Search NASA⌕ Search

SEARCH · Search NASA

Results for “error reduction”

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 955 records · Page 53

The Smoke-Trail Method for Obtaining Detailed Measurements of the Vertical Wind Profile for Application to Missile-Dynamic-Response Problems

Detailed measurements of the vertical profile of the wind are urgently needed f o r studies of the response of rising missiles, but conventional wind-measuring systems cannot provide the accuracy and detail required over the necessary range of altitudes. A method is presented for obtaining a detailed, accurate profile over a large altitude range. The method utilizes photogrammetric measurement of successive positions of the t rail of an ascending rocket. Construction of a rocketborne smoke generator is discussed and the photography, film reading, and data-reduction procedures are described. The accuracy of the method i s examined and error equations are derived. Finally, some examples of wind profiles measured by this method are shown, and their implications for missile response are indicated.

Henry, Robert M.↗

Hydrocarbon Cold-Flow Simulation of LOX/LH2 Rocket Aerodynamics

In FY19, this project developed a test method which employed heated ethane in the physical simulation of rocket plume aerodynamics. Based on this work, EUS partnered with significant leveraged funding. A portable, self-contained test bed was also fabricated to support validation of the new method. In FY20, this project focused on the formal validation of ethane’s performance as an aerodynamic simulant as well as the development of the analytical and hardware infrastructure required to carry out that task. The test bed was activated, a series of heated-ethane tests on one or more rocket/diffuser configurations were conducted, and characterizations of the performance of the ethane simulation compared to hot-fire were performed. A series of 127 tests were conducted on 13 aerodynamic configurations to evaluate ethane’s performance as a simulant. It was found that the new methodology can reproduce rocket plume flow fields with +/- 0.5% pressure error at ~3% physical scale. The result is a cost-per-test reduction of ~99% compared to the hot-fire subscale testing typically used to obtain equivalent data. Some initial efforts have taken place, and future efforts are being planned with LaRC supersonic retro propulsion and MSFC SLS aerodynamics teams, which are interested in using the ethane system to wind-tunnel test human-rated Mars lander retrorockets and SLS Booster. Details of the prototypical experiments were published as NASA TM 2020-5009122,and have been presented in JANNAF Virtual Conference paper, "New Developments in Retro propulsion Testing for Mars Entry, Descent and Landing".

Daniel Jones↗

Reduced Dimensionality Analysis of TEMPO Ozone Profile Retrievals Using the Compact Phase Space (CPSR) Algorithm

TEMPO ozone (O 3 ) profile retrievals are expected to have fidelity in the troposphere due the sensitivities of the associated averaging kernels. However, those averaging kernels are severely rank deficiency meaning that a visual inspection of the vertical structure of the averaging kernel profile sensitivities is misleading due linear dependencies in the profile. The Compact Phase Space Retrieval (CPSR) algorithm use singular value decompositions of the averaging kernels and the ‘compressed’ retrieval solution error covariance to project the transformed averaging kernels into a space that removes the linear dependencies and accounts for the solution error uncertainties. In this oral presentation and poster, we apply the CPSR dimensional reduction analysis to TEMPO and TROPOMI O 3 profile retrievals for 13:45 UTC March 29, 2024 to study the phase space characteristics of the transformed averaging kernels as a function of latitude for North America. Our results show that TEMPO generally has more phase space vertical structure in the troposphere than TROPOMI. TEMPO has four to five dominant modes, and TROPOMI has five to six dominant modes. That means that dimensional reduction can reduce the TEMPO resource requirements by ~77% and the TROPOMI requirements by ~81%. Finally, we found that after removing linear dependences and after accounting for solution uncertainties TEMPO still has sensitivities throughout the troposphere.

