: :

: :E : : 9- : 1779

Metals are an aggregation of atoms that, apart from mercury, are solid at room temperature.These atoms are held together by "metallic bonds" that result from sharing available electrons.A negative electron bond pervades the structure, and heat and electricity can be conducted through the metal by the free movement of electrons. The negative electron bond surrounds the positive ions that make up the crystal structure of the metal. There are three common types of lattice structure that metals belong to: close-packed hexagonal, face-centered cubic, and body-centered cubic. Close-Packed Hexagonal (CPH) Models of crystal structures can be made up of spheres stacked in close-packed layers. Two arrangements are possible, one being hexagonal and the other cubic in basic structure. In the close-packed hexagonal system the spheres repeat the same position every second layer (ABABAB . . . ; Fig 1). Face-Centered Cubic (FCC) Layers can be built up so that the third layer of spheres does not occupy the same position as the spheres in the first row; the structure repeats every third layer (ABCABCABC. .. ) . FCC metals tend to be ductile (i.e., can be mechanically deformed, drawn out into wire, or hammered into sheet).Examples are lead, aluminium, copper, silver, gold, and nickel (Fig. 2). Body-Centered Cubic (BCC) Another common type found in many metals, the body-centered cubic structure, is less closely packed than the FCC or CPH structures and has atoms at the corners and one atom at the center of the cube. The atoms at the corners are shared with each adjoining cube (Fig. 3). Other metals important in antiquity have entirely different lattice structures; for example, arsenic, antimony, and bismuth are rhombohedral, and ordinary tin is body-centered tetragonal. Metals are crystalline solids under normal conditions of working and melting. However, if a metal is cooled very rapidly, as in splat cooling, the normal crystalline structure can be suppressed. In splat cooling, metal droplets are cooled very quickly between chilled metal plates and the structure that develops is similar to glass-a random arrangement of atoms rather than a crystalline array. In the usual crystalline state, metal will consist of a number of discrete grains. The metal is then referred to as being polycrystalline. An important property of metals is that they undergo plastic deformation when stretched or hammered. This is illustrated by a stress-strain diagram using Young's Modulus (YM ) : YM=stress/strain=load applied/cross-sectional area /change in length/original length Before plastic deformation occurs, materials deform by the elastic movement of atoms that hold the structure together. This elastic deformation occurs in metals and in other materials, such as glass, which have no ability to deform at room temperature. Glass will stretch by elastic deformation and then break. Metals can deform plastically because planes of atoms can slip past each other to produce movement. This kind of movement cannot take place in a glassy structure. When metals such as pure copper or iron are stretched they will break or fracture, but only after a certain amount of plastic deformation has occurred (see Fig 4) .

Figure 1.

Figure 4.Relationship between stress and strain

Figure 2.

Figure 3.

Hardness
The hardness of a metal is measured by its resistance to indentation. The metal is indented under a known load using a small steel ball (as in the Brinell test) or a square-based diamond pyramid (as in the Vickers test). In the Vickers test the result is given as the Diamond Pyramid Number DPN (or Hv) .

Dislocations
It is rare for crystals to have a perfect atomic structure;there are usually imperfections present. In metals, edge dislocations and screw dislocations are the most important faults (see Figs. 6, 7). These crystal faults enable deformation to take place at lower applied stress by slip than would be possible if the lattice structure was perfect. When a metal is deformed then slip takes place until a tangle of dislocations builds up which prevents any further working (i.e., dislocation entanglement).As metal is worked, slip planes become thicker and immobile. If the metal is worked further,it must be annealed (heated up to bring about recrystallization) . Before annealing, the metal can be said to be work-hardened. Work-hardening is accomplished by, for example, hammering at room temperature. This increases the metal hardness value, but decreases ductility.

Edge dislocation

Edge dislocation

Movement through a crystal

The structure of carbon steels

The type of srructure seen in Figure 5 1 , below, in low-carbon steel, could arise from heat treatment in the + y region of the phase diagram. At room temperature, the structure will be ferrite + pearlite, and after heating this will become: +( +Fe3C) + and after cooling the gamma phase will revert to: ( +Fe3C)

During cooling some of the ferrite will take a Widmanstatten structure. ferrite()

slighty Widmanstatten () Figure 51. Partialy Widmanstatten steel. If the cooling rate is very fast then martensite can form. Even if the cooling rate is too slow for this transformation to occur the proportions of pearlite and ferrite in the solid will not be present in equilibrium amounts. The presence of more pearlite than ferrite will lead to an overestimation of the carbon content.The structure in Figure 52 shows grain boundary cementite (Fe3C) with a subgrain structure of cementite + pearlite. How can this type of structure arise? When cooling down an alloy from the gamma (y) region, the alloy contains more than 0.8% carbon. Upon decomposition of the gamma phase, Fe3C will be formed at the original austenitic grain boundaries after cooling from temperatures above 900 C. If this alloy, typically with about 1 . 2% carbon content, is cooled down and then subsequently reheated to about 800 C (in the gamma + [alpha + Fe3C] region), then more Fe3C will be formed on a finer scale, while the original austenitic grain boundaries will be preserved, so that on subsequent cooling, the gamma phase will decompose to alpha + Fe3C.

