<<

Copyright © 1986 ASM International® ASM Handbook, Volume 10: Materials Characterizations All rights reserved. R.E. Whan, editor, p 299-308 www.asminternational.org

Optical

M.R. Louthan, Jr., Department of Materials Engineering, Virginia Polytechnic Institute and State University

General Uses Limitations • Imaging of topographic or microstructural features • Resolution limit: Approximately 1 tzm on polished and etched surfaces at magnifications • Limited depth of field (cannot focus on rough of 1 to 1500x surfaces) • Characterization of grain and phase structures and • Does not give direct chemical or crystallographic dimensions information about microstructural features Examples of Applications Estimated Analysis Time • Determination of fabrication and heat-treatment • 30 min to several hours per specimen, including history preparation • Determination of braze- and weld-joint integrity Capabilities of Related Techniques • Failure analysis • Characterization of the effects of processing on • Scantling electron : Provides better and properties resolution (higher magnifications); greater depth of field (can image rough surfaces); qualitative Samples elemental microanalysis

• Form: , , composites, and geologic • Electron probe x-ray microanalysis: Provides materials quantitative elemental microanalysis • Size: Dimensions ranging from 10 -5 to 10 -1 m • Transmission electron microscopy: Provides much • Preparation: Specimens are usually sectioned and better resolution (much higher magnifications) on mounted, ground, and polished to produce a flat, specially prepared specimens; semiquantitative scratch-free surface, then etched to reveal elemental microanalysis; crystallographic microstructural features of interest information on microstructural features

Introduction proximately 50x or higher; macroscopy Because the macro- and microstructure of Optical metallography, one of three gen- (macrostructural examination), 50 x or metals and alloys often determine the behav- eral categories of metallography, entails ex- lower. ior of the material, characterization of the amination of materials using visible light to Optical microscopy and, occasionally, effects of composition, processing, service provide a magnified image of the micro- and SEM are used to characterize structure by conditions, and other such variables on the macrostructure. In scanning electron micros- revealing grain boundaries, phase bound- macro- and microstructure is frequently re- copy (SEM), the second category, the sur- aries, inclusion distribution, and evidence of quired. Typical structure-property relation- face of the specimen is bombarded with a mechanical deformation. Scanning electron ships that have been established using optical beam of electrons to provide information for microscopy is also used to characterize frac- metallography include: producing an image (see the article "Scan- ture surfaces, integrated circuits, corrosion ning Electron Microscopy" in this Volume). products, and other rough surfaces, espe- • A general increase in yield strength and Lastly, transmission electron microscopy cially when elemental microanalysis of small of a with decreasing grain (TEM) consists of passing a beam of elec- features is desired. Transmission electron size trons through a very thin specimen and microscopy is used to examine • A general tendency for a decreased duc- analyzing the transmitted beam for structural arrangements or structures and other small tility with increasing inclusion content information (see the article "Analytical defects in metals and alloys. Second-phase • Correlations of weld penetration, heat- Transmission Electron Microscopy" in this particles not observable using optical metal- affected zone (HAZ) size, and weld- Volume). Microscopy (microstructural ex- lography can frequently be analyzed using defect density with the nature and char- amination) involves magnifications of ap- TEM. acter of the welding 300 / Metallographic Techniques

