Aerospace Structural Metals Handbook - Maechanical Properties

March 2002 (revision) Aerospace Structural Metals Handbook General Discussionof Handbook Contents

Author W. F. Brown, Jr.

Page 1© 2002 by Purdue Research Foundation, West Lafayette, Indiana 47907. All Rights Reserved.U.S. Government License. This material may be used, duplicated or disclosed by United StatesGovernment agencies without the payment of any royalty.

IntroductionThis section introduces the reader to the general objective,contents, and format of this handbook. Definitions of thephysical and mechanical properties are presented as well aspertinent comments on the selected data.The general objective is to provide an unbiased characteriza-tion of alloys of interest to the aerospace designer so that al-loy selection for a particular application may be made with aknowledge not only of typical physical and mechanicalproperties, but especially of those factors which limit theload carrying capacity, whether associated with the serviceconditions or with the metallurgical history of the alloy. Thishandbook contains typical data as contrasted to statisticallyderived design data presented in MIL-HDBK-5 (Ref. 1).However, specified mechanical properties are frequentlypresented as derived from AMS, federal, and industrialsources and specifications. Caution should be used whencomparing data from different sources because mechanicalproperties are sensitive to processing history and testingtechniques. Where possible data in this Handbook are char-acterized as to source, product form, size and pertinent de-tails relating to those processing steps and testing techniqueswhich affect the mechanical or physical properties. How-ever, in some cases such information is not available fromthe published literature and contact with the authors is im-possible or impractical. In such cases judgment is made as tothe worth of the data as useful qualitative information. Anattempt is made to provide an organized record of impor-tant mechanical and selected physical properties. However,it should be realized that a vast amount of information relat-ing to mechanical and physical properties of alloys used inthe aerospace and related industries has been and is still be-ing generated. Some of this is confined to specific pieces ofhardware and/or applications and has limited general use-fulness (e.g., flight spectrum fatigue loading behavior). Thistype of information is not generally included in this hand-book. Also in the case of alloys such as 4340 steel, Ti-6Al-4V,and 2024Al which have been widely used for many yearsand where design allowables are well established, it is notpractical to make a comprehensive record here. Extensivedata bases for such alloys exist in industrial and governmentorganizations. For such well characterized alloys this hand-book presents an overview with emphasis on those factorslimiting local carrying capacity. With respect to chapter revi-sion and deletion, all chapters are identified by their date ofpreparation. While it has always been the desire of the edi-tors to update chapters on some regular basis, limited fundshave forced concentration on new materials and have re-stricted revisions to alloys whose characterization is sub-stantially changed by new information. It should be notedthat the AMS, ASTM and other types of specifications maynot be current and while provided general guidance shouldbe checked against their sources. With respect to chapter re-vision or deletion, no special problem exists when an alloyhas been out of production for several years. In such a casethe alloy is deleted from the Handbook. A special problemoccurs when deciding what data to delete when a Chapter is

revised. This decision is easy to make when there has been aclear change in the composition or the processing practice.On the other hand, some alloys have a rather wide range ofaccepted compositions and are processed in several wayswhich have evolved over a period of years. Under these cir-cumstances data are not deleted during a chapter revisionunless evidence has been presented that it was, for some rea-son, obtained from faulty test or analysis procedures. Com-ments on the significance of data are made on the basis ofthe most recent information in a particular chapter.

Alloy ClassificationAlloys included in the Handbook can be generally classifiedas high strength materials as commonly used in load carryingmembers of aircraft and spacecraft and fit the classificationused in the Handbook for many years. However, in a fewapplications, (e.g. exhaust nozzles for high-performance en-gines) relatively low strength alloys may be used and thesewill be grouped under the heading of Other Alloys. In all casesthe Chapters are classified according to specific alloy categoriesand each category is assigned a code number. Chapters arethen numbered serially under their respective codes as theyare added to the handbook. The chapter code numbers appearat the bottom of each page along with the page number. Anoutline of the categories and their numbers is given below.

Ferrous AlloysCategory Code SeriesCarbon and Low Alloy Steels (FeC) 1100Ultra High Strength Steels (FeUH) 1200Austenitic Stainless Steels (FeA) 1300Martensitic Stainless Steels (FeM) 1400Age Hardening Steels (FeAH) 1500Nickel Chromium Steels (FeNC) 1600

Non-Ferrous AlloysCategory Code SeriesAluminum Alloys; Cast (AlC) 3100Aluminum Alloys; Wrought, HeatTreatable (AlWT) 3200Aluminum Alloys; Wrought, Not HeatTreatable (AlWN) 3300Magnesium Alloys; Cast (MgC) 3400Magnesium Alloys; Wrought, HeatTreatable (MgWT) 3500Magnesium Alloys; Wrought, Not HeatTreatable (MgWN) 3600Titanium Alloys (TiW) 3700Titanium Alloys Cast (TiC) 3800Nickel Base Alloys (<5% Co) (Ni) 4100Nickel Base Alloys (>5% Co) (NiCo) 4200Cobalt Base Alloys (Co) 4300Beryllium Alloys (Be) 5100Columbium (Niobium) Alloys (Cb) 5200Molybdenum Alloys (Mo) 5300Tantalum Alloys (Ta) 5400Tungsten Alloys (W) 5500Vanadium Alloys (V) 5600Zirconium Alloys (Zr) 5700Other Alloys 6000

Page 2

General Discussion Aerospace Structural Metals Handbook March 2002 (revision)Handbook Contents

Alloy IdentificationAlloys are identified by their chemical composition. The ma-jor elements are listed in decreasing order of their percent-age by weight followed by the minor elements listed in thesame way. When two or more elements are present inequal quantities they are listed alphabetically. Minor ele-ments are given only when they are intentionally added toyield special mechanical properties. Impurities are ne-glected. The only exception to this practice is for ferriticsteels; when carbon is the element having the predominateinfluence on the mechanical properties, it is listed immedi-ately after iron. In addition to the above described systemof identification, the most commonly used designations(e.g., AISI Code) are given along with the number from theSAE/ASTM Unified Number System (Ref. 2). To facilitatethe location of a particular alloy an index is provided in theprevious section which gives cross-references for alloy cat-egory codes and commonly used designations.

Measurement SystemThe data presented in this handbook are obtained fromsources in various countries. The original units of data mea-surement vary depending on the source. Where practical,mechanical property data are presented in English units(inch, pounds force, seconds, Fahrenheit) which are custom-arily used by the aerospace industry. Physical property dataare generally given in the units used by the source. To assistin converting SI and other units to the customary units aconversion table is presented in Appendix D.

Data OrganizationThe data for each alloy are presented according to a prop-erty code system designed for this handbook. The numbersof the tables and figures that follow each section correspondto their respective paragraphs in the text. An outline of thiscode follows. (Please note also that the numbering systemused in the earlier handbooks has been altered. For ex-ample, 1.01 is now 1.1, 2011 is 2.1.1, and so forth.)

