As part of the development of the Bar Steel Fatigue Database, AISI undertook a limited investigation to examine the amount of scatter that might be expected from fatigue data. Of particular concern was scatter that might result due to securing steel bars from different sources, and due to using different heat treating concerns for processing. While fatigue testing procedures have followed ASTM standards, it was recognized that variability in test results might still occur due to variations in steel product and processing.
Quenched and tempered SAE 4140 steel was selected for the study. Steel bars within the specified composition range for SAE 4140 were obtained from two different sources, and heat treating was carried out by three different concerns. The aim hardness specified for the test samples was HRC 40.
Table I shows the mechanical properties and hardness values for all six test iterations involved in the study. The steel bars for Group A (Iteration Nos. 64, 65, and 66) were secured from a single steel source and the bars for Group B (Iterations Nos. 67, 68, and 69) were secured from a second steel source. Each of the three heat treating concerns was utilized for one test iteration in each group of bars.
|Group||Iteration No.||Yield Str.
|Red. in Area
All of the microstructures were tempered martensite, and the hardness values actually achieved ranged from HRC 38 to 42. The mechanical properties for all the test iterations fell in a fairly narrow range.
Figure 1 shows the strain-life curves for Group A and Figure 2 shows the strain-life curves for Group B. The data in both Figures 1 and 2 lie in a narrow band. Furthermore, a comparison of Figure 1 with Figure 2 shows that there is little difference between both groups of steels. Thus, all six test iterations exhibit equivalent results even though two steel sources and three heat treating concerns were involved in preparation of the test specimens.
The low scatter in this data suggests that fatigue results in the AISI Bar Steel Fatigue Database will not be influenced by changes in the sources for steel products or processing. These results also provide an increased certainty in using all the data in the database for design purposes.