• Evaluation of such surface treatments as Edition of Metals Handbook). The most Fig. 1 One method of mounting carburizing and induction hardening by widely used sectioning device is the the sample to retain flatness for determinations of the depth and micro- cutoff machine, ranging from units using metallographic examination structural characteristics of the hardened thin -rimmed wafering blades to The mount can also be filled with ground glass, region those using wheels that are more than 1.5 pelletized AI203, or another hard material to • Correlations of fatigue crack growth rates mm (1/16 in.) thick, 30 to 45 cm (12 to 18 in.) maintain flatness. and fracture-toughness parameters with in diameter, containing carbide parti- Test Specimen mounted in Bakelite such structural variables as inclusion con- cles. tent and distribution Heat is generated during abrasive cutting, Test specimen • Association of failure initiation sites with and the material just below the abraded microstructural inhomogeneities, such as surface is deformed. To minimize burning second-phase particles and deformation, a lubricant or coolant is • Correlations of anisotropic mechanical typically used. Wet cutting yields a flat ~_oo -- L behavior with elongated grains and/or relatively smooth surface. However, because preferred grain orientations of the abrasion associated with cutting, the structure of the metal or is damaged to The of metals and alloys a depth of approximately 1 mm (0.04 in.). are determined by composition, solidifica- The exact depth of damage depends on the tion processes, and thermomechanical treat- type of cutoff wheel used, the cutting speed, ment. Therefore, these process variables de- and the hardness of the specimen. The harder termine the response of metals and alloys to the specimen, the shallower the depth of laboratory and service environments. Be- damage. This damaged layer must be re- cause of the relationships between structure moved by grinding. However, before the and properties, metallographic characteriza- specimen can be conveniently ground, it Hardened balls or chilled iron shot (typ) tion is used in materials specification, qual- often must be mounted. ity control, quality assurance, process con- Mounting facilitates handling of the trol, and failure analysis. specimen. A procedure that does not damage Optical metallography is applicable to the specimen should be selected. Because studies ranging from fundamental research to large specimens are generally more difficult production evaluations. This article will dis- to prepare than small ones, specimen size cuss use of optical methods to evaluate should be minimized. Standard or typical structure and to relate that structure to pro- specimen mounts are right circular cylinders cess conditions and/or material behavior. 25 to 50 mm (1 to 2 in.) in diameter. Detailed information on the principles and Mounting mediums should be compatible instrumentation of optical microscopy is with the specimen regarding hardness and ,- / 32 mm (1.25 in.) diam ] -, available in the article "Optical Micros- abrasion resistance. Two common mounting copy" in Volume 9 of the 9th Edition of Bakelite mounting materials are thermosetting phenolics, such Nickel plate, 0.05 mm Metals Handbook. as Bakelite, and thermoplastic materials, (0.002 in.) thick Specimen Preparation such as methyl methacrylate (Lucite). A Section A-A thermosetting develops a rigid The first step in metallographic analysis is three-dimensional structure upon being to select a sample that is representative of the heated and held at 200 to 300 °C (390 to 570 material to be evaluated. This step is critical °F). A thermoplastic polymer softens when thermoplastic materials are relatively soft to the success of any subsequent study. The held at elevated temperatures. mounting materials, and the specimen in second, equally important step is to correctly Mounting involves placing the specimen such a mount must often be surrounded by a prepare a metallographic specimen. in a mold and surrounding it with the appro- hard material, for example, hardened steel The region of the sample that is of interest priate powders. The mold and its contents balls (Fig. 1). This material helps retain the must be sectioned from the component. For are then heated under pressure to the thermal edges of the sample by maintaining a flat example, if a failure occurred because a steel setting or the softening temperature. Once surface during grinding and polishing. Ad- pipe leaked during service, the metallo- the powder sets, thermosetting mounts can ditional information on mounting techniques graphic analysis would probably involve at be removed from the mold without lowering and materials is available in the article least three samples: one removed from the the temperature; thermoplastic mounts must "Mounting of Specimens" in Volume 9 of pipe such that a portion of the leak is be cooled to ambient temperature before the 9th Edition of Metals Handbook. contained in the sample, another removed removal. Mounting pressure or temperature Grinding is generally considered the near the leak, and a third taken far from the may alter the structure of low melting tem- most important step in specimen preparation. leak. Each of the samples would be mounted perature or soft and/or fragile specimens; Care must be taken to minimize mechanical to facilitate handling. Selected surfaces therefore, castable (cold-mounting) tech- surface damage. Grinding is generally per- would then be ground flat, polished, and niques have been developed. formed by the abrasion of the specimen etched to reveal the specific structure or Plastics that set at room temperature are surface against water-lubricated abrasive structures of interest. referred to as castable (cold-mounting) ma- wheels (assuming water does not adversely Sectioning of a metallographic sample terials. The most widely used materials are affect the metal). Grinding develops a flat must be performed carefully to avoid altering resins. resist acids and strong surface with a minimum depth of deformed or destroying the structure of interest (see the solvents effectively, a desirable characteris- metal and usually is accomplished by using article "Sectioning" in Volume 9 of the 9th tic in any mounting material. Epoxies and progressively finer abrasive grits on the Optical Metallography / 301