TEMPO↗

Enforcing constraints for time series prediction in supervised, unsupervised and reinforcement learning

We assume that we are given a time series of data from a dynamical system and our task is to learn the flow map of the dynamical system. We present a collection of results on how to enforce constraints coming from the dynamical system in order to accelerate the training of deep neural networks to represent the flow map of the system as well as increase their predictive ability. In particular, we provide ways to enforce constraints during training for all three major modes of learning, namely supervised, unsupervised and reinforcement learning. In general, the dynamic constraints need to include terms which are analogous to memory terms in model reduction formalisms. Such memory terms act as a restoring force which corrects the errors committed by the learned flow map during prediction. For supervised learning, the constraints are added to the objective function. For the case of unsupervised learning, in particular generative adversarial networks, the constraints are introduced by augmenting the input of the discriminator. Finally, for the case of reinforcement learning and in particular actor-critic methods, the constraints are added to the reward function. In addition, for the reinforcement learning case, we present a novel approach based on homotopy of the action-value function in order to stabilize and accelerate training. We use numerical results for the Lorenz system to illustrate the various constructions.

Stinis, Panagiotis↗

Space–Time Least-Squares Petrov–Galerkin Projection for Nonlinear Model Reduction

This work proposes a space--time least-squares Petrov--Galerkin (ST-LSPG) projection method for model reduction of nonlinear dynamical systems. In contrast to typical nonlinear model-reduction methods that first apply (Petrov--)Galerkin projection in the spatial dimension and subsequently apply time integration to numerically resolve the resulting low-dimensional dynamical system, the proposed method applies projection in space and time simultaneously. To accomplish this, the method first introduces a low-dimensional space--time trial subspace, which can be obtained by computing tensor decompositions of state-snapshot data. The method then computes discrete-optimal approximations in this space--time trial subspace by minimizing the residual arising after time discretization over all space and time in a weighted $\ell^2$-norm. This norm can be defined to enable complexity reduction (i.e., hyper-reduction) in time, which leads to space--time collocation and space--time Gauss--Newton with Approximated Tensors (GNAT) variants of the ST-LSPG method. Advantages of the approach relative to typical spatial-projection-based nonlinear model reduction methods such as Galerkin projection and least-squares Petrov--Galerkin projection include a reduction of both the spatial and temporal dimensions of the dynamical system, and a priori error bounds that bound the solution error by the best space--time approximation error and whose stability constants exhibit slower growth in time. Numerical examples performed on model problems in fluid dynamics demonstrate the ability of the method to generate orders-of-magnitude computational savings relative to spatial-projection-based reduced-order models without sacrificing accuracy for a fixed spatio-temporal discretization.

97 MATHEMATICS AND COMPUTING↗

Variable-fidelity multipoint aerodynamic shape optimization with output-based adapted meshes

This work presents a method to control the discretization error in multipoint aerodynamic shape optimization using output-based adapted meshes. The meshes are adapted via adjoint-based error estimates, taking into account both the objective and constraint output errors. A multi-fidelity optimization framework is then developed by taking advantage of the variable fidelity offered by adaptive meshes. The objective functional and its sensitivity at each design point (operating condition) are first evaluated on the same initial coarse mesh, which is then subsequently adapted for each design point individually as the shape optimization proceeds. The effort to set up the optimization is minimal since the initial mesh can be fairly coarse and easy to generate. As the shape approaches the optimal design, the mesh at each design point becomes finer, in regions necessary for that particular operating condition. The multi-fidelity framework is tightly coupled with the objective error estimation to ensure the optimization accuracy at each fidelity. Computational savings arise from a reduction of the mesh size when the design is far from optimal and avoiding an exhaustive search on low-fidelity meshes. The proposed method is demonstrated on multipoint drag minimization problems of a transonic airfoil with lift and area constraints. Improved accuracy and efficiency are shown compared to traditional fixed-fidelity optimization with a fixed computational mesh.