1.0 GENERAL1.1 Commercial Designation1.2 Other Designations1.3 Specifications1.4 Composition1.5 Heat Treatment and Microstructure1.6 Hardness1.7 Forms and Conditions Available1.8 Melting and Casting Practice1.9 Hydrogen Embrittlement Resistance

2.0 PHYSICAL PROPERTIES AND ENVIRONMENTALEFFECTS

2.1 Thermal Properties2.1.1 Melting Range2.1.2 Phase Changes2.1.3 Thermal Conductivity

2.1.4 Thermal Expansion2.1.5 Specific Heat2.1.6 Thermal Diffusivity2.2 Other Physical Properties2.2.1 Density2.2.2 Electrical Properties2.2.3 Magnetic Properties2.2.4 Emittance2.2.5 Damping Capacity2.3 Chemical Environments2.3.1 General Corrosion2.3.2 Stress Corrosion2.3.3 Oxidation Resistance2.3.4 Hydrogen Resistance2.4 Nuclear Environment

3.0 MECHANICAL PROPERTIES3.1 Specified Mechanical Properties3.2 Mechanical Properties at Room Temperature3.2.1 Tension Stress-Strain Diagrams and Tensile

Properties3.2.2 Compression Stress-Strain Diagrams and

Compression Properties3.3.3 Impact3.3.4 Bending3.3.5 Torsion and Shear3.3.6 Bearing3.3.7 Stress Concentration3.2.7.1 Notch Properties3.2.7.2 Fracture Toughness3.3 Mechanical Properties at Various Temperatures3.3.1 Tensile Properties3.3.2 Compression Stress-Strain Diagrams and

Compression Properties3.3.3 Impact3.3.4 Bending3.3.5 Torsion and Shear3.3.6 Bearing3.3.7 Stress Concentration3.3.7.1 Notch Properties3.3.7.2 Fracture Toughness3.3.8 Combined Loading3.4 Creep and Creep Rupture Properties3.5 Fatigue Properties3.5.1 High-Cycle Fatigue3.5.2 Low-cycle Fatigue3.6 Elastic properties3.6.1 Poisson’s Ratio3.6.2 Modulus of Elasticity3.6.3 Modulus of Rigidity3.6.4 Tangent Modulus3.6.5 Secant Modulus

4.0 FABRICATION4.1 Forming4.2 Machining and Grinding4.3 Joining4.4 Surface Treating

January1992 (revision) Aerospace Structural Metals Handbook General Discussionof Handbook Contents

Page 3

Data PresentationData points are generally shown for physical and mechanicalproperties. Multiple data points for a particular value of theindependent variable are averaged to establish the trendcurves. Where large data scatter is encountered or where avery large number of data points are reported (e.g., fatiguecrack propagation data) a scatter band is shown to representthe upper and lower bounds of the data. However, such bandsshould not be used to represent lower bound properties fordesign purposes.

The mechanical properties and, to a lesser extent, the physicalproperties of metallic alloys are strongly dependent on theprocessing history of the test material. Where possible themost important features of processing are identified andreported in the data representations. Melting practice, prod-uct form, heat treatment, cold work, joining conditions, andrelated information (such as the testing direction and thespecimen location with respect to the surfaces of the part fromwhich they were removed) are included. Unfortunately,such complete characterizations are not available in manycases and, occasionally, so little information of this type isavailable that the data in question are excluded.

1 GeneralThis section contains a brief general description of theselected alloy including a listing of any different grades,special features of interest to the structural designerrelating to the alloy’s advantages and limitations, andits typical applications.

1.1 Commercial DesignationThe most commonly used designation is given first.This may originate with the alloy producer or fromwidely used specifications such as the Society of Auto-motive Engineers - Aerospace Materials Specifications(SAE-AMS). Trade names are generally not used, butin a few cases no other satisfactory designation isavailable. The reader should be aware of the rulesgoverning the use of trade names.

1.2 Alternate DesignationsThe alternate designations include proprietary namesand others encountered in the literature. No attemptis made to be complete here. A list of alloy namesused in the handbook is given in the Index. For fur-ther information, see Ref. 3.

1.3 SpecificationsThis section provides a listing of specifications alongwith a description of the product forms and relatedmetallurgical conditions. The primary sources are theAerospace Material Specifications (AMS) issued by theSAE and the ASTM. Military and federal specificationsmay also be given. In some cases the alloy user maydevelop a specification for a particular application

that also may be reported if it assists in the propercharacterization of the data.

1.4 CompositionThe SAE-AMS specifications are the primary sourceof reported chemical composition limits. In caseswhere an AMS specification has not been published thecomposition limits given by the producer are reported.It should be realized that for some materials a rela-tively small variation in composition limits for certainelements may have a substantial influence on theproperties. These effects may be discovered as appli-cation experience is accumulated. Important composi-tional influences are discussed in Section 1.9, SpecialConsiderations.

1.5 Heat TreatmentThis section defines and discusses the thermal and/ordeformation process applied to the alloy to achievedesired properties. A given heat treatment may bedesignated by either the process or the result. Becausea clear understanding of heat treating terms is necessaryto properly use material presented in this section, aGlossary of Heat Treating Terms is given in Appendix B.

1.6 HardnessHardness, as used herein, refers to an alloy’s resistanceto permanent deformation by indentation. Several in-dentation methods of measuring hardness have beenstandardized by the ASTM, namely Brinell E 10 (Ref.4), Rockwell E 18 (Ref. 5), Vickers E 92 (Ref. 6) andMicrohardness E 384 (Ref. 7). Hardness values reportedin this handbook relate to one of these standardizedprocedures. In general, the hardness is related to thealloy tensile strength; hence, hardness values are oftenused to delineate the response of an alloy to heattreatment or cold work. This section reports the effectof various variables on the hardness. The depth to whichan alloy will harden as a result of heat treatment is animportant consideration in structural applications. Theload carrying capacity of a component will be reducedif the desired hardness cannot be achieved through-out the cross section. The property that relates to thedepth of hardening is called the hardenability. It isstrongly dependent on the composition and, to a lesserextent, on the details of heat treatment. In the case oftitanium alloys and steels, the higher the hardenabilitythe less rapid the cooling rate from the solution oraustenitizing temperature that is required to produce agiven hardness level at the center of the cross section.In steels, insufficient hardenability or too slow aquenching rate can produce undesirable transformationproducts in the microstructure which may embrittle thesteel. Hardenability is measured in several differentways. For steels the end quench hardenability test isfrequently employed. This test, ASTM Standard A 255(Ref. 8), consists of water quenching one end face of asteel cylinder and then determining the distribution of

General Discussion Aerospace Structural Metals Handbook January 1992 (revision)of Handbook Contents

Page 4

hardness along the axis of the cylinder. More completeinformation is obtained by quenching cylinders ofvarious diameters and measuring the hardness distri-bution on a diameter. In some cases the depth of hard-ening is judged by abstracting tensile specimens fromdifferent positions in the cross section or from crosssections of different sizes.

1.7 Forms and Conditions AvailableThe availability of a particular form will change withthe market conditions. Current information on avail-ability of forms and conditions should be obtainedfrom the alloy producer.

1.8 Melting and Casting PracticeMelting practice can have a strong influence on themechanical properties when these are sensitive toimpurities. In such cases the details of melting practiceare given in this section. Many of the alloys includedin this handbook can be cast. However, only a limitednumber have cast properties which are of interest tothe aerospace and related industries. The productionof high quality castings requires very careful control ofpouring and gating techniques as well as cooling ratesin various portions of the cast structure. This sectionwill contain detailed information on such castingtechniques for those alloys where high quality castingsare of interest to the designer.

1.9 Special ConsiderationsThis section summarizes problems which may beencountered in the production of alloys as well as intheir application. Emphasis is placed on those factorswhich limit the load carrying capacity in intendedapplications. Included are such factors as hydrogenembrittlement introduced by plating or cleaning pro-cesses or resulting from exposure to high pressurehydrogen gas. Where appropriate this section mayreference other sections for further information. Inaddition, this section may include comments on par-ticular advantages possessed by an alloy.

