Date of Award
Doctor of Philosophy
Materials Science and Engineering
The temperature dependence of the isobaric specific heat (Cp) and electrical resistivity (p) were simultaneously determined for several metallic alloy systems. The principle method to obtain the data was by using a pulse-heating calorimeter (PHC). The usefulness of the PHC in obtaining these two thermophysical properties was demonstrated. The theoretical basis of operation of the PHC was discussed. In addition, some advantages and limitations of using the PHC were demonstrated, and recommendations to improve the versatility and quality of the data were noted. The primary function of the PHC is to simultaneously obtain Cp and p from approximately 300 K to a maximum of 1500 K. In many cases, the Cp and p data can be further used to deduce other thermophysical properties. The main unique feature of the PHC is that it obtains Cp and p data at relatively high heating rates (typically 25 to 350 K/s). In addition, the data can be obtained within a relatively short time. The high heating rates in many cases make possible measurement on metastable phases, and can often provide more insight into the kinetics of phase transformations than heating under equilibrium conditions. The PHC is also capable of operating in an isothermal mode, which aids in understanding kinetic processes. These attributes were utilized to investigate a variety of solidstate phase transformations. Some of the types of solid-state transformations studied include various ordering, magnetic, and allotropic types. Other types include glass transition and crystallization behavior of an initially amorphous alloy, and defect properties in intermetallic compounds. The metallic systems investigated were the Ni4Mo intermetallic compound, Fe-30 at% Al and Fe-43 at% Al intermetallic alloys, pure Co, a Zrs52.5Ti5Cu17.9Ni14.6Al10 bulk amorphous alloy (BAA), a Co-based commercial alloy (ULTIMET™) and a Ni-base commercial alloy (Hastelloy™ G30). In several studies, a literature review of pertinent research was provided. The kinetics of the long-range order (LRO) - short-range order (SRO) phase transformation that occurs in the Ni4Mo intermetallic compound were investigated from 300 to 1400 K. The equilibrium order-disorder temperature (Tc) occurs at 1141 K. The phase transformation was studied by obtaining Cptemperature data and p-temperature data on both the SRO (α) and LRO (β) phase initial conditions, using average heating rates between approximately 30 and 450 K/s. Temperature coefficient of resistivity (TCR) data were also obtained. The effects of heating rate on the data were discussed. Data on the equilibrium a phase were obtained above Tc, and data on the metastable a phase were obtained from 300 to 950 K. Data on the equilibrium p phase were obtained from 300 K to Tc, but data on the metastable p phase significantly above Tc were not obtained. Some minor superheating (about 50 K) of the p phase was achieved. The temperature dependence of the Gibb's free energy change of the SRO a ➔ LRO p transformation (ΔGα➔ β) was determined. The calculated ΔGα➔ β values closely agree with an estimate of ΔGα➔ β (obtained by neglecting the ΔCp term) in the temperature range 1100 to 1300 K. A time-temperature-transformation (TTT) diagram for the start of the α➔β phase transformation was determined between 923 and 1113 K. The TTT curve for the 3% start of the transformation generally exhibits a 'C' shape, but is much steeper at high temperature, and does not appear to approach the equilibrium disorder-order temperature asymptotically. As the transformation proceeds at higher temperatures there is a slight deviation from a smooth 'C' shape, and there is a minor indication of a double 'nose'. The shortest time of the 3% transformation start is about 85 s at 1053 K, consistent with other TTT data from literature. Between 923 and 1038 K, 1800 s isothermal holds are not sufficient to complete the α➔β transformation. Above 1038 K the transformation may have completed within 1800 s. The isothermal p-time data obtained at temperatures between 923 and 1113 K display a slight maximum prior to the a major drop in the data. The major drop is associated with the α➔β phase transformation. The slight maximum is consistent with results of other researchers who attributed it to the SRO structure obtaining a critical wavelength before β begins to nucleate. Long-range order parameter (S) data for the β phase were determined as functions of temperature and time using a model based on p data. The S-temperature data tend to attain a maximum of S =1 (indicating perfect order) at about 1050 K. The lower temperature S data were evidently not obtained under equilibrium conditions. Above 1050 K, S decreases with increasing temperature, approaching zero at Tc. The temperature dependence of p and Cp were determined for Fe-30 at% Al and Fe-43 at% Al binary alloys from 300 to 1400 K. The heating rates used to obtain the data were between 50 and 350 K/s. Data were contrasted between two pre-treatments. One pre-treatment was a step-cooled condition, which allowed a very low initial vacancy content. The other pre-treatment was naturally cooling in the calorimeter from high temperatures at an average cooling rate of about 4 K/s. In the Fe-30 at% Al alloy, the DO3 to B2 phase transformation was detected in both p and Cp data. The magnetic transformation and the disordering reactions established in the literature were not detected in