Figure 52

The section in Figure 53 shows small grain size in the cutting edge and a larger grain size in the body of a knife. There is some tempered martensite on the edge.

Small grains at edge large grains in body Figure 53 This type of structure cannot originate from heating the tip of the blade followed by quenching. If the tip was made by heat treatment from the same metal then the grain structure would be larger. This indicates that the cutting edge must have been pur in at some point during manufacture; it could not have been made in one piece. When etching quenched structures, some areas of the martensite may etch rather darker than others. This can be due to tempering of the martensite or, if fuzzy nodular shapes appear, it could be due to the presence of troosite. Troosite patches appear because the cooling rate is not quite fast enough to produce a martensitic structure. There is often confusion between the terms troosite and sorbite. It is better to reserve the term sorbite for srructures resulting from quenching followed by heat treatment (for example, martensite going to sorbite upon heating) . Troosite should be used for srructures resulting from the rapid cooling of a carbon steel, but not fast enough for the production of martensite. In the banded structure of the quenched sword blade in Figure 54, the question arises as to why, if the whole structure has been quenched, are there some areas of martensite and some that are apparenrly only a mixture of ferrite + pearlite. There are two possible reasons for the structural differences. First, the difference may be due to carbon content. The amount of carbon present will affect the production of martensite when the heated alloy is quenched. It is possible that the carbon content in the fine ferrite + pearlite regions is not sufficient to produce martensite. Second, the shift may be due to other alloy elements in the regions concerned, which would be apparent with further analysis. Similarly, if in a homogeneous specimen the cooling rate is not sufficiently high to produce a totally martensitic structure, then it is possible to get a series of transitions between martensite, troosite, ferrite, and fine pearlite, with incipiently Widmanstatten ferrite between the pearlite regions. In Figure 5 5 , if the alloy is of composition A,then the structure will consist of alpha at the grain boundaries with pearlite infill, but if the alloy is of composition B , then the structure will be cementite at the grain boundaries with an infill of pearlite. Small changes in the direction of B make little difference to the visible amount of cementite at the grain boundaries, but a small movement in the direction of A makes a large difference in the amount of alpha at the grain boundaries.

0.8%C Figure 55

Metallography :

A micrograph of bronze revealing a cast dendritic structure. Metallography is the study of the physical structure and components of metals, typically using microscopy.Ceramic and polymeric materials may also be prepared using metallographic techniques, hence the terms ceramography, plastography and, collectively, materialography.

Preparing metallographic specimens:

The surface of a metallographic specimen is prepared by various methods of grinding, polishing, and etching. After preparation, it is often analyzed using optical or electron microscopy. Using only metallographic techniques, a skilled technician can identify alloys and predict material properties. Mechanical preparation is the most common preparation method. In a series of steps, successively finer abrasive particles are used to remove material from the sample surface until the desired surface quality is achieved. Many different machines are available for doing this grinding and polishing, able to meet different demands for quality, capacity, and reproducibility. A systematic preparation method is easiest way to achieve the true structure. Sample preparation must therefore pursue rules which are suitable for most materials. Different materials with similar properties (hardness and ductility) will respond alike and thus require the same consumables during preparation. Metallographic specimens are typically "mounted" using a hot compression thermosetting resin. In the past, phenolic thermosetting resins have been used, but modern epoxy is becoming more popular because reduced shrinkage during curing results in a better mount with superior edge retention. A typical mounting cycle will compress the specimen and mounting media to 4,000 psi (28 MPa) and heat to a temperature of 350 F (177 C). When specimens are very sensitive to temperature, "cold mounts" may be made with a two-part epoxy resin. Mounting a specimen provides a safe, standardized, and ergonomic way by which to hold a sample during the grinding and polishing operations. After mounting, the specimen is wet ground to reveal the surface of the metal. The specimen is successively ground with finer and finer abrasive media. Silicon carbide abrasive paper was the first method of grinding and is still used today. Many metallographers, however, prefer to use a diamond grit suspension which is dosed onto a reusable fabric pad throughout the polishing process. Diamond grit in suspension might start at 9 micrometres and finish at one micrometre. Generally, polishing with diamond suspension gives finer results than using silicon carbide papers (SiC papers), especially with revealing porosity, which silicon carbide