2 Physical Properties and EnvironmentalEffectsA concise presentation of those physical propertiesof importance to the designer is given rather than acomprehensive compilation. When values of physicalproperties are reported without a specified test tem-perature, the values are for room temperature only. Insome cases, the determination of the property requiresmeasurements at two or more temperatures (e.g.,thermal expansion). These temperatures, unless other-wise indicated, are in the range from room temperatureto 212F. Usually the difference in property values forany two temperatures in this range is less than theuncertainty in the values themselves. Frequently thesource material does not give the form and the condition

of the alloy for which the physical property determina-tion was made. The influence of form and conditionmay explain some of the differences between publishedvalues for the same alloy. For detailed information onphysical properties (excluding magnetic properties)the reader is referred to CINDAS/Purdue University,West Lafayette, Indiana 47906.

Environmental effects include the influence of variouschemical agents on the mechanical properties and onthe general corrosion behavior. The influence ofnuclear environments is also covered here. Particularattention is given to the environmental factors thataffect the crack propagation characteristics of an alloy.

2.1 Thermal PropertiesThermal properties include the melting range, phasechanges, thermal conductivity, thermal expansion,specific heat, and, where available, thermal diffusivityand dimensional changes on heat treating.

2.1.1 Melting Range. The melting range is bound by thesolidus temperature on the low end and the liquidustemperature on the high end. These two temperaturesare influenced by compositional variations withinspecified limits. The range reported is generally estab-lished by experience with the solidus being set lowenough to avoid incipient melting during hot workingwhich could damage the alloy.

2.1.2 Phase Changes. A phase is any physically discretehomogeneous portion of an alloy system. In somecases a phase change refers to a change in the crystalstructure of the major alloying element. This can beobserved in low alloy steels where face centered cubicaustenite transforms to a body centered cubic structureon cooling below a certain critical temperature. How-ever, precipitations of intermetallic compounds fromsolid solution also constitute phase changes. Phasechanges may be caused or accelerated by cold work.Phase changes may be beneficial or detrimental withrespect to the load carrying capacity of a part. For ex-ample, creep resistance may be increased by precipi-tations occurring during heat treatment or during thecreep process. On the other hand, when quenchingsteels from the austenitizing temperature, too slow acooling rate (slack quenching) may produce embrittlingconstituents in the microstructure. In welded compo-nents, phase changes may result in dimensional changeswhich can give rise to residual stresses. A detaileddiscussion of phase changes and transformationstructures may be found in Ref. 9.

2.1.2.1 Time-temperature transformation diagrams.These are also referred to as isothermal trans-formation diagrams or T-T-T diagrams andenable the user to estimate how an alloy willrespond to cooling through a range of tempera-tures where phase transformations take place.Where appropriate, continuous cooling trans-formation diagrams are shown.


Page 5

2.1.3 Thermal Conductivity. This quantity, more properlycalled the coefficient of thermal conductivity K, propor-tions the heat flow rate H over a conducting distance xto the area of the conductor A and the temperaturedifference ∆T over the conducting distance. K is givenby the following expression:

K = HxADT

A variety of units is used to report the thermal conduc-tivity in the literature. This handbook uses Btu-ft per(hr sq ft F). The Btu is defined in several different ways;however, for the present purposes, the difference inthermal conductivity depending on the definition isnegligible.

2.1.4 Thermal Expansion. This section reports the meancoefficient of linear expansion defined as

where α is the coefficient and ∆L the change in originallength L due to a temperature ∆T = T - To. In thishandbook room temperature is used as the referencetemperature To. Units for the coefficient are in. per in.per F.

2.1.5 Specific Heat. The specific heat at a specified tempera-ture is the amount of heat necessary to raise a unit massof a substance one degree in temperature divided bythat necessary to raise a unit mass of water one degreeat the same temperature. For engineering purposesthe specific heat of metallic alloys may be taken equalto their heat capacity. The heat capacity of alloys willvary with temperature. Unless the reference tempera-ture is specified the values of specific heat given inthis handbook refer to room temperature.

2.1.6 Thermal Diffusivity. This quantity is the thermal con-ductivity divided by the density and the heat capacityat constant pressure. The units used here are sq ft per hr.

2.2 Other Physical Properties

2.2.1 Density. The weight density is reported as either lbsper cu. in. or as gms per cu. cm.

2.2.2 Electrical Properties. This section reports the electricalresistivity on a volume basis (sometimes referred to asthe “volume resistivity”), which proportions the elec-trical resistance to the dimensions of the conductoraccording to the expression

R =ρ1A

where R is the resistance, ρ the resistivity, l the length,and A the cross sectional area of the conductor. Avariety of units is used for the volume resistivity, the

most common being microhm-inch. Conversion factorsfor several units are given in Appendix D.

2.2.3 Magnetic Properties. Materials are classified as para-magnetic, ferromagnetic, or diamagnetic. Bars ofparamagnetic and ferromagnetic materials align theirlong dimensions in the flux direction in a magneticfield, while diamagnetic materials align their longdimensions normal to the field. For the purposes ofelectrical engineering, ferromagnetic materials are ofthe greatest interest. A number of magnetic propertiesfor ferromagnetic materials can be derived from thenormal induction curve or from the normal hysteresisloop. Standard methods for determining these curvesare given in ASTM A 34 (Ref. 10).

In non-magnetic media any change in the magnetizingforce H measured in oersteds is accompanied by anequal change in magnetic induction B measured ingauss. However, in ferromagnetic materials the relationbetween H and B is non-linear and a plot of B vs H isknown as the normal induction curve. This is illus-trated by the heavy line in Figure 2.2.3.1. Increasing Hbeyond a certain value produces only a small furtherincrease in B. The limiting value of B is designated asthe saturation induction Bs. The ratio of B to H at anypoint on the normal induction curve is known as thenormal permeability µ.

The units of permeability are gauss per oersted. How-ever, it is usual to report a dimensionless quantity whichis the ratio of the normal permeability described aboveto that of a vacuum which is unity in the cgs systembut 4π x 10-7 in the SI system. This number is knownas the relative permeability µr. The normal permeabilityhas a maximum value µm near the knee of the inductioncurve and is represented by the slope of a line passingthrough the origin and tangent to the induction curve.The slope of the induction curve at the origin is knownas the initial permeability µo. Other values of perme-ability are sometimes reported being representative ofthe B/H ratio at some particular value of the magne-tizing force. It should be noted that the permeability ofparamagnetic materials is slightly greater than unity,essentially independent of the magnetizing force, andthat diamagnetic materials have permeabilities slightlyless than unity. For diamagnetic materials it is commonto report the susceptibility K. The relation between µand K in the cgs system is

µ = 1+4πK

Referring to Figure 2.2.3.1, if the value of H is reducedfrom any point on the induction curve, hysteresis isevident and it is necessary to reverse the magnetizingforce to produce a condition of zero induction. Theresult is a hysteresis loop. There are two quantities ofspecial interest that can be derived from the hysteresisloop, namely, the residual induction Br remaining afterH has been reduced to zero and Hc which is the nega-tive value of H necessary to reduce B to zero. If the

∆L 1α = ___ ___ Lo ∆T

Rev. 7/92


Page 6

normal hysteresis loop is representative of a conditionwhere the peak magnetizing force was sufficiently highto produce saturation induction (as in Figure 2.2.3.1),the Br is designated as the retentivity Brs and Hc isdesignated as the coercivity Hcs. The details of magneticmeasurements are discussed in Refs. 11 and 12.

2.2.3.1 [Figure] Normal induction curve and hysteresisloop for ferromagnetic materials.

