Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “inverse problems”

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 1,639 records · Page 91
All sourcesNASA NTRSDOE OSTI
DOE OSTI2020

Framework for simulating gauge theories with dipolar spin systems

Gauge theories appear broadly in physics, ranging from the standard model of particle physics to long-wavelength descriptions of topological systems in condensed matter. However, systems with sign problems are largely inaccessible to classical computations and also beyond the current limitations of digital quantum hardware. In this work, we develop an analog approach to simulating gauge theories with an experimental setup that employs dipolar spins (molecules or Rydberg atoms). We consider molecules fixed in space and interacting through dipole-dipole interactions, avoiding the need for itinerant degrees of freedom. Each molecule represents either a site or gauge degree of freedom, and Gauss's law is preserved by a direct and programmatic tuning of positions and internal state energies. This approach can be regarded as a form of analog systems programming and charts a path forward for near-term quantum simulation. As a first step, we numerically validate this scheme in a small-system study of U(1) quantum link models in (1+1) dimensions with link spin S = 1/2 and S =1 and illustrate how dynamical phenomena such as string inversion and string breaking could be observed in near-term experiments. Our work brings together methods from atomic and molecular physics, condensed matter physics, high-energy physics, and quantum information science for the study of nonperturbative processes in gauge theories.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2022

Density matrix renormalization group description of the island of inversion isotopes F 28 – 33

Recent experiments have confirmed that the neutron-rich isotopes 28,29 F belong to the so-called island of inversion (IOI), a region of the nuclear chart around Z = 10 and N = 20 where nuclear structure deviates from the standard shell model predictions due to deformation and continuum effects. However, while the general principles leading to the IOI are relatively well understood, the details of the low-lying structure of the exotic fluorine isotopes 28–33 F are basically unknown. In this study, we perform large-scale shell model calculations including continuum states to investigate the properties of the neutron-rich isotopes 25–33 F, from a core of 24 O and using an effective two-body interaction with a small number of adjustable parameters in the central and tensor channels. We develop two models adjusted on experimentally confirmed states in 25,26 O and 25–27 F based on different assumptions concerning the positions of the neutron 0d 3/2 and 1p 3/2 shells, and solve the many-body problem using the density matrix renormalization group (DMRG) method for open quantum systems in an sd–fp model space. We obtain the first detailed spectroscopy of 25–33F in the continuum and show how the interplay between continuum effects and deformation explains the recent data on 28,29 F. Several deformed one- and two-neutron halo states are predicted in 29,31 F, and we provide some information about the possible structure of the heaviest fluorine isotopes. We also suggest several experimental studies of interest to constraint models and test the present predictions.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗
NASA NTRS1994

Dynamically consistent hydrography and absolute velocity in the eastern North Atlantic Ocean

The problem of mapping a dynamically consistent hydrographic field and associated absolute geostrophic flow in the eastern North Atlantic between 24 deg and 36 deg N is related directly to the solution of the so-called thermocline equations. A nonlinear optimization problem involving Needler's P equation is solved to find the hydrography and resulting flow that minimizes the vertical mixing above about 1500 m in the ocean and is simultaneously consistent with the observations. A sharp minimum (at least in some dimensions) is found, apparently corresponding to a solution nearly conserving potential vorticity and with vertical eddy coefficient less than about 10(exp -5) sq m/s. Estimates of `residual' quantities such as eddy coefficients are extremely sensitive to slight modifications to the observed fields. Boundary conditions, vertical velocities, etc., are a product of the optimization and produce estimates differing quantitatively from prior ones relying directly upon observed hydrography. The results are generally insensitive to particular elements of the solution methodology, but many questions remain concerning the extent to which different synoptic sections can be asserted to represent the same ocean. The method can be regarded as a practical generalization of the beta spiral and geostrophic balance inverses for the estimate of absolute geostrophic flows. Numerous improvements to the methodology used in this preliminary attempt are possible.

