Search NASA⌕ Search

SEARCH · Search NASA

Results for “LU”

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 127 records · Page 7

GSoFa: Scalable Sparse Symbolic LU Factorization on GPUs

Decomposing a matrix $\mathbf {A}$ into a lower matrix $\mathbf {L}$ and an upper matrix $\mathbf {U}$, which is also known as LU decomposition, is an essential operation in numerical linear algebra. For a sparse matrix, LU decomposition often introduces more nonzero entries in the $\mathbf {L}$ and $\mathbf {U}$ factors than in the original matrix. A symbolic factorization step is needed to identify the nonzero structures of $\mathbf {L}$ and $\mathbf {U}$ matrices. Attracted by the enormous potentials of the Graphics Processing Units (GPUs), an array of efforts have surged to deploy various LU factorization steps except for the symbolic factorization, to the best of our knowledge, on GPUs. This article introduces gSoFa, the first GPU-based symbolic factorization design with the following three optimizations to enable scalable LU symbolic factorization for nonsymmetric pattern sparse matrices on GPUs. First, here we introduce a novel fine-grained parallel symbolic factorization algorithm that is well suited for the Single Instruction Multiple Thread (SIMT) architecture of GPUs. Second, we tailor supernode detection into a SIMT friendly process and strive to balance the workload, minimize the communication and saturate the GPU computing resources during supernode detection. Third, we introduce a three-pronged optimization to reduce the excessive space consumption problem faced by multi-source concurrent symbolic factorization. Taken together, gSoFa achieves up to 31× speedup from 1 to 44 Summit nodes (6 to 264 GPUs) and outperforms the state-of-the-art CPU project, on average, by 5×. Notably, gSoFa also achieves up to 47 percent of the peak memory throughput of a V100 GPU in the Summit Supercomputer.

97 MATHEMATICS AND COMPUTING↗

Simulation studies of a full-ring, CZT SPECT system for whole-body imaging of 99m Tc and 177 Lu

Single photon emission computed tomography (SPECT) is an imaging modality that has demonstrated its utility in a number of clinical indications. Despite this progress, a high sensitivity, high spatial resolution, multi-tracer SPECT with a large field of view suitable for whole-body imaging of a broad range of radiotracers for theranostics is not available. With the goal of filling this technological gap, we have designed a cadmium zinc telluride (CZT) full-ring SPECT scanner instrumented with a broad-energy tungsten collimator. The final purpose is to provide a multi-tracer solution for brain and whole-body imaging. Our static SPECT does not rely on the dual- and the triple-head rotational SPECT standard paradigm, enabling a larger effective area in each scan to increase the sensitivity. We provide a demonstration of the performance of our design using a realistic model of our detector with simulated body-sized phantoms filled with 99m Tc and 177 Lu. Our SPECT design can resolve 7.9 mm rods for 99m Tc (140 keV) and 9.5 mm for 177 Lu (208 keV) in a hot-rod Derenzo phantom with a 3-min exposure and reach an image contrast of 78% for 99m Tc and 57% for 177 Lu using the NEMA IQ phantom with a 6-min exposure. Our modified scatter correction shows an improved contrast-recovery ratio compared to a standard correction. In this paper, we demonstrate the good performance of our design for whole-body imaging purposes. This adds to our previous demonstration of improved qualitative and quantitative 99m Tc imaging of brain perfusion and 123 I imaging of dopamine transport with respect to state-of-the-art NaI dual-head cameras. We show that our design provides similar IQ and contrast to the commercial full-ring SPECT VERITON for 99m Tc. Regarding 177 Lu imaging of the 208 keV emissions, our design provides similar contrast to that of other state-of-the-art SPECTs with a significant reduction in exposure. In conclusion, the high sensitivity and extended energy range up to 250 keV makes our SPECT design a promising alternative for clinical imaging and theranostics of emerging radionuclides.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

Temperature dependence of the radiation tolerance of nanocrystalline pyrochlores A 2 Ti 2 O 7 (A = Gd, Ho and Lu)