2.2.4 Emittance. The hemispherical emittance is the ratio ofthe radiance of a body to the radiance of a black bodyat the same temperature. The radiance is the rate ofradiant energy emission from a unit area of the sourcein all radial directions of the overspreading hemisphere.If measurements are made normal to the surface, thequantity determined is called the normal emittance.When the measurements are made at a specific wavelength, the result is designated as spectral emittance.Emittance is a dimensionless quantity with valuesless than unity. For further information the reader isreferred to Ref. 13.

2.2.5 Damping Capacity. If the excitation is removed froma metal specimen subjected to torsional or bendingvibrations the amplitude will decay due to internalfriction which results in the dissipation of energy. Thehigher the damping the larger the rate of amplitudedecay. Various measures have been proposed fordamping and these are reviewed in Ref. 14. One that isfrequently used in engineering is the damping capac-ity or specific damping energy measured as energydissipated per unit volume per cycle.

2.3 Chemical EnvironmentsThis section includes information concerning theresistance of an alloy to various environments whichcan degrade its load carrying capacity by chemicalreaction with its surface or interior. While a largenumber of corrosion tests have been standardizednone of these give information which can directly berelated to the degradation in load carrying capacity ofa structural component. However, valuable compara-tive information can be obtained by selecting testswhich represent the service environment. Corrosiontests can be divided into two generally recognizedcategories, namely, general corrosion and stress corro-sion. This distinction is somewhat blurred by the influ-ence of residual stresses which can affect the resultsof a general corrosion test. When corrosion data arereported, reference is made to an applicable ASTM orother standard for the test method and data evalua-tion. However, as pointed out in ASTM G 31 (Ref. 15),it is not possible to rigorously standardize corrosiontesting because of the great variety of factors, many ofwhich can not be well controlled.

2.3.1 General Corrosion. Smooth specimens are exposed toa variety of environments for various lengths of time.For example, specimens may be exposed to a seacoast

or an industrial atmosphere, to a salt spray, to con-tinuous immersion, or to alternate immersion in a saltsolution (e.g., 3.5% NaCl). Tests of this type are com-monly applied to aluminum and magnesium alloysand to high strength steels. The specimen surface isperiodically examined for evidence of pitting and/orexfoliation and the observed surface conditions maybe graded according to photographs contained inappropriate standards. In some cases, weight loss isdetermined and the result expressed as a time rate, orthe result is converted to a dimensional loss rate (e.g.,mils per year). Aluminum and magnesium alloys arequite susceptible to increased corrosion when electri-cally coupled in a corrosive media to a more noblealloy (e.g., Mg to steel in a salt solution). This galvaniccorrosion is dependent on many variables and is nottreated in this handbook. Some guidelines on predict-ing the behavior of galvanic coupled metals are givenin ASTM G 82 (Ref. 16).

Some alloys (e.g., stainless steels and Ni-rich Cr-bearinghigh temperature alloys) when improperly heat treatedare susceptible to intergranular corrosion. Acceleratedtest methods for evaluating the susceptibility of suchalloys have been developed. The results may beexpressed in terms of weight change or by bendingthe specimen and examining the tension surface forcracks related to intergranular attack.

2.3.2 Stress Corrosion. Most alloys are rendered more sus-ceptible to corrosive attack if subjected to a stress inthe aggressive environment. Stress corrosion tests aredesigned to measure the resistance to this type of be-havior. A large number of tests have been standard-ized which make use of smooth specimens generallysubjected to some fraction of the yield strength (e.g.,tension specimens ASTM G 49 [Ref. 17] and C ringspecimens ASTM G 38 [Ref. 18]). Specimens may becontinuously exposed or alternately exposed to theenvironment and then to air. After an incubationtime, cracks frequently form due to the adsorption ofand diffusion of hydrogen generated by the corrosionreaction. Specimens are examined periodically todetermine the presence of stress corrosion cracks. Intests using tension loading, specimens often fracturecompletely after crack initiation. Failure in tests usingbending loads must be established by careful exami-nation of the surface. The results of stress corrosiontests exhibit large scatter due to the difficulty of estab-lishing the time to first cracking, premature fractureresulting from loss of load carrying area associatedwith general corrosion, and the fact that the cracksrepresent the behavior of a very localized region of thespecimen. Frequently specimens subjected to tensileloading are removed from the aggressive environmentand fractured to determine the retained load carryingcapacity. This type of data has qualitative value inestablishing the degeneration of load carrying capacitydue to the aggressive environment.


Page 7

To establish the structural load carrying capacity of amaterial, fracture mechanics tests can be used to ob-tain a measure of the effect of an aggressive environ-ment. Sustained load tests on cracked specimens areused to establish a value of applied stress intensityfactor K below which cracks will not grow in a particu-lar aggressive environment. This value is known asKIscc and may be well below KIc. (K and KIc are definedin ASTM E 399 [Ref. 19].) At present there are no stan-dardized tests for KIscc; when such values are reportedherein the test conditions are described. If the applica-tion involves alternating loads, the K threshold belowwhich cracks will not grow in an aggressive environ-ment may be below KIscc.

2.4 Nuclear EnvironmentsThis section includes information concerning the resis-tance of metallic alloys to neutron radiation such asmay be encountered in power generating reactors.Neutron radiation can result in large changes inmechanical properties and it is these changes that areemphasized in this section. The mechanical propertiesare represented as a function of the integrated neutronflux multiplied by the exposure time. This product iscalled the fluence, often designated as nvt. The neutronenergy is also reported in millions of electron volts(Mev).

3 Mechanical PropertiesThe properties treated in these sections include thoseconventionally used in design as well as propertiesderived from special tests such as those from notch orprecracked specimens. Information on mechanicalproperties is given concerning the effects of composi-tion, heat treatment, and processing methods. Emphasisis on commercial practice. However, results obtainedfrom experimental procedures are reported wherethese contribute to a better understanding of the effectsof the variables influencing the mechanical properties.

While it is general practice to put all mechanical prop-erties under these sections, where appropriate, mechani-cal properties may also be located in Section 1.5 onHeat Treatment, Section 1.9 on Special Considerations,and in Section 4 on Fabrication.

3.1 Specified Mechanical PropertiesThe specified mechanical properties reported in thissection should not be taken as design values. Designvalues are given in MIL-HDBK-5 and the readershould obtain that handbook as a companion volume.It should be remembered that specifications are some-times subject to frequent change, especially in the caseof new alloys.

Room temperature mechanical properties form thecore of most specifications. However, for heat resis-tant alloys, elevated temperature mechanical proper-

ties are sometimes specified. Specifications generallygive minimum values of strength and ductility prop-erties although occasionally maximum values aregiven. Forming limits in terms of minimum bend radiiare specified in certain cases but this type of informa-tion is usually not included in the handbook.

3.2 Mechanical Properties at Room TemperatureThis section includes data on conventional mechanicalproperties determined at room temperature with theexception of creep, fatigue, and elastic propertieswhich are covered in separate sections. Used in thepresent sense, room temperature is a common butloose way of defining a test temperature. In this hand-book it refers to the temperature range between 60and 90F. In this range there is negligible time depen-dent flow (creep) for the great majority of structuralalloys. Room temperature data will also be found inSection 3.3 where it appears as one of a series of valuesat various temperatures.