Wunsch, Carl↗
NASA NTRS1994

Physical Retrievals of Over-Ocean Rain Rate from Multichannel Microwave Imagery: Theoretical Characteristics of Normalized Polarization and Scattering Indices - Part 1

Microwave rain rate retrieval algorithms have most often been formulated in terms of the raw brightness temperatures observed by one or more channels of a satellite radiometer. Taken individually, single-channel brightness temperatures generally represent a near-arbitrary combination of positive contributions due to liquid water emission and negative contributions due to scattering by ice and/or visibility of the radiometrically cold ocean surface. Unfortunately, for a given rain rate, emission by liquid water below the freezing level and scattering by ice particles above the freezing level are rather loosely coupled in both a physical and statistical sense. Furthermore, microwave brightness temperatures may vary significantly (approx. 30-70 K) in response to geophysical parameters other than liquid water and precipitation. Because of these complications, physical algorithms which attempt to directly invert observed brightness temperatures have typically relied on the iterative adjustment of detailed micro-physical profiles or cloud models, guided by explicit forward microwave radiative transfer calculations. In support of an effort to develop a significantly simpler and more efficient inversion-type rain rate algorithm, the physical information content of two linear transformations of single-frequency, dual-polarization brightness temperatures is studied: the normalized polarization difference P of Petty and Katsaros (1990, 1992), which is intended as a measure of footprint-averaged rain cloud transmittance for a given frequency; and a scattering index S (similar to the polarization corrected temperature of Spencer et al.,1989) which is sensitive almost exclusively to ice. A reverse Monte Carlo radiative transfer model is used to elucidate the qualitative response of these physically distinct single-frequency indices to idealized 3-dimensional rain clouds and to demonstrate their advantages over raw brightness temperatures both as stand-alone indices of precipitation activity and as primary variables in physical, multichannel rain rate retrieval schemes. As a byproduct of the present analysis, it is shown that conventional plane-parallel analyses of the well-known foot-print-filling problem for emission-based algorithms may in some cases give seriously misleading results.

Petty, G. W.↗
NASA NTRS2016

FORWARD: A Toolset for Multiwavelength Coronal Magnetometry

Determining the 3D coronal magnetic field is a critical, but extremely difficult problem to solve. Since different types of multiwavelength coronal data probe different aspects of the coronal magnetic field, ideally these data should be used together to validate and constrain specifications of that field. Such a task requires the ability to create observable quantities at a range of wavelengths from a distribution of magnetic field and associated plasma i.e., to perform forward calculations. In this paper we describe the capabilities of the FORWARD SolarSoft IDL package, a uniquely comprehensive toolset for coronal magnetometry. FORWARD is a community resource that may be used both to synthesize a broad range of coronal observables, and to access and compare synthetic observables to existing data. It enables forward fitting of specific observations, and helps to build intuition into how the physical properties of coronal magnetic structures translate to observable properties. FORWARD can also be used to generate synthetic test beds from MHD simulations in order to facilitate the development of coronal magnetometric inversion methods, and to prepare for the analysis of future large solar telescope data.

Sun: magnetic fields↗
NASA NTRS1995

Electron distributions in quasi-perpendicular shocks: Simulations

One of the outstanding problems of collisionless shock physics is the microphysical description of electron thermalization and associated wave processes. Observations at the Earth's bow shock have shown various features in the electron velocity distribution such as a loss-cone, temperature anisotropies, and beams before a thermalized, flat-topped distribution is reached downstream. It is generally believed that in addition to the macroscopic fields, which determine the overall mapping of the distribution function, electromagnetic whistler waves and lower hybrid waves are responsible for the pitch-angle scattering and thermalization of the electrons. However, a genuine physical explanation requires analysis of the evolution of the electron distribution in the self-consistent wave fields. Particle simulations are ideally suited to do this, and include processes in which both electrons and ions are involved, such as cross-field streaming instabilities. We show results from l-D and 2-D simulations using an implicit code, which allows time steps large compared to the inverse electron plasma frequency. It is shown that in order to capture the correct propagation characteristics of the waves, a large mass ratio between the protons and electrons is generally necessary. Detailed diagnostics of the waves as well as the evolution of the electron velocity distribution will be presented and will be discussed in the frame work of satellite observations.

