Buehler Guide Complete

2

INTRODUCTION

Proper specimen preparation is essential if thetrue microstructure is to be observed, identified,documented and measured. Indeed, whenmetallographers have difficulty seeing or mea-suring microstructures, these problems canalmost always be traced to improper specimenpreparation procedures, or failure to properlyexecute the preparation steps. When a speci-men is prepared properly, and the structureis revealed properly, or a phase or constituent

of interest is revealed selectively and withadequate contrast, then the actual tasks ofexamination, interpretation, documentation andmeasurement are generally quite simple. Expe-rience has demonstrated that getting the requiredimage quality to the microscope is by far thegreatest challenge. Despite this, the specimenpreparation process is often treated as a trivialexercise. However, specimen preparation qual-ity is the determining factor in the value of theexamination. This is in agreement with the clas-sic computer adage, “garbage in = garbage out.”

Written by George F. Vander Voort, Director, Research & Technology, with contributions from Wase Ahmed, George Blann, Paula Didier, Don Fisher,Scott Holt and Gabe Lucas.

Introduction

4

GOALS OF SPECIMENPREPARATION

The preparation procedure and the pre-pared specimen should have the followingcharacteristics:

• Deformation induced by sectioning, grindingand polishing must be removed or be shallowenough to be removed by the etchant.

• Coarse grinding scratches must be removed;very fine polishing scratches may be tolerablefor routine production work.

• Pull out, pitting, cracking of hard particles,smear, and other preparation artifacts, mustbe avoided.

• Relief (i.e., excessive surface height variationsbetween structural features of different hard-ness) must be minimized; otherwise portionsof the image will be out of focus at high magni-fications.

• The surface must be flat, particularly atedges (if they are of interest) or they cannotbe imaged.

• Coated or plated surfaces must be kept flat ifthey are to be examined, measured or photo-graphed.

• Specimens must be cleaned adequatelybetween preparation steps, after preparation,and after etching.

• The etchant chosen must be either general orselective in its action (reveal only the phase orconstituent of interest, or at least producestrong contrast or color differences betweentwo or more phases present), depending uponthe purpose of the investigation, and must pro-duce crisp, clear phase or grain boundaries,and strong contrast.

If these characteristics are met, then the truestructure will be revealed and can be interpreted,measured and recorded. The preparation methodshould be as simple as possible, should yieldconsistent high quality results in a minimum oftime and cost and must be reproducible.

Preparation of metallographic specimens [1-3]generally requires five major operations: (a) sec-tioning, (b) mounting (optional), (c) grinding,(d) polishing and (e) etching (optional).

Goals of Specimen Preparation

3

SAMPLING

The specimens selected for preparation mustbe representative of the material to be exam-ined. Random sampling, as advocated bystatisticians, can rarely be performed by metal-lographers. A good exception is the testing offasteners where a production lot can be sampledrandomly. But, a large forging or casting cannotbe sampled at random as the component wouldbe rendered commercially useless. Instead, testlocations are chosen systematically based onease of sampling. Many material specificationsdictate the sampling procedure. In failure analy-sis studies, specimens are usually removed tostudy the origin of the failure, to examine highlystressed areas, to examine secondary cracks,and so forth. This, of course, is not random sam-pling. It is rare to encounter excessive samplingas testing costs are usually closely controlled.Inadequate sampling is more likely to occur.

Normally, a specimen must be removed from alarger mass and then prepared for examination.This requires application of one or moresectioning methods. For example, in a manufac-turing facility, a piece may be cut from a bar ofincoming metal with a power hacksaw, or anabrasive cutter used dry, i.e., without a coolant.The piece is then forwarded to the laboratorywhere it is cut smaller to obtain a size moreconvenient for preparation. All sectioningprocesses produce damage; some methods(such as flame cutting or dry abrasive cutting)produce extreme amounts of damage. Traditionallaboratory sectioning procedures, using abrasivecut-off saws, introduce a minor amount of dam-age that varies with the material being cut andits thermal and mechanical history. It is gener-ally unwise to use the original face cut in the shopas the starting point for metallographic prepara-tion as the depth of damage at this location maybe quite extensive. This damage must beremoved if the true structure is to be revealed.However, because abrasive gr inding andpolishing steps also produce damage, where thedepth of damage decreases with decreasingabrasive size, the preparation sequence must becarefully planned and performed; otherwise,preparation-induced artifacts will be interpretedas structural elements.

Many metallographic studies require more thanone specimen and sectioning is nearly alwaysrequired to extract the specimens. A classicexample of multiple specimen selection isthe evaluation of the inclusion content of steels.One specimen is not representative of the wholelot of steel, so sampling becomes important.ASTM standards E 45, E 1122 and E 1245 giveadvice on sampling procedures for inclusion stud-ies. To study grain size, it is common to use asingle specimen from a lot. This may or may notbe adequate, depending upon the nature of thelot. Good engineering judgment should guidesampling, in such cases. In many cases, a prod-uct specification may define the procedurerigorously. Grain structures are not alwaysequiaxed and it may be misleading to select onlya plane oriented perpendicular to the deforma-tion axis, a “transverse” plane, for such a study.If the grains are elongated due to processing,which does happen, the transverse plane willusually show that the grains are equiaxed inshape and smaller in diameter than the true grainsize. To study the effect of deformation on thegrain shape of wrought metals, a minimum of twosections are needed — one perpendicular to, andthe other parallel to, the direction of deformation.

HELPFUL HINTS FOR SAMPLING

• When developing a sampling plan for anunfamiliar part or component, determinethe orientation of the piece relative tothe original wrought or cast startingmaterial and prepare sections on thelongitudinal, transverse and planar sur-faces, or radial and transverse surfaces,to reveal the desired information. Re-member that the microstructure maylook much more homogeneous than itis on a transverse plane compared to alongitudinal or planar surface. If thestarting material was cold worked, or notfully recrystallized after hot working,grain structures will appear to beequiaxed and smaller on a transverseplane, but non-equiaxed and larger ona longitudinal or planar surface.

Sampling

5

SECTIONING

Bulk samples for subsequent laboratory section-ing may be removed from larger pieces usingmethods such as core drilling, band or hack-saw-ing, flame cutting, or similar methods. However,when these techniques are used, precautionsmust be taken to avoid alteration of the micro-structure in the area of interest. Laboratoryabrasive-wheel cutting is recommended to es-tablish the desired plane of polish. In the case ofrelatively brittle materials, sectioning may beaccomplished by fracturing the specimen at thedesired location.

Abrasive-Wheel CuttingThe most commonly used sectioning device inthe metallographic laboratory is the abrasivecut-off machine, Figure 1. All abrasive-wheelsectioning should be performed wet. An ampleflow of coolant, with an additive for corrosionprotection and lubrication, should be directed intothe cut. Wet cutting will produce a smoothsurface finish and, most importantly, will guardagainst excessive surface damage causedby overheating. Abrasive wheels should beselected according to the manufacturer’srecommendations. Table 1 summarizes Buehler’srecommendations for our machines. Specimensmust be fixtured securely during cutting, andcutting pressure should be applied carefully toprevent wheel breakage. Some materials, suchas CP (commercial purity) titanium, Figure 2, arequite prone to sectioning damage.

Wheels consist of abrasive particles, chiefly alu-mina or silicon carbide, and filler in a bindermaterial that may be a resin, rubber, or a mix-ture of resin and rubber. Alumina (aluminumoxide) is the preferred abrasive for ferrous al-loys and silicon carbide is the preferred abrasivefor nonferrous metals and minerals. Wheelshave different bond strengths and are recom-mended based on the suitability of their bondstrength and abrasive type for the material to besectioned. In general, as the hardness of a ma-terial increases, abrasives become dull morequickly, and the binder must break-down andrelease the abrasives when they become dullso that fresh abrasive particles are available tomaintain cutting speed and efficiency. Conse-quently, these wheels are called “consumable”wheels because they wear away with usage. Ifthey do not wear at the proper rate, dull abra-sives will rub against the region being cutgenerating heat and altering the existing truemicrostructure. If this heat becomes excessive,it can lead to grain or particle coarsening, soft-ening or phase transformations, and in extremecase, to burning or melting. Different materialshave different sensitivities to this problem. But,the need to balance the wheel break-down ratewith the hardness of the piece being sectioned,produces the various recommendations listedfor cutting different materials and metals withdifferent hardnesses, such as steels.

The size of the cut-off machine also affects wheelparameters. As the diameter of the wheel isincreased, to permit sectioning of larger speci-mens, the wheel thickness is generally increased.Even so, for a given diameter wheel, there maybe a range of thicknesses available. Thickerwheels are stronger but remove more materialin the cut. This may, or may not, be a problem.

Sectioning

Figure 1. Delta Automatic Orbital Cutter

Figure 2. Cutting damage (top) and a “burr” aftersectioning of an annealed CP titanium specimen(mod. Weck’s reagent, 100X, polarized light plussensitive tint).

6

But, thicker wheels do generate more heat dur-ing cutting than a thinner wheel, everything elsebeing held constant. Consequently, for caseswhere the kerf loss or heat loss must be mini-mized, select the thinnest available wheel of theproper bond strength and abrasive. Buehler’sAcu-Thin Cut-off Wheels offer the thinnest pos-sible wheels for delicate abrasive sectioning.Buehler also has diamond cut-off blades withresin bonded or Rimlock1 metal bonded bladeforms in sizes from 8-inch (203 mm) to 12-inch(305 mm) diameters. Resin-bonded diamondblades are ideal for cutting very hard cementedcarbide specimens; Rimlock or continuous rimblades are recommended for cutting rocks andpetrographic specimens.

Historically, the most common cutter design hasbeen the so-called “chop” cutter. Basically, theblade is attached to a motor and the operatorpulls a handle to drive the blade downward intothe work piece. Because of the design, the blademoves through an arc as it is pulled downward,Figure 3a. For efficient cutting, the piece must

be oriented to minimize the contact area withthe wheel. For small parts, this is generally easyto do, but for larger parts, it may not be possibleto orient the specimen properly for optimal cut-ting. When a round wheel is used to cut a roundbar in chop mode, the contact area is initiallyvery small. As the cut progresses, the cut wid-ens until the maximum diameter is reached. Afterthis, the contact area decreases until sectioningis completed. Under application of a constantload, the pressure on the abrasive particles inthe cut decreases as the contact area increases.If the pressure applied to the grains is inadequatefor cutting, then heat is generated which maynot be dissipated by the coolant, which causesdeformation damage, phase changes andpossibly burning or melting.

A chop cutter may be set up to pulse as the loadis applied, that is, the wheel is fed into thespecimen, and then feeding is halted momen-tarily, Figure 3b. Pulsing causes the wheel to bestress shocked which removes both dull andsharp abrasive from the wheel as new abrasive

1Rimlock is a registered trademark of Felker Operations.

Table 1. Buehler’s Abrasive Cutting Recommendations

Sectioning

Available DiametersRecommended Use Bond Abrasive (Inches) (mm)

General Usage Blades

Tools Steels 60 HRC and Rubber Resin Al2O3 9, 10, 12, 230, 250, 300,Above Carburized Steels 14, 16, 18 350, 400, 455

Hard Steel 50 HRC Rubber Resin Al2O3 9, 10, 12, 230, 250, 300,14, 16, 18 350, 400, 455

Medium Hard Steel Rubber Resin Al2O3 9, 10, 12, 230, 250, 300,35-50 HRC 14, 16, 18 350, 400, 455

Soft or Annealed Steel Rubber Al2O3 9, 10, 12, 230, 250, 300,15-35 HRC 46-90 HRB 14, 16, 18 350, 400, 455

Medium Hard Nonferrous Materials, Rubber SiC 9, 10, 12, 230, 250, 300,Uranium, Titanium, Zirconium 14, 16, 18 350, 400, 455

Soft Nonferrous Materials Rubber SiC 9, 10, 12, 230, 250, 300,Aluminum, Brass, etc. 14, 16, 18 350, 400, 455

Thin Blades to Minimize Kerf Loss and Cutting Deformation

Tool, Hard Steel, Rubber Al2O3 5, 9, 10*, 12 130, 230,≤ 45 RC 250*, 300

Medium Hard, Soft Steel Rubber Al2O3 5, 7, 9, 130, 180, 230,≥ 45 HRC 10, 12, 14 250*, 300, 350

Hard or Soft Nonferrous Materials Rubber SiC 7 180

*Rubber Resin BondRefer to Buehler’s Consumables Buyers Guide for ordering information, and exact dimensions of arbor sizes, outer diameter, and thickness ofBuehler Metabrase, Delta, and Acu-Thin Abrasive Cut-off Wheels

7

Wheel Path

Sample

Chop Cutting with PulsingWheel contact still governed by specimensize. The pulsing action gives a shock tothe wheel causing abrasives to breakaway. Wheel wear is generally high withthis method.

3b

Chop CuttingThe traditional form of machineoperation. Wheel contact arc isgoverned by specimen size. Generallya struggle with large/difficult parts.

3a

OscillationThe wheel is reciprocated over asmall distance. This minimizes the contactarc. Available only on manual labor-intensive cutters.

3c

Traverse and IncrementThe wheel contact arc can be preciselycontrolled via depth increment. Thetraverse stroke must always exceed thepart length to avoid a change in wheelcontact arc area. Machine needs to be setfor each part. Action is slow.

3d

OrbitalSimilar in action to traverse and incrementbut on a curved path. Simpler and quickerin operation. Part size is irrelevant as theorbital action produces a minimum contactarc area during cutting.

3e

CUTTING STYLE AND WHEEL PATH

is exposed to the cut. While a better cut may beobtained, cutting time and wheel wear areincreased.

Optimal cutting, with the least damage to thespecimen, is obtained by keeping the pressureon each abrasive particle in the wheel constantand as low as possible by keeping the contactarea between wheel and specimen constant andlow. This idea is called the minimum area of con-tact cutting (MACC) method. Oscillation stylecutters, such as the Oscillimet, minimize the con-tact area by oscillating perpendicular to the cut

contact area, Figure 3c. This approach has theadvantage of permitting larger size pieces to becut properly compared to a chop cutter with thesame wheel diameter.

Another option for MACC is the “traverse andincrement” cutter design, Figure 3d. A strip of ma-terial is removed by moving either the wheel orthe work piece perpendicular to the cut direction.After this, the cutter must move back to the sidewhere the cut began. The wheel is dropped downa fixed increment and another strip is removed.In this approach, the wheel must move to the

Figure 3. Illustration of cutter types and the wheel path during sectioning

Sectioning

8

end of the stroke, return to the starting point,and then be moved an increment downwards ineach cycle. Consequently, cut times are longer.The wheel traverse length must be greater thanthe specimen width. This machine must be pro-grammed for each size part to obtain cutting.

Orbital cutting, Figure 3e, combines the bestfeatures of chop, oscillation and traverse-and-increment cutters, while minimizing cutting time.As the wheel is fed into the specimen, the arboris also rotated in a small ellipse planar with thewheel permitting the wheel to cut a small strip ofmaterial from the specimen. But no time is lost,as with the traverse and increment cutter, as thewheel is ready to cut on the original side. Withthe elliptical orbit and a controlled feed, theconditions for minimum area of contact cuttingare obtained, irrespective of work piece sizeand special programming is not needed. Theorbital cutting concept is available in the Deltacutter from Buehler. These also feature Smartcut,which will sense motor overloading during a cut,if it occurs, and reduce the feed rate automati-cally. When it is safe to return to the original feedrate, Smartcut will do so.

Precision SawsPrecision saws, Figure 4, are commonly used inmetallographic preparation to section materialsthat are small, delicate, friable, extremely hard,or where the cut must be made as close aspossible to a feature of interest, or where the cutwidth and material loss must be minimal. As thename implies, this type of saw is designed tomake very precise cuts. They are smaller in sizethan the usual laboratory abrasive cut-off sawand use much smaller blades, typically from 3-to 8-inch (7.6 to 20.3 mm) in diameter. Theseblades can be of the non-consumable type, madeof copper-based alloys and copper plated steelwith diamond or cubic boron nitride abrasivebonded to the periphery of the blade; or, theycan be consumable blades using alumina orsilicon carbide abrasives with a rubber-basedbond. Blades for the precision saws are muchthinner than the abrasive wheels used in anabrasive cutter and the load applied during cut-ting is much less. Consequently, less heat isgenerated during cutting and damage depthsare reduced. While pieces with a small section

size, that would normally be sectioned with anabrasive cutter can be cut with a precision saw,the cutting time will be appreciably greater butthe depth of damage will be much less. Preci-sion saws are widely used for sectioningsintered carbides, ceramic materials, thermallysprayed coatings, printed circuit boards, elec-tronic components, bone, teeth, etc. Table 2 listsselection criteria for precision saw blades.Buehler’s blades for precision saws are avail-able with different abrasive sizes and bonds toprovide optimum cutting for a wide variety ofapplications.

Sectioning

Figure 4. Isomet 5000 Linear Precision Saw

HELPFUL HINTS FOR SECTIONING

• When cutting a difficult specimen withthe recommended consumable abrasivewheel, if the cutting action has becomevery slow, pulse the applied force.This will help break down the abrasivebonding, exposing fresh, sharp abra-sives to enhance the cutting action.However, if you are using a resin-bonded diamond blade to cut cementedcarbides, or other very hard materials,do not pulse the applied force, as thiswill shorten wheel life.

9

Table 2. Selection Criteria for Precision Saw Blades

3″ x 0.006″ 4″ x 0.012″ 5″ x 0.015″ 6″ x 0.020″ 7″ x 0.025″ 8″ x 0.035″(75mm x (100mm x (125mm x (150mm x (180mm x (200mm x

BLADE SERIES 0.2mm) 0.3mm) 0.4mm) 0.5mm) 0.6mm) 0.9mm)

Diamond Wafering Blades

Series 30HC Diamond, for use X** X** Xwith plastics, polymers, and rubber

Series 20HC Diamond, for aggressive X* X Xgeneral sectioning of ferrous andnonferrous materials includingtitanium alloys

Series 15HC Diamond, for general X X X X X Xuse with ferrous and nonferrousalloys, copper, aluminum, metalmatrix, composites, PC boards,thermal spray coatings, andtitanium alloys

Series 20LC Diamond, for use with X* Xhard/tough materials structuralceramics, boron carbide, boronnitride, and silicon carbide

Series 15LC Diamond, for use with X X X X X Xhard/brittle materials structuralceramics, boron carbide, boronnitride, and silicon carbide

Series 10LC Diamond, for use with X X X* Xmedium to soft ceramics, electronicpackages, unmounted integratedcircuits, GaAs, AIN and glass fiberreinforced composites

Series 5LC Diamond, for use with X Xsoft friable ceramics, electronicpackages, unmounted integratedcircuits, composites with finereinforcing media, CaF2, MgF2,and carbon composites

Isocut Wafering Blades

Cubic Boron Nitride (CBN) abrasiveblades work well for many toughmaterials giving significantly shortercut times

For iron, carbon steels, high alloy X X X X X Xsteels, cobalt alloys, nickel super-alloys, and lead alloys

General Usage Abrasive Cut-off Bond/Wheels 0.03″ (0.8mm) thick Abrasive

For ferrous materials, stainless R/Al2O3 Xsteels, cast irons, andthermal spray coatings

For tough nonferrous metals, R/SiC Xaluminum, copper, titanium,uranium, zirconium

Acu-Thin Abrasive Cut-off Bond/Wheels 0.019″ (0.5mm) thick Abrasive

For sectioning small, delicatespecimens or where minimaldeformation and kerf loss is theprimary concern

Tool, hard steel, ≥ 45 HRC R/Al2O3 X

Medium hard, soft steel ≤ 45 HRC R/Al2O3 X

* Alternate blade thickness of 0.020″ (0.5 mm) ** Alternate blade thickness of 0.030″ (0.8 mm)

For a complete listing of Buehler consumable supplies for use with the Isomet Precision Saws, please refer to Buehler’s Consumables Buyers Guide.

Sectioning

10

MOUNTING OF SPECIMENS

The primary purpose of mounting metallo-graphic specimens is for convenience inhandling specimens of difficult shapes or sizesduring the subsequent steps of metallographicpreparation and examination. A secondary pur-pose is to protect and preserve extreme edgesor surface defects during metallographic prepa-ration. The method of mounting should in no waybe injurious to the microstructure of the speci-men. Pressure and heat are the most likelysources of injurious effects.

Phenolic plastics were introduced to metallog-raphy in 1928 for encapsulating specimens byhot compression mounting. Prior to that time,specimens were prepared unmounted, ormounted in flours of sulfur, wax or low-meltingpoint alloys, such as Wood’s metal. These“mounting compounds” were not without theirproblems. Introduction of polymers was a bigimprovement over these methods. Subse-quently, many polymers were evaluated for useas mounting compounds, as they were intro-duced to the market. Development of castableresins in the 1950’s added new resins to themetallographer’s tool chest. The simplicity ofmounting without using a press, and their lowcuring temperatures, made castable resins anattractive alternative.

Clamp MountingClamps have been used for many years tomount cross sections of thin sheet specimens.Several specimens can be clamped convenientlyin sandwich form making this a quick, convenientmethod for mounting thin sheet specimens. Whendone properly, edge retention is excellent, andseepage of fluids from crevices between speci-mens does not occur. The outer clamp edgesshould be beveled to minimize damage to pol-ishing cloths. If clamps are improperly used sothat gaps exist between specimens, fluids andabrasives can become entrapped and will seepout obscuring edges and can cause cross con-tamination. This problem can be minimized byproper tightening of clamps, by use of plasticspacers between specimens, or by coating speci-men surfaces with epoxy before tightening.

Compression MountingThe most common mounting method usespressure and heat to encapsulate the specimenwith a thermosetting or thermoplastic mountingmaterial. Common thermosetting resins includephenolic (Phenocure resin), diallyl phthalate andepoxy (Epomet resin) while methyl methacry-late (Transoptic resin) is the most commonlyused thermoplastic mounting resin. Table 3 liststhe characteristics of the hot compression mount-

Mounting of Specimens

Table 3. Characteristics of Buehler’s Compression (Hot) Mounting Resins

Best Edge Retention; Near ZeroVery Low Shrinkage; Best Edge electrical

Least Fine Particle Size; Retention; Very Resistance;Materials Expensive Fills Smallest Crevices Low Shrinkage SEM - EDS/WDS Clear

Ceramics Phenocure Epomet F Epomet G Probemet Transoptic

Steels Phenocure Epomet F Epomet G Probemet Transoptic

Plated Phenocure Epomet F Epomet G Probemet TransopticLayers

Aluminum Phenocure Probemet Transoptic

Copper/ Phenocure Probemet TransopticBrass

Black, Red, Black Black Bronze TransparentColor or Green

Temperature 300°F/150°C 300°F/150°C 300°F/150°C 300°F/150°C 300°F/180°C

4200 psi/ 4200 psi/ 3000-4200 psi/ 4200 psi/ 3800 psi/Pressure 290 bar 290 bar 210-290 bar 290 bar 260 bar

11

ing resins. Both thermosetting and thermoplas-tic materials require heat and pressure duringthe molding cycle; but, after curing, mounts madeof thermoplastic resins must be cooled underpressure to at least 70°C while mounts made ofthermosetting materials may be ejected from themold at the maximum molding temperature.However, cooling thermosetting resins underpressure to near ambient temperature beforeejection will significantly reduce shrinkage gapformation. Never rapidly cool a thermosettingresin mount with water after hot ejection fromthe molding temperature. This causes the metalto pull away from the resin producing shrinkagegaps that promote poor edge retention, see Fig-ure 5, because of the different rates of thermalcontraction. Epomet resin, a thermosetting ep-

oxy, provides the best edge retention, Figure 6,of these resins and is virtually unaffected by hotor boiling etchants while phenolic resins arebadly damaged.

Mounting presses vary from simple laboratoryjacks with a heater and mold assembly to fullautomated devices, as shown in Figure 7. Anadvantage of compression mounting is produc-tion of a mount of a predicable, convenient sizeand shape. Further, considerable informationcan be engraved on the backside – this is al-

ways more difficult with unmounted specimens.Manual “hand” polishing is simplified, as thespecimens are easy to hold. Also, placing a num-ber of mounted specimens in a holder for semi-or fully-automated grinding and polishing iseasier with standard mounts than for unmountedspecimens. Mounted specimens are easier onthe grinding/polishing surfaces than unmountedspecimens.

Castable Resins for MountingCold mounting materials require neither pre-ssure nor external heat and are recommendedfor mounting specimens that are sensitive to heatand/or pressure. Acrylic resins, such as Varidurand Sampl-Kwick resins, are the most widelyused castable resins due to their low cost andshort curing time. However, shrinkage can be aproblem with acrylics. Epoxy resins, althoughmore expensive than acrylics, are commonlyused because epoxy will physically adhere tospecimens, have low shrinkage, and can bedrawn into cracks and pores, particularly if avacuum impregnation chamber, Figure 8, isemployed and a low viscosity epoxy, such asEpo-Thin resin, is used. Epoxies are very suit-

Figure 5. Edge retention of this improperly carbur-ized 8620 alloy steel was degraded by a shrinkagegap between the specimen and the phenolic mount:a) top, 500X; b) bottom, 1000X (2% nital).

b

a


Figure 6. Excellent edge retention of a borided42CrMo4 alloy steel specimen mounted in Epometresin (1000X, 2% nital).

Figure 7. Simplimet 3000 Automatic MountingPress

12

able for mounting fragile or friable specimensand corrosion or oxidation specimens. Dyes orfluorescent agents, such as Fluor-Met dye, maybe added to epoxies for the study of porousspecimens such as thermal spray coatedspecimens. Most epoxies are cured at room tem-perature, and curing times can vary from2 to 8 hours. Some can be cured at slightlyelevated temperatures in less time, as long asthe higher temperature does not adverselyaffect the specimen. Table 4 lists the character-istics of Buehler’s castable resins.

Cast epoxy resins provide better edge retentionthan cast acrylic resins, mainly due to the betteradhesion of epoxy to the specimen and theirlower shrinkage. Acrylics usually do not bond tothe specimen and, because they shrink moreduring curing, a gap is inevitability formed be-tween specimen and mount. When a shrinkage

gap is present, edge retention is usuallypoor. To improve edge retention with castablemounts, the specimen can be plated, for ex-ample, with electroless nickel using theEdgemet kit, or Flat-Edge Filler particles can beadded to the resin. To obtain electrical conduc-tivity, Conductive Filler particles can be addedto the castable resin.

When preparing castable resin mounts, particu-larly epoxy mounts by manual (“hand”) methods,the metallographer will observe that the surfacetension between the mount and the workingsurface is much higher than with a compressionmount. This can make holding the mount morechallenging. If automated devices are used, themetallographer may hear “chatter” (noise) dur-ing rough grinding due to the greater surfacetension. The chatter can be reduced or stoppedby changing to contra mode.

Acrylics do generate considerable heat duringcuring and this can be strongly influenced by themolding technique used. Nelson [4] measured theexotherm produced by polymerizing an acrylicresin using two procedures: a glass mold on aglass plate (insulative) and an aluminum moldon an aluminum plate (conductive). Polymeriza-tion produced a maximum exotherm of 132°Cusing the insulative approach but only 42°Cusing the conductive approach. Note that 132°Cis not much less than the 150°C temperature

Table 4. Characteristics of Buehler’s Castable Resins

Peak Shore D Cure RecommendedName Type Temperature Hardness* Time Mold Comments

Epo-Thin Epoxy 80°F (27°C) 78 Best 18-20 Any Low viscosity, low shrinkage,Edge Retention transparent, best for

vacuum impregnation

Epoxicure Epoxy 82°F (28°C) 82 Best 6-8 Any Moderate hardness,(formerly Edge Retention hours transparent,Epoxide) low shrinkage

Epo-Kwick Epoxy 185°F (85°C) 82 Good 30-45 Sampl-Kup Faster epoxy, someEdge Retention minutes shrinkage, transparent

Epo-Color Epoxy 175°F (79°C) 82 Fair 1-2 Sampl-Kup Dye-enhanced epoxy,Edge Retention hours displays red under darkfield

and polarized light

Varidur Acrylic 170°F (77°C) 85 Good 8-10 Any Fast cure, harder acrylic(formerly Edge Retention minutes opaque, abrasive resistantUltra-Mount)

Sampl-Kwick Acrylic 175°F (79°C) 80 Poor 5-8 Any Very fast cure, translucent,Edge Retention minutes some shrinkage

* Hardness differences appear negligible but abrasion resistance has a significant effect on edge rounding


Figure 8. Vacuum Impregnation Equipment

13

used in hot compression mounting! Nelson alsomeasured the exotherm produced when anepoxy resin was cured in a phenolic ring formplaced on a pasteboard base. Although this wasan insulative approach, the maximum tempera-ture during polymerization was only 7°C, a vastimprovement over the acrylics.

Nelson’s work applies to specific acrylic andepoxy resins molded upon specific conditions.While the epoxy that he used exhibited a lowexotherm, this does not imply that all epoxyresins will exhibit such low exotherms in poly-merization. Epoxy resins that cure in short timeperiods develop much higher exotherms, that canexceed that of acrylic resins. In addition to thespeed of curing of the epoxy resin, other factorsdo influence the magnitude of the exothermduring polymerization. The larger the mass ofepoxy in the mount, the faster it will set and thegreater the exotherm. Indeed, very large mountscan generate enough heat to crack extensively.Heating the resin makes it less viscous andspeeds up curing, also generating more heatduring polymerization. The mold material alsocan influence curing time and temperature.For example, Epoxicure cures fastest inSampl-Kup plastic molds, slower in phenolic ringforms, and still slower in the reuseable rubbermounting cups. Consequently, the exotherm willbe greater when using the Sampl-Kup type moldand lowest when using the rubber mountingcups. All of these factors must be considered ifthe exotherm must be minimized.

Edge PreservationEdge preservation is a classic metallographicproblem and many “tricks” have been promoted(most pertaining to mounting, but some to grind-ing and polishing) to enhance edge flatness.These methods [2] include the use of backupmaterial in the mount, the application of coatingsto the surfaces before mounting or the additionof a filler material to the mounting resin. Plating[2] of a compatible metal on the surface to beprotected (electroless nickel, deposited usingthe Edgemet Kit has been widely used) is gen-erally considered to be the most effectiveprocedure. However, image contrast at an inter-face between a specimen and the electrolessnickel may be inadequate for certain evalua-

tions. Figure 9 shows the surface of a specimenof 1215 free-machining steel that was salt bathnitrided. One specimen was plated with electro-less nickel; both were mounted in Epomet G. It ishard to tell where the nitrided layer stops for theplated specimen, Figure 9a, which exhibits poorimage contrast between the nickel and the ni-trided surface. This is not a problem for thenon-plated specimen, Figure 9b.

Introduction of new technology has greatlyreduced edge preservation problems [5,6].Mounting presses that cool the specimen tonear ambient temperature under pressure pro-duce much tighter mounts. Gaps that formbetween specimen and resin are a major con-tributor to edge rounding, as shown in Figure 5.Staining at shrinkage gaps may also be aproblem, as demonstrated in Figure 10. Use ofsemi-automatic and automatic grinding/polish-ing equipment, rather than manual (hand)preparation, increases surface flatness andedge retention. To achieve the best results, how-ever, the position of the specimen holder, relativeto the platen, must be adjusted so that the outeredge of the specimen holder rotates out overthe edge of the surface on the platen duringgrinding and polishing, particularly for 8-inch


Figure 9. Example of (a, top) poor coating edgevisibility due to a lack of contrast between theprotective nickel plating and the salt bath nitridedsurface (arrow) of 1215 free-machining carbon steel;and, (b, bottom) good contrast and visibilitybetween Epomet resin and nitrided surface (arrow)plus excellent edge retention (1000X, 2% nital).

Ni

b

a

14

diameter (200 mm) platens. The use of harder,woven or non-woven, napless surfaces for pol-ishing with diamond abrasives (rather than softercloths such as canvas, billiard and felt) main-tains flatness. Final polishing with low nap clothsfor short times introduces very little roundingcompared to use of higher nap, softer cloths.

These procedures will produce better edgeretention with all thermosetting and thermoplas-tic mounting materials. Nevertheless, there arestill differences among the polymeric materialsused for mounting. Thermosetting resinsprovide better edge retention than thermoplas-tic resins. Of the thermosetting resins, diallylphthalate provides little improvement over themuch less expensive phenolic compounds.

The best results are obtained with Epomet G anepoxy based thermosetting resin that containsa filler material. For comparison, Figure 11 showsmicrographs of the nitrided 1215 specimenmounted in a phenolic resin (Figure 11a),and in methyl methacrylate (Figure 11b), at1000X. These specimens were prepared in thesame specimen holder as those shown in Fig-ure 9, but neither displays acceptable edgeretention at 1000X. Figure 12 shows examplesof perfect edge retention, as also demonstratedin Figure 9.

Very fine aluminum oxide spheres have beenadded to epoxy mounts to improve edge reten-tion, but this is really not a satisfactory solutionas the particles are extremely hard (~2000 HV)and their grinding/polishing characteristics areincompatible with softer metals placed insidethe mount. Recently, a soft ceramic shot(~775 HV) has been introduced that has grind-ing/polishing characteristics compatible withmetallic specimens placed in the mount. Figure13 shows an example of edge retention with theFlat-Edge Filler soft ceramic shot in an epoxymount.

b

a


Figure 10. Staining (thin arrows) due to etchantbleed out from a shrinkage gap (wide arrows)between the phenolic mount and the M2 high speedsteel specimen (500X, Vilella’s reagent).

Figure 11. Poor edge retention (arrows) at the saltbath nitrided surface of 1215 free-machining carbonsteel mounted in (a, top) phenolic resin and in (b,bottom) methyl methacrylate and polished in thesame holder as those shown in Figure 9 (1000X, 2%nital).

b

a

Figure 12. Excellent edge retention for (a, top)complex coated (arrow) sintered carbide insert(1000X, Murakami’s reagent) and for (b, bottom) ionnitrided (arrow points to a white iron nitride layer;an embedded shot blasting particle is to the left ofthe arrow) H13 hot work die steel specimen (1000X,2% nital). Epomet was used in both cases.

15

Following are general guidelines for obtainingthe best possible edge retention. All of thesefactors contribute to the overall success, al-though some are more critical than others.

• Properly mounted specimens yield betteredge retention than unmounted specimens,as rounding is difficult, if not impossible, toprevent at a free edge. Hot compressionmounts yield better edge preservation thancastable resins.

• Electrolytic or electroless Ni plating (e.g., withthe Edgemet Kit) of the surface of interestprovides excellent edge retention. If thecompression mount is cooled too quickly af-ter polymerization, the plating may be pulledaway from the specimen leaving a gap. Whenthis happens, the plating is ineffective for edgeretention.

• Thermoplastic compression mounting mate-rials are less effective than thermosettingresins. The best thermosetting resin for edgeretention is Epomet G, an epoxy-based resincontaining a hard filler material.

• Never hot eject a thermosetting resin after po-lymerization and cool it quickly to ambient (e.g.,by cooling it in water) as a gap will formbetween specimen and mount due to the dif-ferences in thermal contraction rates. Fullyautomated mounting presses cool themounted specimen to near ambient tempera-ture under pressure and this greatly minimizesgap formation due to shrinkage.

• Automated grinding/polishing equipment pro-duces flatter specimens than manual “hand”preparation.

• Use the central force mode (defined later inthe text) with an automated grinder/polisheras this method provides better flatness thanindividual pressure mode (defined later inthe text).

• Orient the position of the smaller diameterspecimen holder so that, as it rotates, its pe-riphery slightly overlaps the periphery of thelarger diameter platen.

• Use PSA-backed SiC grinding paper (if SiC isused), rather than water on the platen and aperipheral hold-down ring, and PSA-backedpolishing cloths rather than stretched cloths.

• Ultra-Prep metal-bonded or resin-bonded dia-mond grinding discs produce excellent flatsurfaces for a wide variety of materials.

• Use “hard” napless surfaces for rough polish-ing (until the final polishing step), such asTexmet 1000, Ultra-Pol or Ultra-Pad cloths,and fine polishing, such as a Trident cloth. Usea napless, or a low- to medium-nap cloth, de-pending upon the material being prepared,for the final step and keep the polishing timebrief.

• Rigid grinding discs, such as the Buehler-Hercules H and S discs, produce excellentflatness and edge retention and should beused whenever possible.


Figure 13. Flat-Edge Filler shot was added toEpoxicure resin to improve the edge retention ofthis annealed H13 hot work die steel specimen(500X, 4% picral).

HELPFUL HINTS FOR MOUNTING

• Epoxy is the only resin that will physi-cally adhere to a specimen. If itsviscosity is low, epoxy can be drawn intopores and cracks by vacuum impregna-tion. Acrylics are too viscous for vacuumimpregnation.

• Castable resins are sensitive to shelf life,which can be extended by keeping themin a refrigerator. It is a good practice todate your containers when you get them.

• A protective nickel plating can bepopped off the specimen if you eject thespecimen after the curing cycle and coolit quickly.

16

Other materials have also been used both forthe planar grinding stage or, afterwards, toreplace SiC paper. For very hard materialssuch as ceramics and sintered carbides, one, ormore, metal-bonded or resin-bonded diamonddisks (the traditional type) with grit sizes fromabout 70 to 9 µm can be used. The traditionalmetal- or resin-bonded diamond disc has dia-mond spread uniformly over its entire surface.An alternate type of disc, the Ultra-Prep disc,has diamond particles applied in small spots tothe disk surface, so that surface tension is less-ened. Ultra-Prep metal-bonded discs areavailable in six diamond sizes from 125 to 6 µmwhile Ultra-Prep resin bonded discs are avail-able in three diamond sizes from 30 to3 µm. Another approach uses a stainless steelwoven mesh Ultra-Plan cloth on a platen chargedwith coarse diamond, usually in slurry form, forplanar grinding. Once planar surfaces have beenobtained, there are several single-step proce-dures available for avoiding the finer SiC papers.These include the use of platens, thick wovenpolyester cloths, silk, or rigid grinding disks. Witheach of these, an intermediate diamond size,generally 9 to 3 µm, is used.

Grinding MediaThe grinding abrasives commonly used in thepreparation of metallographic specimens aresilicon carbide (SiC), aluminum oxide (Al2O3), em-ery (Al2O3 - Fe3O4), composite ceramics anddiamond. Emery paper is rarely used today inmetallography due to its low cutting efficiency.SiC is more readily available as waterproofpaper than aluminum oxide. Alumina papers,such as Planarmet Al 120-grit paper, do have abetter cutting rate than SiC for some metals [3].These abrasives are bonded to paper, polymericor cloth backing materials of various weights inthe form of sheets, discs and belts of varioussizes. Limited use is made of standard grindingwheels with abrasives embedded in a bondingmaterial. The abrasives may be used also inpowder form by charging the grinding surfaceswith the abrasive in a premixed slurry or sus-pension. SiC particles, particularly with the finersize papers, embed readily when grinding softmetals, such as Pb, Sn, Cd and Bi (see Figure14). Embedding of diamond abrasive is also a

GRINDING

Grinding should commence with the finest gritsize that will establish an initially flat surfaceand remove the effects of sectioning within afew minutes. An abrasive grit size of 180 or 240(P180 or P280) is coarse enough to use onspecimen surfaces sectioned by an abrasive cut-off wheel. Hacksawed, bandsawed, or otherrough surfaces usually require abrasive grit sizesin the range of 120- to 180-grit. The abrasiveused for each succeeding grinding operationshould be one or two grit sizes smaller than thatused in the preceding step. A satisfactory finegrinding sequence might involve SiC paperswith grit sizes of 240-, 320-, 400-, and 600-grit(P280, P400, P600 and P1200). This sequenceis used in the “traditional” approach.

As with abrasive-wheel sectioning, all grindingsteps should be performed wet provided thatwater has no adverse effects on any constitu-ents of the microstructure. Wet grindingminimizes specimen heating, and prevents theabrasive from becoming loaded with metalremoved from the specimen being prepared.

Each grinding step, while producing damageitself, must remove the damage from the previ-ous step. The depth of damage decreases withthe abrasive size but so does the metal removalrate. For a given abrasive size, the depth ofdamage introduced is greater for soft materialsthan for hard materials.

For automated preparation using a multiple-specimen holder, the intital step is called planargrinding. This step must remove the damage fromsectioning while establishing a common plane forall of the specimens in the holder, so that eachspecimen is affected equally in subsequent steps.Silicon carbide and alumina abrasive papers arecommonly used for the planar grinding step andare very effective. Besides these papers, thereare a number of other options available. Oneoption is to planar grind the specimens with aconventional alumina grinding stone. Thisrequires a special purpose machine, as the stonemust rotate at a high speed, ≥1500 rpm, to cuteffectively. The stone must be dressed regularlywith a diamond tool to maintain flatness, dam-age depth is high and embedding of aluminaabrasive in specimens can be a problem.