paper sometimes "smear" over. After grinding the specimen, polishing is performed. Typically, a specimen is polished with a slurry of alumina, silica, or diamond on a napless cloth to produce a scratch-free mirror finish, free from smear, drag, or pull-outs and with minimal deformation remaining from the preparation process. After polishing, certain microstructural constituents can be seen with the microscope, e.g., inclusions and nitrides. If the crystal structure is non-cubic (e.g., a metal with a hexagonal-closed packed crystal structure, such as Ti or Zr) the microstructure can be revealed without etching using crossed polarized light (light microscopy). Otherwise, the microstructural constituents of the specimen are revealed by using a suitable chemical or electrolytic etchant. A great many etchants have been developed to reveal the structure of metals and alloys, ceramics, carbides, nitrides, and so forth. While a number of etchants may work for a given metal or alloy, they generally produce different results, in that some etchants may reveal the general structure, while others may be selective to certain phases or constituents.

Analysis :
Prepared specimens should be examined after etching with the unaided eye to detect any visible areas that respond differently to the etchant as a guide to where the microscopical examination should be employed. Light optical microscopy (LOM) examination should always be performed prior to any electron metallographic (EM) technique, as these are more time-consuming to perform and the instruments are much more expensive. Further, certain features can be best observed with the LOM, e.g., the natural color of a constituent can be seen with the LOM but not with EM systems. Also, image contrast of microstructures at relatively low magnifications, e.g., <500X, is far better with the LOM than with the scanning electron microscope (SEM), while transmission electron microscopes (TEM) generally cannot be utilized at magnifications below about 2000 to 3000X. LOM examination is fast and can cover a large area. Thus, the analysis can determine if the more expensive, more time-consuming examination techniques using the SEM or the TEM are required and where on the specimen the work should be concentrated. Light microscopes are designed with either specimen placement of the polished surface on the stage upright or inverted. Each type has advantages and disadvantages. Most LOM work is done at magnifications between 50 and 1000X. However, with a good microscope, it is possible to perform examination at higher magnifications, e.g., 2000X, and even higher, as long as diffraction fringes are not present to distort the image. However, the resolution limit of the LOM will not be better than about 0.2 to 0.3 micrometer. Special objectives can be obtain to use the LOM at magnifications below 50X, which can be very helpful when examining the microstructure of cast specimens where greater spatial coverage in the field of view may be required to observe features such as dendrites. Besides considering the resolution of the optics, one must also maximize visibility by maximizing image contrast. A microscope with excellent resolution may not be able to image a structure, that is there is no visibility, if image contrast is poor. Image contrast depends upon the quality of the optics, coatings on the lenses, and reduction of flare and glare; but, it also requires proper specimen preparation and good etching techniques. So, obtaining good images requires maximum resolution and image contrast. Most LOM observations are conducted using bright field (BF) illumination where the image of any flat feature perpendicular to the incident light path is bright, or appears to be white. But, other illumination methods can be used and, in some cases, may provide superior images with greater detail. Dark field (DF), although not used much today, provides high contrast images and actually

