SEARCH · Search NASA
Results for “Iterative methods”
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.
Scalable fabrication of a tough and recyclable spore-bearing biocomposite thermoplastic polyurethane
Thermoplastic polyurethanes (TPUs) are a class of versatile thermoplastic elastomers, but most of their products lack a proper recycling strategy or have no end-of-life solutions. To pursue a sustainable end-of-life solution for TPU-based products, self-disintegrating biocomposite TPUs have recently been developed by embedding spores of TPU-degrading bacteria into TPUs via melt extrusion. Herein, we improve upon spore-bearing biocomposites and demonstrate industrially relevant manufacturing conditions for fabricating biocomposite TPUs. To minimize the coloration of biocomposite TPUs, spore production was modified. The innate brown color of the resulting materials was diminished by reducing FeSO 4 in sporulation media, generating white spores without compromising spore productivity, viability, morphology or heat-shock tolerance. Biocomposite TPUs containing white spores displayed a 45 % increase in toughness compared to TPUs without spores, while retaining ∼ 90 % spore viability post processing. Furthermore, biocomposite TPU fabrication was demonstrated using a scalable continuous extruder followed by injection molding. Biocomposite TPUs generated by these industry-relevant processes exhibited comparable toughness improvement and spore viability to biocomposite TPU prepared using a lab scale microcompounder, while enhancing productivity by 30-fold. Finally, spore addition significantly improved the recyclability of biocomposite TPUs, enabling 80 % toughness retention after 5 rounds of iterative melt processing. Additionally, no negative effect on the lifespan of the generated TPUs was observed over 1 year of storage. Overall, this study confirms that spore-bearing biocomposite TPUs are promising for practical applications, offering an accessible method to enhance toughness and sustainability of commercial TPUs through the incorporation of spore-based living fillers.
NSFnets (Navier-Stokes flow nets): Physics-informed neural networks for the incompressible Navier-Stokes equations
In the last 50 years there has been a tremendous progress in solving numerically the Navier-Stokes equations using finite differences, finite elements, spectral, and even meshless methods. Yet, in many real cases, we still cannot incorporate seamlessly (multi-fidelity) data into existing algorithms, and for industrial-complexity applications the mesh generation is time consuming and still an art. Moreover, solving ill-posed problems (e.g., lacking boundary conditions) or inverse problems is often prohibitively expensive and requires different formulations and new computer codes. Here, we employ physics-informed neural networks (PINNs), encoding the governing equations directly into the deep neural network via automatic differentiation, to overcome some of the aforementioned limitations for simulating incompressible laminar and turbulent flows. We develop the Navier-Stokes flow nets (NSFnets) by considering two different mathematical formulations of the Navier-Stokes equations: the velocity-pressure (VP) formulation and the vorticity-velocity (VV) formulation. Since this is a new approach, we first select some standard benchmark problems to assess the accuracy, convergence rate, computational cost and flexibility of NSFnets; analytical solutions and direct numerical simulation (DNS) databases provide proper initial and boundary conditions for the NSFnet simulations. The spatial and temporal coordinates are the inputs of the NSFnets, while the instantaneous velocity and pressure fields are the outputs for the VP-NSFnet, and the instantaneous velocity and vorticity fields are the outputs for the VV-NSFnet. This is unsupervised learning and, hence, no labeled data are required beyond boundary and initial conditions and the fluid properties. The residuals of the VP or VV governing equations, together with the initial and boundary conditions, are embedded into the loss function of the NSFnets. No data is provided for the pressure to the VP-NSFnet, which is a hidden state and is obtained via the incompressibility constraint without extra computational cost. Unlike the traditional numerical methods, NSFnets inherit the properties of neural networks (NNs), hence the total error is composed of the approximation, the optimization, and the generalization errors. Here, we empirically attempt to quantify these errors by varying the sampling (“residual”) points, the iterative solvers, and the size of the NN architecture. For the laminar flow solutions, we show that both the VP and the VV formulations are comparable in accuracy but their best performance corresponds to different NN architectures. The initial convergence rate is fast but the error eventually saturates to a plateau due to the dominance of the optimization error. For the turbulent channel flow, we show that NSFnets can sustain turbulence at , but due to expensive training we only consider part of the channel domain and enforce velocity boundary conditions on the subdomain boundaries provided by the DNS data base. We also perform a systematic study on the weights used in the loss function for balancing the data and physics components, and investigate a new way of computing the weights dynamically to accelerate training and enhance accuracy. In the last part, we demonstrate how NSFnets should be used in practice, namely for ill-posed problems with incomplete or noisy boundary conditions as well as for inverse problems. We obtain reasonably accurate solutions for such cases as well without the need to change the NSFnets and at the same computational cost as in the forward well-posed problems. As a result, we also present a simple example of transfer learning that will aid in accelerating the training of NSFnets for different parameter settings.