Grinding

17

and the same standards to calibrate thesedevices (sieving for the coarsest grits, sedimen-tation grading for intermediate grits (240-600 orP280-P1200), and the electrical resistancemethod for very fine grit sizes). The grit sizenumbering systems differ above 180- (P180) grit,but equivalent sizes can be determined usingTable 5. As with many standards, they are notmandatory and manufacturers can, and do,make some of their papers to different mean

problem with these soft metals and with alumi-num, but mainly with slurries when naplesscloths are used, see Figure 15.

Carbimet SiC paper is made in the USAaccording to the ANSI/CAMI standard (B74.18-1996) while SiC papers manufactured inEurope are made according to the FEPAstandard (43-GB-1984, R 1993). Both standardsuse the same methods for sizing the abrasives

Table 5. European/USA Equivalency Grit Guide

European/USA Equivalency Grit Guide

FEPA (Europe) ANSI/CAMI (USA)

Grit Number Size(µµµµµm) Grit Number Size(µµµµµm) Emery Grit

P60 269.0 60 268.0

P80 201.0 80 188.0

P100 162.0 100 148.0

P120 127.0 120 116.0

P180 78.0 180 78.0 3

P240 58.5 220 66.0 2

P280 52.2 240 51.8

P320 46.2

P360 40.5 280 42.3 1

P400 35.0 320 34.3 0

P500 30.2 360 27.3

P600 25.8 400 22.1 00

P800 21.8

P1000 18.3 500 18.2 000

P1200 15.3 600 14.5

P1500 12.6 800 12.2 0000

P2000 10.3 1000 9.2

P2500 8.4 1200 6.5

P4000* 5.0*

The chart shows the midpoints for the size ranges for ANSI/CAMI graded paper according to ANSI standard B74.18-1996 and for FEPA gradedpaper according to FEPA standard 43-GB-1984 (R1993). The ANSI/CAMI standard lists SiC particles sizes ranges up to 600 grit paper. For finergrit ANSI/CAMI papers, the particles sizes come from the CAMI booklet, Coated Abrasive (1996).

*FEPA grades finer than P2500 are not standardized and are graded at the discretion of the manufacturer. In practice, the above standard valuesare only guidelines and individual manufactures may work to a different size range and mean value.

Figure 15. 6µm diamond particles (arrows)embedded in lead (1000X).

Figure 14. SiC grit particle (arrow) embedded in a6061-T6 aluminum weldment (500X, aqueous 0.5%HF).

Grinding

18

60- to 240-grit, and are mainly used for remov-ing burrs from sectioning, for rounding edgesthat need not be preserved for examination, forflattening cut surfaces to be macro-etched, orfor removing sectioning damage. Manual grind-ing work is performed on rotating “wheels”; thatis, motor-driven platens, Figure 18, upon whichthe SiC paper is attached.

Lapping is an abrasive technique in which theabrasive particles roll freely on the surface of acarrier disc. During the lapping process, the discis charged with small amounts of a hard abra-sive such as diamond or silicon carbide. Lappingdisks can be made of many different materials;cast iron and plastic are used most commonly.Lapping produces a flatter specimen surface

than grinding, but it does not remove metal inthe same manner as in grinding. Some platens,referred to as laps, are charged with diamondslurries. Initially the diamond particles roll overthe lap surface (just as with other grinding sur-faces), but they soon become embedded andcut the surface producing microchips.

particle sizes than defined in these specifica-tions. There is a philosophical difference in thetwo systems. ANSI/CAMI papers use a widerparticle size distribution (centered on the meansize) than FEPA papers. A broader size rangeallows cutting to begin faster at lower pressuresthan with a narrower size range, so less heat isgenerated and less damage results. However,the broader size range does produce a widerrange of scratch depths; but, these should beremoved by the next step in the preparation se-quence. Generation of less damage to thestructure is considered to be more importantthan the surface finish after a particular grindingstep, as it is the residual damage in the speci-men that may prevent us from seeing the truemicrostructure at the end of the preparation se-quence.

Grinding EquipmentStationary grinding papers, often used by stu-dents, but uncommon in industrial use, aresupplied in strips or rolls, such as for use withthe Handimet 2 roll grinder. The specimen isrubbed against the paper from top to bottom.Grinding in one direction is usually better formaintaining flatness than grinding in both di-rections. This procedure can be done dry forcertain delicate materials, but water is usuallyadded to keep the specimen surface cool and tocarry away the grinding debris.

Belt grinders, such as the Surfmet I or theDuomet II belt grinders, Figure 16, are usuallypresent in most laboratories. An alternative ap-proach is to use a high speed disc grinder, suchas the Supermet grinder, Figure 17. These de-vices are used with coarse abrasive papers from

Figure 17. Supermet grinder

Grinding

Figure 16. Duomet II belt grinder

Figure 18. Beta grinder/polisher

19

POLISHING

Polishing is the final step, or steps, in producinga deformation-free surface that is flat, scratchfree, and mirror-like in appearance. Such a sur-face is necessary to observe the true micro-structure for subsequent metallographic inter-pretation, both qualitative and quantitative. Thepolishing technique used should not introduceextraneous structures such as disturbed metal(Figure 19), pitting (Figure 20), dragging out of

inclusions, “comet tailing” (Figure 21), staining(Figure 22) or relief (height differences betweendifferent constituents, or between holes andconstituents (Figure 23 and 24). Polishingusually is conducted in several stages. Tradi-tionally, rough polishing generally is conductedwith 6- or 3-µm diamond abrasives charged ontonapless or low-nap cloths. For hard materials,

such as through hardened steels, ceramics andcemented carbides, an additional rough polish-ing step may be required. The initial roughpolishing step may be followed by polishing with1-µm diamond on a napless, low nap, or me-dium nap cloth. A compatible lubricant shouldbe used sparingly to prevent overheating ordeformation of the surface. Intermediate polish-ing should be performed thoroughly so that finalpolishing may be of minimal duration. Manual,or “hand” polishing, is usually conducted usinga rotating wheel where the operator rotates thespecimen in a circular path counter to the wheelrotation direction.

Figure 23. Examples of freedom from relief (a, top)and minor relief (b, bottom) at the edges of largeprimary hypereutectic silicon particles in an as-castAl-19.85% Si specimen (200X, aqueous 0.5% HF).

Figure 21. “Comet tails” at large nitrides in anannealed H13 hot work die steel specimen (200X,DIC).

Figure 22. Staining (arrow) on the surface of a ofTi-6% Al-2% Sn-4% Zr-2% Mo prepared specimen(200X).

a

Polishing

Figure 19. (top) Preparation damage (arrows) inannealed CP titanium (500X, DIC, Kroll’s reagent).Figure 20. (bottom) Pitting (arrow) on the surface ofa cold-drawn brass (Cu-20% Zn) specimen (100X)

20

Figure 24. Differences in relief control in a brazedspecimen containing shrinkage cavities: a) (top)poor control; and, b) (bottom) good control (100X,glyceregia).

Mechanical PolishingThe term “mechanical polishing” is frequentlyused to describe the various polishing proceduresinvolving the use of fine abrasives on cloth. Thecloth may be attached to a rotating wheel or avibratory polisher bowl. Historically, cloths havebeen either stretched over the wheel and held inplace with an adjustable clamp on the platenperiphery, or held in place with a pressure sensi-tive adhesive (PSA) bonded to the back ofthe cloth. If a stretched cloth moves under theapplied pressure during polishing, cutting willbe less effective. If an automated polishing headis used, stretched cloths are more likely to rip,especially if unmounted specimens are beingprepared. In mechanical polishing, the specimensare held by hand, held mechanically in a fixture,or merely confined within the polishing area, aswith the Vibromet 2 polisher.

Electrolytic PolishingElectrolytic polishing can be used to preparespecimens with deformation free surfaces. Thetechnique offers reproducibility and speed. Inmost cases, the published instructions for elec-trolytes tell the user to grind the surface to a600-grit finish and then electropolish for about1 to 2 minutes. However, the depth of damageafter a 600- (P1200) grit finish may be several

micrometres but most electropolishing solutionsremove only about 1µm per minute. In this case,the deformation will not be completely removed.In general, electropolished surfaces tend to bewavy rather than flat and focusing may bedifficult at high magnifications. Fur ther,electropolishing tends to round edges associ-ated with external surfaces, cracks or pores. Intwo-phase alloys, one phase will polish at adifferent rate than another, leading to excessiverelief. In some cases, one phase may be attackedpreferentially and inclusions are usuallyattacked. Consequently, electrolytic polishing isnot recommended for failure analysis or imageanalysis work, except possibly as a very briefstep at the end of a mechanical polishing cycleto remove whatever minor damage may persist.Electropolishing has been most successfulwith soft single-phase metals and alloys, par-ticularly where polarized light response needsto be maximized.

Manual “Hand” PolishingAside from the use of improved polishing clothsand abrasives, hand-polishing techniques stillfollow the basic practice established manyyears ago:

1. Specimen Movement. The specimen is heldwith one or both hands, depending on theoperator’s preference, and is rotated in a direc-tion counter to the rotation of the polishing wheel.In addition, the specimen is continually movedback and forth between the center and the edgeof the wheel, thereby ensuring even distributionof the abrasive and uniform wear of the polish-ing cloth. (Some metallographers use a smallwrist rotation while moving the specimen fromthe center to the edge of one side of the wheel.)After each step, the specimen is rotated 45 to90° so that the abrasion is not unidirectional.

2. Polishing Pressure. The correct amount ofapplied pressure must be determined by experi-ence. In general, a firm hand pressure is appliedto the specimen.

3. Washing and Drying. The specimen iswashed by swabbing with a liquid detergent so-lution, rinsed in warm running water, then withethanol, and dried in a stream of warm air. Alco-hol usually can be used for washing when theabrasive carrier is not soluble in water or if the

b

a

Polishing

21

specimen cannot tolerate water. Ultrasoniccleaning may be needed if the specimens areporous or cracked.

4. Cleanness. The precautions for cleanness,as previously mentioned, must be strictly ob-served to avoid contamination problems. Thisinvolves the specimen, the metallographer’shands, and the equipment.

Automated PolishingMechanical polishing can be automated to ahigh degree using a wide variety of devices rang-ing from relatively simple systems, Figure 25, torather sophisticated, minicomputer, or micropro-

number of specimens per day with a higher de-gree of quality than that of hand polishing andat reduced consumable costs. Automatic pol-ishing devices produce the best surface flatnessand edge retention. There are two approachesfor handling specimens. Central force utilizes aspecimen holder with each specimen held inplace rigidly. The holder is pressed downwardagainst the preparation surface with the forceapplied to the entire holder. Central forceyields the best edge retention and specimenflatness. If the results after etching are inad-equate, the specimens must be placed back inthe holder and the entire preparation sequencemust be repeated. Instead of doing this, mostmetallographers will repeat the final step manu-ally and then re-etch the specimen.

The second method utilizes a specimen holderwhere the specimens are held in place loosely.Force is applied to each specimen by a piston,hence the term “individual force” for this ap-proach. This method provides convenience inexamining individual specimens during the prepa-ration cycle, without the problem of regainingplanarity for all specimens in the holder on thenext step. Also, if the etch results are deemedinadequate, the specimen can be replaced in theholder to repeat the last step, as planarity isachieved individually rather than collectively. Thedrawback to this method is that slight rocking ofthe specimen may occur, especially if the speci-men height is too great, which degrades edgeretention and flatness.

Polishing ClothsThe requirements of a good polishing clothinclude the ability to hold the abrasive media,long life, absence of any foreign material thatmay cause scratches, and absence of any pro-cessing chemical (such as dye or sizing) thatmay react with the specimen. Many cloths ofdifferent fabrics, weaves, or naps are availablefor metallographic polishing. Napless or low napcloths are recommended for rough polishing withdiamond abrasive compounds. Napless, low,medium, and occasionally high nap cloths areused for final polishing. This step should be briefto minimize relief. Table 6 lists Buehler’s lineof polishing cloths, their characteristics andapplications.

cessor controlled devices, Figure 26. Units alsovary in capacity from a single specimen to a halfdozen or more at a time. These systems can beused for all grinding and polishing steps. Thesedevices enable the operator to prepare a large

Polishing

Figure 25. Beta & Vector grinder/polisher andpower head

Figure 26. BuehlerVanguard fully automaticspecimen preparation system

22

Table 6. Cloth Selection Guide

RecommendedPremium µµµµµm Size Abrasive Cloth UsageCloths Range Types Characteristics Guide Applications

Ultra-Pad 15 - 3 Diamond Hard Woven, Used to Replace Ferrouscloth No Nap with High Multiple SiC Materials and

Material Removal Grinding Steps ThermalSpray Coatings

Ultra-Pol 15 - 1 Diamond Hard Woven, Excellent Surface Minerals, Coal,cloth No Nap Finish Used to Retain and Ceramics,

Flatness in Medium Inclusion Retentionto Hard Specimens in Steels, and

Refractory Metals

Nylon 15 - 1 Diamond Medium-Hard Used to Retain Ferrous Materials,cloth Woven, No Nap Flatness and Sintered Carbides

Hard Phases and Cast Irons

Texmet P 15 - 3 Diamond Hardest Perforated Used for Material Ceramics, Carbides,pad Non-woven Pad Removal and Petrographic, Hard

for High Material Flatness of Metals, Glass, MetalRemoval Hard Specimens Matrix Composites

Texmet 2000 15 - 3 Diamond, Hard Non-woven Used for Harder Hard Ferrouspad Al2O3 Pad used for Specimens and Materials

Durability Increased Flatness and Ceramics

Texmet 1000 9 - 0.02 Diamond, Soft Most Popular Cloth Ferrous andpad Al2O3, SiO2 Non-woven for Intermediate Nonferrous Metals,

Pad for Steps, Good for Ceramics, ElectronicSurface Finish Most Materials Packages, PCB’s,

Thermal SprayCoatings, Cast Irons,

Cermets, Minerals,Composites, Plastics

Trident 9 - 1 Diamond Softer, Used to Maximize Ferrous andcloth Durable, Flatness and Nonferrous Metals,

Woven Synthetic, Retain Phases while Microelectronics,No Nap Providing Excellent Coatings

Surface Finish

Whitefelt 6 - 0.02 Diamond, Soft and General Usage Ferrous andcloth Al2O3, SiO2 Durable Matted for Intermediate Nonferrous Materials

Wool Cloth to Fine Steps

Nanocloth 6 - 0.02 Diamond, Synthetic Cloth Softer Final Soft Ferrous andcloth Al2O3, SiO2 with Dense Polishing Nonferrous Metals

Short Nap Cloth also usedwith Diamondto MaximizeFlatness and

Surface Finish

Microcloth 5 - 0.02 Diamond, Synthetic Cloth Most Popular Ferrous andcloth Al2O3, SiO2 with Medium Nap General Nonferrous

Usage Final Metals, Ceramics,Polishing Cloth Composites, PCB’s,

Cast Irons, Cermets, Plastics, Electronics

Mastertex 1 - 0.05 Al2O3, SiO2 Soft Synthetic Softer Final Soft Nonferrous andcloth Velvet with Polishing Cloth Microelectronic

Low Nap Packages

Chemomet 1 - 0.02 Al2O3, SiO2 Soft Synthetic General Usage Titanium, Stainlesscloth Pad, Micro-nap Pad that Removes Steels, Lead/Tin

Smear Metal Solders, Electronicfrom Tough Packages, Soft

Materials during NonferrousChemomechanical Metals, Plastics

Polishing

Polishing

23

Polishing AbrasivesPolishing usually involves the use of one or moreof the following abrasives: diamond, aluminumoxide (Al2O3), and amorphous silicon dioxide(SiO2) in collidal suspension. For certain materi-als, cerium oxide, chromium oxide, magnesiumoxide or iron oxide may be used, although theseare used infrequently. With the exception of dia-mond, these abrasives are normally suspendedin distilled-water, but if the metal to be polishedis not compatible with water, other suspensionssuch as ethylene glycol, alcohol, kerosene orglycerol, may be required. The diamond abra-sive should be extended only with the carrierrecommended by the manufacturer. Besides thewater-based diamond suspension, Buehler alsooffers oil-based suspensions that are needed toprepare materials sensitive to water.

Diamond abrasives were introduced to metallog-raphers commercially in the late 1940s. However,their use in metallography goes back to at leastthe late 1920s, as Hoyt [7] mentions a visit to theCarboloy plant in West Lynn, Massachusetts,where he saw sapphire bearings being polishedwith diamond dust in an oil carrier. He used someof this material to prepare sintered carbides andpublished this work in 1930. Diamond abrasiveswere first introduced in a carrier paste but lateraerosol and slurry forms were introduced. Virgin

Figure 27. Comparison of monocrystalline (a, top)and polycrystalline (b, bottom) synthetic diamondgrain shapes (SEM, 450X).

natural diamond was used initially, and is stillavailable in Metadi diamond paste and suspen-sions. Later, synthetic diamond was introduced,first of the monocrystalline form, similar in mor-phology to natural diamond, and then inpolycrystalline form. Metadi II diamond pastesuse synthetic monocrystalline diamond whileMetadi Supreme suspensions use synthetic poly-crystalline diamonds. Figure 27 shows the shapedifferences between monocrystalline and poly-crystalline diamonds. Studies have shown thatcutting rates are higher for many materials us-ing polycrystall ine diamond compared tomonocrystalline diamond.

Colloidal silica was first used for polishing wa-fers of single crystal silicon where all of thedamage on the wafer surface must be eliminatedbefore a device can be grown on it. The silica isamorphous and the solution has a basic pH ofabout 9.5. The silica particles are actually nearlyspherical in shape, Figure 28, the polishingaction is slow, and is due to both chemical andmechanical action. Damage-free surfaces canbe produced more easily when using colloidal

silica than with other abrasives (for final polish-ing). Etchants can respond differently to surfacespolished with colloidal silica. For example, anetchant that produces a grain contrast etchwhen polished with alumina may instead revealthe grain and twin boundaries with a “flat” etchwhen polished with colloidal silica. Coloretchants frequently respond better when colloi-dal silica is used producing a more pleasingrange of colors and a crisper image. But, clean-ing of the specimen is more difficult. For manualwork, use a tuft of cotton soaked in a detergentsolution. For automated systems, stop addingabrasive about 15 seconds before the cycle is

b

a

Polishing

Figure 28. Amorphous silica particles in colloidalsilica (TEM, 300,000X).

24

to end and, for the last 10 seconds, flush thecloth surface with running water. Then,cleaning is simpler. Amorphous silica willcrystallize if allowed to evaporate. Crystallinesilica will scratch specimens, so this must beavoided. When opening a bottle, clean off anycrystallized particles than may have formedaround the opening. The safest approach is tofilter the suspension before using it. Additivesare used to minimize crystallization, as inMastermet 2 Colloidal Silica. These greatly re-tard crystallization.

For routine examinations, a fine diamondabrasive, such as 1-µm, may be adequate asthe last preparation step. Traditionally, aqueousfine alumina powders and slurries, such as theMicropolish II deagglomerated alumina powdersand suspensions, have been used for final pol-ishing with medium nap cloths. Alpha alumina(0.3-µm size) and gamma alumina (0.05-µmsize) slurries (or suspensions) are popular forfinal polishing, either in sequence or singularly.Masterprep alumina suspension util izesalumina made by the sol-gel process, and itproduces better surface finishes than aluminaabrasives made by the traditional calcinationprocess. Calcined alumina abrasives alwaysexhibit some degree of agglomeration, regard-less of the efforts to deagglomerate them, whilesol-gel alumina is free of this problem. Mastermetcolloidal silica suspensions (basic pH around9.5) are newer final polishing abrasives that pro-duce a combination of mechanical and chemicalaction which is particularly beneficial for difficult

Polishing

to prepare materials. Vibratory polishers, Figure29, are often used for final polishing, particu-larly with more difficult to prepare materials, forimage analysis studies, or for publicationquality work.

Figure 29. Vibromet 2 vibratory polisher

HELPFUL HINTS FOR GRINDING/POLISHING

• Specimens that contain cracks or poresthat are not filled with epoxy may requireultrasonic cleaning to remove abrasiveand debris from the openings to avoidcontaminating the next step.

• Excessive ultrasonic cleaning vibrationscan damage the structure of certain softmetals and alloys, particularly preciousmetals.

• To remove a tightly adhering cloth, soakthe platen under hot water for a few min-utes. This loosens the glue bond andallows you to pull it off easier.

25

EXAMPLES OF PREPARATIONPROCEDURES

The “Traditional” MethodOver the past forty years, a general procedurehas been developed that is quite successful forpreparing most metals and alloys. This methodis based on grinding with silicon carbide water-proof papers through a series of grits, then roughpolishing with one or more diamond abrasivesizes, followed by fine polishing with one or morealumina suspensions of different particle size.This procedure will be called the “traditional”method, and is described in Table 7.

This procedure is used for manual or automatedpreparation, although manual control of the forceapplied to a specimen would not be as consis-tent. Complementary motion means that thespecimen holder is rotated in the same direc-tion as the platen, and does not apply to manualpreparation. Some machines can be set so thatthe specimen holder rotates in the direction op-posite to that of the platen, called “contra.” Thisprovides a more aggressive action but was notpart of the “traditional” approach when auto-mated. The traditional method is not rigid, asother polishing cloths may be substitutedand one or more of the polishing steps might

be omitted. Times and pressures could bevaried, as well, to suit the needs of the work, orthe material being prepared. This is the “art”of metallography.

Contemporary MethodsNew concepts and new preparation materialshave been introduced that enable metallogra-phers to shorten the process while producingbetter, more consistent results. Much of this ef-fort has centered upon reducing or eliminatingthe use of silicon carbide paper in the grindingsteps. In all cases, an initial grinding step mustbe used, but there are a wide range of materialsthat can be chosen instead of SiC paper. Thereis nothing wrong with the use of SiC for the firststep, except that it has a short life. If an auto-mated device is used that holds a number ofspecimens rigidly (central force), then the firststep must remove the sectioning damage on eachspecimen and bring all of the specimens in theholder to a common plane. This first step is oftencalled “planar grinding.” SiC paper can be usedfor this step, although more than one sheet maybe needed. Alternatively, the metallographercould use Planarmet Al alumina paper,conventional metal or resin bonded diamonddiscs or the Ultra-Prep diamond discs, theUltra-Plan stainless steel mesh cloth (diamond

Table 7. The Traditional Procedure for Preparing Most Metals and Alloys

Abrasive/ Load Lb. (N)/ Base Speed TimeSurface Size Specimen (rpm)/Direction (min:sec)

Carbimet 120- (P120) grit 6 (27) 240-300 Untilabrasive discs SiC water Comp. Plane(waterproof paper) cooled

Carbimet 240- (P280) grit 6 (27) 240-300 1:00abrasive discs SiC water cooled Comp.




Canvas 6-µm Metadi 6 (27) 120-150 2:00cloth diamond paste with Comp.

Metadi fluid extender

Billiard or 1-µm diamond paste with 6 (27) 120-150 2:00Felt cloths Metadi fluid extender Comp.

Microcloth Aqueous 0.3-µm 6 (27) 120-150 2:00cloth α − alumina Comp.

Micropolish slurry

Microcloth Aqueous 0.05-µm 6 (27) 120-150 2:00cloth γ - alumina Comp.

Micropolish slurry

Comp. = Complementary (platen and specimen holder both rotate in the same direction)

Examples of Preparation Procedures

26

is applied during use), the Buehler- Hercules Hand S rigid grinding discs (diamond is appliedduring use), or lapping platens of several types(diamond is applied and becomes embeddedin the surface during use). With this huge arrayof products to choose from, how can the metal-lographer decide what to use? Each of theseproducts has advantages and disadvantages,and this is only the first step.

One or more steps using diamond abrasives onnapless surfaces usually follow planar grinding.PSA-backed silk, nylon or polyester cloths arewidely used. These give good cutting rates, main-tain flatness and minimize relief. Silk cloths, suchas the Ultra-Pol cloth, provide the best flatnessand excellent surface finishes relative to the dia-mond size used. Ultra-Pad cloth, a thicker hard,woven cloth, is more aggressive, gives nearlyas good a surface finish, similar excellent flat-ness, and longer life than an Ultra-Pol cloth.Synthetic chemotextile pads, such as the Texmet1000 and 2000 cloths, give excellent flatnessand are more aggressive than silk. They are ex-cellent for retaining second phase particles andinclusions. Diamond suspensions are mostpopular with automated polishers as they canbe added easily during polishing; although it isstill best to charge the cloth initially with dia-mond paste of the same size to get polishingstarted quickly. Final polishing could be per-formed with a very fine diamond size, such as0.1-µm diamond, depending upon the material,your needs and personal preferences. Other-

wise, final polishing is performed with Mastermetcolloidal silica or with Micropolish or Masterprepalumina slurries using napless, or low to me-dium nap cloths. For some materials, such astitanium and zirconium alloys, an attack polish-ing solution is added to the slurry to enhancedeformation and scratch removal and improvepolarized light response. Contra rotation (headmoves in the direction opposite to the platen) ispreferred as the slurry stays on the cloth better,although this will not work it the headrotates at a high rpm. Examples of generic con-temporary preparation practices follow in Tables8 to 10. Specific procedures for a wide range ofmaterials are given in the next section.

The starting SiC abrasive size is chosen basedupon the degree of surface roughness and depthof cutting damage and the hardness of the ma-terial. Never start with a coarser abrasive thannecessary to remove the cutting damage andachieve planar conditions in a reasonable time.

A similar scheme can be developed usingrigid grinding discs, such as the Buehler-Hercules H Disc. These discs are generally re-stricted to materials above a certain hardness,such as 175 HV, although some softer materialscan be prepared using them. This disc can alsobe used for the planar grinding step. An exampleof such a practice, applicable to nearly all steels(results are marginal for solution annealed aus-tenitic stainless steels) is given in Table 9.

Table 8. Generic Four-Step Contemporary Procedure for Many Metals and Alloys


Carbimet 120-, 180- or 240- 6 (27) 240-300 Untilabrasive discs (P120, P180, or P280) Comp. Plane(waterproof paper) grit SiC water cooled

Ultra-Pol 9-µm Metadi 6 (27) 120-150 5:00cloth Supreme diamond Comp.

suspension*

Texmet 1000 or 3-µm Metadi 6 (27) 120-150 4:00Trident cloth Supreme diamond Comp.

suspension*

Microcloth, ~0.05µm 6 (27) 120-150 2:00Nanocloth or Mastermet ContraChemomet colloidal silica orcloths Masterprep

alumina suspensions

* Plus Metadi fluid extender as desiredComp. = Complementary (platen and specimen holder both rotate in the same direction)Contra = platen and specimen holder rotate in opposite directions


27

The planar grinding step could also beperformed using a 45-µm metal bonded, or a30-µm resin bonded, Ultra-Prep disc or with theBuehlerHercules H rigid grinding disc and15- or 30-µm diamond, depending upon thematerial. Rigid grinding discs contain no abra-sive; they must be charged during use andsuspensions are the easiest way to do this. Polycrystalline diamond suspensions are favored overmonocrystalline synthetic diamond suspensionsfor most metals and alloys due to their highercutting rate.

The BuehlerHercules S rigid grinding disc, de-signed for soft metals and alloys, is used in asimilar manner. This disc is quite versatile andcan be used to prepare harder materials as well,

although its wear rate will be greater than the Hdisc when used to prepare very hardmaterials. A generic five step practice is givenbelow for soft metals and alloys.

The planar grinding step can be performed withthe 30-µm resin bonded diamond disc, or witha second BuehlerHercules S disc and 15- or30-µm diamond, depending upon the metal oralloy. For some very difficult metals and alloys,a 1-µm diamond step on a Trident cloth (similarto step 3, but for 3 minutes) could be added,and/or a brief vibratory polish (use the samecloths and abrasives as for step 4) may beneeded to produce perfect publication qualityimages. Four steps may suffice for routine work.

Table 9. Four-Step Contemporary Procedure for Steels Using a Rigid Grinding Disc


Table 10. Four-Step Contemporary Procedure for Nonferrous Metals Using a Rigid Grinding Disc

Abrasive/ Load Lb. (N)/ Base Speed TimeSurface Size Specimen (rpm)/ Direction (min:sec)

Carbimet 240- or 320- 5 (22) 240-300 Untilabrasive discs (P280 or P400) grit Comp. Planewaterproof paper SiC water cooled

BuehlerHercules S 6-µm Metadi 5 (22) 120-150rigid grinding disc Supreme diamond Comp. 5:00

suspension*

Trident 3-µm Metadi 5 (22) 120-150cloth Supreme diamond Comp. 4:00

suspension*

Microcloth, ~0.05-µm 5 (22) 120-150Nanocloth or Mastermet Contra 2:00Chemomet colloidal silicacloths or Masterprep

alumina suspensions

*Metadi fluid extender as desiredComp.= Complementary (platen and specimen holder both rotate in the same direction)Contra = Platen and specimen holder rotate in opposite directions


Carbimet 120-, 180- or 240- 6 (27) 240-300 Untilabrasive discs (P120, P180 or P280) Comp. Planewaterproof paper grit SiC, water cooled

BuehlerHercules H 9-µm Metadi 6 (27) 120-150 5:00rigid grinding disc Supreme diamond Comp.

suspension*

Texmet 1000 or 3-µm Metadi 6 (27) 120-150Trident cloth Supreme diamond Comp. 4:00

suspension*

Microcloth, ~0.05-µm 6 (27) 120-150 2:00Nanocloth or Mastermet ContraChemomet colloidal silicacloths or Masterprep

alumina suspensions

*Plus Metadi fluid extender as desiredComp.= Complementary (platen and specimen holder both rotate in the same direction)Contra = Platen and specimen holder rotate in opposite directions

28

PROCEDURES FOR SPECIFICMATERIALS

Following are our recommendations for prepar-ing a wide range of materials arranged, in thecase of metals, according to common charac-teristics, basically in accordance with their class-ification in the periodic table of the elements.This break down is further modified, especiallyin the case of iron based alloys, as required dueto the wide range of properties that can beexperienced.

The periodic table of the elements categorizesmetallic, nonmetallic and metalloid elementsaccording to similarities in atomic structure andis a good starting point, aided by similarities inphysical properties and behavior, for groupingelements and their alloys that have reasonablysimilar preparation procedures. The periodictable below has the various groups color codedfor convenience.

Of course, metallographers prepare materialsother than metals. These have been groupedand sub-divided according to the nature of thesematerials, as follows:

• Sintered Carbides

• Ceramics

• Composites: Metal Matrix, Polymer Matrixand Ceramic Matrix

• Printed Circuit Boards

• Microelectronic Devices

• Plastics and Polymers

The procedures discussed in this book weredeveloped and tested extensively using ma-chines with an 8-inch (200 mm) diameter platenwith six 1.25-inch (30 mm) diameter specimensin a 5-inch (125 mm) diameter holder. Theseprocedures produced the same results whenthe specimens were prepared using a 10-inch(250 mm) or a 12-inch (300 mm) diameter platen.When using the 8-inch (200 mm) platen and a5-inch (125 mm) diameter specimen holder, themachine’s head position was adjusted so thatthe specimens rotate out over the outer periph-ery of the surface on the platen. This proceduremakes maximum use of the working surface andimproves edge retention. This alignment prac-tice is less critical with the larger diameter platenswhen using the same diameter holder. However,

if the 7-inch (175 mm) diameter specimen holderis used with the 12-inch (300 mm) diameterplaten, the head position should be adjusted sothat the specimens rotate out over the platen’sedge.

In general, it is difficult to make accurate predic-tions about variations in time and load whenusing different format platens and mount sizes.Different mount sizes per se should not be usedas the basis for calculations as the area of thespecimens within the mount is more importantthan the mount area itself. For example, the samespecimen cross sectional area could be presentin 1-, 1.25- or 1.5-inch (25, 30 or 40 mm) diam-eter polymer mounts. Further, the number ofspecimens in the holder can vary substantially.It is necessary to put at least three mounts in aholder to obtain a proper weight balance. But,two specimens could be dummy mounts(mounts without specimens). In general, as thearea of the specimen in the mount increases, asthe diameter and the number of specimens in-creases, the time and pressure must increase toobtain the same degree of preparation perfec-tion.

If the area of the mounts is used, rather than thearea of the embedded specimens, and we as-sume that we are preparing six 1-inch (25 mm),six 1.25-inch (30 mm) or three 1.5-inch (40 mm)diameter specimens, then we can adjust thepressures relative to the standard values in thetables (for six 1.25-inch (30 mm) diameter speci-mens) using an 8 inch (200 mm) diameter platenformat machine. A comparison of the cross-sec-tional areas would suggest that for the six 1-inch(25 mm) specimens, the pressures could be re-duced to 65% of the standard values while forthree 1.5-inch (40 mm) diameter specimens, thepressures and times could be reduced to about70% of the standard values.

Changing the platen diameter increases the dis-tance that the specimens travel in a given time.For example, for the same rpm values, speci-mens travel 1.25 and 1.5 times as far using a10- or 12-inch (250 or 300 mm) diameter platen,respectively, compared to an 8-inch (200 mm)platen. This suggests that the times could bereduced to 80 or 67% of that used for the 8-inch(200 mm) diameter platen when using 10- or12-inch (250 or 300 mm) diameter platens. Whilepreparation times may be shorter using a largerformat machine, consumable cost is greater andlarger machines are needed only when thespecimen throughput needs are very high orwith large specimens.

Procedures for Specific Materials

29

PER

IOD

IC T

AB

LE O

F E

LEM

EN

TS

Procedures for Specific Materials

30

4Be

12Mg

13Al

LIGHT METALS: Al, Mg and Be

LIGHT METALS: Al, Mg and Be

AluminumAluminum is a soft, ductile metal. Deformationinduced damage is a common preparation prob-lem in the purer compositions. After preparation,the surface will form a tight protective oxide layerthat makes etching difficult. Commercial gradescontain many discrete intermetallic particles witha variety of compositions. These intermetallicparticles are usually attacked by etchants be-fore the matrix. Although the response to specificetchants has been used for many years to iden-tify these phases, this procedure requires carefulcontrol. Today, energy-dispersive analysis iscommonly performed for phase identificationdue to its greater reliability.

Light Metals

Five and four-step practices for aluminum al-loys are presented below. While MgO was thepreferred final polishing abrasive foraluminum and its alloys, it is a difficult abrasiveto use and is not available in very fine sizes andcolloidal silica has replaced magnesia as the pre-ferred abrasive for the final step. For coloretching work, and for the most difficult grades ofaluminum, a brief vibratory polish may beneeded to completely remove any trace of dam-age or scratches. The five-step practice isrecommended for super pure (SP) and com-mercially pure (CP) aluminum and for wroughtalloys that are difficult to prepare.

240- or 320-grit (P280 and P400) SiC water-proof paper may also be used for the planargrinding step. An Ultra-Pol cloth produces bet-ter surface finishes than an Ultra-Pad cloth, butthe Ultra-Pad cloth has a longer useful life. A

Table 11. Five-Step Procedure for Aluminum Alloys


Ultra-Prep 30-µm diamond, 5 (22) 240-300 Untilresin-bonded disc water cooled Comp. Plane

Ultra-Pol or 9-µm Metadi 5 (22) 120-150 5:00Ultra-Pad cloths Supreme diamond* Comp.

Trident cloth 3-µm Metadi 5 (22) 120-150 4:00Supreme suspension* Comp.

Trident cloth 1-µm Metadi 5 (22) 120-150 2:00paste Comp.

Microcloth, ~0.05-µm 5 (22) 120-150 1:30Nanocloth or Mastermet ContraChemomet cloth colloidal silica

*Plus Metadi fluid extender as desiredComp. = Complementary (platen and specimen holder both rotate in the same direction)Contra = platen and specimen holder rotate in opposite directions

31

Light Metals

Chemomet cloth is recommended when edgeretention is critical. Pure aluminum and somealloys are susceptible to embedment of fine dia-mond abrasive particles, especially whensuspensions are used. If this occurs, switch todiamond in paste form, which is much less likelyto cause embedding. SP and CP aluminum canbe given a brief vibratory polish (same productsas last step) to improve scratch control, althoughthis is generally not required. Masterprepalumina suspension has been found to be highlyeffective as a final polishing abrasive for alumi-num alloys, however, the standard alumina

observed in aluminum and its alloys and mini-mizes relief. Synthetic napless cloths may alsobe used for the final step with colloidal silicaand they will introduce less relief than a low ormedium nap cloth, but may not remove finepolishing scratches as well. For very pure alu-minum alloys, this procedure could be followed

Table 12. Four-Step Procedure for Aluminum Alloys

Abrasive/ Load Lb. (N)/ Base Speed TimeSurface Size Specimen (rpm)/ Direction (min:sec)

Carbimet 240- or 320- (P280 5 (22) 240-300 Untilabrasive discs or P400) grit SiC Comp. Plane(waterproof paper) water cooled

Ultra-Pol 6-µm Metadi 6 (27) 120-150 6:00cloth Supreme Comp.

diamondsuspension*

Trident cloth 3µm Metadi 6 (27) 120-150 4:00Supreme diamond Comp.

suspension*

Microcloth ~0.05µm 6 (27) 120-150 2:00Nanocloth or Mastermet ContraChemomet colloidal silicacloth or Masterprep

alumina suspensions

*Plus Metadi fluid extender as desiredComp. = Complementary (platen and specimen holder both rotate in the same direction)Contra = platen and specimen holder rotate in opposite directions

Dendritic segregation in as cast 206 aluminumrevealed by color etching with Weck’s reagent(polarized light, 200X).

abrasives made by the calcination process areunsuitable for aluminum.

For many aluminum alloys, excellent results canbe obtained using a four step procedure, suchas shown below in Table 12. This procedure re-tains all of the intermetall ic precipitates

HELPFUL HINTS FOR ALUMINUM

• Many low-alloy aluminum specimensare difficult to polish to a perfect finishusing standard methods. A vibratorypolisher can be used quite effectivelywith Mastermet colloidal silica to pro-duce deformation-free, scratch-freesurfaces for anodizing or color etchingand examination with polarized light orNomarski DIC.

Deformed, elongated grain structure of extruded6061-F aluminum after shearing revealed byanodizing with Barker’s reagent (polarized light,100X).

32

Light Metals

by vibratory polishing to improve the surfacefinish, as these are quite difficult to preparetotally free of fine polishing scratches.

MagnesiumPreparation of magnesium and its alloys is ratherdifficult due to the low matrix hardness and thehigher hardness of precipitate phases that leadto relief problems, and from the reactivity of themetal. Mechanical twinning may result during cut-ting, grinding, or handling if pressures areexcessive. Final polishing and cleaning opera-tions should avoid or minimize the use of waterand a variety of solutions have been proposed.Pure magnesium is attacked slowly by waterwhile Mg alloys may exhibit much higher attackrates. Some authors state that water should notbe used in any step and they use a 1 to 3 mix-ture of glycerol to ethanol as the coolant even inthe grinding steps. Always grind with a coolant,as fine Mg dust is a fire hazard. Because of thepresence of hard intermetallic phases, relief may

be difficult to control, especially if napped clothsare used. Following is a five-step procedure formagnesium and its alloys, see Table 13.

Masterpolish is nearly water-free and yields ex-cellent results as the final abrasive. After the laststep, wash the specimens with ethanol. Clean-ing after the last step, without using water, isdifficult. Holding the specimen under runningwater for about a second eased the cleaningproblem and did not appear to harm the micro-structure. Cosmetic cotton puffs can scratch thesurface when swab etching. For best results, etch-polish-etch cycle may be needed. Magnesiumhas a hexagonal close-packed crystal structureand will respond to polarized light. To enhancethe response, use a brief vibratory polish withthe materials used in the last step.

Mechanical twinning in deformed high-puritymagnesium (99.8% Mg) (acetic-picral etch, crossedpolarized light plus sensitive tint, 50X).

As-cast microstructure of a magnesium alloy with2.5% of rare earth elements, 2.11% Zn and 0.64% Zrshowing a film of the rare earth elements in thegrain boundaries, alloy segregation within grains(“coring” revealed by color variations within grains)and a few mechanical twins in the grains (acetic-picral etch, crossed polarized light plus sensitivetint, 100X).

Table 13. Four-Step Procedure for Magnesium Alloys


Carbimet 240- or 320- (P280 5 (22) 250 Untilabrasive discs or P400) grit SiC* Comp. Plane(waterproof paper) water cooled

Texmet 1000 9-µm Metadi oil-based 5 (22) 150 6:00pad diamond slurry Contra

Texmet 1000 3-µm Metadi oil-based 5 (22) 120-150 5:00pad diamond slurry Contra

Texmet 1000 1-µm Metadi oil-based 5 (22) 120-150 4:00pad diamond slurry Contra

Chemomet ~0.05-µm 3(13) 120-150 1:30-3:00cloth Masterpolish or Contra

Masterprep alumina

*SiC surfaces were coated with wax to minimize embedmentComp. = Complementary (platen and specimen holder both rotate in the same direction)Contra = platen and specimen holder rotate in opposite directions

33

Light Metals

BerylliumBeryllium is also a difficult metal to prepare andpresents a health risk to the metallographer. Onlythose familiar with the toxicology of Be, and areproperly equipped to deal with these issues,should work with the metal. The grinding dust isextremely toxic. Wet cutting prevents aircontamination but the grit must be disposedproperly. As with Mg, Be is easily damaged incutting and grinding forming mechanical twins.Light pressures are required. Although someauthors claim that water cannot be used, evenwhen grinding Be, others report no difficultiesusing water. Attack-polishing agents are fre-quently used when preparing Be and many arerecommended [2]. Table 14 shows a four-steppractice for beryllium.

