Ab initio virial equation of state for argon using a new nonadditive three-body potential

An ab initio nonadditive three-body potential for argon has been developed using quantum-chemical calculations at the CCSD(T) and CCSDT levels of theory. Applying this potential together with a recent ab initio pair potential from the literature, the third and fourth to seventh pressure virial coefficients of argon were computed by standard numerical integration and the Mayer-sampling Monte Carlo method, respectively, for a wide temperature range. All calculated virial coefficients were fitted separately as polynomials in temperature. The results for the third virial coefficient agree with values evaluated directly from experimental data and with those computed for other nonadditive three-body potentials. We also redetermined the second and third virial coefficients from the best experimental pρT data utilizing the computed higher virial coefficients as constraints. Thus, a significantly closer agreement of the calculated third virial coefficients with the experimental data was achieved. For different orders of the virial expansion, pρT data have been calculated and compared with results from high quality measurements in the gaseous and supercritical region. The theoretically predicted pressures are within the very small experimental errors of ±0.02% for p ≤ 12 MPa in the supercritical region near room temperature, whereas for subcritical temperatures the deviations increase up to +0.3%. The computed pressure at the critical density and temperature is about 1.3% below the experimental value. At pressures between 200 MPa and 1000 MPa and at 373 K, the calculated values deviate by 1% to 9% from the experimental results.     

Strength and fatigue of some binary and ternary silicate glasses

A survey represented mainly by means of glasses of the borosilicate system shows the generally valid limits of the applicability of the crack growth model and the essential influence of the corrosion of glasses on fatigue. A close correlation between macroscopic crack growth and fatigue was affirmed for samples with fresh, slightly corroded cracks. If the state of corrosion of the samples is advanced, the influence of region 2 is reduced and the fatigue limit is increased. Processes of corrosion during the measurement influence the fatigue curve prevailingly in the range of great time-to-failure, in particular the slope in region 1. The results indicate the high complexity of the effective chemomechanical processes during the fatigue measurements.