Search NASA⌕ Search

SEARCH · Search NASA

Results for “diffusion models”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 1,369 records · Page 76

SPIDER OPTIMIZATION. II. OPTICAL, MAGNETIC, AND FOREGROUND EFFECTS

SPIDER is a balloon-borne instrument designed to map the polarization of the cosmic microwave background (CMB) with degree-scale resolution over a large fraction of the sky. SPIDER's main goal is to measure the amplitude of primordial gravitational waves through their imprint on the polarization of the CMB if the tensor-to-scalar ratio, r, is greater than 0.03. To achieve this goal, instrumental systematic errors must be controlled with unprecedented accuracy. Here, we build on previous work to use simulations of SPIDER observations to examine the impact of several systematic effects that have been characterized through testing and modeling of various instrument components. In particular, we investigate the impact of the non-ideal spectral response of the half-wave plates, coupling between focal-plane components and Earth's magnetic field, and beam mismatches and asymmetries. We also present a model of diffuse polarized foreground emission based on a three-dimensional model of the Galactic magnetic field and dust, and study the interaction of this foreground emission with our observation strategy and instrumental effects. We find that the expected level of foreground and systematic contamination is sufficiently low for SPIDER to achieve its science goals.

79 ASTRONOMY AND ASTROPHYSICS↗

MODIS Aqua Reflective Solar Bands Calibration for NASA's R2018 Ocean Color Products

Remote sensing ocean color products have stringent requirements on radiometric calibration stability. To address a calibration deficiency in MODIS Aqua in recent years, the NASA Ocean Biological Processing Group (OBPG) developed a new calibration for the reflective solar bands. Prior to the reprocessing of NASA's ocean color products for 2018 (R2018), the OBPG MODIS products had been based on calibration provided by the MODIS Calibration Support Team (MCST). Several modifications were made to the MCST calibration approach to improve the calibration accuracy for ocean color products. These include: 1) apply 936nm detector normalization to solar diffuser stability monitor (SDSM) data to reduce coherent noise; 2) modeling solar diffuser (SD) degradation wavelength dependency to determine SD degradation in near-infrared and shortwave infrared wavelength; 3) detector gains are computed using SD screen-closed data to better match ocean radiance levels in all bands; 4) a simple atmospheric correction is performed to reduce bidirectional reflectance distribution function (BRDF) effects in desert trends; 5) modulated relative spectral response (RSR) impact on ocean data is estimated and used to adjust the calibration coefficients. 6) smoothing is used to characterize the temporal change in calibration, and response versus scan angle (RVS) changes are characterized using 2nd order polynomials to improve spatial/temporal calibration stability. Relative to the previous R2014 ocean color products, the R2018 calibration removed the suspect late-mission global trends in blue-band water-leaving reflectance, and some anomalously large short-term variability (spikes) in the temporal trend of chlorophyll concentration. This paper will describe the OBPG calibration with a focus on the differences between the MCST and OBPG approaches.

calibration↗

Contributions of anharmonic phonon interactions to thermal boundary conductance.

Continued reduction of characteristic dimensions in nanosystems has given rise to increasing importance of material interfaces on the overall system performance. With regard to thermal transport, this increases the need for a better fundamental understanding of the processes affecting interfacial thermal transport, as characterized by the thermal boundary conductance. When thermal boundary conductance is driven by phononic scattering events, accurate predictions of interfacial transport must account for anharmonic phononic coupling as this affects the thermal transmission. In this paper, a new model for phononic thermal boundary conductance is developed that takes into account anharonic coupling, or inelastic scattering events, at the interface between two materials. Previous models for thermal boundary conductance are first reviewed, including the Diffuse Mismatch Model, which only consdiers elastic phonon scattering events, and earlier attempts to account for inelastic phonon scattering, namely, the Maximum Transmission Model and the Higher Harmonic Inelastic model. A new model is derived, the Anharmonic Inelastic Model, which provides a more physical consideration of the effects of inelastic scattering on thermal boundary conductance. This is accomplished by considering specific ranges of phonon frequency interactions and phonon number density conservation. Thus, this model considers the contributions of anharmonic, inelastically scattered phonons to thermal boundary conductance. This new Anharmonic Inelastic Model shows excellent agreement between model predictions and experimental data at the Pb/diamond interface due to its ability to account for the temperature dependent changing phonon population in diamond, which can couple anharmonically with multiple phonons in Pb.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Alveolar ventilation to perfusion heterogeneity and diffusion impairment in a mathematical model of gas exchange

