The growing tension between tensile and Brinell.

The Growing Tension between Tensile and Brinell

Long before he could measure its properties, man learned to take advantage of the benefits of cast iron. Through the ages, craftsmen utilized its properties to create an "artistry" that we recognize now for its sheer skill and experience.

Today we use the same materials with the same union of artistry and utility. The only difference is that we combine modern design with economical manufacturing, newer standards and documented process controls.

In an attempt to define cast irons more narrowly, an inappropriate linking of two basic standards has evolved that can be a point of controversy between foundries and their customers. The tendency for customers to specify both strenght (ASTM specifications) and hardness (SAE specifications) results from design requirements for performance (high strenght), manufacturing requirements for ease of machining (low hardness) and a low cost test method.

In wear applications, or when heat treatment is employed, a microstructure requirement for pearlite also may exist. Often, these requirements conflict. As a result, the competitive advantages both foundry and customer could claim are jeopardized, and the full design advantage of cast iron goes unrealized.

This article attempts to clarify the widely-accepted testing technology currently in use and to provide a new platform for discussion and understanding from which cast iron utilization may progress and grow.

The Cast Iron Family

Cast iron actually comprises a large family of materials with wide-ranging properties. White cast irons are used in mill liners to crush rocks and ores. Cast gray iron engines are machinable, stand up well to piston wear action, contain the forces and temperatures of combustion (over 2000F) and automatically dampen vibration and noise. Ductile iron steering knuckles bend under overloads, much like steel forgings and fabrications.

These different properties of cast iron result not from hardness differences, but rather from (a) microstructural differences in the metal matrix, (b) size, shape and distribution of graphite, (c) intrinsic imperfections and (d) chemical composition.

The family of cast irons can be considered as a compositie material of metal and particles, depicted in Fig. 1 as an oversimplified, three-dimensional grid model. In reality, there are no clear lines of demarcation. Carbon phases are schematically depicted in such a way as to emphazie their influence on properties.

Phases and constituents in cast irons are shown in Table 1, along with their relative hardness, and those of common tool materials. There is no one hardness test method used across this broad range of materials; so the numbers are calculated and extrapolated from what one might expect from a Brinell hardness test.

The wide ranges of hardness possible for a single phase, such as ferrite, come from differences in chemical composition. For example, an increase of 0.20% Si to a ferritic ductile iron having about 2.7% Si, can raise its hardness ten Brinell points. Mixtures of phases are common, and can be at different amounts. This creates many different ways of obtaining a particular hardness value.

Casting design, processing in the foundry and heat treatment all can influence the shape, amount and distribution of the above constituents of cast iron and, therefore, its properties. Through proper process control in the foundry, properties of interest are reproduced.

Hardness Measurements

Hardness measurements are common methods of process control and quality assurance in the materials industry because they relate directly to engineering properties and manufacturing characteristics. However, the metals industry has come to expect too much from the hardness values, especially in view of the composite nature of cast irons noted above.

Most familiar to cast iron producers and users is the Brinell hardness number (HB), followed closely by the Rock-well [(R.sub.c)] the subscript refers to load details.

Brinell hardness is determined by pressing a metallic ball into a material, measuring the indentation or "dimple" and converting the measurement into a Brinell hardness number. The ball and its indentation appear tiny to the human eye, but to the microstructure of the metal, the ball is huge.

As seen in Fig. 2, the ball's contact area covers a wide range of microconstituents. This measurement relates to the bulk system. It cannot accurately indicate the presence of small amounts of any phase that might significantly add to or detract from a material's properties. In addition, many factors influence the accuracy of HB, such as:

* surface flatness and preparation of sample;

* specimen size, thickness, hardness;

* specimen support;

* load variation in the hardness testing machine;

* operator influence;

* definition of the diameter;

* measuring instrument precision;

* amount of material removed from surface (skin effect);

* temperature of sample or part during measurement.

Microhardness testers, such as Knoop or Vickers, can give a better indication of a specific matrix area or particle hardness. However, subsurface differences could influence the reading yet go undetected, as suggested in Fig. 3. Protection from such errors could be ensured by a large number of readings, which would allow for the systematic rejection of the outliers.

When properly applied, hardness numbers are valuable indicators of physical material properties and manufacturing controls. It is important to recognize that hardness can be obtained many ways and that a series of hardness measurements from different areas on one casting can vary significantly. For this reason, additional appropriate measurements and controls are necessary to define specific properties of interest.

Tensile Properties

Tensile properties of cast iron are determined using the same machines as used for steel, but the behavior and results are very different. Specimens are stretched until they break. Progress is plotted on a load vs. elongation curve. Figure 4 shows the curve for a typical steel. Note that the yield point is a well-defined displacement in the curve. After the yield point is reached, the specimen takes on a permanent stretch, the amount of which is related to ductility and toughness.

The curve for cast iron shows no obvious yield point. Figure 5 shows a typical load-elongation curve for gray iron. Gray and white cast irons have no "usable" ductility. On the other hand, ductile iron elongates elastically, then plastically, and necks down before finally rupturing. A typical load-elongation curve for ductile iron is seen in Fig. 6. Unlike steel, however, there is no well-defined yield point.