Fig. 2 The effect of disturbed metal on the metallographic free specimen surface, in which inclusions appearance of a plain steel and other second-phase articles may be vis- (a) A layer of disturbed metal--an artifact structurecaused by grinding damage--covers the polished ible. Polishing damage, such as that illus- surface. (b) The layer of disturbed metal is removed, and the structureis revealed to be lamellar pearlite. trated in Fig. 3, should be recognized and Etched using picral. 1000 × avoided when preparing metallographic specimens. Etching includes any process used to reveal the microstructure of a metal or alloy. Because many microstructural details are not observable on an as-polished specimen, the specimen surface must be treated to reveal such structural features as grains, grain boundaries, twins, slip lines, and phase boundaries. Etchants attack at different rates areas of different crystal orientation, crystal- line imperfections, or different composition. The resulting surface irregularities differenti- ally reflect the incident light, producing contrast, coloration, polarization, etc. Vari- ous etching techniques are available, includ- (a) (b) ing chemical attack, electrochemical attack, thermal treatments, vacuum cathodic etch- ing, and mechanical treatments (see the arti- Fig. 3 The effect of improper polishing on AISI 1010 steel cles "Color Metallography" and "Etching" (a) "Comet tails" from improper polishing. (b) The same material polished correctly, exhibitingsmall in Volume 9 of the 9th Edition of Metals manganese sulfide inclusions Handbook). Chemical and electrochemical attack are the most frequently used. The details of the structure revealed by etching depend on the type of etchant used (Fig. 4). Metallography involves many steps that can obscure or alter the structure observed during examination, leading to erroneous conclusions. Therefore, specimen prepara- tion is not necessarily straightforward, and care must be taken to ensure that the struc- ture observed is not an artifact. Good metal- lography is necessary in developing a corre- lation between the structure and the properties of metals and alloys. Macroanalysls Macrostmctural characterization of metals (a) (b) and alloys is the detailed evaluation of large- scale inhomogeneities in composition, mor- grinding wheels. A typical sequence might Polishing of the metallographic speci- phology, and/or density. These inhomoge- begin with 120- or 180-grit papers and pro- men generally involves rough polishing and neities may develop during such procedures ceed to 240, 320, 400, and 600 grits. fine polishing. In rough polishing, the cloth as casting, extrusion, forging, rolling, and Scratches and damage to the specimen sur- covering on a wheel is impregnated with a welding or during service. Figure 5 shows face from each grit must be. removed by the fine (often as small as 1 txm) diamond paste the macrostructure of a small relatively pure next finer grinding step. or a slurry of powdered ot-A1203 in water, aluminum ingot exhibiting typical cast grain The surface damage remaining, on the and the specimen is held against the rotating structure. To obtain the macrograph, the specimen after grinding must be removed by wheel. The cloth for rough polishing is aluminum ingot was sectioned, then ground polishing. If this disturbed or deformed frequently napless, providing easy access of and polished to produce a flat reflective metal at the surface is not removed, micro- the polishing abrasive to the specimen sur- surface. The polished section was then structural observations may be obscured face. Fine polishing is conducted similarly, etched by immersion in a solution that at- (Fig. 2). Because structure and properties are but with finer (down to 0.05 ~m in tacked the various grain orientations at dif- so closely related, conclusions based on the diameter) on a napped cloth. ferent rates. structure in Fig. 2(a) would lead to incorrect Although often automated, polishing can The etched structure was examined using interpretation of the anticipated behavior of be performed by hand. Vibratory polishing a low-power microscope. The structural el- the metal. Grinding of metallographic spec- and techniques have also ements visible in this macrograph are grains. imens is discussed in the article "Mechani- been developed for many metals and alloys The small grains near the bottom of the ingot cal Grinding, Abrasion, and Polishing" in (see the article "Electrolytic Polishing" in appear relatively equiaxed. This region of Volume 9 of the 9th Edition of Metals Volume 9 of the 9th Edition of Metals small equiaxed grains is the chill zone. Handbook. Handbook). Polishing should yield a scratch- During casting, such macrostructural defects 302 / Metallographic Techniques

