Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “computational efficiency”

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 343 records · Page 19

MAS: A versatile Landau-fluid eigenvalue code for plasma stability analysis in general geometry

We have developed a new global eigenvalue code, Multiscale Analysis for plasma Stabilities (MAS), for studying plasma problems with wave toroidal mode number (n) and frequency (ω) in a broad range of interest in general tokamak geometry, based on a five-field Landau-fluid description of thermal plasmas. Beyond keeping the necessary plasma fluid response, we further retain the important kinetic effects including diamagnetic drift, ion finite Larmor radius, finite parallel electric field (E||), ion and electron Landau resonances in a self-consistent and non-perturbative manner without sacrificing the attractive efficiency in computation. The physical capabilities of the code are evaluated and examined in the aspects of both theory and simulation. In theory, the comprehensive Landau-fluid model implemented in MAS can be reduced to the well-known ideal MHD model, electrostatic ion-fluid model, and drift-kinetic model in various limits, which clearly delineates the physics validity regime. In simulation, MAS has been well benchmarked with theory and other gyrokinetic and kinetic-MHD hybrid codes in a manner of adopting the unified physical and numerical framework, which covers the kinetic Alfv\'en wave (KAW), ion sound wave (ISW), low-n kink, high-n ion temperature gradient mode (ITG) and kinetic ballooning mode (KBM). Moreover, MAS is successfully applied to model the Alfv\'en eigenmode (AE) activities in DIII-D discharge #159243, which faithfully captures the frequency sweeping of reversed shear Alfv\'en eigenmode (RSAE), the tunneling damping of toroidal Alfv\'en eigenmode (TAE), as well as the polarization characteristics of kinetic beta-induced Alfv\'en eigenmode (KBAE) and beta-induced Alfv\'en-acoustic eigenmode (BAAE) being consistent with former gyrokinetic theory and simulation. With respect to the key progress contributed to the community, MAS has the advantage of combining rich physics ingredients, realistic global geometry and high computation efficiency together for plasma stability analysis in linear regime.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Dynamics of rotationally periodic large space structure

The analysis of large area rotationally periodic space structures presented in the paper combines the finite element method, transfer matrix procedures, approximation methods, and periodic structure analysis to obtain computational efficiency. The computations used in the analysis indicate that additive damping mechanisms can be evaluated from the frequency response of the structure. The transient response can also be obtained from the frequency response to complete the dynamic analysis.

Mcdaniel, T. J.↗

Eigenvalue sensitivity analysis of planar frames with variable joint and support locations

Two sensitivity equations are derived in this study based upon the continuum approach for eigenvalue sensitivity analysis of planar frame structures with variable joint and support locations. A variational form of an eigenvalue equation is first derived in which all of the quantities are expressed in the local coordinate system attached to each member. Material derivative of this variational equation is then sought to account for changes in member's length and orientation resulting form the perturbation of joint and support locations. Finally, eigenvalue sensitivity equations are formulated in either domain quantities (by the domain method) or boundary quantities (by the boundary method). It is concluded that the sensitivity equation derived by the boundary method is more efficient in computation but less accurate than that of the domain method. Nevertheless, both of them in terms of computational efficiency are superior to the conventional direct differentiation method and the finite difference method.

Chuang, Ching H.↗

Variational Methods in Sensitivity Analysis and Optimization for Aerodynamic Applications

Variational methods (VM) sensitivity analysis, which is the continuous alternative to the discrete sensitivity analysis, is employed to derive the costate (adjoint) equations, the transversality conditions, and the functional sensitivity derivatives. In the derivation of the sensitivity equations, the variational methods use the generalized calculus of variations, in which the variable boundary is considered as the design function. The converged solution of the state equations together with the converged solution of the costate equations are integrated along the domain boundary to uniquely determine the functional sensitivity derivatives with respect to the design function. The determination of the sensitivity derivatives of the performance index or functional entails the coupled solutions of the state and costate equations. As the stable and converged numerical solution of the costate equations with their boundary conditions are a priori unknown, numerical stability analysis is performed on both the state and costate equations. Thereafter, based on the amplification factors obtained by solving the generalized eigenvalue equations, the stability behavior of the costate equations is discussed and compared with the state (Euler) equations. The stability analysis of the costate equations suggests that the converged and stable solution of the costate equation is possible only if the computational domain of the costate equations is transformed to take into account the reverse flow nature of the costate equations. The application of the variational methods to aerodynamic shape optimization problems is demonstrated for internal flow problems at supersonic Mach number range. The study shows, that while maintaining the accuracy of the functional sensitivity derivatives within the reasonable range for engineering prediction purposes, the variational methods show a substantial gain in computational efficiency, i.e., computer time and memory, when compared with the finite difference sensitivity analysis.