A potentially enhanced radiation resistance of nanocrystalline materials, as a consequence of the high density of interfaces and surfaces, has attracted much attention both to understand the fundamental role of these defect sinks and to develop them for high-radiation environments. Here, irradiation response of nanocrystalline A 2 Ti 2 O 7 (A = Gd, Ho and Lu) pyrochlore powders with grain sizes of 20–30 nm was investigated by 1-MeV Kr 2+ ion bombardment. In situ transmission electron microscopy (TEM) revealed that the critical amorphization fluence for each nanocrystalline compound at room temperature was greater than that for their coarse-grained counterparts, indicating an enhanced amorphization resistance. The effect of temperature on the irradiation response of one of these compounds, nanocrystalline Lu 2 Ti 2 O 7 , was further examined by performing ion irradiation at an elevated temperature range of 480–600 K. The critical amorphization temperature (T c ) was found to be noticeably higher in nanocrystalline Lu 2 Ti 2 O 7 (610 K) than its coarse-grained counterpart (480 K), revealing that nanocrystalline Lu 2 Ti 2 O 7 is less resistant to amorphization compared to its coarse-grained phase under high temperatures. We interpret these results with the aid of atomistic simulations. Molecular statics calculations find that cation antisite defects are less energetically costly to form near surfaces than in the bulk, suggesting that the nanocrystalline form of these materials is generally less susceptible to amorphization than coarse-grained counterparts at low temperatures where defect kinetics are negligible. In contrast, at high temperatures, the annealing efficiency of antisite defects by cation interstitials is significantly reduced due to the sink properties of the surfaces in the nanocrystalline pyrochlore, which contributes to the observed higher amorphization temperature in the nano-grained phase than in coarse-grained counterpart. Altogether, these results provide new insight into the behavior of nanocrystalline materials under irradiation.

36 MATERIALS SCIENCE↗

Isotope harvesting from a NSCL beam blocker: Isolation of 173,174 lu

A decommissioned beam blocker from the National Superconducting Cyclotron Laboratory was processed for recovery of long-lived Lu radioisotopes at Lawrence Livermore National Laboratory. The beam blocker was a W alloy that had built up considerable levels of activation from high energy particle irradiation over the course of 20 years. The isolation of 173,174 Lu from the bulk material required a complex chemical separation scheme due to the large mass and the wide variety of elements present. The final yield of 173,174 Lu was 76.7 ± 4.8% with an extremely high radiopurity and chemical purity. The reduction in W mass from the initial material to the final sample was 8 × 10 10 . In conclusion, the results show great promise for extracting large quantities of Lu radioisotopes from these types of W beam blockers at the Facility for Rare Isotope Beams, which will be irradiated at levels significantly higher than the beam blocker investigated here.

and nuclear chemistry↗

Ternary aluminides R{sub 0.67}Ni{sub 2}Al{sub 6} (R=Sc, Y, Gd-Lu) with partly disordered structures

Ten rare-earth nickel aluminides R{sub 0.67}Ni{sub 2}Al{sub 6} (R=Sc, Y, Gd, Tb, Dy, Ho, Er, Tm, Yb, Lu) were synthesized by arc melting and their crystal structures studied by X-ray powder diffraction. They crystallize in the hexagonal structure type Yb{sub 0.67}Ni{sub 2}Al{sub 6}: Pearson symbol hP11-2.33, space group P-6m2, a=4.18904(2), c=9.09816(6) A, which can be decomposed into three kinds of layer. The sequence of layers along [001] is Yb{sub 0.67}Al-NiAl{sub 2}-Al-NiAl{sub 2}, but stacking faults appear due to relative shifts of consecutive Yb{sub 0.67}Al layers containing two rare-earth metal atoms for one Al-atom triangle. An ordered superstructure is the type ErNi{sub 3}Al{sub 9} (Pearson symbol hR78, space group R32, a={radical}(3)a{sub 0}, c=3c{sub 0}), refined on single crystals. The physical properties of the R{sub 0.67}Ni{sub 2}Al{sub 6} (R=Tb, Er, Tm, Yb, and Lu) compounds were determined by means of magnetic and electrical transport measurements. All these materials but Lu{sub 0.67}Ni{sub 2}Al{sub 6} exhibit Curie-Weiss paramagnetic behavior due to the trivalent R ions, and order magnetically at low temperatures. They show good metallic conductivity. The Yb-containing compound is a dense Kondo system with distinct features in the paramagnetic state of the interplay between Kondo and crystalline electric field interactions. - Graphical abstract: Stacking of layers along [001] in the Yb{sub 0.67}Ni{sub 2}Al{sub 6}-type structure. Highlights: Black-Right-Pointing-Pointer The aluminides R{sub 0.67}Ni{sub 2}Al{sub 6} (R=Sc, Y, Gd-Lu) adopt a partly disordered structure. Black-Right-Pointing-Pointer The structure is characterized by R{sub 2}Al{sub 3} layers with two R atoms for one Al{sub 3} triangle. Black-Right-Pointing-Pointer The compounds with R=Tb, Er, Tm are paramagnets at rt and order magnetically at lt. Black-Right-Pointing-Pointer The Yb-compound exhibits a magnetically ordered Kondo lattice with strong CEF effect.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