Fig. 4 Comparison of nital and picral for revealing a martensite Fig. 5 Macrostructure of as-cast structure aluminum ingot Ca) Specimen etched using nital (nitric acid in ethanol or methanol). (b) Specimen etched using picral (picric Transverse section shows outer chill zone and acid in ethanol). Both 1000 × columnar grains that have grown perpendicularly to the mold faces. Etched using Tucker's reagent. 1.5×

Ca) (b) dye to fluoresce, and the cracks became readily observable. Dye-penetrant tech- Fig. 6 Macrostructure of a continuous-cast copper ingot niques are excellent for examination of Ca) Spider cracks revealed using dye-penetrant inspection. Transverse section at top; longitudinal section at crack-like macrostructural defects in metals: bottom. (b) Same ingot, etched using Waterbury's reagent. Cracks are not revealed. Both approximately However, grains and other microstructural 0.5x features are visible only after etching, which frequently obscures the presence of the cracks. Therefore, different metallographic techniques are necessary to reveal various macrostructural elements. Materials characterization by optical \ macrostructural examination can be divided into three categories. First, examination of the macrostructure of metallographically prepared sections removed from the compo- nent of interest is used to evaluate such structural parameters as:

• Flow lines in wrought products • Solidification structures in cast products • Weld characteristics, including depth of penetration, fusion-zone size and number of passes, size of heat-affected zone and J type and density of weld defects o • General size and distribution of large inclusions and stringers Ca) (b) Fabrication defects, such as laps, cold welds, folds, and seams, in wrought products • Gas and shrinkage porosity in cast prod- as gas or shrinkage porosity and center evaluation of the macrostructure may fail to ucts cracks can develop. Many of these defects reveal this type of structural defect (Fig. 6b). • Depth and uniformity of a hardened layer can be characterized using macrostructural The cracks shown in Fig. 6(a) were re- in a case-hardened product evaluation. vealed by applying a dye penetrant to the Figure 6(a) shows spider cracks in the polished specimen. The dye was drawn into center of a copper specimen. This specimen the cracks by capillary action, and the sur- Second, characterization of the macro- was sectioned, ground, and polished, but not face was then wiped clean. The specimen structural features of fracture surfaces is used etched. Chemical etching and subsequent was then placed under a light that caused the to identify such features as: Optical Metallography / 303