For the final step, mix hydrogen peroxide (30%concentration – avoid physical contact!) with theMastermet colloidal silica in a ratio of one parthydrogen peroxide to five parts colloidal silica.Oxalic acid solutions (5% concentration) havealso been used with alumina for attack polish-ing. For optimal polarized light response, followthis with vibratory polishing using a one-to-tenratio of hydrogen peroxide to colloidal silica.

Grain structure of wrought, P/M beryllium (unetched,cross-polarized light, 100X).

Table 14. Four-Step Procedure for Beryllium


Carbimet 240- or 320- (P280 4 (18) 240-300 Untilabrasive discs or P400) grit SiC Comp. Plane(waterproof paper) water cooled

Ultra-Pol 6-µm Metadi 4 (18) 120-150cloth Supreme Comp. 5:00

diamondsuspension*

Trident 3-µm Metadi 4 (18) 120-150 4:00cloth Supreme Comp.

diamondsuspension*

Microcloth ~0.05-µm 3 (13) 80-120 2:00Nanocloth or Mastermet ContraChemomet colloidal silicacloth

*Plus Metadi fluid extender as desiredComp. = Complementary (platen and specimen holder both rotate in the same direction)Contra = Platen and specimen holder rotate in opposite directions

HELPFUL HINTS FOR MAGNESIUM

• For polishing, apply diamond paste anduse lapping oil, or use Buehler’s oil-based Metadi diamond suspensions,available down to a 0.10-µm size, witha medium nap cloth.

HELPFUL HINTS FOR LIGHT

METALS

• Small diamond abrasive sizes areprone to embedment when applied asa suspension or by spraying. Chargethe cloth using diamond paste toeliminate embedding. Another tech-nique is to use both a fine diamondpaste and Masterprep alumina suspen-sion – simultaneously – to preventembedding.

34

Low Melting Point Metals

30Zn

48Cd

50Sn

51Sb

82Pb

83Bi

LOW-MELTING POINT METALS:Sb, Bi, Cd, Pb, Sn and Zn

LOW-MELTING POINT METALS:Sb, Bi, Cd, Pb, Sn and Zn

As pure metals, antimony, bismuth, cadmium,lead, tin, and zinc are all very soft and difficult toprepare. Pure antimony is quite brittle, but alloyscontaining Sb are more common. Bismuth is asoft metal, but brittle, and not difficult to prepare.However, retaining bismuth particles in freemachining steels is difficult. Cadmium and zinc,both with hexagonal close-packed crystal struc-tures, are quite prone to mechanical twinformation if sectioning or grinding is performedtoo aggressively. Zinc is harder than tin or leadand tends to be brittle. Zinc is widely used to coatsheet steel (galvanized steel) for corrosionprotection, and is a common metallographicsubject. Pure zinc is very difficult to prepare. Leadis very soft and ductile and pure specimens areextremely difficult to prepare; however, leadalloys are considerably easier. Tin, which is allo-tropic with a body-centered tetragonal crystalstructure at room temperature, is soft and mal-leable and less sensitive to twinning. Pure tin,

like pure lead, is very difficult to prepare. Due totheir low melting points, and low recrystalliza-tion temperatures, cold setting resins are usuallyrecommended as recrystallization may occurduring hot compression mounting. Some of thesemetals in the pure, or nearly pure form, will de-form under the pressures used in compressionmounting. Alloys of these metals are easier toprepare, as they are usually higher in hardness.Heating of surfaces during grinding must be mini-mized. Grinding of these metals is always difficult,

Proeutectic dendrites of Cu2Sb surrounded by aeutectic mixture of antimony and Cu2Sb, in an as-cast Sb – 30% Cu specimen (unetched, polarizedlight, 200X).

(top) Cadmium dendrites and a Cd-Bi eutectic in anas-cast Cd – 20% Bi alloy (unetched, crossedpolarized light plus sensitive tint, 50X). (bottom) As-cast microstructure of Zn – 0.1% Ti – 0.2% Cu alloyexhibiting mechanical twins and a three-phaseeutectic (alpha-Zn, Cu-Zn and Zn15Ti) in a cellularpattern (Palmerton reagent, crossed polarized lightplus sensitive tint, 200X).

35

Low Melting Point Metals

as SiC particles tend to embed heavily. Manyauthors have recommended coating the SiC pa-per surface with bees wax, but this does notsolve the problem. Embedding is most commonwith the finer grit size papers. Diamond is not avery effective abrasive with these metals. Alu-mina works quite well. Following is a procedurefor these alloys, see Table 15.

For best results, follow this with a vibratorypolish using Mastermet colloidal sil icaon Microcloth or Nanocloth pads for times up to1-2 hours. This will improve polarized lightresponse for the hcp Cd and Zn and the rhom-bohedral Bi. The 1-µm alumina step will removeany embedded silicon carbide particles muchmore effectively than a diamond abrasive step.

Table 15. Six-Step Procedure for Low-Melting Point Metals


Carbimet 320- (P400) grit SiC 4 (18) 150-250 Untilabrasive discs water cooled Comp. Plane(waterproof paper) (wax coated)*

Carbimet 400- (P600) grit SiC 4 (18) 150-250 0:30abrasive water cooleddiscs (wax coated)*

Carbimet 600- (P1200) grit SiC 4 (18) 150-250 0:30abrasive water cooled Comp.discs (wax coated)*

Carbimet 800- (P1500) grit SiC 4 (18) 150-250 0:30abrasive water cooled Comp.discs (wax coated)*

Microcloth or 1-µm 5 (22) 120-150 4:00-5:00Nanocloth Micropolish II Comp.pads deagglomerated

alumina suspension

Microcloth or Masterprep 4 (18) 80-150 4:00Nanocloth 0.05-µm alumina Contrapads suspension**

*Rub candle wax lightly across revolving disc prior to grinding**See vibratory polish recommendation in textComp. = Complementary (platen and specimen holder both rotate in the same direction)Contra = Platen and specimen holder rotate in opposite directions

HELPFUL HINTS FOR

LOW-MELTING POINT METALS

• Embedment of fine abrasive particles isa common problem when preparing soft,low-melting point specimens. To reduceembedment of SiC abrasive particles,coat the paper with wax before grind-ing. Johnson (Wear, Vol. 16, 1970, p.351-358) showed that candle wax is farmore effective than paraffin. However,if the layer applied is too thick, the metal-removal rate in grinding will be reduceddrastically.

36

Refractory Metals

50Sn

REFRACTORY METALS:Ti, Zr, Hf, Cr, Mo, Nb, Re, Ta, V, and W

22Ti

23V

24Cr

40Zr

41Nb

42Mo

72Hf

73Ta

74W

75Re

REFRACTORY METALS: Ti, Zr,Hf, Cr, Mo, Nb, Re, Ta, V and W

TitaniumPure titanium is soft and ductile, but is very eas-ily damaged by twinning in sectioning andgrinding. Preparation of commercially pure tita-nium, which is a popular grade, is very difficult,while preparation of the alloys is somewhateasier. Some authors have stated that titaniumalloys should not be mounted in phenolic resinsas the alloys can absorb hydrogen from the resin.Further, it is possible that the heat from mount-ing could cause hydrides to go into solution. Thisis also possible with castable resins if the exo-thermic reaction of polymerization generatesexcessive heat. If the hydride phase content is asubject of interest, then the specimens must be

mounted in a castable resin with a very lowexotherm (long curing times favor lower heatgeneration, and vice versa). Titanium is very dif-ficult to section and has low grinding andpolishing rates. The following practice for tita-nium and its alloys demonstrates the use of an

Table 16. Three-Step Procedure for Titanium and Titanium Alloys


Carbimet 320- (P400) grit 6 (27) 240-300 Untilabrasive discs SiC, water cooled Comp. Plane(waterproof paper)

Ultra-Pol 9-µm Metadi 6 (27) 120-150 10:00cloth Supreme diamond Contra

suspension*

Microcloth or ~0.05-µm 6 (27) 120-150 10:00Nanocloth Mastermet Contrapads colloidal silica plus

attack polish agent**

*Plus Metadi fluid extender as desired **See text for agentComp. = Complementary (platen and specimen holder both rotate in the same direction)Contra = Platen and specimen holder rotate in opposite directions

Alpha at the surface of heat treated (1038ºC, waterquench) Ti – 3% Cr alloy after tint etching withBeraha’s reagent (polarized light, 500X).

37

Refractory Metals

Table 17. Four-Step Procedure for Zirconium and Hafnium


Carbimet 320- (P400) 5 (22) 200-250 Untilabrasive discs grit SiC Comp. Plane(waterproof paper) water cooled

Ultra-Pol or 9-µm Metadi Supreme 5 (22) 150-200 5:00Ultra-Pad cloths diamond suspension* Contra

Trident cloth 3-µm Metadi Supreme 5 (22) 150-200 3:00diamond suspension* Contra

Microcloth, ~0.05-µm 6 (27) 120-150 7:00Nanocloth or Mastermet ContraChemomet colloidal silica pluscloths attack polish agent**

*Plus Metadi fluid extender as desired **See text for agentComp. = Complementary (platen and specimen holder both rotate in the same directionContra = Platen and specimen holder rotate in opposite directions

Basket-weave alpha-beta structure of as-cast Ti –6% Al – 4% V revealed by heat tinting (polarizedlight, 100X).

attack-polishing agent added to the final polish-ing abrasive to obtain the best results, especiallyfor commercially pure titanium, a rather difficultmetal to prepare free of deformation for coloretching, heat tinting and/or polarized light ex-amination of the grain structure. Attack polishingsolutions added to the abrasive slurry or sus-pension must be treated with great care to avoidburns. Use good, safe laboratory practices andit is advisable to wear protective gloves. Thisthree step practice could be modified to four stepsby adding a 3- or 1-µm diamond step, but this isusually unnecessary, see Table 16.

A number of attack polishing agents have beenused. The simplest is a mixture of 10 mL hydro-gen peroxide (30% concentration – avoid skincontact) and 50 mL colloidal silica. Some metal-lographers add either a small amount of Kroll’sreagent to this mixture, or a few mL of nitric andhydrofluoric acids. These latter additions maycause the suspension to gel. In general, theseacid additions do little to improve the action of

the hydrogen peroxide (the safer 3% concen-tration is not effective). Polarized light responseof CP titanium can be improved by followingthis procedure with a brief vibratory polish usingcolloidal silica.

Zirconium and HafniumPure zirconium and pure hafnium are soft, duc-tile hexagonal close-packed metals that candeform by mechanical twinning if handled ag-gressively in sectioning and grinding. As withmost refractory metals, grinding and polishingremoval rates are low and eliminating all pol-ishing scratches and deformation can be difficult.It may even be possible to form mechanical twinsin compression mounting. Both can contain veryhard phases that make relief control more diffi-cult. To improve polarized light response, it iscommon practice to chemically polish specimensafter mechanical polishing. Alternatively, attackpolishing additions can be made to the final pol-ishing abrasive slurry, or vibratory polishingmay be employed. Table 17 is a four-step proce-dure that can be followed by either chemicalpolishing or vibratory polishing.

Several attack polishing agents have been usedfor Zr and Hf. One is a mixture of 1-2 partshydrogen peroxide (30% concentration – avoidall skin contact) to 8 or 9 parts colloidal silica.Another is 5 mL of a chromium trioxide solution(20 g CrO3 to 100 mL water) added to 95 mLcolloidal silica or Masterprep alumina slurry.

38

Refractory Metals

Additions of oxalic, hydrofluoric or nitric acidshave also been used. All of these attackpolishing additions must be handled with careas they are strong oxidizers. Skin contact mustbe avoided. Chemical polishing solutions arereviewed in [2]. Cain’s has been popular.Use under a hood and avoid skin contact. AnnKelly developed an excellent chemical polishfor refractory metals, such as Zr, Hf, and Ta. Itcontists of 25 mL lactic acid, 15 mL nitric-acidand 5 mL hydrofluoric acid. Swab vigorously forup to 2 minutes. To prepare ultra-pure Zr and Hf,the above method is unsatisfactory and a pro-cedure, such as in Table 15 (add a 5- or 3-µmalumina step) is needed, as diamond is ineffec-tive. Use the chromium trioxide attack polishwith the alumina and conclude with Kelly’schemical polish.

HELPFUL HINTS FOR REFRACTORY

METALS

• Due to their very low rate of grindingand polishing, the more aggressivecontra rotation gives suitable surfacesin less time.

• Refractory metal preparation is aided byusing attack-polishing additives. Polar-ized light response may be improved byfollowing preparation by swabbing witha chemical polishing solution.

Equiaxed grain structure of wrought zirconium(unetched, crossed polarized light, 100X.)

Equiaxed grain structure of wrought hafnium(unetched, crossed polarized light plus sensitivetint, 100X).

39

Other Refractory Metals

OTHER REFRACTORYMETALS: Cr, Mo, Nb, Re, Ta, Vand W

These refractory metals have body-centeredcubic crystal structures (except for rheniumwhich is hexagonal close packed) and are softand ductile when pure, but some are brittle incommercial form. They can be cold worked eas-ily, although they do not work harden appreciably,so it may be difficult to get completely deforma-tion-free microstructures.

Pure chromium is soft and brittle; but, when en-countered commercially, for example, as aplated layer, it is hard and brittle. Chromium al-loys are relatively easy to prepare, althoughdifficult to etch. Molybdenum may be tough orbrittle depending upon composition. It is sus-ceptible to deformation damage in sectioningand grinding. Pure niobium (columbium) is soft

Microstructure of powder-made W – 5.6% Ni – 2.4%Fe revealing tungsten grains surrounded by a nickel-iron matrix (unetched, Nomarski DIC, 200X).

Table 18. Four-Step Procedure for Refractory Metals


Carbimet 320- (P400) 6 (27) 150-250 Untilabrasive discs grit SiC Comp. Plane(waterproof paper) water cooled

Ultra-Pol or 9-µm Metadi 6 (27) 150-200 10:00Ultra-Pad Supreme diamond Contracloths suspension*

Texmet 1000 3-µm Metadi 6 (27) 150-200 8:00Trident cloth Supreme diamond Contra

suspension*

Microcloth, ~0.05-µm 6 (27) 120-150 5:00Nanocloth or Mastermet ContraChemomet colloidal silica pluscloths attack polish agent**

*Plus Metadi fluid extender as desired **See text for agent


Contra = Platen and specimen holder rotate in opposite directions

and ductile and difficult to prepare while itsalloys are harder and simpler to prepare. Grind-ing and polishing rates have been reported tovary with crystallographic orientation. Rheniumis very sensitive to cold work and will formmechanical twins. Tantalum is softer than nio-bium and more difficult to prepare as it easilyforms damaged layers in sectioning and grind-ing. Tantalum may contain hard phases thatpromote relief control problems. Vanadium is asoft, ductile metal but may be embrittled byhydrogen; otherwise it can be prepared muchlike a stainless steel. Tungsten is not too difficultto prepare, although grinding and polishingrates are low. Hard carbides and oxides may bepresent in these metals that introduce reliefcontrol problems.

Mechanical polishing often incorporates anattack-polishing agent in the final step or is fol-lowed by vibratory polishing or chemicalpolishing. Manual preparation of these metalsand their alloys tends to be very tedious due totheir low grinding and polishing rates. Automatedapproaches are highly recommended, especiallyif attack polishing is performed. Following is ageneric four step practice suitable for thesemetals and their alloys, see Table 18.

Many attack-polishing additives [2] have beensuggested for these metals and their alloys.A good general purpose attack polish consistsof a mixture of 5 mL chromium trioxide

40

Other Refractory Metals

solution (20 g CrO3 in 100 mL water) to 95 mLMastermet colloidal silica. Avoid skin contact asthis is a strong oxidizing solution. A number ofchemical polishing solutions have beensuggested [2]. For Nb, V and Ta, use asolution consisting of 30 mL water, 30 mL nitricacid, 30 mL hydrochloric acid and 15 mL hy-drofluoric acid. Swab or immerse at room temp-erature. An alternative chemical polish for Nb,V and Ta consists of 120 mL water, 6 g ferricchloride, 30 mL hydrochloric acid and 16 mLhydrofluoric acid. Eary and Johnson recommendimmersing the specimen in this solution 1minute for V, 2 minutes for Nb and 3 minutes forTa. Vibratory polishing is also very helpful forthese metals and their alloys.

Deformed alpha grains in W – 25% Re containingsigma phase (black spots) revealed using (top)bright field and (bottom) Nomarski DIC whichreveals the cold work much better (Murakami’sreagent, 500X).

Alpha grains in Mo – 47.5% Re. The small spots aresigma phase (Murakami’s reagent, 200X).

41

Ferrous Metals

FERROUS METALS: Fe

26Fe

FERROUS METALS

Iron-based alloys account for a large portion ofall metals production. The range of compositionsand microstructures of iron-based alloys is farwider than any other system. Pure iron is softand ductile. Development of scratch-free anddeformation-free grain structures is difficult.Sheet steels present the same problem, whichcan be complicated by protective coatings ofzinc, aluminum or Zn-Al mixtures. In general,harder alloys are much easier to prepare. Castirons may contain graphite, which must be re-tained in preparation. Inclusions are frequentlyevaluated and quantified. Volume fractions canvary from nearly 2% in a free machining gradeto levels barely detectable in a premium, doublevacuum melt alloy. A wide range of inclusion,

carbide and nitride phases has been identifiedin steels. Addition of 12 or more percent chro-mium dramatically reduces the corrosion rate ofsteels, producing a wide range of stainless steelalloys. Tool steels cover a wide range of compo-

Table 19. Four-Step Procedure for Hard Steels using the BuehlerHercules H disc


Ultra-Prep 45-µm 6 (27) 240-300 Untilmetal-bonded diamond, Comp. Planedisc water cooled

BuehlerHercules H 9-µm Metadi Supreme 6 (27) 120-150 5:00rigid grinding disc diamond suspension* Comp.

Trident cloth 3-µm Metadi Supreme 6 (27) 120-150 3:00diamond suspension* Comp.

Microcloth, Masterprep 6 (27) 120-150 2:00Nanocloth or 0.05-µm alumina ContraChemomet suspension,cloths or Mastermet

colloidal silica

*Plus Metadi fluid extender desiredComp. = Complementary (platen and specimen holder both rotate in the same direction)Contra = Platen and specimen holder rotate in opposite directions

Lath martensitic microstructure of anunconsolidated powder metallurgy steel gearcontaining substantial porosity (Klemm’s I reagent,polarized light, 200X).

42

Ferrous Metals

sitions and can attain very high hardnesses.Preparation of ferrous metals and alloys is quitestraightforward using the contemporary meth-ods. Edge retention (see guidelines on pages13-15) and inclusion retention are excellent, es-pecially if automated equipment is used. Thefollowing procedures are recommended and areadequate for the vast majority of ferrous metalsand alloys. Table 9 presented a variation of themethod given in Table 19.

This practice is also useful for cast iron speci-mens including graphitic cast irons. 240- (P280)grit SiC paper can be substituted for theUltra-Prep disc or another BuehlerHercules Hdisc can be used for planar grinding with 30-µmMetadi Supreme diamond suspension. Due totheir high silicon content and the potential for

staining problems, it is best to use theMasterprep alumina suspension for the finalpolishing step.

For softer steels, use the 30-µm resin bondedUltra-Prep disc, or 240- (P280) grit SiCpaper, for step 1. The BuehlerHercules S disccan be used in place of the BuehlerHercules Hdisc but this is not usually necessary. However,with either disc, use 6-µm Metadi Supreme forthe second step instead of 9-µm, as shown inTable 20.

This practice is well suited for solution annealedaustenitic stainless steels and for soft sheetsteels. Ultra-Pol or Ultra-Pad cloths could besubstitute for the rigid grinding discs, if desired.For perfect publication quality images, or for

Table 20. Four-Step Procedure for Soft Steels using the BuehlerHercules S disc


Ultra-Prep 30-µm 6 (27) 240-300 Untilresin-bonded diamond, Comp. Planedisc water cooled

BuehlerHercules S 6-µm Metadi 6 (27) 120-150 5:00rigid grinding disc Supreme diamond Comp.

suspension*

Trident cloth 3-µm Metadi 6 (27) 120-150 3:00Supreme diamond Comp.

suspension*


colloidal silica


Table 21. Three-Step Practice for Steels


CARBIMET® 120, 180, 240 or 320 (P120, 6 (27) 240-300 UntilAbrasive Discs P180, P280, or P400) Comp. Plane(Waterproof Paper) grit SiC, water cooled

BUEHLERHERCULES™ H or S 3µm METADI® 6 (27) 120-150 5:00Rigid Grinding Disc SUPREME Comp.

Diamond Suspension*

MICROCLOTH® MASTERPREP™ 0.05µm 6 (27) 120-150 5:00NANOCLOTH™ or Alumina Suspension, ContraCHEMOMET® or MASTERMET®

Cloths Colloidal Silica

*Plus METADI® Fluid extender as desiredComp. = Complementary (platen and specimen holder rotate in the same direction)Contra = Platen and specimen holder rotate in opposite directions

43

Ferrous Metals

color etching, follow this practice with a briefvibratory polish using the cloths and abrasivesin the last step.

Many steels, particularly the harder steels, canbe prepared in three steps with excellentresults. A recommended practice is given inTable 21.

For soft alloys, use 240- or 320- (P280 or P400)grit SiC paper; for harder alloys use 120 (P120),180- (P180), or 240- (P280) grit SiC paper,depending upon the starting surface finish andthe hardness of the alloy. Planar grinding canalso be performed using 45-µm metal-bondedor 30-µm resin-bonded Ultra-Prep diamonddiscs. For softer steels, use the BuehlerHerculesS disc for best results. The Ultra-Pol cloth canalso be used for the second step for steels ofany hardness.

The following practice is recommended for stain-less steels and maraging steels. For solutionannealed austenitic grades and for ferriticstainless grades and for annealed maraginggrades, use the BuehlerHercules S disc or theUltra-Pol cloth for best results. Start with 120-(P120) grit SiC paper only if it is a very hardmartensitic stainless steel, such as type 440C.For the martensitic grades, planar grinding can

be performed using a 45-µm metal-bonded dia-mond Ultra-Prep disc. For softer stainless steels,use the 30-µm resin-bonded diamond Ultra-Prep disc for planar grinding. Another alternativeis a second BuehlerHercules disc, either H or S,depending upon the hardness of the grade, anda 30-µm Metadi Supreme polycrystalline dia-mond suspension. The solution annealedaustenitic stainless steels and the fully ferriticstainless steels are the most difficult to prepare.It may be helpful to add a 1-µm diamond step ona Trident cloth before the last step, or tofollow the last step with a brief vibratory polishusing colloidal silica on Microcloth, Nanoclothor a Chemomet I cloth, see Table 22.

Table 22. Four-Step Contemporary Procedure for Stainless Steels and Maraging Steels


Carbimet 120- or 240- (P120 6 (27) 240-300 Untilabrasive discs or P280) grit SiC, Comp. Plane(waterproof paper) water cooled

BuehlerHercules H or S 9-µm 6 (27) 120-150 5:00rigid grinding disc Metadi Comp.

SupremeDiamond

Suspension*

Trident cloth 3-µm 6 (27) 120-150 5:00Metadi Comp.

Supremediamond

suspension*


colloidal silica


Microstructure of a duplex (ferrite tan, austenitewhite) stainless steel held at 816º C for 48 hourswhich forms sigma phase (orange particles)revealed by electrolytic etching with aqueous 20%NaOH at 3 V dc, 9 s (500X).

44

Copper, Nickel and Cobalt

27Co

28Ni

29Cu


COPPER, NICKEL and COBALT

CopperPure copper is extremely ductile and malleable.Copper and its alloys come in a wide rangeof compositions, including several variants ofnearly pure copper for electrical applications tohighly alloyed brasses and bronzes and toprecipitation hardened high strength alloys.Copper and its alloys can be easily damaged byrough sectioning and grinding practices andthe depth of damage can be substantial. Scratchremoval, particularly for pure copper andbrass alloys, can be very difficult. Following thepreparation cycle with a brief vibratory polishusing colloidal silica is very helpful for scratchremoval. Attack-polish additions have been usedin the past to improve scratch removal butusually are not necessary using the contempo-

rary method followed by vibratory polishing, seeTable 23).

Planar grinding can be performed using the45- or 15-µm metal-bonded or the 30-µm resin-bonded Ultra-Prep discs. Use the resin-bondeddisc for the soft copper grades and copperalloys.

Table 23. Five-Step Procedure for Cu and Cu Alloys

Alpha grains containing annealing twins inphosphorous-deoxidized arsenical bronze that wasannealed and lightly cold drawn (Klemm’s I reagent,polarized light, 50X).


Carbimet abrasive discs 240- or 320- (P280 or 5-6 240-300 Until(waterproof paper) P400) grit SiC water cooled (22-27) Comp. Plane

Ultra-Pol cloth or 6-µm Metadi 5-6 120-150 5:00BuehlerHercules S Supreme diamond (22-27) Comp.rigid grinding disc suspension*

Trident or 3-µm Metadi Supreme 5-6 120-150 3:00Texmet 1000 pads diamond suspension* (22-27) Comp.

Trident 1-µm Metadi Supreme 5-6 120-150 2:00cloth Diamond Suspension* (22-27) Comp.

Microcloth, ~0.05-µm Mastermet 5-6 120-150 1:30-2:00Nanocloth or colloidal silica (22-27) ContraChemomet or Masterpreppads alumina suspensions

*Plus METADI® Fluid extender as desiredComp.= Complementary (platen and specimen holder both rotate in the same direction)Contra = Platen and specimen holder rotate in opposite directions

45


NickelNickel and its alloys have face-centered cubiccrystal structures and are prepared in basicallythe same way as austenitic stainless steels. Purenickel is more difficult to prepare than the alloys.The Ni-Fe magnetic alloys are rather difficult toprepare scratch free unless vibratory polishingis used. The Monel (Ni-Cu) and the highlycorrosion resistant (Ni-Cr-Fe) alloys are moredifficult to prepare than the nickel-basedsuperalloys. Solution annealed superalloys arealways more difficult to prepare than agehardened superalloys. Age hardened superal-loys can be prepared using the BuehlerHerculesH disc; for all other nickel alloys, use theBuehlerHercules S disc for best results. The fol-lowing practice works well for nickel basedsuperalloys (and Fe-Ni based super alloys) andthe highly corrosion resistant Ni-Cr-Fe alloys,see Table 24.

If color etching is to be performed, follow the laststep with a brief vibratory polish using the samematerials as in the last step. This step is alsohelpful for the most difficult to prepare solutionannealed alloys. Alternatively, for the most diffi-cult specimens, or when color etching is beingperformed, a 1-µm diamond step on a Tridentcloth can be added before the final step.

For pure nickel, nickel-copper and nickel-ironalloys, a five step practice is preferred, as givenbelow. The planar gr inding step can beperformed using either the 30-µm resin-bondedUltra-Prep diamond disc or with 240- (P280) or320- (P400) grit SiC papers with equal success,see Table 25.

Attack-polishing agents are not often used withthese alloys to eliminate fine polishing scratchesor residual damage. If this is a problem, and some

Table 25. Five-Step Procedure for Ni, Ni-Cu and Ni-Fe Alloys


Ultra-Prep 30-µm diamond 5 (22) 200-300 Untilresin-bonded disc water cooled Comp. Plane

Ultra-Pol cloth or 9-µm Metadi 6 (27) 100-150 5:00BuehlerHercules S Supreme diamond Comp.rigid grinding disc suspension*

Trident or 3-µm Metadi Supreme 6 (27) 100-150 3:00Texmet 1000 pads diamond suspension* Comp.

Trident 1-µm Metadi Supreme 6 (27) 100-150 2:00cloth diamond suspension* Comp.

Microcloth, ∼ 0.05-µm Mastermet 6 (27) 80-150 1:30-2:00Nanocloth or colloidal silica or ContraChemomet pads Masterprep

alumina suspensions


Table 24. Four-Step Procedure for Nickel-Based Superalloys and Ni-Cr-Fe Alloys


Carbimet abrasive discs 240-(P280) grit SiC 6 (27) 240-300 Until(waterproof paper) water cooled Comp. Plane

BuehlerHercules H 9-µm Metadi 6 (27) 120-150 5:00or BuehlerHercules S Supreme diamond Comp.rigid grinding disc suspension*

Trident cloth 3-µm Metadi Supreme 6 (27) 120-150 5:00diamond suspension* Comp.

Microcloth, Masterprep 0.05-µm 6 (27) 120-150 2:00-5:00Nanocloth alumina suspension or ContraChemomet pads Mastermet colloidal silica


46


of these grades are very difficult to get perfectlyfree of scratches and deformation damage, abrief vibratory polish, using the same materialsas in the last step, will provide the needed im-provement. Masterprep alumina may give betterresults than colloidal silica for the more purenickel compositions.

CobaltCobalt and its alloys are more difficult toprepare than nickel and its alloys. Cobalt is atough metal with a hexagonal close-packedcrystal structure and is sensitive to deformationdamage by mechanical twinning. Grinding andpolishing rates are lower for Co than for Ni, Cuor Fe. Preparation of cobalt and its alloys issomewhat similar to that of refractory metals.Despite its hcp crystal structure, crossedpolarized light is not very useful for examiningcobalt alloys compared to other hcp metals and

alloys. Following is a practice for preparing Coand its alloys, see Table 26.

Two steps of SiC paper may be needed to getthe specimens co-planar. If the cut surface is ofgood quality, start with 320- (P400) grit paper.Cobalt and its alloys are more difficult to cut thanmost steels, regardless of their hardness. Attackpolishing has not been reported but chemicalpolishing has been used after mechanical pol-ishing. Morral (2) has recommended twochemical polishing solutions: equal parts of ace-tic and nitric acids (immerse) or 40 mL lacticacid, 30 mL hydrochloric acid and 5 mL nitricacid (immerse). A wide variety of Co-based al-loys have been prepared with the above methodwithout need for chemical polishing. The 1-µmdiamond step could be eliminated for routinework.

Alpha grains containing annealing twins of solutionannealed and double aged Waspaloy nickel-basedsuperalloy (Beraha’s reagent, 100X).

Table 26. Five-Step Procedure for Cobalt


Carbimet 240- (P280) or 6 (27) 250-300 Untilabrasive discs 320- (P400) grit Contra Plane(Waterproof Paper) SiC water cooled

Ultra-Pol cloth or 9-µm Metadi 6 (27) 100-150 5:00BuehlerHercules S Supreme diamond Contrarigid grinding disc suspension*

Texmet 1000 3-µm Metadi Supreme 6 (27) 100-150 5:00pad diamond suspension* Contra

Trident or 1-µm Metadi Supreme 6 (27) 100-150 3:00Texmet 1000 pads diamond suspension* Contra

Microcloth, ~0.05-µm Mastermet 6 (27) 80-120 2:00-3:00Nanocloth or colloidal silica or ContraChemomet Masterpreppads alumina suspensions


Equiaxed grain structure of Elgiloy (Co – 20% Cr –15% Fe – 15% Ni – 7% Mo – 2% Mn – 0.15% C –0.05% Be) after hot rolling and annealing revealingannealing twins (Beraha’s reagent, crossedpolarized light plus sensitive tint, 100X).

47

Precious Metals

PRECIOUS METALS: Au, Ag, Ir,Os, Pd, Pt, Rh, and Ru

44Ru

45Rh

46Pd

47Ag

77Ir

78Pt

79Au

76Os

PRECIOUS METALS: Au, Ag, Ir,Os, Pd, Pt, Rh and Ru

Relatively few metallographers work with pre-cious metals, other than those used in electronicdevices. Preparing precious metals within anintegrated circuit is discussed later (see Micro-electronic Devices). The precious metals are verysoft and ductile, deform and smear easily, andare quite challenging to prepare. Pure gold is verysoft and the most malleable metal known. Alloys,which are more commonly encountered, areharder and somewhat easier to prepare. Gold isdifficult to etch. Silver is very soft and ductile andprone to surface damage from deformation. Em-bedding of abrasives is a common problem withboth gold and silver and their alloys. Iridium ismuch harder and more easily prepared. Osmiumis rarely encountered in its pure form, even its

alloys are infrequent subjects for metallogra-phers. Damaged surface layers are easilyproduced and grinding and polishing rates arelow. It is quite difficult to prepare. Palladium ismalleable and not as difficult to prepare as mostof the precious metals. Platinum is soft andmalleable. Its alloys are more commonly encoun-

Eutectic microstructure of Ag – 28% Cu whereKlemm’s reagent has colored the copper particlesand the silver phase is uncolored (500X).

Table 27. Stewart’s Manual Procedure for Precious Metals


Carbimet abrasive 240- (P280) grit SiC moderate 400 Untildiscs (waterproof paper) water cooled Plane

Carbimet 320- (P400) grit SiC moderate 400 1:00abrasive discs water cooled



Metcloth 6-µm Metadi Supreme moderate 400 2:00pad diamond suspension

Mastertex 1-µm Metadi Supreme moderate 400 2:00cloth diamond suspension


48

Precious Metals

tered. Abrasive embedment is a problem with Ptand its alloys. Rhodium is a hard metal and isrelatively easy to prepare. Rh is sensitive to sur-face damage in sectioning and grinding.Ruthenium is a hard, brittle metal that is not toodifficult to prepare.

Stewart (Tech-Notes, Vol. 2, Issue 5) hasdescribed a method for preparing jewelry al-loys. Invariably, these are small pieces, due totheir cost, and must be mounted. Stewart usesEpomet G resin for most specimens. If transpar-ency is need, he uses Transoptic resin. Fragilebeads and balls are mounted in castable resins.Grinding and polishing was conducted at400 rpm. His practice is shown in Table 27.

This can be followed by a brief vibratory polishusing colloidal silica on Mastertex, Chemomet,Microcloth or Nanocloth pads to further enhancethe quality of the preparation; but, a 1-µm dia-mond finish is adequate for most work.Attack-polishing has been used to prepare goldand its alloys and chemical polishing has beenperformed after mechanical polishing of silver,but neither practice is commonly performed. Al-ternate etch-polish cycles may be needed toremove fine polishing scratches, especially forannealed specimens.

A procedure for an automated system wasdeveloped after experimentation with a numberof precious metal specimens. Most of thesemetals and alloys are quite soft, unless they havebeen cold worked, and they are susceptible toembedding of abrasives. In this method, only onesilicon carbide step is used. Texmet 1000 pads

are used for the diamond steps, as it will holdthe abrasive in its surface well, which minimizesembedding. Only diamond paste is used, as slur-ries will be more prone to embedding, as weobserved in high gold alloys. Use only a smallamount of distilled water as the lubricant. Do notget the cloth wet. Final polishing is with aChemomet I cloth and Masterprep alumina.Due to their excellent corrosion resistance, col-loidal silica is not effective as an abrasive forprecious metals. The Chemomet pad has lots offine pores to hold the abrasive. The cycle is givenin Table 28.

For 18 karat gold and higher (≤75% Au), it isnecessary to use an attack polish agent in thefinal step. An aqueous solution of 5 g CrO3 in100 mL water works well. Mix 10 mL of the at-tack polish agent with 50 mL of Masterprepsuspension. This will thicken, so add about 20-30 mL water to make it thinner. A 3 to 6 minuteattack- polish step will remove the fine polish-ing scratches. But, wear protective gloves as thechromium trioxide solution is a strong oxidizer.

Table 28. Five-Step Automated Procedure for Precious Metals


Carbimet 240- or 320- (P280 or 3 (13) 150-240 Untilabrasive discs P400) grit water cooled Comp. Plane

Texmet 1000 9-µm Metadi II 3 (13) 150-240 5:00pad diamond paste* Comp.



Chemomet Masterprep 2 (9) 100-150 2:00pad alumina suspension Comp.

* Use water as the lubricant, but keep the pad relatively dryComp. = Complementary (platen and specimen holder both rotate in the same direction)

Equiaxed grain structure of cold rolled and annealed18-karat gold (Neyoro 28A: 75% Au – 22% Ag – 3%Ni) revealing annealing twins (equal parts 10% NaCNand 30% H2O2, 50X).

49

Thermally-Spray Coated Specimens

THERMALLY-SPRAY COATEDSPECIMENS

Thermally sprayed coatings (TSC) and thermalbarrier coatings (TBC) are widely used on manymetal substrates. Invariably, these coatings arenot 100% dense but contain several types ofvoids, such as porosity and linear detachments.Hot compression mounting is not recommendedas the molding pressure can collapse the voids.Use a low-viscosity castable epoxy and usevacuum infiltration to fill the connected voids withepoxy. Fluorescent dyes, such as Flour-Met dye,

Table 30. Four-Step Procedure for TSC and TBC Specimens with Metallic Coatings


BuehlerHercules H 30-µm Metadi 5 (22) 240-300 Untilrigid grinding disc Supreme diamond Contra Plane

suspension*

Ultra-Pol 9-µm Metadi 5 (22) 150-200 4:00cloth Supreme diamond Contra

suspension*

Trident 3-µm Metadi 6 (27) 150 3:00cloth Supreme diamond Comp.

suspension*

Microcloth, Masterprep 4 (18) 120-150 2:00Nanocloth or 0.05-µm alumina ContraChemomet suspension orcloths Mastermet

colloidal silica


Table 29. Four-Step Procedure for TSC and TBC Specimens with Metallic Coatings


BuehlerHercules H 30-µm Metadi 5 (22) 240-300 Untilrigid grinding disc Supreme diamond Contra Plane

suspension*

BuehlerHercules H 6-µm Metadi 5 (22) 150-200rigid grinding disc Supreme diamond Contra 4:00

suspension*

Trident 3-µm Metadi 6 (27) 150 3:00cloth Supreme diamond Comp.

suspension*


colloidal silica


may be added to the epoxy. When viewedwith fluorescent illumination, the epoxy-filledvoids appear bright yellow green. This makes iteasy to discriminate between dark holes anddark oxides, as would be seen with bright fieldillumination. Fill ing the pores with epoxyalso makes it easier to keep the pore wallsflat to the edge during preparation. Aside fromthis mounting requirement, TSC and TBAspecimens are prepared using all of the factorsneeded for good edge retention (see pages 13-15). A variety of procedures can be used. TheBuehlerHercules H rigid grinding disc produces

50

exceptional edge flatness for these specimens.Two four step procedures for TSC and TBA speci-mens using the BuehlerHercules H rigid grindingdisc for specimens with metallic coatings andone procedure for specimens with ceramic coat-ings are given, see Tables 29-31.

The 30-µm resin bonded, or the 45-µm metalbonded Ultra-Prep diamond discs, can be sub-stituted for the planar grinding step. (Table 29and 30).

Ultra-Pol or Ultra-Pad cloths can be used in thesecond step. (Table 29 and 31).

Table 31. Four-Step Procedure for TSC and TBA Specimens with Ceramic Coatings


Ultra-Prep 45-µm disc 5 (22) 240-300 Untilmetal-bonded disc water cooled Contra Plane

BuehlerHercules H 9-µm 5 (22) 150-180 4:00rigid grinding disc Metadi Contra

Supremediamond

suspension*

Texmet 1000 or 2000 3-µm 6 (27) 120-150 3:00or Trident cloth Metadi Comp.

Supremediamond

suspension*


colloidal silica


Thermally-Spray Coated Specimens

NiCrAlY thermally-spray coated steel specimenrevealing a small amount of porosity (black spots),linear detachments (elongated black lines), andinclusions (gray particles) (unetched, 100X).

Microstructure of a steel substrate covered by twothermally-sprayed layers, a NiAl bond coat andyittria-zirconia top coat (unetched, 100X). The bondcoat contains pores, linear detachments andinclusions while the top coat is quite porous.

51

SINTERED CARBIDES

Sintered carbides are very hard materials madeby the powder metallurgy process and may byreinforced with several types of MC-typecarbides besides the usual tungsten carbide(WC). The binder phase is normally cobaltalthough minor use is made of nickel. Moderncutting tools are frequently coated with avariety of very hard phases, such as alumina,titanium carbide, titanium nitride and titaniumcarbonitride. Sectioning is normally performedwith a precision saw, so surfaces are very goodand rough abrasives are not usually required.Listed below are two alternate proceduresfor preparing sintered carbides using metal-bonded Ultra-Prep discs or the BuehlerHercules

H rigid grinding discs. A further option would bethe use of the traditional metal-bonded diamonddiscs.