An orange emitting phosphor Lu{sub 2−x}CaMg{sub 2}Si{sub 2.9}Ti{sub 0.1}O{sub 12}:xCe with pure garnet phase for warm white LEDs

A new silicate garnet phosphor, Lu{sub 2−x}CaMg{sub 2}Si{sub 2.9}Ti{sub 0.1}O{sub 12}:xCe was synthesized by a high temperature solid-state reaction under reductive atmosphere. X-ray diffraction (XRD) showed that the powder was pure garnet phase. The emission and excitation spectrum indicated that the Lu{sub 2−x}CaMg{sub 2}Si{sub 2.9}Ti{sub 0.1}O{sub 12}:xCe phosphors could absorb blue light in the spectral range of 400–550 nm efficiently and exhibit bright yellow–orange emission in the range of 520–750 nm. With the increase of Ce{sup 3+} concentration, the emission band of Ce{sup 3+} showed a red shift. Interestingly, the concentration quenching occurred when the Ce{sup 3+} concentration exceeded 4 mol%. The temperature-dependent luminescent properties of the phosphors were discussed and the Lu{sub 1.96}CaMg{sub 2}Si{sub 2.9}Ti{sub 0.1}O{sub 12}:0.04Ce phosphors showed good performances in color temperature (2430 K) and potential applications for warm white LEDs. - Graphical Abstract: This image shows that the phosphor of Lu{sub 1.96}CaMg{sub 2}Si{sub 2.9}Ti{sub 0.1}O{sub 12}:0.04Ce can generate a uniform yellow tint under natural light illumination and emit orange–red light when excited by blue light. With a fixed 467 nm emission light, warm white light can be produced by this phosphor, which indicates that the phosphor is potentially applicable in warm white light emitting diodes based on GaN chips. - Highlights: • A new silicate garnet phosphor was synthesized by solid-state method. • Secondary phases can be avoided when a small amount of Si{sup 4+} were replaced by Ti{sup 4+}. • A broad emission band of Ce{sup 3+} in the phosphors was described. • The phosphors are potentially applicable in warm white light emitting diodes.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Sol-gel synthesis, structure and photoluminescence properties of nanocrystalline Lu{sub 2}MoO{sub 6}:Eu

Highlights: {yields} Monoclinic-structural Lu{sub 2}MoO{sub 6} has strong absorption of near-UV light. {yields} The absorbed energy can be transferred from MoO{sub 5} groups to doped Eu{sup 3+} and luminescence. {yields} The absorption edge and the peak of the MoO{sub 5} excitation band shift in reverse ways. {yields} The decline of the crystallinity and the introduced defect respond for the variation. -- Abstract: In this paper, we reported the obtention of Eu{sup 3+} ion doped Lu{sub 2}MoO{sub 6} powders synthesized by a sol-gel method, and followed by annealing at different temperatures. The structure and photoluminescence properties of these powders were investigated. The X-ray diffraction pattern suggests that Lu{sub 2}MoO{sub 6} powder has a monoclinic structure. It was observed that the UV-visible and photoluminescence spectra of Lu{sub 2}MoO{sub 6}:Eu nanocrystallines varied systematically with the calcination temperature. The near-UV absorption edge shifts to long wavelength direction with the decreasing of the calcination temperature, while the peak of MoO{sub 5} excitation band shifts in an opposite way. The decline of the crystallinity and the introduced lattice defect were considered to respond for these variations. Additionally, due to the efficient red light emission under near-UV light excitation, the powder can be a candidate as red phosphor for white-light-emitting diodes.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Tiny Sc Allows the Chains to Rattle: Impact of Lu and Y Doping on the Charge-Density Wave in ScV 6 Sn 6