Ibrahim, A. H.↗

Variational Methods in Design Optimization and Sensitivity Analysis for Two-Dimensional Euler Equations

Variational methods (VM) sensitivity analysis employed to derive the costate (adjoint) equations, the transversality conditions, and the functional sensitivity derivatives. In the derivation of the sensitivity equations, the variational methods use the generalized calculus of variations, in which the variable boundary is considered as the design function. The converged solution of the state equations together with the converged solution of the costate equations are integrated along the domain boundary to uniquely determine the functional sensitivity derivatives with respect to the design function. The application of the variational methods to aerodynamic shape optimization problems is demonstrated for internal flow problems at supersonic Mach number range. The study shows, that while maintaining the accuracy of the functional sensitivity derivatives within the reasonable range for engineering prediction purposes, the variational methods show a substantial gain in computational efficiency, i.e., computer time and memory, when compared with the finite difference sensitivity analysis.

Ibrahim, A. H.↗

Molecular Dynamics Calculations

The development of thermodynamics and statistical mechanics is very important in the history of physics, and it underlines the difficulty in dealing with systems involving many bodies, even if those bodies are identical. Macroscopic systems of atoms typically contain so many particles that it would be virtually impossible to follow the behavior of all of the particles involved. Therefore, the behavior of a complete system can only be described or predicted in statistical ways. Under a grant to the NASA Lewis Research Center, scientists at the Case Western Reserve University have been examining the use of modern computing techniques that may be able to investigate and find the behavior of complete systems that have a large number of particles by tracking each particle individually. This is the study of molecular dynamics. In contrast to Monte Carlo techniques, which incorporate uncertainty from the outset, molecular dynamics calculations are fully deterministic. Although it is still impossible to track, even on high-speed computers, each particle in a system of a trillion trillion particles, it has been found that such systems can be well simulated by calculating the trajectories of a few thousand particles. Modern computers and efficient computing strategies have been used to calculate the behavior of a few physical systems and are now being employed to study important problems such as supersonic flows in the laboratory and in space. In particular, an animated video (available in mpeg format--4.4 MB) was produced by Dr. M.J. Woo, now a National Research Council fellow at Lewis, and the G-VIS laboratory at Lewis. This video shows the behavior of supersonic shocks produced by pistons in enclosed cylinders by following exactly the behavior of thousands of particles. The major assumptions made were that the particles involved were hard spheres and that all collisions with the walls and with other particles were fully elastic. The animated video was voted one of two winning videos in a competition held at the meeting of the American Physical Society's Division of Fluid Dynamics, held in Atlanta, Georgia, in November 1994. Of great interest was the result that in every shock there were a few high-speed precursor particles racing ahead of the shock, carrying information about its impending arrival. Most recently, Dr. Woo has been applying molecular dynamics techniques to the problem of determining the drag produced by the space station truss structure as it flies through the thin residual atmosphere of low-Earth orbit. This problem is made difficult by the complex structure of the truss and by the extreme supersonic nature of the flow. A fully filled section of the truss has already been examined, and drag predictions have been made. Molecular dynamics techniques promise to make realistic drag calculations possible even for very complex partially filled truss segments flying at arbitrary angles.

