Search NASA⌕ Search

SEARCH · Search NASA

Results for “Nonlinear 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 1,243 records · Page 69

Using adaptive grid in modeling rocket nozzle flow

The mechanical behavior of a rocket motor internal flow field results in a system of nonlinear partial differential equations which cannot be solved analytically. However, this system of equations called the Navier-Stokes equations can be solved numerically. The accuracy and the convergence of the solution of the system of equations will depend largely on how precisely the sharp gradients in the domain of interest can be resolved. With the advances in computer technology, more sophisticated algorithms are available to improve the accuracy and convergence of the solutions. An adaptive grid generation is one of the schemes which can be incorporated into the algorithm to enhance the capability of numerical modeling. It is equivalent to putting intelligence into the algorithm to optimize the use of computer memory. With this scheme, the finite difference domain of the flow field called the grid does neither have to be very fine nor strategically placed at the location of sharp gradients. The grid is self adapting as the solution evolves. This scheme significantly improves the methodology of solving flow problems in rocket nozzles by taking the refinement part of grid generation out of the hands of computational fluid dynamics (CFD) specialists and place it into the computer algorithm itself.

Chow, Alan S.↗

Numerical Investigations of High Pressure Acoustic Waves in Resonators

This presentation presents work on numerical investigations of nonlinear acoustic phenomena in resonators that can generate high-pressure waves using acoustic forcing of the flow. Time-accurate simulations of the flow in a closed cone resonator were performed at different oscillation frequencies and amplitudes, and the numerical results for the resonance frequency and fluid pressure increase match the GRC experimental data well. Work on cone resonator assembly simulations has started and will involve calculations of the flow through the resonator assembly with and without acoustic excitation. A new technique for direct calculation of resonance frequency of complex shaped resonators is also being investigated. Script-driven command procedures will also be developed for optimization of the resonator shape for maximum pressure increase.

Athavale, Mahesh↗

Forecasting of Storm Surge Floods Using ADCIRC and Optimized DEMs

Increasing the accuracy of storm surge flood forecasts is essential for improving preparedness for hurricanes and other severe storms and, in particular, for optimizing evacuation scenarios. An interactive database, developed by WorldWinds, Inc., contains atlases of storm surge flood levels for the Louisiana/Mississippi gulf coast region. These atlases were developed to improve forecasting of flooding along the coastline and estuaries and in adjacent inland areas. Storm surge heights depend on a complex interaction of several factors, including: storm size, central minimum pressure, forward speed of motion, bottom topography near the point of landfall, astronomical tides, and most importantly, maximum wind speed. The information in the atlases was generated in over 100 computational simulations, partly by use of a parallel-processing version of the ADvanced CIRCulation (ADCIRC) model. ADCIRC is a nonlinear computational model of hydrodynamics, developed by the U.S. Army Corps of Engineers and the US Navy, as a family of two- and three-dimensional finite element based codes. It affords a capability for simulating tidal circulation and storm surge propagation over very large computational domains, while simultaneously providing high-resolution output in areas of complex shoreline and bathymetry. The ADCIRC finite-element grid for this project covered the Gulf of Mexico and contiguous basins, extending into the deep Atlantic Ocean with progressively higher resolution approaching the study area. The advantage of using ADCIRC over other storm surge models, such as SLOSH, is that input conditions can include all or part of wind stress, tides, wave stress, and river discharge, which serve to make the model output more accurate.

Valenti, Elizabeth↗

Orion Parachute Riser Cutter Development

This paper presents the tests and analytical approach used on the development of a steel riser cutter for the CEV Parachute Assembly System (CPAS) used on the Orion crew module. Figure 1 shows the riser cutter and the steel riser bundle which consists of six individual cables. Due to the highly compressed schedule, initial unavailability of the riser material and the Orion Forward Bay mechanical constraints, JSC primarily relied on a combination of internal ballistics analysis and LS-DYNA simulation for this project. Various one dimensional internal ballistics codes that use standard equation of state and conservation of energy have commonly used in the development of CAD devices for initial first order estimates and as an enhancement to the test program. While these codes are very accurate for propellant performance prediction, they usually lack a fully defined kinematic model for dynamic predictions. A simple piston device can easily and accurately be modeled using an equation of motion. However, the accuracy of analytical models is greatly reduced on more complicated devices with complex external loads, nonlinear trajectories or unique unlocking features. A 3D finite element model of CAD device with all critical features included can vastly improve the analytical ballistic predictions when it is used as a supplement to the ballistic code. During this project, LS-DYNA structural 3D model was used to predict the riser resisting load that was needed for the ballistic code. A Lagrangian model with eroding elements shown in Figure 2 was used for the blade, steel riser and the anvil. The riser material failure strain was fine tuned by matching the dent depth on the anvil with the actual test data. LS-DYNA model was also utilized to optimize the blade tip design for the most efficient cut. In parallel, the propellant type and the amount were determined by using CADPROG internal ballistics code. Initial test results showed a good match with LS-DYNA and CADPROG simulations. Final paper will present a detailed roadmap from initial ballistic modeling and LS-DYNA simulation to the performance testing. Blade shape optimization study will also be presented.