Fig. 7 Flow lines in a forged Fig. 8 Case-hardened layer in tures of a part or component selected for 4140 steel hook Wl tool steel process characterization or quality assur- Specimen was etched using 50% HCI. 0.5 × Specimenswere austenitizedat 800 °C (1475 °F), ance. The selected region of the specimen brine quenched, and tempered 2 h at 150 °C (300 must then be removed from the component °F). Black rings are hardened zones. Etched using using techniques that do not damage or 50% hot HCI. Approximately 0.5 X distort the features of interest. The section of interest is then prepared metallographically, and the prepared section is characterized using macroscopic examination. Macroscopic examination generally does not require the extreme surface smoothness needed for microscopic examinations. Such surface preparation techniques as etching are frequently prolonged such that surface fea- tures are greatly enhanced; therefore, quan- titative measurements should not be con- ducted on macroetched samples. Heavy etching accentuates any microstructural inhomogeneity (Fig. 7). The flow lines show the direction of metal flow during processing incomplete, represents the wide variety of and frequently represent paths for easy frac- features that can be evaluated and character- ture. ized using optical macroscopy. One of the Figure 8 shows the use of similar macro- major constraints of optical macroscopy is its scopic techniques to illustrate the depth of limited depth of focus when the surfaces case hardening in a tool steel; Fig. 9, for examined are very rough. If this lack of examination of an arc weld. The weld depth of focus is a problem, use of SEM is macrograph shows the different etching recommended. characteristics of the various areas of the weld. The existence of the HAZ illustrates the effect of welding on the structure. The Macroscopy of 2% nital etchant used to reveal the weld IVletallographic Sections macrostructure is much less aggressive than the 50% hydrochloric acid etchants used on Preparation of a metallographic section for the specimens shown in Fig. 7 and 8 and examination requires careful selection of the reveals more structural detail. area to be characterized. This area must be Macroscopic examination of cast struc- chosen to represent the unique features of a tures can be used to reveal various casting specific zone of interest or the general fea- conditions. Solidification patterns are appar-

• Fracture initiation site Fig. 9 Section through an arc butt weld joining two 13-ram (0.5-in.) • Changes in crack propagation process thick ASTM A517, grade J, steel plates The schematicshows the fusion zone, the heat-affected zone, and base metal. Etched using 2% nital. 4 x Third, characterization of surfaces and surface defects on parts and coupons is accomplished for purposes such as: • Estimations of surface roughness, grind- ing patterns, and honing angles • Evaluation of coating integrity and uni- formity • Determination of extent and location of wear • Estimation of plastic deformation associ- ated with various mechanical processes • Determination of the extent and form of corrosive attack; readily distinguishable types of attack include pitting, uniform,' crevice, and erosion corrosion • Evaluation of tendency for oxidation • Association of failure with welds, sol- ders, and other processing operations The above listing of uses for macrostruc- tural characterization of materials, though 304 / Metallographic Techniques

Fig. 10 Sketch of grains in a Fig. 12 Example of how fracture surface features can point to the typical cast ingot failure origin (a) Fractograph of a high-velocityfracture in steel plate showingchevron pattern indicatingthe origin Chill Columnar Central Columnar Chill (left). (b) Schematicview of (a) zone zone zone zone zone

...... ~~i~i~ii' >~ Mold Mold wall wall le)

Fibrous Fig. 11 Transverse fracture of an AISI 1075 steel railroad rail Fracture nucleus(dark area near top of railhead) initiated a fatigue crack (large light area around nucleus).

(b)