3.2.1 Tension Stress-strain Diagrams and Tensile Properties.Unless otherwise stated, stress is based on the originalcross-section area. Elongation and reduction of arearefer to percent changes from an original dimension.Occasionally, true stress and true strain are reportedwhere the stress and strains are related to instantaneousvalues. Stress-strain curves generally encompass therange between zero and one percent strain. Suchcurves are useful in determining the static modulusand the yield strength. The tensile strength Ftu is thestress at maximum load. The yield strength Fty, some-times called proof stress, is the stress at a specifiedamount of plastic strain. In the United States this isgenerally 0.2 percent. The deformation values, elonga-tion e and the reduction of area RA, are also reported.A standard method for measuring these strength anddeformation values at room temperature is given inASTM E 8 (Ref. 20). The tensile strength of brittlematerials is influenced by any bending stresses intro-duced by the loading and gripping devices. Thesebending stresses may vary from test to test and resultin substantial scatter of the strength values (ASTM E1012, Ref. 21). It should be appreciated that the elon-gation of an alloy exhibiting necking before fracturemay vary with the gage length. Neither elongationnor reduction of area are satisfactory predictors offracture toughness (Ref. 22).

3.2.2 Compression Stress-strain Diagrams and CompressionProperties. As in tension the stress and strain refer tothe original dimension of the specimen and the com-pression yield strength Fcy is analogous to the tensileyield strength. Compression loads are influenced byfriction at the interface between the anvils of the testmachine and the specimen ends. Various means havebeen used to overcome these effects (e.g., large lengthto diameter ratios and/or lubrication; see ASTM E 9,Ref. 23). Unfortunately, information on specimen di-


Page 8

mensions and means to minimize friction are seldomreported with compression data. In the absence offrictional effects the compression yield strength formost materials is essentially equal to the tensile yieldstrength.

3.2.3 Impact, see Section 3.3.3.

3.2.4 Bending. This section reports data from slow (as dis-tinct from impact) bend tests on smooth specimens.These types of tests are sometimes used to evaluatethe ability of welds to resist cracking at high deforma-tions (ASTM E 290, Ref. 24). They are also used todetermine the modulus of elasticity and yield strengthin bending (ASTM E 855, Ref. 25). The yield strengthso determined may not be equal to the yield strengthin pure tension. The strength and deformation capac-ity of very brittle materials are frequently evaluated inthree point bend tests commonly called modulus ofrupture tests (ASTM B 406, Ref. 26). These tests havelimited application in evaluation of aerospace alloys.

3.2.5 Torsion and Shear. The property reported here is theultimate strength in shear (Fsu) as determined from atorsion test or from one of a number of special sheartests.

3.2.6 Bearing. Bearing tests measure the yield (Fbry) andultimate strength (Fbru) of a material when subjected tobearing loads. In a pinned or riveted joint the effectivebearing area is the product of the hole diameter andthe thickness of the bearing member. Specimen geom-etry must be considered when evaluating the resultsfrom these tests. Pin type bearing tests of metallicmaterials have been standardized and are describedin ASTM E 238 (Ref. 27).

3.2.7 Stress Concentration, see Section 3.3.7.

3.2.7.1 Notch properties, see Section 3.3.7.1.

3.2.7.2 Fracture toughness, see Section 3.3.7.2.

3.2.8 Combined Loading (formerly Combined Properties).This section reports data from tests involving combi-nations of static loads such as combined bending andtension or multiaxial stress tests. Tests of this naturehave not been standardized and only a limited amountof data from such tests is reported in this handbook.

3.3 Mechanical Properties at Various Temperatures(see also Section 3.2).This section includes only the so-called short-timeproperties determined in conventional tests. The rateof deformation can influence the strength values attemperatures where time dependent flow occurs. Atsufficiently high temperatures the time necessary tostabilize the specimen temperature may be sufficientlylong to cause metallurgical changes in the material.For these reasons the scatter observed in elevatedtemperature mechanical properties is sometimes con-siderably larger than in the room or low temperatureproperties. At cryogenic temperatures some alloys

become quite brittle and may break before the yieldstrength is reached. Under such circumstances theultimate strength values may exhibit considerablescatter.

3.3.1 Tension Stress-strain Diagrams and Tensile Properties,see also Section 3.2.1. Methods recommended for con-ducting high temperature tensile tests are describedin ASTM E21 (Ref. 28). This recommended practiceemphasizes the importance of loading rate effectswhich depend on the testing techniques used.

3.3.2 Compression Stress-strain Diagrams and CompressionProperties, see also Section 3.2.2. Compression proper-ties, like tension properties, are sensitive to loading andheating rate effects at sufficiently high temperatures.ASTM E 209 (Ref. 29) gives recommended practices forconducting compression tests at elevated temperaturesfor conventional or rapid heating rates and strain rates.

3.3.3 Impact. Impact tests provide a measure of toughnessin terms of an alloy’s resistance to fracture when theload is applied at relatively high rates. Such tests donot directly provide a result that can be used in deter-mining structural load carrying capacity. However,they are useful for ranking alloys in terms of theirresistance to impact loading and in some cases mayprovide safe temperature limits for avoiding brittlefracture. Various types of impact tests have beendescribed in the literature and several have been stan-dardized. This handbook will report data only fromstandard test methods (Refs. 30, 31, 32, 33). Thesemethods are listed in Table 3.3.3.1 which identifies thetest method, the specimen type, and the indices oftoughness reported. The standardized impact testsfall into two general classes, namely, those employingsmall specimens (Charpy and Izod) and those employ-ing large specimens (NDT, Drop Weight Tear, andDynamic Tear). The Charpy and Izod tests have beenused for many years in the evaluation of a variety ofalloys. The most widely used Charpy specimen has aV notch with a 0.01 in. radius. Charpy specimenswithout a notch and those with mild notches (Key-hole and U) have limited applications to special situa-tions where the relatively sharp V notch is consideredundesirable. The large specimens are designed tobring the test conditions more in line with structuralapplications. In general, the larger the specimen andthe sharper the notch the greater is the sensitivity tofactors producing brittle behavior. Consequently, theresults of impact tests from different types of specimensshould not be co-mingled. The trend of the impactenergy with test temperature is strongly dependenton crystallographic structure, composition, and heattreatment. Body center cubic structures are particu-larly sensitive to temperature. Ferritic steels and betatitanium alloys generally show a marked decrease inimpact energy within a relatively narrow temperaturerange. The center temperature of this range is thetransition temperature. At temperatures outside this


Page 9

range the impact energy tends to level out. At tem-peratures above the transition range the nearly levelvalue of impact energy is known as the upper shelfvalue and has been given special significance in defin-ing tough behavior. At sufficiently high temperaturesthe impact energy will decrease due to loss of strength.At subtransition temperatures the impact energy mayfall to a few ft-lbs indicating very brittle behavior.Heat treated high strength steels do not exhibit strongtransitional behavior and the impact energy is a lessuseful quantity than for the ferritic steels. The impactenergy values may be supplemented by measurementof the lateral expansion and/or percent shear. Lateralexpansion refers to a measurement of the maximumthickness increase after fracture which results fromthe development of shear lips. The percent shear refersto percentage of the fracture projected area consumedby shear lips. This measurement procedure is neces-sarily arbitrary because the shear zone generally variesas the crack progresses through the specimen. Fractureappearance measurements are likely to be more easilydetermined from large specimens. Both lateral expan-sion and percent shear generally follow the same tem-perature trend as the impact energy with the brittleconditions exhibiting very low values. The NDT andDrop Weight Tear tests do not require measurementof an energy value. The NDT test is designed to deter-mine the highest temperature where a specimen breakswhen subjected to a small deflection (approximatingyield point loading) in the presence of cleavage crackoriginating in a brittle weld bead. This temperature isdefined as the nil ductility temperature. The DropWeight Tear test is designed to determine the fractureappearance in terms of percent shear as a function oftest temperature for impact values at least sufficientto completely fracture the specimen.

3.3.3.1 [Table] Summary of ASTM selected standardimpact texts.