Source record↗

Higher Order Chemistry Models in the CFD Simulation of Laser-Ablated Carbon Plumes

Production of single-walled carbon nanotubes (SWNT) has taken place for a number of years and by a variety of methods such-as laser ablation, chemical vapor deposition, and arc-jet ablation. Yet, little is actually understood about the exact chemical kinetics and processes that occur in SWNT formation. In recent time, NASA Johnson Space Center has devoted a considerable effort to the experimental evaluation of the laser ablation production process for SWNT originally developed at Rice University. To fully understand the nature of the laser ablation process it is necessary to understand the development of the carbon plume dynamics within the laser ablation oven. The present work is a continuation of previous studies into the efforts to model plume dynamics using computational fluid dynamics (CFD). The ultimate goal of the work is to improve understanding of the laser ablation process, and through that improved understanding, refine the laser ablation production of SWNT. Fig. 1 shows a basic schematic of the laser-ablation oven at NASA-JSC. Construction of the facility is simple in concept. Two concentric quartz tubes of 1.5 mm thickness form the inner and outer tubes with inside diameters of 2.2 and 5.08 cm respectively. At one end of the inner tube are located two 60 Hz pulsed lasers operating at 1064 nm and 532 nm wavelength with beam diameters of 5 mm aligned coaxially with the longitudinal axis of the inner quartz tube. For standard nanotube production runs, a 10 ns 532 nm pulse is followed 50 ns later by a 10 ns 1064 nm pulse. Each pulse is of 300 mJ energy. A target of carbon graphite with approximately 1% nickel and cobalt catalysts is located at the other end of the inner quartz tube. In the ordinary processing of SWNT, a base flow of 100 sccm of argon is maintained from the laser location and exits past the carbon target at a pressure of 66.7 kPa. These conditions yield a baseline mass flow through the chamber of 2.723x10(exp -6)kg/s of argon. The whole oven facility is heated to a temperature of 1473 K prior to nanotube production runs. Upon laser irradiation, part of the carbon target ablates immediately and forms a carbon vapor plume that penetrates into the argon base flow towards the laser initially at supersonic velocities. In the previous studies either a single carbon species, C3, was used to model the plume development, or a simplified 11 species carbon model reduced down from the Krestinin and Moravsky model for full fullerene chemistry was used. While both of these options yielded significant results, it was felt that the actual chemistry occ&g in the carbon plume might have a greater affect on the plume than assumed. Indeed, in the earlier attempts at modeling the carbon plume, several thermophysical characteristics could never be matched to experimental observations of plume development - mainly the propagation distance of the plume itself. In the present study, two additional chemistry models will be used to duplicate the previous studies simulations of the carbon plume. The first chemistry model used in this study is again a reduced form of the Krestinin and Moravsky rates. However the highest order carbon species allowed has been increased from C6 to C30 - therefore allowing the simulation of up to a half of the standard C60 fullerene. The second chemistry model investigated is a reduced form of a full carbon nanotube model developed at NASA-JSC. The C30 studies have already been accomplished at the present time, and the reduced SWNT model studies are currently underway. To pursue the current study, one sacrifice had to be made in that the simulation grid spacing had to be increased from 0.5 mm spacing to 1 mm spacing for the sake of computational efficiency since computational effort is proportional to the square of the number of grid points multiplied by the number of species considered. propagation that is far more in line with the experimental results observed by Puretzky et al as shown in Fig. 2. e C6 studies had yielded a far greater propagation in previous studies. In addition, chemical species development with the C30 model indicates that many higher order carbon species are produced outside of the plume proper (indicated by plotting contours of the background argon concentrations in Fig. 3) - this result was not observed in previous studies. In fact, some species primarily occurred outside of the plume itself - as shown for C27 in Fig. 4 when it is compared to Fig. 3. It could be asserted that this has occurred because all of the C27 in the plume had already been consumed in the formation of C30, but this does not seem to be indicated over time. Several other factors that arose in the previous studies have also been made more clear by the use of the higher order chemical models - one being that the use of c6 as an indicator species was mistaken. C6 is the only carbon species in the previous studies that was not injected into the flowfield as a boundary condition; it was therefore hoped that this species would provide insight into the formation of higher order carbon species for comparison to full SWNT production. But, when the plot of total mass in the plume is examined on a species by species basis in Fig. 5, it is seen that Cg was a fairly insignificant contributor to the total carbon mass in the plume and would not provide information on higher order carbon formation. the thermophysical characteristics of the carbon plume as well as simulate the carbon plume using the reduced SWNT model to provide an even better simulation of full chemistry effects upon plume propagation.

