Imagens da página
PDF
ePub

The following basic formula used by Sectional Committee B36 for computing its pipe wall thickness schedules was derived from the modified Barlow formula then used in the ASME Boiler Code:

Where P

[blocks in formation]

maximum internal service pressure in pounds per square inch gage Sallowable stress in material due to internal pressure, at the operating temperature, pounds per square inch

t = nominal wall thickness of pipe, in.

In adjusting the Barlow formula to this purpose, 2 was changed to 1.75 to look after the 1212 per cent mill tolerance in pipe manufacture, and the allowance C for corrosion, etc, was rounded out to 0.1 in. The formula was not applied to nominal sizes smaller than one inch, these thicknesses being selected according to traditional usage from commercial lists.

Using the above formula as a basis, Sectional Committee B36 endeavored to evolve its principal pipe schedules around the then commonly used "standard weight" and "extrastrong" pipe lists. In this way much of the "standard-weight" list was used for Schedule 40 and "extrastrong" for Schedule 80. Double-extra-strong pipe could not be fitted into the schedule system inasmuch as it did not follow any regular P/S pattern, but it has been retained in the B36 standard because it is still manufactured by the pipe mills and fills certain needs as illustrated later in this article.

As stated in the American Standard for Wrought Iron and Wrought-Steel Pipe, ASA B36.10, formula (3) is not intended to be used as a basis of design. It is contemplated that the user will compute the exact value of wall thickness suitable for the conditions for which the pipe is required as described in detail in the ASME Boiler Code, the American Standard Code for Pressure Piping, or similar safety regulations. From the schedules of nominal thicknesses given in B36.10, a thickness may then be selected to fulfill the conditions for which the pipe is desired. If pipe is

ordered by its nominal weight or wall thickness as is customary in trade practice, the manufacturing tolerance on wall thickness must be added to the minimum thickness computed by code formulas. The next heavier commercial thickness may then be selected from the standard thickness schedules of B36.10.

The following formula for pipe wall thickness appears in current editions of the ASME Boiler Code and the Power Section of the American Standard Code for Pressure Piping:

tmin =

+ C

(4)

PD 2S+0.8P Where tmin is the minimum pipe wall thickness in inches after allowing for manufacturing tolerance, and C is an allowance, in the case of plain end pipe, for "mechanical strength and/or corrosion" of 0.065 in. for pipe sizes above 1 in.

In order to put the code formula on a basis comparable to that of the B36 formula, it is only necessary to substitute the numerical value of C and allow for the 1212 per cent manufacturing tolerance by putting tmin = 0.875 t, where t is the nominal thickness.

[blocks in formation]

Selection of Proper Pipe Wall Thick

ness Assisted By Use of Chart

The accompanying chart is offered as a means of assisting designers in the selection of the proper pipe wall thickness for any given service conditions in power piping using plain. end steel pipe. In some cases the 1000 P/S value determined for the given conditions can be accommodated for all sizes by a single pipe schedule of those available. As illustrated in the first example given later in this article, this fortunate situation can be identified at once from the

chart as soon as the 1000 P/S value is determined.

In other attempts to select the proper pipe wall thicknesses for a given set of power plant service con ditions it sometimes develops tha: no single schedule number will sat isfy code requirements in all pipe sizes. Under these conditions it be comes necessary to pick and choose from available schedules in order to make a series of selections that will be satisfactory for each of the respective sizes to be used. While this could be done through repeated solutions of the pipe wall thickness formulas given in the ASME Power Boiler Code and in the Power Pip ing and District Heating Sections of the American Standard Code for Pressure Piping, the process would involve considerable time and effort with the chance of errors in the rather involved computations. It is simpler and more dependable, there fore, to read the answer from a char such as the one accompanying this article. In this way the problem ca be visualized as a whole so that ir consistencies become apparent. Fr thermore, the chance of making. error is reduced in proportion to the ease of solution.

The graphs shown on the char were plotted from equation (6) for pipe sizes larger than one inch, and using a similar formula with C = 0.050 for sizes one inch and smaller As a result the graphs can be use directly for determining allowabl P/S ratios for plain-end pipe of the thickness schedules shown witho having to make further adjustments for manufacturing tolerance and C allowance. In other words, manufacturing tolerances and C allow ances are already provided for in computing and plotting the graphs

Examples of Using the Chart

The following examples are give for the purpose of explaining how to use the graph.

Example 1-Simple Case Where Oni Schedule Suffices

Problem: What pipe schedule

should be used for a steam system

(Continued on page 194)

[merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][ocr errors][merged small][merged small][merged small]

A

A Guide for Measurement of Color

LL our knowledge comes to us through our senses. Even measurements and specifica

[ocr errors]

tions must "make sense," else they tell us very little. Yardsticks, thermometers, weights, and clocks are only indirect means of recording facts that are revealed to us directly by our senses.