The energy confinement response of DIII-D plasmas to Resonant Magnetic Perturbations
Here, Resonant Magnetic Perturbations (RMPs) are a leading method for edge localized modes (ELMs) Control in fusion plasmas. However they can also cause a rapid degradation in energy confinement. In this paper we show that the energy confinement in low collisionality (v* e < 0.3) DIII-D ITER Similar Shape (ISS) plasmas often recovers after several energy confinement times for RMP amplitudes up to the threshold for ELM suppression. Immediately following the application of the RMP, the plasma stored energy decreases in proportion to the decrease in the line-averaged density during density "pump-out". Later in the discharge confinement recovery is observed in the thermal ion channel and is correlated with the increase in the ion temperature at the top of the H-mode pedestal. A correlation between the inverse scale length of the ion temperature (α/L Ti ) and the E x B shearing rate at the top of the pedestal is seen during the confinement recovery phase. Transport analysis reveals that the confinement improvement in the ion channel results from the self-similarity in the ion temperature profiles in the plasma core combined with the observed increase in α/L Ti in the plasma edge following density pump-out. In contrast the electron temperature scale length (α/L Ti ) remains essentially unchanged in response to the application of the RMP. At significantly higher RMP levels the edge EXB shearing rate and α/L Ti does not increase and the confinement does not recover following density pump-out.
A Three-Phase Microgrid Restoration Model Considering Unbalanced Operation of Distributed Generation
Recent severe outages highlight the urgency of improving grid resiliency in the U.S. Microgrid formation schemes are proposed to restore critical loads after outages occur. Most distribution networks have unbalanced configurations that are not represented in sufficient detail by single-phase models. This study provides a microgrid formation plan that adopts a three-phase network model to represent unbalanced distribution networks. The problem formulation has a quadratic objective function with mixed-integer linear constraints. The three-phase network model enables us to examine the three-phase power outputs of distributed generators (DGs), preventing unbalanced operation that might trip DGs. Because the DG unbalanced operation constraint is non-convex, an iterative process is presented that checks whether the unbalanced operation limits for DGs are satisfied after each iteration of optimization. We also develop a relatively conservative linear approximation on the unbalanced operation constraint to handle larger networks. Compared with the iterative solution process, the conservative linear approximation is able to accelerate the solution process at the cost of sacrificing optimality to a limited extent. Simulation in the IEEE 34 node and IEEE 123 test feeders indicate that the proposed method yields more practical microgrid formations results. In addition, this paper explores the coordinated operation of DGs and energy storage (ES) installations. The unbalanced three-phase outputs of ESs combined with the relatively balanced outputs of DGs could supply unbalanced loads. In conclusion, the case study also validates the DG-ES coordination.
Multigrid Optimization for Large-Scale Ptychographic Phase Retrieval
Ptychography is a popular imaging technique that combines diffractive imaging with scanning microscopy. The technique consists of a coherent beam that is scanned across an object in a series of overlapping positions, leading to reliable and improved reconstructions. Ptychographic microscopes allow for large fields to be imaged at high resolution at the cost of additional computational expense. Herein, we propose a multigrid-based optimization framework to reduce the computational burdens of large-scale ptychographic phase retrieval. Our proposed method exploits the inherent hierarchical structures in ptychography through tailored restriction and prolongation operators for the object and data domains. Our numerical results show that our proposed scheme accelerates the convergence of its underlying solver and outperforms the ptychographical iterative engine, a workhorse in the optics community.
