Kutz M. Handbook of materials selection

Подождите немного. Документ загружается.

454 HOW TO FIND MATERIALS PROPERTIES DATA

time to call on colleagues (if you have not before). Then searching the web

might bring up a new lead. Or then it may simply be time to stop and go with

what you have. Only the materials engineer can make that decision.

3 DATABASES

Big Five Literature Databases. All of the major science literature databases

provide indexing and abstracting of information from the primary literature from

around the world. All contain full bibliographic information that is retrieved

through subject indexing and free text searching of titles and abstracts. Each

database contains some references to non-English language materials. The types

of primary literature covered vary by discipline. All of the databases are available

through major database vendors such as DIALOG, STN International, OCLC

FirstSearch, Axiom (from IOP), and OVID. Many also have front ends and web

search engines direct from the publisher. Different search features are available

through different search interfaces. Most databases include title, abstract, index-

ing, and references. Many provide links to full-text information. The most up-

to-date information about access options can be found on the databases or

database vendors home pages. One way to evaluate which database to use is to

look at the list of titles covered in a particular database. Note which titles are

in a database ‘‘cover to cover’’ (i.e., all articles are entered) and which are

selectively added.

1. BIOSIS—Biological Abstracts. BIOSIS, Two Commerce Square, 2001

Market Street, Suite 700, Philadelphia, PA 19103-7095. Phone: 1–800–523–

4806 or

⫹1–215–587–4800. Fax: ⫹1–215–587–2016. Web: http://www.biosis.

org. E-mail: info@mail.biosis.org . Indexes and abstracts in the biological, bio-

chemical, and life sciences articles, books, dissertations, and conference pro-

ceedings. It has an extensive biological classiﬁcation scheme. Available through

most major database vendors. An added feature through STN: addition of CAS

registry numbers to the indexes for compound listed in the article title. An

excellent resource for biomaterials.

2. INSPEC—Physics Abstracts, Electrical and Electronics Abstracts, Com-

puter and Control Abstracts. Institute for Electrical Engineers, Michael Faraday

House, Six Hills Way, Stevenage, Herts, SG1 2AY, United Kingdom. Phone:

(

⫹44) 1438/313311, US 732-321-5575. Fax: (⫹44) 1438/742840, US 732-321-

5702. E-mail: inspec@iee.org.uk, or inspec@inspecinc.com. 1969 to present.

International in scope, the database contains over 2 million records for articles,

conference proceedings, and books. The materials engineer should use this da-

tabase to access information, pertaining to the physical properties of materials,

in areas of physics, polymers, ceramics, computational methods, and biophysics.

An interesting feature of this database is the element and physical property ﬁelds

that can allow the materials engineer to specify what type of chemical compo-

nents are discussed in the primary literature. An excellent resource for infor-

mation on sold-state materials.

3. MEDLINE—Index Medicus. National Library of Medicine, MEDLARS

Management Section, 8600 Rockville Pike, Bethesda, MD 20894. Phone: 1–

800–638–8480. Fax:

⫹1–301–496–0822. Web: http://www.nlm.nih.gov. E-

3 DATABASES 455

mail: mms@nlm.nih.gov. Indexes and abstracts the ﬁelds of medicine, nursing,

dentistry, veterinary medicine, the health care system, and the preclinical sci-

ences. MEDLINE contains bibliographic citations and author abstracts from

more than 4000 biomedical journals published in the United States and 70 other

countries. Available through many database vendors and directly (at no cost)

from the National Library of Medicine. PubMed also provides access and links

to the integrated molecular biology databases included in NCBI’s Entrez retrieval

system. These databases contain DNA and protein sequences, 3-D protein struc-

ture data, population study data sets, and assemblies of complete genomes in an

integrated system.

4. CAS—Chemical Abstracts and the Registry File. Chemical Abstract Ser-

vice, 2540 Olentangy River Road. Columbus, Ohio 43202. Phone: 1–614–447–

3700. Fax: 1–614–447–3767. Web: http://www.cas.org. Email: help@cas.org.

CAS indexes and abstracts in chemistry, biochemistry, and other chemical sci-

ences papers, books, conference proceedings, patents, technical reports, and dis-

sertations worldwide (1907 to current). Chemical name, chemical structure,

genome, and protein sequence searching are available in the Registry. Other CAS

ﬁles provide detailed synthesis and reaction information. For ceramics, inorganic

materials, and organometalic materials searching molecular formula, element

counts and ring systems in the Registry ﬁle allow a high level of speciﬁcity as

well as broad class identiﬁcation of compounds and materials. The most com-

plete versions of these ﬁles is available through STN International; however, the

basic databases are available through all major database vendors.