3.3.4 Bending, see Section 3.2.4.

3.3.5 Torsion and Shear, see Section 3.2.5.

3.3.6 Bearing, see Section 3.2.6.

3.3.7 Stress Concentration. This section is concerned withthe effects of stress concentrations on the staticstrength of alloys. Stress concentrations are of specialconcern when high strength alloys are being used insevere service. In these circumstances large losses inload carrying capacity can accompany the presence ofhigh stress concentrations. The majority of tests usedto investigate the effect of stress concentrations may beclassified either as notch tests or crack tests. Detaileddiscussion of both types of these tests may be foundin Appendix C: Fracture Properties (under revision).The effect of stress concentrations on the static strengthcan be strongly influenced by the metallurgical struc-ture as determined by the fabrication history and heattreatment, as well as by the external environment.

3.3.7.1 Notch properties. The results from notch testsare generally presented in terms of the notchstrength computed from the maximum loadand the original dimensions of the notchedsection (e.g., ASTM E 602, Ref. 34). Occasion-ally the strength reported is based on theunnotched section dimensions and in suchcases the strength value is referred to as theresidual strength. Strength values determinedfrom notched specimens are dependent onthe geometry of the notch and specimen size.For this reason it is not possible to comparedata from different sources unless the speci-men geometry and sizes are identical. Resultsfrom notch specimen tests reported in thishandbook are accompanied by informationconcerning the notch geometry and pertinentspecimen dimensions.

3.3.7.2 Fracture toughness. Fracture toughness in thiscontext is concerned with the influence ofcracks on the strength of alloys. Evaluation offracture toughness requires the use of speci-mens containing so called “natural cracks.”These are generally produced by fatigueloading of specimens provided with sharpnotch crack starters. Results from such testsmay be reported as crack strength, analogousto the notch strength, or as an index of frac-ture toughness which can be used to give anindication of structural load carrying capacityin the presence of cracks or crack-like flaws.Various indices have been developed and testmethods for defining several of them havebeen standardized, namely, KIc (ASTM E 399,Ref. 19), JIc (ASTM E 813, Ref. 35), and CTOD(ASTM E 1290, Ref. 36). These methods areprimarily concerned with determining criticalvalues corresponding to crack initiation understatic loading. Each method contains certaincriteria which must be met to qualify the testresult as a valid value. Invalid results arefrequently published as indicators of tough-ness trends (e.g., changes in toughness withtempering temperature). Such values havedubious value for this purpose (Ref. 22) andshould be used with caution. The plane strainfracture toughness KIc is considered to be alower bound of fracture toughness wherecrack initation is primarily controlled by thelinear elastic stress field at the crack tip. It hasspecial significance for predicting the fracturebehavior of high strength materials underhighly constrained loading (e.g., thick sections).The JIc and CTOD indices of fracture tough-ness are designed to characterize materialswhere crack initiation is accompanied by anamount of plastic flow sufficient to invalidate

Rev. 7/92


Page 10

the determination of KIc. CTOD tests haveproven useful in cases where bursts of cleavagecrack extension are observed in the testrecords. This type of behavior is commonlyobserved in ferritic steels. Under dynamicloads the fracture toughness may be well belowthe value determined from the static tests.This effect of rapid loading is most pronouncedin ferritic steels but has also been observed inheat treated low alloy steels (Ref. 37). Thereappears to be a limiting value of the stressintensity factor below which increases inloading rate produce no further decrease intoughness. Thus, a fast running crack wouldarrest if the K level falls below this value. Thearrest value KIa is determined according toASTM E 1221 (Ref. 38) which relates specifi-cally to ferritic steels.

Stable crack growth resistance under staticloads following initiation can be determinedin terms of K (ASTM E 561, Ref. 39) or interms of J (ASTM E 1152, Ref. 40). The resis-tance is expressed as a curve relating K or J tocrack extension. The K resistance curves aredependent on material thickness and criteriaare provided to ensure that the uncrackedligament is always predominantly elastic. TheJ resistance curves are presumably indepen-dent of thickness if the validity criteria of themethod are met. Resistance curves can beused to estimate the amount of stable crackgrowth (tearing) which precedes unstablefracture in very ductile alloys.

Other types of tests have been proposed tomeasure fracture toughness either directly orby correlations with valid values from KIc, JIc,or CTOD tests. The chevron notch specimen(ASTM E 1304, Ref. 41) measures a plane straintoughness quantity KIv which may be equalto KIc for materials of low toughness but willoverestimate KIc for materials having a risingcrack growth resistance curve (e.g., mostaluminum alloys, Ref. 42). Fracture toughnessvalues from Chevron notch specimens arereported only when valid KIc or JIc values arenot available. An equivalent energy method(ASTM E 992, Ref. 43) yields a K-EE valuefrom invalid ASTM E 399 tests. However, theequivalent energy method does not transforminvalid KIc values to valid values. K-EE valuesare useful only in restricted circumstancesand are not reported in this handbook.

Many attempts have been made to correlateCharpy V impact test results with KIc. Areview of these (Ref. 44) indicates that thereis no general relation which can be appliedto a wide range of alloys. An example of this

lack of generality is shown in Ref. 22 for avariety of steels where valid KIc values wereavailable. The difficulty of establishing generalrelationships is not surprising consideringthe differences in strain rate and notch acuityin the two types of tests. The slow bendprecracked Charpy test (ASTM E 812, Ref. 45)has been used to establish correlations betweenKIc and the crack strength of the Charpyspecimens. These correlations are limited byplastic yielding in the small Charpy specimen(Ref. 46) and should never be extrapolated.

3.3.8 Combined Loading (formerly Combined Properties),see Section 3.2.8.

3.4 Creep and Creep Rupture PropertiesIn most cases, creep tests are performed in tensionwith temperature and the applied load held constant.Stress values generally are based on the original areaof the specimen. In creep rupture tests the deformationis not reported and the data are represented as stressvs time to failure (rupture). ASTM E 139 (Ref. 47)gives recommended practices for creep and creep-rupture tests which are followed by most investigators.A wide variety of specimen sizes are employed and,except for very brittle materials or for tests on notchedspecimens, the size has little effect on the results. Onthe other hand, the atmosphere in contact with thespecimen can have a substantial influence on the resultsdepending on the surface reactions which may occur.Since most tests are carried out in air, caution shouldbe used in applying such data to situations where theenvironment may be more or less aggressive.

Creep data are reported in terms of either total strain oras only the plastic component of the strain. The basicdata are represented as strain as a function of time forvarious applied stress levels at a constant temperature.Creep deformation curves of this type usually give moreinformation than is needed for practical applicationsand the data are often replotted in different waysusing the normal stress as an independent variablebut plotting it on the ordinate: (a) stress vs time withstrain and temperature held constant, (b) stress vsminimum creep rate at a constant temperature,(c) stress for a specified creep strain as a function oftemperature with time constant, and (d) isochronousstress vs strain curves at constant temperature. Toobtain an isochronous curve the total creep strain or theplastic portion of the strain is plotted as the abscissawith the stress necessary to obtain that strain as anordinate. Time is then the parameter. The plastic portionof the creep strain is obtained by subtracting the elasticpart from the total strain. This procedure is somewhatindefinite because of the uncertainty regarding themodulus of elasticity which is indicated by the tangentat the origin of the isochronous curve. In graphicalrepresentation of creep data, logarithmic stress and


Page 11

time scales are generally used although no specialphysical significance is attributed to the trends of thedata on log-log representations. If the data are verylimited or cover only a small range of the variables,tabular representations may be employed. The creeprupture strength of notched specimens is sensitive tothe time-temperature dependent embrittlement thatdevelops in many high temperature alloys. Theseembrittlements are frequently associated with micro-structural changes, some of which enhance the smoothcreep strength. Generally, the sensitivity to embrittlingreactions increases with increasing notch sharpness.For this reason creep tests using very sharp notchesor cracks have special significance in the selection ofmaterials for use in critical components which operateat elevated temperatures for extended periods of time(e.g., steam and gas turbines). The notch rupturestrength is calculated on the basis of the initial areaof the notched section and, to be a useful indicator ofembrittlement, must be compared with the smoothstrength at the corresponding rupture time.

