Key words: Abrasion; brittle-ductile transition; erosion; fracture location; ring cracking; shaping; size-effect; statistical nature of strength; ultrasonic cutting; Weibull distribution.

In mechanical shaping processes for brittle solids, materials are removed by the propagation and intersection of cracks. To control the resulting surface finish, the extent of cracking is localized by loading the surface in very small regions, usually with small abrasive grains. Thus, in analyzing such processes as erosion and ultrasonic cutting, we have to examine the fracture patterns produced by small hard indenters.

As a starting point, it is shown that fracture loads for indenters subjected to normal force may be predicted from conventional strength tests by using Weibull's statistical treatment of brittle strength. It is then shown that Weibull's procedures may be extended to predict the distribution of fracture location in brittle solids. For the case of a spherical indenter under normal and tangential forces, many cracks may form as the load is increased. It is shown that the location of the outermost of these cracks may be predicted. Estimates may also be made of the depth of cracking and thus the extent of the cracking produced by a single contracting particle may be described.

These results may be applied directly to study the process of erosive shaping. The influence on volume removal of changes in the velocity and size of the impacting particles is predicted and shown to be in accord with experiment. Many other features of the erosion process may be described including a transition to "ductile" behavior in the case of glass when the impacting particles are small enough. The analysis of ultrasonic cutting presents greater difficulty because the distribution in size of the abrasive grains is of great importance. However, predictions may be made for the influence of load, grain size, and material properties on the rate of volume removal and these are shown to be in general accord with experiment.

Sonic machining of ceramics, W. B. Campbell, SP348, pp. 133-139 (May 1972).

Key words: Abrasive; ceramic machining; sonic machining; vibrational machining; wheel machining.

The effect of 10 kHz sonic vibrations on the shaping and finishing of fired ceramic shapes was investigated to identify optimum machining conditions commensurate with product and surface quality. Sonic motors, developed at The Ohio State University and with outputs up to 15 horsepower, were used to activate the machine tool. Significant improvements in tool wear were observed in impact-tuned systems that produced removal rates greater than 45 cu in per min. Template and profile tooling were successfully used to obtain open cutting surfaces. Over 3,000 test cuts provided the data necessary to identify optimal parameters of operation. Environment-sensitive machining behavior of nonmetals, A. R. C. Westwood and R. M. Latanision, SP348, pp. 141-153 (May 1972).

Key words: Adsorption; ceramics; drilling; environmental effects; fracture; glasses; machining; mechanical behavior.

Liquid environments can influence the efficiency of machining nonmetallic solids in a variety of ways, e.g., by serving as lubricants, coolants, or particle dispersants. More importantly, however, certain environments can markedly increase or decrease the hardness of the near-surface regions of such solids, and thereby exert a profound influence both on the rate of material removal and on tool life.

Because of its technological potential, the characteristics and possible mechanisms of this latter phenomenon - the Rebinder effect-receive primary consideration in this paper.

For crystalline ceramics, Rebinder effects in machining arise because of the influence of the environment on nearsurface dislocation behavior. Effects resulting from adsorption-induced changes in the surface free energy of the solid are of minor importance. Rebinder effects can also occur in noncrystalline solids, however, and recent observations on such effects in various glasses are described. The possibility that these effects are caused by a stress-plus-chemisorptioninduced redistribution of sodium ions in the near-surface region is discussed.

The importance of considering the total cutting system, environment-solid-tool, in any account of environment-sensitive machining is stressed, for environments which facilitate material removal when one type of tool is used can be detrimental to the effectiveness of another tool with a different cutting action.

Shaping or figuring ceramic surfaces by ion-beam bombardment, P. W. Levy, SP348, pp. 155-168 (May 1972).

Key words: Figuring; ion-beams; optical surfaces; polishing; radiation damage; sputtering.