42 ENGINEERING↗

A Neural Optimizer With Decision-Focused Learning for Optimal Energy Storage Operation

Here, this article introduces a neural optimizer-based framework for optimizing battery energy storage system (BESS) control for grid services, including demand charge and energy cost reduction. By leveraging decision-focused learning (DFL), the proposed framework ensures seamless integration and adaptation, significantly enhancing control performance. A patch time-series transformer is employed for peak load forecasting, incorporating aleatoric uncertainty quantification to account for forecasting uncertainties within the decision-making process. The framework utilizes a solver-in-the-loop approach to generate optimal BESS actions, which are then used to train the neural optimizer-based agent. By co-optimizing both BESS operational modes and output power within the NN, the system achieves improved performance and robustness. After initial training, the forecasting and control models are jointly fine-tuned to account for forecasting errors, further improving decision precision and efficiency through DFL. Case studies are performed to validate the performance of the framework using multiple real-world datasets, demonstrating superior performance in monthly peak load forecasting compared to state-of-the-art models. In addition, the results are compared against existing decision-making approaches. The results demonstrate a reduction in monthly peak forecasting error by approximately 15% across various performance measures and achieve an optimization gap for BESS operation that is about three times smaller compared to existing methods.

Kim, Hyeonjin [Pacific Northwest National Laborato↗

Mitigating the noise of DESI mocks using analytic control variates

In order to address fundamental questions related to the expansion history of the Universe and its primordial nature with the next generation of galaxy experiments, we need to model reliably large-scale structure observables such as the correlation function and the power spectrum. Cosmological N-body simulations provide a reference through which we can test our models, but their output suffers from sample variance on large scales. Fortunately, this is the regime where accurate analytic approximations exist. To reduce the variance, which is key to making optimal use of these simulations, we can leverage the accuracy and precision of such analytic descriptions using Control Variates (CV). The power of control variates stems from utilizing inexpensive but highly correlated surrogates of the statistics one wishes to measure. The stronger the correlation between the surrogate and the statistic of interest, the larger the variance reduction delivered by the method. We apply two control variate formulations to mock catalogs generated in anticipation of upcoming data from the Dark Energy Spectroscopic Instrument (DESI) to test the robustness of its analysis pipeline. Our CV-reduced measurements offer a factor of 5-10 improvement in the measurement error compared with the raw measurements. We explore the relevant properties of the galaxy samples that dictate this reduction and comment on the improvements we find on some of the derived quantities relevant to Baryon Acoustic Oscillation (BAO) analysis.

79 ASTRONOMY AND ASTROPHYSICS↗

Symmetry reduction of tensor networks in many-body theory: I. Automated symbolic evaluation of SU(2) algebra

Abstract The ongoing progress in (nuclear) many-body theory is accompanied by an ever-rising increase in complexity of the underlying formalisms used to solve the stationary Schrödinger equation. The associated working equations at play in state-of-the-art ab initio nuclear many-body methods can be analytically reduced with respect to angular-momentum, i.e. SU (2), quantum numbers whenever they are effectively employed in a symmetry-restricted context. The corresponding procedure constitutes a tedious and error-prone but yet an integral part of the implementation of those many-body frameworks. Indeed, this symmetry reduction is a key step to advance modern simulations to higher accuracy since the use of symmetry-adapted tensors can decrease the computational complexity by orders of magnitude. While attempts have been made in the past to automate the (anti-) commutation rules linked to Fermionic and Bosonic algebras at play in the derivation of the working equations, there is no systematic account to achieve the same goal for their symmetry reduction. In this work, the first version of an automated tool performing graph-theory-based angular-momentum reduction is presented. Taking the symmetry-unrestricted expressions of a generic tensor network as an input, the code provides their angular-momentum-reduced form in an error-safe way in a matter of seconds. Several state-of-the-art many-body methods serve as examples to demonstrate the generality of the approach and to highlight the potential impact on the many-body community.