the PHC data. The general shape of the p-temperature curve for the Fe-30 at% Al alloy agrees with other researchers. The p-temperature data for the 30 % Al alloy was compared to long-range order parameter (S)-temperature data of alloys with similar compositions found in the literature. The onset of the maximum and the cusp in the p-temperature data seem to correspond to the sharp drop in S when approaching the DO3 disorder temperature. Thus the p behavior in this region may be due to disorder. The p-temperature data for the 43 at% Al alloy generally increased with continuously decreasing slope. There were slight effects occurring around 1100 K, which may have been attributed to vacancy generation and/or dissolution. The enthalpy of formation of the triple defect structure in the Fe-43 at% Al alloy was determined to be 110 kJ/mol, in agreement with the literature. The triple defect and vacancy concentrations were determined as a function of temperature for the 43 at% Al alloy. Both properties agree well with those on similar compositions from various researchers. The temperature dependence of Cp and p of pure Co from approximately 300 to 1550 K were determined, and in good agreement with the literature values. Transformation temperatures were determined to be about 705 K for the HCP ε to FCC α allotropic transformation and about 1370 K for the Curie temperature. These values are well within the large spread of values reported in the literature. The heating rate varied during the tests from about 1 0 K/s at 300 K to 50 K/s K at 1500 K. There was a sharp maximum in the Cp-temperature data associated with the HCP E to FCC a phase transformation that was in agreement with only one of several other studies. The detection of the peak in the Cp-temperature data from the present investigation may be attributed to the ability of the PHC to sample several data in a very short temperature interval. The Cp-temperature data in the region of the Curie temperature exhibited a sharp peak, in close agreement with other data in the literature. Isothermal annealing below the allotropic transformation temperature had the effect of raising the transformation temperature on a subsequent pulse-heating test, based on p-temperature measurements. The Cp-temperature and p-temperature data of UL TIMET™ were obtained from 400 to 1300 K with the PHC using heating rates between 35 and 150 K/s. The Cp-temperature data obtained with the PHC increase monotonically with temperature, and show no indication of a phase change. There is a slight change in the slope in the p-temperature curve near 975 K. In contrast to the Cp-temperature data obtained with the PHC, the Cp-temperature data obtained by DSC using a much lower heating rate (0.33 K/s) exhibit deviation from smooth behavior between 825 and 975 K, which is attributed to the formation and then dissolution of the HCP ε phase. The higher heating rate viii of the pulse-heating calorimeter prevents this from occurring. Simplified estimates of the E to a transformation temperature based on phase relations and solute-effect data in the literature do not agree with the transformation temperature based on Gp-temperature results from the DSC and p-temperature data from the PHC. The temperature dependence of Cp and p on the alloy Hastelloy™ G-30 were determined from 300 to 1400 K heating rates between approximately 20 and 100 K/s. The results are briefly discussed in terms of approximate Ni-Cr-Fe ternary and Ni-Cr binary phase equilibria. In addition, p-time data were obtained isothermally between 775 and 1475 K and for times between 240 and 4200 s with the objective of detecting the formation of a-phase. If any a-phase did form, it was not detected with the PHC. The p-temperature data on Hastelloy™ G-30 obtained with the PHC are in good agreement with literature values up to 875 K. Both Gp-temperature and p-temperature data indicate a structural change between approximately 875 and 1100 K, the details of which are unexplained. The p-temperature data show a distinct plateau of nearly constant p that begins between 875 and 925 K. The effect of heating rate on p-temperature data was insignificant when using heating rates between about 20 and 85 K/s. Cptemperature data exhibit a very distinct inflection in slope between 875 and 1100 K, followed by a sharp upswing. At the higher heating rates, the inflection at intermediate temperatures appears to be partially suppressed. The effect of increased heating rate is to shift the sharp increase to higher temperatures. The crystallization behavior of a Zrs52.5Ti5Cu17.9Ni14.6Al10 bulk amorphous alloy (BAA) was investigated by pulse-heating the as-cast (amorphous) alloy from room temperature to near the melting temperature (1069 K). The alloy was also pulse-heated in the initial crystalline condition using similar heating rates. The PHC simultaneously determined the temperature dependence of Cp and p at high average heating rates between 25 and 220 K/s. The actual maximum heating rates obtained during the tests were during crystallization, and ranged ix from about 1800 K/s to 2200 K/s in the crystallization temperature range. The melting temperature obtained from temperature-time data on specimens that melted was found to be between 1058 and 1069 K, the latter of which agrees exactly with a value reported in the literature for this particular alloy composition. The glass transition temperatures, crystallization temperatures, supercooled liquid regions, and