Scott, C. D.↗

Raster-Based Approach to Solar Pressure Modeling

An algorithm has been developed to take advantage of the graphics processing hardware in modern computers to efficiently compute high-fidelity solar pressure forces and torques on spacecraft, taking into account the possibility of self-shading due to the articulation of spacecraft components such as solar arrays. The process is easily extended to compute other results that depend on three-dimensional attitude analysis, such as solar array power generation or free molecular flow drag. The impact of photons upon a spacecraft introduces small forces and moments. The magnitude and direction of the forces depend on the material properties of the spacecraft components being illuminated. The parts of the components being lit depends on the orientation of the craft with respect to the Sun, as well as the gimbal angles for any significant moving external parts (solar arrays, typically). Some components may shield others from the Sun. The purpose of this innovation is to enable high-fidelity computation of solar pressure and power generation effects of illuminated portions of spacecraft, taking self-shading from spacecraft attitude and movable components into account. The key idea in this innovation is to compute results dependent upon complicated geometry by using an image to break the problem into thousands or millions of sub-problems with simple geometry, and then the results from the simpler problems are combined to give high-fidelity results for the full geometry. This process is performed by constructing a 3D model of a spacecraft using an appropriate computer language (OpenGL), and running that model on a modern computer's 3D accelerated video processor. This quickly and accurately generates a view of the model (as shown on a computer screen) that takes rotation and articulation of spacecraft components into account. When this view is interpreted as the spacecraft as seen by the Sun, then only the portions of the craft visible in the view are illuminated. The view as shown on the computer screen is composed of up to millions of pixels. Each of those pixels is associated with a small illuminated area of the spacecraft. For each pixel, it is possible to compute its position, angle (surface normal) from the view direction, and the spacecraft material (and therefore, optical coefficients) associated with that area. With this information, the area associated with each pixel can be modeled as a simple flat plate for calculating solar pressure. The vector sum of these individual flat plate models is a high-fidelity approximation of the solar pressure forces and torques on the whole vehicle. In addition to using optical coefficients associated with each spacecraft material to calculate solar pressure, a power generation coefficient is added for computing solar array power generation from the sum of the illuminated areas. Similarly, other area-based calculations, such as free molecular flow drag, are also enabled. Because the model rendering is separated from other calculations, it is relatively easy to add a new model to explore a new vehicle or mission configuration. Adding a new model is performed by adding OpenGL code, but a future version might read a mesh file exported from a computer-aided design (CAD) system to enable very rapid turnaround for new designs

Wright, Theodore W. II↗

Arithmetic Primitives for Efficient Neuromorphic Computing

Neuromorphic computing is steadily gaining popularity in many scientific and engineering disciplines. However, one of the biggest problems that has prevented widespread usage of neuromorphic computing is the lack of efficient encoding methods. Traditional encoding methods such as binning, rate encoding, and temporal encoding are based on unary encoding and generate a large number of spikes for certain applications, making them less energy efficient. Lack of better encoding methods has also prevented preprocessing operations from being carried out on neuromorphic computers. As a result, over 99% of the time can be spent on data preprocessing and data transfer operations in some cases, leading to an inefficient workflow. In this paper, we present preliminary results that would enable us to efficiently encode data and perform basic arithmetic operations on neuromorphic computers. First, we present a neuromorphic approach for the two’s complement encoding of numbers and leverage it to devise addition and multiplication circuits, which could be used in preprocessing operations on neuromorphic computers. We test our approach on the SuperNeuroMAT simulator. Our results indicate that two’s complement is a highly efficient encoding method in terms of time, space, and energy complexity and that the addition and multiplication circuits produce accurate results on two numbers having arbitrary precision.

Wurm, Ahna↗

Evaluation of fluxon synapse device based on superconducting loops for energy efficient neuromorphic computing

With Moore’s law nearing its end due to the physical scaling limitations of CMOS technology, alternative computing approaches have gained considerable attention as ways to improve computing performance. Here, we evaluate performance prospects of a new approach based on disordered superconducting loops with Josephson-junctions for energy efficient neuromorphic computing. Synaptic weights can be stored as internal trapped fluxon states of three superconducting loops connected with multiple Josephson-junctions (JJ) and modulated by input signals applied in the form of discrete fluxons (quantized flux) in a controlled manner. The stable trapped fluxon state directs the incoming flux through different pathways with the flow statistics representing different synaptic weights. We explore implementation of matrix–vector-multiplication (MVM) operations using arrays of these fluxon synapse devices. We investigate the energy efficiency of online-learning of MNIST dataset. Our results suggest that the fluxon synapse array can provide ~100× reduction in energy consumption compared to other state-of-the-art synaptic devices. This work presents a proof-of-concept that will pave the way for development of high-speed and highly energy efficient neuromorphic computing systems based on superconducting materials.