This study describes a two-compartment model of pulmonary gas exchange in which alveolar ventilation to perfusion (VA/Q) heterogeneity and impairment of pulmonary diffusing capacity (D) are simultaneously taken into account. The mathematical model uses as input data measurements usually obtained in the lung function laboratory. It consists of two compartments and an anatomical shunt. Each compartment receives fractions of alveolar ventilation and blood flow. Mass balance equations and integration of Fick's law of diffusion are used to compute alveolar and blood O2 and CO2 values compatible with input O2 uptake and CO2 elimination. Two applications are presented. The first is a method to partition O2 and CO2 alveolar-arterial gradients into VA/Q and D components. The technique is evaluated in data of patients with chronic obstructive pulmonary disease (COPD). The second is a theoretical analysis of the effects of blood flow variation in alveolar and blood O2 partial pressures. The results show the importance of simultaneous consideration of D to estimate VA/Q heterogeneity in patients with diffusion impairment. This factor plays an increasing role in gas alveolar-arterial gradients as severity of COPD increases. Association of VA/Q heterogeneity and D may produce an increase of O2 arterial pressure with decreasing QT which would not be observed if only D were considered. We conclude that the presented computer model is a useful tool for description and interpretation of data from COPD patients and for performing theoretical analysis of variables involved in the gas exchange process.

NASA Discipline Environmental Health↗

Diffusion processes in Al2O3 scales - Void growth, grain growth, and scale growth

The internal microstructure and growth kinetics of Al2O3 scales on Ni-15Cr-13Al (wt percent) are investigated by TEM and analyzed in relation to models of diffusivity. Polished arc-melted specimens were oxidized in 1-atm air at 1100 C for 0.1, 1.0, and 20 hours and ion-thinned for TEM at 100 kV. The frequency distribution of void size and grain size is determined for different oxidation times and scale depths. The kinetics of microvoid growth and of grain and scale growth are plotted and related via simplified models to lattice and grain-boundary oxygen diffusivity, respectively. Good agreement is found between model predictions and data obtained by Oishi and Kingery (1960) on oxygen diffusion in bulk Al2O3. The further implications and limitations of these findings are discssed.

Smialek, J. L.↗

CFD Modeling of Boron Removal from Liquid Silicon with Cold Gases and Plasma

The present study focuses on a specific step of the metallurgical path of purification to provide solar-grade silicon: the removal of boron through the injection of H{sub 2}O(g)-H{sub 2}(g)-Ar(g) (cold gas process) or of Ar-H{sub 2}-O{sub 2} plasma (plasma process) on stirred liquid silicon. We propose a way to predict silicon and boron flows from the liquid silicon surface by using a CFD model (©Ansys Fluent) combined with some results on one-dimensional diffusive-reactive models to consider the formation of silica aerosols in a layer above the liquid silicon. The comparison of the model with experimental results on cold gas processes provided satisfying results for cases with low and high concentrations of oxidants. This confirms that the choices of thermodynamic data of HBO(g) and the activity coefficient of boron in liquid silicon are suitable and that the hypotheses regarding similar diffusion mechanisms at the surface for HBO(g) and SiO(g) are appropriate. The reasons for similar diffusion mechanisms need further enquiry. We also studied the effect of pressure and geometric variations in the cold gas process. For some cases with high injection flows, the model slightly overestimates the boron extraction rate, and the overestimation increases with increasing injection flow. A single plasma experiment from SIMaP (France) was modeled, and the model results fit the experimental data on purification if we suppose that aerosols form, but it is not enough to draw conclusions about the formation of aerosols for plasma experiments.

36 MATERIALS SCIENCE↗

ANALYSIS OF MAGNETOROTATIONAL INSTABILITY WITH THE EFFECT OF COSMIC-RAY DIFFUSION

We present the results obtained from the linear stability analysis and 2.5 dimensional magnetohydrodynamic (MHD) simulations of magnetorotational instability (MRI), including the effects of cosmic rays (CRs). We took into account the CR diffusion along the magnetic field but neglected the cross-field-line diffusion. Two models are considered in this paper: the shearing box model and differentially rotating cylinder model. We studied how MRI is affected by the initial CR pressure (i.e., energy) distribution. In the shearing box model, the initial state is uniform distribution. Linear analysis shows that the growth rate of MRI does not depend on the value of the CR diffusion coefficient. In the differentially rotating cylinder model, the initial state is a constant angular momentum polytropic disk threaded by a weak uniform vertical magnetic field. Linear analysis shows that the growth rate of MRI becomes larger if the CR diffusion coefficient is larger. Both results are confirmed by MHD simulations. The MHD simulation results show that the outward movement of matter by the growth of MRI is not impeded by the CR pressure gradient, and the centrifugal force that acts on the concentrated matter becomes larger. Consequently, the growth rate of MRI is increased. On the other hand, if the initial CR pressure is uniform, then the growth rate of the MRI barely depends on the value of the CR diffusion coefficient.