Primary in importance though they are, senses are too inaccurate, too unstable, and too uncertain in memory for the needs of these times. A foot cannot be estimated consistently to better than about one quarter of an inch. Fever temperature, to which we are most sensitive, cannot be judged with certainty much closer than one degree. Judgment of a pound cannot be relied upon to much better than an ounce, and it is almost impossible to signal a minute's duration with an error of less than two seconds. After concentrating on tiny things, such as watchparts, our estimate of a foot is likely to be short as much as an inch. After soaking our hands in hot water, we are likely to "feel" that a fevered person's brow is cool. After moving a number of heavy boxes, a pound of sugar seems short-weight, and forty seconds is likely to seem like a long minute. Coming out of a motion picture theatre, we are likely to find daylight unbearably bright, and it is not wise to trust the unaided memory of the colors of furnishings when selecting draperies.

On the other hand, the senses yield great precision when two similar things are directly compared, or when the position of a pointer is to be read on a scale. To be sure of the length of an object within 1/100 of an inch or better, we need only place it in coincidence with a good ruler and look. To be sure of a temperature within 1/10 of a degree. or better, we need only place a thermometer in the object, wait for equilibrium, and look at the position of the top of a mercury column on a

by David L. MacAdam

Eastman Kodak Company Chairman, Subcommittee 7 ASA Sectional Committee Z58 on Optics

scale. To be sure of a weight within a small fraction of an ounce, we need only place the object on one side of a balance and place known weights on the opposite side until the position of a pointer indicates equality or a small calibrated deviation from it. To be sure of a duration within a few tenths of a second, we need only press the controls of a stop watch at the beginning and end of an event and see the position of the pointer on the dial. The accuracy

of any of these schemes, may be greatly increased by use of mechanical, electrical, and optical refinements. But in all cases, the final step involves one or another of our senses. For example, in the most accurate measurements of length it is

necessary to note the position of an interference fringe, or to read a photoelectric counter. We must note the position of the optical image of a hair-line on the scale of a microanalytical balance or of an oscillo scope pattern in some modern measurements of time. Use of the senses is the only permanent requirement of all measurements.

As measurements become more precise, standards become essential. The definition of measurements in terms of basic principles become more important when the highest accuracy is sought.

All the reasons for relying on measurements-of length, weight. temperature, and time-in preference to sense, apply with even greater force to color. We cannot estimate color in the sense that we can esti mate length, weight, temperature. or time because there is no generally understood manner of expressio like foot, degree, pound, or minute Nor is such a manner of expression likely to become common, because

[graphic][merged small][merged small][merged small]

color involves three independent variables which are blended in a manner too subtle for commonplace distinction. Furthermore, the color sense is perhaps more subject to variation, with conditions of observation and previous experience, than any other sense. Finally, the absence of any generally understood manner of describing color handicaps our memory for colors, or perhaps difficult memory is partly responsible for poor communication.

On the other hand, our capacity for noticing small differences of adjacent colors is phenomenal and taxes to the utmost the accuracy of the best methods of measuring color. Like all measurements, color measurements must "make sense"; that is, they must detect the smallest color differences that can be seen, and should do better than that if they are to be useful for anything other than communication and record.

Although a sufficiently accurate technique for measuring color has been developed only recently, the use of so-called "color standards" has been common for over thirty years. These consist of materials which are visually compared with the sample. This may be simple side-byside comparison or may be facilitated by optical devices for juxtaposing the colors of the standard and sample. To cover the wide range of colors (several million different colors being distinguishable), sets consisting of fairly large numbers of "color standards" have been used. To bring order out of the chaos of several hundred such standards and to insure against serious gaps, they have been arranged in a systematic manner by their originators.

Although all such systems of arrangement agree in using three independent variables, the choice of specific variables is subject to wide differences, and the several popular "color systems" differ primarily in this choice. The names of Munsell, Ostwald, and Lovibond refer to dif ferent systems of arranging "color standards."

These standards have not been adopted by any authoritative bodies having broad bases of representation,

[graphic][graphic][subsumed][merged small]

nor are they the subject of the project carried on under the procedure of the American Standards Association which is under discussion here.

Hundreds of other "color standards" are in use in commerce and industry. Usually each represents a desired color with which product. samples are compared. Sometimes "tolerance colors" are used to bracket the desired color. Thus, a certain yellow carton color is specified between four tolerance limits, indicating how red or how green the yellow may be, and how light and how dark. These tolerance samples, and "color standards," in general, should perhaps be regarded as gages rather than as standards. There is little to be

gained by standardizing gages, each of which is of use only in specific applications, often confined to a single product of a single manufacturer, and subject to frequent change with fashions.