The shaping or figuring of glass and ceramic surfaces by ion-beam erosion is a relatively new technique. All operations are carried out in a vacuum. Usually the optic is bombarded by a collimated beam of rare gas ions whose energy is between 1-100 keV. Selected areas can be eroded by electrostatically deflecting the beam or, less conveniently, by moving the optic. The erosion area can be controlled by focusing the beam with electrostatic or magnetic lenses. The basis of the erosion process is the physical phenomenon known as sputtering. This occurs when the incident ions scatter elastically with atoms near the surface of the target and some of the displaced atoms, after participating in one or more collisions, acquire sufficient velocity to escape from the solid. Other dislodged atoms become interstitials and, together with vacancies created in the recoil process, create a radiation-damaged layer just below the eroding surface, whose thickness is roughly the range of the bombarding ions. For ceramic materials, typical erosion rates are on the order of angstroms/μamp/cm2/min and increase with increasing angle of incidence. Ion-beam currents up to 100 μamp/cm2 are practical. Under these conditions, the material struck by the beam may reach temperatures between 300 and 700 °C. The maximum lattice disruption occurs at a depth of one incident ion range from the original surface just as the eroding surface reaches this point. Additional bombardment does not increase the damage beyond this maximum. Furthermore, the damage is minimized by using bombardment conditions which maximize the temperature in the target. It is not known if there are any radiation damage effects which reduce the usefulness of ion-beam figured optics. At the moment, it would appear that ion-beam figuring is a practical technique for "touching up" conventional optics and, more importantly, for fabricating nonspherical or nonaxially symmetrical optics. Also, the ion-beam erosion process is particularly amenable to computer control.

The effect of sputtering on surface topography and strength of ceramics, R. W. Rice, SP348, pp. 171-187 (May 1972). Key words: Carbides; glass, ion beam sputtering; nitrides; oxides; RF sputtering; strength; surface finish; twins.

The topography developed as a result of RF sputtering on a variety of ceramic surfaces is described showing that sur

face cracks, scratches, pits, etc., are fairly rapidly rounded out. Generally, this occurs without differential grain boundary sputtering. However, some bodies develop a rough finish due to nonuniform microstructural removal. Limited trials with ion beam sputtering show it also rounds cracks, etc., but does not develop a rough finish on bodies that will with RF sputtering.

Because of the rounding of stress concentrating features and generally similar effects on single and polycrystalline bodies, sputtering was investigated as a substitute for flame polishing, but with much wider applicability. Flame polished strengths were not obtained. However, some materials (e.g., MgAl2O4) do show greater improvements in strengths than others (e.g., Al2O4) as a result of sputtering. These differences will be discussed and results compared with other methods of surface finish.

Both the detailed nature of the sputtered surface as well as enhancement of grain boundaries and other microstructural features under certain conditions indicate that sputtering may also be useful as an etching technique.

Computer controlled ionic polishing of optical surfaces, J. W. Douglass, SP348, pp. 189-192 (May 1972).

Key words: Computer controlled process; ion beam technology; ionic polishing; optical fabrication; optical scatter; sputtering.

Results of the development of an ionic polishing investigation are described. During this investigation many controlled polishing experiments were performed. The results of two of these experiments are presented. In the first of these, a 0.203 m diameter optical flat of fused silica was polished to a diffraction limited surface quality of 95A (9.5 nm) rms. The second experiment described was the conversion of a 0.203 m diameter Cervit sphere into an f/5 paraboloid of diffraction limited quality by using analytical expressions to generate the removal distribution. The ionic polishing process was computer controlled for this experiment. The temporal stability and optical scatter performance of ionically polished surfaces has been determined to be as good as or better than conventional surfaces.

Arc, laser, and electron beam machining of ceramics, R. W. Rice, SP348, pp. 193-195 (May 1972).

Key words: Arc machining; cutting; drilling; electron beam machining; laser machining; shaping; thermal shock control.

The various types of arc, laser, and electron beam machining, the considerations that are required in their operations, and some of the applications for which they have been tried or used, are briefly reviewed. The most general consideration is whether thermal shock and stress cracking will be a problem. The environmental and material parameters, which vary substantially with the types of these different machining processes, can also be quite important. Meeting the requirements of all these considerations does limit the versatility and applicability of these techniques. Nevertheless, they deserve consideration for a variety of applications.

