Engineering Papers⌕ Search

Engineering topics

Henry, A.

Publications and source records attributed to Henry, A..

Exploring model complexity in machine learned potentials for simulated properties

Abstract Machine learning (ML) enables the development of interatomic potentials with the accuracy of first principles methods while retaining the speed and parallel efficiency of empirical potentials. While ML potentials traditionally use atom-centered descriptors as inputs, different models such as linear regression and neural networks map descriptors to atomic energies and forces. This begs the question: what is the improvement in accuracy due to model complexity irrespective of descriptors? We curate three datasets to investigate this question in terms of ab initio energy and force errors: (1) solid and liquid silicon, (2) gallium nitride, and (3) the superionic conductor Li $$_{10}$$ 10 Ge(PS $$_{6}$$ 6 ) $$_{2}$$ 2 (LGPS). We further investigate how these errors affect simulated properties and verify if the improvement in fitting errors corresponds to measurable improvement in property prediction. By assessing different models, we observe correlations between fitting quantity (e.g. atomic force) error and simulated property error with respect to ab initio values. Graphical abstract

Rohskopf, A. (ORCID:0000000227128296)↗

No correlation of the Lyman continuum escape fraction with spectral hardness

The properties that govern the production and escape of hydrogen-ionizing photons (Lyman continuum, LyC; with energies > 13.6 eV) in star-forming galaxies are still poorly understood, but they are key to identifying and characterizing the sources that reionized the Universe. Here we empirically explore the relation between the hardness of ionizing radiation and the LyC leakage in a large sample of low- z star-forming galaxies from the recent Hubble Space Telescope Low- z Lyman Continuum Survey. Using Sloan Digital Sky Survey stacks and deep X-shooter observations, we investigate the hardness of the ionizing spectra ( Q He + / Q H ) between 54.4 eV (He + ) and 13.6 eV (H) from the optical recombination lines He II 4686 Å and Hβ 4861 Å for galaxies with LyC escape fractions spanning a wide range, f esc (LyC) ≃ 0–90%. We find that the observed intensity of He II/Hβ is primarily driven by variations in the metallicity, but is not correlated with LyC leakage. Both very strong (< f esc (LyC)> ≃ 0.5) and nonleakers (< f esc (LyC)> ≃ 0) present similar observed intensities of He II and Hβ at comparable metallicity, between ≃0.01 and ≃0.02 for 12 + log(O/H)> 8.0 and < 8.0, respectively. Our results demonstrate that Q He + / Q H does not correlate with f esc (LyC), which implies that strong LyC emitters do not show harder ionizing spectra than nonleakers at similar metallicity.

79 ASTRONOMY AND ASTROPHYSICS↗

Ceramic–metal composites for heat exchangers in concentrated solar power plants

The efficiency of generating electricity from heat using concentrated solar power plants (which use mirrors or lenses to concentrate sunlight in order to drive heat engines, usually involving turbines) may be appreciably increased by operating with higher turbine inlet temperatures, but this would require improved heat exchanger materials. By operating turbines with inlet temperatures above 1,023 kelvin using closed-cycle high-pressure supercritical carbon dioxide (sCO 2 ) recompression cycles, instead of using conventional (such as subcritical steam Rankine) cycles with inlet temperatures below 823 kelvin, the relative heat-to-electricity conversion efficiency may be increased by more than 20 per cent. The resulting reduction in the cost of dispatchable electricity from concentrated solar power plants (coupled with thermal energy storage) would be an important step towards direct competition with fossil-fuel-based plants and a large reduction in greenhouse gas emissions. However, the inlet temperatures of closed-cycle high-pressure sCO 2 turbine systems are limited by the thermomechanical performance of the compact, metal-alloy-based, printed-circuit-type heat exchangers used to transfer heat to sCO 2 . We present a robust composite of ceramic (zirconium carbide, ZrC) and the refractory metal tungsten (W) for use in printed-circuit-type heat exchangers at temperatures above 1,023 kelvin. This composite has attractive high-temperature thermal, mechanical and chemical properties and can be processed in a cost-effective manner. We fabricated ZrC/W-based heat exchanger plates with tunable channel patterns by the shape-and-size-preserving chemical conversion of porous tungsten carbide plates. The dense ZrC/W-based composites exhibited failure strengths of over 350 megapascals at 1,073 kelvin, and thermal conductivity values two to three times greater than those of iron- or nickel-based alloys at this temperature. Corrosion resistance to sCO 2 at 1,023 kelvin and 20 megapascals was achieved by bonding a copper layer to the composite surface and adding 50 parts per million carbon monoxide to sCO 2 . Finally, techno-economic analyses indicate that ZrC/W-based heat exchangers can strongly outperform nickel-superalloy-based printed-circuit heat exchangers at lower cost.

