Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Numerical optimization”

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 1,279 records · Page 71

Learning high-dimensional parametric maps via reduced basis adaptive residual networks

We propose a scalable framework for the learning of high-dimensional parametric maps via adaptively constructed residual network (ResNet) maps between reduced bases of the inputs and outputs. When just few training data are available, it is beneficial to have a compact parametrization in order to ameliorate the ill-posedness of the neural network training problem. By linearly restricting high-dimensional maps to informed reduced bases of the inputs, one can compress high-dimensional maps in a constructive way that can be used to detect appropriate basis ranks, equipped with rigorous error estimates. A scalable neural network learning framework is thus to learn the nonlinear compressed reduced basis mapping. Unlike the reduced basis construction, however, neural network constructions are not guaranteed to reduce errors by adding representation power, making it difficult to achieve good practical performance. Inspired by recent approximation theory that connects ResNets to sequential minimizing flows, we present an adaptive ResNet construction algorithm. This algorithm allows for depth-wise enrichment of the neural network approximation, in a manner that can achieve good practical performance by first training a shallow network and then adapting. We prove universal approximation of the associated neural network class for $L^2_v$ functions on compact sets. Our overall framework allows for constructive means to detect appropriate breadth and depth, and related compact parametrizations of neural networks, significantly reducing the need for architectural hyperparameter tuning. Numerical experiments for parametric PDE problems and a 3D CFD wing design optimization parametric map demonstrate that the proposed methodology can achieve remarkably high accuracy for limited training data, and outperformed other neural network strategies we compared against.

42 ENGINEERING↗

Co-optimized machine-learned manifold models for large eddy simulation of turbulent combustion

Many modeling approaches in large eddy simulation (LES) of turbulent combustion employ a projection of the thermochemical state onto a low-dimensional manifold within state space to reduce the number of transported variables and hence computational cost. Flamelet-generated manifolds (FGM) is an example of a well-established, physics-based approach, but increasingly, principal component analysis (PCA) is being used as a data-driven method for generating manifold models. For both approaches, the nonlinear relationship between the location on the predefined manifold and the outputs of interest, such as reaction rates, can be tabulated or encoded in a neural network. This work proposes a new approach for manifold modeling that extends these existing approaches. A modified neural network structure simultaneously encodes the definition of the manifold variables, the nonlinear mapping, and the subfilter closure for LES. This allows all three of these aspects of the model to be co-optimized, generating a model from any source of combustion thermochemical state data. The manifold parameterizing variables are constrained to be linear combinations of species, as in FGM and PCA-based models, to aid in interpretability and implementation. For LES, subfilter variances of the manifold variables are also included as inputs. Two types of a priori analysis are performed to evaluate the new approach. In the first, the model is trained on data from one-dimensional premixed flames. In this case, the approach recovers the behavior of flamelet-based manifold approaches, and in fact slightly improves performance by identifying an optimized progress variable. The approach is also applied to data from direct numerical simulations of spherical ignition kernels in isotropic turbulence. For any specified manifold dimensionality, the new approach provides substantially lower prediction errors than a PCA-based model developed from the same data set. Additionally, the LES formulation of the new approach can provide accurate predictions for filtered reaction rates across a variety of filter widths.

97 MATHEMATICS AND COMPUTING↗

A direct-adjoint approach for material point model calibration with application to plasticity

Here, this paper proposes a new approach for the calibration of material parameters in local elastoplastic constitutive models. The calibration is posed as a constrained optimization problem, where the constitutive model evolution equations for a single material point serve as constraints. The objective function quantifies the mismatch between the stress predicted by the model and corresponding experimental measurements. To improve calibration efficiency, a novel direct-adjoint approach is presented to compute the Hessian of the objective function, which enables the use of second-order optimization algorithms. Automatic differentiation is used for gradient and Hessian computations. Two numerical examples are employed to validate the Hessian matrices and to demonstrate that the Newton–Raphson algorithm consistently outperforms gradient-based algorithms such as L-BFGS-B.

