Search

Sorting of materials at different stages of manufacturing processes is very important to ensure the final product quality. In this paper, the machine learning algorithms are used on the magnetic Barkhausen emission (MBE) signals for the classification of martensitic stainless-steel specimens, which are subjected to different heat treatments. The variation in MBE parameters, such as root mean square, peak height, peak position, and full width at half maximum, was found to be marginal for normalized and quenching and partitioned specimens even though their x-ray diffraction analysis showed distinct microstructural conditions. The conventional MBE parameters are used to train the machine learning models. The performance of decision tree and ensemble learning techniques, such as Bagging, Random Subspace, AdaBoost, RUSBoost, Total Boost, and LP Boost classifiers, is compared and found that the AdaBoost classifier provides the maximum accuracy of 98% in the classification of complex materials. The results showed that the machine learning approach using a limited number of features is sufficiently sensitive toward the classification of specimens that are otherwise indistinguishable in their conventional MBE response. [ABSTRACT FROM AUTHOR]

Many isothermal equations of state lead to numerical problems under tension conditions due to negative bulk modulus. We summarize the properties of four existing models and devise a continuous non-piecewise reference curve to address this issue. The final functional form contains three fitting parameters that are directly related to the bulk modulus and its pressure derivative at the reference state and at infinite pressures. We also include bounds that guarantee the monotonicity of its pressure derivative in all volume space. To demonstrate the qualitative features of each isotherm, we conduct calibrations against experimental data for cyclotetramethylene tetranitramine, magnesium oxide, gold, and platinum. Pressure residuals and root mean square deviations show that the proposed functional form is capable of adequately fitting the data compared to other three-parameter models. [ABSTRACT FROM AUTHOR]

The wet etching of germanium-tin (Ge1-xSnx) alloys (4.2%

Atomic- or nanometer-scale surface roughening has been investigated during Si etching in inductively coupled Cl2 plasmas, as a function of rf bias power or ion incident energy Ei, by varying feed gas flow rate, wafer stage temperature, and etching time. The experiments revealed two modes of surface roughening which occur depending on Ei: one is the roughening mode at low Ei<200–300 eV, where the root-mean-square (rms) roughness of etched surfaces increases with increasing Ei, exhibiting an almost linear increase with time during etching (t<20 min). The other is the smoothing mode at higher Ei, where the rms surface roughness decreases substantially with Ei down to a low level<0.4 nm, exhibiting a quasi-steady state after some increase at the initial stage (t<1min). Correspondingly, two different behaviors depending on Ei were also observed in the etch rate versus Ei√ curve, and in the evolution of the power spectral density distribution of surfaces. Such changes from the roughening to smoothing modes with increasing Ei were found to correspond to changes in the predominant ion flux from feed gas ions Clx+ to ionized etch products SiClx+ caused by the increased etch rates at increased Ei, in view of the results of several plasma diagnostics. Possible mechanisms for the formation and evolution of surface roughness during plasma etching are discussed with the help of Monte Carlo simulations of the surface feature evolution and classical molecular dynamics simulations of etch fundamentals, including stochastic roughening and effects of ion reflection and etch inhibitors. [ABSTRACT FROM AUTHOR]