Featured Product
This Week in Quality Digest Live
Metrology Features
Master Gage and Tool Co.
Why it matters for accurate measurements
Scott A. Hindle
Part 4 of our series on SPC in the digital era
NIST
Having more pixels could advance everything from biomedical imaging to astronomical observations
Tara Fortier
It will likely change in the next decade
Douglas C. Fair
Part 3 of our series on SPC in a digital era

More Features

Metrology News
New KMR-Mx Series video inspection system to be introduced at the show
Study of intelligent noise reduction in pediatric study
Easy to use, automated measurement collection
High-end microscope camera for life science and industrial applications
Three new models for nondestructive inspection
Machine learning identifies flaws in real time
Advancing additive manufacturing
ABB robot charger automatically detects boreholes, fills them with charges, with no humans present
Two awards annually for students studying precision metrology

More News

Sharbari Banerjee

Metrology

Avoiding Sample Preparation Pitfalls in Materials Analysis

A material’s microstructure acts as a record of its process history

Published: Wednesday, January 26, 2011 - 07:55

A material’s internal structure or microstructure is defined as one that is viewed with either a metallurgical microscope at magnifications in the range of 25X up to 1000X or a scanning electron microscope (SEM) at higher magnifications. Features observed in microstructures include phases and precipitates in processed materials, dendritic structures in cast alloys, heat-affected zones in welds, layer thicknesses in coated materials, and corrosion modes and depths in affected parts.

Microstructural examination yields valuable information on the mechanical and thermal treatments performed on the material and is important in characterizing materials for quality purposes. We can say that the microstructure of a material acts as a record of its process history.  Because many material properties such as tensile strength and ductility are influenced by process history, microstructures can be used to estimate such properties. As such, microstructural analysis is an important screening tool in the research and development of new materials. Failure analysis investigations rely on microstructures to determine failure origins and property changes in the material while in service.

The surface of a processed material typically obscures the internal structure due to the presence of surface defects such as oxide layers, oil, dirt, and handling damage. Sample preparation is the process by which the surface defects are removed to an extent that allows microscopic examination of the internal structure of materials. In other words, metallographic sample preparation allows us to view the microstructure of a material.

Surface material removal is achieved by sectioning, mounting, grinding, polishing, and, occasionally, etching. Each step is critically important in obtaining true microstructures. Selecting improper samples or introducing artifacts during grinding or polishing will lead to inaccurate microstructural interpretation. This article describes some common problems that occur during sample preparation and ways to correct them.

Sectioning

Sectioning is the process by which materials are sized down to samples appropriate for mounting, grinding, and polishing—typically smaller than 2 in. The most frequent mistake made in sectioning is using aggressive methods that produce much mechanical and thermal damage.  For example, a shear cutter can mechanically damage up to a 5 mm depth from the sample surface (figures 1a, 1b, and 2).  Similarly, plasma cutting produces deep thermal damage, up to several millimeters in depth. Removing damage to such a large extent would require prolonged grinding, and therefore any surface cut with one of those methods is best recut with an abrasive or precision saw for further metallographic preparation.


Figure 1a:  Thermal damage in hardened steel  


Figure 1b:
  Normal hardened steel microstructure 


Figure 2:  Thermal and mechanical damage in zinc 

Abrasive saws yield fast, burn-free cuts on a wide variety of materials. However, a number of factors, if not controlled, can trigger problems such as resistance to cutting, burnt cut surfaces, and glazed or broken wheels. Many of these troubles occur due to an incorrect choice of a wheel for the material being cut. Wheels are available in a variety of abrasive particle types and sizes as well as bond strengths, and they break down during cutting. Hard materials require soft bonded wheels that break down rapidly, thus exposing fresh abrasive. A stronger bonded wheel will resist cutting a hard material, become glazed at the edges, and eventually fracture. Wheel fracture can also result from either the wheel or sample not being clamped properly. Insufficient lubrication and cooling can overheat the sample, leading to a burnt cut surface and a potentially altered microstructure.