36 MATERIALS SCIENCE↗

The evolutionary rate dynamically tracks changes in HIV-1 epidemics

Large-sequence datasets provide an opportunity to investigate the dynamics of pathogen epidemics. Thus, a fast method to estimate the evolutionary rate from large and numerous phylogenetic trees becomes necessary. Based on minimizing tip height variances, we optimize the root in a given phylogenetic tree to estimate the most homogenous evolutionary rate between samples from at least two different time points. Simulations showed that the method had no bias in the estimation of evolutionary rates and that it was robust to tree rooting and topological errors. We show that the evolutionary rates of HIV-1 subtype B and C epidemics have changed over time, with the rate of evolution inversely correlated to the rate of virus spread. For subtype B, the evolutionary rate slowed down and tracked the start of the HAART era in 1996. Subtype C in Ethiopia showed an increase in the evolutionary rate when the prevalence increase markedly slowed down in 1995. Thus, we show that the evolutionary rate of HIV-1 on the population level dynamically tracks epidemic events.

59 BASIC BIOLOGICAL SCIENCES↗

Development of lean, efficient, and fast physics-framed deep-learning-based proxy models for subsurface carbon storage

In this work, we present deep-learning-based surrogate models for CCUS developed with four different algorithms and a physics-framed two-phase flow problem involving displacement of water by CO 2 . The deep-learning models were trained using 3D datasets describing the pressure plume, CO 2 saturation plume, and water extraction rate generated by numerical simulation. The hyperparameters defining the architecture of the neural networks were optimized to determine the slimmest network size and training parameters that give the most efficient performance at the least training cost. To develop a robust model that closely mimics the governing physical laws, the discretized form of the two-phase fluid transport equation was used to formulate the supervised deep-learning task. The algorithms investigated in this study predicted the data to above 95% accuracy, with the multi-layer perceptron model demonstrating the best performance by balancing training speed, prediction time, and prediction accuracy with lean network capacity. Furthermore, the surrogate models simultaneously predict reservoir pressure and CO 2 saturation in every grid block, including the surface well extraction rate and bottomhole pressure, at all simulation times for a given static model realization in just a few seconds on a standard desktop computer. A key outcome of this study is that limits can be placed on network design parameters to avoid over designing neural networks, with associated efficiencies in training and prediction times. This is very useful because large volumes of data may be generated in CCUS projects and over-design of neural network architectures imposes penalties that are antithetical to the goal of near-real time forecasting.

58 GEOSCIENCES↗

Hybrid immersed interface-immersed boundary methods for AC dielectrophoresis

Dielectrophoresis, a nonlinear electrokinetic transport mechanism, has become popular in many engineering applications including manipulation, characterization and actuation of biomaterials, particles and biological cells. In this paper, we present a hybrid immersed interface–immersed boundary method to study AC dielectrophoresis where an algorithm is developed to solve the complex Poisson equation using a real variable formulation. An immersed interface method is employed to obtain the AC electric field in a fluid media with suspended particles and an immersed boundary method is used for the fluid equations and particle transport. The convergence of the proposed algorithm as well as validation of the hybrid scheme with experimental results is presented. In this paper, the Maxwell stress tensor is used to calculate the dielectrophoretic force acting on particles by considering the physical effect of particles in the computational domain. Thus, this study eliminates the approximations used in point dipole methods for calculating dielectrophoretic force. A comparative study between Maxwell stress tensor and point dipole methods for computing dielectrophoretic forces are presented. The hybrid method is used to investigate the physics of dielectrophoresis in microfluidic devices using an AC electric field. The numerical results show that with proper design and appropriate selection of applied potential and frequency, global electric field minima can be obtained to facilitate multiple particle trapping by exploiting the mechanism of negative dielectrophoresis. Our numerical results also show that electrically neutral particles form a chain parallel to the applied electric field irrespective of their initial orientation when an AC electric field is applied. This proposed hybrid numerical scheme will help to better understand dielectrophoresis and to design and optimize microfluidic devices.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

A cell-centered AMR-ALE framework for 3D multi-material hydrodynamics. Part II: linesweep ALE rezoning for nonconformal block-structured AMR meshes