greater resolution than bright field. In dark field, the light from features perpendicular to the optical axis is blocked and appears dark while the light from features inclined to the surface, that look dark in BF, appear bright, or "self luminous" in DF. Grain boundaries, for example, are more vivid in DF than BF. Polarized light (PL) is very useful when studying the structure of metals with noncubic crystal structures (mainly metals with hexagonal close-packed (hcp) crystal structures). If the specimen is prepared with minimal damage remaining at the surface, the structure can be seen vividly in crossed polarized light (the optic axis of the polarizer and analyzer are 90 degrees to each other, i.e., crossed). In some cases, an hcp metal can be chemically etched and then examined more effectively with PL. Tint etched surfaces, where a thin film (such as a sulfide, molybdate, chromate or elemental selenium film) is grown epitaxially on the surface to a depth where interference effects are created when examined with BF producing color images, can be improved with PL. If it is difficult to get a good interference film with good coloration, the colors can be improved by examination in PL using a sensitive tint (ST) filter. Another useful imaging mode is differential interference contrast (DIC), where the most common, and best detail, is obtained with a system designed by Nomarski. DIC converts minor height differences on the plane-of-polish, invisible in BF, into visible detail. The detail in some cases can be quite striking and very useful. If an ST filter is used along with the Wollaston prism, color is introduced. The colors are controlled by the adjustment of the Wollaston prism, and have no specific physical meaning, per se. But, visibility may be better. DIC has largely replaced the older oblique illumination (OI) technique available on reflected light microscopes prior to about 1975. In OI, the vertical illuminator is offset from perpendicular producing shading effects that reveal height differences. But, this procedure reduces resolution and yields uneven illumination across the field of view. Nevertheless, OI was useful when people needed to know if a second phase particle was standing above or was recessed below the plane-of-polish, and is still available on a few microscopes. OI can be created on any microscope by placing a piece of paper under one corner of the mount so that the planeof-polish is no longer perpendicular to the optical axis. If a specimen must be observed at higher magnification, it can be examined with a scanning electron microscope (SEM), or a transmission electron microscope (TEM). When equipped with an energy dispersive spectrometer (EDS), the chemical composition of the microstructural features can be determined. The ability to detect low-atomic number elements, such as C, O and N, depends upon the nature of the detector used. But, quantification of these elements by EDS is difficult and their minimum detectable limits are higher than when a wavelength-dispersive spectrometer (WDS) is used. But quantification of composition by EDS has improved greatly over time. The WDS system has historically had better sensitivity, that is ability to detect low amounts of an element, and ability to detect low-atomic weight elements, and better quantification of compositions, compared to EDS, it was slower to use. Again, in recent years, the speed required to perform WDS analysis has improved substantially. Historically, EDS was used with the SEM while WDS was used with the electron microprobe analyzer (EMPA). But, today EDS and WDS is used with both the SEM and the EMPA. However, a dedicated EMPA is not as common to find in laboratories as an SEM. Characterization of microstructures has also been performed using x-ray diffraction (XRD) techniques for many years. XRD can be used to determine the percentages of various phases present in a specimen if they have different crystal structures. For example, the amount of retained austenite in a hardened steel is best measured using XRD (ASTM E 975). If a particular phase can be chemically extracted from a bulk specimen, it can be identified using XRD based on the crystal structure and lattice dimensions. This work can be complemented by EDS and/or WDS analysis where the chemical composition is quantified. But EDS and WDS are difficult to apply to particles <2-3 micrometers in diameter. For smaller particles, diffraction techniques can be

performed using the TEM for identification and EDS can be performed on small particles if they are extracted from the matrix using replication methods to avoid detection of the matrix along with the precipitate.

Quantitative metallography :
A number of techniques exist to quantitatively analyze metallographic specimens. These techniques are valuable in the research and production of all metals and alloys and nonmetallic or composite materials. Microstructural quantification is performed on a prepared, two-dimensional plane through the three-dimensional part or component. Measurements may involve simple metrology techniques, e.g., the measurement of the thickness of a surface coating, or the apparent diameter of a discrete second-phase particle, e.g., spheroidal graphite in ductile iron; or, it may require application of stereological methods to assess matrix and second-phase structures. Stereology is the field of taking 0-, 1- or 2dimensional measurements on the two-dimensional sectioning plane and estimating the amount, size, shape or distribution of the microstructure in three dimensions. These measurements may be made using manual procedures with the aid of templates overlaying the microstructure, or with automated image analyzers. In all cases, adequate sampling must be made to obtain a proper statistical basis for the measurement. Efforts to eliminate bias is required. Some of the most basic measurements include determination of the volume fraction of a phase or constituent, measurement of the grain size in polycrystalline metals and alloys, measurement of the size and size distribution of particles, assessment of the shape of particles and spacing between particles. ASTM Committee E-4 on Metallography, and some other national and international standards organizations, have developed standard test methods describing how to characterize microstructures quantitatively. For example, the amount of a phase or constituent, that is, its volume fraction, is defined in ASTM E 562; manual grain size measurements are described in ASTM E 112 (equiaxed grain structures with a single size distribution) and E 1182 (specimens with a bi-modal grain size distribution); while ASTM E 1382 describes how any grain size type or condition can be measured using image analysis methods. Characterization of nonmetallic inclusions using standard charts is described in ASTM E 45 (historically, E 45 covered only manual chart methods and an image analysis method for making such chart measurements was described in ASTM E 1122. The image analysis methods are currently being incorporated into E 45). A stereological method for characterizing discrete second-phase particles, such as nonmetallic inclusions, carbides, graphite, etc., is presented in ASTM E 1245.

Hot mounting: The specimens are placed in the mounting press, and the resin is added. The specimens are mounted under heat and high pressure.