Pantellini, F.↗
NASA NTRS1985

MAGSAT scalar anomaly map of South America

A scalar magnetic anomaly map was prepared for South America and adjacent marine areas directly from original MAGSAT orbits. The preparation of the map poses special problems, notably in the separation of external field and crustal anomalies, and in the reduction of data to a common altitude. External fields are manifested in a long-wavelength ring current effect, a medium-wavelength equatorial electrojet, and short-wavelength noise. The noise is reduced by selecting profiles from quiet periods, and since the electrojet is confined primarily to dusk profiles, its effect is minimized by drawing the data set from dawn profiles only. The ring current is corrected through the use of the standard ring current equation, augmented by further filtering with a Butterworth bandpass filter. Under the assumption that the time-variant ring current is best removed when a replication of redundant profiles is achieved, a test set of 25 groups of 3 nearly coincident orbits per group is set up for filtering with a range of long-wavelength cutoffs to determine which cutoff best replicates the residual profiles. Altitude differences are then normalized by an inversion of the profile data onto a grid of equivalent point dipoles, and recalculated at an altitude of 350 km. The resulting map, when compared to the 2 deg averaged map, shows more coherent anomalies, with notable differences in the region affected by the electrojet.

Ridgway, J. R.↗
NASA NTRS2004

Highlights of the Version 8 SBUV and TOMS Datasets Released at this Symposium

Last October was the 25th anniversary of the launch of the SBUV and TOMS instruments on NASA's Nimbus-7 satellite. Total Ozone and ozone profile datasets produced by these and following instruments have produced a quarter century long record. Over time we have released several versions of these datasets to incorporate advances in UV radiative transfer, inverse modeling, and instrument characterization. In this meeting we are releasing datasets produced from the version 8 algorithms. They replace the previous versions (V6 SBUV, and V7 TOMS) released about a decade ago. About a dozen companion papers in this meeting provide details of the new algorithms and intercomparison of the new data with external data. In this paper we present key features of the new algorithm, and discuss how the new results differ from those released previously. We show that the new datasets have better internal consistency and also agree better with external datasets. A key feature of the V8 SBUV algorithm is that the climatology has no influence on inter-annual variability and trends; it only affects the mean values and, to a limited extent, the seasonal dependence. By contrast, climatology does have some influence on TOMS total O3 trends, particularly at large solar zenith angles. For this reason, and also because TOMS record has gaps, md EP/TOMS is suffering from data quality problems, we recommend using SBUV total ozone data for applications where the high spatial resolution of TOMS is not essential.

Bhartia, Pawan K.↗
DOE OSTI2019

A cloud computing approach to massively-parallel distributed stochastic optimization with application to geomechanical reservoir characterization

Many search strategies used to optimize parameters during hydrologic model calibration are inherently limited in that simulations must be performed sequentially in order for the algorithm to converge. As a result, the overall time required to reach a solution increases linearly with the number of steps necessary. In contrast, sampling strategies that attempt to approximate the posterior distribution of the parameters do not require the model simulations to be run in sequence and therefore invite the use of a massively parallel approach across heterogeneous computing environments. We apply a cloud computing methodology to this problem where a head node running on Amazon Web Services continuously produces candidate models that are distributed to arbitrary compute environments to run the simulation, in our case Clemson’s Palmetto cluster computer. Thus the time required to obtain a sufficient number of samples to produce an adequate approximation of the posterior distribution of the parameters is limited by the number of available compute sites. Monte Carlo sampling of the posterior space would be duplicative and inefficient, thus we combine Latin hypercube methods with genetic algorithms to ensure that no large regions of the parameter space are left unexplored, and that high probability regions are investigated thoroughly. We independently assess the fit to different types of data sets without the need to construct a single objective function. Sets of Pareto optimal models are particularly valuable when the quality and information content of data is poorly known or conceptual model errors exist in our simulation. We illustrate our approach for a problem where in situ strain data are used to calibrate the parameters of an oil reservoir in the Avant oil field in Oklahoma. We use well injection and shut-in to stress a deep (500m) confined formation, and observe the geomechanical response using borehole strainmeters installed in the shallow subsurface (30m). These data were then used to calibrate the poroelastic properties of the reservoir and caprock along with the geometry of an elliptic channel associated with the formation. Using our approach we are able to identify issues with model error as well as substantial tradeoffs that occur within the parameters.

inversion, stochastic, geomechanics, poroelastic, ↗
NASA NTRS1989

A design philosophy for multi-layer neural networks with applications to robot control