The techniques and mechanisms of chemical, electrochemical, and electrical discharge machining of ceramic materials, D. W. Lee and G. Feick, SP348, pp. 197-210 (May 1972).

Key words: Ceramic materials; chemical machining (CHM); electrical discharge machining (EDM); electrochemical machining (ECM); intermetallic compounds.

Electrochemical Machining (ECM), Electrical Discharge Machining (EDM), and Chemical Machining (CHM)

techniques have not been extensively used as a material removal process for ceramics. The fact that these processes are not dependent upon the hardness of the workpiece should make them attractive for the shaping of ceramics. However, ECM and EDM techniques are limited to material with reasonably good electrical conductivity while CHM may be limited by the availability of effective etchants and the nature of the ceramic itself.

The theory and techniques of chemical, electrochemical, and electrical discharge processes will be reviewed. The variables and important parameters of the material removal processes will be discussed, including etchants and mask techniques for chemical milling; current-voltage characteristics, dielectric fluids, electrode materials, etc., for EDM; electrolytes, gap effects, current density, etc., for ECM. The available information and experiences developed on metals and alloys will be used to examine the applications and limitations of these processes to ceramic materials. Where possible, the effects of the material removal process on the surface condition, microstructure, and subsequent material properties will be pointed out. Finally, comments will be made on the technical and economic feasibility of ECM, EDM, and CHM processes for the shaping, cutting, and finishing of crystalline ceramics.

Improvements in the surface finish of ceramics by flame polishing and annealing techniques, M. J. Noone and A. H. Heuer, SP348, pp. 213-232 (May 1972).

Key words: a-Al2O3; annealing; flame-polishing; reinforcements; ruby; sapphire; strength; surface perfection.

Improvements in the surface perfection (i.e., "smoothness") of ceramics can be achieved by thermal treatments at temperatures where material transport may be induced at the surface. Such treatments include annealing in controlled environments, where surface diffusion, vapor transport mechanisms, etc., may be active, and flame polishing, where a thin layer at the surface of the material is melted, allowed to flow freely, and then resolidified.

The effects of these treatments on the surface of single crystals of a-Al2O, are reviewed in this paper. The most sensitive assessment of the degree to which surface perfection is attained is the measurement of the strength of the treated crystals; for this reason, strength data are used extensively to characterize the thermal treatments described. In the case of flame polishing, it found that a sufficient degree of surface perfection can be attained so that the strength is no longer limited by surface defects (the usual experience with ceramics) but by defects within the material. The technological significance of high strength Al2O, filaments for reinforcement of metals and the unfortunate deterioration of the surface in these applications is briefly described.

Healing of surface cracks in ceramics, F. F. Lange, SP348, pp. 233-236 (May 1972).

Key words: Abrasive machining; ceramics; crack healing; sintering; surface damage.

Grinding, cutting, thermal shocking, impacting, and rough handling all introduce surface cracks that degrade the strength of ceramics. Within the last year, it has been shown that, once introduced, these cracks can be eliminated by either resintering the damaged ceramic component for the case of oxides or oxidizing the component for the case of materials such as SiC.

Results of crack healing experiments, performed on Al2O3, ZnO, and SiC will be reviewed. The technical impli

cations of these results will be discussed as related to abrasive machining.

Flame polishing of flat oxide bars, P. F. Becher and R. W. Rice, SP348, pp. 237-244 (May 1972).

Key words: Flame polishing; glass; rutile; sapphire; spinel; strength; surface characterization; twinning.

Some of the problems and limitations of flame polishing flat bars are discussed. Results are presented for single crystal, a-Al2O3, as well as for more limited trials of MgAl,O,, TiO,, and soda lime glass. The wide variability of strength is partially related to variations in surface, but twinning also appears to be important in sapphire and TiO2. Preliminary results on twin sources and their effect on strength of sapphire are discussed.

Continuous flame polishing of sapphire filament, J. T. A. Pollock, SP348, pp. 247-256 (May 1972).

Key words: Characterization; continuous flame polishing; sapphire filament; strength enhancements.

