Geometry impacts aluminum tensile test bar results.

Designers of cast aluminum automotive parts rely a great deal on material property data from test bars taken directly from castings. Many of these castings are complex and involve a wide range of section thicknesses and cast features. As has been determined in many studies, material properties can vary widely across any given casting. Parameters such as cooling rate and pouring temperature play a significant role in final part properties by directly affecting the microstructure.

Tensile and fatigue tests are usually conducted on test bars from castings during development and for periodic checks in production. Materials engineers are faced with the challenge of choosing a location in the casting from which an ASTM standard test bar can be extracted. Typically, subsize ASTM or nonstandard test bars are the only choices in complex castings. Engineers must also decide whether to use a flat or round test specimen. While it is generally agreed that specimen size and shape influence test results, there is little agreement or documentation, however, of the degree of this influence.

Procedure and Results

The GM Powertrain Casting Technology Center became interested in the influence of test specimen design and location for aluminum castings as more and more of these parts were produced. To gather background data, a series of special parts were production-cast over 2 mo. Then, 20 of these castings were randomly selected to yield a more respective sample variation in production over time.

Both a flat and round bar were selected according to the material available in the sampling locations. Both test bar designs are regularly used at various GM facilities, however, only the flat bar design meets ASTM B557 specifications. Each bar was chosen to use the maximum amount of material from the sampling location - which resulted in a 20% larger cross sectional area for the round bar vs. the flat bar. Two thick-bolt boss sections (locations A and B) were marked for sampling.

One flat bar was machined from location A or B from each casting in a random manner. A round bar was machined from the remaining thick boss section, A or B. This sampling scheme resulted in a total of 20 flat bars and 20 round bars. The tensile tests were conducted according to ASTM B557.

Table 1 summarizes the averages and standard deviations for each of the sample groups.

Table 1. Average and Standard Deviations for Flat and Round Test
Bars

Test Group       Test Bar     Average     Standard Deviation

Location A&B       Flat       22,983            3211
Location A&B      Round       27,156            1701

Examination of the test bars revealed shrinkage and gas porosity on the fracture surfaces from locations A and B.

Shape and Size

The study suggests that a cast-in feature in the thick sections affected the difference in tensile results between the flat and round specimens in two ways. First, the average tensile strength was significantly lower for the flat tensile bars. Second, the round tensile bars produced a more consistent result, with a standard deviation nearly one-half that of the flat bars.

Failure of aluminum castings, as well as failure of test bars taken from such castings, begins at a discontinuity or surface defect. Evaluation of the data suggests that the variance between test groups in this study relates directly to discontinuities. In order to properly represent the material properties in a particular area of an aluminum casting, the test bar cross section and shape must be such that the base material strength, and not a discontinuity, is the major factor in the final outcome of the test.

Since it is impossible to control the size and location of the discontinuities (if we could do this, we wouldn't need testing), it is crucial to use a test bar design that gives a high degree of confidence in the resulting data. Both the flat and round bar designs chosen in this study were appropriate for the casting section to be sampled. However, the bar with the larger cross sectional area showed higher average properties and more consistent results from test to test.

Furthermore, fundamental geometric differences between flat and round bar designs bring out other considerations. For example, a round bar will have a smaller perimeter (circumference) than a flat bar with the same cross sectional area. Couple this fact with the random, unpredictable nature of casting discontinuities (such as networked shrinkage porosity), and the flat bar has a greater chance of containing a surface or near-surface discontinuity, which acts as a notch. It can be argued that the discontinuities at or near the corners of the rectangular-shaped gage section of a flat bar may have a much different effect on test results than a discontinuity of identical size in any other location near or at the flat or round bar gage surface

Location Isn't Everything

Test bar size and shape are just as important as sample location when gathering material properties from aluminum castings. Maximizing the test bar cross-sectional area - easily accomplished by choosing a round design - should be an issue in each testing program. Paying close attention to factors such as this one will help ensure more meaningful and consistent data on which to base design and development decisions.