A system is proposed which receives input information from many sensors that may have diverse scaling, dimension, and data representations. The proposed system tolerates sensory information with faults. The proposed self-adaptive processing technique has great promise in integrating the techniques of artificial intelligence and neural networks in an attempt to build a more intelligent computing environment. The proposed architecture can provide a detailed decision tree based on the input information, information stored in a long-term memory, and the adapted rule-based knowledge. A mathematical model for analysis will be obtained to validate the cited hypotheses. An extensive software program will be developed to simulate a typical example of pattern recognition problem. It is shown that the proposed model displays attention, expectation, spatio-temporal, and predictory behavior which are specific to the human brain. The anticipated results of this research project are: (1) creation of a new dynamic neural network structure, and (2) applications to and comparison with conventional multi-layer neural network structures. The anticipated benefits from this research are vast. The model can be used in a neuro-computer architecture as a building block which can perform complicated, nonlinear, time-varying mapping from a multitude of input excitory classes to an output or decision environment. It can be used for coordinating different sensory inputs and past experience of a dynamic system and actuating signals. The commercial applications of this project can be the creation of a special-purpose neuro-computer hardware which can be used in spatio-temporal pattern recognitions in such areas as air defense systems, e.g., target tracking, and recognition. Potential robotics-related applications are trajectory planning, inverse dynamics computations, hierarchical control, task-oriented control, and collision avoidance.

Vadiee, Nader↗
DOE OSTI2024

In-depth analysis of solar models with high-metallicity abundances and updated opacity tables

Context. As a result of the high-quality constraints available for the Sun, we are able to carry out detailed combined analyses using neutrino, spectroscopic, and helioseismic observations. These studies lay the ground for future improvements of the key physical components of solar and stellar models because ingredients such as the equation of state, the radiative opacities, or the prescriptions for macroscopic transport processes of chemicals are then used to study other stars in the Universe. Aims. We study the existing degeneracies in solar models using the recent high-metallicity spectroscopic abundances by comparing them to helioseismic and neutrino data and discuss the effect on their properties of changes in the micro and macro physical ingredients. Methods. We carried out a detailed study of solar models computed with a high-metallicity composition from the literature based on averaged 3D models that were claimed to resolve the solar modelling problem. We compared these models to helioseismic and neutrino constraints. Results. The properties of the solar models are significantly affected by the use of the recent OPLIB opacity tables and the inclusion of macroscopic transport. The properties of the standard solar models computed using the OPAL opacities are similar to those for which the OP opacities were used. We show that a modification of the temperature gradient just below the base of the convective zone is required to remove the discrepancies in solar models, particularly in the presence of macroscopic mixing. This can be simulated by a localised increase in the opacity of a few percent. Conclusions. We conclude that the existing degeneracies and issues in solar modelling are not removed by using an increase in the solar metallicity, in contradiction to what has been suggested in the recent literature. Therefore, standard solar models cannot be used as an argument for a high-metallicity composition. While further work is required to improve solar models, we note that direct helioseismic inversions indicate a low metallicity in the convective envelope, in agreement with spectroscopic analyses based on full 3D models.

74 ATOMIC AND MOLECULAR PHYSICS↗
DOE OSTI2021

Estimation of constituent properties of concrete materials with an artificial neural network based method

Multi-scale models are developed for heterogeneous concrete materials to estimate their macroscopic mechanical properties in terms of micro-structural data. One crucial challenge of those models is the identification of local properties of constituent phases. In this paper, we present an efficient method based on Artificial Neural Networks (ANN). Typical concrete materials are taken as example. A macroscopic analytical strength criterion is established from three steps of nonlinear homogenization procedure. The macroscopic strength of materials is determined as a function of the frictional coefficient and cohesion of solid cement particles at nanometer scale, intra-particle pores, inter-particle pores and aggregates (inclusions). The objective is to identify the nanoscopic frictional coefficient and cohesion of cement particle from measured macroscopic values of uniaxial compression and tensile strengths. For this purpose, a numerical method based on the ANN is developed. With the analytical macroscopic strength criterion, sensitivity studies are first realized to identify the most important micro-structural parameters influencing the macroscopic strength of concrete. A simplified analytical macroscopic strength criterion is then proposed. A large dataset is further constructed through the inversion of the analytical strength criterion by using the aggregates volume fraction, porosity, macroscopic uniaxial tensile and compressive strengths as input variables and the frictional coefficient and cohesion of cement particles as output unknowns. An ANN model containing four hidden layers and 100 neurons in each layer is constructed and trained by using this dataset. Various types of validation of the ANN model are performed. It is found that the proposed ANN based model can effectively predict the frictional coefficient and cohesion of porous cement paste at the microscopic scale with a very good accuracy.