on the fracture surface (Fig. 12) known as These two types of metallography are com- chevrons or herringbone marks. The tip of plementary, but examination should always the V generally points to the origin of the begin at low magnification and work up- failure. The origin of fatigue failures can wards. Detailed information on sample prep- also be isolated using macroscopic examina- aration, equipment and etchants used in tion. For example, Fig. 13 shows a failure of macroanalysis, and interpretation of results a steel housing tube initiated in four regions. is available in Volume 9 of the 9th Edition of Each initiation region is observable in the Metals Handbook. ent in the cross sections of macroetched macrographs, as shown by the four arrows. ingots (Fig. 5 and 6). The outer chill zone The position of the crack fronts at various depth, shape and size of the columnar or times during the failure process is also visi- Microanalysis dendritic grains perpendicular to the mold ble as the so-called beach marks that are The importance of microstructure to the wall, and size of the central equiaxed zone in initially fairly concentric to the origin. The properties of metals and alloys has long been a casting can be established (Fig. 10). One major problem with optical macroscopic or recognized. Grain size, twins, and the size, benefit of the macroscopy of cast structures microscopic examination of fracture surfaces shape, and distribution of second-phase par- is the ability to reveal the structure and is the technique's inability to obtain favor- ticles are important in determining the be- associated defects. able focus over the entire surface if the havior of most structural metals. Therefore, Optical macroscopic examination of a magnification exceeds 5 to 10 x. Therefore, characterization of the various microstruc- fracture surface may reveal features that will SEM has become a standard metallographic tural elements in a metal or alloy is often help establish the failure process. For exam- tool in failure analysis. necessary. Process-control parameters are ple, Fig. 11 illustrates a fracture in a railroad Macroscopic evaluation of corroded parts established to provide specific grain sizes. rail. The relatively smooth region in the can also provide considerable insight into The number, size, and distribution of sec- photograph represents crack growth because corrosion processes. Each type of corrosion- ond-phase particles, such as inclusions, are of cyclic or fatigue loading. The dark spot induced failure causes a characteristic frequently specified, and quantitative metal- (or fish eye) in the center of the reflective macrographic appearance; therefore, evalua- lographic procedures have been developed to area represents the fracture-initiation site. tion using optical macroscopy is generally describe microstructure. The remainder of the surface failed by over- effective. A frequent mistake in failure anal- The upper limit of useful magnification in load. This macroscopic observation of the ysis is to neglect examination of the broken the is approximately failed rail shows that the failure was initiated pieces at low magnifications. Too frequently 1500 x, and the fundamental limitations of because of the dark-appearing defect in the the component is sectioned immediately, and light optic systems limit resolution to fea- rail. the failure, casting, or other type of speci- tures which are --1 Izm or larger. Although The macroscopic nature of many fractures men examined at high magnification. Opti- this value is small, many microstructural is such that the fracture origin is easily cal microscopic evaluation clearly is signif- features influencing the properties of metals recognizable. Brittle or low-ductility frac- icant in any structural evaluation, but should and alloys are too small to be observed using tures have characteristic V-shaped markings not replace characterization by macroscopy. optical microscopy. , numerous Optical Metallography / 305

Fig. 13 Froctographs of a typical fatigue crack in a clamp (a) The fatigue crack origin is marked by the arrow. The crack propagatedto the right by continuousfatigue cracking (light) region, then continuedalternately by rapid tearing and slow fatigue cracking. 2 ×. (b) Higher magnificationview of the region near the arrow in (a). 10 ×

(a) (b)

Fig. 14 Copper alloy 26000 (cartridge , 70%) sheet, hot rolled to a thickness of 10 mm (0.4 in.), annealed, cold rolled to a thickness of 6 mm (0.239 in.), and annealed to a grain size of 0.120 mm (0.005 in.) At this reduction, grains are basicallyequiaxed. Compare with Fig. 15. Diagram in lower left of each microgroph indicatesorientation of the view relative to the rolling plane of the sheet. Etched using NH40H plus H202. 75 ×

(a) (b) (c)

Fig. 15 Some alloy and processing as in Fig. 14, but reduced 50% by cold rolling from 6 mm (0.239 in.) to 3 mm (0.120 in.) Grains are elongatedin the rolling direction. Diagramsindicate same orientationof view as in Fig. 14. Etched using NH40H plus H202. 75 ×

(a) (b) (c) 306 / Metallographic Techniques

Fig. 16 Dendritic solidification structure in a Ni-5Ce (at.%) alloy Nickel dendrites (light in b and c) are surrounded by a matrix of nickel-cerium eutectic. (a) 25 x. (b) 75 x. (c) 2,50 x