Continuous flame polishing of nominal 2.5 × 10 m diameter single crystal sapphire filament oriented with the c-direction parallel to the filament axis has resulted in considerable enhancement of the tensile fracture strength. Optimum increases of approximately 1.1 x 10° N/m2 (1.6 × 10 psi) have been obtained on flame polishing many lengths of filament exhibiting as-grown tensile strength of 2.2 -2.8 × 10 N/m2 (3.2-4.0 × 10 psi). Flame polishing was carried out in a continuous manner at 6.3 x 10 m/sec using an oxygen/hydrogen flame. Data are presented which suggest that the enhancement is not entirely dependent on the production of a more perfect ALO, surface. Increases in strength of approximately 5.5x108 N/m2 (8×10 psi) are reported for filament which has passed through flames not sufficiently hot for surface melting to occur. Maximum strength enhancement is obtained when the oxygen/hydrogen flame temperature is such that an axial molten zone two to three times the filament diameter is produced at the filament surface. When the flame temperature is too high, the geometric integrity of the filament is lost and an apparent fall in tensile strength from the peak value is observed.

Metallographic evidence is presented indicating that the molten zone has a radial depth of less than 6.5 × 10-6 m. Scanning electron microscopy and related surface analysis studies are reported which confirm that optimum polishing results in a smoother filament surface. Experiments to determine the state of stress in the filament before and after polishing are described.

The relative contributions to the reported tensile strength enhancement of thermal strain relieving, thermally activated atomic diffusion leading to blunting of possible fracture sources, and the creation of a more perfect AbO3 surface are discussed.

Preparation of smooth crystalline, damage free, sapphire surfaces by gaseous etching, W. A. Schmidt and J. E. Davey, SP348, pp. 259-265 (May 1972).

Key words: Crystalline sapphire surfaces; fluorinatedhydrocarbon etch; fluorotrichloromethane; freon etching; hydrogen annealing; sapphire; sapphire substrates; sapphire surface preparation.

Sapphire has a combination of properties such as relative chemical inertness, exceedingly high resistivity, optical transparency, and crystalline structure that suggests its use as a substrate material for various semiconductor devices. Its successful use for silicon vapor phase epitaxy substrates indicates that similar results might be possible with other

semiconductors. Previous epitaxial studies have shown that surface crystalline disorder and topographic imperfections inhibit epitaxial growth; therefore, highly polished, well ordered sapphire surfaces were needed. Mechanical polishing alone is insufficient as it results in cold flow and work damage.

Initial experiments in removing surface damage were performed with hot phosphoric acid and with hydroxide etches. Undesirable preferential etching was observed for both processes and the processes did not seem amenable to the routine production of high quality surfaces. Two other surface treatments for which successful results have recently been reported were investigated.

The first was a hydrogen firing technique in which the substrates were heated in hydrogen in an Mo wound resistance furnace at 1500 °C. The second was a simple laboratory process in which the sample is heated by means of RF heating of a carbon susceptor to 1350 °C in an atmosphere of helium and fluorotrichloromethane (Freon-11). Commercial reagents were used throughout and a fused quartz tube was used for the reaction chamber for the second technique.

Reflection electron diffraction (RED) was used to determine the surface crystalline order, and electron microscopy (EM), using high resolution replication techniques, was used to examine the topography structure of the surfaces. A number of different substrates from various industrial sources, with different surface topographies and different orientations, were used.

Hydrogen firing at 1500 °C results in an etch rate of 0.1 10- m/min. Firing for times up to 45 min did not produce consistent surfaces on the 0, 60 or 90° orientations. While hydrogen firing did produce high quality, well ordered crystallograph surfaces by RED, their topographic condition was poor. Resolvable surface structures could be detected on some with standard optical microscopy (200x, dark field) and for all high resolution EM measurements. The final surface finish quality was related to the quality of the prefired surface, indicating that complete damage removal was not accomplished.

Freon firing at 1350 °C resulted in an etch rate of 1.5 10-$ m/min. This technique consistently produced well ordered, high quality surfaces for the 60 and 90° oriented surfaces. EM viewed 60 and 90° freon-fired surfaces had no resolvable surface structure after a 5 min etch and were well ordered crystallographically as measured by RED. The same etch produced strongly etched surfaces on 0° oriented material.