The kagome metals display an intriguing variety of electronic and magnetic phases arising from the connectivity of atoms on a kagome lattice. A growing number of these materials with vanadium–kagome nets host charge–density waves (CDWs) at low temperatures, including ScV 6 Sn 6 , CsV 3 Sb 5 , and V 3 Sb 2 . Curiously, only the Sc version of the RV 6 Sn 6 materials with a HfFe 6 Ge 6 -type structure hosts a CDW (R = Gd–Lu, Y, Sc). In this study, we investigate the role of rare earth size in CDW formation in the RV 6 Sn 6 compounds. Magnetization measurements on our single crystals of (Sc,Lu)V 6 Sn 6 and (Sc,Y)V 6 Sn 6 establish that the CDW is suppressed by substituting Sc by larger Lu or Y. Single-crystal X-ray diffraction reveals that compressible Sn–Sn bonds accommodate the larger rare earth atoms within loosely packed R–Sn–Sn chains without significantly expanding the lattice. We propose that Sc provides extra room in these chains crucial to CDW formation in ScV 6 Sn 6 . Our rattling chain model explains why both physical pressure and substitution by larger rare earth atoms hinder CDW formation despite opposite impacts on lattice size. Here, we emphasize the cooperative effect of pressure and rare earth size by demonstrating that pressure further suppresses the CDW in a Lu-doped ScV 6 Sn 6 crystal. Our model not only addresses why a CDW only forms in the RV 6 Sn 6 materials with tiny Sc but also advances our understanding of why unusual CDWs form in the kagome metals.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Structure and magnetic properties of Y{sub 1-x}Lu{sub x}FeO{sub 3} (0 {<=} x {<=} 1) ceramics

Polycrystalline Lu-doped YFeO{sub 3} samples with perovskite structure were synthesized by solid-state reaction. Powder X-ray diffraction, scanning electron microscopy, and energy dispersive X-ray analysis manifest the high quality of the samples. Room temperature {sup 57}Fe Moessbauer spectra indicate that only the Fe{sup 3+} exists in the samples, which excludes the formation of multiple valence of Fe. The large effective magnetic moments of Fe{sup 3+} obtained from the temperature dependence of magnetization data can be explained in terms of the formation of ferromagnetic clusters. Field-dependent magnetization reveals that all the samples show weak ferromagnetic property due to the small canting of the Fe{sup 3+} moments. The field-induced spin-reorientation was detected on YFeO{sub 3} and was gradually suppressed by Lu{sup 3+} doping. Polarization hysteresis loops of Y{sub 1-x}Lu{sub x}FeO{sub 3} (0 {<=} x {<=} 1) were observed at room temperature. Our results suggest that the multiferroic properties may exist in the Y{sub 1-x}Lu{sub x}FeO{sub 3} (0 {<=} x {<=} 1) ceramics.

36 MATERIALS SCIENCE↗

Three-dimensional lattice modulations in the charge density wave system Lu 2 Ir 3 Si 5

Using total and resonant x-ray scattering coupled to large-scale computer modeling, we study the lattice modulations in the complex charge density wave (CDW) material Lu 2 ⁢Ir 3 ⁢Si 5 . Here, we find that it is a unique quantum system where periodic lattice modulations related to emergent CDW order occur in three orthogonal atomic planes of the crystal lattice, leading to the emergence of an unusual three-dimensional (3D) pattern of short and long Ir-Ir and Lu-Lu bonds. The 3D character of observed lattice modulations explains the largely isotropic character of the changes in the electronic properties occurring when the CDW order sets in, demonstrating the strong electron-lattice coupling in Lu 2 ⁢Ir 3 ⁢Si 5 . The result is supported by DFT calculations based on the experimental structure data. Altogether, our work provides strong evidence for the presence of a relationship between the dimensionality of emergent lattice distortions and that of concurrent changes in the electronic properties of CDW materials. The relationship may need to be accounted for when these materials are explored for practical applications.

36 MATERIALS SCIENCE↗

Thermodynamic Modeling of Complex Solid Solutions in the Lu-H-N System via Graph Neural Network Accelerated Monte Carlo Simulations