5. COMPENDEX (or EI or Engineering Index). Engineering Information

Inc., EI Headquarters, 1 Castle Point Terrace, Hoboken, NJ 07030-5996. Tel:

(

⫹1) 201–216–8500. Fax: (⫹1) 201–216–8532. Toll free: (⫹1) 800–221–1044

US and Canada. Web: http://www.ei.org/. E-mail: ei@ei.org. Indexes and ab-

stracts information from the engineering and technological literature including

over 2600 international journals, conference papers, and technical reports. Of

special interest to materials engineers is the chemical, biomedical, and metals

formation.

Other Databases

1. Science Citation Index (part of Web of Science). ISI 3501, Market Street,

Philadelphia, PA 19104. Phone: 800–336–4474 or 215–386–0100. Fax: 215–

386–6362. Web: http://www.isinet.com/isi/. A citation index provides ready

access to information on what papers are routinely being cited. They provide a

method to broaden searches by ﬁnding out how seminal papers have been cited.

Several database producers are now adding citation searching within their subject

areas.

Science Citation Index provides access to current and retrospective biblio-

graphic information, author abstracts, and cited references found in 3500 of the

world’s leading scholarly science and technical journals covering more than 150

disciplines. Its focus is on the most heavily used materials in all of science, thus

allowing cross-disciplinary searching and retrieval of the most important refer-

ences. (A related product Materials Science Citation Index may also be of in-

terest as it provides a more focused group of journals.) Keyword and title word

456 HOW TO FIND MATERIALS PROPERTIES DATA

searching is available, making this a useful but limited tool for subject searching.

This database is best used in conjunction with more in-depth and formally in-

dexed subject database.

2. ProceedingsFirst. OCLC Online Computer Library Center, Inc., 6565

Frantz Road, Dublin, Ohio 43017-3395. Phone: 1–800–848–5878. Fax:

⫹1–

614–764–6096. Web: http://www.oclc.org. E-mail: support@oclc.org. Provides

citations of every congress, symposium, conference, exposition, workshop, and

meeting received at The British Library. Contains in each record a list of the

papers presented at each conference.

3. Derwent World Patents Index. Derwent Information, 1725 Duke Street,

Suite 250, Alexandria, VA 22314. Telephone:

⫹ 1–703–706–4220. Fax: ⫹1–

800–457–0850. Web: http://www.derwent.com/index.html. E-mail: custserv@

derwentus.com. Derwent reviews and processes more than a million patents a

year from over 40 international patent issuing authorities, as well as research

papers and articles from more than 1200 other sources. The World Patents Index

is one of the ﬁles they offer. Extensive chemical and engineering related indexing

terminology as well as Markush structure searching is available in these re-

sources.

REFERENCES

1. J. B. Park, ‘‘Biomaterials,’’ in The Biomedical Engineering Handbook; 2nd ed., Joseph D. Bron-

zino (Ed.), CRC Press, Boca Raton, FL, 2000; online, ENGnetBASE, 4 / 24 / 01.

2. R. B. Martin, ‘‘Biomaterials,’’ in The Engineering Handbook, Richard C. Dorf (Ed.), CRC Press,

Boca Raton, FL, 2000, p. 2032.

3. C. G. Gebelein, ‘‘Prosthetic and Biomedical Devices,’’ in Kirk-Othmer Encyclopedia of Chemical

Technology, 3rd ed., Vol. 19, Wiley, New York, 1982, p. 277.

4. G. S. Brady and H. R. Clauser, Materials Handbook, 13th ed., McGraw Hill, New York, 1991 p.

93.

5. Dorland’s Illustrated Medical Dictionary, 29th ed., Saunders, Philadelphia, 2000, p. 212.

6. D. J. Plazek and K. L. Ngai, ‘‘The Glass Temperature,’’ in Physical Properties of Polymers Hand-

book, James E. Mark (Ed.), American Institute of Physics, Woodbury, NY, 1996, p. 153.