Table 32. Four-Step Procedure for Sintered Carbides using Ultra-Prep Discs


Ultra-Prep 45-µm diamond 6 (27) 240-300 5:00metal-bonded disc water cooled Contra

Ultra-Prep 9-µm diamond 6 (27) 240-300 4:00metal-bonded disc water cooled Contra

Trident, Ultra-Pol, 3-µm Metadi 6 (27) 120-150 3:00Texmet 1000 Supreme diamond Contraor Nylon cloths suspension*

Trident, ~0.05-µm Mastermet 6 (27) 120-150 2:00Ultra-Pol colloidal silica or Contraor Nylon cloths Masterprep alumina

suspensions


Table 33. Four-Step Procedure for Sintered Carbides using BuehlerHercules H Discs


BuehlerHercules H 30-µm Metadi 6 (27) 240-300 5:00rigid grinding disc Supreme diamond Contra

suspension*

BuehlerHercules H 9-µm Metadi 6 (27) 240-300 4:00rigid grinding disc Supreme diamond Contra

suspension*

Trident , Ultra-Pol, 3-µm Metadi 6 (27) 120-150 3:00Texmet 1000 Supreme diamond Comp.or Nylon cloths suspension*

Microcloth, ~0.05-µm Mastermet 6 (27) 120-150 2:00Nanocloth or colloidal silica or ContraChemomet cloths Masterprep alumina

suspensions


Surface of a WC – 11.5% Co cutting tool enrichedwith 1.9% Ta and 0.4% Nb (form complex carbides,the dark regions in the matrix) and coated withalumina (arrows) for enhanced tool life (Murakami’sreagent, 1000X).

Sintered Carbides

52

If a greater amount of material must be removedin the planar grinding step, use either the30-µm resin bonded or the 45-µm metal-bondedUltra-Prep disc for step one, in Table 33.

The final step in these two procedures, Tables32 and 33, employs either Mastermet colloidalsilica or Masterprep alumina suspensions asthey will produce the best definition of the struc-ture, particularly if the surfaces have a complexseries of coatings for improved wear resistance,as used in some coated carbide inserts. How-ever, if such a coating is not present, the matrixgrain structure can be observed quite clearlyafter the 3-µm diamond step. For routine exami-nation there is little need to go beyond the thirdstep. It is also possible to use a 1-µm diamondstep, similar to step three, but for 2 minutes, asan alternative fourth step.

Sintered Carbides

Microstructure at the surface of a multilayered,CVD-coated WC - 8% Co cutting tool. The arrowspoint to the CVD layers of TiCN, TiN, TiC andalumina. The region below the coatings is madehigher in Co to improve crack resistance andcomplex carbide forming elements (Ta, Ti and Nb)are added to the matrix (dark spots at bottom) forwear resistance (Murakami’s reagent, 1000X).

HELPFUL HINTS FOR SINTERED

CARBIDES

• The ability to see the boundaries be-tween the WC particles and cobaltbinder in the as-polished condition de-pends on the surface of the polishingcloth used in the last step. To see noboundaries, which makes it easier tosee graphite or eta phase, use a nap-less surface, such as Trident, Texmet1000 or Ultra-Pol pads. To see the phaseboundaries between WC particles andthe cobalt binder, use a medium napsurface, such as Microcloth, Mastertexor Nanocloth pads.

53

CERAMICS

Ceramic materials are extremely hard and brittleand may contain pores. Sectioning must be per-formed using diamond blades. If the specimen isto be thermally etched, then it must be mountedin a resin that permits easy demounting, andvacuum infiltration of epoxy into the voids shouldnot be done. Deformation and smearing are notproblems with ceramics due to their inherentcharacteristics. But, it is possible to break outgrains or produce cracking during preparation.

Ceramics

Table 34. Four-Step Procedure Suitable for Most Ceramic Materials


Ultra-Prep 45-µm diamond 8 (36) 120-150 Untilmetal-bonded disc Metadi fluid Comp. Flat

for coolant

BuehlerHercules H 9-µm Metadi 6 (27) 120-150 4:00rigid grinding disc Supreme diamond Comp.

suspension withMetadi fluid

BuehlerHercules S 3-µm Metadi 6 (27) 120-150 3:00rigid grinding disc Supreme diamond Comp.

suspension*

Texmet 1000, 1-µm Metadi 10 (44) 120-150 5:00Nylon or Supreme diamond ContraUltra-Pol suspension withcloths Metadi fluid or

Masterpolish 2suspension


Table 35. Four-Step Procedure Suitable for Most Ceramic Materials


Ultra-Prep 45-µm diamond 8 (36) 120-150 Untilmetal-bonded Metadi fluid Contra Flatdisc for coolant

Ultra-Pad 15-µm Metadi 6 (27) 150 6:00cloth Supreme diamond Comp.

suspension*

Ultra-Pad or 6-µm Metadi 6 (27) 150 4:00Texmet 1000 Supreme diamond Comp.or 2000 cloths suspension*

Texmet 1000 or 3-µm Metadi 8 (36) 120-150 4:00Texmet 2000, Supreme diamond ContraNylon or suspension andUltra-Pol Mastermetcloths colloidal silica

(50:50 simultaneously)


Microstructure of hot-pressed silicon nitride (binderunknown) (unetched, 500X).

54

Ceramics

P u l l o u t sare a major problem to control as they can bemisinterpreted as porosity. Mechanical prepa-ration has been done successfully with laps,metal- bonded diamond discs, rigid grindingdiscs, or hard cloths. SiC paper is rather ineffec-tive with most ceramics, as they are nearly ashard or as hard as the SiC abrasive. Conse-quently, diamond is commonly used for nearlyall preparation steps. Automated preparation ishighly recommended when preparing ceramicmaterials as very high forces arise between thespecimen and the working surface, often toohigh for manual preparation. Listed are two ge-neric procedures for preparing ceramics usingdifferent approaches, see Tables 34 and 35.

Masterpolish 2 suspension is specifically for-mulated for final polishing ceramic materials andwill yield a better surface finish than 1-µm dia-mond.

HELPFUL HINTS FOR CERAMICS

• Cut specimens as close as possible tothe desired examination plane using asaw and blade that produces the leastpossible amount of damage, producingthe best possible surface for subsequentpreparation. Rough grinding is slow andtedious and does considerable damageto the structure and should be avoidedas much as possible.

• Coarse abrasives for grinding may beneeded at times. However, severe grainpull-outs can result from use of coarseabrasives and it is recommended touse the finest possible coarse abrasivefor grinding.

Microstructure of an alumina (matrix), zirconia(gray), silica (white) refractory material (unetched,200X).

55

COMPOSITES

Composites cover a wide range of compositions,general grouped under the subcategoriesmetal-matrix composites (MMC), polymer-ma-trix composites (PMC) and ceramic-matrixcomposites (CMC). Consequently, preparationschemes are quite difficult to generalize due tothe extreme range of materials employed, dif-ferences in hardness and grinding/polishingcharacteristics; so, relief control is a big issue.Pull out problems are also very common, espe-cially with PMCs. Sectioning frequently producesconsiderable damage that must be removed inthe initial preparation step. Mounting withcastable epoxy resin along with vacuum impreg-nation is frequently performed.

Microstructure of a metal-matrix composite, SiC ina CP titanium matrix (unetched, Nomarski DIC,100X).

(top) Microstructure of a polymer-matrixcomposite, graphite fabric-reinforced polysulfone(unetched, polarized light, 100X).(bottom) Microstructure of alumina fibers in analuminum - lithium alloy matrix (500X, as-polished,Nomarski DIC).

HELPFUL HINTS FOR COMPOSITES

• Sectioning damage can propagatedown the fibers and can be difficult toremove. Prevent damage by using pre-cision saw sectioning and diamondwafering blades with a very find abra-sive, specifically like 5 LC and 10 LCSeries. Thinner blades produce lessdamage. Section as close as possibleto the plane of interest.

• Composites can be damaged by com-pression mounting. Castable mountingresins with low peak temperature arepreferred.

• Avoid aggressive grinding steps by us-ing longer times and finer abrasives.

56

Table 37. Four-Step Procedure for Polymer-Matrix Composites

Table 38. Four-Step Procedure for Ceramic-Matrix Composites

Composites


Ultra-Prep 45-µm diamond 6 (27) 240-300 Untilmetal-bonded Metadi fluid Contra Flatdisc for coolant

Ultra-Pol, 15-µm Metadi 6 (27) 150 4:00Ultra-Pad, or Supreme ContraTexmet 1000 or 2000 diamondcloths suspension*

Ultra-Pol, 6-µm Metadi 6 (27) 150 3:00Trident, or Supreme ContraTexmet 1000 diamondor 2000 cloths suspension*

Ultra-Pol, 1-µm Metadi 6 (27) 120-150 2:00Chemomet Supreme diamond Contraor Nylon suspension andcloths ~0.05-µm Mastermet

colloidal silica(50:50 simultaneously)

*Plus Metadi Fluid extender as desiredContra = Platen and specimen holder rotate in opposite directions


Carbimet 320- (P400) grit 4 (18) 150-240 Untilabrasive discs SiC water cooled Comp. Flat(waterproof paper)

Ultra-Pol, or 6-µm Metadi 4 (18) 120-150 4:00Ultra-Pad Supreme Comp.cloths diamond

suspension*

Texmet 1000, 3-µm Metadi 5 (22) 120-150 4:00Trident or Supreme Comp.Nylon cloths diamond

suspension*

Chemomet or ~0.05-µm Mastermet 6 (27) 120-150 2:00-4:00Microcloth Colloidal Silica or Comp.pads MASTERPREP

Alumina Suspension

*Plus METADI® Fluid extender as desiredComp. = Complementary (platen and specimen holder both rotate in the same direction)

57

PRINTED CIRCUIT BOARDS

The vast majority of printed circuit boards(PCB’s) are of a rigid variety, the bulk of whichare composed of layers of woven glass fibercloth in a polymeric matrix. Flex circuits, whichare becoming quite common, do not typicallycontain glass fiber, but instead, their bulk is of-ten composed of layers of polyimide. Thecircuitry in both types of boards is composed ofplated and/or foil metal. The metal used is gen-erally copper, while in a few cases, gold and/ornickel plating may be present. Furthermore, de-pending upon whether the boards haveundergone assembly or shock testing, soldersof various compositions might also be present.

Luckily for the metallographer, the variety of ma-terials present in PCB’s generally do notcomplicate the preparation methods due to the

fact that extremely hard and brittle materials arenot commonly found in the boards. This changes,however, when ‘packed’ boards with ceramic orsemiconductor components must be sectioned.Please refer to methods of preparation forceramics or microelectronic materials for infor-mation on these cases. A generalized methodfor nonpacked PCB’s of both the rigid and flexvarieties is presented in Table 39.

Quality control of manufactured PCB’s oftendemands statistical analysis based on platingthickness measurements taken from the centersof plated through-holes. However, to generateenough data for a representative sampling,numerous specimens must be prepared. This iseasiest when coupons are taken from PCB’s andare ganged together in precise alignment. In thismanner, the through holes in all of the couponscan be sectioned at one time, with one setof consumables. The Buehler Nelson-ZimmerSystem allows this type of automation with fewadjustments to the method listed below.

Printed Circuit Boards

Table 39. Five-Step Procedure for Non-Packed Printed Circuit Boards


Carbimet abrasive 180- (P180) grit SiC 5 (22) 120-150 Untildiscs (waterproof paper) water cooled Contra Flat

Carbimet 320- (P400) grit SiC 5 (22) 120-150 1:30abrasive discs water cooled Contra

Carbimet 600- (P1200) grit SiC 5 (22) 120-150 1:30abrasive discs water cooled Contra

Texmet 1000, 3-µm Metadi II 5 (22) 150-240 1:30or Trident cloth diamond paste* Contra

Microcloth 0.05-µm Masterprep 5 (22) 100-150 1:15cloth alumina suspension Contra

*Plus Metadi Fluid extender as desiredContra = Platen and specimen holder rotate in opposite directions

Micrograph showing a plated-through hole in aprinted circuit board (resin and fibers) with copperfoil, electroless-plated copper and Pb-Sn solder(equal parts ammonium hydroxide and 3% hydrogenperoxide, 200X).

HELPFUL HINTS FOR PRINTED

CIRCUIT BOARDS

• To improve penetration of acrylic mount-ing material into the through-holes of aPCB specimen, dip the “pinned” PCBcoupons in liquid hardener immediatelyprior to mounting in acrylic. This will“wick” the acrylic into the through-holesand minimize the presence of problem-atic bubbles in these critical locations.

58

ELECTRONIC MATERIALS

The term, ‘microelectronic materials’ encom-passes an extremely wide range of materials.This is due to the fact that most microelectronicdevices are composites, containing any numberof individual components. For example, presentday microprocessor failure analysis might requirethe metallographer to precisely cross sectionthrough a silicon chip plated with multiple thin-film layers of oxides, polymers, ductile metalssuch as copper or aluminum, and refractorymetals such as tungsten and/or titanium-tung-sten. In addition, the packaging of such a devicemight contain materials of such varying me-chanical properties as toughened aluminumoxide and solder. The solder materials may havecompositions ranging up to 97% lead. With sucha vast number of materials incorporated into asingle device, and with these materials havingsuch highly disparate mechanical properties, itis virtually impossible to develop a generalmethod for achieving perfect metallographic re-sults. Instead, we must focus on a few individual

materials, and develop a philosophy of prepa-ration in which we give our attention specificallyto the materials of interest.

First and foremost in the class of ‘microelectronicmaterials’ is silicon. Silicon is a relatively hard,brittle material, which does not respond well togrinding with large silicon carbide abrasives.Silicon carbide papers contain strongly bondedabrasive particles which, when they collide withthe leading edge of a piece of silicon, createsignificant impact damage. In addition, they cre-ate tensile stresses on the trailing edge ofsilicon, which results in deep and destructivecracking. Cutting close to the target area is pref-erable to grinding, but to accurately approach thetarget area within a silicon device, fine grindingis still necessary. This is the point at which sili-con preparation divides into two distinct classes.The first is preparation through traditional metal-lographic techniques. The second is preparationof unencapsulated silicon chips (or die), usingspecial fixturing and abrasives. This second cat-egory is a specialized field of metallography,which is discussed in other texts [8] and is be-yond the scope of this book. Therefore, we willfocus on standard metallographic techniquesthroughout this discussion.

Standard preparation of epoxy encapsulatedsilicon is similar to general metallographicpreparation, with the exception that only veryfine grades of silicon carbide paper are used.Table 40 illustrates a typical process of prepa-ration.

Electronic Materials

Table 40. Three-Step Procedure for Preparing Silicon in Microelectronic Devices


Ultra-Prep, Type A 15-µm diamond Manual 100-150 Untildiamond lapping film water cooled Flaton Trident cloth

Texmet 1000, 3-µm diamond Metadi II 7 (31) 240-300 2:00diamond paste/ Comp

Masterprep as lubricant

Chemomet 0.02-µm 5 (22) 100-150 1:30pad Mastermet 2 Contra

colloidal silica

Comp. = Complementary(platen and specimen holder both rotate in the same direction)Contra = Platen and specimen holder rotate in opposite directions

Aluminum circuitry on silicon adjacent to mineralfilled epoxy packaging, (1000x).

59

Using silicon carbide grinding paper with abra-sive sizes coarser than 600- (1200) grit maycause serious damage to the silicon. Therefore,this method often requires precision cutting orcleaving near to the target area of the specimenprior to the initial grinding step. Care must betaken to assure that only precision saw bladesdesigned especially for microelectronic materi-als are used when cutting.

Silicon responds very well to the chemo-me-chanical polishing effects of colloidal silica, butnot as well to aluminum oxide. However, thereare instances when an aluminum oxide prod-uct, such as the Masterprep suspension, shouldbe used for final polishing. This relates back tothe philosophy of preparation mentioned ear-lier. An example of the appropriate use ofMasterprep alumina would be: when a silicondie is attached to a lead frame material such asnickel-plated copper, and the lead frame anddie attach materials are the materials of interest.

In this case, we are not so concerned with thesurface finish of the silicon, but instead, we wantto be sure that the nickel doesn’t smear, andthat we are able to discern the true microstruc-ture of the lead frame materials (refer to sectionlisting preparation methods for copper andnickel).

When preparing a silicon device for the purposeof inspecting the metalized, thin film circuitry,the techniques would be the same as those listedabove. Again, the choice of final polishingagent would be determined entirely by the fea-tures of interest. For instance, aluminum circuitryresponds extremely well to the chemomechanicaleffects of colloidal silica, but tungsten and tita-nium-tungsten are not as effectively polishedwith colloidal silica as are the materials by whichthey are surrounded. Therefore, colloidal silicaoften causes such refractory metals to stand inrelief against a well polished background. Thisresults in edge rounding of the refractory metalsand can make interface analysis quite difficult.An alternative is to use an extremely fine dia-mond suspension for final polishing in order toreduce these effects.

When lead alloy solders are the materials ofinterest in a microelectronic package, we canoften separate them into two general catego-ries. The first is the eutectic, or near-eutecticsolders. These tend to be ductile, but they re-spond well to the same general preparation listedin Table 40. (During the 3-µm polishing step, the


Table 41. Procedure for High-Temperature Solder on Ceramic Devices


BuehlerHercules S 9-µm Metadi II 7 (31) 100-150 Untilrigid grinding disc diamond paste/ Flat

Masterprepas lubricant*

Texmet 3-µm Metadi II 7 (31) 240-300 Until1000 diamond paste/ Comp all embedded

Masterprep 9-µm diamondas lubricant is removed

Chemomet 0.02-µm 5 (22) 100-150 2:00pad Mastermet 2 Contra

colloidal silica

*Apply sparinglyComp. = Complementary(platen and specimen holder both rotate in the same direction)Contra = Platen and specimen holder rotate in opposite directions

High temperature (96% Pb) solder under flip chipon ceramic substrate, (200x).

60

best results are obtained by using diamondpaste. Pastes contain a waxy base that reducesdiamond embedding in ductile metals. Limit theuse of extenders in such cases. A few drops ofwater lubrication does not break down the waxybase, and therefore it retains the desired prop-erties of diamond paste.)

The second category of lead solders is the hightemperature solders (in the range of 90% to 97%lead). These are difficult to prepare, and requirespecial consideration. This is especially truewhen they are incorporated into a device thatcontains a ceramic package. The mechanicalproperties of these toughened ceramics requireaggressive grinding techniques, which causeceramic debris (and often, the abrasive used forthe grinding operation) to became embeddedin the solder. Silicon carbide abrasives are apoor choice for grinding in such cases, as sili-con carbide is only slightly harder (andsignificantly more brittle) than the typical pack-aging material they are intended to grind. As thesilicon carbide abrasive fractures, it produceselongated shards, which embed deeply intohigh-temperature solders. In addition, they donot grind the ceramic package effectively, andtherefore produce extreme edge rounding at theinterface with the solder. Diamond grinding pro-duces a more desired result since diamondabrasives are blockier in shape, and are there-fore easier to remove after they have embeddedthemselves. In addition, they are capable of cre-ating a flat surface when ceramic packages arepresent.

High removal rates and good surfacefinishes have been obtained on ceramic pack-ages when using diamond paste on theBuehlerHercules S discs.

Polishing of high-temperature solders attachedto ceramics often produces undesired edgerounding. However, this cannot be completelyavoided. The cushioning effects of polishing padswill tend to cause the abrasive to preferentiallyremove material from ductile materials faster thanfrom hard, brittle materials. By using stiff, flatpolishing pads, one can reduce this effect, but atthe expense of having additional material em-bedded in the high-temperature solder. Oneeffective preparation technique is to perform agrind-polish-grind-polish type of procedure. Inthis technique, a polishing step, which utilizes anapped polishing cloth, is used in conjunctionwith a step-down in abrasive size from the pre-vious grinding step. This process is used to polishout the embedded material. The next step is agrinding process, utilizing another step-downin abrasive size, to flatten the specimen again.This continues to a sufficiently fine polishingstage at which final polishing will produce thedesired result. Table 41 illustrates an exampleof such a process.


HELPFUL HINTS FOR ELECTRONIC

MATERIALS

• Consider what area of the specimen ismost critical when choosing a finalpolishing suspension. For example,the chemo-mechanical nature of colloi-dal sil icas (such as Mastermetand Mastermet 2) make them idealfor polishing silicon, glass, oxides,aluminum, and to some degree copper.However, polishing nickel platings andgold ball bonds with these agentsoften results in smearing. For such non-reactive materials, use Masterprepaluminum oxide final polishing suspen-sion to produce a flat, scratch-freesurface.

Silicon flip chip, solder balls, on PC board, (200x).

61

POLYMERS

Plastics and polymers are normally quite soft.Many different sectioning methods have beenused. A sharp razor blade or scalpel, or even apair of scissors, can be used. Microtomes havebeen used, and a surface cut after refrigerationin liquid nitrogen requires very little subsequentpreparation. A jeweler’s saw may be used, al-though the surface is somewhat rough. Theprecision saw produces excellent surfaces, whilean abrasive cut-off saw yields a bit more dam-age and a rougher surface. Blades and wheelsfor sectioning polymers are available fromBuehler. Damage from sectioning is quite easyto remove.

Surface quality can be degraded by abrasionfrom the debris produced during grinding andpolishing. Mounted specimens are much easierto prepare than nonmounted specimens.Castable resins are preferred as the heat from a

mounting press may damage or alter the struc-ture of the specimen. However, there may be avisibility problem if a transparent, clear poly-meric specimen is mounted in a clear,transparent epoxy. In this case, Epo-Colorresin, with its deep red color, will produce ex-cellent color contrast between specimen andmount in darkfield or polarized light. Due totheir light weight, a specimen may float in theepoxy filled mold. To prevent this, put doublesided adhesive tape on the bottom of the moldcup and set the specimen on top of the tape.Always use practices that minimize theexotherm during polymerization.

Preparation of plastics and polymers for mi-crostructural examination follows the samebasic principles as for other materials. Auto-mated polishing devices are preferred tomanual preparation. Rough grinding abrasivesare unnecessary, even for the planar grinding

Table 42. Generic Procedure for Preparing Plastic and Polymeric Specimens


Carbimet 320- (P400) grit SiC 4 (18) 240 Untilabrasive discs water cooled Contra Flat(waterproof paper)

Carbimet 400- (P600) grit SiC 4 (18) 240 0:30abrasive discs water cooled Contra



Texmet 1000 or 3-µm Metadi II 5 (22) 120 4:00Trident diamond paste Comp.cloths and Metadi fluid

Mastertex 0.05-µm Masterprep 3 (13) 120 4:00cloth alumina suspension Contra

Comp. = Complementary (platen and specimen holder both rotate in the same direction)Contra = Platen and specimen holder rotate in opposite directions

Polymers

Micrograph of high-density polyethylene containing afiller (white particles)(unetched, Nomarski DIC,100X).

Micrograph showing multi-layered polymer meatcasing mounted in epoxy (dark field, 100X).

62

step. Pressures should be lighter than used formost metals. Water is generally used as the cool-ant, although some plastics and polymers mayreact with water. In such cases, use a fluid thatwill not react with the particular plastic or poly-mer. Embedding can be a problem with plasticsand polymers. ASTM E 2015 (Standard Guidefor Preparation of Plastics and Polymeric Speci-mens for Microstructural Examination) describesprocedures for preparing several types of plas-tics and polymers. Table 42 contains a genericprocedure for preparing many plastic and poly-meric materials. Depending upon the materialand the surface roughness after sectioning, itmay be possible to start grinding with 400-(P600) or even 600- (P1200) grit SiC paper.

If flatness is critical, the final step can be altered.Use a Texmet pad with Masterprep aluminaabrasive slurry at 10 lbs. pressure (44N)per specimen, 120 rpm, contra rotation, for 3minutes.

HELPFUL HINTS FOR POLYMERS

• Plastics often have little contrast fromthe mounting media when viewed un-der a microscope. This makes thicknessmeasurement and discerning edgesdifficult. Mounting in Epo-Color dyeenhanced epoxy will resolve this situa-tion by producing excellent colorcontrast between the specimens andmount.

• Polymers and plastics may react withfluids during sectioning and prepara-tion. Always check your particularpolymer and choose the proper fluid,usually a water, or oil based solution toavoid a reaction.

Polymers

63

lution is not improved beyond the limit of 0.2-0.3-µm for the light microscope. Microscopicexamination of a properly prepared specimenwill clearly reveal structural characteristics suchas grain size, segregation, and the shape, size,and distribution of the phases and inclusionsthat are present. Examination of the microstruc-ture will reveal prior mechanical and thermaltreatments that the metal has received. Many ofthese microstructural features are measured ei-ther according to established image analysisprocedures, e.g., ASTM standards, or internallydeveloped methods.

Etching is done by immersion or by swabbing(or electrolytically) with a suitable chemical solu-tion that essentially produces selective corrosion.Swabbing is preferred for those metals andalloys that form a tenacious oxide surface layerwith atmospheric exposure such as stainlesssteels, aluminum, nickel, niobium, and titaniumand their alloys. It is best to use surgical gradecotton that will not scratch the polished surface.Etch time varies with etch strength and can onlybe determined by experience. In general, for highmagnification examination the etch depth shouldbe shallow; while for low magnification examina-tion a deeper etch yields better image contrast.Some etchants produce selective results in thatonly one phase will be attacked or colored. A vastnumber of etchants have been developed; thereader is directed to references 1-3, 9 and ASTME 407. Table 43 lists some of the most commonlyused etchants for the materials described in thisbook.

Etchants that reveal grain boundaries are veryimportant for successful determination of thegrain size. Grain boundary etchants are given in[1-3, 9]. Problems associated with grain bound-ary etching, particularly prior austenite grainboundary etching, are given in [2, 10 and 11].Measurement of grain size in austenitic or face-centered cubic metals that exhibit annealing twinsis a commonly encountered problem. Etchantsthat will reveal grain boundaries, but not twinboundaries, are reviewed in [2].

ETCHING

Metallographic etching encompasses all pro-cesses used to reveal particular structuralcharacteristics of a metal that are not evident inthe as-polished condition. Examination of aproperly polished specimen before etching mayreveal structural aspects such as porosity, cracks,and nonmetallic inclusions. Indeed, certainconstituents are best measured by image analy-sis without etching, because etching will revealadditional, unwanted detail and make detectiondifficult or impossible. The classic examples arethe measurement of inclusions in steels andgraphite in cast iron. Of course, inclusions arepresent in all metals, not just steels. Many inter-metall ic precipitates and nitrides can bemeasured effectively in the as polishedtion.

In certain nonferrous alloys that have non-cubiccrystallographic structures (such as beryllium,hafnium, magnesium, titanium, uranium and zir-conium), grain size can be revealed adequatelyin the as polished condition using polarized light.Figure 30 shows the microstructure of cold-drawn zirconium viewed in cross-polarized light.This produces grain coloration, rather than a“flat etched” appearance where only the grainboundaries are dark.

Etching ProceduresMicroscopic examination is usually limited to amaximum magnification of 1000X — the approxi-mate useful limit of the light microscope, unlessoil immersion objectives are used. Many imageanalysis systems use relay lenses that yieldhigher screen magnifications that may makedetection of fine structures easier. However, reso-

Etching

Figure 30. Mechanical twins at the surface of hotworked and cold drawn high-purity zirconium viewedwith polarized light (200X).

64

either the anodic or cathodic constituent in amicrostructure. Tint etchants are listed and illus-trated in several publications [1-3, 16-21].

A classic example of the different behavior ofetchants is given in Figure 31 where low-car-bon sheet steel has been etched with thestandard nital and picral etchants, and also acolor tint etch. Etching with 2% nital reveals theferrite grain boundaries and cementite. Note thatmany of the ferrite grain boundaries are miss-ing or quite faint; a problem that degrades theaccuracy of grain size ratings. Etching with 4%picral reveals the cementite aggregates (onecould not call this pearlite as it is too nonlamellarin appearance and some of the cementite existsas simple grain boundary films) but no ferritegrain boundaries. If one is interested in theamount and nature of the cementite (which caninfluence formability), then the picral etch is farsuperior to the nital etch as picral revealed onlythe cementite. Tint etching with Beraha’s solu-tion (Klemm’s I could also be used) colored thegrains according to their crystallographic orien-tation. With the development of color imageanalyzers, this image can now be used quiteeffectively to provide accurate grain size mea-surements since all of the grains are colored.

Figure 32 shows a somewhat more complex ex-ample of selective etching. The micrographshows the ferrite-cementite-iron phosphideternary eutectic in gray iron. Etching sequentially

with picral and nital revealed the eutectic, Fig-ure 32a, surrounded by pearlite. Etching withboiling alkaline sodium picrate, Figure 32b, col-ored the cementite phase only, including in thesurrounding pearlite (a higher magnification isrequired to see the very finely spaced cement-ite that is more lightly colored). Etching with

Selective EtchingImage analysis work is facilitated if the etchantchosen improves the contrast between the fea-ture of interest and everything else. Thousandsof etchants have been developed over the years,but only a small number of these are selective innature. Although the selection of the best etchant,and its proper use, is a very critical phase of theimage analysis process, only a few publicationshave addressed this problem [12-14]. Selectiveetchants, that is, etchants that preferentiallyattack or color a specific phase, are listed in[1-3, 9, 13 and 14] and illustrated in 13 and 14.Stansbury [15] has described how potentiostaticetching works and has listed many preferentialpotentiostatic etching methods. The potentiostatoffers the ultimate in control over the etching pro-cess and is an outstanding tool for this purpose.Many tint etchants act selectively in that they color

Figure 32a. The ternary eutectic (ααααα-Fe3C-Fe3P) ingray cast iron revealed by etching (a, top) in picraland nital to “outline” the phases

b

a

a

Etching

Figure 31. Microstructure of low-carbon sheet steeletched with (a, top) 2% nital, (b, middle) 4% picral;and (c, bottom) Beraha’s reagent (100 mL water,10 g Na2S2O3 and 3 g K2S2O5) at 100X.

c

65

boiling Murakami’s reagent, Figure 32c, colorsthe iron phosphide darkly and will lightly colorthe cementite after prolonged etching. The fer-rite could be colored preferentially if Klemm’s Iwas used.

Figure 33. Microstructure of a duplex stainless steelrevealed by electrolytic etching with (c) aqueous60% HNO3 (1 V dc, 20 s); with (d) aqueous 10%oxalic acid (6 V dc, 75 s); with (e) aqueous 10%CrO3 (6 V dc, 30 s); and, with (f) aqueous 2% H2SO4(5 V dc, 30 s) at 200X.

Figure 33. Microstructure of a duplex stainless steelrevealed using (a, top) alcoholic 15% HCl byimmersion (30 min.); and, with (b, bottom)glyceregia by swabbing (2 min.) at 200X.

Selective etching has been commonly appliedto stainless steels for detection, identificationand measurement of delta ferrite, ferrite in dual

phase grades, and sigma phase. Figure 33shows examples of the use of a number of popu-lar etchants to reveal the microstructure of a dualphase stainless steel in the hot rolled and an-nealed condition. Figure 33a shows a welldelineated structure when the specimen wasimmersed in ethanolic 15% HCl for 30 minutes.All of the phase boundaries are clearly revealed,

Figure 32b. (top) boiling alkaline sodium picrate tocolor the cementite. Figure 32c. (bottom) boilingMurakami’s reagent to color the phosphide (200X).

b

a

e

d

c

f

b

c

Etching

66

Figure 33. Selective coloring of ferrite in a duplexstainless steel by electrolytic etching with (g)aqueous 20% NaOH (4 V dc, 10 s) and with (h)aqueous 10 N KOH (3 V dc, 4 s); and by immersioncolor etching (i) with Beraha’s reagent (100 mLwater, 10 mL HCl and 1 g K2S2O5) at 200X.

but there is no discrimination between ferriteand austenite. Twin boundaries in the austeniteare not revealed. Glyceregia, a popular etchantfor stainless steels, was not suitable for thisgrade, Figure 33b, as it appears to be ratherorientation sensitive. Many electrolytic etchantshave been used for etching stainless steels, butonly a few have selective characteristics. Of thefour shown in Figures 33c to f, only aqueous60% nitric acid produced any gray level discrimi-nation between the phases, and that was weak.All revealed the phase boundaries nicely, how-ever. Two electrolytic reagents, shown in Figures33g and h, are commonly used to color ferrite indual phase grades and delta ferrite in martensi-tic grades. Of these, aqueous 20% sodiumhydroxide, Figure 33g, usually gives more uni-form coloring of the ferrite. Murakami’s andGroesbeck’s reagents have also been used for

this purpose. Tint etchants have been developedby Beraha that color the ferrite phase nicely, asdemonstrated in Figure 33i.

Selective etching techniques are not limited toiron-based alloys, although these have morethoroughly developed than for any other alloysystem. Selective etching of beta phase inalpha-beta copper alloys has been a popular

Figure 34. Beta phase colored in Naval Brass (Cu-39.7% Zn-0.8% Sn) by immersion color etching withKlemm’s I reagent (200X).

subject. Figure 34 illustrates coloring of betaphase in Naval Brass (UNS 46400) usingKlemm’s I reagent. Selective etching has a longhistorical record for identification of intermetallicphases in aluminum alloys. This method wasused for many years before the development ofenergy-dispersive spectroscopy. Today, it is still

useful for image analysis work. Figure 35 showsselective coloration of theta phase, CuAl2, in theAl-33% Cu eutectic alloy. As a final example,Figure 36 illustrates the structure of a simpleWC-Co sintered carbide, cutting tool. In the as-polished condition, Figure 36a, the cobalt bindercan be seen faintly against the more grayishtungsten carbide grains. A few particles of graph-ite are visible. In Figure 36b, light relief polishinghas brought out the outlines of the cobalt binder

Figure 35. Theta phase, CuAl2, colored in the á-Al –è eutectic in an as-cast Al-33% Cu specimen byimmersion color etching with the Lienard andPacque etch (200 mL water, 1 g ammoniummolybdate, 6 g ammonium chloride) at 1000X.

h

g

i

Etching

67

Electrolytic Etching and AnodizingThe procedure for electrolytic etching is basi-cally the same as for electropolishing exceptthat voltage and current densities are consider-ably lower. The specimen is made the anode,and some relatively insoluble but conductivematerial such as stainless steel, graphite, orplatinum is used for the cathode. Direct currentelectrolysis is used for most electrolytic etching.Electrolytic etching is commonly used with stain-less steels, either to reveal grain boundarieswithout twin boundaries, or for coloring ferrite(as illustrated) in Figure 33, delta ferrite, sigmaor chi phases. Anodizing is a term applied to elec-trolytic etchants that develop grain colorationwhen viewed with crossed-polarized light, as inthe case of aluminum, tantalum, titanium, tung-sten, uranium, vanadium and zirconium [2].Figure 37 shows the grain structure of 5754 alu-minum revealed by anodizing with Barker’sreagent, viewed with crossed-polarized light.Again, color image analysis makes thisimage useful now for grain size measurements.

Heat TintingAlthough not commonly utilized, heat tinting[2] is an excellent method for obtaining colorcontrast between constituents or grains. Anunmounted polished specimen is placed faceup in an air-fired furnace at a set temperatureand held there as an oxide film grows on thesurface. Interference effects, as in tint etching,create coloration for film thicknesses within acertain range, about 30-500 nm. The observedcolor is a function of the film thickness. Natu-rally, the thermal exposure cannot alter themicrostructure. The correct temperature must be

Figure 36. Microstructure of WC-Co cutting tool: a)as-polished revealing graphite particles; b) reliefpolished revealing the cobalt binder phase; c) afterimmersion in Chaporova’s etch (HCl saturated withFeCl3) to attack and “ darken” the cobalt; and, d)after (c) plus Murakami’s reagent to outline the WCgrains (1000X).

phase, but this image is not particularly usefulfor image analysis. Etching in a solution ofhydrochloric acid saturated with ferric chloride,Figure 36c, attacks the cobalt and provides gooduniform contrast for measurement of the cobaltbinder phase. Following this with Murakami’sreagent at room temperature reveals the edgesof the tungsten carbide grains, useful for evalua-tion of the WC grain size, Figure 36d.

b

a

c

d

Etching

Figure 37. Grain structure of annealed 5754aluminum sheet revealed by anodizing with Barker’sreagent (30 V dc, 2 min.) and viewing with polarizedlight plus sensitive tint (100X).

68

HELPFUL HINTS FOR ETCHING

• Many etchants can be used by swab-bing or by immersion. Swabbing ispreferred for those specimens that forma tight protective oxide on the surfacein air, such as Al, Ni, Cr, stainless steels,Nb (Cb), Ti and Zr. However, if theetchant forms a film, as in tint etchants,then immersion must be used as swab-bing will keep the film from forming.Keller’s reagent reveals the grain sizeof certain aluminum alloys by forming afilm. This will not occur if the etch is usedby swabbing.

• Many etchants, and their ingredients, dopresent potential health hazards to theuser. ASTM E 2014, Standard Guide onMetallography Laboratory Safety, de-scribes many of the common problemsand how to avoid them.

determined by the trial-and-error approach, butthe procedure is reproducible and reliable. Fig-ure 38 shows the grain structure of CP titaniumrevealed by heat tinting.

Interference Layer MethodThe interference layer method [2], introducedby Pepperhoff in 1960, is another procedure forobtaining a film over the microstructure thatgenerates color by interference effects. In thismethod, a suitable material is deposited on thepolished specimen face by vapor deposition toproduce a low-absorption, dielectric film with ahigh refractive index at a thickness within therange for interference. Very small differences inthe natural reflectivity between constituents and

Figure 38. Grain structure of annealed CP titaniumrevealed by heat tinting on a hot plate (100X,polarized light plus sensitive tint).

the matrix can be dramatically enhanced bythis method. Suitable materials for the produc-tion of evaporation layers have been summarizedin [22,23]. The technique is universally applicable,but does require a vacuum evaporator. Its mainweakness is difficulty in obtaining a uniformlycoated large surface area for image analysismeasurements.

Etching

69

Table 43. Commonly Used Etchants for Metals and Alloys3.

LIGHT METALS - Aluminum and AlloysComposition Comments

1. 95 mL water Keller’s reagent, very popular general purpose reagent for Al and Al alloys,2.5 mL HNO3 except high-Si alloys. Immerse sample 10-20 seconds, wash in warm water.1.5 mL HCI Can follow with a dip in conc. HNO3. Outlines all common constituents,1.0 mL HF reveals grain structure in certain alloys when used by immersion.

2. 90-100 mL water General-purpose reagent. Attacks FeAl3, other constituents outlined. The0.1-10 mL HF 0.5% concentration of HF is very popular.

3. 84 mL water Graff and Sargent’s etchant, for grain size of 2XXX, 3XXX,6XXX, and 7XXX15.5 mL HNO3 wrought alloys. Immerse specimen 20-60 seconds with mild agitation.0.5 mL HF3 g CrO3

4. 1.8% fluoboric acid in water Barker’s anodizing method for grain structure. Use 0.5-1.5 A/in2, 30-45 V dc.For most alloys and tempers, 20 seconds at 1 A/in2 and 30 V dc at 20°C issufficient. Stirring not needed. Rinse in warm water, dry. Use polarized light;sensitive tint helpful.

Magnesium and AlloysComposition Comments

5. 25 mL water Glycol etch, general purpose etch for pure Mg and alloys. Swab specimen 3-575 mL ethylene glycol seconds for F and T6 temper alloys, 1-2 minutes for T4 and 0 temper alloys.1 mL HNO3

6. 19 mL water Acetic glycol etchant for pure Mg and alloys. Swab specimen 1-3 seconds for60 mL ethylene glycol F and T6 temper alloys, 10 seconds for T4 and 0 temper alloys. Reveals grain20 mL acetic acid boundaries in solution-treated castings and most wrought alloys.1 mL HNO3

7. 100 mL ethanol For Mg and alloys. Use fresh. Immerse specimen for 15-30 seconds.10 mL water Produces grain contrast.5 g picric acid

LOW MELTING POINT METALS - Sb, Bi, Cd, Pb, Sn and ZnComposition Comments

8. 100 mL water For Sb, Bi and alloys. Immerse specimen up to a few minutes.30 mL HCI2 g FeCI3

9. 100 mL water For Sb-Pb, Bi-Sn, Cd-Sn, Cd-Zn, and Bi-Cd alloys. Immerse specimen up to a25 mL HCI few minutes.8 g FeCI3

10. 95-99 mL ethanol For Cd, Cd alloys, Sn, Zn alloys, Pb and alloys, Bi-Sn eutectic alloy and Bi-Cd1-5 mL HNO3 alloys. Can add a few drops of zephiran chloride. Immerse sample. For Pb

and alloys, if a stain forms, wash in 10% alcoholic HCI.