Metal hydrides are important across diverse applications, such as hydrogen storage, batteries, gas sensors, nuclear reactions, and high-temperature superconductivity. Previous computational studies of metal hydrides under extreme pressures, e.g., 𝑂⁡(10 2 ) ⁢GPa, usually treat them as stoichiometric compounds without considering interstitial lattice disorder. As pressures become more moderate in the 𝑂⁡(10 0 ) ⁢GPa and below range, hydrogen disorder at interstitial lattice sites becomes prominent, e.g., in the N-doped Lu hydride that was recently claimed superconducting near 1 GPa. Further adding compositional complexity from alloying and/or multielement interstitial occupation makes elucidating pressure- and temperature-dependent observables intractable by first-principles calculations alone. We therefore propose a lattice graph neural-network surrogate modeling approach to predict configuration- and pressure-dependent equation-of-state properties. Their efficiency permits Monte Carlo simulations to calculate Gibbs energies and pressure-dependent phase diagrams, thereby revealing insights into the synthesis conditions required for achieving desired phase equilibria. We demonstrate this concept for the compositionally complex cubic Lu(H,N,Va) 3 system where three constituents (hydrogen, nitrogen and vacancy) have disordered multielement interstitial occupancies and insights into pressure-dependent phase equilibria are critically needed, e.g., N-doping levels can significantly lower dehydrogenation temperatures and provide a new strategy to optimize hydrogen-storage alloys. This work can improve the thermodynamic understanding of the Lu-H-N system and help rational synthesis of N-doped Lu hydrides, but more generally demonstrates an efficient approach to model pressure-dependent thermodynamics of multicomponent solid solutions.

Monte Carlo methods↗

Weak trimerization in the frustrated two-dimensional triangular Heisenberg antiferromagnet Lu y Y 1-y MnO 3

To understand the 2D triangular Heisenberg antiferromagnetic system, we investigated the magnetic structures and the dynamics of Lu y Y 1-y MnO 3 in detail. The substitutions are adjusted to the Mn atomic position close to x Mn =$\frac{1}{3}$. The neutron powder diffraction data claims that the magnetic structure of Lu y Y 1-y MnO 3 is described as a mixture of Γ 3 (P6' 3 cm') and Γ 4 (P6' 3 c'm) at the x Mn position for y=0.15, 0.30, and 0.45. The ratio of Γ 3 and Γ 4 depends on temperature and composition and the fraction of Γ 3 increases upon cooling, while no clear trimerization was observed at the xMn position. We estimated exchange parameters from the analysis of the low-energy part of the spin waves. The results showed a weak trimerization effect on cooling because the nearest-neighbor exchange interaction is slightly enhanced. The temperature dependence of the spin-wave dispersion around the Γ point shows that the spin gap closes with increasing temperature because the exchange interactions in the nearest Mn-Mn neighbor become smaller. Gapless diffusive magnetic excitation from a Mn triangular lattice has been observed in a wide range in Q and E space of Lu y Y 1-y MnO 3 . We found that Lu 0.3 Y 0.7 MnO 3 could be an ideal case to investigate the trimerization, frustrated magnetism, and magnetoelastic coupling often observed in two-dimensional triangular lattice Heisenberg antiferromagnet systems.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

Sci—Thur AM: YIS - 03: irtGPUMCD: a new GPU-calculated dosimetry code for {sup 177}Lu-octreotate radionuclide therapy of neuroendocrine tumors

In peptide receptor radionuclide therapy (PRRT), huge inter-patient variability in absorbed radiation doses per administered activity mandates the utilization of individualized dosimetry to evaluate therapeutic efficacy and toxicity. We created a reliable GPU-calculated dosimetry code (irtGPUMCD) and assessed {sup 177}Lu-octreotate renal dosimetry in eight patients (4 cycles of approximately 7.4 GBq). irtGPUMCD was derived from a brachytherapy dosimetry code (bGPUMCD), which was adapted to {sup 177}Lu PRRT dosimetry. Serial quantitative single-photon emission computed tomography (SPECT) images were obtained from three SPECT/CT acquisitions performed at 4, 24 and 72 hours after {sup 177}Lu-octreotate administration, and registered with non-rigid deformation of CT volumes, to obtain {sup 177}Lu-octreotate 4D quantitative biodistribution. Local energy deposition from the β disintegrations was assumed. Using Monte Carlo gamma photon transportation, irtGPUMCD computed dose rate at each time point. Average kidney absorbed dose was obtained from 1-cm{sup 3} VOI dose rate samples on each cortex, subjected to a biexponential curve fit. Integration of the latter time-dose rate curve yielded the renal absorbed dose. The mean renal dose per administered activity was 0.48 ± 0.13 Gy/GBq (range: 0.30–0.71 Gy/GBq). Comparison to another PRRT dosimetry code (VRAK: Voxelized Registration and Kinetics) showed fair accordance with irtGPUMCD (11.4 ± 6.8 %, range: 3.3–26.2%). These results suggest the possibility to use the irtGPUMCD code in order to personalize administered activity in PRRT. This could allow improving clinical outcomes by maximizing per-cycle tumor doses, without exceeding the tolerable renal dose.