Precision saws, which use diamond or cubic boron nitride wafering blades, allow for highly precise positioning and sectioning of specialty materials such as ceramics, electronics, and medical device materials. As with abrasive wheels, correct blade selection for the material being cut eliminates problems such as prolonged cutting times and poor cut surface finish. Periodically dressing wafering blades with a ceramic stick maintains their cutting ability. Soft, ductile materials pose an additional challenge by building up on the blade and causing it to stall. Reducing the load or feed rate and lightly dressing the blade during the cut help to minimize this situation. Sometimes, it can be prevented by simply using a thinner blade.

Mounting

Errors in the mounting process include using the wrong method—compression or castable—or an inappropriate compound for the sample being mounted. Delicate, friable samples and those made of low-melting-point materials may be damaged by heat and pressure in the mounting press and are best mounted in a castable compound. However, all castable compounds cure with an exotherm, and peak temperatures may be relatively high. This is a factor to be considered when selecting a compound for very delicate samples. Most other samples can be safely compression-mounted.

Relief and rounding of sample edges can result from selecting mounting compounds that either do not match the sample hardness or shrink significantly from the sample edges while curing, leaving the latter unsupported and creating gaps that may trap grinding and polishing debris. These defects are particularly problematic when the sample edge is the area of interest—for example, when a coating or corroded surface is to be examined. Mineral- or silica-filled mount media are associated with improved abrasion resistance and reduced shrinkage compared to wood-flour-filled media.

Radial splits in compression mounts indicate a sharp sample corner or a too large sample for the mount size (figure 3). Grinding off the corner and reducing sample size will eliminate this defect. An unfused, chalky appearance indicates failure to reach the molding temperature, while a surface bulge indicates insufficient pressure during solidification or moisture in the powder (figure 4).


Figure 3: Radial crack due to sample corner at edge 


Figure 4:  Bulged mounts

All castable mounting compounds shrink to some extent, with acrylics showing considerably larger gaps than epoxies. Castable epoxy shrinkage effects can be mitigated with the use of ceramic edge-filler particles or by plating the edge with nickel. Acrylics filled with silica show reduced shrinkage and higher hardness.

Soft, uncured castable mounts can occur due to either incorrect ratios of epoxy resin to hardener or inadequate mixing of resin and hardener. Accurately weighing resin and hardener quantities to represent their recommended ratio of mixing, and mixing them thoroughly, ensures there are no unpolymerized sites to prevent hardening of the mount. Mixing too small a quantity of epoxy does not produce a sufficiently high exotherm for curing to initiate and is another possible cause for a soft mount. Too large a volume, on the other hand, produces excessive exothermic heat, causing the mixture to become unstable. A total mix weight between 25 gm and 100 gm is an optimal range.

Grinding and polishing

The purpose of the first grinding step is to remove all sectioning damage and establish a perfectly flat surface with a uniform scratch pattern that is suitable for further polishing. An insufficiently aggressive initial grinding surface or inadequate time allowed for grinding may produce a flat surface but leave subsurface damage that is usually not removed in subsequent polishing steps. To grind effectively, the abrasive must be one and one-half times harder than the material being ground. Although silicon carbide grinds most materials efficiently, hard tool steels and ceramics are ground faster with diamond. Poor lubrication can lead to both mechanical and thermal grinding damage.

Ground sample surfaces are further refined by polishing on cloths that are charged with abrasives, typically diamond or alumina. Polishing consists of one to three steps that produce progressively finer scratch patterns and lesser depths of subsurface damage. It is important to accomplish these functions without introducing any artifacts that can be misinterpreted as microstructural features. Also important is to maintain flatness and clean edges throughout the preparation process.

Parameters to be controlled during polishing with a semiautomatic grinder or polisher are abrasive type and size, cloth surface, platen- and head-rotation speed, and rotation direction relative to each other, amount and method of force application, and time at each step. While there is considerable latitude in parameter selection for a variety of materials, polishing challenges can often be solved by addressing one or more of these parameters.