36 MATERIALS SCIENCE↗
NASA NTRSDate not supplied

Study of Coronal Heating in Solar Active Regions Using Wide-Field Imaging Spectroscopy: Hinode EIS Slot Observations

Understanding the frequency of heating events that keep the coronal plasma at several million Kelvin above the photospheric temperature of~ 6000K, is one of the most important problems in solar astrophysics. Spectroscopic observations of the Sun in the extreme ultraviolet (EUV) indicate that the coronal plasma reaches temperatures from 1 to 5 MK in active regions. It is also established that temperature in active regions can vary strongly with time and, moreover, contain sub-regions that evolve and develop separately. Tracking the spatio-temporal evolution of temperature requires continuous observation of the entire active region via imaging and spectroscopy. Traditional slit imaging spectroscopy probes plasma heating in solar active regions through observations of diagnostic emission lines and the resulting data are spectrally pure. Here, imaging is performed through rastering process, which severely limits co-temporal observations and often can be slow to miss events that evolve at other portions of the active region. In contrast, wide-field imaging spectroscopy offer simultaneous coverage of a large field of view as well as obtain spectral information in the same direction. This data suffers from spatial-spectral confusion, and are called spectroheliograms. Using the state-of-the-art inversion techniques that are developed recently, now spectroheliogram data can be unfolded to yield spectrally pure maps of large fields over long duration of observations. We use wide slit data, usually referred as ‘slot’, from the EUV Imaging Spectrometer (EIS) onboard Hinode satellite, focusing on active region observations. Here, we present our study of coronal heating in an active region using a long duration Hinode EIS slot observation.

Active region heating↗
DOE OSTI2023

Trust-Enhancing Probabilistic Transfer Learning for Sparse and Noisy Data Environments

There is an increasing aspiration to utilize machine learning (ML) for various tasks of relevance to national security. ML models have thus far been mostly applied to tasks and domains that, while impactful, have sufficient volume of data. For predictive tasks of national security relevance, ML models of great capacity (ability to approximate nonlinear trends in input-output maps) are often needed to capture the complex underlying physics. However, scientific problems of relevance to national security are often accompanied by various sources of sparse and/or incomplete data, including experiments and simulations, across different regimes of operation, of varying degrees of fidelity, and include noise with different characteristics and/or intensity. State-of-the-art ML models, despite exhibiting superior performance on the task and domain they were trained on, may suffer detrimental loss in performance in such sparse data environments. This report summarizes the results of the Laboratory Directed Research and Development project entitled Trust-Enhancing Probabilistic Transfer Learning for Sparse and Noisy Data Environments. The objective of the project was to develop a new transfer learning (TL) framework that aims to adaptively blend the data across different sources in tackling one task of interest, resulting in enhanced trustworthiness of ML models for mission- and safety-critical systems. The proposed framework determines when it is worth applying TL and how much knowledge is to be transferred, despite uncontrollable uncertainties. The framework accomplishes this by leveraging concepts and techniques from the fields of Bayesian inverse modeling and uncertainty quantification, relying on strong mathematical foundations of probability and measure theories to devise new uncertainty-aware TL workflows.

97 MATHEMATICS AND COMPUTING↗
NASA NTRS2021

Multiobjective Multidisciplinary Optimization of Low-Boom Supersonic Transports Using Multifidelity Models

A multidisciplinary optimization (MDO) method has been developed to design a computational fluid dynamics (CFD) based low-boom configuration that can be obtained from a Pareto solution of a low-fidelity multiobjective MDO problem with mission constraints. This paper refines the developed MDO method by using multifidelity models for CFD-based multiobjective MDO and a better method for the system-level trade between the target low boom ground noise level and the overland range. The refined MDO method can generate a low-boom configuration that satisfies the mission requirements, has the lowest takeoff gross weight and the longest range for the low-boom mission, trims the low-boom cruise flight with fuel redistributions, and has a reversed equivalent area distribution closely matching a low-boom target with ground noise level below 70 PLdB. The validity of the refined MDO method is demonstrated by a design study of a low-boom supersonic transport that carries 40 passengers, flies a low-boom mission with cruise Mach of 1.7 and range of 3500 nm, and cruises overwater at Mach 1.8 with range of 3882 nm. Moreover, the refined MDO method eliminates the difference between the assumed cruise weight for CFD-based low-boom inverse design optimization and the estimated cruise weight of the optimal inverse design solution with respect to the mission requirements.