PYRANDA
Pyranda Is fortran 90 code which solves compact finite difference equations on a three dimensional fortran arrays, domain decomposed using the Message Passing Interface (MPI). The fortran source solves the linear system of equations which forms a pentadtagona1 matrix. Derivatives are used to construct hyperbolic partlal differential equation (PDE) operators, namely gradient, divergence, Laplacian, and curl operators. 5. Solution Method: The library provides a python Interface which facilitates problem setup and description. However, the core solver Is written in the fortran 90 language. The solver couples MPI communication and linear solves together so that a direct (non-iterative) solution to the linear system of equations can be computed with minimal communication. Domain decomposition is limited to a uniform data size but provides perfect load balancing. 6. Computer for which the code Is written:
Removal of correlated background in a high-order harmonic transient absorption spectra with principal component regression
We demonstrate a 40x mean noise power reduction (NPR) in core-to-valence extreme ultraviolet (XUV) femtosecond transient absorption spectroscopy with a high harmonic generation (HHG) light source. An adaptive iteratively reweighted principal component regression (airPCR) is used to analyze and suppress spectrally correlated HHG intensity fluctuations. The technique requires significantly less user input and leads to a higher mean NPR than a previously introduced edge-pixel PCR method that relies on the manual identification of signal-free spectral regions. Both techniques are applied in a time-resolved XUV absorption study of the 2 snp 1 P o ( n ≥ 2) autoionizing Rydberg states of helium, demonstrating sub-10 −3 optical density sensitivity.
Spread Spectrum Time Domain Reflectometry and Steepest Descent Inversion to Measure Complex Impedance
In this paper, we present a method for estimating complex impedances using reflectometry and a modified steepest descent inversion algorithm. We simulate spread spectrum time domain reflectometry (SSTDR), which can measure complex impedances on energized systems for an experimental setup with resistive and capacitive loads. A parametric function, which includes both a misfit function and stabilizer function, is created. The misfit function is a least squares estimate of how close the model data matches observed data. The stabilizer function prevents the steepest descent algorithm from becoming unstable and diverging. Steepest descent iteratively identifies the model parameters that minimize the parametric function. We validate the algorithm by correctly identifying the model parameters (capacitance and resistance) associated with simulated SSTDR data, with added 3 dB white Gaussian noise. With the stabilizer function, the steepest descent algorithm estimates of the model parameters are bounded within a specified range. Furthermore, the errors for capacitance (220pF to 820pF) and resistance (50 Ω to 270 Ω) are < 10%, corresponding to a complex impedance magnitude |R +1/jωC| of 53 Ω to 510 Ω.
Method For Production of Diamond Hardfaced Titanium Carbide/Titanium Bits For Geothermal Drilling
During Phase I, Diamond Materials, Inc. developed compositions and methodology to produce dense superhard D-TiC/Ti composites and performed feasibility studies to characterize structure, hardness, strength, toughness, thermal stability, corrosive wear and oxidation resistance for rock drilling and other applications. As expected multiple iterations were required to optimize the composition and microstructure of the D-TiC/Ti composite to achieve the desired super-hardness, drilling efficiency and high temperature corrosive-wear resistance. It is now experimentally established that D-TiC/Ti composites can be fabricated at much lower pressures and temperatures than D-WC/Co composites, thus manufacturing costs can be significantly reduced. It is also evident that this new process will enable large monolithic articles of superhard materials, including drill bits, to be made using conventional hot pressing equipment of very large working volume.
Scattering amplitudes and N -body post-Minkowskian Hamiltonians in general relativity and beyond
We present a general framework for calculating post-Minskowskian, classical, conservative Hamiltonians for N non-spinning bodies in general relativity from relativistic scattering amplitudes. Novel features for N > 2 are described including the subtraction of tree-like iteration contributions and the calculation of non-trivial many-body Fourier transform integrals needed to construct position space potentials. A new approach to calculating these integrals as an expansion in the hierarchical limit is described based on the method of regions. As an explicit example, we present the O(G 2 ) 3-body momentum space potential in general relativity as well as for charged bodies in Einstein-Maxwell. The result is shown to be in perfect agreement with previous post-Newtonian calculations in general relativity up to O(G 2 v 4 ). Furthermore, in appropriate limits the result is shown to agree perfectly with relativistic probe scattering in multi-center extremal black hole backgrounds and with the scattering of slowly-moving extremal black holes in the moduli space approximation.