For example, diamond suspensions are widely used in the intermediate polishing stages for their convenience (figure 5). However, the free rolling particles in a diamond suspension may embed in softer materials like copper or aluminum and some of their alloys. This problem can be alleviated by replacing the suspension with a paste that is worked into the cloth surface, thus making the particles semifixed. Pastes, of course, have no inherent cooling or lubrication and are typically used in conjunction with a lubricating fluid.


Figure 5:  Diamond particles embedded in pure niobium

Very soft, ductile materials such as the lower-melting-point metals—tin, lead, zinc, and their alloys—present their own set of challenges. Diamond being too hard for these materials, they are prepared with a number of silicon carbide steps followed by one or two steps of polishing on napped cloths with alumina suspensions. However, silicon carbide has a tendency to embed in these materials, and prolonged polishing on napped cloths can cause relief or smearing. These issues are alleviated by wax-coating the silicon-carbide paper surfaces and minimizing polishing times.

Electronic samples often contain stacks of materials of such varying hardness as ceramics, copper or gold, silicon, PZT, and polymers. Artifacts that can appear in these samples include cracks in the ceramic and silicon, ill-defined edges, and relief between layers (figures 6 and 7). Cracks can be avoided by encapsulating the samples prior to sectioning and using gentle options such as slower speeds and complementary rotation during grinding and polishing. Relief can be prevented and edge definition improved by preparing in several short steps instead of a few longer ones and by polishing with non-napped cloths.


Figure 6:  Edge rounding in pin connector                       


Figure 7:  Relief between layers in SAC305 

Second phase particle retention can often be a problem in both retaining graphite in cast irons and inclusions in steels and other alloys. Inclusion pullout is particularly insidious because the loose inclusion particles, being hard, can scratch the sample surface. The even pressure applied during semiautomatic grinding and polishing is preferable to manual preparation in minimizing inclusion pullout; slower speeds and higher loads also help retain them. A series of silicon carbide grinding steps followed by polishing with fine diamond particles on flat cloths produces superior results. Final polish is achieved with alumina on a napped cloth; however, the time should be minimized.

Scratches remaining on a prepared surface, while not always perceived as a problem, can interfere with certain analytical techniques, for example, electron back-scattered diffraction (EBSD) in a scanning electron microscope. There are two possible sources of scratches—those that are not completely removed in prior grinding or polishing steps, and those that are introduced due to external factors. The first type can be eliminated by allowing sufficient time at each preparation step to remove all scratches and subsurface deformation from the previous step. The time at individual steps can be reduced by reducing the particle size increment between steps, thus increasing the number of steps. Using high-quality diamond products ensures fast and superior results.

Numerous factors can be responsible for scratches from external sources. These include cross-contaminated cloths, improper cleaning of samples and sample holders between steps, and generally unclean lab conditions. Cloth contamination can be avoided by rinsing cloths after each use and storing them in individual drawers in storage cabinets. Washing samples and holders thoroughly between grinding and polishing steps is essential to prevent abrasives and debris from being transferred to subsequent surfaces. Debris trapped in mount-shrinkage gaps and second-phase particle pullout also contributes to these types of scratches. Ultrasonic cleaning may be necessary to remove such debris if it proves too tenacious during regular washing.

To conclude, sample preparation can be painless if care is taken at each step. Key factors to avoid are excessive sectioning damage, mounting-medium hardness mismatch and shrinkage, and inappropriate abrasives and surfaces during grinding and polishing. Finally, maintaining a clean environment throughout the process minimizes preparation artifacts.

Discuss

About The Author

Sharbari Banerjee’s picture

Sharbari Banerjee

Sharbari Banerjee, Ph.D., currently holds the position of materials engineer, research and development (R&D) and education with Buehler, a division of Illinois Tool Works Inc. In this role, Banerjee identifies new consumable products for metallographic preparation of materials. She is additionally responsible for developing and teaching Buehler’s education courses in the field of metallurgical sample preparation. Sharbari’s past employment includes research and development for an aluminum supplier, Alumax Inc. (now part of Alcoa Inc.) and failure analysis for a packaging manufacturer, Ball Packaging. Banerjee earned a Ph.D. in materials science and engineering from Vanderbilt University after having obtained bachelor’s and master’s degrees in metallurgical engineering.