Oguz, Sirri↗

Assessment of Radiated Fan Noise Prediction Capabilities Using Static Engine Test Data

This paper describes further assessment of the CDUCT-LaRC code via comparison with static engine test data. In an effort to improve confidence in the use of CDUCT-LaRC for liner optimization studies addressing realistic three-dimensional geometries, inlet radiated fan noise predictions were performed at 54% and 87% engine speed settings. Predictions were then compared with far-field measurements to assess the approach and implementation. The particular configurations were chosen to exercise the three-dimensional capability of CDUCT-LaRC and it s applicability to realistic configurations and conditions. At the 54% engine speed setting, the predictions capture the general directivity and acoustic treatment effects quite well. Comparisons of the predicted and measured directivity at the 87% power setting were more problematic. This was likely due in part to the difficulties in source specification and possibly the nonlinear nature of buzz-saw tones at this engine operating condition. Overall, the approach captured the basic trends and provided a conservative estimate of liner effects from which relative performance metrics could be inferred.

Nark, Douglas M.↗

Applications of CONMIN to wing design optimization with vortex flow effect

Slender wings on supersonic cruise configurations are expected to be thin and highly swept. As a result, edge-separated vortex flow is inevitable and must be accounted for in aerodynamic analysis and design. The present method is based on the method of suction analogy to calculate the total aerodynamic characteristics. The method requires the solution of the attached flow problem, the latter being solved by a low-order panel method in subsonic and supersonic flow. In essence, the lifting pressure is calculated by using a pressure-doublet distribution satisfying the Prandtl-Glauert equation. From the pressure distribution, the leading-edge suction is calculated. The latter is assumed to be the vortex lift through the method of suction analogy. For a cambered wing, the location of vortex-lift action point is important in predicting the aerodynamic characteristics. It is also seen that the effect of camber shape appears nonlinearly in all aerodynamic expressions. To design the camber shape, the camber slope is represented by a cosine Fourier series at each of several spanwise stations. The Fourier coefficients are the design variables. To design a leading-edge flap in the vortex flow (i.e., a vortex flap), the coordinates of corner points and the deflection angle are the design variables. The process of wing design is to determine the camber shape and twist distribution such that an objective function, typically the drag, is minimized, subject to various constraints.

Lan, C. E.↗

Concurrent algorithms for transient nonlinear FE analysis

A two-parameter class of time-stepping algorithms for nonlinear structural dynamics is investigated. What sets the present method apart from other concurrent algorithms is the fact that it can be used to some advantage in sequential machines as well. Thus, substantial speed-ups are obtained on a single processor as the number of subdomains is increased. An additional O(p) speed-up is obtained when p processors are utilized. The test case discussed is being repeated for a mesh comprising four times as many elements, in an effort to understand how the large scale asymptotic speed-ups are attained. A three dimensional example involving finite deformations and free body motions is also being pursued. A code optimized for concurrency in the Alliant FX8 computer is being finalized. This will provide the means for testing the performance of the algorithm in a multiprocessor environment.

Ortiz, M.↗

Operation of a H4RG-10 in the NASA Goddard Astrophyiscs Divison IR Detector Lab Testbed

Future space and ground based missions in the near infrared are planning to or will utilize the next generation of Teledyne's HxRG detectors, the HgCdTe 4K x 4K array (H4RG). The science cases of such missions will require optimal stability and noise performance. To assess the detailed performance of the H4RG, we have developed a small single detector testbed in NASA Goddard's Astrophysics Division IR detector lab. The testbed operates a H4RG array inside a large dewar using a room temperature Leach controller. The dewar will include two integrating spheres with controlled apertures using NIR LEDs as light sources as well as a calibrated photodiode to precisely measure flux. We present preliminary results of a banded H4RG-10 array on the bench. In the near future, we plan to use the test bed to investigate the specific origins of electronic noise in the test bed, persistence, and other flux dependent nonlinearities.