For some applications (e.g., steam and gas turbinecomponents) it is desired to characterize alloy behaviorfor service times of many years. Very long time creepor creep rupture tests are very expensive and subjectto interruptions due to power failure or equipmentbreakdown. It has been recognized for many years thatit would be advantageous to substitute temperature fortime in such a way that the long time behavior couldbe predicted from time-temperature parameters. Oneof the first and simplest proposed is the Larson-Millerparameter (Ref. 48). Many others followed havingincreased complexity. It is not yet clear which, if any,have general applicability and none have been stan-dardized. This handbook prefers to present creep andrupture directly and will report data in terms of atime-temperature parameter only if no direct data isavailable. Time-temperature parameters should beused with caution particularly when extrapolationoutside the data range is attempted.

3.5 Fatigue PropertiesThis section presents the effects of repeated loading onthe strength and deformation properties. In some casesthe temperature may vary as well as the load. For thepurposes of this discussion the information on fatigueproperties may be divided into (1) conventional high-cycle fatigue, (2) low-cycle fatigue, and (3) fatiguecrack propagation. In conventional fatigue the strainsthroughout the test are predominantly elastic and con-sequently the stresses are relatively low and the numberof cycles to failure large. In low-cycle fatigue the plas-tic component of the total strain is larger than theelastic component, the stresses are relatively high, andthe cycles to failure relatively short. Both conventionalhigh-cycle and low-cycle fatigue tests are concernedwith total life including the crack initiation and propa-

gation phases, and in some cases data are reported foreach phase. Fatigue-crack growth is most commonlyobtained from specimens with pre-existing cracks.

3.5.1 Conventional High-cycle Fatigue. The basic types oftests used are (1) rotating beam where a specimen ofcircular cross-section rotates under an applied bendingmoment, (2) reverse bending where a specimengenerally of rectangular crossection is subjected toalternating bending, and (3) axial load in which thealternating load is parallel to the specimen axis. Bothsmooth and notched specimens are employed in thesetests. Generally the load is allowed to fluctuate betweentwo definite limits although random load spectrumtests are frequently employed in the design of a par-ticular component. Random load test data have limitedapplication and are not reported in this handbook.The stresses are calculated by conventional elasticformulae. The maximum stress is referred to as Fmaxand the minimum stress as Fmin. The stress ratio R isdefined as follows:

R = Fmin/Fmax

Another definition of the stress relationship is the Aratio defined as follows:

A = Falt/Fmf

where Falt = 1⁄2(Fmax - Fmin) is the alternating stress and

where Fmf = 1⁄2(Fmax+ Fmin) is the mean stress. The stress

range is 2 Falt. Where only one stress ratio is involved,it is common to report this ratio and plot the fatiguestrength as a function of the number of cycles to fail-ure. Where more than one stress ratio is used the datamay be presented in the form of a stress range diagram.Each curve in this diagram gives the alternating stressas a function of the mean stress for a given number ofcycles to failure. The fatigue strength is derived fromthese diagrams by the following relation:

Fmax = Fmf + FaltWhen reporting the results of notch fatigue tests anattempt is made to give the dimensions of the notchspecimen as well as the stress concentration factor Kt.It should be noted that the fatigue strength of anotched specimen will, in some cases, be dependenton the absolute size of the specimen while the elasticstress concentration factor will be independent of sizeproviding the proportions remain unchanged. Forboth smooth and notched specimens, the fatigueproperties will be dependent on the type of loading,specimen type, stress ratio and, in many cases, thesurface condition of the specimen.

Two types of behavior are noted in high-cycle fatiguetests at room temperature. In one case a definite fatigueor endurance limit exists. Stresses below this limitproduce no failure in an extreme number of cycles.

Rev. 7/92


Page 12

This behavior is most clearly noted in ferritic steelsand is less pronounced in heat treated low alloy steels.In the other case no definite limit is observed. This typeof behavior characterizes aluminum alloys and otherface center cubic materials. In this case the fatiguelimit is arbitrarily assigned, usually at 106 or 107 cycles.

3.5.2 Low-cycle Fatigue. The most commonly used low-cyclefatigue tests employ cylindrical specimens subjectedto axial loads. Attention is focused primarily on thestrain and strain range rather than on the stress andstress range. Tests are generally carried out betweenfixed strain limits. It is important to recognize thatunder these conditions the alloy may either cyclicallyharden or soften. When hardening occurs, progres-sively higher stresses are produced under a constantstrain range. If strain softening occurs, the stresseswill progressively decrease under a constant strainrange. In order to simplify the material characterizationit is common to express the strain range in terms ofthe stress range developed at the half-life (half thecycles to failure) point. Such a plot of strain vs stressat a given number of cycles is called the cyclic stress-strain curve. For cyclically strain hardening materials,the cyclic stress-strain curve is above the monotoniccurve; for cyclically softening materials it is below theconventional stress-strain curve. Cyclic stress-straincurves are more useful in design against low-cyclefatigue than are the conventional monotonic curves.A simple criterion for establishing whether the materialis cyclically strain hardening or softening (Ref. 49) isthe ratio Ftu/Fty. When this ratio is greater than 1.4the material is likely to be cyclically strain hardening;if less than 1.2, the material is likely to be strain soften-ing. Ratios between 1.2 and 1.4 do not clearly identifyeither strain hardening or strain softening. A morecomplex and somewhat more accurate criterion fordetermining whether softening or hardening takesplace under cyclic loading is given in Ref. 50. The basicrelation between fatigue life and plastic strain in theabsence of creep, first formulated by Manson (Ref. 51)and Coffin (Ref. 52), is

∆εp = MN zf

where εp is the plastic component of the cyclic strain,Nf is the number of cyles to failure, and M and z arematerial constants. Thus, the low-cycle fatigue behaviorof a material is often characterized by a plot of plasticstrain vs cyclic life on log-log coordinates. Accordingto the Manson-Coffin relation such a plot should be astraight line; however, curvature is sometimes notedin the high-cycle range.

General relationships. The fatigue limit generallyincreases with tensile strength and sometimes is esti-mated as one-half the tensile strength. However, atbest, this relationship is a very rough approximation.Steels heat treated to hardness above about 40 RC can

not be characterized by such a simple relation (Ref. 53),probably because at high hardness levels steels becomevery sensitive to crack initiation at surface irregularitiesand/or inclusions. Much more useful estimates offatigue lives from tensile properties have been madeby developing relationships based on the total strain(elastic plus plastic). The following relation has beenproposed (Ref. 54) for estimating the fatigue life interms of the total strain and the tensile properties.

∆ε = (Nf/D)-0.6+3.5(Ftu/E)Nf-0.12

where D is the true diametrical strain = ln(1 - RA),where RA is the reduction in area at fracture and Nfis the fatigue life at a total strain range of ∆ε. For mostmaterials this relation gives useful results in the rangebetween 10 and 107 cycles.