79 ASTRONOMY AND ASTROPHYSICS↗

Atoms in carbon cages as a source of interstellar diffuse lines

A model to describe the resonance absorption lines of various atoms trapped in closed carbon cages is presented. These systems may be responsible for some of the as yet unexplained diffuse interstellar bands. Model potentials for possible atom-C60 systems are obtained and used to calculate the resonance lines. The trapped atoms considered are O, N, Si, Mg, Al, Na, and S, and in all cases the resonance lines are shifted toward the red as compared to the isolated atoms. The calculated wavelengths are compared to the range of wavelengths observed for the diffuse interstellar bands, and good agreement is found for Mg and Si resonance lines. Other lines may be caused by other than resonance transitions or by trapped molecules. The oscillator strengths and the abundances are evaluated and compared with observation. Mechanisms to explain the observed band width of the lines and the existence of certain correlated pairs of lines are discussed.

Ballester, J. L.↗

3D simulations of spinlike flames in Co/Al multilayers with enhanced conduction losses

Reactive Co/Al multilayers are uniformly structured materials that may be ignited to produce rapid and localized heating. Prior studies varying the bilayer thickness (i.e., sum of two individual layers of Co and Al) have revealed different types of flame morphologies, including: (a) steady/planar, (b) wavy/periodic, and (c) transverse bands, originating in the flame front. These instabilities resemble the “spin waves” first observed in the early studies of solid combustion (i.e., Ti cylinder in a N 2 atmosphere), and are likewise thought to be due to the balance of heat released by reaction and heat conduction forward into the unreacted multilayer. However, the multilayer geometry and three-dimensional (3D) edge effects are relatively unexplored. In this work, a new diffusion-limited reaction model for Co/Al multilayers was implemented in large, novel 3D finite element analysis (FEA) simulations, in order to study the origins of these spinlike flames. This reaction model builds upon previous work by introducing three new phase-dependent property models for: (1) the diffusion coefficient, (2) anisotropic thermal conductivity tensor, and (3) bulk heat capacity, as well as one additional model for the bilayer-dependent heat of reaction. These novel 3D simulations are the first to predict both steady and unsteady flames in Co/Al multilayers. Moreover, two unsteady modes of flame propagation are identified, which depend on the enhanced conduction losses with slower flames, as well as flame propagation around notched edges. Finally, future work will consider the generality of the current modeling approach and also seek to define a more generalized set of stability criteria for additional multilayer systems.