The strength of gas polished sapphire and rutile, R. W. Rice, P. F. Becher, and W. A. Schmidt, SP348, pp. 267-269 (May 1972).

Key words: Chemical polishing; rutile; sapphire; spinel; strength; twinning.

Surface polishing of a-Al2O, using a helium-freon gas mixture at elevated temperatures is shown to yield bend strengths comparable to flame polished sapphire. However, the gas polishing process proved to be more versatile as evidenced by the success with flat bar surfaces. Further, the process can be readily applied to some other materials as demonstrated by the substantial improvements in the strength of TiO, after gas polishing. Limited attempts at polishing other ceramic materials are also discussed, as well as observations on strength variations and fracture behavior.

Analysis and characterization of ceramic surfaces for electronic applications, R. C. Sundahl and L. Berrin, SP348, pp. 271-290 (May 1972).

Key words: Aluminum oxide; Auger electron spectroscopy; crystallographic texture; ion microprobe; scanning electron microscopy; substrates; surface defects; surface segregation; thin film adhesion; thin film circuit imperfections.

The criteria which are used to evaluate the surfaces of ceramic substrates for use in the electronics industry must be related to the specific application of the surface. This article emphasizes one such application- the use of a ceramic surface as a support for complex thin film conductor patterns which serve to interconnect silicon integrated circuit chips. Those ceramic surface parameters which are found to be critical to this application are (1) topographical properties, (2) chemical properties, and (3) crystallographic properties.

These properties are characterized using such tools and techniques as optical, transmission electron and scanning electron microscopy, profilometry, electron microprobe analysis, ion microscope analysis, Auger electron spectroscopy, and electron diffraction. Special emphasis is placed on the relationship between the results of such analyses and the performance of the surface as part of the SIC interconnect system.

Surface characteristics of ceramic substrates for hybrid and microwave electronic circuits, J. K. Emery, SP348, pp. 293-300 (May 1972).

Key words: Camber; ceramic; fired; flatness; ground; lapped; microwave; polished; roughness; substrate; surface; thick film; thin film; waviness.

Several methods currently used for finishing ceramic substrates in ready-for-deposition state are discussed, including firing, glazing, burnishing, grinding, lapping, and polishing. Each method tends to exhibit a different and typical combination of surface characteristics, such as variations in thickness, parallelism, camber, waviness, roughness, and localized defects; each of these in turn may influence the deposition process, performance, and reliability. Analysis of some typical surface details, particularly roughness and waviness, as significant surface factors capable of affecting process or function, is presented with comments on instrumentation and measurement procedures.

Significant differences exist in the use of terms relating to substrate surfaces and their characteristics; a need is indicated for more particularized use of such terms and for recognition of uniform definitions. Clarification of terminology may

facilitate improved communications and measurements, which in turn may correlate with current and future requirements in circuit production. Several existing applicable standards are referenced, and definitions of terms are suggested. A number of choices of surface characteristics are presented which may be related with process controls and function, but no recommendations for particular values of surface details are intended.

Ceramic surface texture by reflective replica technique, W. C. Lo, SP348, pp. 301-307 (May 1972).

Key words: Alumina substrates; surface replica; surface texture.

Surface texture is a parameter of general concern in the study of thin film metallizing adherence on ceramic substrates. It refers to the geometrical character of the surface irregularities recurring many times across the ceramic surface tending to form a pattern. In this context, surface texture is determined by the size, shape, arrangement, and distribution of the surface constituents.

This note presents a convenient method by which the surface texture of ceramic substrates can be visually assessed

and recorded photographically using a simple metallographic microscope.

Examples in applying this technique for distinguishing and differentiating surface textures of different types of ceramic substrates are given. Use of the technique for identifying surface defects is also shown.

The simplicity of the method should warrant its use for monitoring the physical surface quality of ceramic substrates.

Quantitative surface finish characterization by CESEMI, E. W. White, H. A. McKinstry, and A. Diness, SP348, pp. 309316 (May 1972).