(a) (b) (c) types of second-phase particles, spinodal and Optical characterization of the microstruc- microstructure in several directions. Figure ordered structures, and many aspects of mar- tures of metals and alloys involves determi- 14 shows an annealed microstructure exhib- tensitic structures can be categorized as too nation of the size and shape of the grains, the iting similar grain shapes in all three views. small for optical microscopy. Therefore, me- extent of twinning, and some of the charac- Grain size is characterized by placing a line tallographic observations of these very fine teristics of grain boundaries and other ob- of known length (or preferably a circle of structural features is generally restricted to servable defects. Solidification, -state known circumference) on the magnified im- electron microscopy. Optical microscopy, transformation, deformation, and annealing age of the microstructure and counting the then, is used primarily to examine grain microstructures are the four basic types in number of intersections between the line and structures and the morphology of large metals and alloys. Each of these has distinct grain boundaries in the microstructure. The second-phase particles. Specialized optical characteristics. number of intersections, N, can be converted metallographic techniques, such as polarized Microstructural features exist in three di- to a measure of grain size, d, using: light microscopy and interference micros- mensions, and in a typical metallographic copy, can add significantly to the informa- observation, only two dimensions are ob- d- L tion obtained in a microscopic investigation, served. Therefore, effective microscopy fre- NM (Eq 1) and interference microscopy can be used to quently requires microstructural observa- identify height differences on a sample sur- tions in two or more directions. Figures 14 face that are far smaller than 0.2 ixm. and 15 illustrate the value of viewing the where M is the magnification of the image

Fig. 17 Typical defects observable using optical microscopy (a) Shrinkage porosity in an aluminum alloy 5052 ingot. Note angularity. ,SO ×. (b) Coarse primary CrAI 7 crystal in aluminum alloy 7075 ingot. 100 ×. (c) Oxide stringer inclusion in a rolled aluminum alloy 1100 sheet. 250 x. All as-polished