42 ENGINEERING↗

PREPARATION AND TESTING OF CORROSIONAND SPALLATION-RESISTANT COATINGS

This Energy & Environmental Research Center (EERC) project is designed to determine if plating APMT®, a specific highly oxidation-resistant oxide dispersion-strengthened FeCrAl alloy made by Kanthal, onto nickel-based superalloy turbine parts is a viable method for substantially improving the lifetimes and maximum use temperatures of the parts. The method for joining the APMT plate to the superalloys is called evaporative metal bonding and involves placing a thin foil of zinc (Zn) between the plate and the superalloy, clamping them together, and heating in an atmosphere-controlled furnace. Upon heating, the Zn melts and dissolves the oxide skins of the alloys at the bond line, allowing the two alloys to diffuse into each other. The Zn then diffuses through the alloys and evaporates from their surfaces. Laboratory testing to determine the diffusion rate of Zn through the alloys has been completed. However, an analytical solution does not exist to model the diffusion of zinc through the alloys. For this reason, a finite difference algorithm using MATLAB was developed. It makes use of the hopscotch algorithm. The model allows the user to specify the dimensions of the metal parts, the Zn concentration at the bondline, the mesh size, time step, and Zn diffusivity. The experimentally measured values of diffusivity for Zn in APMT and Rene 80/CM 247LC are approximately 2.7 × 10-12 and 4 × 10-14 m2/s, respectively. While the qualitative behavior of the model appears correct, a comparison of the diffusion predictions with the experimental results from earlier in the project indicates that the expected Zn concentration is significantly higher than that measured experimentally. The difference depends on the assumed initial concentration, which is difficult to quantify exactly under experimental conditions for t = 0. In addition to the diffusion work, the coefficients of thermal expansions were determined for each of the alloys as a function of temperature. This information has been entered into a finite element model using ANSYS so that appropriate force-applying structures can be designed for use in joining structures composed of APMT and the nickel alloys. Finite element modeling has been performed to finalize the fabrication geometry for the corrosion-testing phase. The addition of another bolt increases stress uniformity away from the region where the clamping is applied. It appears that a bolt spacing of approximately 25 mm in each jig is appropriate. This will allow the fabrication of 50-mm-wide sections of joints for the corrosion-testing task. Gasifier sampling activities continue to determine what types of trace contaminants may occur in cleaned syngas that could lead to corrosion or deposition in turbines firing coal syngas. The EERC has several pilot-scale gasifiers that are continually used in a variety of test configurations as determined by the needs of the projects that are funding the tests. We are sampling both noncombusted and combusted syngas produced during some of the pilot-scale gasifier tests. This year sampling was performed of both syngas and combusted syngas while the entrained-flow gasifier (EFG) was firing subbituminous coal from the Antelope Mine in Wyoming. Results of scanning electron microscope analyses of the syngas before combustion showed no submicron particles, only flakes of iron oxide that had likely formed on steel surfaces inside the combustor. As shown in the 2013 annual report, soot was also collected from the syngas when the much-lower-temperature fluid-bed gasifier (FBG) was fired, indicating that the much higher temperature of the EFG prevented soot formation. However, particles collected from the combusted syngas consist almost entirely of submicron soot, and little to no vaporized metals made it past the warm-gas filters and scrubbers in the high-temperature EFG system which could then deposit in a turbine system burning a higher hydrogen syngas. These results are consistent with the analyses of the particulates collected from combusted syngas when the lowertemperature FBG system is used.

Electric↗

Prediction of Cooling Curves for Controlled Unidirectional Solidification Under the Influence of Shrinkage: A Semi-analytical Approach

Quality of casting products can be significantly improved by controlling the crystal growth rate. Both dendritic and equiaxed grain structures depend strongly on crystal growth rate along the crystallographic direction. Uncontrolled solidification processes may lead to the formation of porous and columnar material with an extremely non-homogeneous composition distribution. These undesirable micro-structures can be avoided by adopting an optimal cooling rate (K/s) during the crystal growth. The present work focuses on determining suitable cooling curves to obtain desired unidirectional crystal growth rates, by implementing a semi-analytical model. Here, we defined cooling curve as the time history of temperature evolution at the cold boundary. The proposed semi-analytical heat transfer model is diffusion driven and accounts for shrinkage effect during solidification. The proposed model is found capable of predicting cooling curves for prescribed interface velocities. The model is self-sufficient in validating itself by means of solving inverse and forward problems. The cooling curve prediction by the model is validated by using it as the transient boundary condition for an existing enthalpy updating scheme-based numerical model. The proposed model is also applied to empirical data reported in literature, to validate the model against experimental results. All the validations showcased excellent fit. The proposed model will facilitate controlled unidirectional crystal growth rate by providing cooling rate (K/s) as an input for the experiments involving directional solidification.

36 MATERIALS SCIENCE↗

The exit-time problem for a Markov jump process

The purpose of this paper is to consider the exit-time problem for a finite-range Markov jump process, i.e, the distance the particle can jump is bounded independent of its location. Additionally, such jump diffusions are expedient models for anomalous transport exhibiting super-diffusion or nonstandard normal diffusion. We refer to the associated deterministic equation as a volume-constrained nonlocal diffusion equation. The volume constraint is the nonlocal analogue of a boundary condition necessary to demonstrate that the nonlocal diffusion equation is well-posed and is consistent with the jump process. A critical aspect of the analysis is a variational formulation and a recently developed nonlocal vector calculus. Finally, this calculus allows us to pose nonlocal backward and forward Kolmogorov equations, the former equation granting the various moments of the exit-time distribution.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Reactive transport modelling of cement paste leaching in brines

We propose and analyse a unified reactive transport model for leaching of cement paste in brines, with and without CO{sub 2}. We investigate the impact of the boundary conditions on the rate of propagation of the reaction. In agreement with experiments, we observe distinct mechanisms as a function of the pH and the CO{sub 2} concentration of the leaching brine. In particular, clogging is observed at intermediate concentration of CO{sub 2} while higher concentration accelerates the reaction. In addition, we leverage the flexibility of the simulator to explore the impact of different models on the simulations. In particular, aqueous species, phase assemblage and diffusion coefficient model are varied. We demonstrate that in order to be able to obtain quantitative predictions, it is important to accurately capture the feedback between chemical reactions, microstructure and macroscopic transport parameters.

