the National Bureau of Standards Lewis V. Judson The methods used at the National Bureau of Standards in calibrating line standards of length and measuring tapes submitted for standardization are outlined. The equipment used is described briefly. There is a discussion of some considerations that should be given as to whether or not a standard should be submitted to the Bureau. Instructions are given for submitting items to the Bureau for calibration. The appendix contains useful information on the use of steel tapes. 1. Introduction The Bureau receives many requests from entists, industry, and governmental agencies to librate line standard of length and measuring pes. The Circular is issued as a guide for anycontemplating the submission of such items the Bureau for standardization. It replaces relar 332, Testing of line standards of length, Circular 328, Testing of measuring tapes at National Bureau of Standards. Line Standards of Length at NBS The primary standard of length in the United ates is the National Prototype Meter 27 (fig. 1), ich is identical in form and material with the ternational Prototype Meter deposited at the temational Bureau of Weights and Measures at res, near Paris, and also with the other national ototype meters distributed in 1889 in accordance ith the treaty known as the Convention of the eter, dated May 20, 1875. In the United States, ace 1893, the yard has been defined in terms of e meter by the relation Proposals have been made to adopt, as the definition of the meter, a specified number of wavelengths of a suitable isotope of an element such as mercury (for example, the 5461-A wavelength of mercury 198). Interferometric calibrations of end standards of length have been made for many years. Such calibrations of line standards of length have been made in Germany, and the method is being followed with considerable interest at the Bureau. 3. Facilities for Calibrating Line For the calibration of line standards of length the most precise measurements now being made at the Bureau are on its 1-m longitudinal comparator (fig. 2). This has been briefly described by Page. It is especially suited for the calibration of the subintervals of a bar, using any one of the several methods that have been developed. Although this comparator is nominally a 1-m comparator, it is readily adaptable for standardizing 48-in. bars, and can be used for calibrating even longer bars. For comparing two bars with an accuracy not better than 1 micron (0.001 mm or 0.00004 in.) a simple transverse comparator with no thermal insulation is commonly used. In this instrument the bars are mounted parallel to each other, and the carriage supporting the two bars moves back arallel to the axis of each of the two bars. The nicrometer microscopes are not moved during a omparison with this transverse comparator. Short bars and small scales are often more coneniently calibrated by means of a linear dividing ngine having two microscopes conveniently ounted so that they can be focused on the bars ounted on the longitudinally moving carriage. The instrument is then, in effect, a longitudinal omparator. Micrometer microscopes having a pair of parllel "cross hairs" moving in the focus of a posive-type ocular are probably the most important arts of any instrument for comparing line standrds of length. The movement of the cross wires measured by a precision screw and a graduated rum. In the comparators in the length section the Bureau the axes of the microscopes are ertical. . Facilities for Calibrating Measuring Tapes At the Bureau there is a temperature-controlled boratory about 212 ft long (fig. 3). Two distinct stallations have been made in this laboratory. ne is a 200-ft steel bench for calibrating steel pes used in general surveying and engineering actice, as shown at the right in figure 3. The her is a geodetic comparator designed especially r calibrating the 50-m invar base-line tapes, such are used by the U. S. Coast and Geodetic Sury. Its use is not limited, however, to a 50-m gth. It is shown at the left in figure 3. 4.1. Steel-Tape Bench The steel-tape bench has an over-all length of 0 ft 1 in., a width of 21⁄2 in., and a thickness of in., and is constructed of stainless-steel bars proximately 12 ft long, with individually lapped d fitted ends. Stainless-steel conical dowel pins ving threaded lower ends for locking nuts urely hold adjoining bars in the correct position. e supports for the bench are attached firmly to wall, but can be adjusted whenever necessary. e bench is graduated at intervals ordinarily quired for testing tapes graduated in the metric in the United States customary system. The ipment is furnished with the necessary supports the tape when it is to be supported at specified nts, with apparatus for applying the tension, rmometers for observing the temperature, and necessary clamps and other auxiliary equip nt. Comparisons of a tape with the bench standard ordinarily made with a low-power microscope da precision steel scale graduated either to in. or to % mm. When unusually high accuracy required and the character of the graduation es warrants, comparisons with the bench can be de with a micrometer microscope. A 5-m bar, packed in melting ice when measurements are being taken, is the working standard used as the basis of measurements in the geodetic tape comparator. Piers bearing microscopes are spaced 5 m apart for a total length of 50 m. Auxiliary piers 1 m apart, also bearing microscopes, are placed between the 20-m and the 25-m piers. These provide intervals so that the 5-m bar can be standardized by means of a calibrated 1-m bar. By use of a special bracket at the 15-m pier, a double pier at the 30-m location, and an additional pier near the 45-m pier, it is possible to mount microscopes at the 50-, 100-, and 150-ft points. In figure 4 are shown the 15-m and 50-ft microscopes at the extreme left, then the covered trough and carriages for the 5-m bar, immediately beyond that the six piers with their microscopes spaced 1-m apart, and other piers beyond them. In addition to the basic equipment of this comparator, there is the necessary auxiliary equipment, such as tape clamps, thermometers, etc. Most of the work done with this comparator is with the 50-m interval, and the use of this interval will be assumed for the remainder of this section. After the 50-m interval is established by moving the 5-m bar to measure the distance between each of the 10 possible 5-m intervals between microscopes, and the positions of the 0- and 50-m points are transferred from the focuses of the microscopes to centers of two hemispheres in piers at floor level, the 5-m bar is moved outside the 50-m interval. An invar tape to be calibrated is then mounted in the comparator, supported on ballbearing wheels, and proper tension is applied to the tape by means of a calibrated weight. The difference between the interval on the tape. and the established 50-m interval is obtained by the micrometer microscopes. 