Mosby, Gregory↗

NASA Langley's Approach to the Sandia's Structural Dynamics Challenge Problem

The objective of this challenge is to develop a data-based probabilistic model of uncertainty to predict the behavior of subsystems (payloads) by themselves and while coupled to a primary (target) system. Although this type of analysis is routinely performed and representative of issues faced in real-world system design and integration, there are still several key technical challenges that must be addressed when analyzing uncertain interconnected systems. For example, one key technical challenge is related to the fact that there is limited data on target configurations. Moreover, it is typical to have multiple data sets from experiments conducted at the subsystem level, but often samples sizes are not sufficient to compute high confidence statistics. In this challenge problem additional constraints are placed as ground rules for the participants. One such rule is that mathematical models of the subsystem are limited to linear approximations of the nonlinear physics of the problem at hand. Also, participants are constrained to use these models and the multiple data sets to make predictions about the target system response under completely different input conditions. Our approach involved initially the screening of several different methods. Three of the ones considered are presented herein. The first one is based on the transformation of the modal data to an orthogonal space where the mean and covariance of the data are matched by the model. The other two approaches worked solutions in physical space where the uncertain parameter set is made of masses, stiffnesses and damping coefficients; one matches confidence intervals of low order moments of the statistics via optimization while the second one uses a Kernel density estimation approach. The paper will touch on all the approaches, lessons learned, validation 1 metrics and their comparison, data quantity restriction, and assumptions/limitations of each approach. Keywords: Probabilistic modeling, model validation, uncertainty quantification, kernel density

Horta, Lucas G.↗

Dual-shaped offset reflector antenna designs from solutions of the geometrical optics first-order partial differential equations

In obtaining solutions to the first-order nonlinear partial differential equations (PDEs) for synthesizing offset dual-shaped reflectors, it is found that previously observed computational problems can be avoided if the integration of the PDEs is started from an inner projected perimeter and integrated outward rather than starting from an outer projected perimeter and integrating inward. This procedure, however, introduces a new parameter, the main reflector inner perimeter radius p(o), when given a subreflector inner angle 0(o). Furthermore, a desired outer projected perimeter (e.g., a circle) is no longer guaranteed. Stability of the integration is maintained if some of the initial parameters are determined first from an approximate solution to the PDEs. A one-, two-, or three-parameter optimization algorithm can then be used to obtain a best set of parameters yielding a close fit to the desired projected outer rim. Good low cross-polarization mapping functions are also obtained. These methods are illustrated by synthesis of a high-gain offset-shaped Cassegrainian antenna and a low-noise offset-shaped Gregorian antenna.

Galindo-Israel, V.↗

Finite Element Modeling of the Bulk Magnitization of Railroad Wheels to Improve Test Conditions for Magnetoacoustic Residual Stress Measurements

The magnetoacoustic measurement technique has been used successfully for residual stress measurements in laboratory samples. However, when used to field test samples with complex geometries, such as railroad wheels, the sensitivity of the method declines dramatically. It has been suggested that the decrease in performance may be due, in part, to an insufficient or nonuniform magnetic induction in the test sample. The purpose of this paper is to optimize the test conditions by using finite element modeling to predict the distribution of the induced bulk magnetization of railroad wheels. The results suggest that it is possible to obtain a sufficiently large and uniform bulk magnetization by altering the shape of the electromagnet used in the tests. Consequently, problems associated with bulk magnetization can be overcome, and should not prohibit the magnetoacoustic technique from being used to make residual stress measurements in railroad wheels. We begin by giving a brief overview of the magnetoacoustic technique as it applies to residual stress measurements of railroad wheels. We then define the finite element model used to predict the behavior of the current test configuration along with the nonlinear constitutive relations which we obtained experimentally through measurements on materials typically used to construct both railroad wheels and electromagnets. Finally, we show that by modifying the pole of the electromagnet it is possible to obtain a significantly more uniform bulk magnetization in the region of interest.