36 MATERIALS SCIENCE↗

FERMI LARGE AREA TELESCOPE FIRST SOURCE CATALOG

We present a catalog of high-energy gamma-ray sources detected by the Large Area Telescope (LAT), the primary science instrument on the Fermi Gamma-ray Space Telescope (Fermi), during the first 11 months of the science phase of the mission, which began on 2008 August 4. The First Fermi-LAT catalog (1FGL) contains 1451 sources detected and characterized in the 100 MeV to 100 GeV range. Source detection was based on the average flux over the 11 month period, and the threshold likelihood Test Statistic is 25, corresponding to a significance of just over 4{sigma}. The 1FGL catalog includes source location regions, defined in terms of elliptical fits to the 95% confidence regions and power-law spectral fits as well as flux measurements in five energy bands for each source. In addition, monthly light curves are provided. Using a protocol defined before launch we have tested for several populations of gamma-ray sources among the sources in the catalog. For individual LAT-detected sources we provide firm identifications or plausible associations with sources in other astronomical catalogs. Identifications are based on correlated variability with counterparts at other wavelengths, or on spin or orbital periodicity. For the catalogs and association criteria that we have selected, 630 of the sources are unassociated. Care was taken to characterize the sensitivity of the results to the model of interstellar diffuse gamma-ray emission used to model the bright foreground, with the result that 161 sources at low Galactic latitudes and toward bright local interstellar clouds are flagged as having properties that are strongly dependent on the model or as potentially being due to incorrectly modeled structure in the Galactic diffuse emission.

79 ASTRONOMY AND ASTROPHYSICS↗

Ocean Turbulence: One-Point Closure Model Momentum, Heat and Salt Vertical Diffusivities in the Presence of Shear - Paper 2

We develop and test a 1-point closure turbulence model with the following features: 1) we include the salinity field and derive the expression for the vertical turbulent diffusivities of momentum K(sub m) , heat K(sub h) and salt K(sub s) as a function of two stability parameters: the Richardson number R(sub i) (stratification vs. shear) and the Turner number R(sub rho) (salinity gradient vs. temperature gradient). 2) to describe turbulent mixing below the mixed layer (ML), all previous models have adopted three adjustable "background diffusivities" for momentum, heat and salt. We propose a model that avoids such adjustable diffusivities. We assume that below the ML, the three diffusivities have the same functional dependence on R( sub i) and R(sub rho) as derived from the turbulence model. However, in order to compute R(sub i) below the ML, we use data of vertical shear due to wave-breaking.measured by Gargett et al. The procedure frees the model from adjustable background diffusivities and indeed we employ the same model throughout the entire vertical extent of the ocean. 3) in the local model, the turbulent diffusivities K(sub m,h,s) are given as analytical functions of R(sub i) and R(sub rho). 5) the model is used in an O-GCM and several results are presented to exhibit the effect of double diffusion processes. 6) the code is available upon request.

Canuto, V. M.↗

Climate stability for a Sellers-type model

We study a diffusive energy-balance climate model governed by a nonlinear parabolic partial differential equation. Three positive steady-state solutions of this equation are found; they correspond to three possible climates of our planet: an interglacial (nearly identical to the present climate), a glacial, and a completely ice-covered earth. We consider also models similar to the main one studied, and determine the number of their steady states. All the models have albedo continuously varying with latitude and temperature, and entirely diffusive horizontal heat transfer. The diffusion is taken to be nonlinear as well as linear. We investigate the stability under small perturbations of the main model's climates. A stability criterion is derived, and its application shows that the 'present climate' and the 'deep freeze' are stable, whereas the model's glacial is unstable. A variational principle is introduced to confirm the results of this stability analysis. For a sufficient decrease in solar radiation (about 2%) the glacial and interglacial solutions disappear, leaving the ice-covered earth as the only possible climate.

Ghil, M.↗

SU-F-T-118: Characterization of Change in Fractional Anisotropy After Radiation Therapy: Does Nearby Disruption Predict for White Matter Damage?

