era of the Measurement Assurance Pro- and volume measurements in this field. The grams. In the early days of this era, the first specifications are now in the ballot- Director of IBS established the Office of ing stage, eventually to become ANSI Measurement Services as an operating Divi- standards for the industry [16, 17]. sion, and for the first time, NBS was able In early 1972, a precision kilogram bal- to present one "face" to the users of the ance, NBS-2, was shipped to BIPM [18]. As a calibration services. Efforts to unify the part of the transfer, BIPM personnel spent activities within the Metrology Division by approximately 1 month at NBS learning about the establishment of branches were not suc- our procedures, methods of data analysis, cessful. In early 1970, the Metrology Divi- and the concepts of characterizing the per- sion was disbanded, and the activities were formance of a measurement process. Over the combined with those of the past few years this balance has been used to Division to form a new division, the Optical evaluate the stability of Pt-Ir kilograms in Physics Division. Immediately following, preparation for planned intercomparisons of the Mass and Volume Section absorbed the prototype kilograms [19]. In 1973, a new Length Section and the Engineering Metrology version of the report of mass calibration Section, becoming the Mass, Length and Vol- was introduced, a laboratory notebook type ume Section. of report with a complete documentation of From the formulation of the new section the measurements, and the status of the mea- until mid 1971, the activities were largely surement process at the time the measurements directed toward cleaning up a large backlog were made. Much interest in these procedures in the engineering metrology area, and in has been shown by both Italy and Sweden. the establishment of the Measurement Assur- In late 1974 the gage block measurement ance Program in the area of gage block mea- processes were operating in a state of sta- surements. In mid 1971, the Dimensional tistical control, with new reporting proce- Technology Section was formed, assuming res- dures, and the development of detailed MAP ponsibility for both engineering metrology procedures were under discussion. Bowman et and length measurements (with the exception al had completed the determination of of tapes). The responsibility for the com- pure single crystal silicon, the basis for a pletion of the MAP program in gage blocks, solid density standard. With the publica- however, remained as a responsibility of the tion of procedures for the calibration of Chief of the Mass and Volume Section, acting volumetric glassware, all of the procedures as a Special Assistant to the Division Chief now in use at NBS are in the literature for Measurement Assurance Programs. [20]. The marriage of one of the most conserva- These brief remarks indicate the history tive areas of NBS with one of the most high- of the section and the manner in which it ly specialized scientific areas was to be an has been operated over the years. The iso- experiment. It is perhaps too early to lation with respect to the Office of Weights evaluate the merits of the action. From the and Measures remains (primarily at the man- metrologist's viewpoint, it certainly pro- agement level since the key working techni- vided a wealth of resources, however, a cal staff of OWM are former members of the great effort had to be made in order to for- Mass and Volume Section). mulate the myriad of problems in such a man- The important contributions, other than ner that (1) the problem is of interest to the direct support to certain specific areas the scientist, and (2) that the results from of science and that obviously associated the scientist can be utilized in a practical with providing access to the mass unit, have sort of way. The measurement expertise of been in the contributions to the understand- the section has assisted in certain areas, ing of the mass measurement process, and in e.g., the redetermination of Avogadro's num- the applications of measurement assurance ber, but, with few exceptions, there was no procedures. The problems facing the section real interest in metrology to the depth at this time are (1) those associated with necessary for assistance in problem the activities of the 0IML, which are regula- formulation and solution. tory in nature and obviously in direct con- flict with the directions of the section activity over the past 15 years; (2) the problems associated with the anticipated recal ibration of the prototype kilograms by BIPM, scheduled tentatively to take place within the next several years; and (3) the extension of the mass unit to practical mass measurement scales. The present environment of NBS reflects several distinctly different philosophies. Perhaps the largest group is the highly spe- cialized who, after a cursory look at the role of measurement, has decided that the "way to go" is to organize, or legalize, measurement. The concepts of "certified

standards," and "traceabi 1 i ty" are the "ob- vious solution to all measurement problems" and with the problem taken care of in this manner, one can go back to the more inter- esting things. This philosophy was strongly advocated by the top administration up to the middle of the "transition era," at about which time it reached its peak. When car- ried to the extreme, such a policy becomes self-serving, that is, having created the organization, one exists to support that or- ganization. In the meantime, there is little, if any, direct support to the user. The opposing philosophy, originating in about the middle of the "transition era" concerns the adequacy of the measurements which have to be made by the user. Measure- ment assurance is merely doing what has to be done in order to be sure that the mea- surement is adequate for the intended use. To support this, one must look for the pro- blems, then provide the most economical solutions. Of course there is a need for good standards, Standard Reference Materials and the like, but they are means to an end, not the end itself. Success is measured by the status of the measurement system itself, and not by the number of "certified stan- dards" provided by NBS. To carry out such a program requires a willingness to learn, and a substantial commitment of time on the part of both NBS and the user. Needless to say, such programs are strongly supported from that segment of the user community which has real measurement problem but not in a coordinated way. It is unfortunate that coordinated op- position comes from those who see NBS ac- tivities as a simple means to meet contractual requirements.

65 [11] Macurdy, L. B. and Alrner, H. E., J. Res APPENDIX B Nat. Bur. Stand. (U.S.), 68C (Engr. and REFERENCES Instr.), No. 3, pp. 135-140 (1964).