Tichai, A.↗

WE-D-204-03: A Six-Year Longitudinal Evaluation of the DICOM GSDF Conformance Stability of LCD Monitors

Purpose: The calibration of monitors in radiology is critical to ensure a standardized reading environment. If left unchecked, monitors initially calibrated to follow the DICOM Grayscale Standard Display Function (GSDF) can fall out of calibration. This work presents a quantitative evaluation of the stability of a cohort of monitors with similar deployment times and clinical utilization. Methods: Fifty-four liquid crystal display (LCD) monitors (NEC L200ME) were deployed for clinical use in 2009. At that time, a subset of eight of these monitors were used to generate a look-up table (LUT) using the open-source software pacsDisplay. The software was used to load the LUT to the graphics card of the computer in order to make the monitors compliant with the GSDF. The luminance response of the monitors was evaluated twice over six years, once in 2011 and again in 2015. Results: As expected, the maximum luminance of the monitors decreased over time, with an average reduction from 2009 of 35% in 2011, and 53% in 2015. The luminance ratio (maximum luminance divided by the minimum) also decreased, with the all of the decrease occurring in the first two years (average 20%). There was an overall increase in relative error compared with the DICOM GSDF from measurement to measurement, indicating that deviation from the GSDF increases with monitor luminance reduction. Along with changes in luminance, several other issues were identified during the testing, including non-uniformities, bad pixels, and missing calibration software. Conclusion: From the initial installation of these monitors, most of the degradation occurred during the first two years, highlighting the importance of routine clinical testing of displays. Following such quality assurance, displays could be either re-calibrated or replaced depending on different thresholds. In addition, other issues not related to luminance could be identified and corrected.

60 APPLIED LIFE SCIENCES↗

Mercury's Global Topography from Radar Ranging Data

When Mercury's radius is expanded in Legendre functions to the second degree and order, the systematic error in radar ranging data is reduced substantially. Previously, data spanning an observing interval from 1966 to 1990 were used to infer an equatorial ellipticity (a - b)/a = (540 +/- 54) X 10(exp -6) and a center-of-figure minus center-of-mass offset of (640 +/- 78) m. The magnitude of this equatorial center of figure offset implies an excess crustal thickness of 12 km or less, comparable to the Moon's excess. By comparing the equatorial ellipticity with the Mariner 10 gravity field, and assuming Airy isostatic compensation, bounds on crustal thickness can be derived. Mercury's crustal thickness is in the range from 100 to 300 km. The Mercury radar ranging observing interval has been extended from 1966 to the present. In addition, improvements in data reduction techniques have resulted in a set of Mercury ranging data less affected by systematic error, in particular the biases introduced by local topographic variations. We use this new set of reduced ranging data to improve Mercury's global topography and center-of-figure minus center-of-mass offset. New results on crustal thickness are derived, and prospects for further improvement with Mercury Orbiter data are discussed.

Anderson, J. D.↗

Enhanced CO 2 Reactive Capture and Conversion Using Aminothiolate Ligand–Metal Interface

Metallic catalyst modification by organic ligands is an emerging catalyst design in enhancing the activity and selectivity of electrocatalytic carbon dioxide (CO 2 ) reactive capture and reduction to value-added fuels. However, a lack of fundamental science on how these ligand–metal interfaces interact with CO 2 and key intermediates under working conditions has resulted in a trial-and-error approach for experimental designs. With the aid of density functional theory calculations, we provided a comprehensive mechanism study of CO 2 reduction to multicarbon products over aminothiolate-coated copper (Cu) catalysts. Our results indicate that the CO 2 reduction performance was closely related to the alkyl chain length, ligand coverage, ligand configuration, and Cu facet. The aminothiolate ligand–Cu interface significantly promoted initial CO 2 activation and lowered the activation barrier of carbon–carbon coupling through the organic (nitrogen (N)) and inorganic (Cu) interfacial active sites. Experimentally, the selectivity and partial current density of the multicarbon products over aminothiolate-coated Cu increased by 1.5-fold and 2-fold, respectively, as compared to the pristine Cu at –1.16 V RHE , consistent with our theoretical findings. Furthermore, this work highlights the promising strategy of designing the ligand–metal interface for CO 2 reactive capture and conversion to multicarbon products.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