60 APPLIED LIFE SCIENCES↗

Features of the band structure and conduction mechanisms of n-HfNiSn semiconductor heavily Lu-doped

The crystal and electronic structures, energy, kinetic, and magnetic characteristics of n-HfNiSn semiconductor heavily doped with a Lu acceptor impurity in the ranges T = 80–400 K and N{sub A}{sup Lu} ≈ 1.9 × 10{sup 20}−1.9 × 10{sup 21} cm{sup −3} (x = 0.01–0.10) at H ≤ 10 kG is studied. The nature of the structural-defect generation mechanism leading to changes in the band gap and the degree of semiconductor compensation is determined. Its essence is the simultaneous reduction and elimination of donor-type structural defects due to the displacement of ∼1% of Ni atoms from the Hf (4a) site, the generation of acceptor-type structural defects by substituting Ni atoms with Lu atoms at the 4c site, and the generation of donor-type defects such as vacancies at the Sn (4b) site. The results of calculations of the electronic structure of Hf{sub 1−x}Lu{sub x}NiSn are in agreement with experimental data. The results are discussed within the model of a heavily doped and compensated Shklovskii-Efros semiconductor.

36 MATERIALS SCIENCE↗

Yb:Lu 2 O 3 single-crystal fiber: spectroscopy, amplification, and lasing

For the first time, to our knowledge, a lutetium oxide (Lu 2 O 3 ) single-crystal fiber (SCF) laser is demonstrated. The laser heated pedestal growth (LHPG) technique was used to pull Yb-doped Lu 2 O 3 SCFs between 10 and 50 mm long and with diameters between 150 and 225 μm. Spectroscopic properties are first reported in detail, as the two-site nature of the host demands careful attention. Short 10 mm long, unclad fibers were used as amplifier media in a single pass copropagating configuration. Then, a 50 mm long 0.1%Yb:Lu 2 O 3 SCF with a 180 μm diameter was configured to lase by butt-coupling mirrors on the ends and pumping at 976 nm. Lasing occurred at the 1033 nm peak of Yb, and a maximum output of around 300 mW is reported. Finally, the results indicate there is no, at least obvious, fundamental reason that should deter future interest in Lu 2 O 3 as a SCF platform, which has been considered to have high potential for power scaling based on its beneficial intrinsic properties.

47 OTHER INSTRUMENTATION↗

A diagonally inverted LU implicit multigrid scheme

A new Diagonally Inverted LU Implicit scheme is developed within the framework of the multigrid method for the 3-D unsteady Euler equations. The matrix systems that are to be inverted in the LU scheme are treated by local diagonalizing transformations that decouple them into systems of scalar equations. Unlike the Diagonalized ADI method, the time accuracy of the LU scheme is not reduced since the diagonalization procedure does not destroy time conservation. Even more importantly, this diagonalization significantly reduces the computational effort required to solve the LU approximation and therefore transforms it into a more efficient method of numerically solving the 3-D Euler equations.

Yokota, Jeffrey W.↗

A diagonally inverted LU implicit multigrid scheme

A new Diagonally Inverted LU Implicit scheme is developed within the framework of the multigrid method for the three-dimensional unsteady Euler equations. The matrix systems that are to be inverted in the LU scheme are treated by local diagonalizing transformations that decouple them into systems of scalar equations. Unlike the Diagonalized ADI method, the time accuracy of the LU scheme is not reduced since the diagonalizing procedure does not destroy time conservation. Even more importantly, this diagonalization significantly reduces the computational effort required to solve the LU approximation and therefore transforms it into a more efficient method of numerically solving the three-dimensional Euler equations.

Yokota, J. W.↗

Block LU factorization

Many of the currently popular 'block algorithms' are scalar algorithms in which the operations have been grouped and reordered into matrix operations. One genuine block algorithm in practical use is block LU factorization, and this has recently been shown by Demmel and Higham to be unstable in general. It is shown here that block LU factorization is stable if A is block diagonally dominant by columns. Moreover, for a general matrix the level of instability in block LU factorization can be founded in terms of the condition number kappa(A) and the growth factor for Gaussian elimination without pivoting. A consequence is that block LU factorization is stable for a matrix A that is symmetric positive definite or point diagonally dominant by rows or columns as long as A is well-conditioned.

Demmel, James W.↗