Date of Award


Degree Type


Degree Name

Doctor of Philosophy


Engineering Science

Major Professor

C. R. Brooks

Committee Members

Carl McHayes, B. S. Bowie, E. E. Stansbury, J. E. Spruiell, Lewis Nelson, Harry C. Jacobsen


The specific heat capacity at constant pressure Cp of a ferromagnetic solid is assumed separable into six contributions - that due to the dilation of the lattice Cd, that due to the harmonic oscillations of the atoms Cvh, that due to the anharmonic oscillations of the electrons Cva, that due to changes in energy distributions of the electrons Cve, that due to changes in the magnetic coupling among the electrons Cvm, and that due to atomic rearrangements Cvw. Expressions appearing in the literature for these contributions, which are assumed to be independent, are reviewed and the validity of each is discussed.

Specific heat capacity measurements performed in one of two pulse calorimeters are presented for iron, nickel, and ordered and disordered Ni3Fe from 300 to 1400 K. Pulse Calorimeter II, having been developed during this research, is described in detail. Primary attributes of this apparatus are accurate and simultaneous measurement of the specific heat capacity (±1.05%) and electrical resistivity (±0.48%) at temperature intervals of 0.2 degree while the specimen is heated 100 to 600 degrees above its initial temperature at rates of 5 to 60 degrees/second. All specimens are in the form of rods (0.3-cm nominal diameter) and are of purity greater than 99.8%, contain no voids, and have a minute trace of second phase - possibly an oxide.

Comparison of the Cp values on nickel and iron with those in the literature indicates two regions of significant disagreement - at high temperatures and near the Curie temperature Tc of each. Both of these discrepancies are attributed to experimental errors in thermometry and/or methodology. Examples for nickel in iron encountered in this work are given to illustrate these errors.

Expressions for Cvd, Cvh, Cva, and Cve are chosen from the literature and used in conjunction with the Cp data of this investigations to calculate Cvm of nickel and iron as a function of temperature T. Some of the parameters requisite to these expressions are taken from the literature and some are obtained from the data treatment. The latter parameters are found to be in reasonable agreement with theory. The Cvm values calculated for nickel and iron are within experimental error of literature values based on experimental calculations. Appropriate integrations of the Cvm: T relationships yield values of the magnetic energy Uvm and entropy Svm which are in agreement with experimental and theoretical treatments of the literature of both materials. Values of the discontinuity ΔCvm in Cvm at the Curie temperature Tc agree with theoretical predictions for this second-order transformation. Plots of Cvm for nickel and iron versus log10 |T - Tc| do not demonstrate the behavior expected of critical phenomena.

The Cp of Ni3Fe is found to be a function of the thermal history and temperature of the specimen and of the experimental heating rate for temperatures between 750 and 1050 K. Below 750 K, the Cp decreases as the long-range order of the alloy increases. From theoretical interpretations of Cp data of this investigation for ordered and disordered Ni3Fe between 1.2 and 4.4 K, this change is attributed to a decrease of the density of electron states and to an increase of the Debye temperature upon ordering. Use of rapid heating rates allows measurements on ordered structures of Ni3Fe at temperatures where they are metastable, These measurements reveal that the Curie transformation of a highly ordered metastable structure of Ni3Fe is 70 degrees above that of the disordered structure and that the order-disorder transformation can be suppressed to temperatures 230 degrees above the equilibrium transformation temperature using heating rates of 60 degrees/second. In spite of the suppression, it is possible to calculate Cvm only for the disordered structure of Ni3Fe because the ordered structures could not be maintained to high enough temperatures for completion of the Cvm analysis.

The Cvm:T relationship for the disordered structure is obtained in the manner analogous to that for nickel and iron. Values of Uvm and Svm found from these calculations are in excellent agreement with theoretical predictions, whereas the ΔCvm value is not. The energy associated with transforming completely ordered Ni3Fe to completely disordered Ni3Fe at the equilibrium temperature is calculated. Only estimates of the entropy of this transformation could be made since the transformation is carried out in an irreversible manner. The values of the energy and entropy are within experimental error of those found in the literature.

Recommendation for further research are listed.

Files over 3MB may be slow to open. For best results, right-click and select "save as..."

Included in

Metallurgy Commons