Low-Cost HTS Based Magnet System with an Inductively Coupled Pulsed Energy Extraction Protection System
In this project, RadiaBeam Technologies was tasked with developing a novel solution for a cost effective quench protection based on fast expansion of the normal zone. By inductively coupling a strong electromagnetic pulse via a resonant LC circuit, we attempted to demonstrate accelerated normal zone propagation. The AC field induces currents in the superconducting layer with the current density exceeding that of the critical current density, J c . This creates a large normal zone, uniformly distributing the dissipation through the magnet body. The method does not rely on thermal heating of the conductor, thus enabling nearly instantaneous protection. Through the course of the Phase II project, RadiaBeam Technologies continued extensive numerical modeling of the inductive quench system, re-designed and built several iterations of the POC system for testing and observed evidence of a transient partial quench being induced. However the final device was not fabricated. This was a consequence of the fundamentally complex nature of the energy extraction process and the challenges associated even with demonstrating the proof of concept in a bench top device.
Testing, Characterization, and Modeling of the Resonant Plate Test Environment
The resonant plate shock test is a dynamic test of a mid-field pyroshock environment where a projectile is struck against a plate. The structure undergoing the simulated field shock is mounted to the plate. The plate resonates when struck and provides a two sided shock that is representative of the shock observed in the field. This test environment shock simulates a shock in a single coordinate direction for components looking to provide evidence that they will survive a similar or less shock when deployed in their operating environment. However, testing in one axis at a time provides many challenges. The true environment is a multi-axis environment. The test environment exhibits strong off-axis motion when only motion in one axis is desired. Multiple fixtures are needed for a single test series. It would be advantageous if a single test could be developed that tests the multi-axis environment simultaneously. In order to design such a test, a model must be developed and validated. The model can be iterated in design and configuration until the specified multi-axis environment is met. The test can then execute the model driven test design. This report discusses the resonant plate model needed to design future tests and the steps and methods used to obtain the model. This report also details aspects of the resonant plate test discovered during the process of model development that aids in our understanding of the test.
Modeling and Design of the Triton OWC System (Final Report)
Two versions of the Triton Oscillating Water Column type device will be modeled in WEC Sim. First, a model of a wave-tank scale device which can be tuned and validated against tank test data. Second, a model of the deployment-scale device will be constructed following the method of the tank-scale device to ensure that relevant physics are captured. In the latter case, the geometry, PTO architecture, and other design details are not yet finalized, so the model will serve as a platform for design iteration as time and budget allows. A subset of this iteration will be automated using existing WEC-Sim capabilities. A primary focus of this work will be familiarizing Triton personnel with the WEC-Sim workflow and model details so that the model of the deployment device can continue to be enhanced after project end.
On the Formalization of Development and Assessment Process for Digital Twins in the Nearly Autonomous Management and Control System
In recent years, the autonomous control system has been encouraged in advanced reactors for restoring economic viability, simplifying the operation and maintenance, and enabling remote-site power generations [1]. Since the reactor is expected to be operated for a long period of time with a limited number of individuals onsite, it is recommended that the autonomous control system should have access to very realistic models of the state of processes in the whole lifecycle, together with these process behaviors in interaction with their environment in the real world. As a result, digital twin (DT) technology is suggested in autonomous control systems. DT is defined as a digital representation of a physical object or system, which contains a record for the histories of loads, operation and maintenance status, predictions for the near-term transient of important state variables, and decision-making process [2]. Since machine learning (ML) can recognize patterns within a complex system in real-time applications, it has been used to build DTs in the autonomous control systems for advanced reactors. Meanwhile, due to the rareness of operation data in accident scenarios, the development and assessment of DTs is expected to be mainly driven by simulations. Although the capability and feasibility of ML-based DTs are recognized in improving the safety and efficiency of reactor control, a major concern from the regulatory commission and the nuclear industry is whether the information from a DT is developed and assessed in accordance with expectation and requirements by the target decision. Such concerns not only affect the acceptance criteria for DTs, but also values that can be extracted from DTs and autonomous control system during operations. Inspired by the success of formal methods in improving the reliability and robustness of computer programming and software development, it is suggested that the development and assessment process (DAP) for both separate DTs and integral control system should be formalized in a transparent, consistent, and improvable manner. In this study, a digital-twin development and assessment process (DT-DAP) is proposed by adapting the evaluation model development and assessment process (EMDAP) [3] to requirements by the autonomous control system, ML algorithms, and DT technology. To demonstrate the framework, a baseline nearly autonomous management and control (NAMAC) system with ML-based DTs for diagnosis and prognosis is developed and assessed based on the framework. It is found that with selected testing methods and techniques, the DT-DAP can help identify errors in DTs and NAMAC which would otherwise be left unverified. Meanwhile, it is found that the DT-DAP can improve the DTs and NAMAC by continuously learning and iterating through different elements.