It is much more important to standardize methods for measuring the colors of such gages. Material samples, especially when handled frequently in making comparisons with products, become soiled, faded, damaged, and lost. It is important to measure and record their original colors in some unambiguous, communicable way so that equivalent new inspection "standards" or "tolerances" can be selected even after the passage of years. This is the intent

of three new American Standards on methods of measuring and specifying color, just published.1

Suppose we have selected a set of samples of a product, illustrating the extreme limits of permissible color variations; for example, a greenishyellow limit, a reddish-yellow limit, a dark and a light limit. Of course, we will try to make a supply of such sets sufficient for all foreseeable needs of all concerned. But to guard against loss, soiling, fading, and "doubting Thomases" ten years hence, we put each of the tolerance samples in an optical instrument called a spectrophotometer. This produces for us a permanent, quantitative record of the proportion of the light reflected by the sample for each portion of the visible spectrum. Because the measurement is complicated by irregularities of surface, by degree of glossiness, by fluorescence, and by many other details too numerous to list, our results will mean much more to us ten years from now, and to our suppliers and customers, if we use a spectrophotometer which conforms to functional quirements established by American Standard Z58.7.1-1951.

re

As the years go by, the dyes or pig. ments used in the original lots may be hard to get; or better and cheaper materials may become available. With them we may have no trouble matching the appearance of the color, but may find it impossible to make samples having spectophotometric curves like those of the original samples or of any of the tolerances. At this point we turn to well-established data concerning the behavior of hu man vision. Arithmetical methods are well known which enable us to decide whether samples having different spectrophotometric curves will look alike in daylight or other prescribed light. There has been a rather high degree of uniformity of American and even of international

1 American Standard Method of Spectrophotometric Measurement of Color, Z58.7.11951; American Standard Method for Determination of Color Specifications, Z58.7.2-1951; American Standard Alternative Methods for Expressing Color Specifications, Z58.7.3-1951.

practice in this connection for the past two decades. The basic principles and data for this work are the object of the second of the new standards, American Standard Z58.7. 2-1951. With them we can check samples made with new materials against the original tolerance limits, and if the mill will make enough variations, spectrophotometry supplemented with these "colorimetric" calculations enable us to select new tolerance samples made of the new materials. They are preferable to the old tolerance samples for use in dayto-day visual checking of products. Their appearance relative to goods made with the new materials is less dependent on variations of the quality of light at the inspector's station, and less dependent on the inspector having exactly normal color vision. Experience has shown that after the

• A course in Industrial Standardization was given in Philadelphia this Spring by Madhu S. Gokhale and Fred M. Oberlander, both of the RCA Victor Division, Radio Corporation of America. The 11-session course was sponsored by the Management Service Division of Temple University's Community College and Technical Institute, of which Harry C. Rountree is Dean.

Students in the course represented leading Philadelphia organizations, such as Minneapolis-Honeywell Regulator Company, Belfield Valve Division; Michael Flynn Manufacturing Company; Sharples Corporation; U.S. Navy; International Resistance Company; and RCA Victor Division.

In addition to Mr Gokhale and Mr Oberlander, three guest speakers addressed the class: Colonel Thew Ice, Armed Services Electronic Standards Agency, on electrical and electronic components; I. W. Markowitch, In Charge of Plating, U.S. Gauge Company, on surface coatings; and S. H. Watson, Manager, Standardizing, RCA, on administrative aspects.

Subjects covered were:

Introduction-Definition, aims and scope; types of standards; history of standardization: determination of the proper time for

standardization.

spectrophotometer and colorimetry have been used for a few years in this manner to record permanently the colors of the tolerance limits. part of the task of selecting tolerance limits for new products is, quite nat urally, shifted to that instrument and method. Then it becomes essential to know quantitatively what constitutes. for instance, the difference called "greener," without confusing it with either "lighter" or "stronger." For many other purposes, too, such as in terpreting spectrophotometric and colorimetric results to persons rela tively unfamiliar with the subject, it is desirable and even necessary to be able to translate from the working data to the traditional language of color. The third new standard group American Standard Z58.7.3-1951. provides for uniformity of American practice in such translation.

Development of Standardization-Activ ities of Government agencies, America Standards Association, Technical Societi and Trade Associations. Method of devel oping new company standards.

Standardization of Basic Materials Advantages of a purchasing specificati system; types of specifications: factors? the selection of standard materials.

Standardization of Electrical Components -Basis for standardization; electrical ty tests and their significance.

Standardization of Mechanical PartsSimplification and classification of fre quently used items; advantages of using standardized parts.

Standardization of Protective and Dec rative Chemical Treatments and Surface Coatings-Tests for durability, uniformity and protection; economic and estheti considerations affecting choice of finishes

Drafting Standards-Selection of form and size; function of a drawing; sequen tial vs. significant numbering systems types of drawings; dimensional tolerances,

Application of Standards-Preferred numbers, sizes and gauges: publication and distribution of standards; compliance with standards and responsibility for devi ations; coding practices; consultation serv ice; educational programs.

[blocks in formation]

1

« AnteriorContinuar »