Fulton, J. P.↗

Quasi-Static and Dynamic Analysis of Composite Panels: Challenges in Engineering-Level Commercial Software

In engineering applications, where analysis time and efficiency are essential, analytical models and commercially available finite element (FE) material models are the two mainstream methods used to analyze composite crush scenarios. A reliable numerical model addressing the crushing response of composite materials requires a nonlinear progressive damage analysis which includes appropriate failure mechanisms such as matrix cracking, fragmentation, delamination, and fiber breakage. One might be able to address all these interacting failure responses through implementation/validation of a user defined material model or by utilization of a commercially available model, the latter being the most efficient engineering approach. In this paper, the quasi-static and dynamic crushing response of composite specimens, consisting of a C-channel and a corrugated geometry manufactured with carbon fiber/epoxy matrix fabric, are modeled using the progressive composite damage model within ABAQUS. The novel part of this work, which has not been previously addressed in the literature, is the utilization of a detailed parametric study to optimize composite damage model parameters for a given quasi-static crush simulation in ABAQUS. C-channel experiments were used to calibrate the model and then refine/optimize a set of material parameters that best fit the experimental results. These identical parameters were then used to simulate the crush response of the corrugated channel specimens, thereby making it a pure prediction and not a double recalibration. Next, the dynamic response of both panels were predicted, further confirming the predictive capability of the developed numerical model. The effect of parameter variation on the crush response is fully documented for future use in engineering applications utilizing this specific ABAQUS damage model.

Paria Naghipour↗

Predictive and Neural Predictive Control of Uncertain Systems

Accomplishments and future work are:(1) Stability analysis: the work completed includes characterization of stability of receding horizon-based MPC in the setting of LQ paradigm. The current work-in-progress includes analyzing local as well as global stability of the closed-loop system under various nonlinearities; for example, actuator nonlinearities; sensor nonlinearities, and other plant nonlinearities. Actuator nonlinearities include three major types of nonlineaxities: saturation, dead-zone, and (0, 00) sector. (2) Robustness analysis: It is shown that receding horizon parameters such as input and output horizon lengths have direct effect on the robustness of the system. (3) Code development: A matlab code has been developed which can simulate various MPC formulations. The current effort is to generalize the code to include ability to handle all plant types and all MPC types. (4) Improved predictor: It is shown that MPC design using better predictors that can minimize prediction errors. It is shown analytically and numerically that Smith predictor can provide closed-loop stability under GPC operation for plants with dead times where standard optimal predictor fails. (5) Neural network predictors: When neural network is used as predictor it can be shown that neural network predicts the plant output within some finite error bound under certain conditions. Our preliminary study shows that with proper choice of update laws and network architectures such bound can be obtained. However, much work needs to be done to obtain a similar result in general case.

Kelkar, Atul G.↗

Wavelet Approximation in Data Assimilation

Estimation of the state of the atmosphere with the Kalman filter remains a distant goal because of high computational cost of evolving the error covariance for both linear and nonlinear systems. Wavelet approximation is presented here as a possible solution that efficiently compresses both global and local covariance information. We demonstrate the compression characteristics on the the error correlation field from a global two-dimensional chemical constituent assimilation, and implement an adaptive wavelet approximation scheme on the assimilation of the one-dimensional Burger's equation. In the former problem, we show that 99%, of the error correlation can be represented by just 3% of the wavelet coefficients, with good representation of localized features. In the Burger's equation assimilation, the discrete linearized equations (tangent linear model) and analysis covariance are projected onto a wavelet basis and truncated to just 6%, of the coefficients. A nearly optimal forecast is achieved and we show that errors due to truncation of the dynamics are no greater than the errors due to covariance truncation.

Tangborn, Andrew↗

Quantifying Wrinkle Features of Thin Membrane Structures

For future micro-systems utilizing membrane based structures, quantified predictions of wrinkling behavior in terms of amplitude, angle and wavelength are needed to optimize the efficiency and integrity of such structures, as well as their associated control systems. For numerical analyses performed in the past, limitations on the accuracy of membrane distortion simulations have often been related to the assumptions made. This work demonstrates that critical assumptions include: effects of gravity, supposed initial or boundary conditions, and the type of element used to model the membrane. In this work, a 0.2 m x 02 m membrane is treated as a structural material with non-negligible bending stiffness. Finite element modeling is used to simulate wrinkling behavior due to a constant applied in-plane shear load. Membrane thickness, gravity effects, and initial imperfections with respect to flatness were varied in numerous nonlinear analysis cases. Significant findings include notable variations in wrinkle modes for thickness in the range of 50 microns to 1000 microns, which also depend on the presence of an applied gravity field. However, it is revealed that relationships between overall strain energy density and thickness for cases with differing initial conditions are independent of assumed initial conditions. In addition, analysis results indicate that the relationship between wrinkle amplitude scale (W/t) and structural scale (L/t) is independent of the nonlinear relationship between thickness and stiffness.