The simulation of flows presenting contact discontinuities, vorticity, and large variations in spatial scales can be performed in a framework coupling Arbitrary Lagrangian Eulerian (ALE) algorithms and Adaptive Mesh Refinement (AMR). This coupling requires adaptation of ALE rezoning techniques to meshes containing nonconformal nodes arising from both the AMR topology and the junction of mesh blocks. Here, in this paper, we present an ALE rezoning strategy that is compatible with such meshes, and that can also act as a disentangling algorithm. Emphasis is put on an algorithm that respects intrinsic Lagrangian mesh properties in order to preserve accuracy around discontinuities. To that end, we adapt the weighted linesweep algorithm to nonconformal block-structured AMR meshes. Then, we present control parameters introduced in the method for it to be applicable in practical situations. Notably, the method is coupled to a specific metric optimization in order to palliate some shortcomings of the linesweep method. Finally, numerical test cases are presented that feature the capabilities of the ALE-AMR algorithm for flows that present discontinuities, vorticity, and a variety of scales. Notably, we show that our ALE-AMR algorithm gives results at least similar to Euler-AMR, but provides better accuracy in cases where discontinuities are involved, thanks to a method that respects the Lagrangian features of the mesh. Additionally, it enables Euler-AMR-like computations on domains with temporally varying domain boundaries.

Adaptive mesh refinement↗

Integration of discrete-event dynamics and machining dynamics for machine tool: Modeling, analysis and algorithms

Machining dynamics research lays a solid foundation for machining operations by providing stable combinations of spindle speed and depth of cut. Furthermore, machine learning has been applied to predict tool life as a function of cutting speed. However, the existing research does not consider the discrete-event dynamics in machine shop, i.e., the machine tool needs to process a series of parts in queue under various practical production requirements. This paper addresses the integration of discrete-event dynamics and machining dynamics to achieve cost savings in machining. A learning-based cost function is first proposed for the studied integrated optimization problem of machine tool. The proposed cost function utilizes the predicted tool life under different stable cutting speeds for further optimizing speed selection of machine tool to deal with the discrete-event dynamics in machine shop. Then, according to the practical production requirements, effective mathematical optimization models are developed for the related integrated optimization problems with the consideration of cost, makespan and due date, respectively. Numerical results show the effectiveness of our proposed methods and also the potential to be used in practice.

Ma, Mason↗

Moderate deviations-based importance sampling for stochastic recursive equations

Subsolutions to the Hamilton–Jacobi–Bellman equation associated with a moderate deviations approximation are used to design importance sampling changes of measure for stochastic recursive equations. Analogous to what has been done for large deviations subsolution-based importance sampling, these schemes are shown to be asymptotically optimal under the moderate deviations scaling. Here, we present various implementations and numerical results to contrast their performance, and also discuss the circumstances under which a moderate deviation scaling might be appropriate.

97 MATHEMATICS AND COMPUTING↗

In Situ Velocity‐Strain Sensitivity Near the San Jacinto Fault Zone Analyzed Through Train Tremors

Abstract We utilize train tremors as P‐wave seismic sources to investigate velocity‐strain sensitivity near the San Jacinto Fault Zone. A dense nodal array deployed at the Piñon Flat Observatory is used to detect and identify repeating train energy emitted from a railway in the Coachella valley. We construct P‐wave correlation functions across the fault zone and estimate the spatially averaged dt / t versus strain sensitivity to be 6.25 × 10 4 . Through numerical simulations, we explore how the sensitivity decays exponentially with depth. The optimal solution reveals a subsurface sensitivity of 1.2 × 10 5 and a depth decay rate of 0.05 km −1 . This sensitivity aligns with previous findings but is toward the higher end, likely due to the fractured fault‐zone rocks. The depth decay rate, previously unreported, is notably smaller than assumed in empirical models. This raises the necessity of further investigations of this parameter, which is crucial to study stress and velocity variations at seismogenic depth.