Key words: Ceramic materials; CESEMI (Computer evaluation of scanning electron microscope images); digital magnetic tape recordings; profilometer; profilometer rasters; quantitative characterization; scanning electron microscope; surface finish analysis; surface morphology; surface roughness; surface topography; three-dimensional images.

A feasibility for quantitative characterization of surface finishes based on the computer evaluation of scanning electron microscope images (CESEMI) has been established. Measurement of the number, size, shape, and orientation of isointensity (or selected intensity interval) regions constitute the basic analysis. Brightness in SEM images varies as a function of the steepness of slope of the surface at a given point. In their analysis, the characteristics of grooves, pits, ridges, etc., can be studied separately. Detailed interpretation of the results is not attempted in this preliminary study. A feasibility for recording and analyzing profilometer rasters has also been demonstrated. Such recordings are, in essence, true three-dimensional images of surface topography insofar as the profilometer stylus faithfully follows the surface morphology. Analysis of these profilometer recordings by the basic CESEMI computer programs yields direct descriptions of the topography. The SEM image and profilometer "image" analyses appear to be complementary techniques. One obvious advantage of the SEM image analysis is that it is a no-contact technique; hence, there are no problems introduced by stylus damage.

An assessment of surface and subsurface damage introduced in ceramics by semifinish grinding operations, B. G. Koepke, SP348, pp. 317-332 (May 1972).

Key words: Ceramics; etch pits; fracture; grinding damage; plastic deformation; surface condition; surface grinding.

The nature and extent of grinding damage introduced by surface grinding a number of ceramics having widely varying mechanical properties have been studied. Surface damage has been characterized by optical and electron microscopy, and subsurface damage has been observed using etch pit techniques. Grinding damage was found to depend strongly on wheel type and rate of material removal as well as on the mechanical behavior and microstructure of the workpiece. The results indicate that grinding damage is mainly composed of mixtures of three types depending on the mechanism of material removal.

When material is removed efficiently from low impact resistance ceramics (e.g., magnesium oxide and ferrite), the surfaces are generated by brittle fracture and are composed of regions of transgranular and intergranular fracture. When grinding is inefficient, i.e., when the grinding wheel loads up, material is removed by plastic flow. The resultant surfaces on deformable ceramics (e.g., magnesium oxide, ferrite, and silicon) are smooth and burnished but may contain thermal cracks due to the heat generated. In this instance, subsurface damage consists of a discrete, highly deformed layer containing cracks in most cases.

When material is removed efficiently from high impact resistance, nondeformable ceramics (e.g., alumina and boron carbide) material is removed by plastic flow and by transgranular and intergranular fracture. The presence of extensive plastically flowed regions on the ground surfaces of extremely hard ceramics is surprising and points out the extremes of stress and temperature existing under the wheelworkpiece interface during a grinding pass.

Observations on mechanically abraded aluminum oxide crystals by transmission electron microscopy, B. J. Hockey, SP348, pp. 333-339 (May 1972).

Key words: Abrasion; aluminum oxide; dislocations; ionbombardment; microtwins; scanning electron microscopy; subsurface damage; transmission electron microscopy.

Use of the argon ion-bombardment thinning technique has made possible the examination of the near surface regions of mechanically abraded aluminum oxide by transmission electron microscopy. Observations on diamond-polished (0.25 μm), alumina-polished (0.3 μm), and diamond-ground specimens have shown that subsurface damage as well as surface damage is typically produced.

Specifically, mechanical polishing introduces relatively high densities of dislocations to a depth of approximately 1 μm from the original surface. The dislocations are generally in the form of half-loops and are clearly associated with surface scratches produced by individual abrasive particles.

The magnitude of residual surface stresses and the irregular surface topography produced by grinding necessitated the removal of at least 2-4 μm from the original surfaces. At this depth, both the nature and extent of subsurface damage in polycrystals varied from grain to grain., Most grains contained either tangled dislocation arrays or microtwins (either basal or rhombohedral). A number of grains, however, were found to be completely free of damage and may correspond to regions below fracture surfaces which are apparent in observations of ground surfaces by scanning electron microscopy.