Reinforcement Learning for Distance Maximization in Energy-Constrained Robots
Our goal is to teach a robot to walk in an energy efficient way. Inspired by nature [1], we create a reinforcement learning method that balances distance traveled and energy consumption, which we call “utility.” Further, each step is symmetric to ensure equal strides while walking straight. We achieve our goal by rewarding decisions that lead to the best utility over several training iterations.
ClimGen: Learning the Forcing-Response Relationship in Climate System
Solar Radiation Management (SRM) is emerging as a potential geoengineering strategy to address the anthropogenic impact on climate, but its effective implementation requires an iterative and large ensemble of highly accurate and efficient climate projections. Traditional climate projections rely on executing computationally demanding and time-consuming numerical climate models. Recent advances in machine learning (ML) aim to enhance these approaches by emulating traditional methods. In this work, we propose a novel framework for directly learning the relationship between solar radiation flux at the top of the atmosphere and the corresponding surface temperature response. To evaluate the feasibility of this direct ML-based projection, we developed a dataset using an intermediate complexity model, incorporating a comprehensive suite of different forcing patterns and evaluation metrics to rigorously assess the ML model’s performance. We introduce a Conditional Denoising Diffusion Probabilistic Model (cDDPM) for this task, which demonstrates encouraging skill in representing climate statistics under previously unseen forcing patterns. This approach provides a promising pathway for direct climate projections by accurately learning the forcing-response relationship, with a wide range of applications in impact mitigation, emissions policy design, and SRM strategies.
Development of an inexpensive matrix-assisted laser desorption—time of flight mass spectrometry method for the identification of endophytes and rhizobacteria cultured from the microbiome associated with maize
Many endophytes and rhizobacteria associated with plants support the growth and health of their hosts. The vast majority of these potentially beneficial bacteria have yet to be characterized, in part because of the cost of identifying bacterial isolates. Matrix-assisted laser desorption-time of flight (MALDI-TOF) has enabled culturomic studies of host-associated microbiomes but analysis of mass spectra generated from plant-associated bacteria requires optimization. In this study, we aligned mass spectra generated from endophytes and rhizobacteria isolated from heritage and sweet varieties of Zea mays. Multiple iterations of alignment attempts identified a set of parameters that sorted 114 isolates into 60 coherent MALDI-TOF taxonomic units (MTUs). These MTUs corresponded to strains with practically identical (>99%) 16S rRNA gene sequences. Mass spectra were used to train a machine learning algorithm that classified 100% of the isolates into 60 MTUs. These MTUs provided >70% coverage of aerobic, heterotrophic bacteria readily cultured with nutrient rich media from the maize microbiome and allowed prediction of the total diversity recoverable with that particular cultivation method. Acidovorax sp., Pseudomonas sp. and Cellulosimicrobium sp. dominated the library generated from the rhizoplane. Relative to the sweet variety, the heritage variety contained a high number of MTUs. The ability to detect these differences in libraries, suggests a rapid and inexpensive method of describing the diversity of bacteria cultured from the endosphere and rhizosphere of maize.