Reduction of prostate intrafraction motion using gas-release rectal balloons

Purpose: To analyze prostate intrafraction motion using both non-gas-release (NGR) and gas-release (GR) rectal balloons and to evaluate the ability of GR rectal balloons to reduce prostate intrafraction motion. Methods: Twenty-nine patients with NGR rectal balloons and 29 patients with GR balloons were randomly selected from prostate patients treated with proton therapy at University of Florida Proton Therapy Institute (Jacksonville, FL). Their pretreatment and post-treatment orthogonal radiographs were analyzed, and both pretreatment setup residual error and intrafraction-motion data were obtained. Population histograms of intrafraction motion were plotted for both types of balloons. Population planning target-volume (PTV) margins were calculated with the van Herk formula of 2.5{Sigma}+ 0.7{sigma} to account for setup residual errors and intrafraction motion errors. Results: Pretreatment and post-treatment radiographs indicated that the use of gas-release rectal balloons reduced prostate intrafraction motion along superior-inferior (SI) and anterior-posterior (AP) directions. Similar patient setup residual errors were exhibited for both types of balloons. Gas-release rectal balloons resulted in PTV margin reductions from 3.9 to 2.8 mm in the SI direction, 3.1 to 1.8 mm in the AP direction, and an increase from 1.9 to 2.1 mm in the left-right direction. Conclusions: Prostate intrafraction motion is an important uncertainty source in radiotherapy after image-guided patient setup with online corrections. Compared to non-gas-release rectal balloons, gas-release balloons can reduce prostate intrafraction motion in the SI and AP directions caused by gas buildup.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

Improving Simulations of Fine Dust Surface Concentrations over the Western United States by Optimizing the Particle Size Distribution

To improve estimates of remote contributions of dust to fine particulate matter (PM2.5) in the western United States, new dust particle size distributions (PSDs) based upon scale-invariant fragmentation theory (Kok_PSD) with constraints from in situ measurements (IMP_PSD) are implemented in a chemical transport model (GEOS-Chem). Compared to initial simulations, this leads to reductions in the mass of emitted dust particles with radii <1.8 mm by 40%-60%. Consequently, the root-mean-square error in simulated fine dust concentrations compared to springtime surface observations in the western United States is reduced by 67%-81%. The ratio of simulated fine to coarse PM mass is also improved, which is not achievable by reductions in total dust emissions. The IMP_PSD best represents the PSD of dust transported from remote sources and reduces modeled PM2.5 concentrations up to 5 mg/m3 over the western United States, which is important when considering sources contributing to nonattainment of air quality standards. Citation: Zhang, L., J. F. Kok, D. K. Henze, Q. Li, and C. Zhao (2013), Improving simulations of fine dust surface concentrations over the western United States by optimizing the particle size distribution, Geophys. Res. Lett., 40, 3270-3275, doi:10.1002/grl.50591.

distribution↗

Quantitatively accurate activity measurements with a dedicated cardiac SPECT camera: Physical phantom experiments