7. R. B. Fox, ‘‘Glass Transition Temperature for Selected Polymers,’’ in CRC Handbook of Chemistry

and Physics, 80th ed., David R. Lide (Ed.), CRC Press, Ohio, 1999, Section 13, p. 4.

8. http:/ / www.ncbi.nlm.nih.gov:80 / entrez / query / static / overview.html#Database Coverage, May 1,

2001.

457

CHAPTER 16

SOURCES OF MATERIALS DATA

J. G. Kaufman

Kaufman Associates, Inc.

Columbus, Ohio

1 INTRODUCTION AND SCOPE 457

2 INTENDED USES FOR DATA 458

2.1 Modeling Material and /or

Product Performance 458

2.2 Materials Selection 458

2.3 Analytical Comparisons 459

2.4 Preliminary Design 460

2.5 Final Design 460

2.6 Material Speciﬁcation 461

2.7 Manufacturing 461

2.8 Quality Assurance 462

2.9 Maintenance 462

2.10 Failure Analysis 463

3 TYPES OF DATA 463

3.1 Textual Data 463

3.2 Numeric Databases 464

3.3 Metadata 464

4 SUBJECTS OF DATA SOURCES 465

5 DATA QUALITY AND

RELIABILITY 465

6 PLATFORMS: TYPES OF

DATA SOURCES 467

7 SPECIFIC DATA SOURCES 468

7.1 ASM International 468

7.2 STN International 469

7.3 Internet 470

REFERENCES 472

1 INTRODUCTION AND SCOPE

It is the purpose of this chapter to aid engineers and materials scientists in

locating reliable sources of high-quality materials property data. While sources

in hardcopy form are referenced, the main focus is on electronic sources that

provide well-documented searchable property data.

To identify useful sources of materials data, it is important to have clearly in

mind at the outset (a) the intended use of data, (b) the type of data required,

and (c) the quality of data required. These three factors are key in narrowing a

search and improving its efﬁciency of a search for property data. Therefore, as

an introduction to the identiﬁcation of some speciﬁc potentially useful sources

of materials data, we will discuss those three factors in some detail, and then

describe the options available in types of data sources.

It is beyond the scope of this publication to attempt to provide a comprehen-

sive list of all of the several thousand sources of materials data in various forms

and formats. Readers interested in a more comprehensive list of sources and of

more discussion of the technology of material property data technology and

terminology are referred to Westbrook’s extensive treatment of these subjects in

HandbookofMaterialsSelection.EditedbyMyerKutz

458 SOURCES OF MATERIALS DATA

Refs. 1 and 2 and to ASM International Directory of Material Property Data-

bases.

2 INTENDED USES FOR DATA

Numeric material property data are typically needed for one of the following

purposes by individuals performing the respective functions as part of their jobs:

●

Modeling material or product performance

●

Materials selection (ﬁnding candidate materials for speciﬁc applications)

●

Analytical comparisons (narrowing the choices)

●

Preliminary design (initially sizing the components)

●

Final design (assuring performance; setting performance speciﬁcations)

●

Material speciﬁcation (deﬁning speciﬁcations for purchase)

●

Manufacturing (assuring processes to achieve desired product)

●

Quality assurance (monitoring manufacturing quality)

●

Maintenance (repairing deterioration/damage)

●

Failure analysis (ﬁguring out what went wrong)

It is useful to note some of the differing characteristics of data needed for these

different functions.

2.1 Modeling Material and/or Product Performance

To an increasing extent, mathematic modeling is used to establish the ﬁrst es-

timates of the required product performance and material behavior, and even in

some cases the optimum manufacturing processes that should be used to achieve

the desired performance. The processes and/or performance analyzed and rep-

resented may include any of the issues addressed in the following paragraphs,

and so the nature of the types of data described under the various needs are the

same as those needed for the modeling process itself.

2.2 Materials Selection

The needs of materials specialists and engineers looking for materials data to

aid in the selection of a material for some speciﬁc application are likely to be

inﬂuenced by whether they are (a) in the early stages of their process or (b)

have already narrowed the options down to two or three candidates and are trying

to make the ﬁnal choice. The second situation is covered in Section 2.3, Ana-

lytical Comparisons.