At elevated temperatures creep may be superimposedon fatigue and the interpretation of the data becomesrather complicated. Frequency of cycling is an importantvariable under these circumstances. If temperatureand mean stress are sufficiently high, creep rupturebecomes the primary mode of failure and the times tofailure are often reported rather than the number ofcycles to failure. For much of the elevated temperaturefatigue test data, it is not possible to clearly classifythe information as conventional high-cycle or as low-cycle fatigue.

3.5.3 Fatigue Crack Propagation. This section contains in-formation on propagation of so-called natural cracksunder cyclic loading. With few exceptions these cracksare produced from a starter notch by fatigue cyclingat stress intensity levels well below KIc. ASTM E 647(Ref. 59) gives details concerning the preparation ofspecimens, testing methods, and validity criteria.Logarithmic coordinates are most frequently used toplot crack propagation rates (da/dN) as a function ofthe stress intensity range ∆K. Such plots often definetwo major regions of interest: (1) region where the rateis a nearly linear function of the stress intensity rangeand (2) a value of the stress intensity range, designatedas ∆Kth, below which fatigue crack propagation pre-sumably will not occur under the test conditions. It isimportant to realize that the crack propagation ratesand the Kth value will be a function of the test environ-ment and the mean value of the stress intensity range.In come cases the value of Kth will be below KIscc. Atelevated temperatures and in the presence of aggressiveenvironments, the test results can be a function of thecyclic load frequency. Appendix C (under revision)gives further details.

3.6 Elastic PropertiesThis section includes not only the classical elastic con-stants but also the tangent and secant moduli whichextend into the plastic region. A high degree of sensi-tivity and precision in strain measurements is requiredin determining the elastic constants. Values reported


Page 13

in the literature often exhibit considerable scatter dueto poor testing techniques. Some annealed alloys haveno true proportional limit at room temperature, andat sufficiently high temperature no alloy has a welldefined proportional limit. Under these circumstancestruly elastic properties do not exist and attempts todetermine the elastic constants will result in valueswhich depend strongly on the testing technique.

3.6.1 Poisson’s Ratio. Poisson’s ratio is the absolute value ofthe ratio of the transverse strain to the correspondingaxial strain resulting from uniformly distributed axialstress below the proportional limit of the material.Poisson’s ratio is related to the elastic modulus E andshear modulus G as follows:

µ =E

2G - 1

ASTM D 132 (Ref. 55) is a standard method for deter-mination of Poisson’s ratio at room temperature.

3.6.2 Modulus of Elasticity. For the purposes of this hand-book the modulus of elasticity refers to Young’smodulus. The static modulus of elasticity is the ratio ofthe normal stress to the corresponding strain measuredat stresses below the proportional limit of the material.The determination may be made in either tension orcompression (ASTM E 111, Ref. 56). The tensionmodulus E is theoretically equal to the compressionmodulus Ec; however, residual stresses may producesome difference in these values. The modulus mayalso be determined by resonance methods (Ref. 57)and when so determined is generally referred to asthe dynamic modulus. At temperatures where timedependent deformation is absent the static and dynamicmoduli are essentially equal with the dynamic valuein some materials being slightly lower than the staticvalue. At temperatures where time dependent defor-mation would occur in static tests the dynamic modu-lus is useful where vibratory loads accompany creepdeformation. The modulus of elasticity decreases withincreasing temperature and is much less sensitive tothe metallurgical structure than is the flow stress.However, the modulus will depend on the testingdirection in cases where a bulk crystallographicanisotropy is present.

3.6.3 Modulus of Rigidity. The modulus of rigidity or theshear modulus is the ratio of the shear stress to thecorresponding shear strain for shear stresses belowthe proportional limit in shear of the material. Themodulus of rigidity is related to Young’s modulusand Poisson’s ratio as follows:

G = E2(1+ µ)

ASTM E 143 (Ref. 58) gives a standard method fordetermining the shear modulus.

3.6.4 Tangent Modulus. The tangent modulus is the slope ofthe stress strain curve at each stress value considered.Reported values are subject to considerable variationbecause of the basic difficulty in determining the slopeof any curve. The tangent modulus may be determinedin either tension or compression.

3.6.5 Secant Modulus. The secant modulus is the slope of aline from the origin to the stress values considered. Thesecant modulus may be determined in either tensionor compression.

4 FabricationThe term fabrication is used here to mean all of theprocesses, except heat treatment, which may normallybe employed in the manufacture of components frommaterial supplied by the alloy producers. These pro-cesses include forming (forging, rolling, drawing, etc.),metal removal (machining, chem milling, etc.), joining(welding, soldering, brazing, etc.), and post fabricationtreatments that may be required (cleaning and othersurface treatments). In many cases the fabrication pro-cess may profoundly alter the mechanical propertiesof the material through the effect on the microstructureby the introduction of residual stresses or the produc-tion of crack like defects. This handbook can devoteonly a limited amount of space to the very large amountof information that exists on the so-called fabricabilityof an alloy or the details of recommended fabricationprocedures. Emphasis is placed on the effect of fabri-cation processes which can adversely affect the loadcarrying capacity in service. In this connection par-ticular attention is given to welding and other joiningprocesses.

4.1 FormingThe information presented in this section relatesprimarily to the effects of various forming processeson the mechanical properties. Forming includes theprocesses of permanent metal deformation (e.g., forg-ing, rolling, drawing, etc.). Forging temperatures arereported as the maximum starting and the minimumfinishing temperature. These apply to closed die orblacksmith forgings in the weight range from 5 to1000 lbs. Forging temperatures for small parts areapproximately the same; however, for small forgingsa great deal of care must be exercised to avoid criticalstrains which can induce grain growth on reheatingfor a subsequent operation or during heat treating.

4.2 Machining and GrindingOnly a very limited amount of information on thesesubjects is presented. The reader is referred to theMachinability Data Center, Metcut Research Associates,Cincinnati, Ohio, for detailed information on themachinability of the alloys in this handbook.


Page 14

4.3 JoiningThe joining processes of welding and brazing and, to alesser extent, soldering can produce damaged regionsin a fabricated component which may severely limitits load carrying capacity. Thus, the deposited fillermetal or heat affected zone of the parent metal maycontain embrittled regions or cracks. Special attentionis given to the effects of welding and brazing param-eters on the fracture properties and to recommendedprocedures which will produce sound joints.

4.4 Surface TreatingAttention here is focused on cleaning procedures orspecial surface treatments (e.g., case hardening) thatcould influence the mechanical properties.

Fig. 2.2.3.1 Normal induction curve and hysteresis loop for ferromagnetic materials

Bs

Brs

µo

µm

Hcs

Magnetizing Force, H – Oersteds

Indu

ctio

n,

B –

Gaus

s

Table 3.3.3.1 Summary of ASTM selected standard impact tests

Name Specimen Notch Index of Toughn Reference

Charpy and Izod Simple Beam None Absorbed EnergyASTM E 23Ref. 30

Charpy and IzodNotched Beam 10x10 mm (a)

V, Keyhole or U

Absorbed EnergyLateral ExpansionTransition Temp.Fracture Appearance

ASTM E 23Ref. 30

NDT Surface Crack PlateBrittle WeldCrack Starter

Nil DuctilityTransition Temp.

ASTM E 208Ref. 31

Drop Weight TearEdge NotchPlate 3x12 in. full thick

Pressed VR < 0.001 in.

Fracture AppearanceASTM E 436Ref. 32

Dynamic TearEdge NotchPlate 7x1.6 in.3/16<t<5/8 in.

Pressed VR < 0.001 in.

Absorbed EnergyFracture Appearance

ASTM E 604Ref. 33

(a) Subsized Charpy specimens permitted

Documents

Aerospace Structural Metals Handbook - Maechanical Properties