42 ENGINEERING↗

Efficient stream distributions in radiative transfer theory

This paper discusses a new, computationally-efficient method for approximating the integro-differential equation of radiative transfer with a finite set of coupled differential equations for discrete streams. The method uses recommended spatial distributions of streams that are quite different from those typically used in that they are based on the symmetry of several regular Platonic solids. To facilitate the use of such distributions, an explicit, one-parameter relationship between the physical radiance and the abstract stream is formulated. The parameter is used to determine the minimum number of streams required in the radiative transfer model. Accuracy and computational efficiency are shown to be served best by choosing a stream distribution that is invariant to a large number of three space rotations. For various values of the above-mentioned parameter, the resulting recommended stream distribution is shown to be more computationally efficient than more conventional stream distributions. Finally, the incorporation of polarization in the stream definition is described.

Whitney, C.↗

Uncertainty Quantification of Geophysical Inversion Using Stochastic Partial Differential Equations (LDRD #218329)

This report summarizes work completed under the Laboratory Directed Research and Development (LDRD) project "Uncertainty Quantification of Geophysical Inversion Using Stochastic Differential Equations." Geophysical inversions often require computationally expensive algorithms to find even one solution, let alone propagating uncertainties through to the solution domain. The primary purpose of this project was to find more computationally efficient means to approximate solution uncertainty in geophysical inversions. We found multiple computationally efficient methods of propagating Earth model uncertainty into uncertainties in solutions of full waveform seismic moment tensor inversions. However, the optimum method of approximating the uncertainty in these seismic source solutions was to use the Karhunen-Love theorem with data misfit residuals. This method was orders of magnitude more computationally efficient than traditional Monte Carlo methods and yielded estimates of uncertainty that closely approximated those of Monte Carlo. We will summarize the various methods we evaluated for estimating uncertainty in seismic source inversions as well as work toward this goal in the realm of 3-D seismic tomographic inversion uncertainty.

58 GEOSCIENCES↗

Hybrid discrete/continuum algorithms for stochastic reaction networks

Direct solutions of the Chemical Master Equation (CME) governing Stochastic Reaction Networks (SRNs) are generally prohibitively expensive due to excessive numbers of possible discrete states in such systems. To enhance computational efficiency we develop a hybrid approach where the evolution of states with low molecule counts is treated with the discrete CME model while that of states with large molecule counts is modeled by the continuum Fokker–Planck equation. The Fokker–Planck equation is discretized using a 2nd order finite volume approach with appropriate treatment of flux components. The numerical construction at the interface between the discrete and continuum regions implements the transfer of probability reaction by reaction according to the stoichiometry of the system. The performance of this novel hybrid approach is explored for a two-species circadian model with computational efficiency gains of about one order of magnitude.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Using hybrid implicit Monte Carlo diffusion to simulate gray radiation hydrodynamics

This work describes how to couple a hybrid Implicit Monte Carlo Diffusion (HIMCD) method with a Lagrangian hydrodynamics code to evaluate the coupled radiation hydrodynamics equations. This HIMCD method dynamically applies Implicit Monte Carlo Diffusion (IMD) [1] to regions of a problem that are opaque and diffusive while applying standard Implicit Monte Carlo (IMC) [2] to regions where the diffusion approximation is invalid. We show that this method significantly improves the computational efficiency as compared to a standard IMC/Hydrodynamics solver, when optically thick diffusive material is present, while maintaining accuracy. Two test cases are used to demonstrate the accuracy and performance of HIMCD as compared to IMC and IMD. The first is the Lowrie semi-analytic diffusive shock [3]. The second is a simple test case where the source radiation streams through optically thin material and heats a thick diffusive region of material causing it to rapidly expand. We found that HIMCD proves to be accurate, robust, and computationally efficient for these test problems.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

