Characterization The microstructure of each sample set was characterized using both optical (Olympus GX51) and scanning electron (Quanta 200 and Sirion) microscopy. Sections were taken both transverse to the pin length (across the pin diameter) and longitudinally along the pin length. Serial sectioning performed by successive grinding was used to reveal the microstructure gradient from the pin surface to the pin center. This technique was particularly useful for determining the extent of crack backfilling in certain compositions. All samples were then polished and electrolytically etched in a 10% chromic acid solution for 5–10 s at 5 V and 2 A with a tungsten anode and stainless steel cathode. The area fraction eutectic was determined via image analysis using ImageJ software. The procedure was developed from work by Payton et. al. (Ref. 47) and conformed to ASTM E1245-03. SEM/EDS was used to determine the composition of the matrix and eutectic constituents. An accelerating voltage of 25 kV, spot size 6, and probe current 10 nA were utilized at magnifications ranging from 200 to 6000¥. EDS analysis was performed at 800¥ on samples that had all been etched in chromic acid in order to reveal the eutectic constituents and solidification structure in the samples. Etching prior to EDS analysis facilitated locating and analyzing the eutectic constituents. Results Solidification Modeling Results from Scheil simulations for all compositions are shown in Table 2. Partition coefficients were calculated by dividing the composition of the first solid to form by the initial, nominal composition in terms of either Nb or Mo. The available literature does not report partition coefficients for Alloy 690, but the calculated values used seem to be consistent with other studies of Nb- and Mo-bearing Nibased alloys (Refs. 8, 9, 48–51). A small under/over estimation of the partition coefficients may decrease the accuracy of the fraction eutectic calculations, but the effect should be minor. As both the Nb and Mo alloying additions increase, the partition coefficients increase slightly (less partitioning). The fraction eutectic was predicted to increase with increasing alloy content, particularly with respect to Nb. Molybdenum additions produced only a small change in the calculated fraction eutectic, but did result in a large change in the final calculated composition of the eutectic phase since Mo partitions to the liquid during solidification. The calculated fraction eutectic values are reported as mass fraction while the measured fraction eutectic values are volume fraction (based on area fraction measurements). These mass fraction values are reportedly comparable to volume fraction values as the densities of the phases involved are similar (Ref. 9). Regardless, the calculated fractions correlate well with the measured values. The calculated average weight fractions of niobium and molybdenum were later compared to Nb and Mo levels measured in backfilled regions using EDS. Cast Pin Tear Test The critical cracking thresholds for the niobium-bearing samples are shown in Fig. 2. The base composition, Alloy 690 with no Nb, and the alloy with 8 wt-% Nb exhibited the lowest susceptibility to solidification cracking, as indicated by threshold (no cracking) pin lengths of 1.5 and 1.375 in., respectively. The alloy with 4 wt-% Nb addition showed the highest susceptibility with a threshold pin length of 0.75 in. The scatter in the CPTT data tends to increase with an increase in pin length above the threshold and with increased Nb content. This scatter is likely due to some variation in the amount of eutectic that is present and the ability of cracks to heal by a backfilling mechanism when greater stress (pin length) is imposed on the system. For that reason, the threshold pin length at which there is no cracking is used as the most reliable indicator of resistance to solidification cracking. Samples representing the critical cracking thresholds determined in Fig. 2 were then prepared for image analysis. Comparison of their measured volume fraction eutectic as well as calculated mass fractions as a function of Nb content is shown in Fig. 3. In order to be consistent with other data showing the effect of composition on cracking susceptibility in eutectic systems, such as that shown in Fig. 1, susceptibility is represented by the reciprocal of the critical cracking threshold (Refs. 29, 30, 34, 36). The larger this value, the more susceptible to cracking relative to the other plotted compositions. This plot clearly illustrates that the highest susceptibility is exhibited by the 4 wt-% Nb composition. When considering the fraction eutectic, it appears that the highest cracking susceptibility occurs at levels just below 10%. Above 10%, cracking susceptibility decreases due to an apparent crack healing effect, as will be shown by characterization results in the next section. Additional data at 3 and 5 wt-% Nb are needed to more ful- WELDING RESEARCH 234-s WELDING JOURNAL / JULY 2016, VOL. 95 Fig. 11 — Cracking susceptibility versus solidification temperature range based on Varestraint data from DuPont et al. (Ref. 54). Fig. 12 — Cracking susceptibility and predicted solidification temperature range for the Nbbearing alloys at 0.95 fraction solidified.
Welding Journal | July 2016
To see the actual publication please follow the link above