11. 100 mL water Pollack’s reagent for Pb and alloys. Immerse specimen 15-30 seconds. Other10 g ammonium molybdate compositions used are: 100 mL: 9 g: 15 g and 100 mL: 10 g: 25 g.10 g citric acid

12. 100 mL water For Sn-based babbitt metal. Immerse specimen up to 5 minutes.2 mL HCI10 g FeCI3

13. 200 mL water Palmerton reagent for pure Zn and alloys. Immerse specimen up to 340 g CrO3 minutes. Rinse in 20% aqueous CrO3.3 g Na2SO4

14. 200 mL water Modified Palmerton reagent for Zn die-casting alloys. Immerse specimen for10 g CrO3 several seconds, rinse in 20% aqueous CrO3.1 g Na2SO4

3 When water is specified, always use distilled water. It is best to use reagent grade chemicals. Etchants can be quite dangerous and it is advisableto consult with a chemist when dealing with new etchants that may be potentially dangerous. ASTM E 2014, Standard Guide on MetallographicLaboratory Safety, is a valuable reference.

Etching

70


REFRACTORY METALS: Ti, Zr, Hf, Cr, Mo, Re, Nb, Ta, W and VComposition Comments

15. 100 mL water Kroll’s reagent for Ti alloys. Swab specimen 3-10 seconds or immerse1-3 mL HF specimen 10-30 seconds.2-6 mL HNO3

16. 200 mL water For Ti, Zr and alloys. Swab or immerse specimen. Higher concentrations can1 mL HF be used but are prone to staining problems.

17. 30 mL lactic acid For Ti alloys. Swab specimen up to 30 seconds. Decomposees, do not store.15 mL HNO3 Good for alpha-beta alloys.30 mL HF

18. 30 mL HCI For Zr, Hf, and alloys. Swab specimen 3-10 seconds, or immerse specimen15 mL HNO3 up to 120 seconds.30 mL HF

19. 45 mL H2O (H2O2 or glycerol) Cain’s chemical polish and etch for Hf, Zr, and alloys. Can dilute aqueous45 mL HNO3 solution with 3-5 parts water to stain the structure (swab specimen) after8-10 mL HF chemical polishing. Chemically polish and etch specimen by swabbing 5-20

seconds. Use polarized light.

20. 60 mL HCI Aqua regia. For Cr and alloys. Immerse or swab specimen up to 1 minute.20 mL HNO3 Use under a hood with care, do not store.

21. 30 mL HCI Modified “Glyceregia”. For Cr and alloys. Immerse specimen up to a few45 mL glycerol minutes.15 mL HNO3

22. 100 mL water Murakami’s reagent. For Cr, Mo, Re, Ta-Mo, W, V and alloys. Use fresh, can10 g KOH or NaOH immerse sample for up to 1 minute.10 g K3Fe(CN)6

23. 70 mL water For Mo alloys. Immerse specimen 2 minutes. Wash with water and dry;20 mL H2O2 (30%) immersion produces colors, swabbing produces grain-boundary etch.10 mL H2SO4

24. 10-20 mL glycerol For Mo and Mo-Ti alloys. Immerse specimen for up to 5 minutes.10 mL HNO3

10 mL HF

25. 100 mL water For Mo-Re alloys. Use at 20°C by immersion.5 g K3Fe(CN)6

2 g KOH

26. 50 mL acetic acid For Nb, Ta and alloys. Swab specimen 10-30 seconds.20 mL HNO3

5 mL HF

27. 50 mL water DuPont Nb reagent. For Nb-Hf and Nb alloys.14 mL H2SO4

5 mL HNO3

28. 50 mL water For Nb-Zr and Nb-Zr-Re alloys. Swab specimen.50 mL HNO3

1 mL HF

29. 30 mL lactic acid For Re and W-Re alloys. Swab specimen.10 mL HNO3

5 mL HF

30. 10 mL HF For V and alloys; grain-boundary etch for Ta alloys. Swab specimen. Equal10 mL HNO3 parts used for Ta and high Ta alloys.10-30 mL glycerol

IRON and STEELComposition Comments

31. 90-99 mL methanol or ethanol Nital. Most common etchant for Fe, carbon and alloy steels, cast iron.1-10 mL HNO3 Reveals alpha grain boundaries and constituents. Excellent for martensitic

structures. The 2% solution is most common, 5-10% used for high alloysteels (do not store). Use by immersion or swabbing of sample for up toabout 60 seconds.

Etching

71

32. 100 mL ethanol Picral. Recommended for structures consisting of ferrite and carbide. Does4 g picric acid not reveal ferrite grain boundaries. Addition of about 0.5-1% zephiran

chloride improves etch rate and uniformity.

33. 100 mL ethanol Vilella’s reagent. Good for ferrite-carbide structures. Produces grain contrast5 mL HCI for estimating prior austenite grain size. Results best on martensite tempered1 g picric acid at 572-932° (300-500°C). Occasionally reveals prior-austenite grain

boundaries in high alloy steels. Outlines constituents in stainless steels.Good for tool steels and martensitic stainless steels.

34. Saturated aqueous picric acid Bechet and Beaujard’s etch, most successful etchant for prior-austenite grainsolution plus small amount of a boundaries. Good for martensitic and bainitic steels. Many wetting agentswetting agent have been used, sodium tridecylbenzene sulfonate is one of most successful

(the dodecyl version is easier to obtain and works as well). Use at 20-100°C.Swab or immerse sample for 2-60 minutes. Etch in ultrasonic cleaner (see ref.2, pg. 219-223). Additions of 0.5g CuCl2 per 100mL solution or about 1% HCIhave been used for higher alloy steels to produce etching. Room temperatureetching most common. Lightly back polish to remove surface smut.

35. 150 mL water Modified Fry’s reagent. Used for 18% Ni maraging steels, martensitic and PH50 mL HCI stainless steels.25 mL HNO3

1 g CuCl2

36. 100 mL water Alkaline sodium picrate. Best etch for McQuaid-Ehn carburized samples.25 g NaOH Darkens cementite. Use boiling for 1-15 minutes or electrolytic at 6 V dc, 0.52 g picric acid A/in2, 30-120 seconds. May reveal prior-austenite grain boundaries in high

carbon steels when no apparent grain boundary film is present.

37. 3 parts HCI “Glyceregia”. For austenitic stainless steels. Reveals grain structure, outlines2 parts glycerol sigma and carbides. Mix fresh, do not store. Use by swabbing.1 part HNO3

38. 100 mL ethanol Kalling’s no. 2 (“waterless” Kalling’s) etch for austenitic and duplex stainless100 mL HCI steels. Ferrite attacked readily, carbides unattacked, austenite slightly5 g CuCl2 attacked. Use at 20°C by immersion or swabbing. Can be stored.

39. 15 mL HCI Acetic glyceregia. Mix fresh; do not store. Use for high alloy stainless steels.10 mL acetic acid5 mL HNO3

2 drops glycerol

40. 100 mL water Murakami’s reagent. Usually works better on ferritic stainless grades than10 g K2Fe(CN)6 on austenitic grades. Use at 20°C for 7-60 seconds: reveals carbides sigma10 g KOH or NaOH faintly attacked with etching up to 3 minutes. Use at 80°C (176°F) to

boiling for 2-60 minutes: carbides dark, sigma blue (not always attacked),ferrite yellow to yellow-brown, austenite unattacked. Do not always getuniform etching.

41. 100 mL water Use for stainless steels at 6 V dc, 25-mm spacing. Carbides revealed by10 g oxalic acid etching for 15-30 seconds, grain boundaries after 45-60 seconds, sigma

outlined after 6 seconds. 1-3 V also used. Dissolves carbides, sigma stronglyattacked, austenite moderately attacked, ferrite unattacked.

42. 100 mL water Used to color ferrite in martensitic, PH or dual-phase stainless steels. Use at20 g NaOH 3-5 V dc, 20°C, 5 seconds, stainless steel cathode. Ferrite outlined and

colored tan.

43. 40 mL water Electrolytic etch to reveal austenite boundaries but not twin boundaries in60 mL HNO3 austenitic stainless steels (304, 316, etc.). Voltage is critical. Pt cathode

preferred to stainless steel. Use at 1.4 V dc, 2 minutes (see ref. 2, pgs. 235,238 and 239).

COPPER, NICKEL and COBALT: Copper and AlloysComposition Comments

44. 25 mL NH4OH General purpose grain contrast etch for Cu and alloys (produces a flat etch25 mL water (optional) for some alloys). Use fresh, add peroxide last. Use under a hood. Swab25-50 mL H2O2 (3%) specimen 5-45 seconds.

45. 100 mL water General purpose etch for Cu and alloys. Immerse or swab for 3-60 seconds.10 g ammonium persulfate Reveals grain boundaries but is sensitive to crystallographic orientation.


Etching

72


46. 100 mL water General purpose etch for Cu and alloys, particularly Cu-Be alloys.3 g ammonium persulfate1 mL NH4OH

47. 70 mL water Excellent general purpose etch, reveals grain boundaries well. Immerse5 g Fe(NO3)3 specimen 10-30 seconds.25 mL HCI

Nickel and AlloysComposition Comments

48. 5 g FeCl3 Carapella’s etch for Ni and Ni-Cu (Monel) alloys. Use by immersion2 mL HCI or swabbing.99 mL methanol

49. 40-80 mL ethanol Kalling’s no.2 etch (“waterless” Kalling’s) for Ni-Cu alloys and superalloys.40 mL HCI Immerse or swab specimen up to a few minutes.2 g CuCI2

50. 50 mL water Marble’s reagent for Ni, Ni-Cu, and Ni-Fe alloys and superalloys. Immerse or50 mL HCI swab sample 5-60 seconds. Reveals grain structure of superalloys.10 g CuSO4

51. 15 mL HCI “Glyceregia”, for superalloys and Ni-Cr alloys. Swab specimen for 5-6010 mL glycerol seconds. Mix fresh. Do not store. Use under a hood.5 mL HNO3

52. 60 mL glycerol Modified Glyceregia for superalloys. Reveals precipitates. Use under hood;50 mL HCI do not store. Add HNO3 last. Discard when dark yellow. Immerse or swab10 mL HNO3 specimen 10-60 seconds.

Cobalt and AlloysComposition Comments

53. 60 mL HCI For Co and alloys. Mix fresh and age 1 hour before use. Immerse specimen15 mL water for up to 30 seconds. Do not store.15 mL acetic acid15 mL HNO3

54. 200 mL methanol General etch for Co and alloys. Immerse specimen 2-4 minutes.7.5 mL HF2.5 mL HNO3

55. 50 mL water Marble’s reagent, for Co high temperature alloys. Immerse or swab specimen50 mL HCI for up to 1 minute.10 g CuSO4

56. 80 mL lactic acid For Co alloys. Use by swabbing.10 mL H2O2 (30%)10 mL HNO3

PRECIOUS METALS: Au, Ag, Ir, Os, Pd, Pt, Rh and RuComposition Comments

57. 60 mL HCI For gold, silver, palladium and high noble metal alloys. Use under hood.40 mL HNO3 Immerse specimen up to 60 seconds. Equal parts of each acid also used.

58. 60 mL HCI Aqua regia for pure gold, Pt and alloys, some Rh alloys. Use boiling for up to20 mL HNO3 30 minutes.

59. 1-5 g CrO3 For Au, Ag, Pd and alloys. Swab or immerse specimen up to 60 seconds.100 mL HCI

60. 30 mL water For pure Pt. Use hot, immerse specimen up to 5 minutes.25 mL HCI5 mL HNO3

61. Conc. HCI For Rh and alloys. Use at 5 V ac, 1-2 minutes, graphite cathode,Pt lead wires.

Etching

73


62. Solution a For Ru. Mix 4 parts solution a to 1 part solution b, use 5-20 V ac, 1-2 100 mL water minutes, graphite cathode, Pt-lead wires. 40 g NaCISolution b Conc. HCI

63. 50 mL NH4OH For pure Ag, Ag solders, and Ag-Pd alloys. Mix fresh. Swab specimen up to20 mL H2O2 (3%) 60 seconds. Discard after etching. Use a 50:50 mix for sterling silver; for fine

silver, use 30% conc. hydrogen peroxide.

64. Solution a For gold alloys up to 18 karat. Mix equal amounts of a and b directly on 10 g NaCN the specimen using eye droppers. Swab and replenish the etchants 100 mL water until the desired etch level is obtained. If a brown stain forms, swab withSolution b a to remove it. H2O2 (30%)

SINTERED CARBIDESComposition Comments

65. 100 mL water Murakami’s reagent, for WC-Co and complex sintered carbides. Immerse10 g KOH or NaOH specimen seconds to minutes. Use 2-10 seconds to identify eta phase10 g K3Fe(CN)6 (colored). Longer times attack eta. Reveals phase and grain boundaries.

Normally used at 20°C.

66. 97 mL water For WC, MO2C, TiC, or Ni in sintered carbides. Use boiling for up to 603 mL H2O2 (30%) seconds. For coarse carbides or high Co content, use short etch time.

67. 15 mL water For WC, TiC, TaC, and Co in sintered carbides. Use at 20°C for 5-30 seconds.30 mL HCI15 mL HNO3

15 mL acetic acid

CERAMICS and NITRIDESComposition Comments

68. Phosphoric acid For alumina, Al2O3, and silicon nitride, Si3N4. Use by immersion at 250°C forup to a few minutes for Al2O3 or up to 15 minutes for Si3N4.

69. 100 mL water For magnesia, MgO. Use by immersion at 25-60°C for several minutes.15 mL HNO3

70. 100 mL water For titania, TiO2. Immerse up to a few minutes.5 g NH4 FHF4 mL HCI

71. 50 mL water For zirconia, ZrO2. Use by immersion in boiling solution for up to 5 minutes.50 mL H2SO4

72. HCI For calcia, CaO, or MgO. Immerse for up to 6 minutes.

73. HF For Si3N4, BeO, BaO, MgO, ZrO2 and Zr2O3. Immerse for up to 6 minutes.

PLASTICS and POLYMERSComposition Comments

74. 100 mL water For polypropylene (PP). Immerse for up to several hours at 70°C.60 g CrO3

75. HNO3 For polyethylene (PE). Immerse for up to a few minutes.

76. Xylol Reveals spherolites in polyethene (PE). Immerse for up to a few days at70°C.

77. 70 mL water For polyoxymethylene (POM). Immerse up to 20 seconds.30 mL HCI

78. Xylene For polyamid (PA) and polyethylene (PE). Use at 70°C for 60 seconds.For Nylon 6, use at 65-70°C for 2-3 minutes. For Nylon 6, use at 75°C for3-4 minutes.

Etching

74

LIGHT OPTICAL MICROSCOPY

Metallurgical microscopes differ from biologicalmicroscopes primarily in the manner by whichthe specimen is illuminated due to the opacity ofmetals. Unlike biological microscopes, metallur-gical microscopes must use reflected light, ratherthan transmitted light. Figure 39 is a simplifiedlight ray diagram of a metallurgical microscope.The prepared specimen is placed on the stagewith the surface perpendicular to the optical axisof the microscope and is illuminated throughthe objective lens by light from a lamp or arcsource. This light is focused by the condenserlens into a beam that is made approximatelyparallel to the optical axis of the microscope bythe half silvered mirror. The light then passesthrough the objective onto the specimen. It isthen reflected from the surface of the specimen,back through the objective, the half silvered mir-ror, and then to the eyepiece to the observer’seye, or to a camera port or a film plane.

After the specimen has been properly sectionedand polished, it is ready for examination. How-ever, in many cases, it is important to begin theexamination with an as-polished specimen,particularly for quality control or failure analysiswork. Certain microstructural constituents, such

as nonmetallic inclusions, graphite, nitrides,cracks or porosity, are best evaluated in theas-polished condition as etching reveals othermicrostructural features that may obscure suchdetails. In the as-polished condition, there is aminimum of extra information for the observer todeal with which makes examination of these fea-tures most efficient. Some materials do notrequire etching and, in a few cases, examina-tion is less satisfactory after etching.

For most metals, etching must be used to fullyreveal the microstructure. A wide range ofprocesses are used, the most common beingetching with either acidic or basic solutions. Formost metals and alloys, there are a number ofgeneral-purpose etchants that should be usedfirst to reveal the microstructure. Subsequentexamination may indicate that other morespecialized techniques for revealing the micro-structure may be useful. For example, theexamination may reveal certain constituents thatthe metallographer may wish to measure.Measurements, particularly those performed withautomatic image analyzers, will be simpler toperform and more reliable if the desired constitu-ent can be revealed selectively, and numerousprocedures are available to produce selectivephase contrasting.

Light Optical Microscopy

Figure 39. Schematic diagram showing the light path through an upright reflected light microscope operating withbright field illumination.

75

The Light MicroscopeWhen Henry Clifton Sorby made his first exami-nation of the microstructure of iron on 28 July1863, he was using a transmitted light petro-graphic microscope made by Smith, Beck andBeck of London that he had purchased in 1861.Although capable of magnifications up to 400X,most of his examinations were conducted at 30,60 or 100X and his first micrographs wereproduced at only 9X. The objective lenses of thismicroscope were equipped with Lieberkühnssilvered concave reflectors for focusing light onopaque specimens.

Sorby quickly realized that reflected lightproduced with the Lieberkühns reflectors wasinadequate and he developed two alternatemethods for this purpose. Subsequently, othersdeveloped vertical illuminators using prisms orplane glass reflectors and Sorby’s systems werenot further developed. In 1886, Sorby reportedon the use of “very high power” (650X) for thestudy of pearlite. This was accomplished using a45º cover glass vertical illuminator made for himby Beck.

For many years, photomicroscopy was con-ducted using specially built reflected microscopesknown as metallographs. These devices repre-sented the “top-of-the-line” in metallurgicalmicroscopes and were essential for quality work.In the late 1960’s and early 1970’s, manufactur-ers developed metallographs that were easier touse from the eyepiece viewing position. Thetemperamental carbon arc light sources werereplaced by xenon arc lamps. The unwieldybellows systems for altering magnification werereplaced by zoom systems. Vertical illuminators,previously using single position objectives, wereequipped with four to six position rotatable nose-piece turrets to minimize objective handling. Thelight path was deflected so that the film planewas conveniently near at hand. Universal typevertical illuminators and objective lenses wereintroduced so that the illumination mode couldbe readily switched from bright field to darkfield, polarized light or differential interferencecontrast. Such systems were very attractive inthat separate vertical illuminators and objectiveswere no longer needed for each illuminationmode and handling was eliminated. Exposuremeters were also added at this time, chiefly as a

result of the rising popularity of instant filmswhere such devices are needed to minimize filmwastage. But, by the late 1970s, these largemetallographs had become very expensive, tooexpensive for most laboratories.

In 1979, microscope manufacturers began tointroduce very high quality, reasonably priced,compact metallographs. These microscopescan be obtained with a wide variety of objectivelens types and auxiliary accessories to meet themetallographer’s needs. They are available withuniversal vertical illuminators that permit easyswitching from one illumination mode to anotherusing the same set of objective lenses. Further-more, the manufacturers have introduced newglass compositions and lens formulations,generally by computer-aided design, for im-proved image contrast and brightness. Tungsten-halogen filament lamps have largely replacedxenon arc lamps as the preferred light source.

Microscope ComponentsLight Sources The amount of light lost duringpassage from the source through a reflecting typemicroscope is appreciable because of the intri-cate path the light follows. For this reason, it isgenerally preferable that the intensity of thesource be high, especially for photomicroscopy.Several types of light sources are used includingtungsten-filament lamps, tungsten-halogenlamps, quartz-halogen lamps, and xenon arcbulbs. Tungsten-filament lamps generally oper-ate at low voltage and high current. They arewidely used for visual examination because oftheir low cost and ease of operation.

Tungsten-halogen lamps are the most popularlight source today due to their high light inten-sity. They produce good color micrographs whentungsten-corrected films are employed. Lightintensity can be varied easily to suit the viewingconditions by adjusting a rheostat, a distinctadvantage over arc lamps.

Xenon arc lamps produce extremely high inten-sity light, and their uniform spectra and daylightcolor temperature makes them suitable for colorphotomicrography. The first xenon lamps pro-duced ozone, but modern units have overcomethis problem. Light output is constant and canonly be reduced using neutral density filters.


76

Condenser. An adjustable lens free of spheri-cal aberration and coma is placed in front of thelight source to focus the light at the desired pointin the optical path. A field diaphragm is placedin front of this lens to minimize internal glareand reflections within the microscope. The fielddiaphragm is stopped down to the edge of thefield of view.

A second adjustable iris diaphragm, the aper-ture diaphragm, is placed in the light path beforethe vertical illuminator. Opening or closing thisdiaphragm alters the amount of light and theangle of the cone of light entering the objectivelens. The optimum setting for the aperturevaries with each objective lens and is a compro-mise among image contrast, sharpness, anddepth of field. As magnification increases, theaperture diaphragm is stopped down. Openingthis aperture increases image sharpness, butreduces contrast; closing the aperture increasescontrast, but impairs image sharpness. Theaperture diaphragm should not be used forreducing light intensity. It should be adjusted onlyfor contrast and sharpness.

Filters. Light filters are used to modify the lightfor ease of observation, for improved photo-microscopy, or to alter contrast. Neutral densityfilters are used to reduce the light intensity uni-formly across the visible spectrum. Variousneutral density filters, with a range of approxi-mately 85 to 0.01% transmittance, are available.They are a necessity when using an arc-lampbut virtually unnecessary when using a filamentlamp.

Selective filters are used to balance the colortemperature of the light source to that of the film.They maybe needed for faithful reproduction ofcolor images, depending on the light sourceused and the film type. A green or yellow-greenfilter is widely used in black and white photog-raphy to reduce the effect of lens defects onimage quality. Most objectives, particularly thelower cost achromats, require such filtering forbest results.

Polarizing filters are used to produce planepolarized light (one filter) or crossed-polarizedlight (two filters rotated 90° to each other to pro-duce extinction) for examinations of anisotropicnoncubic (crystallographic) materials.

Objectives. The objective lens forms the primaryimage of the microstructure and is the mostimportant component of the optical microscope.The objective lens collects as much light aspossible from the specimen and combines thislight to produce the image. The numerical aper-ture (NA) of the objective, a measure of thelight-collecting ability of the lens, is defined as:

NA = n sin α

where n is the minimum refraction index of thematerial (air or oil) between the specimen andthe lens, and α is the half-angle of the mostoblique light rays that enter the front lens of theobjective. Light-collecting ability increases withα. The setting of the aperture diaphragm willalter the NA of the condenser and therefore theNA of the system.

The most commonly used objective is theachromat, which is corrected spherically for onecolor (usually yellow green) and for longitudinalchromatic aberration for two colors (usually redand green). Therefore, achromats are not suit-able for color photomicroscopy, particularly athigher magnifications. Use of a yellow-greenfilter yields optimum results. However, achromatsdo provide a relatively long working distance, thatis, the distance from the front lens of the objec-tive to the specimen surface when in focus. Theworking distance decreases as the objectivemagnification increases. Most manufacturersmake long-working distance objectives for spe-cial applications, for example, in hot stagemicroscopy. Achromats are usually strain free,which is important for polarized light examina-tion. Because they contain fewer lenses thanother more highly corrected lenses, internalreflection losses are minimized.

Semiapochromatic or fluorite objectives providea higher degree of correction of spherical andchromatic aberration. Therefore, they producehigher quality color images than achromats.The apochromatic objectives have the highestdegree of correction, produce the best results,and are more expensive. Plano objectives haveextensive correction for flatness of field, whichreduces eyestrain, and are usually found onmodern microscopes.


77

With parfocal lens systems, each objective onthe nosepiece turret will be nearly in focus whenthe turret is rotated, preventing the objective frontlens from striking the specimen when lenses areswitched. Many objectives also are spring loaded,which helps prevent damage to the lens. This ismore of a problem with high magnificationobjectives, because the working distance can bevery small.

Certain objectives are designed for use with oilbetween the specimen and the front lens of theobjective. However, oil-immersion lenses arerarely used in metallography, because thespecimen and lens must be cleaned after use.However, they do provide higher resolution thancan be achieved when air is between the lensand specimen. In the latter case, the maximumpossible NA is 0.95; oil-immersion lensesproduce a 1.3 to 1.45 NA, depending on thelens and the oil-immersion used. Objective mag-nifications from about 25 to 200X are available,depending upon the manufacturer. Use of oilalso improves image contrast, which is valuablewhen examining low reflectivity specimens, suchas coal or ceramics.

Eyepieces. The eyepiece, or ocular, magnifiesthe primary image produced by the objective; theeye can then use the full resolution capability ofthe objective. The microscope produces avirtual image of the specimen at the point of mostdistinct vision, generally 250 mm (10 inch) fromthe eye. The eyepiece magnifies this image,permitting achievement of useful magnifications.The standard eyepiece has a 24 mm diameterfield of view; wide field eyepieces for plano-type objectives have a 30 mm diameter field ofview, which increases the usable area of theprimary image. Today, the wide field plano-objective is standard on nearly all metallurgicalmicroscopes.

The simplest eyepiece is the Huygenian, whichis satisfactory for use with low and medium powerachromat objectives. Compensating eyepiecesare used with high NA achromats and the morehighly corrected objectives. Because some lenscorrections are performed using these eyepieces,the eyepiece must be matched with the type ofobjective used. The newer, infinity-correctedmicroscopes do not perform corrections in theeyepieces, but in the tube lens. Eyepieces,

therefore, are simpler in a infinity-correctedmicroscopes.

Eye clearance is the distance between the eyelens of the ocular and the eye. For most eye-pieces, the eye clearance is 10 mm or less –inadequate if the microscopist wears glasses.Simple vision problems, such as near-sightedness, can be accommodated using thefine focus adjustment. The microscope cannotcorrect vision problems such as astigmatism, andglasses must be worn. High eyepoint eyepiecesare available to provide an eye clearance ofapproximately 20 mm, necessary for eyeglasses.

Eyepieces are commonly equipped with vari-ous reticles or graticules for locating, measuring,counting or comparing microstructures. The eye-piece enlarges the reticle or graticule image andthe primary image. Both images must be infocus simultaneously. Special eyepieces are alsoproduced to permit more accurate measure-ments than can be made with a graticule scale.Examples are the filar-micrometer ocular orscrew-micrometer ocular. Such devices can beautomated to produce a direct digital readout ofthe measurement, which is accurate to approxi-mately 1-µm.

A 10X magnification eyepiece is usually used;to obtain standard magnifications, some systemsrequire other magnifications, such as 6.3X.Higher power eyepieces, such as 12, 15, 20, or25X, are also useful in certain situations. Theoverall magnification is found by multiplying theobjective magnification, Mo, by the eyepiecemagnification, Me. If a zoom system or bellowsis also used, the magnification should be alteredaccordingly.

Stage. A mechanical stage is provided for focus-ing and moving the specimen, which is placedon the stage and secured using clips. The stageof an inverted microscope has replaceablecenter stage plates with different size holes. Thepolished surface is placed over the hole forviewing. However, the entire surface cannot beviewed, unless the specimen is smaller than thehole and it is mounted. At high magnificationsit may not be possible to focus the objectivenear the edge of the hole due to the restrictedworking distance.


78

Using the upright microscope, the specimen isplaced on a slide on the stage. Because thepolished surface must be perpendicular to thelight beam, clay is placed between the specimenbottom and the slide. A piece of lens tissue isplaced over the polished surface, and the speci-men is pressed into the clay using a levelingpress. However, pieces of tissue may adhere tothe specimen surface. An alternative, particularlyuseful with mounted specimens, is to use a ringinstead of tissue to flatten the specimen. Alumi-num or stainless steel ring forms of the samesize as the mounts (flattened slightly in a vise)will seat on the mount rather than the specimen.

The upright microscope allows viewing of theentire surface with any objective, and theoperator can see which section of the specimenis being viewed – a useful feature when examin-ing specific areas on coated specimens, welds,and other specimens where specific areas areto be examined. For mounted specimens, anautoleveling stage holder can eliminate levelingspecimens with clay.

The stage must be rigid to eliminate vibrations.Stage movement, controlled by x and y microme-ters, must be smooth and precise; rack and piniongearing is normally used. Many stages havescales for measuring distances in the x and ydirections. The focusing controls oftencontain rulings for estimating vertical movement.Some units have motorized stages and focuscontrols.

A circular, rotatable stage plate may facilitatepolarized light examination. Such stages, com-mon for mineralogical or petrographic studies,are graduated to permit measuring the angle ofrotation. A rectilinear stage is generally placedon top of the circular stage.

Stand. Bench microscopes require a rigid stand,particularly if photomicroscopy is performed onthe unit. The various pieces of the microscopeare attached to the stand when assembled.In some cases, the bench microscope is placedon a separate stand that also holds the photo-graphic system.

ResolutionTo see microstructural detail, the optical systemmust produce adequate resolution, or resolvingpower, and adequate image contrast. If resolu-tion is acceptable but contrast is lacking, detailcannot be observed. In general, the ability toresolve two points or lines separated by distanced is a function of the wavelength, λ , of theincident light and the numerical aperture, NA, ofthat objective.

where k is 0.5 or 0.61. Figure 40 illustrates thisrelationship for k = 0.61 and four light wave-lengths. Other formulas have also been reported.This equation does not include other factors thatinfluence resolution, such as the degree ofcorrection of the objectives and the visual acuityof the microscopist. It was based on the work ofAbbe under conditions not present in metallog-raphy, such as self-luminous points, perfectblack-white contrast, transmitted light examina-tion, an ideal point-light source, and absence oflens defects.

Using the above equation, the limit of resolutionfor an objective with an NA of 1.3 is approxi-mately 0.2-µm. To see lines or points spaced0.2-µm apart, the required magnification is de-termined by dividing by the resolving power ofthe human eye, which is difficult to determineunder observation conditions. Abbe used a valueof 0.3 mm at a distance of 250 mm – the dis-tance from the eye for optimum vision. This gives1500x. For light with a mean wavelength of 0.55-µm, the required magnification is 1100 timesthe NA of the objective. This is the origin of the1000·NA rule for the maximum useful magnifi-cation. Any magnification above 1000·NA istermed “empty”, or useless.

Strict adherence to the 1000·NA rule should bequestioned, considering the conditions underwhich it was developed, which are certainly fardifferent from those encountered in metallogra-phy. According to the Abbe analysis, for amicroscopist with optimum 20/20 vision and foroptimum contrast conditions and a mean lightwavelength of 550 nm, the lowest magnificationthat takes full advantage of the NA of the objec-

kλ

NAd =


79

that depth of field increases as the NA decreasesand when longer wavelength light is used, asshown in Figure 41.

Imaging ModesMost microscopical studies of metals are madeusing bright field illumination. In addition to thistype of illumination, several special techniques(oblique illumination, dark field illumination,differential interference contrast microscopy andpolarized light microscopy) have particularapplications for metallographic studies.

Nearly all microscopes using reflected or trans-mitted light employ Köhler illumination becauseit provides the most intense, most even illumina-tion possible with standard light sources. Thereflected light microscope has two adjustablediaphragms: the aperture diaphragm and the fielddiaphragm, located between the lamp housingand the objective. Both are adjusted to improveillumination and the image. To obtain Köhler il-lumination, the image of the field diaphragm mustbe brought into focus on the specimen plane.This normally occurs automatically when the mi-

tive is 500 times the NA. This establishes a use-ful minimum magnification to use with a givenobjective. It has been suggested that the upperlimit of useful magnification for the averagemicroscopist is 2200·NA, not 1000·NA.

Depth of FieldDepth of field is the distance along the opticalaxis over which image details are observed withacceptable clarity. Those factors that influenceresolution also affect depth of field, but in theopposite direction. Therefore, a compromisemust be reached between these two parameters,which is more difficult as magnification increases.This is one reason light etching is preferred forhigh-magnification examination. The depth offield, df, can be estimated from:

where n is the refractive index of the mediumbetween the specimen and the objective (n ~1.0 for air), λ is the wavelength of light, and NAis the numerical aperture. This equation shows

Figure 40. Influence of objective numerical aperture and light wavelength on the resolution of the lightmicroscope.

λ (n2 −NA2)½

NA2df =


RESOLUTION CHART

Sodium (589 nm)

Green (546 nm)

Blue (436 nm)

Ultraviolet(365 nm)

80

crostructural image is brought into focus. Thefilament image must also be focused onthe aperture diaphragm plane. This producesuniform illumination of the specimen imaged atthe intermediate image plane and magnified bythe eyepiece.

Bright Field. In bright field illumination, the sur-face of the specimen is normal to the optical axisof the microscope, and white light is used. A raydiagram for bright-field illumination is illustratedin Figure 42. Light that passes through theobjective and strikes a region of the specimensurface that is perpendicular to the beamwill be reflected back up the objective throughthe eyepieces to the eyes where it will appear tobe bright or white. Light that strikes grainboundaries, phase boundaries, and otherfeatures not perpendicular to the optical axis willbe scattered at an angle and will not be collectedby the objective. These regions will appear tobe dark or black in the image. Bright field isthe most common mode of illumination usedby metallographers.

Oblique Illumination. The surface relief of ametallographic specimen can be revealed us-ing oblique illumination. This involves offsettingthe condenser lens system or, as is more usu-ally done, moving the condenser aperture to aposition slightly off the optical axis. Although itshould be possible to continually increase thecontrast achieved by oblique illumination bymoving the condenser farther and farther fromthe optical axis, the numerical aperture of a lensis reduced when this happens because only a

Figure 41. Influence of the objective numerical aperture and light wavelength on the depth of field of thelight microscope.

Sodium (589 nm)

Green (546 nm)

Blue (436 nm)

Index of Refraction = 1.0

DEPTH OF FIELD CHART


Figure 42. Schematic diagram showing the light pathin bright field illumination.

81

ter) before the eyepiece, as illustrated in Figure44. The polarizer produces plane polarized lightthat strikes the surface and is reflected throughthe analyzer to the eyepieces. If an anisotropicmetal is examined with the analyzer set 90° tothe polarizer, the grain structure will be visible.However, viewing of an isotropic metal (cubicmetals) under such conditions will produce adark, “extinguished” condition (complete dark-ness is not possible using Polaroid filters).Polarized light is particularly useful in metallog-raphy for revealing grain structure and twinningin anisotropic metals and alloys (see the ap-pendix for crystal structure information) and foridentifying anisotropic phases and inclusions.

Differential Interference Contrast (DIC). Whencrossed polarized light is used along with adouble quartz prism (Wollaston prism) placedbetween the objective and the verticalilluminator, Figure 45, two light beams are pro-duced which exhibit coherent interference in theimage plane. This leads to two slightly displaced(laterally) images differing in phase (λ/2) thatproduces height contrast. The image producedreveals topographic detail somewhat similar tothat produced by oblique illumination but with-out the loss of resolution. Images can be viewedwith natural colors similar to those observed inbright field, or artificial coloring can be intro-duced by adding a sensitive tint plate.

As an example of the use of these differentimaging modes, Figure 46 shows the micro-

portion of the lens is used. For this reason, thereis a practical limit to the amount of contrast thatcan be achieved. Illumination also becomesuneven as the degree of “obliqueness” in-creases. Since differential interference contrastsystems have been available, oblique illumina-tion is rarely offered as an option on newmicroscopes.

Dark Field. Another method that can be used todistinguish features not in the plane of thepolished and etched surface of a metallographicspecimen is dark field (also called dark ground)illumination. This type of illumination (seeFigure 43 for a ray diagram) gives contrastcompletely reversed from that obtained withbright-field illumination — the features that arelight in bright field will be dark in dark field, andthose that are dark in bright field will belight, appearing to be self luminous in dark field.This highlighting of angled surfaces (pits, cracks,or etched grain boundaries) allows morepositive identification of their nature than can bederived from a black image with bright fieldillumination. Due to the high image contrastobtained and the brightness associated withfeatures at an angle to the optical axis, it is oftenpossible to see details not observed with brightfield illumination.

Polarized Light. Because many metals and me-tallic and nonmetallic phases are opticallyanisotropic, polarized light is particularly usefulin metallography. Polarized light is obtained byplacing a polarizer (usually a Polaroid filter) infront of the condenser lens of the microscopeand placing an analyzer (another Polaroid fil-


Figure 44. Schematic diagram showing the lightpath in polarized light with an optional lambdaplate (sensitive tint plate).

Figure 43. Schematic diagram showing the lightpath in dark field illumination.

82

structure of an aluminum bronze specimen thatwas water quenched from the beta field formingmartensite. The specimen was prepared but notetched. A minor amount of relief was introducedin final polishing. Figure 46a shows the surfacein bright field illumination. Due to the reliefpresent, a faint image of the structure is visible.Dark field, Figure 46c, and differential interfer-


ence contrast, Figure 46d, make use of this re-lief to reveal much more detail than in brightfield. However, as the martensite is noncubic incrystal structure, it responds well to crossed-polarized light, Figure 46b. Of course, not everymaterial can be examined using all fourillumination modes, but there are many caseswhere two or more modes can be effectivelyutilized to reveal more information than givenby bright field alone.

Figure 46. Martensitic microstructure of heat treated,eutectoid aluminum bronze (Cu–11.8% Al) in theunetched condition viewed using: a) bright fieldillumination, b) polarized light

Figure 46. Martensitic microstructure of heattreated, eutectoid aluminum bronze (Cu–11.8% Al) inthe unetched condition viewed using: c) dark fieldillumination; and, d) differential interferencecontrast illumination (200X).

a

b

c

d

HELPFUL HINTS FOR LIGHT

OPTICAL MICROSCOPY

• It can be difficult to determine whichobjective you are using with an invertedmicroscope due to the limited view fromabove, as the magnification value maynot be easily observed. However, eachobjective has a color-coded ring on it thatis usually visible. The colors and corre-sponding magnifications are: red – 5X,yellow – 10X, green – 20X, blue – 50X,and white – 100X.

Figure 45. Schematic diagram showing the light pathfor differential interference contrast (DIC)illumination.

83

MICROINDENTATIONHARDNESS TESTING

Microindentation hardness testing, more com-monly (but incorrectly) called microhardnesstesting, is widely used to study fine scale changesin hardness, either intentional or accidental. Heattreaters have utilized the technique for manyyears to evaluate the success of surface hard-ening treatments or to detect and assessdecarburization. Metallographers and failureanalysts use the method for a host of purposesincluding evaluation of homogeneity, character-ization of weldments, as an aid to phaseidentification, or simply to determine the hard-ness of specimens too small for traditional bulkindentation tests.

Metallurgists and metallographers tend todevelop their own jargon, often as a matter oflinguistic simplicity, which is not always as rigor-ously correct as it could be. Although the term“microhardness” is generally understood by itsusers, the word implies that the hardness isextremely low, which is not the case. The ap-plied load and the resulting indent size are smallrelative to bulk tests, but the same hardnessnumber is obtained. Consequently, ASTM Com-mittee E-4 on Metallography recommends useof the term “microindentation hardness testing”which could be given the acronym MHT. ASTMStandard E 384 fully describes the two most com-mon microindentation tests - the Vickers and theKnoop tests.

The Vickers TestIn 1925, Smith and Sandland of the UK devel-oped a new indentation test for metals that weretoo hard to evaluate using the Brinell test. Thehardened steel ball of the Brinell test limitedthe test to steels with hardnesses below ~450HBS (~48 HRC). (The harder tungsten carbideball was not available in 1925. The WC indenterextends the Brinell test to metals up to 615 HBW(~58 HRC). The WC ball has now replacedthe steel ball for the Brinell test.) In designingthe new indenter, a square-based diamondpyramid (see Figure 47), they chose a geometrythat would produce hardness numbers nearly

identical to Brinell numbers in the range whereboth tests could be used. This was a very wisedecision as it made the Vickers test very easyto adopt. The ideal d/D ratio (d = impressiondiameter, D = ball diameter) for a sphericalindenter is 0.375. If tangents are drawn to theball at the impression edges for d/D = 0.375,they meet below the center of the impressionat an angle of 136°, the angle chosen for theVickers indenter.

Use of diamond allowed the Vickers test to beused to evaluate any material (except diamond)and, furthermore, had the very important advan-tage of placing the hardness of all materials onone continuous scale. This is a major disadvan-tage of Rockwell type tests where different scales(15 standard and 15 superficial) were developedto evaluate materials. Not one of these scalescan cover the full hardness range. The HRA scalecovers the broadest hardness range, but it is notcommonly used.

In the Vickers test, the load is applied smoothly,without impact, forcing the indenter into the testpiece. The indenter is held in place for 10 or 15seconds. The physical quality of the indenter andthe accuracy of the applied load (defined in E384) must be controlled in order to get thecorrect results. After the load is removed, thetwo impression diagonals are measured, usu-ally to the nearest 0.1-µm with a filar micrometer,

Microindentation Hardness Testing

Figure 47. Schematic of the Vickers indenter and theshape of an impression.

136°

136°

84

and averaged. The Vickers hardness (HV) is cal-culated using:

where the load L is in gf and the average diago-nal d is in µm (this produces hardness numberunits of gf/µm2 although the equivalent units kgf/mm2 are preferred; in practice the numbers arereported without indication of the units). Theoriginal Vickers testers were developed for testloads of 1 to 120kgf that produce rather largeindents. Recognizing the need for lower testloads, the National Physical Laboratory (UK)reported on use of lower test loads in 1932. Lipsand Sack developed the first low-load Vickerstester in 1936.