Cold mounting: The specimens are placed in a mounting cup and mounting material is then poured over the specimens. A vacuum impregnation unit (photo) is used for mounting of porous materials.

The MD-System is an example of a reusable pad for use with diamond suspension. A single magnetic platen is positioned on the grinding and polishing machine to support the preparation pads.

macro etched copper disc

Metallography offers one of the most useful means for the examination of ancient metals. It is the study of polished sections of metallic materials using a special microscope that reflects light passing through the objective lens onto the specimen surface. The reflected light passes back through the objective to the eyepiece, which enables the surface structure of the section to be studied (see Fig. 80 for a typical example of a metallurgical microscope) .Reflected light microscopy is used for metallographic examination because metals cannot transmit light in thin sections in the same way as ceramic or mineral materials. Apart from gold foil, which, if very thin, can transmit a greenish light through grain boundaries, metals are opaque substances. It is usually necessary to take a sample, which can be quite small, from the object being studied. It is possible to polish a small area of a relatively flat surface on an object using a minidrill and fine polishing discs, but this method is quite difficult. With many antiquities, useful information can be obtained from samples as small as 1 mm 3 , which can be removed from the object with a minimal amount of damage. The sample itself need not be metallic: corrosion products, paint layers, core material from castings, niello, deteriorated enamels, patinations, and an array of other compositional materials are of considerable interest. Examination with a metallurgical microscope is often a useful first step in characterizing a particular component of an object. Metallography is, then, an important tool that may provide clues to the fabrication technology of the object or may assist in answering questions that arise during the treatment of an object by a conservator. It may be possible to provide information on the following range of topics: 1. The manufacturing processes used to produce the object. 2. The thermal history of the object. Quenching and tempering processes may produce definite changes in the microstructure that can be seen in section. 3. The nature of the metal or alloy employed to make the object. 4. The nature of corrosion products. Much useful and varied information can be obtained from corroded fragments or pieces of corrosion crusts. The principal difficulties with metallography applied to ancient material lie in the problems associated with sampling the object and the selectionof the sample. If any damage is to be caused,the owner should always be consulted first so that necessary permission is obtained for the work. It should b e made clear to the owner whether the sample can be returned when the examination is completed, and the kind of information that the sample is expected to provide. The owner should receive a copy of any written report that is prepared as a result of the examination, together with any photomicrographs of the structure, clearly identified with the magnification of the print, the etching solution used, and the laboratory number assigned to the sample. It may be necessary to examine carefully a group of objects in the museum or laboratory before coming to a decision as to the number of samples to be taken or the objects from which it may be permissible to take a sample. If the damage is such that it can be repaired by filling the area concerned, the owner should be given the option of deciding whether or not to have the object gap-filled. It is preferable to fill small holes or missing corners with a synthetic resin colored to match the surface color of the object. A fairly viscous epoxy resin can be used for this purpose and can be colored with powder pigments. In many museum collections there are examples where large slices, pieces, or drillings have been taken from metallic antiquities causing gross damage to the objects concerned. Missing parts are sometimes filled with unsuitable materials such as Plasticine (normally a mixture based on putty, whiting, and linseed oil) , which usually creates severe corrosion of exposed metal surfaces over a period of years in the museum collection. Sulfur-containing fillers such as Plasticine should never be used for mounting, display, or gap-filling of metallic objects. Great improvements in metallographic techniques over the last 20 years mean that it is no longer necessary to remove as much sample bulk from the object as many earlier investigations

required. The extent of the damage is therefore greatly reduced; indeed, with much corroded metalwork or fragmentary objects, the loss may be insignificant and the resulting information may be very important indeed. The only difficulty with taking small samples is the problem of the representative nature of the sample compared with the overall structure of the object. It does not follow, for example, that because the area sampled shows an undistorted dendritic structure (indicating that the piece was cast) that the whole of the object will reveal the same structure. Obviously there are cases in which this difficulty would be very unlikely to occur: worked, thin sheet-metal, pieces of wire, small items of jewelry, and so on, but with artifacts such as axes, knives, swords, large castings, etc., it may present a very difficult problem. If a complete interpretation of the metallographic structure of an artifact such as an axe is required, then it is almost essential to take two samples from the axe (Fig. 8 1 ) . Another example is that of a piece of wire or rod where both the longitudinal and transverse sections are of potential interest. Here the two sections can be obtained from a single sample by cutting the wire into two pieces and mounting them in a mold (Fig. 82). The procedures for preparing metallographic samples are as follows:

wire: longitudinal Figure 81 Figure 82

Mold wire : transverse