Geology↗

Mechanical characterization of thin TiO{sub 2} films by means of microelectromechanical systems-based cantilevers

The measurement of mechanical parameters by means of microcantilever structures offers a reliable and accurate alternative to traditional methods, especially when dealing with thin films, which are extensively used in microfabrication technology and nanotechnology. In this work, microelectromechanical systems (MEMS)-based piezoresistive cantilevers were realized and used for the determination of Young's modulus and residual stress of thin titanium dioxide (TiO{sub 2}) deposited by sputtering from a TiO{sub 2} target using a rf plasma discharge. Films were deposited at different thicknesses, ranging from a few to a hundred nanometers. Dedicated silicon microcantilevers were designed through an optimization of geometrical parameters with the development of analytical as well as numerical models. Young's modulus and residual stress of sputtered TiO{sub 2} films were assessed by using both mechanical characterization based on scanning profilometers and piezoresistive sensing elements integrated in the silicon cantilevers. Results of MEMS-based characterization were combined with the tribological and morphological properties measured by microscratch test and x-ray diffraction analysis.

36 MATERIALS SCIENCE↗

Design of a magnetic shielding system for the time of flight enhanced diagnostics neutron spectrometer at Experimental Advanced Superconducting Tokamak

The novel neutron spectrometer TOFED (Time of Flight Enhanced Diagnostics), comprising 90 individual photomultiplier tubes coupled with 85 plastic scintillation detectors through light guides, has been constructed and installed at Experimental Advanced Superconducting Tokamak. A dedicated magnetic shielding system has been constructed for TOFED, and is designed to guarantee the normal operation of photomultiplier tubes in the stray magnetic field leaking from the tokamak device. Experimental measurements and numerical simulations carried out employing the finite element method are combined to optimize the design of the magnetic shielding system. The system allows detectors to work properly in an external magnetic field of 200 G.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Ab initio determination of Bose-Hubbard parameters for two ultracold atoms in an optical lattice using a three-well potential

We calculate numerically the energy spectrum of the six dimensional problem of two interacting Bosons in a three-well optical lattice. The particles interact via a full Born-Oppenheimer potential, which can be adapted to model the behavior of the s-wave scattering length at Feshbach resonances. By adjusting the parameters of the corresponding Bose-Hubbard (BH) Hamiltonian the deviation between the numerical energy spectrum and the BH spectrum is minimized. This defines the optimal BH parameter set, which we compare to the standard parameters of the BH model. The range of validity of the BH model with these parameter sets is examined, and an analytical prediction of the interaction parameter is introduced. Furthermore, an extended BH model and implications due to the energy dependence of the scattering length and couplings to higher Bloch bands at a Feshbach resonance are discussed.

74 ATOMIC AND MOLECULAR PHYSICS↗

Search on a hypercubic lattice using a quantum random walk. II. d=2

We investigate the spatial search problem on the two-dimensional square lattice, using the Dirac evolution operator discretized according to the staggered lattice fermion formalism. d=2 is the critical dimension for the spatial search problem, where infrared divergence of the evolution operator leads to logarithmic factors in the scaling behavior. As a result, the construction used in our accompanying article [A. Patel and M. A. Rahaman, Phys. Rev. A 82, 032330 (2010)] provides an O({radical}(N)lnN) algorithm, which is not optimal. The scaling behavior can be improved to O({radical}(NlnN)) by cleverly controlling the massless Dirac evolution operator by an ancilla qubit, as proposed by Tulsi [Phys. Rev. A 78, 012310 (2008)]. We reinterpret the ancilla control as introduction of an effective mass at the marked vertex, and optimize the proportionality constants of the scaling behavior of the algorithm by numerically tuning the parameters.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Data-Conforming Data-Driven Control: Avoiding Premature Generalizations Beyond Data

Data-driven and adaptive control approaches face the problem of introducing sudden distributional shifts beyond the distribution of data encountered during learning. Therefore, they are prone to invalidating the very assumptions used in their own construction. This is due to the linearity of the underlying system, inherently assumed and formulated in most data-driven control approaches, which may falsely generalize the behavior of the system beyond the behavior experienced in the data. This article seeks to mitigate these problems by enforcing consistency of the newly designed closed-loop systems with data and slowing down any distributional shifts in the joint state-input space. This is achieved through incorporating affine regularization terms and linear matrix inequality constraints to data-driven approaches, resulting in convex semi-definite programs that can be efficiently solved by standard software packages. We discuss the optimality conditions of these programs and then conclude this article with a numerical example that further highlights the problem of premature generalization beyond data and shows the effectiveness of our proposed approaches in enhancing the safety of data-driven control methods.