Because the shape of the Vickers indentation isgeometrically similar at all test loads, the HVvalue is constant, within statistical precision, overa very wide test load range as long as the testspecimen is reasonably homogeneous. Numer-ous studies of microindentation hardness testresults conducted over a wide range of test loadshave shown that test results are not constant atvery low loads. This problem, called the “inden-tation size effect”, or ISE, has been attributed tofundamental characteristics of the material. How-ever, the same effect is observed at the low loadtest range (1-10kgf) of bulk Vickers testers [2]and an ASTM interlaboratory “round robin” ofindents made by one laboratory but measuredby twelve different people, reported all threepossible ISE responses [24,25] for the same in-dents!

Since the 1960s, the standard symbol forVickers hardness per ASTM E 92 and E 384,has been HV. This should be used in preferenceto the older, obsolete symbols DPN or VPN. Thehardness is expressed in a standard format. Forexample, if a 300 gf load is used and the testreveals a hardness of 375 HV, the hardness isexpressed as 375 HV300. Rigorous application ofthe SI system results in hardness units ex-pressed not in the standard, understandable kgf/mm2 values but in GPa units that are meaning-less to most engineers and technicians. ASTMrecommends a ‘soft” metric approach in thiscases.

In the Vickers test, it is assumed that elasticrecovery does not occur once the load isremoved. However, elastic recovery does occur,and sometimes its influence is quite pronounced.Generally, the impression (Figure 48) appearsto be square, and the two diagonals have similarlengths. As with the Brinell test, the Vickers hard-ness number is calculated based on the surfacearea of the indent rather than the projected area.If the impression shape is distorted due to elas-tic recovery (very common in anisotropicmaterials), Figure 49, should the hardness bebased on the average of the two diagonals? It ispossible to calculate the Vickers hardness basedon the projected area of the impression, whichcan be measured by image analysis. Whilerigorous studies of this problem are scant in theliterature, the diagonal measurement is thepreferred approach even for distorted indents,at this time.

The Knoop TestAs an alternative to the Vickers test, particularlyto test very thin layers, Frederick Knoop and hisassociates at the former National Bureau ofStandards developed a low-load test using arhombohedral-shaped diamond indenter, Fig-ure 50. The long diagonal is seven times (7.114actually) as long as the short diagonal. With thisindenter shape, elastic recovery can be held to a

1854.4L

d2HV =


Figure 48. (top) Example of a well-formed Vickersindentation (400X). Figure 49. (bottom) Example of adistorted Vickers indentation (400X).

85

minimum. Some investigators claim there is noelastic recovery with the Knoop indent, but thiscannot be true as measurements of the ratio oflong to short diagonal often reveal results sub-stantially different than the ideal 7.114 value.

The Knoop test is conducted in the same man-ner, and with the same tester, as the Vickers test.However, only the long diagonal is measured.This, of course, saves some time. The Knoophardness is calculated from

HK = 14229L d2

where the load L is in gf and the long diagonal dis in µm. Again, the symbol HK was adopted inthe early 1960’s while other terms; e.g., HKN orKHN, are obsolete and should not be used. TheKnoop hardness is expressed in the same man-ner as the Vickers hardness; i.e., 375 HK300 meansthat a 300 gf load produced a Knoop hardnessof 375. (The kgf/mm2 unit information is no longerreported).

Aside from a minor savings of time, one chiefmerit of the Knoop test is the ability to test thinlayers more easily. For surfaces with varyinghardness, such as case hardened parts, Knoopindents can be spaced closer together thanVickers indents. Thus, a single Knoop traversecan define a hardness gradient more simply thana series of two or three parallel Vickers traverseswhere each indent is made at different depths.

Furthermore, if the hardness varies stronglywith the depth, the Vickers indent will be dis-torted by this change; that is, the diagonalparallel to the hardness change will be affectedby the hardness gradient (i.e., there is a sub-stantial difference in the lengths of the two halvesof the diagonal), while the diagonal perpendicu-lar to the hardness gradient will be unaffected(both halves of this diagonal of the sameapproximate length).

The down side of the Knoop indent is that thethree dimensional indent shape will change withtest load and, consequently, HK varies withload. At high loads, this variation is not substan-tial. Conversion of HK values to other test scalescan only be done reliably for HK values performedat the standard load, generally 500gf, used todevelop the correlations. All hardness scaleconversions are based on empirical data.Conversions are not precise but are estimates.

Factors Affecting Accuracy, Precisionand BiasMany factors (see Table 44) can influence thequality of microindentation test results [26]. In theearly days of low-load (< 100gf) hardness test-ing, it was quickly recognized that improperspecimen preparation may influence hardnesstest results. Most text books state that improperpreparation will yield higher test results becausethe surface contains excessive preparationinduced deformation. While this is certainly true,there are other cases where improper prepara-tion can create excessive heat that will lower thehardness and strength of many metals andalloys. So either problem may be encountereddue to faulty preparation. For many years, it wasconsidered necessary to electrolytically polishspecimens so that the preparation-induced dam-age could be removed thus permitting bias-freelow-load testing. However, the science behindmechanical specimen preparation, chiefly dueto the work of Len Samuels [3], has led to devel-opment of excellent mechanical specimenpreparation procedures, and electropolishing isno longer required.

There are several operational factors that mustbe controlled in order to obtain optimum testresults. First, it is a good practice to inspect the


Figure 50. Schematic of the Knoop indenter and theshape of an impression

130°

172°30´

86

indenter periodically for damage, for example,cracking or chipping of the diamond. If you havemetrology equipment, you can measure the faceangles and the sharpness of the tip. Specifica-tions for Vickers and Knoop indenter geometriesare given in E 384.

A prime source of error in the tests is the align-ment of the specimen surface relative to theindenter. The indenter itself must be properlyaligned perpendicular (±1°) to the stage plate.Next, the specimen surface must be perpendicu-lar to the indenter. Most testers provide holdersthat align the polished face perpendicular to theindenter (parallel to the stage). If a specimen issimply placed on the stage surface, its backsurface must be parallel to its polished surface.Tilting the surface more than 1° from perpendicu-lar results in nonsymmetrical impressions andcan produce lateral movement between speci-men and indenter. The occurrence of non-symmetrical indents is generally easily detectedduring measurement.

In most cases, errors in indenting with moderntesters are not the major source of error, althoughthis can occur [26]. It is important to check theperformance of your tester regularly using acertified test block. It is safest to use a test blockmanufactured for microindentation testing andcertified for the test (Vickers or Knoop) and theload that you intend to use. Strictly speaking, ablock certified for Vickers testing at 300 or500 gf (commonly chosen loads) should yieldessentially the same hardness with loads fromabout 50 to 1000 gf. That is, if you take the aver-age of about five indents and compare the

average at your load to the average at the cali-brated load (knowing the standard deviation ofthe test results), statistical tests can tell you (atany desired confidence level) if the differencebetween the mean values of the tests at the twoloads is statistically significant or not.

Because of the method of defining HV and HK(equations given above) where we divide by d2,measurement errors become more critical as dget smaller; that is, as L decreases and thematerial’s hardness increases (discussed later).So departure from a constant hardness for theVickers or Knoop tests as a function of load willbe a greater problem as the hardness increases.For the Knoop test, HK increases as L decreasesbecause the indent geometry changes with in-dent depth and width. The degree of the changein HK also varies with test load being greater asL decreases.

The greatest source of error is in measuring theindent as has been documented in an ASTMinterlaboratory test [24,25]. Place the indent inthe center of the measuring field, if it is not al-ready there, as lens image quality is best in thecenter. The light source should provide adequate,even illumination to provide maximum contrastand resolution. The accuracy of the filar mi-crometer, or other measuring device, should beverified using a stage micrometer.

Specimen preparation quality becomes moreimportant as the load decreases, and it must beat an acceptable level. Specimen thickness mustbe at least 2.5 times the Vickers diagonal length.

Table 44. Factors Affecting Precision and Bias in Microindentation Hardness Testing

Instrument Factors Measurement Factors Material Factors

Accuracy of the applied Calibration of the measurement Heterogenity of theload system specimen

Inertia effects, Numerical aperture Strength of crystallographicspeed of loading of the objective texture, if present

Lateral movement of the Magnification Quality of specimenindenter or specimen preparation

Indentation time Inadequate image quality Low reflectivity or transparency

Indenter shape deviations Uniformity of illumination Creep during indentation

Damage to the indenter Distortion in optics Fracture during indentation

Inadequate spacing between Operator’s visual acuity Oil, grease or dirt onindents or from edges indenter or specimen

Angle of indentation Focusing of the image


87

Because the Knoop indent is shallower than theVickers at the same load, somewhat thinnerspecimens can be tested. Spacing of indents isimportant because indenting produces plasticdeformation and a strain field around the indent.If the spacing is too small, the new indent will beaffected by the strain field around the last indent.ASTM recommends a minimum spacing (centerto edge of adjacent indent) of 2.5 times theVickers diagonal. For the Knoop test where thelong diagonals are parallel, the spacing is 2.5times the short diagonal. The minimum recom-mended spacing between the edge of thespecimen and the center of the indent should be2.5 times. Again, Knoop indents can be placedcloser to the surface than Vickers indents.

Several studies have stated that the resolutionof the optical system is the major factor limitingmeasurement precision. They state that thiscauses indents to be undersized by a constantamount. However, as undersizing would increasethe measured hardness, not decrease it,resolution limits per se cannot explain thisproblem. For the Knoop indent, the chiefproblem is image contrast at the indent tips thatresults in undersized indents. This problem, plusthe variable indent shape, both result inincreasing HK with decreasing test load. Forthe Vickers test, one would assume thatundersizing or oversizing would be equallylikely; but, experience suggests that oversizingis much more commonly encountered for lowloads and small indents.

AutomationMicroindentation hardness testing is tedious, soanything that can be done to simplify testing isvaluable, especially for laboratories that do sub-stantial testing. Many adjuncts to indentmeasurement have been tried and a variety ofsuch systems are available. There has been con-siderable interest in applying image analyzersto the indent measurement task. Further, withstage automation, it is possible to automate theindenting process itself, and with the sameequipment. Figure 51 shows the BuehlerOmnimet Mht automated microindentation hard-ness testing system. This system can be used inthe fully automatic, semiautomatic or manualmode, depending upon the nature of the test-ing.

An automated system can be programmed tomake any number of indents in either a definedpattern (x number of equally spaced indents, orx number between two chosen points) or arandom pattern (locations selected at randomwith a mouse). Curved patterns can be done, notjust straight-line patterns. The system then makesthe indents at the requested load and locations,measures each indent, calculates the hardness(and desired conversions to other scales), andprints/plots the results. Statistical analysis canalso be done.

In general, Vickers indents are easier to mea-sure by image analysis than Knoop indents dueto the lower contrast at the tips of the Knoopindents (leads to undersizing and higher HKvalues). Metal flow (plastic deformation) aroundthe indent edges can interfere with measurementprecision. Vickers indents, like the ones shownin Figure 48, exhibit excellent contrast and shapeand are easily measured. If the magnification istoo high (and this may be influenced by thenumerical aperture of the objective) for a givenindent size, image contrast suffers and correctdetection of the indent will be very difficult. Onthe other hand, if the indent is very small on thescreen, it will be hard for the system to detect itautomatically. In this case, use a higher magnifi-cation, or use the semiautomatic measurementmode if this is not possible. The operator fits abox around the indent with a mouse to measurethe indent.

Microindentation hardness testing is a veryvaluable tool for the materials engineer but itmust be used with care and a full understanding


Figure 51. Omnimet Mht

88

of the potential problems that can occur. Alwaystry to use the highest possible load dependingupon the indent spacing or closeness to edges.Try to keep indents greater than 20-µm in diam-eter, as recommended in ASTM standard E 384.If you have not read E 384, or it has been sometime since you read it (all ASTM standards mustbe reviewed and revised, if necessary, everyfive years), get the latest copy and go over it.The Precision and Bias section (and appendixX1 of E 384) contains a wealth of practicaladvice. Automation of indentation and measure-ment is possible and, for laboratories that dosubstantial testing, greatly reduces operator fa-tigue while reducing testing times.

HELPFUL HINTS FOR

MICROINDENTATION HARDNESS

TESTING

• If you suspect that the test resultsobtained on a specimen are question-able, verify the tester using a certifiedtest block.

• When testing thin coatings, use theKnoop indenter with the long axis paral-lel to the surface. If you have a specimenwhere the hardness changes rapidly,Vickers indents may be distorted badlyin the direction of the hardness gradi-ent. Switch to a Knoop indenter andplace the long axis perpendicular to thehardness gradient.


97

Laboratory Safety

LABORATORY SAFETY

The metallographic laboratory is a relatively safeworking environment; however, there are dan-gers inherent to the job. Included in these dangersis exposure to heat, acids, bases, oxidizers andsolvents. Specimen preparation devices, such asdrill presses, shears and cutoff saws, also presenthazards. In general, these dangers can be mini-mized if the metallographer consults documentssuch as ASTM E 2014 (Standard Guide on Met-allographic Laboratory Safety) and relevantMaterial Safety Data Sheets (MSDS) beforeworking with unfamiliar chemicals. Commonsense, caution, training in basic laboratory skills,a laboratory safety program, access to safetyreference books – these are some of the ingre-dients of a recipe for laboratory safety. Table 48lists the main requirements for a comprehensivesafety program.

Safe working habits begin with good housekeep-ing. A neat, orderly laboratory promotes safeworking habits, while a sloppy, messy work areainvites disaster. Good working habits include suchobvious, commonsense items as washing thehands after handling chemicals or before eating.Simple carelessness can cause accidents (seeFigure 63). For example, failure to clean glass-ware after use can cause an accident for the nextuser. Another common problem is burns due tofailure to properly clean acid spills or splatter.

Most reagents, chemical or electrolytic polishingelectrolytes, and solvents should be used undera ventilation hood designed for use with chemi-cals. Many of these chemicals used inmetallography can cause serious damage oncontact. It is best to assume that all chemicalsare toxic and all vapors or fumes will be toxic ifinhaled or will be damaging to the eyes. A hoodwill prevent the working area from being contami-nated with these fumes. However, you will notbe protected when your head is inside the hood.In most cases, a protective plastic or shatterproofglass shield can be drawn across the front of thehood for further protection from splattering or anyunexpected reactions.

All laboratories should be equipped with a showerand eyewash for emergency use. This equipmentshould be near the work area so that the injuredcan reach it quickly and easily. Fire alarms and

Figure 63. Carelessness in the laboratory can resultin dangerous chemical spills and a clean-upnightmare.

1. Emergency Responsea. First Respondersb. Emergency Phone Numbersc. First Aidd. Spills

2. Chemical Procurement, Distribution, and Storagea. MSDS Managementb. Labeling/Storage

3. Laboratory SOPsa. Handling Chemicalsb. Mixing Chemicalsc. Rulesd. Hygiene

4. Employee Health and Exposurea. Personal Protective Equipmentb. Monitoring of exposure levelsc. Medical Evaluationsd. Additional protections for employees working withparticularly hazardous substances (carcinogens, toxins)

5. Disposal of Hazardous Materialsa. Etchants and Chemicalsb. Recirculating Tanksc. Lubricants and Abrasives

6. Equipment Use (JSAs)a. Sectioningb. Mountingc. Grinding and Polishingd. Electropolishinge. Etching

7. Facilitya. Facility and Equipment Maintenanceb. Housecleaningc. Fume Hood Air Flowd. Fire Extinguishers and Fire Protectione. Spill Responsef. Showers and Eye Washesg. Air Lines and Filtersh. Plumbingi. Electricals

8. Employee Traininga. Safetyb. Indications and Symptoms of Exposurec. Job Functions

9. Scheduled and Documented Safety Inspections andMeetings

Table 48. Elements of a Comprehensive Metallography Laboratory Safety Plan

98

Laboratory Safety

fire extinguishers (CO2 type) should be availableand tested periodically. A good first aid kit and achemical spill treatment kit should be readilyavailable.

Laboratory EquipmentSpecimen preparation devices used in metallo-graphic laboratories are generally quite safe touse. Information supplied by the manufacturerusually describes safe operating procedures. Itis good laboratory practice to prepare a job safetyanalysis detailing potential hazards and describ-ing the safe operating procedure for each pieceof equipment. This information should be pro-vided to all users and it must be revised andreviewed periodically.

Band saws or abrasive cutoff saws are commonlyused by metallographers. The cutting area ofband saws is exposed and potentially danger-ous. Your hands should never be used to guidethe work piece during cutting. A guiding deviceor block of wood should always be used betweenthe work piece and your hands. After cutting iscompleted, the saw should be turned off beforepieces near the blade are removed. Samplesshould be handled carefully, because consider-able heat can be generated. In addition, sharpburrs are often present, which should be care-fully removed by filing or grinding. Abrasive cutoffsaws are safer to use because the cutting areais closed off during use. The chief danger is fromflying pieces from a broken wheel. Fortunately,the closed cover contains these pieces within thecutting chamber. Wheel breakage usually occurswhen the part is not firmly clamped in place or ifexcessive pressure is applied, a bad practicefrom the standpoint of specimen damage as well.

Dust produced during grinding of metals is al-ways dangerous. For certain metals, likeberyllium, magnesium, lead, manganese, and sil-ver, the dusts are extremely toxic. Wet grindingis preferred both for dust control and for prevent-ing thermal damage to the specimen. Benchgrinders must be firmly mounted to prevent sud-den movement. Care must be exercised to avoidgrinding one’s fingers or striking the edge of agrinding belt, which will cause painful lacerations.With nearly all materials, wet grinding is preferredand produces best results. For routine handling

of dangerous metals, grinding should be donewet under a ventilation hood. Waste must behandled carefully and disposed of properly.Radioactive materials require special remote-handling facilities and elaborate safety precau-tions.

A drill press is frequently used in the laboratory.Drilling holes in thin sections requires secureclamping; otherwise the sample can be grabbedby the drill and spun around, inflicting seriouslacerations. Hair, ties, and shirt cuffs can becometangled in a drill, inflicting serious injuries. Safetyglasses should always be worn when using drillpresses or when cutting or grinding. Mountingpresses or laboratory heat-treatment furnacespresent potential burn hazards. Gloves shouldbe worn when working with these devices. Mod-ern mounting presses that cool the cured resinback to near room temperature dramatically re-duce the potential for burns. It is a good practiceto place a “hot” sign in front of a laboratory fur-nace when it is in use.

It is occasionally necessary to heat solutionsduring their preparation or use. Although Bun-sen burners are commonly employed for thispurpose, it is much safer to use a hot plate orwater bath and thus avoid the use of an openflame. If a Bunsen burner is used, the flameshould never be applied directly to a flask, bea-ker, or dish. Plain- or asbestos-centered wiregauze should always be placed between theflame and the container.

Personal Protective Equipment (PPE)Metallographer must take certain precautions toinsure their personal safety. A laboratory coat isuseful for protecting the operator’s clothing, andshould be changed regularly and cleaned pro-fessionally. When handling caustics, a rubberizedor plastic coated apron provides better protec-tion. Gloves should be worn when handling bulksamples, working with hot material, or using haz-ardous solutions. Lightweight surgeons’ glovesare very popular, because the operator retainsthe ability to “feel.” Many metallographers wearthese to protect their skin when mounting withepoxies, and polishing with oxide suspensions.When using these gloves with chemicals, alwaysinspect for holes, as they are easily punctured.

99

Laboratory Safety

Thick rubber gloves are often used for handlingspecimens during macroetching, chemical pol-ishing, pickling, etc. The gloves should alwaysbe checked first for small holes or cracks,because gloves can impart a false sense of se-curity. The operator’s hands generally perspirewhen using rubber gloves and it is sometimesdifficult to tell if the moisture is due solely to per-spiration or to leakage. Safety glasses should beworn during processes that generate particulatematter. Goggles are appropriate for use withchemicals, and a chemical face shield is recom-mended when handling large quantities ofhazardous liquids. The appropriate PPE will bespecified on the MSDS for most laboratory chemi-cals and products.

Chemicals, Storage and HandlingMany of the chemicals used in metallographyare toxic, corrosive, flammable, or potentiallyexplosive. Whenever possible, purchase smallquantities that are likely to be used within a rea-sonably short time. Flammable solvents shouldbe stored in fireproof steel cabinets. Acids andbases should be stored separately, again in fire-proof steel cabinets. Strong oxidants must notbe stored along with acids, bases or flammablesolvents.

Reagent-grade chemicals or solvents of highestpurity are recommended. Although more expen-sive, the amounts used are small and the gain insafety and reliability compensates for the costdifference. Chemicals may deteriorate duringstorage. Exposure to light can accelerate dete-rioration of some chemicals. Hence, they shouldbe stored in a closed metal cabinet.

EtchantsMost laboratories mix commonly used reagentsin quantities of 250 to 1000 mL and then storethem as stock reagents. Many reagents can besafely handled in this manner. It is best to storeonly those reagents that are used regularly.Glass-stoppered bottles are commonly used asstock reagent bottles. If these bottles are openedregularly, the stopper will not become “frozen”.However, if they are used infrequently, a frozenstopper often results. Holding the neck of thebottle under a stream of hot water will usually

loosen them. If thermal expansion does not freethe stopper, the stopper can be gently tappedwith a piece of wood. Glass bottles with plasticscrew-on tops can be used so long as the solu-tion does not attack the plastic. These bottles areuseful for holding solutions, such as nital, thatcan build up gas pressure within a tightly stop-pered bottle. A small hole can be drilled throughthe cap top to serve as a pressure relief vent.Tightly stoppered bottles of nital and some othersolutions have exploded as the result of pres-sure buildup. Be certain that the reagent is safeto store and store only small quantities. All bottlesshould be clearly labeled. Polyethylene bottlesare required for etchants containing hydrofluoricacid, which attacks glass.

Most recipes for etchants or electrolytes list theingredients by weight if they are solids and byvolume if they are liquids. In a few cases, allamounts are given in weight percentages. In mostcases, reagent compositions are not extremelycritical. An ordinary laboratory balance providesadequate weighing accuracy, while graduatedcylinders provide acceptable accuracy for volu-metric measurements. These devices should becleaned after use to prevent accidents to the nextuser. For weight measurements, a clean pieceof filter paper, or a cup, should be placed on thebalance pan to hold the chemical, to protect thepan surface, and to facilitate transfer to the mix-ing beaker. A large graduated beaker is usuallyemployed for mixing solutions.

With many etchants, the mixing order is impor-tant, especially when dangerous chemicals areused. When water is specified, distilled watershould always be used, because most tap watercontains minerals or may be chlorinated orfluorinated. Tap water can produce poor resultsor unexpected problems. Cold water should al-ways be used, never warm or hot water, whichcan cause a reaction to become violent. In mix-ing, one should start with the solvents, such aswater and alcohol; then dissolve the specifiedsalts. A magnetic stirring device is of great value,as shown in Figure 64. Then, the dangerouschemicals, such as acids, should be addedcarefully and slowly while the solution is beingstirred. Whenever sulfuric acid (H2SO4) is speci-fied, it should be added last. It should be addedslowly, while stirring, and it should be cooled, if

100

Laboratory Safety

necessary, to minimize heating. Never just pourone liquid into another, as shown in Figure 65. Ifsulfuric acid is added to water without stirring, itcan collect at the bottom of the beaker andenough local heating can occur to throw the con-tents out of the beaker.

The literature contains references to a greatmany formulas for etchants, chemical polishes,and electrolytes that are potentially dangerousor extremely dangerous. Few of these referencescontain comments regarding safe handling pro-cedures or potential hazards. For tunately,metallographic applications involve small quan-tities of these solutions, and accidents do notusually produce catastrophic results. However,

even with small solution volumes considerabledamage can be, and has been, done. Table 49lists examples from the literature (27-29) ofchemical polishing solutions, electrolytic polish-ing solutions and etchants that have beeninvolved in accidents. Table 50 lists a number ofcommonly used chemicals and incompatiblechemicals.

SolventsNumerous organic solvents are used for clean-ing or are ingredients in chemical or electrolyticpolishing solutions or etchants, in which they areused to control ionization or the speed and modeof attack. Commonly employed solvents includewater, acetone, ethyl ether, ethylene glycol, glyc-erol (glycerin), kerosene, petroleum ether,trichloroethylene, butyl cellosolve, and alcohols,such as amyl alcohol, ethanol, methanol, andisopropyl alcohol. Most are flammable and theirvapors can form explosive mixtures with air. Theyshould be kept closed when not in use and shouldbe stored in a cool place away from heat andopen flames.

Acetone (CH3COCH3) is a colorless liquid with afragrant mintlike odor. It is volatile and highly flam-mable. It is an irritant to the eyes and the mucousmembranes. Its vapor is denser than air and cantravel along the ground and can be ignited at adistance. Acetone can form explosive peroxideson contact with strong oxidizers such as aceticacid, nitric acid and hydrogen peroxide. It is anirritant to the eyes and respiratory tract, and willcause the skin to dry and crack.

Butyl cellosolve (HOCH2CH2OC4H9), or ethyleneglycol monobutyl ether, is a colorless liquid witha rancid odor that is used in electropolishing so-lutions. It is combustible and may form explosiveperoxides. It is toxic in contact with the skin, canbe absorbed through the skin, can cause seri-ous damage to the eyes, and irritation to the skin,and respiratory tract.

Carbitol (C2H5OCH2CH2OCH2CH2OH), or diethyl-ene glycol monoethyl ether, is a colorless,viscous solvent that is compatible with water andis used in electropolishing solutions. It irritatesthe skin, eyes, mucous membranes, and upperrespiratory tract, and is harmful if inhaled orswallowed.

Figure 64. When mixing etchants, use a magneticstirring plate with a magnetic bar for stirring. Slowlyadd the liquid ingredients to the solvent by drippingthem down a glass stirring rod. If the solution ismore dangerous than this one, wear protectivegloves and use a face shield. If mixing generatessubstantial heat, it is a good practice to place acooling jacket around the beaker.

Figure 65. Illustration of a bad mixing practice. Theacid ingredient was poured into an empty acid bottleand the solvents were added without stirring orcooling. The solution may erupt in the metallo-grapher’s face at any moment. It is recommended tokeep the hands clear of the area and remove rings.

101

Laboratory Safety

Table 50. Some Incompatible Chemicals

Chemical Use in Metallography Do Not Mix With the Following:

Acetic Acid Chemical polishing Chromic acid, glycol, hydroxol compounds,electrolytic polishing nitric acid, peroxides, permanganates

Acetone Degreasing, cleaning, Concentrated solutions of nitric andetchants sulfuric acid

Chromic Acid Electropolishing Acetic acid, flammable liquid, glycerol

Hydrogen Peroxide Chemical polishing, etchants Flammable liquid, organic materials

Nitric Acid (conc) Chemical polishing, etchants Acetic acid, chromic acid, flammable liquids,isopropyl alcohol

Perchloric Acid Electropolishing Acetic anhydride, alcohol, some organics,oil, grease

Sulfuric Acid Etchants Methly alcohol, potassium chlorate,potassium perchlorate and potassiumpermanganate

Table 49. Chemical and Electrolytic Polishing Solutions and Etchants Known to be Dangerous

Solution Use Problems

5 parts Lactic Acid Chemical polishing Lactic and nitric acids react autocatalytically.5 parts HNO3 solution for Zr Explosion can occur if stored.2 parts Water1 part HF

50 parts Lactic Acid Chemical polishing Lactic and nitric acids react autocatalytically.30 parts HNO3 solution for Ta, Nb Explosion can occur if stored.2 parts HF and alloys

3 parts Perchloric Acid Electropolishing Mixture is unstable and can, and has, exploded, with1 part Acetic Anhydride solution for Al heating, or in the presence of organic compounds

adding to the potential hazard.

60-90 parts Perchloric Acid Electropolishing Solution will explode at room temperature.40-10 parts Butyl Cellosolve solution Solutions with ≤30% HCIO4 will be safe if T is <20°C.

100g CrO3 Electropolishing CrO3 was dissolved in water, cooled to about 20°C;200mL Water solution the acetic anhydride was added very slowing with700mL Acetic Anhydride stirring. The solution became warm to the touch.

About 20 seconds later it erupted from the beaker.

1 part HNO3 Electropolishing Mixture is unstable and cannot be stored2 parts Methanol solution for Muntz

(Cu-40% Zn) metal

20mL HF Etchant for Nb, Ta, This etchant is unstable. At 20°C, it reacted after 1810mL HNO3 Ti, V, Zr hours. At 30-35°C, it reacted after 8 hours with30mL Glycerol violence. At 100°C, it will react after 1 minute.

20-30mL HCI Etchant for Ni and Two incidents occurred when a violent reaction10mL HNO3 stainless steels resulted producing NO2 and a spray of acid30mL Glycerol after the etch was left standing for 2-3 hours.

40mL Acetic Acid Etchant for Ni A closed bottled exploded about 4 hours after40mL Acetone it was mixed. Authors state that solutions40mL HNO3 without the acetic acid also cannot be stored.

10mL HNO3 Etchant The solution reacted spontaneously about 2 minutes10mL Acetic Acid after mixing with evolution of heat and fumes (nitrous20mL Acetone and nitric oxides). The mixed acids were poured into

the acetone. The beaker was externally cooled.

50mL Nitric Acid Etchant Mixtures have exploded violently after mixing950mL Isopropyl Alcohol or about 20 minutes after mixing.

102

Laboratory Safety

Ethylene glycol (HOCH2CH2OH) is a colorless,hygroscopic liquid with a sweet taste (but do notswallow as it is poisonous) that reacts with strongoxidants and strong bases. It is slightly flam-mable. The substance irritates the eyes, skin, andthe respiratory tract.

Glycerol (glycerin) (CH2OHCHOHCH2OH) is acolorless or pale yellow, odorless, hygroscopic,syrupy liquid with a sweet, warm taste. It is rela-tively nontoxic and nonvolatile but can causeiritis (inflammation of the iris). It is combustibleand is a moderate fire hazard. Glycerol shouldnever be used in anhydrous solutions contain-ing nitric and sulfuric acids, becausenitroglycerin can form. Glycerol should not beused with strong oxidizing agents, such as chro-mium trioxide and potassium permanganate, asan explosion may occur. Glycerol is often addedto aqua regia (glyceregia). This mixture decom-poses readily and should be discardedimmediately after use. This etchant should notbe allowed to stand for more than about 15 min-utes after mixing.

Kerosene is occasionally employed in grindingsamples and with diamond paste as a lubricant.Only the deodorized form should be used. It isflammable, but the vapors do not readily explode.Contact defattens the skin and can cause der-matitis, irritation or infections.

Trichloroethylene (CHCl:CCl2) is a stable, color-less liquid with a chloroform-like odor. Effectivelaboratory ventilation is necessary. At ambienttemperatures it is nonflammable and nonexplo-sive, but becomes hazardous at higher temp-eratures. In the presence of strong alkalies, withwhich it can react, it can form explosive mixtures.In the presence of moisture, the substance canbe decomposed by light to corrosive hydrochlo-ric acid. It is probably carcinogenic to humans,and toxic when inhaled or ingested, which maycause acute poisoning.

Amyl alcohol (CH3(CH2)4OH), or 1-Pentanol, is acolorless liquid with noxious odor. It is flammableand the vapors may form explosive mixtures atelevated temperatures. The substance reacts vio-lently with strong oxidants and attacks alkalinemetals. The fumes are irritating to the eyes, up-per respiratory tract, and skin. The substance is

toxic through ingestion, inhalation, or absorp-tion through the skin.

Ethyl alcohol (CH3CH2OH), or ethanol, is a color-less, inoffensive solvent commonly used inmetallography. Ethanol is miscible with water andrapidly absorbs up to 5% water from the air. Thedenatured version is less expensive and contains5% absolute methonal and is suitable for anyrecipe requiring ethyl alcohol. It is a dangerousfire hazard, and its vapors are irritating to the eyesand upper respiratory tract. High concentrationsof its vapor can produce intoxication. Becauseethanol is completely burned in the body, it is nota cumulative poison like methanol.

Methyl alcohol (CH3OH) is an excellent, non-hy-groscopic solvent, but it is a cumulative poison.Ingestion, inhalation or absorption through theskin in toxic levels can damage the central ner-vous system, kidneys, liver, heart, and otherorgans. Blindness has resulted from severe poi-soning. It is particularly dangerous becauserepeated low-level exposures can also causeacute poisoning as a result of accumulation.Thus, whenever possible, ethanol should beused. When using methanol, always work undera ventilation hood. Mixtures of methanol and sul-furic acid can form dimethyl sulfate, which isextremely toxic. Solutions of methanol and nitricacid are more stable than mixtures of nitric acidand higher alcohols.

Isopropyl alcohol [CH3CH(OH)CH3], also knownas 2-propanol, is a clear, colorless liquid that, likeethanol, does not accumulate in the body, al-though it does have a strong narcotic effect. It isa flammable liquid and is a dangerous fire haz-ard. Metallographers have used it as a substitutefor ethanol but isopropyl alcohol has quite differ-ent characteristics and should not be used. Fatalinjuries and explosions have been reported dueto its use.

AcidsInorganic and organic acids are common con-stituents in chemical and electrolytic polishingsolutions and in etchants. The inorganic, or min-eral acids, including the very familiar acids suchas hydrochloric, nitric, perchloric, phosphoric, andsulfuric, are highly corrosive and poisonous.Theyshould be stored in a cool, well-ventilated loca-

103

Laboratory Safety

tion away from potential fire hazards and, ofcourse, away from open flames. They shouldnot be stored in a location that receives directsunlight. When the pure acids contact metals,most liberate hydrogen gas – a fire and explo-sion hazard. The organic acids are naturallyoccurring substances in sour milk, fruits, andplants and include the following acids: acetic,lactic, citric, oxalic, and tartaric.

Hydrochloric acid (HCl), commonly used in met-allography, is a colorless gas or fuming liquid witha sharp, choking odor. It is very dangerous tothe eyes and irritating to the nose and throat. Itattacks the skin strongly causing severe burns.

Nitric acid (HNO3), also commonly used in met-allography, is a colorless or yellowish fumingliquid, highly toxic and dangerous to the eyes. Ifit contacts organic material or other easily oxi-dizable materials, it can cause fires and possiblyexplosions. When it reacts with other materials,toxic oxides of nitrogen are produced. The ox-ides, which vary with the conditions, includenitrous acid, nitrogen dioxide, nitric oxide, nitrousoxide, and hydroxylamine. A commonly encoun-tered problem involves pouring nitric acid into agraduated cylinder that contains some methanolor ethanol from prior use. The brown fumes givenoff are quite harmful. Mixtures of nitric acid andalcohols higher than ethanol should not be stored.Mixtures of concentrated nitric and sulfuric acidsare extremely dangerous, while strong mixturesof nitric acid and glycerin or glycols can be ex-plosive. Aqua regia, a mixture of one part nitricacid and two to four parts hydrochloric acid, formsseveral products including nitrosyl chloride, anexceptionally toxic gas. Aqua regia is a popularetchant but must be used with care under a hood.

Ethanol with additions of up to 3% nitric acid(nital) can be safely mixed and stored in smallquantities. Higher concentrations result in pres-sure buildup in tightly stoppered bottles.Explosions of 5% nitric acid in ethanol have oc-curred as a result of failure to relieve thepressure. If higher concentrations are desired,they can be mixed daily, placed in an open dish,and used safely. Discard the etchant at the endof the day. Mixtures of methanol with up to 5%nitric acid are safe to use and store in smallquantities. Mixtures of methanol with more than5% nitric acid if heated are subject to violent

decomposition. Mixtures of 33% nitric acid inmethanol have decomposed suddenly and vio-lently.

Never add nitric acid to isopropyl alcohol. Ander-son (27) reported that a litre bottle of 5% nitricacid in isopropyl alcohol was mixed and placedin a cabinet. Although this had been done manytimes in the past without problems, twenty min-utes later the bottle exploded destroying thecabinet, other stored bottles, and throwing de-bris up to 20 feet away. Anderson (27) alsoreported that a metallographer was pouring afreshly mixed litre of 5% nitric acid in isopropylalcohol into another bottle when it exploded. Theperson died within three hours without being ableto tell anyone what happened. As with the otherexplosion, this same procedure had been per-formed many times previously without mishap.Anderson recommends avoiding use of isopro-pyl alcohol completely.

Most metallographers consider nital to be verysafe to use, and indeed it is. However, even withsuch an apparently safe solution, one can haveaccidents. One such accident occurred when anemployee, not a skilled metallographer, was re-plenishing a stock of 5% nitric acid in ethanolusing a procedure that he had claimed to haveperformed many times previously (he was nottaught the safe way to mix nital as it was a unionchemist’s job to mix nital; i.e., not his job). Theworker began by adding the desired volume ofconcentrated nitric acid into the container thatcontained a small residual amount of stale 5%nital. To his surprise, the contents began “boil-ing” and spewing out of the container along withdense, brown fumes. The acid splashed theworker, resulting in burns on his forehead, faceand eyes. The small amount of aged nitric acidsolution (the concentration may have been in-creased due to evaporation of the alcohol),present in the container when the fresh acid wasadded, created a dangerous chemical reaction.An experiment also showed that a similar reac-tion can occur when nitric acid is poured into agraduated cylinder containing only remnants ofethanol or methanol.

Sulfuric acid (H2SO4) is a colorless, oily liquidthat is highly corrosive, a strong oxidizing agent,and dangerously reactive. It reacts violently with

104

Laboratory Safety

bases and is corrosive to most metals formingflammable/explosive hydrogen gas. It reacts vio-lently with water and organic materials withevolution of heat. Upon heating, toxic sulfur ox-ides are formed. One should add sulfuric acidvery slowly to water with constant stirring. If addedwithout stirring, it will produce a pocket of steamin the bottom of the vessel, throwing the con-tents out of the vessel. Concentrated sulfuric acidcan cause severe, deep burns on contact withthe skin, and permanent vision loss on contactwith the eyes. Tissue is destroyed by the acid’sdehydrating action. Lungs may be affected bylong-term or chronic exposure to aerosol. Skinlesions, tooth erosion, and conjunctivitis are otherlong-term effects.

Hydrofluoric acid (HF) is a clear, colorless, fum-ing liquid or gas with a sharp, penetrating odor.It is very dangerous to the eyes, skin, and upperrespiratory tract. The substance can be absorbedinto the body by inhalation, through the skin andby ingestion. A harmful concentration of the gasin air can be reached quickly, making it very dan-gerous to handle. Exposure by ingestion,inhalation or contact, can be fatal. UndissociatedHF poses a unique threat in that it can destroysoft tissues and result in decalcification of thebone. Moreover, the effects may be delayed.Laboratories where HF is used should stock anantidote kit to be used in case of exposure. Al-though it is a relatively weak mineral acid, HFwill attack glass or silicon compounds, and shouldbe measured, mixed, and stored in polyethylenevessels. HF reacts with many compounds, in-cluding metals, and will liberate explosivehydrogen gas.

Orthophosphoric acid (H3PO4), a colorless thickliquid or hygroscopic crystal, is a medium strongacid. It is corrosive to the skin, eyes and respira-tory tract. Phosphoric acid decomposes oncontact with alcohols, aldehydes, cyanides, sul-fides, ketones, and can react with halogenatedorganic compounds forming organophosphorusnerve-gas type compounds that are extremelytoxic. It reacts violently with bases, and will gen-erate hydrogen gas when it reacts with metals.

Perchloric acid (HClO4) is a colorless, fuming, hy-groscopic liquid. It is extremely unstable inconcentrated form and may explode by shock

or concussion when dry or drying, so commer-cially available perchloric acids come inconcentrations of 65 to 72%. In this form, con-tact with perchloric acid will cause irritation andburns, while its fumes are highly irritating to themucous membranes. Contact with organic, orother easily oxidized material, can form highlyunstable perchlorates, which can ignite andcause explosions. Regular use of perchloric acidrequires that the ventilation system must be spe-cifically designed and maintained for perchloricacid. Special fume hoods with a waterfall-typefume washer will remove the perchlorate fumesbefore they can enter the exhaust system.

Perchloric acid is very useful in electropolishingsolutions. However, never electropolish samplesmounted in phenolic (Bakelite) or other plasticsin perchloric acid solutions as explosions canresult. The mixture of perchloric acid and aceticanhydride, which was developed by Jacquet, isdifficult to prepare and highly explosive. Jacquethas reviewed the accidents involving perchloricacid and has described safety procedures (30).The worst accident occurred on February 20,1947, in an electroplating factory in Los Ange-les. In this accident 17 people were killed and150 were injured (29). Medard, Jacquet, and Sar-torius have prepared a ternary diagram showingsafe compositions of perchloric acid, acetic an-hydride, and water, Figure 66. Anderson,however, states that accidents have still occurredwith solutions in the “safe” region of this diagram(27). Thus, electropolishing solutions composedof perchloric acid and acetic anhydride are notrecommended. Indeed, many companies forbidthe use of such mixtures, and some cities havebanned their use. Electropolishing solutions ofperchloric acid and alcohol, with or without or-ganic additions, and mixtures of perchloric acidand glacial acetic acid are safe to use. Never-theless, in using these “safe” mixtures, oneshould follow the formula instructions carefully,mix only small quantities, keep the temperatureunder control, and avoid evaporation. These so-lutions should not be stored.