Purpose: We investigated predictors of fractional anisotropy (FA) change in the corticospinal white matter tract (CST) following radiation therapy (RT). Methods: Diffusion tensor imaging (DTI) is a non-invasive modality which models water diffusion properties. FA quantifies the extent of directional bias—a decrease indicates disrupted white matter integrity. Fifteen patients with high-grade glioma underwent DTI scans before, and ten months after RT to 59.4–60 Gy. The CST was segmented using an automated atlas-based algorithm on all DTI images. Treatment planning CT and DTI images were aligned using non-linear registration allowing for baseline FA, follow-up FA, and absorbed dose to be determined in each voxel. Relative FA change was dichotomized into a binary outcome using 25% decrease as cutoff. Three metrics were assessed as predictors: voxel dose, distance from the voxel to the center of the CST (Rc), and the number of neighboring voxels (Nadj from 0 to 26) with ≥25% decrease in FA. Logistic regression and the area under the receiver-operating characteristics curve (AUC) analysis were performed for each patient. Results: Median age of the cohort was 59 years (range: 40–85). The average number of voxels in the CST amongst all patients was 1181 (±172, SD). In logistic regression, the probability of FA change was highly associated with Nadj in all 15 patients with corresponding AUCs between 0.81 and 0.97. With all three metrics included in the logistic regression models, Nadj was highly significant (p<0.001) in all patients, voxel dose significant (p<0.05) in 3/15 patients, and Rc significant in 12/15 patients (p<0.05). Conclusion: The number of neighboring voxels with change in FA was the dominant predictor of FA change at any given voxel. This suggests that the microenvironment of surrounding white matter disruption after radiation therapy may drive local effects along a white matter tract. Pettersson and Cervino are funded by a Varian Medical Systems grant.

60 APPLIED LIFE SCIENCES↗

Synergistic Effect of Nitrogen and Refractory Material on TiN Formation and Equiaxed Grain Structure of Ferritic Stainless Steel

The effect of nitrogen content on the formation of an equiaxed solidification structure of Fe-16Cr steel was investigated. Moreover, two different kinds of refractory materials, i.e., alumina and magnesia, were employed to control the type of oxide inclusion. The characteristics of TiN(-oxide) inclusions were quantitatively analyzed in both molten steel and solidified samples. When the melting was carried out in the alumina refractory, the grain size continuously decreased with increasing nitrogen content. However, a minimum grain size was observed at a specific nitrogen content (approx. 150 ppm) when the steel was melted in the magnesia refractory. Most of the single TiN particles had a cuboidal shape and fine irregularly shaped particles were located along the grain boundary due to the microsegregation of Ti at the grain boundary during solidification. The type of TiN-oxide hybrid inclusion was strongly affected by the refractory material where Al{sub 2}O{sub 3}-TiN and MgAl{sub 2}O{sub 4}-TiN hybrid-type inclusions were obtained in the alumina and magnesia refractory experiments, respectively. The formation of oxide inclusions was well predicted by thermochemical computations and it was commonly found that oxide particles were initially formed, followed by the nucleation and growth of TiN. When the nitrogen content increased, the number density of TiN linearly increased in the alumina refractory experiments. However, the number of TiN exhibits a maximum at about [N] = 150 ppm, at which a minimum grain size was obtained in the magnesia refractory experiments. Therefore, the larger the number density of TiN, the smaller the primary grain size after solidification. The number density of TiN in the steel melted in the magnesia refractory was greater than that in the steel melted in the alumina refractory at given Ti and N contents, which was due to the lower planar lattice disregistry of MgAl{sub 2}O{sub 4}-TiN interface rather than that of Al{sub 2}O{sub 3}-TiN interface. When ∆T{sub TiN} (= difference between the TiN precipitation temperature and the liquidus of the steel) was 20 K to 40 K, the number density of effective TiN was maximized and thus, the grain size was minimized after solidification. Finally, although most of the TiN particles were smaller than 1 μm in the molten steel samples irrespective of the nitrogen content, TiN particles larger than 10 μm were observed in the solidified samples when the nitrogen content was greater than 150 ppm. The growth of TiN particles during melting and solidification was well predicted by the combinatorial simulation of the ‘Ostwald ripening model’ based on the Lifshitz–Slyozov–Wagner theory in conjunction with the ‘Diffusion controlled model’ using Ohnaka’s microsegregation equation.

36 MATERIALS SCIENCE↗