MDO↗
NASA NTRS2000

The Use of Vegetation For Estimating Broken-Cloud Optical Properties From Surface Measurements

It is known that because of complex three-dimensional (3D) radiative effects of broken clouds, the retrieval of cloud optical properties from upward measurements based on a one-dimensional (1D) inversion technique almost surely fails. To remove radiative effects of 3D cloud structure, we have developed a new technique that retrieves cloud optical thickness for broken clouds above green vegetation from simultaneous surface measurements in the VIS and Near Infrared (NIR) spectral regions. The theoretical basis of the method is the very different spectral behavior of cloud liquid water drops and green vegetation. For example, cloud optical properties, and hence cloud reflectivities, change little between 650 and 860 nm, while the vegetated surface albedo changes from 0.05 to 0.5 between the same two wavelengths. This spectral contrast in surface albedo suggests using ground measurements at both wavelengths not independently, but as an algebraic combination (a spectral index). For a spectral band in the NIR region, the green vegetation acts as a powerful reflector that "illuminates" horizontally inhomogeneous clouds from below. This provides the extra information needed to largely remove the 3D radiative effects, especially in the case of broken clouds; this in turn allows the retrieval of cloud optical depth using traditional 1D radiative transfer theory. This approach is similar to the so-called Green's function problem for radiative transfer where a laser beam illuminates clouds and the resulting "spot-size" of the reflected light around the beam characterizes cloud properties. We generalize Green's function theory to surf ace-cloud interaction and develop new spectral indices from which broken-cloud optical depth can be retrieved.

Marshak, A.↗
DOE OSTI2020

Soft de Sitter Effective Theory

Calculating the quantum evolution of a de Sitter universe on superhorizon scales is notoriously difficult. To address this challenge, we introduce the Soft de Sitter Effective Theory (SdSET). This framework holds for superhorizon modes whose comoving momentum is far below the UV scale, which is set by the inverse comoving horizon. The SdSET is formulated using the same approach that yields the Heavy Quark Effective Theory. The degrees of freedom that capture the long wavelength dynamics are identified with the growing and decaying solutions to the equations of motion. The operator expansion is organized using a power counting scheme, and loops can be regulated while respecting the low energy symmetries. For massive quantum fields in a fixed de Sitter background, power counting implies that all interactions beyond the horizon are irrelevant. Alternatively, if the fields are very light, the leading interactions are at most marginal, and resumming the associated logarithms using (dynamical) renormalization group techniques yields the evolution equation for canonical stochastic inflation. The SdSET is also applicable to models where gravity is dynamical, including inflation. In this case, diffeomorphism invariance ensures that all interactions are irrelevant, trivially implying the all-orders conservation of adiabatic density fluctuations and gravitational waves. We briefly touch on the application to slow-roll eternal inflation by identifying novel relevant operators. This work serves to demystify many aspects of perturbation theory outside the horizon, and has a variety of applications to problems of cosmological interest.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
NASA NTRS2003

Aerodynamic Design Using Neural Networks

The design of aerodynamic components of aircraft, such as wings or engines, involves a process of obtaining the most optimal component shape that can deliver the desired level of component performance, subject to various constraints, e.g., total weight or cost, that the component must satisfy. Aerodynamic design can thus be formulated as an optimization problem that involves the minimization of an objective function subject to constraints. A new aerodynamic design optimization procedure based on neural networks and response surface methodology (RSM) incorporates the advantages of both traditional RSM and neural networks. The procedure uses a strategy, denoted parameter-based partitioning of the design space, to construct a sequence of response surfaces based on both neural networks and polynomial fits to traverse the design space in search of the optimal solution. Some desirable characteristics of the new design optimization procedure include the ability to handle a variety of design objectives, easily impose constraints, and incorporate design guidelines and rules of thumb. It provides an infrastructure for variable fidelity analysis and reduces the cost of computation by using less-expensive, lower fidelity simulations in the early stages of the design evolution. The initial or starting design can be far from optimal. The procedure is easy and economical to use in large-dimensional design space and can be used to perform design tradeoff studies rapidly. Designs involving multiple disciplines can also be optimized. Some practical applications of the design procedure that have demonstrated some of its capabilities include the inverse design of an optimal turbine airfoil starting from a generic shape and the redesign of transonic turbines to improve their unsteady aerodynamic characteristics.

Rai, Man Mohan↗