[1] C. Eisenhart, "Some Antecedents of Modern [12] Aimer, H. E . , Bowman, H. A., Jensen, Experiment Design", Invited address be- M. W., Macurdy, L. B., Peiser, H. S., fore the Section of Statistics, Thirty- and Wasko, B., "A Direct-Reading Two- Third Annual Meeting of the Virginia Knife 50-1 b Balance of High Precision Academy of Science, Madison College, Suitable for State Weight and Measures Harrisonburg, Virginia, May 13, 1955. Laboratories", J. Res. Mat. Bur. Stand. (U.S.), 68C (Engr. and Instr.), No. 3, [2] C. Eisenhart, "The Background and pp. 141-154 (1964). Evolution of the Method of Least Squares" 34th Session of the International Statis- [13] Eisenhart, C, "Realistic Evaluation of tical Institute, Ottawa, Canada, 21-29 the Precision and Accuracy of Instru- August 1963. ment Calibration Systems, J. Res. Nat. Bur. Stand. (U.S.), 67C (Engr. and In- [3] S. W. Stratton, "The Bureau of Standards str), No. 2, pp. 161-187 (1963). and its Relation to the United States Coast and Geodetic Survey", Centennial [14] Pontius, P. E., "Metrology of Nuclear Celebration of the United States Coast Materials", Nat. Bur. Stand. (U.S.) Re- and Geodetic Survey, pp 33-34, Govern- port 9994 (Feb. 1969). ment Printing Office, Washington 1916. [15] Pontius, P. E., "On the Calibration of [4] J. F. Hayford, "On the Least Square Ad- Large Tanks", Nat. Bur. Stand. (U.S.) justment of Weighings", United States Report 9993 (Feb. 1969). Coast and Geodetic Survey, Report from 1892, Appendix No. 10, Government [16] American National Standard N15.18, Mass Printing Office, Washington 1893. Calibration Techniques for Nuclear Materials Control. [5] J. P. Benoit, "Standardization of Sets of Weights", Annalos des Poids et [17] American National Standard N15.19, Vol- Measures, Vol. 13, 1907. ume Calibration Techniques for Nuclear Materials Control. [6] H. H. Russell, "The National Bureau of Standards 1000 lb Balance", Scale [18] Aimer, H. E., "National Bureau of Stan- Journal, pp 6-7 (Aug. 1955). dards Kilogram Balance", NBS No. 2, J. Res. Nat. Bur. Stand. (U.S.), 76C (Engr [7] H. E. Aimer, "Response of Microchemical and Instr.), pp. 1-10 (1972). Balances to Changes in Relative Hum- idity, J. Res. Nat. Bur. Stand. (U.S.), [19] Girard, Georges, Private Communication, 64C (Engr. and Instr.), No. 4, pp 281- BIPM. 285 (1960). [20] Lembeck, J., "The Calibration of Small [8] Bowman, H. A., and Macurdy, L. B., "A Volumetric Laboratory Glassware, Nat. Photoelectric Follow-up and Recording Bur. Stand. (U.S.), NBSIR 74-467 (1974) System and its Application to Remote Observations of the Beam in High- Precision Balances", J. Res. Nat. Bu. Stand. (U.S.) 63C (Engr. and Instr.), No. 2, pp. 91-96 (1959).

[9] Bowman, H. A., and Macurdy, L. B., "Gimbal Device to Minimize the Effects of Off- Center Loading on Balance Pans", J. Res. Nat. Bu. Stand. (U.S.) 64C (Engr. and Instr.), No. 4, pp 277-279 (1960).

[10] Bowman, H. A. , and Aimer, H. E., "Mini- mization of the Arrestment Error in One- Pan, Two-Knife Balance System", J. Res. Nat. Bu. Stand. (U.S.), 67C (Engr. and Instr.), No. 3, pp 227-235 (1963). 66 NBS-114A rev. n-77)

U.S. DEPT. OF COMM. 1. PUBLICATION OR REPORT NO. 2. Gov't Accession 3. Recipient's Accession No. BIBLIOGRAPHIC DATA No. SHEET NBSIR 74-928

4. TITLE AND SUBTITLE 5. Publication Date The National Measurement System for Mass, Volume and Density M-.., | Q~7Q

6. Performing Organization Code

S. Performing Organ. Report No. Paul E. Pontius, James R. Whetstone, and John A. Simpson 9. PERFORMING ORGANIZATION NAME AND ADDRESS 10. Project/Task/Work Unit No.

T70fl i r\r\ NATIONAL BUREAU OF STANDARDS DEPARTMENT OF COMMERCE 11. Contract .'Grant No. WASHINGTON, D.C. 20234

12. Sponsoring Organization Name and Complete Address (Street, City, State, ZIP) 13. Type of Report & Period Covered

14. Sponsoring Agency Code

15. SUPPLEMENTARY NOTES

16. ABSTRACT (A 200-word or less factual summary of most significant information. If document includes a significant bibliography or literature survey, mention it here.)

This document, one of several similar documents resulting from the NBS study of the National Measurement System, reflects the results of intensive studies carried out from 1972-1975 relative to Mass, Volume and Density measurements in the United States. The history and the current status of these measurements is discussed as well as the relative role of NBS. A bibliography is included.

17. KEY TORDS (six to twelve entries; alphabetical order; capitalize only the first letter of the first key word unless a proper name; separated by semicolons)

Density measurements; mass measurements; national measurement system; volume measurements

18. AVAILABILITY \X2 Unlimited 19. SECURITY CLASS [21. NO. OF PAGET (THIS REPORT)

1 1 For Official Distribution. Do Not Release to NTIS 71 UNCL ASSIFIED

1 1 Order From Sup. of Doc, U.S. Government Printing Office 20. SECURITY CLASS 22. Price Washinetoo. D.C. 20402. SD Stock No. SNO03-O03 (THIS PAGET

IX ' Order From National Technical Information Serrice (NTIS) $5.25 Springfield, Virginia 22151 UNCLASSIFIED U3COMM-DC 6S035-P76