Mixtures of acetic acid and 5-10% perchloric acidhave been commonly used to electropolishiron-based alloys and are reasonably safe. Donot use these solutions to electropolish bismuth,arsenic or tin, as explosions have occurred.Anderson suggests that arsenic, antimony, and

105

Laboratory Safety

tin may also be incompatible with perchloricelectrolytes (27). Do not store these electrolytesfor more than a few days. Discard them whenthey become colored by dissolved metallic ions(from electropolishing). Always keep these solu-tions cool; increasing the temperature increasesthe oxidizing power of perchloric acid.

Comas et al. have studied the hazards associ-ated with mixtures consisting of butyl cellosolveand from 10 to 95% of 70% perchloric acid (31).Mixtures with 60 to 90% acid were explosive atroom temperature. Acid concentrations of 30%or less were inflammable but were judged to besafe to use as long as the operating temperaturedoes not exceed 20°C.

Acetic acid (CH3COOH) is a clear, colorless liq-uid with a pungent odor. It is a weak acid thatreacts with strong oxidizers, bases and metals.It is flammable and is not easily ignited, althoughwhen heated, it releases vapors that can be ig-nited and can travel some distance to an ignition

source. Contact with the skin results in seriousburns. Inhalation of the fumes irritates the mu-cous membranes. Anderson states that aceticacid is a good solvent for nitric acid and that a50% solution can be prepared, but not stored,without danger (27). Sax, however, states thatmixtures of nitric and acetic acids are danger-ous (32).

Acetic anhydride [(CH3CO)2O], or acetic oxide,is a colorless liquid with a very strong acetic odor.It can cause irritation and severe burns to theskin and eyes. Acetic anhydride decomposes onheating producing toxic fumes. It reacts violentlywith boiling water, steam, strong oxidants (spe-cifically sulfuric acid), alcohols, amines, strongbases, and others. It attacks metals, and is verycorrosive, especially in presence of water ormoisture. It is extremely flammable and shouldbe avoided. The electrolytic polishing mixturesof acetic anhydride and perchloric acid (4-to-1to 2-to-1 mixtures) developed by Jacquet, asmentioned above, are exceptionally dangerous

Figure 66. Diagram developed by Médard, Jacquet and Sartorius showing safe (A and 1 to 9) nonexplosiveelectropolishing solutions and explosive (C, E and Los Angeles) compositions of perchloric acid, acetic anhydrideand water (brough in by the perchloric acid, but not corrected for the effect of acetic anhydride).

89

Image Capture and Analysis

IMAGE CAPTURE ANDANALYSIS

The progress in computer and video technol-ogy has created a movement toward electronicimage acquisition. These images can be usedin software applications such as word process-ing or desktop publishing programs allowing forfast report generation and electronic distribu-tion. In addition, associated data can be storedwith the images in a database allowing for easyretrieval of information through searches.

When adding imaging capabilities to the labora-tory, it is important to consider your goals.Oftentimes, images are just another step in thedocumentation process. For example, in a fail-ure investigation it is useful to capture the imageof a complete component before the sectioningprocess. Or, an image of the microstructure mightbe attached to a report as an indication of a pass/fail condition. Additional functionality of an imag-ing system might include a scale marker overlayand point-to-point or other operator interactivemeasurements. Fully automated image analysisincludes detection of the features of interestbased on grey level or color differences as wellas morphological characteristics such as size andshape. Automated imaging applications arebased on a fundamental series of steps shownin Table 45. Depending on your goals, some orall of these steps may be utilized.

Aligned with these imaging goals, Buehler offersa series of upgradeable Omnimet imagingproducts (Figure 52). The Omnimet ImageCapture and Report System focuses on the ba-sics of image capture and generating singleimage reports. The Omnimet Archive imagingproduct has additional capturing features that

enable image comparison and multi-layerfocusing as well as a powerful database for track-ing projects and measurements. The OmnimetExpress and Omnimet Enterprise products in-clude automated measurements in addition tothe operator interactive measurements found inthe other products. Omnimet Express has spe-cific application modules while the OmnimetEnterprise product enables the user to tackleany application.

AcquisitionImage capture is a term used to describe imageacquisition by means of a camera and framegrabber or a digital camera. Because of themany choices of camera types, a video micros-copy system must be flexible. Analog CCDcameras, black and white or color, are most fre-quently used. Component video (Y/C or S-Video)and composite video signals and a number ofcolor video standards such as NTSC, PAL andSECAM are generally supported. Images ac-quired in the materials laboratory are optimizedin real time by adjusting brightness, contrast,and color saturation. The analog output camerasignal is then digitized utilizing an analog inputframe grabber board.

Table 45. Primary Imaging Steps

Imaging Steps Description

Acquisition Capture, load, or import an image

Clarification Develop the necessary contrast and clarity for detecting the features of interest

Thresholding Detect the features of interest

Binary Operations Clean-up any detection discrepancies, categorize features, and overlay grids

Measurements Conduct field or feature measurements

Data Analysis Evaluate statistics and relevance of the measurements

Archive Store images, annotations, and associated measurements in a database

Distribution Printing or electronic distribution of images and results

Figure 52. Omnimet Imaging product line

90


Alternatively, digital cameras may be integratedwith a SCSI or USB connection. Since the late1990’s digital video cameras have become in-creasingly popular for scientif ic imagingapplications. This is largely accounted for by twofactors: 1) higher pixel densities in the acquiredimages potentially provide higher spatial resolu-tion and 2) the cost range of digital cameras hasmerged with the combined cost of an analog cam-era and specialized capture board. Capturing animage with a digital camera is often a two-stageprocess, first a preview image is viewed for fo-cusing and selecting the field of interest and thena snapshot of the image is taken. The previewimage window displays the image at a lowerpixel density and the refresh rate of the camerawill determine the amount of lag time betweenthe operator adjusting the microscope and thenew image being displayed. In order to be com-parable to an analog camera, the refresh rateshould approach 25 frames per second.

The resulting image files are based on two pri-mary formats, bit-map and vector. The majorityof scientific imaging programs are based on bit-mapped images. Bit-mapped graphics areessentially grids, consisting of rows and columnsof pixels. When the image is zoomed to observedetails, the individual pixels will become vis-ible. Pixels, or picture elements, are the smallestpart of an image that a computer, printer or moni-tor display can control. An image on a monitorconsists of hundreds of thousands of pixels, ar-ranged in such a manner that they appear to beconnected. Figure 53 displays a cast iron imagewith a small segment zoomed such that the pix-els are observed.

Each pixel in a bitmap is stored in one or moredata bits. If an image is monochromatic (blackand white), then one bit is enough to store infor-mation for one pixel. If an image is colored oruses various shades of grey, then additional bitsare required to store color and shading informa-tion. Table 46 lists some of the more common bitdepths referred to in imaging software and cam-era literature.

When working in grayscale, 256 values are usedto represent each shade of grey, starting withblack at 0 and transitioning to white at 255. Forexample, the information stored for the image inFigure 53 would be a combination of the x and ypositions and a number from 0 to 255. For colorimages there are two common models used tonumerically represent color, RGB and HLS. Eachrequires storing three values, in addition to the xand y position of the pixel, in order to recreatethe image. In general, you will notice the file sizesfor color images are larger than grey scale im-ages of the same pixel density.

The RGB model is based on three primary col-ors – red, green and blue – that can becombined in various proportions to produce dif-ferent colors. This scheme is considered additivebecause, when combined in equal portions, theresult is white. Likewise, an absence of all threeis black. The three primary colors are each mea-sured as a value from 0-255. The colorsproduced by combining the three primaries area result of the relative values of each primary.For example, pure red has a red value of 255, agreen value of 0, and a blue value of 0. Yellowhas a red value of 255, a green value of 255,and a blue value of 0. This system is modeled asa cube with three independent directions: red,green, and blue, spanning the available space.The body diagonal of the cube contains lumi-nance, in other words, colorless informationabout how light or dark the signal is.

The HLS model uses the concept of Hue, Lumi-nance, and Saturation. Hue is the color tone. For

Table 46. Bit Depth

Bits

1 21 or 2 tones, black and white

8 28 or 256 grey scale shades

24 224 or 16.7 million colors

32 232 or 4.29 billion colors

Figure 53. A small segment of an image zoomed todisplay the individual pixels..

91

example, an etchant might highlight the differ-ent phases as unique shades of brown or blue.Saturation is complementary to the hue. It de-scribes how brilliant or pure – as opposed towashed out or grayish a particular color is.Luminance is correlated with the light intensityand viewed as comparable to the greyscale val-ues where zero is black and 255 is white. TheHLS color space can be viewed as a doublecone, in which the z-axis of the cone is thegrayscale progression from black to white, dis-tance from the central axis is the saturation, andthe direction or angle is the hue.

The illustration in Figure 54 represents pixels ofnine different color values. These values areshown in Table 47 for the greyscale, RGB, andHLS models. When the pixels are not truegrayscale tones, the greyscale number is notgiven.

Resolution is primarily the imaging system’s abil-ity to reproduce object detail by resolving closelyspaced features. In digital microscopy, the reso-lution is examined in terms of both themicroscope limitations and the pixel array of thecamera. The clarity and definition of a digitalimage depends largely on the total number ofpixels used to create it. Typical pixel array densi-ties range from 640 x 480 to 3840 x 3072.Multiplying the two numbers gives you the totalnumber of pixels used to create the image. Be-cause a large number of pixels leads to sharperimages, you might conclude that it is always de-sirable to capture images with the largest pixelarray possible. That is not necessarily alwaystrue. The problem with large array images is thatthey consume a lot of storage space. As a result,these images are not recommended for posting


Table 47. Values based on color models

HLS Method RGB MethodPixels Greyscale Hue Luminance Saturation Red Green Blue

1 192 0 192 0 192 192 192

2 150 0 150 0 150 150 150

3 128 0 128 0 128 128 128

4 0 127 255 255 0 0

5 85 127 255 0 255 0

6 170 127 255 0 0 255

7 60 98 100 255 255 0

8 310 36 79 148 31 129

9 25 100 176 169 113 31

on web sites or email use. It is important to takeinto account what you intend to do with yourimages when selecting the appropriate camera.

In order to perform measurements, the image orthe image source must first be calibrated. Cali-bration is achieved by assigning a knowndistance to a pixel count. When using a light mi-croscope this is accomplished with a stagemicrometer. Each of the objectives will have itsown unique calibration factor. If the aspect ratioof the camera pixels is unknown, it can be deter-mined by calibrating in both the x and y direction.If the aspect ratio is known, then calibration onlyneeds to be completed in the x direction. Thiscalibration technique is repeated for each objec-tive or working magnification. It is also possible

Figure 54. A representation of nine pixels withdifferent color values.

92

tions can be as simple as a label that is refer-enced in the image caption or a complete textdescription of a feature within the image. A rec-ommended annotation is a scale marker in thelower corner of the image. This is particularlyimportant for distributing the image electroni-cally because the final size of the printout is atthe discretion of the recipient.

Automated MeasurementsWhen more detailed measurements are re-quired or a large quantity of measurements isimportant, it is useful to automate the process.For example, consider the coating that was mea-sured in Figure 56. It is possible to duplicate thatmeasurement as well as to evaluate any varia-tion across the coating and determine thepercentage porosity within the coating. A typi-cal process for determining coating thickness isoutlined in Figure 57 and is described below.

ThresholdingThresholding is the way of representing rangesof pixel grey or color values with different colorbitplane overlays. The bitplanes are a binary layerplaced over the image plane superimposing thephase(s) of interest. Most of the measurementsare accomplished on this layer not on the actualimage.

The first step of the thresholding process is thedevelopment of a threshold histogram. Each pixelis assigned a value and then the frequency isgraphed. When working in grayscale with an 8-bit image, 256 values are used to represent eachshade of grey, starting with black at 0 andtransitioning to white at 255. If you are working

to calibrate images that have been importedfrom another source. The only requirement is tohave a feature within the image that is of a knowndimension. A scale marker is usually the bestchoice.

ClarificationImage clarification is also referred to as imageenhancement and is largely achieved through theuse of greyscale filters. Filters perform one ofseveral standard functions: edge detection,photographic enhancement, or grey level modi-fication. Image clarifications, which modify thevalues of the pixels across an entire image, areconducted in one of two ways: adjusting limitssuch as contrast, brightness and gamma or com-paring the values of neighboring pixels. Thelatter is referred to as a convolution or kerneloperation. The neighborhood sizes are typicallysquares of pixels from 3x3, 5x5, 7x7, etc. A typi-cal need is to increase the local contrast at thephase boundaries. Often times, the use of aneighborhood transformation filter results in amore narrow gray scale distribution of a givenphase, making the subsequent thresholding ordetection process easier.

Operator Interactive MeasurementsInteractive measurements require the operatorto use a mouse to select the start and end pointsfor each measurement. The most common ofthese are linear measurements (point-to-point),parallel, radius or diameter, curvilinear, and angle(Figure 55). An operator will typically make sev-eral measurements over the area of interest.These are either electronically transferred to afile or transcribed by hand. This method is ap-propriate for a limited number of measurementsor where an average is of interest rather thanvariation. Figure 56 represents the most directway to take an average measurement of thick-ness using a parallel line tool. One line is alignedwith the interface and then the opposite line isaligned with the median of the rough surface.

In addition to measurements, other annotationsare often placed on electronic images. Annota-


Figure 55. Operator interactive tools founds inimaging software.

Figure 56. An example of an operator estimatingcoating thickness with the use of a parallel tool.

93


Figure 57. A flowchart of the logic used when creating an automated measurement routine for coatingthickness.

94

with a color image, the same 256 values corre-spond to luminance. Figure 59 demonstrates ayellow bitplane being assigned to the coating(Figure 58) based on the HLS values of the pix-els as discussed previously.

Binary OperationsAfter the thresholding process has been accom-plished, the different phases and features in animage are represented by different bitplane coloroverlays. It is possible that more than one phaseor feature of interest was detected by the samebitplane color because of a similar gray levelrange. A binary operation is the process thatseparates and classifies features within thesame bitplane, based on morphology or size.Additionally, not all images are detected as de-sired. There may be over-detected orunder-detected regions that need to modified.In general, binary operations allow the operatorto isolate the features that are of interest.

Two common measurements on this type of coat-ing would include the relative area fraction ofporosity as well as the average thickness. Be-fore either of these measurements can beautomated it is necessary to define the totalarea of the coating. A combination of the binarycommands Fill and Close are used to eliminatethe holes (undetected areas) that are totallyenclosed by the detected bitplane. The thicknessis determined by averaging a series of chordmeasurements across the coating, rather than asingle measurement of the yellow bitplane. Anadditional binary layer is superimposed in theform of a grid using a Grid command. Figure 60demonstrates the effects of these commands onthe coating.

To define the individual chords, where the gridoverlaps the coating, a Boolean command isemployed. Boolean operations perform logicalfunctions between bitplanes. The two most com-mon are OR and AND. AND takes those portions


Figure 58. A TSC color image captured at 1200 x1600 pixels.

Figure 60. The effect of the Binary commands,Close, Fill, and Grid on the detected coating.

Figure 59. Thresholding the coating with ayellow bitplane overlay using the HLS colormodel.

95


of two bitplanes that overlap and integrates theminto a single destination bitplane. In the coat-ing, for example, AND considers only thosesegments of the red grid bitplane that overlapthe yellow coating bitplane. Figure 61 showsthe red chords which result from the AND com-mand. OR combines two source bitplanestogether to form a single destination bitplane.For example, if two different inclusion types,oxides and sulfides, are detected initially, theycan be combined to measure the total inclusioncontent.

One of the most powerful binary functions is theability to categorize individual features basedon shape or size. It can be a single specification(>5-µm) or multiple specifications (>2 and <5)separated by an AND or an OR. If a maximumallowable pore size is specified, all pores thatexceed that limit can be flagged. The limit itselfcan be based on average diameter, maximumdiameter, area, aspect ratio, etc. Additionally, theshape of the pore may have some significance.For example, in cast materials, a gas bubbletends to be circular while a shrinkage cavity willhave a more complex shape. Common shape fac-tors examine the relationship between theperimeter and area of the feature. For this coat-ing, the pores were separated from the elongatedoxides to insure that the porosity level was notoverestimated.

Returning to Figure 61, where grid l inesare placed across the coating, the length of allthe individual grid lines can be measured. Thedata can be shown as a histogram offering theminimum, maximum, mean and standard devia-tion of the measurements across the coating(Figure 62).

Most measurements fall into one of twocategories: field or feature-specific measure-ments. The grid lines across the coating aretypical of feature measurements. Each individualfeature is measured and then plotted based onfrequency in the histogram and contributes to theoverall statistics. In the case of grain or particlesize measurements, features on the edge of theimage are eliminated because the complete sizeof the feature is unknown.

Field measurements are performed on the en-tire field of view or a framed portion of the image,providing the sum of the individual measurementsin the selected area. Statistical information forfield measurements is only generated if multiplefields are analyzed. This can be employed toobserve microstructural variations within differ-ent fields of a specimen. A fairly common fieldmeasurement is area percentage. An area per-centage measurement results from dividing thenumber of pixels in the bitplane of interest by thetotal number of pixels in the image. In the caseof the coating, there is a need to measure thearea percent porosity relative to the coating areaand not the entire image area. This is calculatedby dividing the number of pixels in the bitplaneof interest, i.e. the pores in the coating, by thetotal number of pixels representing the coating.

Common ApplicationsThroughout the past 20 years of Buehler’s in-volvement in image analysis, a number ofapplications stand out and may be of interest tomost image analysis users. Below is a listing ofthese applications:

Grain Size. Image analysis provides a rapid andaccurate means for determining grain size ac-cording to ASTM E 112. Even if etching is unable

Figure 61. The chords that result from theoverlapping coating and grid lines.

Figure 62. The histogram of results representingthe chords shown in Figure 61.

96


to produce complete grain boundaries, or if thereare twins that could skew the data, binary modi-fications can be employed to make corrections.If a specification cites maximum grain size limi-tations, the excessively large grains may betransferred to a different bitplane color to pro-vide visual and numerical feedback.

Porosity. Porosity is detrimental to the physicalproperties of most engineering materials. Imageanalysis is able to characterize the pores accord-ing to the total number of pores, number per unitarea, maximum size, average size and the sizedistribution in the form of a histogram.

Linear Measurements. Simple point-to-pointmeasurements are widely used for making oc-casional measurements; however, in caseswhere a high quantity of measurements andmore statistics are required, automated imageanalysis is time saving. A coating or layer isdetected based on the pixel values and then,after binary isolation of the coating, grid linesare superimposed. Using Boolean logic to evalu-ate the common pixels between the layer andgrid lines, the result is many chords represent-ing the thickness at given points in the coating.

Feature Shape and Size. The shape of the graph-ite constituent in ductile irons is critical. Ductileiron was developed such that the graphite wouldoccur in the form of spherical nodules with theresult of dramatically improved mechanical prop-erties. However, variations in chemistry and otherfactors can cause the nodules to be irregular,leading to some degradation of the properties.The ability to monitor the graphite shape or de-termine “nodularity” is another ability of imageanalysis. These same techniques are applicableto any constituent that can be detected.

Phase Percentage. The area percent of vari-ous phases in a microstructure influences theproperties. The tensile strength of grey iron, forexample, is directly related to the percentage ofpearlite in its microstructure. In addition, in asingle image, multiple phases can be detected,measured and presented in one graph. For ex-ample, when evaluating the inclusion content ofsteels, it would be useful to examine the overallinclusion content as well as isolating particularinclusion types, such as oxides and sulfides.

HELPFUL HINTS FOR IMAGE

CAPTURE AND ANALYSIS

• Determine the required pixel density ofa camera system based on the size ofthe features of interest and acceptablefile size

• When distributing an image, always adda scale bar

• When saving images for analysis at alater time, verify that the image is storedas captured, i.e., without any filtersapplied.

106

and should never be used. Dawkins (33) re-ported an accident involving a mixture ofchromium trioxide and acetic anhydride that hadbeen used for electropolishing (Table 49).

Citric acid [C3H4(OH)(COOH)3·H2O] comes as col-orless, odorless crystals that are water-soluble.It is an irritant to the skin, eyes and respiratorytract; no unusual problems are encountered ex-cept for occasional allergic reactions.

Lactic acid (CH3CHOHCOOH) is a yellow or col-orless thick liquid. It is damaging to the eyes.

Oxalic acid (COOHCOOH·2H2O) comes as trans-parent, colorless crystals. It is poisonous ifingested and irritating to the upper respiratorytract and digestive system if inhaled. Skin con-tact produces caustic action and will discolor andembrittle the fingernails. It is not compatible withnitric acid, as it reacts violently with strong oxi-dants. It can also form explosive compounds dueto reacts with silver.

Picric acid [(NO2)3C6H2OH], or 2,4,6-trinitrophe-nol, comes as yellow crystals that are wet with10 to 35% water. When picric acid is dry, it is adangerous explosive. It is toxic and stains theskin. It is incompatible with all oxidizable sub-stances. Picrates, which are metal salts of picricacid, are explosive. When picrates are dry, theycan detonate readily, possibly spontaneously.Purchase in small quantities, keep it moist, andstore it in a safe, cool place. If it starts to dry out,add a small amount of water to keep it moist.The maximum solubilities of picric acid inwater and in ethanol are about 1.3 and 8 g per100 mL, respectively. Picral can be stored safely.During use, the solution should not be allowedto dry out. The etching residue should be dis-carded at the end of the day to avoid potentialexplosions.

BasesBases, such as ammonium hydroxide (NH4OH),potassium hydroxide (KOH), and sodium hydrox-ide (NaOH), are commonly used in metallo-graphy, chiefly in etchants.

Ammonium hydroxide is a colorless liquid witha strong, obnoxious odor. Solutions are ex-tremely corrosive and irritating to the skin, eyes,

and mucous membranes. It reacts exothermi-cally with sulfuric acid and other strong mineralacids, producing boiling solutions.

Sodium and potassium hydroxides are strongbases, available as white deliquescent pelletsthat are soluble in water. They can rapidly ab-sorb carbon dioxide and water from the air.Solutions stored in flasks with ground glass stop-pers may leak air and freeze the stoppers,making re-opening difficult. Dissolving NaOH orKOH in water will generate considerable heat.Do not dissolve either in hot water. Never pourwater onto these hydroxides; always add thepellets slowly to the water. Alkali metal hydrox-ides react violently with acid, and are corrosivein moist air to metals like zinc, aluminum, tinand lead forming flammable/explosive hydro-gen gas. They are very corrosive to skin, eyesand respiratory tract. Long-term exposure maylead to dermatitis. Potassium hydroxide is some-what more corrosive than sodium hydroxide.

Other ChemicalsHydrogen peroxide (H2O2) is available as a liquidin concentrations of either 3 or 30%. The 3% so-lution is reasonably safe to use, while the 30%solution is a very powerful oxidant whose effecton the skin is about as harmful as that producedby contact with sulfuric acid. Hydrogen peroxideby itself is not combustible, but if brought in con-tact with combustible materials, it can produceviolent combustion. Hydrogen peroxide is verydamaging to the eyes. Because release of oxy-gen can cause high pressures to develop withinthe container, the container caps are vented.

Bromine (Br2), a fuming reddish brown liquid witha pungent, suffocating odor, is commonly usedin deep-etching solutions. It is very corrosive, re-acting violently with easily oxidized substances,including some metals. Bromine is a dangerousliquid that should only be handled by well-quali-fied personnel. Its vapors are extremely irritatingto the eyes, skin, and mucous membranes. Skincontact produces deep, penetrating burns thatare slow to heal. Contact with organic matter cancause fires.

Chromic acid (H2CrO4) is formed when chromiumtrioxide (CrO3) is dissolved in water. CrO3 is usedin electropolishing solutions (see previous com-

Laboratory Safety

107

ment and Table 49 about explosive nature ofmixtures with acetic anhydride). Dilute aqueoussolutions are widely used for attack polishing. Itis a powerful oxidant; always wear gloves whenusing it for attack polishing or use automatic de-vices and avoid contact potential. Chronic orlong-term inhalation exposure may produceasthma-like reactions.

Potassium permanganate (KMnO4), a black crys-talline powder, is a powerful oxidant used inetchants. It is a dangerous fire and explosion haz-ard, especially when in contact with organicmaterials. Ingestion produces serious damage.KMnO4 and sulfuric acid should never be mixedtogether because a violent explosion can result.

Potassium dichromate (K2Cr2O7), a bright orangecrystalline powder, is another powerful oxidantthat is used in etchants. Contact can cause ul-ceration of the hands, severe damage to nasaltissue, or asthma and allergies with long-termexposure.

Cyanide compounds are occasionally used inmetallographic applications. Potassium cyanide(KCN) and sodium cyanide (NaCN) are extremelydangerous and highly toxic. Exposure by eye orskin contact, or by ingestion, is fatal. NaCN andKCN vapors are intensely poisonous. They areparticularly hazardous when brought in contactwith acids or acid fumes because of liberation ofhydrogen cyanide, which is extremely toxic andhighly flammable. Potassium ferricyanide(K3Fe(CN)6), a ruby-red crystalline powder andan ingredient in Murakami-type reagents, is poi-sonous but stable and reasonably safe to use.

A number of nitrates, such as ferric nitrate[Fe(NO3)3·6H2O], lead nitrate [Pb(NO3)6], and sil-ver nitrate (AgNO3), are employed by metallo-graphers. Since they are powerful oxidizers, theypose a dangerous fire hazard, especially whenin contact with organic materials. They mayevolve toxic fumes, such as oxides of nitrogenand lead, and are poisonous and corrosive to theeyes, skin and respiratory tract.

SummaryIn general, the metallographic laboratory is areasonably safe environment. However, depend-ing on the materials being prepared, dangerous

Laboratory Safety

situations can arise. Some hazards, such as thepreparation of radioactive or reactive metals, arequite obvious, while others are not. In the pre-ceding discussion, some of the potential hazardsthat can be encountered are summarized; oth-ers undoubtedly exist that are not covered.

Most accidents can be prevented by simplecommonsense rules. It is best to assume that allmetal dust and all chemicals are hazardous. In-halation of dust and fumes, ingestion, or bodilycontact should be avoided. Personal protectiveequipment is very useful, but it should not be usedas a substitute for good laboratory procedures.The use of such equipment does not guaranteefreedom from injury.

The metallographic literature contains many ref-erences to the use of dangerous materials, oftenwithout any mention of the dangers involved orsafe handling procedures. This is unfortunate be-cause the unwary may be injured. Many of usare tempted to experiment when a recommendedprocedure does not work as claimed. The devel-opment of electrolytes, chemical polishingagents, or etchants should be left to those whoare fully versed in the potential dangers. Metal-lographic laboratories should have some of thereferenced safety publications (see bibliography)readily available in the laboratory and thesesafety publications should be consulted whenworking with new or infrequently used materials.

HELPFUL HINTS FOR LABORATORY

SAFETY

• Purchase realtively small quantities ofchemicals that are used routinely, es-pecially those that have a short shelf life,or are dangerous.

• After pouring acids, carefully replace thecap, wash the exterior of the bottle un-der running water, and dry off thesurface with a paper towel before replac-ing the acid bottle in its fireproof cabinet.

• Schedule regular time periods for labo-ratory clean-ups. These should beweekly for high volume laboratories. Aclean lab is an essential ingredient in alllaboratory safety programs.

108

SUMMARY

Preparation of metallographic specimens isbased upon scientific principles that are easilyunderstood. Sectioning creates damage thatmust be removed by the grinding and polishingsteps if the true structure is to be examined. Eachsectioning process exhibits a certain amount ofdamage, thermal and/or mechanical. Conse-quently, choose a procedure that produces theleast possible damage and use the correct wheelor blade. Grinding also causes damage, with thedepth of damage decreasing as the abrasive sizedecreases. Materials respond differently to the

same size abrasive, so one cannot generalizeon removal depths. Removal rates also decreaseas the abrasive size decreases. Through an un-derstanding of these factors, good, reproduciblepreparation procedures, such as presented in thisguide, can be established for the materials be-ing prepared. Automation in specimenpreparation offers much more than reduced la-bor. Specimens properly prepared usingautomated devices consistently exhibit much bet-ter flatness, edge retention, relief control andfreedom from artifacts such as scratches, pull out,smearing and comet tailing, than specimens pre-pared manually.

Summary

109

REFERENCES

1. Metallography and Microstructures,Vol. 9, Metals Handbook, 9th ed., AmericanSociety for Metals, Metals Park, OH, 1985.

2. G. F. Vander Voort, Metallography:Principles and Practice, ASM International,Materials Park, OH, 1999.

3. L. E. Samuels, Metallographic Polishingby Mechanical Methods, 3rd Ed., AmericanSociety for Metals, Metals Park, OH, 1982.

4. J. A. Nelson, “Heating of MetallographicSpecimens Mounted in ‘Cold Setting’ Resins”,Practical Metallography, Vol. 7, 1970,pp. 188-191.

5. G. F. Vander Voort, “Trends in SpecimenPreparation,” Advanced Materials &Processes, Vol. 157, February 2000,pp. 45-49.

6. G. F. Vander Voort, “Metallography forEdge Retention in Surface Treated SinteredCarbides and Steels,” Industrial Heating,Vol. 67, March 2000, pp. 85-90.

7. S. L. Hoyt, Men of Metals, ASM, MetalsPark, OH, 1979, pp. 99-100.

8. J. Benedict, R. Anderson and S. J. Klepeis,“Recent Developments in the Use of the TripodPolisher for TEM Specimen Preparation,”Specimen Preparation for TransmissionElectron Microscopy of Materials – III,Materials Research Society, Pittsburgh, PA,1992, Vol. 254, pp. 121-140.

9. G. Petzow and V. Carle, MetallographicEtching, 2nd ed., ASM International, MaterialsPark, OH, 1999.

10. G. F. Vander Voort, “Grain Size Measure-ment,” Practical Applications of QuantitativeMetallography, STP 839, American Society forTesting and Materials, Philadelphia, 1984, pp.85-131.

11. G. F. Vander Voort, “Wetting Agents inMetallography,” Materials Characterization,Vol. 35, No. 2, September 1995, pp. 135-137.

12. A. Skidmore and L. Dillinger, “EtchingTechniques for Quantimet Evaluation,”Microstructures, Vol. 2, Aug./Sept. 1971,pp. 23-24.

13. G. F. Vander Voort, “Etching Techniquesfor Image Analysis,” Microstructural Science,Vol. 9, Elsevier North-Holland, NY, 1981,pp. 135-154.

14. G. F. Vander Voort, “Phase Identificationby Selective Etching,” Applied Metallography,Van Nostrand Reinhold Co., NY, 1986,pp. 1-19.

15. E. E. Stansbury, “Potentiostatic Etching,”Applied Metallography, Van NostrandReinhold Co., NY, 1986, pp. 21-39.

16. E. Beraha and B. Shpigler, ColorMetallography, American Society for Metals,Metals Park, OH, 1977.

17. G. F. Vander Voort, “Tint Etching,”Metal Progress, Vol. 127, March 1985,pp. 31-33, 36-38, 41.

18. E. Weck and E. Leistner, MetallographicInstructions for Colour Etching byImmersion, Part I: Klemm Colour Etching,Vol. 77, Deutscher Verlag für SchweisstechnikGmbH, Düsseldorf, 1982.

19. E. Weck and E. Leistner, MetallographicInstructions for Colour Etching byImmersion, Part II: Beraha Colour Etchantsand Their Different Variants, Vol. 77/II,Deutscher Verlag für Schweisstechnik GmbH,Düsseldorf, 1983.

20. E. Weck and E. Leistner, MetallographicInstructions for Colour Etching byImmersion, Part III: Nonferrous Metals,Cemented Carbides and Ferrous Metals,Nickel Base and Cobalt Base Alloys,Vol. 77/III, Deutscher Verlag fürSchweisstechnik GmbH, Düsseldorf, 1986.

21. P. Skocovsky, Colour Contrast inMetallographic Microscopy, Slovmetal,• ilina, 1993.

22. H. E. Bühler and H. P. Hougardy, Atlas ofInterference Layer Metallography, DeutscheGesellschaft für Metallkunde, Oberursel,Germany, 1980.

23. H. E. Bühler and I. Aydin, “Applications ofthe Interference Layer Method,” AppliedMetallography, Van Nostrand Reinhold Co.,NY, 1986, pp. 41-51.

24. G. F. Vander Voort, “Results of an ASTME-4 Round-Robin on the Precision and Bias ofMeasurements of Microindentation HardnessImpressions,” Factors that Affect thePrecision of Mechanical Tests, ASTM STP1025, ASTM, Philadelphia, 1989, pp. 3-39.

References

110

References

25. G. F. Vander Voort, “Operator Errors in theMeasurement of Microindentation Hardness,”Accreditation Practices for Inspections,Tests, and Laboratories, ASTM STP 1057,ASTM, Philadelphia, 1989, pp. 47-77.

26. G. F. Vander Voort and Gabriel M. Lucas,“Microindentation Hardness Testing,”Advanced Materials & Processes, Vol. 154,No. 3, September 1998, pp. 21-25.

27. R. L. Anderson, “Safety in MetallographicLaboratory,” Westinghouse ResearchLaboratory Science Paper, No. 65 – 1P30 –METLL – P2, March 29, 1965.

28. G. F. Vander Voort, Metallography:Principles and Practice, ASM International,Materials Park, OH., 1999, pp. 148-159.

29. R. C. Nester and G. F. Vander Voort,“Safety in the Metallographic Laboratory,”Standardization News, Vol. 20, May 1992,pp. 34-39.

30. P. A. Jacquet, “The Safe Use of Perchloric-Acetic Electropolishing Baths,” Met. Finish,Vol. 47, 1949, pp. 62-69.

31. S. M. Comas, R. Gonzalez Palacin and D.Vassallo, “Hazards Associated with PerchloricAcid-Butylcellosolve Polishing Solutions,”Metallography, Vol. 7, 1974, pp. 47-57.

32. N. I. Sax, Dangerous Properties ofIndustrial Materials, 5th ed., Van NostrandReinhold Co., Inc., N.Y., 1979.

33. A.E. Dawkins, “Chromic Acid-AceticAnhydride ‘Explosion,’” J. Iron and SteelInstitute, Vol. 182, 1956, p. 388.

Selected Bibliography of Books onLaboratory SafetyASTM E-2014, “Standard Guide on Metallo-graphic Laboratory Safety”.

N. Van Houten, Laboratory Safety Standardsfor Industry, Research, Academe, Sci. Tech.Pubs., Lake Isabella, CA. 1990.

E. Gershey, A. Wilkerson and E. Party,Laboratory Safety in Practice, Van NostrandReinhold, N.Y., 1991.

A. A. Fuscaldo et al. (eds), Laboratory Safety:Theory & Practice, Academic Press, SanDiego, CA. 1980.

S. B. Pal (ed.), Handbook of LaboratoryHealth & Safety Measures, Kluwer Academic,Norwell, MA., 1985.

Norman V. Steere (ed.), Handbook ofLaboratory Safety, 2nd ed., CRC Press,Boca Raton, FL., 1971.

A. Keith Furr (ed.), Handbook of LaboratorySafety, 3rd ed., CRC Press, Boca Raton, FL.,1989.

L. J. Diberardinis et al., Guidelines forLaboratory Design: Health & SafetyConsiderations, J. Wiley & Sons, N.Y., 1987.

S. R. Rayburn, The Foundations ofLaboratory Safety, Brock-Springer Series inContemporary Bioscience, Springer-Verlag,N.Y., 1989.

C. A. Kelsey and A. F. Gardner (eds.),Radiation Safety for LaboratoryTechnicians, Warren H. Green, Inc.,St. Louis, MO., 1983.

N. I. Sax, Dangerous Properties of IndustrialMaterials, 5th ed., Van Nostrand Reinhold Co.,N.Y., 1979.

Prudent Practices for Handling HazardousChemicals in Laboratories, NationalAcademy Press, Washington, D.C., 1981.

Prudent Practices for Disposal ofChemicals from Laboratories, NationalAcademy Press, Washington, D.C., 1983.

N. Proctor and J. Hughes, Chemical Hazardsin the Workplace, J. B. Lippincott Co.,Philadelphia, 1978.

F. A. Patty (ed.), Industrial Hygiene andToxicology, Volume II – Toxicology, 3rd ed.,Wiley-Interscience, N.Y., 1980.

L. Bretherick, Handbook of Reactive Chemi-cal Hazards, 2nd ed., Butterworths, London,1979.

R. J. Lewis, Sr. (ed.), Hawley’s CondensedChemical Dictionary, 12th ed., Van NostrandReinhold, New York, 1993.

Useful Web sites on laboratory safety:http://ull.chemistry.uakron.edu/erd/chemicals/1-500/0392.html

http://www.hhmi.org/research/labsafe/lcss/lcss.html

http://www.cdc.gov/niosh/ipcsnengnengsyn.html

http://www.state.nj.us/health/eoh/rtkwebrtkhsfs.htm

111

APPENDICES

Appendices

Atomic Atomic Density Melting CrystalGroup Element Symbol Number Weight g/cc@20°C Point,°C Structure

Alkaline Beryllium Be 4 9.015 1.848 1280 cphEarth, IIA Magnesium Mg 12 24.312 1.74 650 cph

Coinage Copper Cu 29 63.545 8.94 1083 fccMetals, IB Silver Ag 47 107.87 10.49 961.9 fcc

Gold Au 79 197 19.32 1064.4 fcc

Group IIB Zinc Zn 30 65.38 7.1 419.57 cphCadmium Cd 48 112.4 8.65 320.9 cph

Group III Aluminum Al 13 26.98 2.7 660 fcc

Group IVA Tin Sn 50 118.69 7.29 232 bctLead Pb 82 207.19 11.35 327.5 fcc

Group IVB Titanium Ti 22 47.9 4.507 1670 cphZirconium Zr 40 91.22 6.44 1853 cphHafnium Hf 72 178.49 13.21 2225 cph

Group V Antimony Sb 51 121.75 6.79 630.5 rhombohedralBismuth Bi 83 208.98 9.78 271.3 rhombohedral

Group VB Vanadium V 23 50.94 6.1 1895 bccNiobium/ Nb/

Columbium Cb 41 92.91 8.66 2468 bccTantalum Ta 73 180.95 16.63 2996 bcc

Group VIB Chromium Cr 24 52 7.14 1867 bccMolybdenum Mo 42 95.94 9.01 2615 bcc

Tungsten W 74 183.85 19.3 3410 bcc

Group VIIB Manganese Mn 25 54.94 7.3 1244 complex cubicRhenium Re 75 186.22 21.04 3175 cph

Group VIII Iron Fe 26 55.85 7.874 1536 bccCobalt Co 27 58.93 8.832 1495 cphNickel Ni 28 58.71 8.8 1452 fcc

Platinum Ruthenium Ru 44 101.1 12.1 2310 cphMetals, Rhodium Rh 45 102.91 12.44 1963 fccGroup VIII Palladium Pd 46 106.4 12.16 1551 fcc

Osmium Os 76 190.2 22.48 3015-3075 cphIridium Ir 77 192.2 22.42 2410 fcc

Platinum Pt 78 195.09 21.37 1772 fcc

Classification of Metals Using Common Characteristics and the Periodic Table

Transformation Temperatures for Allotropic Metals Listed Above

Element Transformation Temperature (C°/F°)

Beryllium (Be) α↔β 1256/2293

Cobalt (Co) ε↔α 427/801

Hafnium (Hf) α↔β 1742/3168

Iron (Fe) Curie 770/1418α↔γ 912/1674γ↔δ 1394/2541

Manganese (Mn) α↔β 707/1305β↔γ 1088/1990γ↔δ 1139/2082

Nickel (Ni) Curie 358/676

Tin (Sn) α↔β 13.0/55.4

Titanium (Ti) α↔β 883/1621

Zirconium (Zr) α↔β 872/1602

Note: The Curie temperature is the temperature where a magnetic material losses its magnetism upon heating above and regains it uponcooling below.