J

~ ~ iii¸ iiii!iiii ¸

~1 ~ h

(a) (b) (c) Optical Metallography / 307

Fig. 18 Continuous Fig. 20 The effect of prior cold require controlling defects to regulate their grain-boundary precipitate in work on recrystallized grain size number, size, and shape in a particular U-700 nickel,base heat-resistant manner. For example, a component having a 0.4 stringer distribution such as that shown in alloy ~ o 700 °C '" Fig. 17(c) would have better ductility if Etched usingHCI, ethanol, and H202. 500 × E |/ • 600 °C I Recrystallization / " 500 °C '- specimens or components were tested with E r~ • 450 °C / temperature _ 0.3 ..... the major stresses parallel to the stringer than o 400 °C J if specimens were oriented with the major c stresses perpendicular to the stringer. '~ 0.2 Transformation structures almost always contain two phases. In such structures, the major phase is typically termed the matrix, or base structure, and the minor phase is ~ 0.1 termed the second phase. The size, shape, I c-t I t °--~ and distribution of second-phase particles are important in determining the properties of 0 0 10 20 30 40 metals and alloys. Characterization of Amount of cold work, % second-phase morphology can sometimes be accomplished using optical metallography. However, the second phase is sometimes so small that the resolution necessary to char- render the grain structure--and the resulting acterize the phase morphology exceeds the mechanical properties---anisotropic. Be- limits of the optical microscope. In these cause of the interrelationships between grain cases, transmission electron microscopy Fig. 19 Discrete precipitates morphology (size and shape) and mechanical must be used. Age-hardenable or preci- along grain boundaries in a properties, characterization of the grain pitation-hardened metals and alloys gener- structure is a typical metallurgical function. ally must be characterized using electron nickel-base heat-resistant alloy The most commonly observed solidifica- microscopy. Specimenwas etchedusing 35 mL HCI, 65 mL ethanol, and 7 drops H202. 500 × tion structure is dendritic. A dendritic struc- High-temperature phase transformations ture usually exhibits compositional varia- frequently nucleate at grain boundaries. The tions, with the dendrite arms containing less grain-boundary structures can be discrete or alloying element or impurity than interden- continuous. Continuous grain-boundary con- dritic regions. Because of such composi- stituents (Fig. 18) provide easy fracture tional changes (termed coring), the rate of paths when the grain-boundary phase is less etching at interdendritic regions differs from ductile than the matrix phase. For the mate- that at dendrite arms. If the alloying element rial shown in Fig. 18, the expected failure or impurity content is high, interdendritic would be fracture along the grain-boundary regions may develop a two-phase structure carbides. Figure 19 also shows discrete (Fig. 16). Because dendrite arm spacing second-phase precipitates at grain bound- tends to decrease with increasing cooling aries. Comparison of the microstructures rates, the properties of as-cast metals depend shown in Fig. 18 and 19 reveals differing on the solidification rates. second phase morphologies in two similar Most metals shrink during solidification. alloys. Therefore, the properties of these two Therefore, the liquid trapped between den- alloys are also different, with the structure drite arms during solidification is frequently illustrated in Fig. 18 exhibiting the highest insufficient to fill the space between the arms strength but the least ductility. when solidification is complete. This inabil- The microscopic details of deformation ity to fill the remaining space leads to shrink- structures typically cannot be established age porosity, which can be observed micro- using optical metallography. Deformation scopically. Porosity is generally easier to changes number and arrangement of disloca- observe on as-polished specimens than on tions (crystal defects) in the metal on an polished and etched ones. Figure 17(a) atomic scale. This dislocation substructure is shows a typical example of shrinkage poros- best characterized using TEM. Optical met- ity. allography can be used to supplement TEM Other structural defects, such as inclu- through characterization of the grain size and sions and stringers (Fig. 17b and 17c), can anisotropy in grain shape and distribution. observed, and L is the length of line on the also be observed microscopically in Microstructural changes due to annealing image. as-polished specimens. Such defects as those may be studied using TEM or optical micros- The microstructure of the cold-rolled cop- shown in Fig. 17 can serve as failure- copy. The most important structural changes per alloy shown in Fig. 15 differs from that initiation sites in metals and alloys; there- that occur during annealing are recovery, of the annealed metal. Rolling elongated the fore, characterization of their size, shape, recrystallizafion, and grain growth. grains in the rolling direction and flattened and distribution is necessary to establish Recovery is the rearrangement and anni- the grains in directions transverse or normal material properties and engineering reliabil- hilation of imperfections (primarily vacan- to the rolling directions. This change can ity. Quality-assurance programs frequently cies and interstitials) within each grain of a 308 / Metallographic Techniques

cold-worked polycrystalline component. Be- amount of cold work, the finer the grain size SELECTED REFERENCES cause recovery deals mainly with point de- (Fig.20). Because grain boundaries are a fects, any microstructural observations of it crystalline defect, continued annealing will * J.L. McCall and W.M. Mueller, Ed., are difficult, and optical microscopy cannot cause this array of grains to be unstable, and Metallographic Specimen Preparation, be used because of its limited resolution. grain growth will take place. Grain growth Plenum Press, 1974 Recrystallization is the formation of new in a recrystallized specimen decreases the * Metallography and Microso-uctures, Vol strain-free grains within the previously cold- grain-boundary surface area to specimen 9, 9th ed., Metals Handbook, American worked (strained) grains. The initial stages volume ratio because the average grain size Society for Metals, 1985 of recrystallization occur on such a fine scale increases as grain growth takes place. The • Metallography, Nondestructive Testing, that TEM is necessary; however, oPtiCal rate of grain growth depends on temp- Vol 03.03, Annual Book of ASTM Stan- metallography can be readily used to study erature and time. Detailed information on dards, ASTM, Philadelphia, 1984 most of the recrystallization. The size of the all of these types of structures and on the • G. Petzow, Metallographic Etching, recrystallized grains depends on the amount metallography and microstructures of specific American Society for Metals, 1978 of cold working of the specimen before the metals and alloys is available in Volume 9 • G.F.Vander Voort, Metallography: Prin- recrystallization anneal. The greater the of the 9th Edition of Metals Handbook. ciples and Practice, McGraw-Hill, 1984