Statistical Comparisons & Correlations of Thermal Analysis, Ultrasonic Velocity and Image Analysis Metallography Methods for Quantification of Ductile Iron Microstructure

This paper will present results of a thorough evaluation of the effects of (1) 'plain' thermal analysis (TA) cup (aka 'eutectometer' cup, as coined by Heine in AFS Trans 78:187-193, 1970; AFS Cast Metals Res J 7:49-54, 1971; AFS Trans 80:462-470, 1973) design/geometry (square vs. round), (2) 'before' and 'after' paper cup inoculation of plain TA cup, (3) inoculant type, and (4) inoculant dosage on the resultant TA metrics and microstructure characteristics as quantified by image analysis metallography of treated and inoculated ductile iron. Since production castings have near-infinite shapes and sizes, there is often a large variation in the types and dosages of inoculants employed by ductile iron casting producers. Thus, the evaluations of this paper are important in quantifying the proportions of gage variation (discriminating capability) for all independent and interactive combinations of the foregoing four predictor variables. That is, the gauge variation for any combination of predictor variables must be sufficiently low to make any valid assessment(s) of them having significant effects on the responses of TA metrics and microstructure characteristics. Further analysis was done in probing for any significant correlations between TA metrics and the ultrasonic velocities and quantified microstructure characteristics of both mold-cooled and water-quenched coupons after casting, with presentation of results for the significant correlations. When GR&R variation for the effects of any combination of all the TA predictor variables is sufficiently low, we can then make valid assessment(s) of their effects on and correlations to ultrasonic velocity and digital metallographic image analysis metrics of test coupons that were cast simultaneously with TA cups in multiple sets cast from a single sample spoon.