If the materials engineers are in the early stages of ﬁnding candidate materials

for the application, they are likely to be looking for a wide variety of properties

for a number of candidate materials. More often, however, they may decide to

focus on two or three key properties that most closely deﬁne the requirements

for that application and search for all possible materials providing relatively

favorable combinations of those key properties. In either case, they may not be

as much concerned about the quality and statistical reliability of the data at this

stage as much as the ability to ﬁnd a wide variety of candidates and to make

direct comparisons of the performance of those candidates.

2 INTENDED USES FOR DATA 459

In cases where there is interest in including relatively newly developed ma-

terials in the survey, it may be necessary to be satisﬁed with only a few repre-

sentative test results or even educated ‘‘guesstimates’’ of how the new materials

may be expected to perform. The engineers will need to be able to translate

these few data into comparisons with the more established materials, but at this

stage they are probably most concerned with not missing out on important new

materials.

Thus, at this early stage of materials selection, the decision makers may be

willing to accept data rather widely ranging in type and quality, with few re-

strictions on statistical reliability. They may even be satisﬁed with quite limited

data to identify a candidate that may merit further evaluation.

2.3 Analytical Comparisons

If on the other hand, the task is to make a ﬁnal decision on which of two or

three candidate materials should be selected for design implementation (the pro-

cess deﬁned here as ‘‘analytical comparison’’), the quality and reliability of the

data become substantially more important, particularly with regard to the key

performance requirements for the application. It will be important that all of

those key properties, e.g., density, tensile yield strength, and plane-strain fracture

toughness, are available for all of the candidate materials that may be the next

cut in the list.

The search will also be for data sources where the background of the data

are well deﬁned in terms of the number of tests made, the number of different

lots tested, and whether the numbers included in the data source are averages

or the result of some statistical calculation such as that to deﬁne three

␴

limits,

for example. It would not be appropriate at this stage to be uncertain whether

the available data represent typical, average values or statistically minimum

properties; it may not be important which they are, but the same quality and

reliability must be available for all of the ﬁnal candidates for a useful decision

to be made.

In addition, the ability to make direct comparisons of properties generated by

essentially the same, ideally standard, methods is very important. The decision

maker will want to be able to determine if the properties reported were deter-

mined from the same or similar procedures and whether or not those procedures

conformed to ASTM International, International Organization for Standardiza-

tion (ISO),

or other applicable standard test methods.

One ﬁnal requirement is added at this stage: The materials themselves for

which the data are presented must all represent to the degree possible comparable

stages of material production history. It would be unwise to base serious deci-

sions on comparisons of data for a laboratory sample on one hand and a com-

mercial size production lot on the other. Laboratory samples have a regrettable

history of promising performance seldom replicated in production-size lots.

Thus, for analytical comparisons for ﬁnal candidate material selection, spe-

cialists need databases for which a relatively complete background of metadata

(i.e., data about the data) are included and readily accessible.

Incidentally, it is not unusual at this point in the total process to decide that

more data are needed for a particular candidate than are available in any existing

460 SOURCES OF MATERIALS DATA

database, and so a new series of tests are needed to increase conﬁdence in the

comparisons being made.

2.4 Preliminary Design

Once a decision is made on a candidate material (or sometimes two) for an

application, the task of designing a real component out of that material begins.

The requirement for statistical reliability steps up, and the importance of the

availability of a data source that provides applicable metadata covering quality,

reliability, and material history becomes even more important.

At this stage, the statistical reliability required includes not only a minimum

value but one based upon a statistically signiﬁcant sample size, ideally something

comparable to the standards required in the establishment of MIL-HDBK-5’s A

or B values.

In MIL-HDBK-5 terminology, an A value is one that would be

expected to be equaled or exceeded by 99% of the lots tested with 95% conﬁ-

dence; the B value provides for 90% of lots tested equaling or exceeding the

value with 95% conﬁdence. Furthermore, the MIL-HDBK-5 guidelines require

that A and B values be based upon predeﬁned sample sizes, representing a

minimum number of lots (normally 100 or more) and compositions (normally

at least 3) of a given alloy. The provision of such statistical levels needs to be

a part of data sources used for design purposes, and the description of the

statistical quality needs to be readily available in the data source.

For preliminary design, then, the data sources sought will include both sta-

tistically reliable data and well-deﬁned metadata concerning the quality and re-

liability provided.