Yield strength calculations are based on an estimate of the coordinates where the material yields take a permanent set. Usually, the intercept of a 0.2% offset line parallel to the slope of the load-elongation curve is used for the calculation. However, slope selection is subjective, providing an inherent source of measurement variation in yield strength determination. Other sources of variation exist. Figure 7 illustrates possible slope variations that can easily result in 5% difference in yield strength calculations, i.e., the difference between YSA and YSB.

ASTM Standards

ASTM standards for cast iron were written to provide a basis for conducting business between foundries and customers. ASTM cast iron standards vary, but have similarities. In almost every case, the tensile strength standard is written for a separately cast test bar poured from metal representative of that poured into castings. It is a "metal qualification" standard.

The sample casting, procedure and method is constant. Properly made metal will meet the minimum test bar properties. Wisely and correctly, the standards are tempered with a qualification that the relationship between the tensile properties of a casting and a test bar are not absolute. But, rather, are dependent on casting design, section size, processing, procedures, practices and controls. Relationships can only be established empirically. Specifications in Tables 2 and 3 are representative of common gray and ductile irons. Hardness is not specified.

SAE Standards

SAE cast iron standards are based on the Brinell hardness of castings, as shown in Tables 4 and 5 for gray and ductile irons. These are automotive standards for high volume production. Interest is in both assurance of properties (performance) and machining (manufacturability). Notice that the code in the tables contains numbers related to the ASTM grades.

An absolute relationship is implied beween HB and tensile strength in the tables. Wording in the text, however, correctly emphasizes the word "typical" properties for the hardness range. Relationships between HB and any properties or characteristics of interest should be statistically verified, not assumed.

Tensile/Brinell Relationship

Some books show a linear relationship between tensile strength and Brinell. Other, more practical books show the scatter band that exists in reality because of chemical, processing and procedural variations and measurement errors.

Figure 8 shows a typical scatter band between the tensile properties of test bars cut from standard ASTM keel blocks and the hardness of the same broken test bar shoulder. A much wider band would be expected if the plot were obtained by comparing the HB of a casting and the tensile properties of a separately cast test bar, an often implied relationship.

Statistical Process Control

Application of statistical process control to all aspects of the materials business has quickly brought into focus the difficulty inherent with combining inapppropriate tensile/Brinell relationships. Consider the data band in Fig. 8, which represents random samples from a few month's production of ductile iron in one plant and the ASTM standard for 80-55-06 ductile iron material:

* tensile strenght: 80,000 psi minimum;

* yield strength: 55,000 psi;

* elongation: 6% minimum.

Arrow 1 shows HB could be no less than 207 to ensure the 80,000 psi minimum tensile strength at all times. Arrow 2 shows HB could be no less than 232 to ensure a 55,000 psi minimum yield strength. Arrow 3 shows the HB could be no more than 187 to ensure 6% elongation at all times. This is an impossible situation!

Again, looking at Fig. 8, note the range of tensile or yield strength at any HB and the range of HB for any tensile or yield strength. Consider an HB spec range of 187-255. Using the data bands in Fig. 7, a wide range of tensile properties could be expected:

* tensile strength: 75,000-125,000 psi;

* yield strength: 45,000-75,000 psi;

* elongation: 2-18%.

The spread is a cumulative result of measurement error and variations in metal microstructure, chemical composition and processing.

Because of data bands like those in Fig. 8, major national standard organizations have always avoided directly linking HB of castings to the tensile properties of separately cast test bars. Yet, pressures of business demand such relationships to the point that "typical" properties become absolute expected properties. Tensile/HB relationships are not absolute, and may in fact be quite different from foundry to foundry.

Recommendations

It is time to recognize and address the limitations of hardness measurements. Wide discrepancies exist between expected and realized properties based on HB strain relationships between foundries and their customers. Designers either avoid cast irons or will assign large safety factors resulting in heavy and often uncompetitive castings. Both foundries and users lose in such a situation. Better measurement of cast iron properties is necessary for increased utilization of cast iron.

It is easier to be critical of a situation than it is to solve it. The intent of this article is to define a problem in the hope that the industry can begin moving toward a solution. In the meantime, some interim recommendations are offered:

* recognize and stop immediately the inapproriate linking of SAE Brinell hardness numbers and ASTM tensile properties;

* use SAE standards when Brinell hardness is the main property of interest, when "typical" property wll suffice or when the customer has statistically significant confidence that the property of interest is ensured by the Brinell hardness;

* use ASTM standards to establish "metal qualification," i.e., that the correct metal has been produced. Any relationship to casting properties should be established by statistically significant data relating test bars to castings or to samples from, or attached to, castings;

* aggressively pursue the development andincorporation of nondestructive methods to measure and verify the specific properties and characteristics of interest to designers and manufacturers of cast iron components;

* work to improve the property/HB relationships by reducing measurement error and using better process control.

References

R. W. lobenhofer, "Beware of Uninformed Application of SPC Hardness," modern casting, Jan 1988.

J. F. Janowak, "Cast Iron Metallurgy for Improved Machinability," ASM International Conference on High Productivity Machining, Materials and Processing, 8503-006, 1985.

J. F. Janowak, R. B. Gundlach, K. Rohring, "Technical Advances in Cast iron Metallurgy," International Foundry Congress Official AFS Exchange Paper, 1981.

A. Alagarsamy, Unpublished data, Grede Foundries, Inc.