A Medium-/Low-Voltage Joint State Estimator Through Linear Uncertainty Propagation

Traditionally, distribution system state estimations (DSSE) are challenged by the lack of measurements at both primary and secondary sides of the system. The widely available cable television (CATV) voltage sensors installed in low-voltage (LV) networks bring opportunities to achieve higher quality DSSE covering a broader area of the distribution network. This study proposes a medium-/low-voltage (MV/LV) joint distribution system state estimation approach using the untapped CATV measurements. It aims at addressing the need for system situational awareness at the grid edge while improving the estimation accuracy at both the primary and secondary sides compared to its disjointed counterpart. Linearized measurement functions and boundary condition uncertainty propagation rules are derived to ensure the computational efficiency and accuracy of the joint state estimator. Numerical experiments are conducted on an IEEE test feeder to demonstrate the efficacy of the proposed method and the value of CATV measurements.

joint state estimation↗

Adaptive nonlocal means filtering based on local noise level for CT denoising

Purpose: To develop and evaluate an image-domain noise reduction method based on a modified nonlocal means (NLM) algorithm that is adaptive to local noise level of CT images and to implement this method in a time frame consistent with clinical workflow. Methods: A computationally efficient technique for local noise estimation directly from CT images was developed. A forward projection, based on a 2D fan-beam approximation, was used to generate the projection data, with a noise model incorporating the effects of the bowtie filter and automatic exposure control. The noise propagation from projection data to images was analytically derived. The analytical noise map was validated using repeated scans of a phantom. A 3D NLM denoising algorithm was modified to adapt its denoising strength locally based on this noise map. The performance of this adaptive NLM filter was evaluated in phantom studies in terms of in-plane and cross-plane high-contrast spatial resolution, noise power spectrum (NPS), subjective low-contrast spatial resolution using the American College of Radiology (ACR) accreditation phantom, and objective low-contrast spatial resolution using a channelized Hotelling model observer (CHO). Graphical processing units (GPU) implementation of this noise map calculation and the adaptive NLM filtering were developed to meet demands of clinical workflow. Adaptive NLM was piloted on lower dose scans in clinical practice. Results: The local noise level estimation matches the noise distribution determined from multiple repetitive scans of a phantom, demonstrated by small variations in the ratio map between the analytical noise map and the one calculated from repeated scans. The phantom studies demonstrated that the adaptive NLM filter can reduce noise substantially without degrading the high-contrast spatial resolution, as illustrated by modulation transfer function and slice sensitivity profile results. The NPS results show that adaptive NLM denoising preserves the shape and peak frequency of the noise power spectrum better than commercial smoothing kernels, and indicate that the spatial resolution at low contrast levels is not significantly degraded. Both the subjective evaluation using the ACR phantom and the objective evaluation on a low-contrast detection task using a CHO model observer demonstrate an improvement on low-contrast performance. The GPU implementation can process and transfer 300 slice images within 5 min. On patient data, the adaptive NLM algorithm provides more effective denoising of CT data throughout a volume than standard NLM, and may allow significant lowering of radiation dose. After a two week pilot study of lower dose CT urography and CT enterography exams, both GI and GU radiology groups elected to proceed with permanent implementation of adaptive NLM in their GI and GU CT practices. Conclusions: This work describes and validates a computationally efficient technique for noise map estimation directly from CT images, and an adaptive NLM filtering based on this noise map, on phantom and patient data. Both the noise map calculation and the adaptive NLM filtering can be performed in times that allow integration with clinical workflow. The adaptive NLM algorithm provides effective denoising of CT data throughout a volume, and may allow significant lowering of radiation dose.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

Gluon field digitization via group space decimation for quantum computers

Efficient digitization is required for quantum simulations of gauge theories. Schemes based on discrete subgroups use fewer qubits at the cost of systematic errors. We systematize this approach by deriving a single plaquette action for approximating general continuous gauge groups through integrating out field fluctuations. This provides insight into the effectiveness of these approximations, and how they could be improved. We accompany the scheme by simulations of pure gauge over the largest discrete subgroup of SU(3) up to the third order.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