2.5 Final Design

Setting the ﬁnal design parameters for any component or structure typically

requires not only data of the highest level of statistically reliability but also, in

many cases, data that have been sanctioned by some group of experts for use

for the given purpose. It is also not unusual that at this stage the need is identiﬁed

for additional test data generated under conditions as close as possible to the

intended service conditions, conditions perhaps not available from any com-

mercial database.

Databases providing the level of information required at this stage often con-

tain what are characterized as ‘‘evaluated’’ or ‘‘certiﬁed’’ values. Evaluated data

are those that, in addition to whatever analytical or statistical treatment they

have been given, have been overviewed by an expert or group of experts who

make a judgment as to whether or not the data adequately and completely rep-

resent the intended service conditions, and if necessary incorporate their own

analysis into the ﬁnal ﬁgures. This technique has been widely used in digesting

and promulgating representative physical property data for many years; examples

are the thermophysical property data provided by Thermophysical Properties

Research Center (TPRC).

Other databases may be said to provide ‘‘certiﬁed’’ data. In this case, the

database or set of data going into a database have been evaluated by a group of

experts and certiﬁed as the appropriate ones to be used for the design of a

particular type of structure. Two examples of this are the aforementioned MIL-

HDBK-5 values,

which are approved for aircraft design by the MIL-HDBK-5

2 INTENDED USES FOR DATA 461

Coordination Committee consisting of aerospace materials experts; and the

American Society of Mechanical Engineers (ASME) Boiler and Pressure Vessel

Code,

with properties certiﬁed by materials experts in that ﬁeld for the design

of pressure vessels and companion equipment for high-temperature chemical

processes.

As noted, it is often the case when designers reach this stage (if not the earlier

preliminary design stage) that they ﬁnd it necessary to conduct additional tests

of some very specialized type to ensure adequate performance under the speciﬁc

conditions the component or structure will see in service but for which reliable

databases have not previously been identiﬁed or developed. The net result is the

creation of new materials databases to meet highly specialized needs in a manner

that provides the appropriate level of statistical conﬁdence.

2.6 Material Speciﬁcation

Material speciﬁcations typically include speciﬁc property values that must be

equaled or exceeded in tests of those materials that are being bought and sold.

The properties that one requires in this case may differ from those needed for

other purposes in two respects. First and foremost, they must be properties that

will ensure that the material has been given the desired mechanical and thermal

processing to consistently achieve the desired performance. The second require-

ment is that, while in most cases there may be only one or a very few properties

required (most often tensile properties), they are required at a very high level of

precision and accuracy, similar to or better than that required for the MIL-

HDBK-5 A properties deﬁned earlier.

Examples are the material speciﬁcations required for the purchase of com-

mercial aluminum alloy products.

These are usually only the chemical com-

position and the tensile properties. So while many of the other properties needed

for design are not required as part of the purchase speciﬁcation, those properties

that are required are needed with very high reliability. In the case of aluminum

alloys, the requirements for tensile strength, yield strength, and elongation are

normally that 99% of lots produced must have properties that equal or exceed

the published purchase speciﬁcation values with 95% conﬁdence, and they must

have been deﬁned from more that 100 different production lots from two or

more producers.

In many cases, the databases needed to generate material speciﬁcation prop-

erties are proprietary and are contained within individual companies or within

the organizations that set industry speciﬁcations. However, the resultant statis-

tically reliable speciﬁcation properties are resident and more readily available in

industry or ASTM material standards.

4,8

2.7 Manufacturing

The properties required for manufacturing purposes may be the most difﬁcult to

ﬁnd in commercially available databases because they typically involve the spe-

cialized treatments or processes utilized by speciﬁc producers or suppliers of the

speciﬁc products in question. Sometimes these processes are proprietary and

closely held for competitive advantage. An added complication is that once some

semifabricated component (e.g., aircraft sheet) has been purchased, it will require

forming to very tight tolerances or ﬁnishing at some relatively high temperature.

462 SOURCES OF MATERIALS DATA

The fabricator may require data to enable the process to be carried through

without otherwise damaging or changing the properties of the component, but

may have difﬁculty representing the fabricating conditions in meaningful tests.

That fabricator may well have to carry out its own tests and build the needed

database to provide the desired assurance of quality, and to provide a source of

information to which its own employees can refer to answer speciﬁc questions.