36 MATERIALS SCIENCE↗

The Application of Reactor Kinetics to the Analysis of Experiments

The conventional kinetics equations for reactors in which fuel is stationary are derived without approximation from the time-dependent transport equation. The utility of the precise form which results is discussed, first generally, then with reference to the detailed analysis of two typical experiments. In this study, it is shown that the analysis required for a precise interpretation of kinetics experiments can generally be made using the same techniques employed in the computation of criticality for stationary cases.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

THE ROLE OF QUENCHING TIME IN THE EVOLUTION OF THE MASS–SIZE RELATION OF PASSIVE GALAXIES FROM THE WISP SURVEY

We analyze how passive galaxies at z ∼ 1.5 populate the mass–size plane as a function of their stellar age, to understand if the observed size growth with time can be explained with the appearance of larger quenched galaxies at lower redshift. We use a sample of 32 passive galaxies extracted from the Wide Field Camera 3 Infrared Spectroscopic Parallel (WISP) survey with spectroscopic redshift 1.3 ≲ z ≲ 2.05, specific star formation rates lower than 0.01 Gyr{sup −1}, and stellar masses above 4.5 × 10{sup 10} M {sub ⊙}. All galaxies have spectrally determined stellar ages from fitting of their rest-frame optical spectra and photometry with stellar population models. When dividing our sample into young (age ≤2.1 Gyr) and old (age >2.1 Gyr) galaxies we do not find a significant trend in the distributions of the difference between the observed radius and that predicted by the mass–size relation. This result indicates that the relation between the galaxy age and its distance from the mass–size relation, if it exists, is rather shallow, with a slope α ≳ −0.6. At face value, this finding suggests that multiple dry and/or wet minor mergers, rather than the appearance of newly quenched galaxies, are mainly responsible for the observed time evolution of the mass–size relation in passive galaxies.

79 ASTRONOMY AND ASTROPHYSICS↗

Notes on the plasma resonance peak employed to determine doping in SiC

In this study, the doping level of a semiconductor material can be determined using the plasma resonance frequency to obtain the carrier concentration associated with doping. This paper provides an overview of the procedure for the three most common polytypes of SiC. Results for 3C-SiC are presented and discussed. In phosphorus doped samples analysed, it is submitted that the 2nd plasma resonance cannot be detected due to high values of the free carrier damping constant γ.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗

IMPACT OF DIELECTRIC PARAMETERS ON THE REFLECTIVITY OF 3C-SiC WAFERS WITH A ROUGH SURFACE MORPHOLOGY IN THE RESTSTRAHLEN REGION

A layer-on-substrate model is used to obtain the infrared reflectance for 3C-SiC with a rough surface morphology. The effect of varying dielectric parameters of the “damaged layer” on the observed reflectivity of the 3C-SiC in the reststrahlen region is assessed. Different simulated reflectance spectra are obtained to those if the dielectric parameters of the “substrate” were varied. Most notable changes in the shape of the simulated reststrahlen peak are observed for changes in the high frequency dielectric constant, the phonon damping constant, the phonon frequencies and “thickness” of damaged surface layer.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗

Monte Carlo investigation of I-125 interseed attenuation for standard and thinner seeds in prostate brachytherapy with phantom validation using a MOSFET

Purpose: In permanent seed implant prostate brachytherapy the actual dose delivered to the patient may be less than that calculated by TG-43U1 due to interseed attenuation (ISA) and differences between prostate tissue composition and water. In this study the magnitude of the ISA effect is assessed in a phantom and in clinical prostate postimplant cases. Results are compared for seed models 6711 and 9011 with 0.8 and 0.5 mm diameters, respectively. Methods: A polymethyl methacrylate (PMMA) phantom was designed to perform ISA measurements in a simple eight-seed arrangement and at the center of an implant of 36 seeds. Monte Carlo (MC) simulation and experimental measurements using a MOSFET dosimeter were used to measure dose rate and the ISA effect. MC simulations of 15 CT-based postimplant prostate treatment plans were performed to compare the clinical impact of ISA on dose to prostate, urethra, rectum, and the volume enclosed by the 100% isodose, for 6711 and 9011 seed models. Results: In the phantom, ISA reduced the dose rate at the MOSFET position by 8.6%-18.3% (6711) and 7.8%-16.7% (9011) depending on the measurement configuration. MOSFET measured dose rates agreed with MC simulation predictions within the MOSFET measurement uncertainty, which ranged from 5.5% to 7.2% depending on the measurement configuration (k= 1, for the mean of four measurements). For 15 clinical implants, the mean ISA effect for 6711 was to reduce prostate D90 by 4.2 Gy (3%), prostate V100 by 0.5 cc (1.4%), urethra D10 by 11.3 Gy (4.4%), rectal D2cc by 5.5 Gy (4.6%), and the 100% isodose volume by 2.3 cc. For the 9011 seed the mean ISA effect reduced prostate D90 by 2.2 Gy (1.6%), prostate V100 by 0.3 cc (0.7%), urethra D10 by 8.0 Gy (3.2%), rectal D2cc by 3.1 Gy (2.7%), and the 100% isodose volume by 1.2 cc. Differences between the MC simulation and TG-43U1 consensus data for the 6711 seed model had a similar impact, reducing mean prostate D90 by 6 Gy (4.2%) and V100 by 0.6 cc (1.8%). Conclusions: ISA causes the delivered dose in prostate seed implant brachytherapy to be lower than the dose calculated by TG-43U1. MC simulation of phantom seed arrangements show that dose at a point can be reduced by up to 18% and this has been validated using a MOSFET dosimeter. Clinical simulations show that ISA reduces DVH parameter values, but the reduction is less for thinner seeds.

60 APPLIED LIFE SCIENCES↗

Thermal interface conductance in Si/Ge superlattices by equilibrium molecular dynamics

We provide here a derivation allowing the calculation of thermal conductance at interfaces by equilibrium molecular dynamics simulations and illustrate our approach by studying thermal conduction mechanisms in Si/Ge superlattices. Thermal conductance calculations of superlattices with period thicknesses ranging from 0.5 to 60 nm are presented as well as the temperature dependence. Results have been compared to complementary Green-Kubo thermal conductivity calculations demonstrating that thermal conductivity of perfect superlattices can be directly deduced from interfacial conductance in the investigated period range. This confirms the predominant role of interfaces in materials with large phonon mean free paths.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