reduced glass transition temperatures (TG, Tx, ΔTx, and TRG respectively) were obtained from the Cp-temperature data obtained on material in the as-cast condition at the different heating rates. The TG appeared to increase from about 620 to 690 K with increased heating rates. The crystallization temperature (Tx1, determined as the start of the major minimum in Cp-temperature data) was about 810 K, relatively independent of heating rate at these¬ high heating rates. The Tx2 temperature (selected as the temperature at the sharp increase in the Cp data at the end of the crystallization minimum) was also relatively independent of heating rate at about 1010 K. The ΔTx obtained at these higher heating rates ranged from about 125 to 195 K. This is much wider than most other values reported in the literature. In addition, ΔTx was found to decrease with increased heating rate. The inverse temperature relation is in contrast to data in the literature, which were obtained at much lower heating rates (between 0.083 and 1.33 K/s). The TRG ranged from about 0.58 to 0.65 and was heating rate dependent, with magnitude and heating rate dependence consistent with the literature. The values of p at 300 K were relatively high (about 180 µΩ cm) in the as-cast condition. The p-temperature data for all four tests on the initial as-cast material in the temperature range agreed within about 5% between 300 about 750 K. The average temperature coefficient of resistivity (TCR) between 300 and 750 K is small and negative ((-8.5 +- 0.7) x 10-5 K-1 ). The p for initially crystalline material at 300 K is about 115 µΩ cm, which is about 40 % below p of the material in the as-cast condition. With increased temperature, p increases with a continuously decreasing slope. In the crystalline state, the material also has a small (but positive and variable) TCR of approximately +3 x 10-4 K-1 between 300 and 1000 K. In the initial as-cast condition a broad plateau and/or a minimum in the p-temperature data occurs between about 800 and 1000 K, but p during the minima from the highest two heating rates tests is much larger than p from the slower heating rate tests. Electrical resistivity data above 1000 K from the slower two heating rate tests agree closely with the data obtained on material in the initially crystalline condition at these temperatures. The unexpected sharp upswing in the p-temperature behavior of the higher heating rate tests is interpreted as being due to some type of (unidentified) precursor structure that formed prior to crystallization. The precursor structure that formed has a larger p than both the amorphous and the crystalline structures. The two slower heating rates allowed sufficient time for crystallization, and the alloy did not form the precursor structure. The Cp-temperature data between 350 and 625 K remains almost constant at about 0.36 J/gK with a slight increase with temperature and the data in this temperature range are independent of whether the alloy is in the as-cast or crystalline condition. Above 625 K, the Cp-temperature curve of the initial crystalline material exhibits a smooth increase in Cp with increased temperature. Cp-temperature data in the as-cast condition above 625 K exhibit a general broad type of maxima between about 700 and 820 K, although the individual shapes of the maxima vary greatly. Beyond the wide maxima in the Cptemperature curves there is an abrupt decrease in Cp and then a broad minimum occurs between about 820 and 1000 K. The minimum in Cp drops to very low values (almost zero) in each case. The minima are followed in each case with a sharp upswing in the curves, which occurs at about 1010 K. Microstructure analysis (SEM) and microhardness data were obtained subsequent to pulse-heating and cooling on one specimen, one end of which melted and re-solidified, and the other end of which never exceeded ambient temperature, and thus was assumed to remain in the as-cast condition. For material assumed to be in the amorphous (as-cast) condition, the hardness is about 560 VHN (5.9 GPa). The hardness increased to about 720 VHN (7.6 GPa) in the crystalline (or partially crystalline) state. The hardness of material near the melted end had a hardness of about 605 VHN (6.4 GPa). The quantitative temperature gradient was unknown, but various microstructures were formed at different locations on the specimen. SEM images in the region near the end of the specimen that melted indicated both rod-like particles and plate-like particles with the rod-like particles being about 2-3 µm long, and 0.1 to 1 µm in diameter. Images obtained in another location appear to be an intertwined two-phase structure. In other regions, spherical like discrete particles were observed and but their size and frequency differed with location, indicating possible nucleation and coarsening processes have occurred. The spherical particle sizes observed ranged from about 0.05 and 0.2 µm. Since these particles are more discrete and spherical than the inner-twined structure, this indicates that coarsening processes did not form the structure in regions subjected to higher temperature, and there thus appears to be a change in mechanism that occurred. Images in one particular narrow location additionally show some relatively large particles present, ranging from 1 to 10 µm in size.
Falcon, Douglas Stewart, "Thermophysical properties of metallic systems determined by pulse-heating calorimetry. " PhD diss., University of Tennessee, 2003.