5. Calibration of Line Standards of Length The measurement of line standards of length undertaken by the Bureau is classified under the several headings that follow. A complete calibration of this type of length standard includes the determination of the length at a known temperature and also of the expansivity usually expressed by the average coefficient of linear thermal expansion over a small range of temperature that embraces the temperatures at which the standard is likely to be used. For most work with any standard except one of the highest grade it is sufficient to assume a coefficient of expansion derived from a knowledge of the composition of the materials of which it is made. If the standard is subdivided, a calibration of the subdivisions may also be necessary. The number of intervals that it is advisable to compare will depend on the character of the standard and on the use to which it is to be put. In many cases the subdivisions; a proportionate part of the errors for the total length may be assumed. Where it is necessary to measure a greater or less distance than the full length of the standard it is only necessary, in many cases, to test a few subdivisions; for example, any distance in even feet. may be accurately measured with a 100-ft tape if the corrections are known for the entire length of the tape, for each 10-ft subdivision, and for each foot of the first 10 ft. It should be noted that while platinum-iridium may be the best material for fundamental length standards, other materials are better suited to the needs of manufacture and industry; for example, steel rules and tapes are better for determining the sizes of steel and iron machine parts and structural members, because changes in length with changes in temperature are practically the same in the steel rule or tape as in the steel or iron part being measured. Metric standards should be graduated to be correct at either 0° or 20° C, the latter temperature being preferred for everyday use because it eliminates to a great extent the question of differences in expansion. In this country, standards in the customary units of yards, feet, and inches are made to be correct at 68° F (20° C). Requests for calibrations to be made by the Bureau should state the use to which the results are to be applied and also the accuracy desired, in order that the tests may be adequate for the end in view, avoiding unnecessary labor and expense. Where the highest attainable accuracy is needed, it is sometimes advisable to have the standard verified immediately before and again immediately after the important measurements in order to guard against possible changes due to undetected injury or structural alteration. 5.1. Reference Standards This class includes standards of the highest type suitable for reference standards for makers of precision apparatus and for use in the most. exact scientific investigations. One of the best constructions is that conforming to the specifications adopted for the prototype meters by the International Committee on Weights and Measures; but less expensive constructions are sufficient for all but exceptional purposes. The bars should be of the X-shaped or of the H-shaped cross section with all rulings in the plane of the neutral axis. The graduations should be accurately perpendicular to the longitudinal axis of the standard and lines must be fine, with sharp, smooth edges, and ruled on plane surfaces that have been given a faultless mirror or dull polish, depending upon the type of illumination used. The lines on the prototype meters are about 0.006 mm wide. Since the time of the construction of these meters there has been a very marked microscopes of higher magnification than those used in the latter part of the last century. But, whatever may be the width of the line, it is most important that the line be symmetrical, that is. that the bottom of the groove made by the tracing mechanism be symmetrical with the edges of the groove line. The portion of the graduations to be used should be defined by two parallel longitudinal lines about 0.2 mm apart. A material should be used that does not oxidize or otherwise tarnish on exposure to air or moisture. The graduations must not be covered by varnish or other protective covering. A very useful alloy is one made of approximately 36 percent of nickel and 64 percent of iron, known as invar. Invar possesses a coefficient of expansion that is almost negligible at ordinary temperatures; in addition, it does not rust or tarnish readily on exposure to the atmosphere. It is important to remember that most materials, alloys in particular, undergo slight changes in the course of time, espe cially if subjected to considerable changes of temperature or to mechanical disturbances. Hence, whenever the highest accuracy is desired, standards should occasionally be verified. Standards of pure nickel and 42 percent of nickel have been found to be sufficiently stable for use as standards where verifications of the length of such bars can be made at suitable intervals of time. The owner of a length standard of the reference class is usually justified in having a calibration made by the Bureau. 5.2. Working Standards This class includes standards suitable for all ordinary precision work and suitable for the needs of college laboratories, manufacturers of the better grades of scientific apparatus, State superintendents of weights and measures, and most scientific work. Either an H-shaped cross section or a rectangular cross section with provision for supports at definite positions should be used. Short bars, such as decimeter bars, may be of rectangular cross section with the graduations on the upper surface and with the bars supported directly on a flat surface. The lines of the graduations should be sharp and less than 0.03 mm wide, and they should be ruled on a plane, well-polished surface that will not tarnish readily on exposure to the atmosphere. No varnish or other protective covering should be used. If the metal of the bar tarnishes readily, the lines should be ruled on plugs or strips of nontarnishing metal. Some such means as a pair of parallel longitudinal lines about 0.2 mm apart should be provided for defining the portion of the graduations to be used, and to facilitate the alinement of the bar. The graduation lines should be accurately perpendicular to the longitudinal axis of the standard. Standards of this class are compared with working standards of the Bureau and are certified to 0.001 mm if their quality justifies it. |