Acoustic emission monitoring of surface-damaged ceramic materials, L. R. Bunnell, J. C. Crowe, and P. E. Hart, SP348, pp. 341-342 (May 1972).

Key words: Acoustic emission; glass; Lucalox; Pyroceram; surface damage.

The acoustic emission response of Lucalox, fused silica, single

crystal alumina, Pyroceram, and soda-lime-silica glass to surface damage is reported. Low temperature annealing at 200 °C reduces the response levels; annealing at 400 °C removes all of the induced acoustic activity.

Effects of surface finishing on mechanical and other physical properties of ceramics, R. J. Stokes, SP348, pp. 343-352 (May 1972).

Key words: Ceramics; cracks; dislocations; electrical properties; electro-optical; ferrite; machining; magnetic properties; mechanical properties; optical properties; residual stress.

This paper reviews the effects of mechanical finishing operations on the physical properties of ceramics. Ceramic machining results in a defective surface containing cracks, dislocations, point defects, and residual stresses. The relative significance of these defects depends on the physical property of interest. The mechanical properties of single crystals are sensitive to dislocations (semibrittle) and surface cracks (brittle); the mechanical properties of polycrystals are sensitive to surface cracks; the electrical properties of semiconductors are sensitive to surface

trapping sites; magnetic, piezoelectric, and optical properties are particularly sensitive to residual stresses. To optimize physical properties, these defects must be eliminated by mechanical lapping, chemical etching, or thermal annealing.

Strength effects resulting from simple surface treatments, H. P. Kirchner, R. M. Gruver, and R. E. Walker, SP348, pp. 353-363 (May 1972).

Key words: Abrasive machining; alumina ceramic; chemical polishing; crack healing; delayed fracture; glazing; humidity; refiring; strength; surface flaws; surface treatments; thermal shock.

In well made polycrystalline alumina ceramics subjected to external loads, fracture originates, in almost every case, at surface flaws rather than at volume flaws. This susceptibility to surface flaw failure has been reliably established in experiments in which compressive surface layers were used to obtain substantial improvements in strength. Knowing that the failures occur at surface flaws, the artificial introduction of surface flaws can be used to obtain an understanding of the ways in which surface flaws affect the strength. Additional information can be gained by using simple treatments to change these artificial surface flaws and by observing the effect of these treatments on the strength.

In the present investigation, artificial flaws were introduced by single point tools, abrading, thermal shock, and abrasive machining. The flaws were treated by refiring, chemical and flame polishing, chemical etching, glazing, and prolonged storage in various environments.

The changes in the flaws were characterized by microscopy and to a limited extent by profilometry. Both flexural and tensile strengths were measured. The effects of the treatments are discussed in terms of changes in the average strength and variations in the distributions of the individual strengths.

The effect of grinding direction on the strength of ceramics, R. W. Rice, SP348, pp. 365-375 (May 1972).

Key words: Carbides ceramics; fracture; grain size; grinding; hardness; mechanical testing; nitrides; oxides; strengths.

It is shown that grinding bars in a direction parallel with the tensile axis has limited effects on their strength while grinding perpendicular to the tensile axis may have no effect or can reduce strengths as much as 50 percent. In single crystals, the reduction in strength generally increases with the hardness of the material but can also depend on crystallographic orientation. In polycrystalline bodies, the reduction also depends a great deal on grain size, with the effect generally increasing with decreasing grain size. Also the effect tends to be greater in polycrystalline bodies that are weaker than normal. The effect of grinding direction on strength is attributed to stress concentrations due to grinding stria, whose continuity and depth generally increase with hardness. The severity of these stria tends to increase as the amount of plastic deformation in the machined surface decreases. The amount of plastic deformation is shown to vary approximately inversely with hardness. The effect of varying strength with grinding direction on the interpretation of various tests is discussed.

The influences of material removal on the strength and surface of an alumina, H. S. Starrett, SP348, pp. 377-388 (May 1972).

Key words: Alumina; surface finish; surface treatments; Weibull.