97 MATHEMATICS AND COMPUTING↗

WE-AB-303-09: Rapid Projection Computations for On-Board Digital Tomosynthesis in Radiation Therapy

Purpose: To facilitate fast and accurate iterative volumetric image reconstruction from limited-angle on-board projections. Methods: Intrafraction motion hinders the clinical applicability of modern radiotherapy techniques, such as lung stereotactic body radiation therapy (SBRT). The LIVE system may impact clinical practice by recovering volumetric information via Digital Tomosynthesis (DTS), thus entailing low time and radiation dose for image acquisition during treatment. The DTS is estimated as a deformation of prior CT via iterative registration with on-board images; this shifts the challenge to the computational domain, owing largely to repeated projection computations across iterations. We address this issue by composing efficient digital projection operators from their constituent parts. This allows us to separate the static (projection geometry) and dynamic (volume/image data) parts of projection operations by means of pre-computations, enabling fast on-board processing, while also relaxing constraints on underlying numerical models (e.g. regridding interpolation kernels). Further decoupling the projectors into simpler ones ensures the incurred memory overhead remains low, within the capacity of a single GPU. These operators depend only on the treatment plan and may be reused across iterations and patients. The dynamic processing load is kept to a minimum and maps well to the GPU computational model. Results: We have integrated efficient, pre-computable modules for volumetric ray-casting and FDK-based back-projection with the LIVE processing pipeline. Our results show a 60x acceleration of the DTS computations, compared to the previous version, using a single GPU; presently, reconstruction is attained within a couple of minutes. The present implementation allows for significant flexibility in terms of the numerical and operational projection model; we are investigating the benefit of further optimizations and accurate digital projection sub-kernels. Conclusion: Composable projection operators constitute a versatile research tool which can greatly accelerate iterative registration algorithms and may be conducive to the clinical applicability of LIVE. National Institutes of Health Grant No. R01-CA184173; GPU donation by NVIDIA Corporation.

60 APPLIED LIFE SCIENCES↗

Teko: A block preconditioning capability with concrete example applications in Navier--Stokes and MHD

This study describes the design of Teko, an object-oriented C++ library for implementing advanced block preconditioners. Mathematical design criteria that elucidate the needs of block preconditioning libraries and techniques are explained and shown to motivate the structure of Teko. For instance, a principal design choice was for Teko to strongly reflect the mathematical statement of the preconditioners to reduce development burden and permit focus on the numerics. Additional mechanisms are explained that provide a pathway to developing an optimized production capable block preconditioning capability with Teko. Finally, Teko is demonstrated on fluid flow and magnetohydrodynamics applications. In addition to highlighting the features of the Teko library, these new results illustrate the effectiveness of recent preconditioning developments applied to advanced discretization approaches.

97 MATHEMATICS AND COMPUTING↗

optimas v0.1

optimas is a Python library that can coordinate a large number of numerical simulations on high-performance computing resources, with the aim of optimizing a given simulation result. More specifically, this library is currently used for design optimization of laser-plasma particle accelerators. In this context, the performance of particle accelerator designs are often evaluated with large-scale simulation codes, and many separate simulations need to be run - with different design parameters - in order to find the most performant design. optimas facilitates this process by providing a convenient interface to advanced optimizers (e.g. Bayesian optimization), and by coordinating the execution of the different simulations on HPC resources. (This is done by leveraging the library libensemble.) Compared to other open-source optimization libraries (e.g. Ax), optimas is more tailored towards execution on DOE HPC resources (e.g. Perlmutter, Summit, etc.) and is specialized for the type of simulation codes and workflows that are used in the community of laser-plasma acceleration.

Lehe, Remi↗