Typically such databases never become commercially available, and new situa-

tions will require compiling new data sources.

Some processing data sources are available, of course. The Aluminum As-

sociation, Inc., for example, provides to all interested parties a data source de-

ﬁning standard solution heat treating, artiﬁcial aging, and annealing treatments

for aluminum alloys that will assure the proper levels of properties will be

obtained.

In some cases, ASTM and the American National Standards Institute

(ANSI) material speciﬁcations will also contain such information.

2.8 Quality Assurance

Quality assurance may be considered to be the ﬂip side of material speciﬁcations,

and so the types of data and the data sources themselves required for the two

functions are essentially the same.

Purchasers of materials, for example, may chose to do their own testing of

the materials once delivered to their facilities by materials producers. If so, they

will use exactly the same tests and refer to exactly the same data sources to

determine compliance. The one difference may be that such purchasers may

chose to gradually accumulate the results of such tests and build their own

databases for internal use, not only by their quality control experts, but also by

their designers and materials experts who must establish safe levels of perform-

ance of the structures. These types of databases also tend to be proprietary, of

course, and are seldom made available to the outside world, especially compet-

itors.

2.9 Maintenance

The principal value of material data in connection with maintenance concerns

will likely be for reference purposes when problems show up with either dete-

rioration of surface conditions or the suspicion of the development of fatigue

cracks at local stress raisers.

In both cases, the important features of the types of data desired to address

such issues are more likely to be those based upon individual exposures or prior

service experience, and so the user may be more concerned about the degree of

applicability than upon data quality and statistical reliability (though both fea-

tures would be desired if available). Typically such data are hard to ﬁnd in any

event, and once again are more likely to be buried in proprietary ﬁles than in

published databases.

In the case of engineers and technicians needing databases comprised of ser-

vice experience, they may well be faced with building their own data sources

based upon their organization’s production and service experience than expecting

to be able to locate applicable external sources.

3 TYPES OF DATA 463

2.10 Failure Analysis

The occurrence of unexpected failure of components in a structure usually calls

for some follow-up study to determine the cause and possible ways to avoid

further loss. In such cases it is inevitable that such failure analysis will involve

both (a) a review of the old databases used to design the part and (b) a search

for or the development of new data sources that may shed more light on the

material’s response to conditions that developed during the life of the structure

that had not been anticipated beforehand.

The types of databases sought in this case will likely be those containing

statistically reliable data, but recognizing the unexpected nature of some prob-

lems, an interest in a wider range of data sources and a willingness to consider

a lower level of data quality may result. Databases for failure analysis studies

may need to be wider in scope and to cover subjects such as corrosion that are

not always easily treated by statistical means. In fact, sources covering failure

experience may be the most valuable, though hardest to ﬁnd because historically

engineers and scientists do not publish much detail about their mistakes.

The net result is that when dealing with failure analysis, the search may be

quite broad in terms of data quality, and the focus most likely will be more on

applicability to the problem than on the quality and structure of the compilation.

As in the case of maintenance engineers and technicians needing databases com-

prised of service experience, failure analysts may well be faced with building

new data sources based upon their organization’s production and service expe-

rience.

3 TYPES OF DATA

It is useful at this stage to note that there are several basic types of materials

databases, i.e., databases containing signiﬁcantly different types of information

and, hence, different data formats. Note that this is different from the type of

platform or presentation format (e.g., hardcopy, CD, online, etc.); these will be

discussed in Section 6.

The two fundamental types of databases discussed here are textual data and

numeric data. In fact, many databases represent a combination of both types,

but there are some basic differences worth noting as in the next two paragraphs.

The concept of metadata will also be described in more detail in this section.

3.1 Textual Data

The terminology textual data is generally applied to data entries that are purely

alphabetic in nature with numbers used only as necessary to complete the

thoughts. Textual databases are typically searched with alphabetical strings, e.g.,

by searching for all occurrences of a term such as aluminum or metallurgy or

whatever subject is of interest.

The subjects of textual databases are predominantly bibliographies and ab-

stracts of publications, but they may reﬂect other speciﬁc subjects such as textual

descriptions of failures of components, for example. Bibliographic databases

seldom reﬂect the ﬁnal answers to whatever inquiry is in mind but rather ref-

erences to sources where the answers may be found.