112

Appendices

Abrasive Cutting Troubleshooting Guides

Problem Possible Cause Suggested Remedy

Burning (bluish discoloration) Overheated specimen Increase coolant flow rate;reduce cutting pressure; select awheel with softer bonding (fasterbreakdown)

Rapid wheel wear Wheel bonding breaks down Select a wheel with harder bonding;too rapidly reduce cutting pressure

Frequent wheel breakage Uneven coolant distribution; Adjust coolant flow to be even onloose specimen fixturing; both sides of the wheel; clamp theabrupt contact with specimen specimen more tightly; start cut

contact carefully

Resistance to cutting Slow wheel bond breakdown Select a wheel with softer bonding;use a “pulse” cutting mode; usecutter with oscillating motion orwith minimal area of contactcutting ability

Stalled wheel Inadequate cutter capacity; Use cutter with greater horsepower;pinched blade due to movement tighten the clamp on one side lessof specimen than on the other side; reduce

pressure or feed rate; use acutter with oscillating motion orwith minimal area of contactcutting ability

* Minimal Area of Contact Cutting like on the Delta orbital abrasive cutter

Pointed:Wheel too hard

Chisel:Nonuniform coolant flow

Glazed Surface:Wrong wheel

Normal:For structural or

medium wall tubing

Normal:For solid materials

Normal:For light tubing or

other thin wall section

Abrasive Wheel Troubleshooting Guides

113

Appendices

Compression (Hot) Mounting Troubleshooting Guide

Thermosetting Resins (Epoxies, Diallyl Phthalates, and Phenolics)

Defect Probable Cause Suggested Remedy

Radial splits Specimen cross sectional area Use a larger molde size; decreaseor cracks too large; specimen with the specimen size; bevel sharp

sharp corners corners, if possible

Shrinkage gaps Specimen surfaces dirty; specimen Clean and dry specimens carefully;cooled quickly after polymerization; after polymerization, cool underwrong resin used pressure to near ambient; use

Epomet G or Epomet F resins

Circumferential cracks Resin contained moisture Keep resins dry during storage;keep containers closed when notusing; dry resin by baking at38-49°C (100-120°F)

Bulging or soft mount Inadequate curing Increase polymerization time(polymerization) time and pressure

Mount appears Time at temperature too short; Increse polymerization time,grainy and unfused temperature for polymerization temperature and pressure

too low; molding pressure too low

Thermoplastic Resins (Acrylics)


Cottonball Incomplete polymerization of resin; Use less resin; use longernot enough time at temperaure; heating and cooling periods;cooling too fast use controlled linear cooling

Crazing Relief of internal stresses Cool mount to a lower temperatureupon ejection of mount before ejection; use controlled

linear cooling

Comparison of Six Edge Retention Compression Mounting Compounds

Micrographs showing the as-forged surface of a hardened modified 5130 alloy steel part mounted using avariety of resins showing different degrees of edge retention. The specimens were polished simultaneously inthe same holder and were etched with 2% nital. The magnification bars are 20 µm long. Best results wereobtained with Epomet, Probemet and Epoxicure resin with the Conductive Filler particles.

Phenocure phenolic resin Epomet thermosetting epoxyresin

Probemet Cu-filled conductiveresin

Konductomet C-filled conductiveresin

Epoxicure cast epoxy resin withConductive Filler particles

Sampl-Kwick cast acrylic resinwith Conductive Filler particles

114

Appendices

Name Typical Cure Time Uses and Restrictions

Sampl-Kwick Acrylic Resin 5-8 minutes Least expensive resin; used for mounting printedcircuit boards and polymers; good for x-raydiffraction work (no peak interferences); highshrinkage (not recommended for edge retention)

Varidure Acrylic Resin 10 minutes Inexpensive; (good edge rentention for acrylic);used with many metals when the edge detail isnot important

Epo-Kwick Epoxy Resin 90 minutes Fast curing epoxy; more shrinkage than otherepoxies; viscosity 250-400 cps at 25 °C; notrecommended for vacuum infiltration of voidsdue to fast curing time; physically adheres tospecimen; gives acceptable edge retention; goodfor most specimens

Epo-Color Epoxy Resin 90 minutes Provides color contrast between mount andspecimen that can be helpful when studyingedges; viscosity 400-700 cps at 25 °C; goodfor most specimens; but not recommended forvacuum impregnation

Epoxicure Epoxy Resin 6 hours General purpose epoxy; viscosity 400-600 cps at25 °C, good adherence to specimen; can be used tovacuum impregnate voids (viscosity can be reducedby warming to 50 °C); good for heat-sensitivespecimens (very low exotherm during polymerization)

Epo-Thin Epoxy Resin 9 hours Lowest viscosity (can be further reduced by warmingresin to 50 °C) and best penetration of voids duringvacuum impregnation; good adherence to specimen;good for heat-sensitive specimens (lowest exothermduring polymerization)

Curing time is not a precise quantity as it can be influenced by the nature of the mold (how well, or how poorly, the exothermic heat of polymerization isextracted), by the volume of the cast resin used, by the volume and nature of the specimen being encapsulated, by the temperature of the mix, by roomtemperature and air circulation and by the age of the resin and hardener (i.e., curing time is longer for products beyond their normal shelf life). The valueslisted are typical for metallic specimens cast in 1.25-inch (30-mm) diameter Sampl-Kup molds at 21 °C (70 °F) using fresh resin and hardener.

Castable Mounting Resin Selection Guide

Greek Letter Name English Equivalent

Α, α Alpha A, a

Β, β Beta B, b

Γ, γ Gamma G, g

∆, δ Delta D, d

Ε, ε Epsilon E, e

Ζ, ζ Zeta Z, z

Η,η Eta E, e

Θ, θ Theta Th

Ι, ι Iota I, i

Κ, κ Kappa K, k

Λ, λ Lambda L, l

Μ, µ Mu M, m

Greek Alphabet and its English Equivalents

Greek Letter Name English Equivalent

Ν, ν Nu N, n

Ξ, ξ Xi X, x

Ο, ο Omicron O, o

Π, π Pi P, p

Ρ, ρ Rho R, r

Σ, σ Sigma S, s

Τ, τ Tau T, t

Υ, υ Upsilon U, u

Φ, ϕ Phi Ph

Χ, χ Chi Ch

Ψ, ψ Psi Ps

Ω, ω Omega O, o

115

Appendices

Cast Resin (“Cold” Mounting) Troubleshooting Guide


Non-Curing Expired resin and/or hardener; Measure weights of resin andWrong resin – to – hardener ratio hardener; after prescribed time, and

it has not hardener, heat the mold,liquid and specimen to 50° C(120° F) to force curing; check agebefore using resin and hardener

Slow Curing Incomplete mixing; too little hardener Mix thoroughly; measure weightsused; curing at too low a temperature of resin and hardener; place in

warm area*

Rapid Curing Too much hardener; curing at too high Measure weights of resin anda temperature; mold size too large hardener; mix smaller quantities;(excessive resin volume) increase air circulation; place in

a cooler place

Gas Tunnels Excessive hardener used; mold size Measure weights of resin andtoo large (excessive resin volume); hardener; increase air circulation;use of a low thermal conductivity mold use higher thermal conductivity

molds

Mount Stuck in Mold Did not use mold release agent Apply Silicon Mold Release orRelease Agent to interior surfacesof mold

Solvent Softening Poor resistance to solvents; Heat the mold, liquid and specimenmount may not be fully cured to 50° C (120° F) to force curing

Excessive Shrinkage Polymerization is too fast; too much Use an epoxy rather than anhardener used; wrong resin selected acrylic; use a slow curing epoxy;

add Flat-Edge Filler particles;use the correct resin – to –hardener ratio

* Placing in an oven at 120°F (50°C) for up to two hours may accelerate the curing of epoxies

116

Appendices

Name Characteristics Applications

Micropolish 1-, 0.3- and 0.05-µm alumina powder Inexpensive general-purpose abrasivealumina powder for polishing most minerals and metals

Micropolish II 1-, 0.3- and 0.05-µm deagglomerated Higher quality alumina suitable for polishingdeagglomerated alumina powders and 5-, 1-, 0.3- and most minerals and metals; produces aalumina powder 0.05-µm deagglomerated alumina better surface finish than standard aluminaand suspensions suspensions abrasives of the same size

Masterprep 0.05-µm sol-gel alumina suspension Unique seeded-gel alumina produces betteragglomerate-free finishes than calcined alumina abrasives;seeded-gel alumina excellent with minerals, metals, carbides,suspension PCBs. Good for percious metals.

Mastermet 0.06-µm amorphous silica colloidal Chemo-mechanical action producescolloidal silica suspension with a pH of ~10 producing a excellent surfaces for metals, minerals,suspension chemo-mechanical polishing action polymers, PCBs, electronic devices,

etc., especially soft metals, but not forprecious metals

Mastermet 2 0.02-µm non-crystallizing amorphous Non-crystallizing version of colloidal silicanon-crystallizing silica colloidal suspension with a pH yields similar results as Mastermet butcolloidal silica of ~10.5 producing chemo-mechanical without the danger of crystallized silicasuspension polishing action particle contamination

Masterpolish Proprietary viscous blend of high Proprietary blend has little water and canblended alumina purity 0.05-µm alumina and colloidal be used with materials that are sensitiveand colloidal silica with a slightly basic pH (can to water, such as Mg and most of its alloys;silica suspension be diluted with water) good for most Fe-, Ni- or Co-based alloys

and MMCs

Masterpolish 2 Proprietary 0.06-µm abrasive suspension Proprietary blend designed for final polishingproprietary with a pH of ~10 for chemo-mechanical of ceramic materials, nitrides, carbideschemo-mechanical polishing action and compositessuspension

Guide to Fine Polishing Abrasives

117

Appendices

Preparation Troubleshooting Guide

Troubleshooting Definitions

Comet Tails: Dull, broad depressed lines emanating from a hard particle or from a hole in a specimen in a pattern that resemblesthe tail of a comet. The hard particles may be nonmetallic inclusions or nitrides. This appears to be a material-specific phenomenonas repeated efforts to produce this defect in randomly chosen specimens fail. Historically, with manually-prepared specimens, itwas claimed to be caused by unidirectional grinding; however, repeated efforts to create comet tails with only unidirectionalgrinding fail. Comet tails have been observed on specimens prepared with automated machines using only complementary grindingand polishing. In a specimen prone to comet tailing, using contra rotation in the final step will enhance the problem compared tocomplementary rotation.

Edge Rounding: The edge or edges of a specimen are abraded at a faster rate than the interior leading to a change inflatness at the edge such that the edge cannot be brought into focus along with the interior with the light microscope (thiscondition is more critical as the magnification is raised due to the higher numerical aperture of the objective which reduces thedepth of focus).

Embedding: Hard abrasive particles that become fixed in the surface of the softer specimen. This is a common problem withthe low melting point alloys (e.g., Pb, Sn, Bi, Cd, Zn), aluminum (mainly fine diamond), and precious metals, but has beenobserved with refractory metals such as titanium.

Pull Out: Removal of second phase particles (either softer or harder than the matrix) during preparation. Pull out may beenhanced if the interface between the particle and matrix is weak or if the particle is particularly brittle.

Relief: Excessive height differences between second-phase particles and the matrix due to differences in grinding andpolishing rates between the matrix and the particles. Soft particles abrade faster than the matrix and are recessed below thematrix while the reverse is observed for hard particles.

Scratches: A linear cut along the surface of a specimen due to contact with a correctly oriented abrasive particle; a groovecaused by the abrasion process. The width and depth of the cut is a function of the size of the abrasive particle, its angle tothe specimen surface, the applied pressure, and other factors. Scratches become a problem when they are excessivelycoarse for the abrasive size used, indicating a problem with the preparation method. Very fine scratches may not be a problemin routine examination for grain size or inclusions but could be a problem in failure analysis if critical fine detail is obscured.

Smear: Matrix flow over voids, cracks, or second phase particles that make detection of these features and measurement oftheir extent difficult or impossible.

Stain: A contamination residue on the surface of a specimen, that may form around second-phase particles due tointeractions between the specimen and abrasives and/or lubricants, or may form on the matrix due to inadequate cleaning ordrying or may form due to interactions between the specimen and solvents after preparation or etching.

118

Appendices


Comet Tails Poorly bonded very hard phase in softer Use hard, napless cloths; reduce applied pressure;matrix; or, pores in matrix results in reduce step times; impregnate pores with epoxy orunidirectional grooves emanating from wax; use complementary rotation. In manualparticles or hole. preparation, avoid unidirectional grinding.

Edge Rounding Shrinkage gaps between specimen and Avoid gaps by cooling thermosetting mount undermount are main problem and, when pressure afer polymerization. Use Epomet resinpresent, avoid napped cloths and long (least prone to shrinkage gaps); of castable resins,polishing times. epoxy is best (adding Flat-Edge Filler particles to

epoxy improves edge retention); use central forcerather than individual force; use the BuehlerHerculesrigid grinding discs; use napless cloths.

Embedding Hard abrasive tends to embed in soft Coat SiC abrasive paper with paraffin wax (candlewaxmetals; finer particles are more likely is best, soaps are also effective); beeswax is lessto embed than larger particles. effective; reduce applied pressure, rpm and times;

avoid finer SiC abrasive papers; avoid using diamondslurries with fine diamond sizes (for ≤ 3-µm diamond,paste embeds less than slurries). SiC embedded insoft metals (e.g., Pb) can be removed by polishingwith alumina slurries.

Pull Outs Excessive grinding time on worn SiC Grind no more than 60 seconds on a sheet of SiCpaper; excessive polishing time on paper; use napless cloths; reduce appliednapped cloths; excessive pressure; pressure; use proper degree of lubrication.and inadequate lubrication lead to pullout of second-phase particles.

Relief Use of napped cloths, long polishing times, Use napless cloths, higher pressures and shorterand low pressure create height differences times. If relief is observed using contra rotation inbetween matrix and second-phase particles. last step, repeat the last step using complementary

rotation.

Scratches Contamination of grinding or polishing Maintain clean operating conditions, washing handssurfaces by coarser abrasives; pull out and equipment between steps; for cracked or porousof hard particles from matrix or broken specimens, use ultrasonic cleaning after each steppieces of brittle materials; or (especially polishing);execute each step thoroughlyinadequate grinding or polishing times (avoid short cutting methods);when usingthe Apexand pressures can leave a heavy magnetic disc system, store the discs in a cleanscratch pattern. drawer or cabinet and scrub surfaces if contamination

is believed to have occurred.

Smear Flow of softer metals may be caused by Use proper degree of lubrication, lower pressuresinadequate lubrication, excessive and rpms; use a medium napped cloth on final step;pressures or rpms. lightly etch the specimen after final polishing and

repeat the last step; use the vibratory polisher toremove smeared surface metal.

Stain Stains around second-phase particles can Some staining issues are unique to the specimen butbe induced by interactions between most result from using impure tap water (switch toabrasive and particle, perhaps affected distilled water; sometimes de-ionized water is needed)by water quality; or they can occur due or inadequate drying. When using colloidal silica,to inadequate cleaning, or improper which can be hard to remove, direct the water jetsolvents used after the abrasive step. onto the cloth and wash the specimens (and cloth) for

the last 6-10 seconds of the polishing cycle. Thismakes subsequent cleaning easy. Otherwise, scrubthe surface with water containing a detergent solutionusing cotton. Then, rinse with alcohol and blow drywith hot air, such as with the Torramet dryer.Compressed air systems can contain impurities, suchas oils, that can stain the surface.

Preparation Troubleshooting Guide Cont.

119

ASTM Metallography Standards

ASTM METALLOGRAPHYSTANDARDSA 247: Visual Classification of Graphite in theMicrostructure of Cast Iron

A 892: Defining and Rating the Microstructureof High Carbon Bearing Steels

B 390: Evaluating Apparent Grain Size andDistribution of Cemented Tungsten Carbides

B 588: Measurement of the Thickness ofTransparent or Opaque Coatings by Double-Beam Interference Microscope Technique

B 657: Metallographic Determination ofMicrostructure in Cemented Tungsten Carbide

B 681: Measurement of Thickness of AnodicCoatings on Aluminum and of Other Transpar-ent Coatings on Opaque Surfaces Using theLight-Section Microscope

B 748: Measurement of the Thickness ofMetallic Coatings by Measurement of CrossSection with a Scanning Electron Microscope

B 795: Determining the Percentage of Alloyedor Unalloyed Iron Contamination Present inPowder Forged Steel Parts

B 796: Nonmetallic Inclusion Level of PowderForged Steel Parts

B 847: Measurement of Metal and OxideCoating Thickness by MicroscopicalExamination of a Cross section

C 664: Thickness of Diffusion Coating

E 3: Preparation of Metallographic Specimens

E 7: Standard Terminology Relating toMetallography

E 45 Determining the Inclusion Contentof Steel

E 82: Determining the Orientation of aMetal Crystal

E 112: Determining Average Grain Size

E 340: Macroetching Metals and Alloys

E 381: Macroetch Testing Steel Bars, Billets,Blooms, and Forgings

E 384: Microindentation Hardness of Materials

E 407: Microetching Metals and Alloys

E 562: Determining Volume Fraction bySystematic Manual Point Count

E 766: Calibrating the Magnification of aScanning Electron Microscope

E 768: Preparing and Evaluating Specimensfor Automatic Inclusion Assessment of Steel

E 807: Metallographic Laboratory Evaluation

E 883: Reflected-Light Photomicrography

E 930: Estimating the Largest Grain Observedin a Metallographic Section (ALA Grain Size)

E 975: X-Ray Determination of RetainedAustenite in Steel with Near RandomCrystallographic Orientation

E 986: Scanning Electron Microscope BeamSize Characterization

E 1077: Estimating the Depth ofDecarburization of Steel Specimens

E 1122: Obtaining JK Inclusion Ratings UsingAutomatic Image Analysis

E 1180: Preparing Sulfur Prints forMacrostructural Examination

E 1181: Characterizing Duplex Grain Sizes

E 1245: Determining the Inclusion or Second-Phase Constituent Content of Metals byAutomatic Image Analysis

E 1268: Assessing the Degree of Banding orOrientation of Microstructures

E 1351: Production and Evaluation of FieldMetallographic Replicas

E 1382: Determining Average Grain Size UsingSemiautomatic and Automatic Image Analysis

E 1508: Quantitative Analysis by Energy-Dispersive Spectroscopy

E 1558: Electrolytic Polishing of MetallographicSpecimens

E 1920: Metallographic Preparation of ThermalSpray Coatings

E 1951: Calibrating Reticles and LightMicroscope Magnifications

E 2014: Metallographic Laboratory Safety

E 2015: Preparation of Plastics and PolymericSpecimens for Microstructural Examination

F 1854: Stereological Evaluation of PorousCoatings on Medical Implants

120

ASTM/ISO Standards

ASTM HARDNESS TESTINGSTANDARDS

B 578: Microhardness of ElectroplatedCoatings

B 721: Microhardness and Case Depth ofPowder Metallurgy Parts

C 730: Knoop Indentation Hardness of Glass

C 849: Knoop Indentation Hardness ofCeramic Whitewares

C 1326: Knoop Indentation Hardness ofAdvanced Ceramics

C 1327: Vickers Indentation Hardness ofAdvanced Ceramics

E 10: Brinell Hardness of Metallic Materials

E 18: Rockwell Hardness and RockwellSuperficial Hardness of Metallic Materials

E 92: Vickers Hardness of Metallic Materials

E 140: Standard Hardness Conversion Tablesfor Metals

E 448: Scleroscope Hardness Testing ofMetallic Materials

E 1842: Macro-Rockwell Hardness Testing ofMetallic Materials

ISO METALLOGRAPHYSTANDARDS

ISO 643: Steels – Micrographic Determinationof the Ferritic or Austenitic Grain Size

ISO 945: Cast Iron: Designation ofMicrostructure of Graphite

ISO 1083: Spheroidal Graphite CastIron – Classification

ISO 1463: Metallic and Oxide Coatings –Measurement of Coating Thickness –Microscopical Method

ISO 2624: Copper and Copper Alloys –Estimation of Average Grain Size

ISO 2639: Steel – Determination andVerification of the Effective Depth ofCarburized and Hardened Cases

ISO 3754: Steel – Determination of EffectiveDepth of Hardening After Flame or InductionHardening

ISO 3763: Wrought Steels – MacroscopicMethods for Assessing the Content ofNon-Metallic Inclusions

ISO 3887: Steel, Non-Alloy and Low Alloy –Determination of Depth of Decarburization

ISO 4499: Hardmetals: MetallographicDetermination of Microstructure

ISO 4524-1: Metallic Coatings – TestMethods for Electrodeposited Gold and GoldAlloy Coatings – Part 1: Determination ofCoating Thickness

ISO 4964: Steel – Hardness Conversions

ISO 4967: Steel – Determination of Content ofNon-Metallic Inclusions – MicrographicMethod Using Standard Diagrams

ISO 4968: Steel – Macrographic Examinationby Sulphur Print (Baumann Method)

ISO 4970: Steel – Determination of Totalor Effective Thickness of Thin Surface-Hardened Layers

ISO 5949: Tool Steels and Bearing Steels –Micrographic Method for Assessing theDistribution of Carbides Using ReferencePhotomicrographs

ISO 9042: Steels – Manual Point CountingMethod for Statistically Estimating the VolumeFraction of a Constituent with a Point Grid

ISO 9220: Metallic Coatings – Measurement ofCoating Thickness – Scanning ElectronMicroscope Method

ISO 14250: Steel – MetallographicCharacterization of Duplex Grain Sizeand Distribution

ISO HARDNESS STANDARDS

ISO 146: Metallic Materials – HardnessTest – Verification of Vickers Hardness TestingMachines HV 0,2 to HV 100

ISO 146-2: Metallic Materials – Hardness Test– Verification of Vickers Hardness TestingMachines – Part 2: Less Than HV 0,2

ISO 640: Metallic Materials – Hardness Test –Calibration of Standardized Blocks to be Usedfor Vickers Hardness Testing Machines HV 0,2to HV 100

121

ISO Standards

ISO 640-2: Metallic Materials – Hardness Test– Calibration of Standardized Blocks to beUsed for Vickers Hardness Testing Machines –Part 2: Less than HV 0,2

ISO 674: Metallic Materials – Hardness Test –Calibration of Standardized Blocks to be Usedfor Rockwell Hardness Testing Machines(Scales A, B, C, D, E, F, G, H, K)

ISO 716: Metallic Materials – Hardness Test –Verification of Rockwell Hardness TestingMachines (Scales A, B, C, D, E, F, G, H, K)

ISO 1024: Metallic Materials – Hardness Test –Rockwell Superficial Test (Scales 15N, 30N,45N, 15T, 30T and 45T)

ISO 1355: Metallic Materials – Hardness Test –Calibration of Standardized Blocks to be Usedfor Rockwell Superficial Hardness TestingMachines (Scales 15N, 30N, 45N, 15T, 30Tand 45T)

ISO 4516: Metallic and Related Coatings –Vickers and Knoop Microhardness Tests

ISO 4545: Metallic Materials – Hardness Test –Knoop Test

ISO 4546: Metallic Materials – HardnessTest – Verification of Knoop HardnessTesting Machines

ISO 4547: Metallic Materials – Hardness Test –Calibration of Standardized Blocks to be Usedfor Knoop Hardness Testing Machines

ISO 6441-1: Paints and Varnishes – Determi-nation of Microindentation Harness – Part 1:Knoop Hardness by Measurement ofIndentation Length

ISO 6441-1: Paints and Varnishes – Determi-nation of Microindentation Harness – Part 2:Knoop Hardness by Measurement of Indenta-tion Depth Under Load

ISO 6507-1: Metallic Materials – Hardness Test– Vickers Test – Part I: HV 5 to HV 100

ISO 6507-2: Metallic Materials – HardnessTest – Vickers Test – Part 2: HV 0, 2 to LessThan HV 5

ISO 6508: Metallic Materials – Hardness Test –Rockwell Test (Scales A, B, C, D, E, F, G, H, K)

ISO 9385: Glass and Glass-Ceramics – KnoopHardness Test

ISO 10250: Metallic Materials – HardnessTesting – Tables of Knoop Hardness Values forUse in Tests Made on Flat Surfaces

ISO 14271: Vickers Hardness Testing ofResistance Spot, Projection and Seam Welds(low load and microhardness)

ISO 674: Metallic Materials – Hardness Test –Calibration of Standardized Blocks to be Usedfor Rockwell Hardness Testing Machines(Scales A, B, C, D, E, F, G, H, K)

ISO 716: Metallic Materials – Hardness Test –Verification of Rockwell Hardness TestingMachines (Scales A, B, C, D, E, F, G, H, K)

ISO 1024: Metallic Materials – Hardness Test –Rockwell Superficial Test (Scales 15N, 30N,45N, 15T, 30T and 45T)

ISO 1355: Metallic Materials – Hardness Test –Calibration of Standardized Blocks to be Usedfor Rockwell Superficial Hardness TestingMachines (Scales 15N, 30N, 45N, 15T, 30Tand 45T)

ISO 4545: Metallic Materials – Hardness Test –Knoop Test

ISO 4546: Metallic Materials – HardnessTest – Verification of Knoop HardnessTesting Machines

ISO 4547: Metallic Materials – Hardness Test –Calibration of Standardized Blocks to be Usedfor Knoop Hardness Testing Machines

ISO 6507/1: Metallic Materials – Hardness Test– Vickers Test – Part I: HV 5 to HV 100

ISO 6507/2: Metallic Materials – HardnessTest – Vickers Test – Part 2: HV 0,2 to LessThan HV 5

ISO 6508: Metallic Materials – Hardness Test –Rockwell Test (Scales A, B, C, D, E, F, G, H, K)

ISO 10250: Metallic Materials – HardnessTesting – Tables of Knoop Hardness Values forUse in Tests Made on Flat Surfaces

108

Other National Standards

OTHER NATIONALSTANDARDS

FranceNF A04-10: Determination of the Ferritic andAustenitic Grain Size of Steels

GermanyDIN 50150: Testing of Steel and Cast Steel;Conversion Table for Vickers Hardness,Brinell Hardness, Rockwell Hardness andTensile Strength

DIN 50192: Determination of the Depth ofDecarburization

DIN 50600: Testing of Metallic Materials;Metallographic Micrographs; Picture Scalesand Formats

DIN 50601: Metallographic Examination;Determination of the Ferritic or AusteniticGrain Size

DIN 50602: Metallographic Examination;Microscopic Examination of Special SteelsUsing Standard Diagrams to Assess theContent of Non-Metallic Inclusions

SEP 1510: Microscopic Test of Steels for GrainSize by Comparison with Standard Charts

SEP 1570: Microscopical Examination ofSpecial Steels for Non-Metallic InclusionsUsing Standard Micrograph Charts

SEP 1572: Microscopic Testing of Free-CuttingSteels for Non-Metallic Sulphide Inclusions byMeans of a Series of Pictures

ItalyUNI 3137: Extraction and Preparation ofSamples

UNI 3138: Macrographic Analysis

UNI 3245: Microscopic Examination of FerrousMaterials - Determination of Austenitic orFerritic Grain Size of Plain Carbon and Low-Alloy Steels

UNI 4227: Determination of MetallographicStructures

UNI 4389: Nonferrous Metals and Their Alloys:Determination of the Dimension of TheirCrystal Grains

JapanJIS B 7725: Vickers Hardness – Verification ofTesting Machines

JIS B 7734: Knoop Hardness Test –Verification of Testing Machines

JIS B 7735: Vickers Hardness Test –Calibration of the Reference Blocks

JIS G 0551: Methods of Austenite Grain SizeTest for Steel

JIS G 0552: Method of Ferrite Grain Size Testfor Steel

JIS G 0553: Macrostructure Detecting Methodfor Steel, Edition 1

JIS H 0501: Methods for EstimatingAverage Grain Size of Wrought Copper andCopper Alloys

JIS R 1610: Testing Method for VickersHardness of High Performance Ceramics

JIS R 1623: Test Method for Vickers Hardnessof Fine Ceramics at Elevated Temperatures

JIS Z 2244: Vickers Hardness Test –Test Method

JIS Z 2251: Knoop Hardness Test –Test Method

JIS Z 2252: Test Methods for Vickers Hardnessat Elevated Temperatures

PolandPN-57/H-04501: Macroscopic Examination ofSteel. The Deep Etching Test.

PN-61/H-04502: Reagents for MacrostructureTests of Iron Alloys

PN-61/H-04503: Reagents for MicrostructureTests of Iron Alloys

PN-63/H-04504: Microstructure Testing ofSteel Products. Iron Carbide. Ghost Structure.Widmanstätten’s Structure.

PN-66/H-04505: Microstructure of SteelProducts. Templates and Evaluation.

PN-75/H-04512: Nonferrous Metals. Reagentsfor Revealing Microstructure.

PN-75/H-04661: Gray, Spheroidal Graphiteand Malleable Cast Iron. MetallographicExamination. Evaluation of Microstructure

109

Other National Standards

Poland (continued)PN-76/H-04660: Cast Steel and Iron. Micro-scopic Examination. Sampling and Preparationof Test Pieces.

PN-84/H-04507/01: Metals. MetallographicTesting of Grain Size. Microscopic Methods forDetermination of Grain Size.

PN-84/H-04507/02: Metals. MetallographicTesting of Grain Size. Methods of Revealingthe Prior-Austenitic Grains in Non-AusteniticSteels.

PN-84/H-04507/03: Metals. MetallographicTesting of Grain Size. Macroscopic Method ofRevealing the Prior-Austenitic Grain Size bythe Fracture Method.

PN-84/H-04507/04: Metals. MetallographicTesting of Grain Size. A Method of Testing forOverheating of Steel.

PN-87/H-04514: Steel, Cast Steel, Cast Iron.Macrostructure Examination. Baumann’s Test.

RussiaGOST 801: Standard for Ball and RollerBearing Steel

GOST 1778: Metallographic Methods ofDetermination of Nonmetallic Inclusions

GOST 5639: Grain Size Determination

SwedenSIS 11 11 01: Estimating the Average GrainSize of Metals

SIS 11 11 02: Estimating the Austenitic GrainSize of Ferritic and Martensitic Steels

SIS 11 11 11: Methods for Assessing theSlag Inclusion Content in Steel:Microscopic Methods

SIS 11 11 14: Determination of SlagInclusions – Microscopic Methods –Manual Representation

SIS 11 11 16: Steel – Method for Estimation ofthe Content of Non-Metallic Inclusions –Microscopic Methods – Jernkontoret’sInclusion Chart II for the Assessment ofNon-Metallic Inclusions

SIS 11 03 40: Hardness Test – Vickers TestHV 0, 2 to HV 100 – Direct Verification ofTesting Machines

SIS 11 03 41: Hardness Test – Vickers Test HV0,2 to HV 100 – Indirect Verification of TestingMachines Using Standardized Blocks

SIS 11 03 42: Hardness Test – Vickers TestHV 0,2 to HV 100 – Direct Verification ofStandardizing Machine for Calibration ofStandardized Blocks

SIS 11 03 43: Hardness Test – VickersTest HV 0,2 to HV 100 – Calibration ofStandardized Blocks

SIS 11 25 16: Metallic Materials – HardnessTest – Vickers Test HV 5 to HV 100

SIS 11 25 17: Metallic Materials – HardnessTest – Vickers Test HV 0,2 to Less Than HV 5

SIS 11 70 20: Determination of the Depth ofDecarburization in Steel

United KingdomBS 860: Tables for Comparison of HardnessScales.

BS 4490: Methods for Micrographic Determina-tion of the Grain Size of Steel

BS 5710: Macroscopic Assessment of theNon-Metallic Inclusion Content of WroughtSteels

BS 6285: Macroscopic Assessment of Steel bySulphur Print

BS 6286: Measurement of Total or EffectiveThickness of Thin Surface-Hardened Layers inSteel

BS 6479: Determination and Verification ofEffective Depth of Carburized and HardenedCases in Steel

BS 6481: Determination of Effective Depth ofHardening of Steel after Flame or InductionHardening

BS 6533: Macroscopic Examination of Steel byEtching with Strong Mineral Acids

BS 6617: Determination of Decarburisation inSteel

124

REGISTERED TRADEMARKSOF BUEHLER LTD.:

Buehler

Carbimet

Chemomet

Duomet

Edgemet Kit

Epo-Kwick

Epomet

Epo-Thin

Fluor-Met

Handimet

Isocut

Isomet

Mastermet

Masterpolish

Mastertex

Metadi

Metadi Supreme

Microcloth

Micropolish

Nelson-Zimmer

Omnimet

Oscillimet

Sampl-Kup

Sampl-Kwick

Simplimet

Supermet

Surfmet

Texmet

Torramet

Vibromet

UNREGISTERED TRADEMARKSOF BUEHLER LTD.:

Acu-Thin

Apex

BuehlerHercules

Delta

Epo-Color

Epoxicure

Masterprep

Nanocloth

Phenocure

Phoenix Beta

Phoenix Vector

Planarmet

Probemet

Transoptic

Trident

Ultra-Pad

Ultra-Plan

Ultra-Pol

Ultra-Prep

Whitefelt

Buehler Trademarks

125

Index

AAbrasive cut-off machines:

chop cutters, 6

minimum area of contact cutting, 7

orbital cutting, 7-8

oscillating cutting, 7

pulse cutting mode, 6-7

traverse-and-increment cutting, 7-8

Abrasive cut-off wheels:

bond strength, 5

recommendations, 6

thickness, 5

troubleshooting guide, 112

Abrasive sectioning, 3, 5-9, 112

Acids, 103-106

Acrylic resins, 11-13, 113-115

Alumina, 16, 23, 24, 31, 36

Alumina grinding paper, 16, 25

Aluminum and alloys:

etchants, 66, 67, 69

preparation, 30-31

ASTM Standards, 119-120

Attack polishing:

Au, 48

Be, 33

Co, 46

Hf, 37-38

Nb, V and Ta, 39-40

safety, 37

Ti, 37

Zr, 37-38

BBases, 106

Beryllium

preparation, 33

Bright field illumination, 79,80,82

Buehler trademarks, 124

CCastable resins:

acrylic resins, 11-13, 114, 115

Conductive Filler particles, 12

epoxy resins, 11-13, 114, 115

exotherm, 12-13

Flat-Edge Filler particles, 12,14,15

recommendations, 12, 144

troubleshooting guide, 115

Central force, 15, 25

Ceramics:

etchants, 73

preparation, 53-54

Chemical polish, 37-38

Clamp mounting, 10

Cobalt and alloys:

etchants, 72

preparation, 46

Cold mounting (see castable resins)

Colloidal silica, 23-24, 30, 116

“Comet tails,” 19, 117-118

Composites:

preparation, 55-56

Compression mounting, 10-11, 13-15, 49, 113

Conductive mounts, 12, 113

Contra rotation, 25, 26-27

Coolant flow, 5

Copper and alloys:

etchants, 66, 71-72

preparation, 44

Cutting damage, 3 - 5, 112

DDamage, sectioning, 3-5, 112

Dark field illumination, 61, 81, 82

Diamond, 23, 24

Differential interference contrast, 19, 39, 40,55, 61, 81, 82

Digital imaging:

acquisition, 89

archiving, 89

calibration, 91-2

cameras, 89-90

clarification, 92

color, 90

pixels, 90, 91, 92

resolution, 91

EEdge preservation, 10, 11, 13-15, 117-118

Electronic materials:

preparation, 58-60

Embedded abrasive, 16, 17, 31, 35, 60,117-118

126

Index

E (CONT.)Epoxy resin, 11-13, 113-115

Etchants, 23, 69-73

chemicals, 106-107

dangerous solutions, 101, 103, 104, 105

mixing, 99-100

storage, 99

Etching, 63-73, 74

anodizing, 67

electrolytic, 67

heat tinting, 67-68

interference layer method, 68

procedure, 63

selective etching, 64-67

FFerrous metals:

etchants, 64-66, 70-71

preparation, 27, 41-43

Fluorescent dyes, 12, 49

GGrain size measurement, 3, 95-96

Grinding:

alumina paper, 16

diamond, 15, 16

embedding, 16,17

emery paper, 16

machines, 18

planar, 16

SiC paper, 16

HHafnium and alloys:

etchants, 70

preparation, 37-38

IImage analysis

binary operations, 94-95

feature-specific measurements, 95

field measurements, 95

measurements, 92, 95-96

thresholding, 92, 95-96

Inclusion measurement, 3

Individual force, 15

Iron and steel:

etchants, 64-66, 70-71

preparation, 27, 41-43

ISO standards, 120-121

K-LKnoop hardness, 84-85

Light microscopy, 74-82

Low-melting point metals:

etchants, 69

preparation, 34-35, 58-60

MMagnesium and alloys:

etchants, 69

preparation, 32

Magnesium oxide, 30

Metals:

allotropic transformations, 112

periodic table, 29

properties, 112

Microindentation hardness:

accuracy, precision, bias, 85-88

automation, 87-88

indentation size effect, 84

Knoop test, 84-85

Vickers test, 83-84

Microscope, reflected light:

condenser, 76

depth of field, 79-80

diaphragms, 76

“empty” magnification, 78-79

eye clearance, 77

eye pieces, 77

filters, 76

graticules, 77

illumination modes, 75, 79-82

Köhler illumination, 79-80

light sources, 74, 75

magnification, 77, 78

numerical aperture, 76, 78, 79

objective lens, 74, 75, 76

ray diagram, 74

resolution, 78, 79

reticles, 77

stage, 74, 77-78

stand, 78

working distance, 76, 77

127

Index

M (CONT.)Mounting:

castable resins, 10, 11-13

clamp mounting, 10

compression mounting, 10-11, 13-15, 113

conductive resins, 12, 113

edge retention (preservation), 10, 11, 13-15,113, 117-118

electroless nickel plating, 13, 15

low-melting point metals, 34

phenolic resins, 10

presses, 13, 15

purposes, 10

recommendations, 10, 12

shrinkage gaps, 10, 11, 13, 14, 15

thermoplastic resins, 110-11, 14, 113

thermosetting resins, 10-11, 13-15, 113

troubleshooting guides, 113

Mount size, 28

NNickel and alloys:

etchants, 72

preparation, 45

Nomarski DIC, 15, 39, 40, 55, 61, 81, 82

O-POblique illumination, 80-81

Periodic table of the elements, 29

Polarized light, 5, 31, 32, 33, 34, 36, 37, 38,41, 44, 46, 55, 63, 67, 81, 82

Polishing:

abrasives, 23-24, 116

automated, 21

cloths, 15, 20, 21-22, 24

electrolytic, 20

head position, 15, 28

lubricants, 23

manual, 20

recommendations, 116

vibratory, 24, 31-32, 33, 35, 37

Polymers:

etchants, 73

preparation

(see also: castable resins)

Precision saws, 8-9

Precious metals:

etchants, 72-73

preparation, 47-48

Preparation problems:

“comet tails,” 19, 117-118

disturbed metal, 19

edge retention (preservation), 10, 11, 13-15,113, 117-118

embedded abrasive, 16, 17, 31, 35, 60, 117-118

excessive relief, 19, 20, 117-118

pitting, 19

pull outs, 4, 117-118

scratches, 117-118

smear, 59, 60, 117-118

troubleshooting guide, 117-118

water sensitivity, 32, 33

Preparation procedures

contemporary methods, 25-27

Al and alloys, 30-31

Be, 33

ceramics, 53-54

Co and alloys, 46

composites, 55-56

Cu and alloys, 44

electronic devices, 58-60

Fe and alloys, 41-43

Hf and alloys, 37-38

low-melting point alloys, 34-35, 58-60

Mg and alloys, 32

Ni and alloys, 45

polymers, 61-62

precious metals, 47-48

printed circuit boards, 57

Refractory metals, 36-40

Sintered carbides, 51-52

Ti and alloys, 36-37

TSC and TBA coated metals, 49-50

Zr and alloys, 37-38

specific procedures, 28-62

Development, 28

Load, 28

platen size, 28

time, 28

“traditional” method, 25

128

Index

Printed circuit boards:

etchant, 57

preparation, 57

Pull outs, 4, 117-118

RRefractory metals:

etchants, 70

preparation, 36-40

Relief, 4, 19, 20

Rigid grinding discs, 15, 26-27, 41-43, 44-46,49-50, 51, 53, 59

SSafety:

acids, 103

bases, 106

chemical storage, 99

dust, 98

equipment operation, 98

fume hoods, 97

MSDS, 97, 99

personal protective equipment, 98-99

publications, 110

safety plan, 97

solvents, 100, 102

web sites, 110

Sampling, 3

Sectioning, 5-9

Sectioning damage, 3, 4, 5

Sectioning planes, 3

Shrinkage gaps, 10, 11, 13, 14, 15, 118

Silicon

preparation, 58-59

Silicon carbide paper, 16-18

ANSI/CAMI grit size, 17-18

FEPA grit size, 17-18

Silicon dioxide, 23

Sintered carbides:

etchants, 66-67, 73

preparation, 51-52

Smartcut, 8

Solvents, 100, 102

Sorby, Sir Henry Clifton, 75

Specimen preparation:

Goals, 2, 4

Methods

Contemporary, 25-27

Specific materials, 28-62

traditional, 25

(see: preparation procedures)

Problems, 4, 117-118

(see: preparation problems)

Stains, staining, 13, 14, 19, 117-118

Standards, metallography and hardness,119-123

Steels (see: Fe and alloys or Irons and steels)

TThermally spray-coated metals

preparation, 49-50

Thermoplastic resins, 10-11, 14

Thermosetting resins, 10-11, 13-15

Titanium and alloys:

etchants, 70

heat tinting, 68

preparation, 36-37

Trademarks, Buehler Ltd., 124

U-ZUltrasonic cleaning, 24

Vacuum impregnation, 11-12, 49, 53, 55

Vickers hardness, 83-84

Zirconium and alloys:

etchants, 70

preparation, 37-38

Documents

Buehler Guide Complete