Jacobson, Mindy B.↗

Identification and Control of Aircrafts using Multiple Models and Adaptive Critics

We compared two possible implementations of local linear models for control: one approach is based on a self-organizing map (SOM) to cluster the dynamics followed by a set of linear models operating at each cluster. Therefore the gating function is hard (a single local model will represent the regional dynamics). This simplifies the controller design since there is a one to one mapping between controllers and local models. The second approach uses a soft gate using a probabilistic framework based on a Gaussian Mixture Model (also called a dynamic mixture of experts). In this approach several models may be active at a given time, we can expect a smaller number of models, but the controller design is more involved, with potentially better noise rejection characteristics. Our experiments showed that the SOM provides overall best performance in high SNRs, but the performance degrades faster than with the GMM for the same noise conditions. The SOM approach required about an order of magnitude more models than the GMM, so in terms of implementation cost, the GMM is preferable. The design of the SOM is straight forward, while the design of the GMM controllers, although still reasonable, is more involved and needs more care in the selection of the parameters. Either one of these locally linear approaches outperform global nonlinear controllers based on neural networks, such as the time delay neural network (TDNN). Therefore, in essence the local model approach warrants practical implementations. In order to call the attention of the control community for this design methodology we extended successfully the multiple model approach to PID controllers (still today the most widely used control scheme in the industry), and wrote a paper on this subject. The echo state network (ESN) is a recurrent neural network with the special characteristics that only the output parameters are trained. The recurrent connections are preset according to the problem domain and are fixed. In a nutshell, the states of the reservoir of recurrent processing elements implement a projection space, where the desired response is optimally projected. This architecture trades training efficiency by a large increase in the dimension of the recurrent layer. However, the power of the recurrent neural networks can be brought to bear on practical difficult problems. Our goal was to implement an adaptive critic architecture implementing Bellman s approach to optimal control. However, we could only characterize the ESN performance as a critic in value function evaluation, which is just one of the pieces of the overall adaptive critic controller. The results were very convincing, and the simplicity of the implementation was unparalleled.

Principe, Jose C.↗

Tutorial: MATLAB Implementation of a Successive Convexification Algorithm for 3 DoF Rocket Landings

The primary objective of this work is to fill in gaps and explore an alternate way of solving the 3 DoF rocket-powered landing problem presented in the 2016 AIAA paper by Szmuk, Ackimese, and Berning using successive convexification (SCvx). In the original paper, CVX, an automatic parsing package, was used to transcribe the high-level trajectory optimization problem into a format that could be read by a conic solver. The parsing step, generally computationally intensive, is hidden from the user. The use of CVX is sufficient for the generation of trajectories off-line due to the lack of runtime and flight software implementation constraints. For on-line applications, it is necessary to parse the problem for flight software implementation. References on hand-parsing powered descent guidance (PDG) problems are sparse. In this Tech Memo, the process of transcribing the 3 DoF PDG problem into the format required by MATLAB’s built-in second-order cone solver, coneprog.m, is presented in detail. Due to the abridged 3 DoF dynamics and the relatively simple nonlinearities, this reference is the natural starting point for anyone interested in grasping the concepts behind SCvx pertaining to PDG and the parsing step. Simulation results shown in this report were independently created by solving the problem using coneprog.m. The intent of this memo is to serve as a supplemental material to the original paper by breaking down the concept behind successive convexification and shed light into the parsing process. Readers are encouraged to first familiarize themselves with the material laid out in the original reference.

Alex Hayes↗

Adaptive optics and optical structures; Proceedings of the Meeting, European Congress on Optics, 3rd, The Hague, Netherlands, Mar. 12-14, 1990

The present conference on adaptive optics (AO) and optical structures addresses AO systems and controls, AO components, nonlinear optics applications to AO, astronomical applications of AO, large telescopes and optical alignment, as well as the wavefront control experiment for the use of AO in beam propagation. Specific references are made to applications of electromagnetic theory in optics, theoretical studies of system performance and design parameters, Hartmann-Shack wavefront sensing, the use of ray-based techniques in cophasing segmented mirrors, the use of a phase-conjugating mirror for real-time phase visualization, and the absolute instability of oppositely directed waves with respect to a high-reflectivity phase-conjugate mirror. Also addressed are automatic control systems, precision segmented reflectors, AO system creation, the VLT's 8.2-m primary mirrors, an optical 12-m telescope, alignment optimization via the Talbot effect, and a combination of interferometry and ray-tracing analysis.

Tyson, Robert K.↗