Purpose: Recently, there has been increased interest in dedicated cardiac single photon emission computed tomography (SPECT) scanners with pinhole collimation and improved detector technology due to their improved count sensitivity and resolution over traditional parallel-hole cameras. With traditional cameras, energy-based approaches are often used in the clinic for scatter compensation because they are fast and easily implemented. Some of the cardiac cameras use cadmium-zinc-telluride (CZT) detectors which can complicate the use of energy-based scatter correction (SC) due to the low-energy tail—an increased number of unscattered photons detected with reduced energy. Modified energy-based scatter correction methods can be implemented, but their level of accuracy is unclear. In this study, the authors validated by physical phantom experiments the quantitative accuracy and reproducibility of easily implemented correction techniques applied to {sup 99m}Tc myocardial imaging with a CZT-detector-based gamma camera with multiple heads, each with a single-pinhole collimator. Methods: Activity in the cardiac compartment of an Anthropomorphic Torso phantom (Data Spectrum Corporation) was measured through 15 {sup 99m}Tc-SPECT acquisitions. The ratio of activity concentrations in organ compartments resembled a clinical {sup 99m}Tc-sestamibi scan and was kept consistent across all experiments (1.2:1 heart to liver and 1.5:1 heart to lung). Two background activity levels were considered: no activity (cold) and an activity concentration 1/10th of the heart (hot). A plastic “lesion” was placed inside of the septal wall of the myocardial insert to simulate the presence of a region without tracer uptake and contrast in this lesion was calculated for all images. The true net activity in each compartment was measured with a dose calibrator (CRC-25R, Capintec, Inc.). A 10 min SPECT image was acquired using a dedicated cardiac camera with CZT detectors (Discovery NM530c, GE Healthcare), followed by a CT scan for attenuation correction (AC). For each experiment, separate images were created including reconstruction with no corrections (NC), with AC, with attenuation and dual-energy window (DEW) scatter correction (ACSC), with attenuation and partial volume correction (PVC) applied (ACPVC), and with attenuation, scatter, and PVC applied (ACSCPVC). The DEW SC method used was modified to account for the presence of the low-energy tail. Results: T-tests showed that the mean error in absolute activity measurement was reduced significantly for AC and ACSC compared to NC for both (hot and cold) datasets (p < 0.001) and that ACSC, ACPVC, and ACSCPVC show significant reductions in mean differences compared to AC (p ≤ 0.001) without increasing the uncertainty (p > 0.4). The effect of SC and PVC was significant in reducing errors over AC in both datasets (p < 0.001 and p < 0.01, respectively), resulting in a mean error of 5% ± 4%. Conclusions: Quantitative measurements of cardiac {sup 99m}Tc activity are achievable using attenuation and scatter corrections, with the authors’ dedicated cardiac SPECT camera. Partial volume corrections offer improvements in measurement accuracy in AC images and ACSC images with elevated background activity; however, these improvements are not significant in ACSC images with low background activity.

61 RADIATION PROTECTION AND DOSIMETRY↗

Solution methods for very highly integrated circuits.

While advances in manufacturing enable the fabrication of integrated circuits containing tens-to-hundreds of millions of devices, the time-sensitive modeling and simulation necessary to design these circuits poses a significant computational challenge. This is especially true for mixed-signal integrated circuits where detailed performance analyses are necessary for the individual analog/digital circuit components as well as the full system. When the integrated circuit has millions of devices, performing a full system simulation is practically infeasible using currently available Electrical Design Automation (EDA) tools. The principal reason for this is the time required for the nonlinear solver to compute the solutions of large linearized systems during the simulation of these circuits. The research presented in this report aims to address the computational difficulties introduced by these large linearized systems by using Model Order Reduction (MOR) to (i) generate specialized preconditioners that accelerate the computation of the linear system solution and (ii) reduce the overall dynamical system size. MOR techniques attempt to produce macromodels that capture the desired input-output behavior of larger dynamical systems and enable substantial speedups in simulation time. Several MOR techniques that have been developed under the LDRD on 'Solution Methods for Very Highly Integrated Circuits' will be presented in this report. Among those presented are techniques for linear time-invariant dynamical systems that either extend current approaches or improve the time-domain performance of the reduced model using novel error bounds and a new approach for linear time-varying dynamical systems that guarantees dimension reduction, which has not been proven before. Progress on preconditioning power grid systems using multi-grid techniques will be presented as well as a framework for delivering MOR techniques to the user community using Trilinos and the Xyce circuit simulator, both prominent world-class software tools.

99 GENERAL AND MISCELLANEOUS↗

Ten Years of VIIRS On-Orbit Geolocation Calibration and Performance

The first innovative Visible Infrared Imaging Radiometer Suite (VIIRS) sensor aboard the Suomi National Polar-orbiting Partnership (SNPP) satellite has been in operation for 10 years since its launch on 28 October 2011. The second VIIRS sensor aboard the first Join Polar Satellite System (JPSS-1) satellite has been in operation for 4 years since its launch on 18 November 2017, which became NOAA-20. Well-geolocated and radiometrically calibrated Level-1 sensor data records (SDRs) from VIIRS are crucial to numerical weather prediction (NWP) and Level-2+ environmental data record (EDR) algorithms and products. The high quality of Level-2+ EDRs is a requirement for the continuity of NASA Earth science data records (ESDRs) and climate data records (CDRs), one of the two objectives of the SNPP mission and one of the three elements in the JPSS mission objective. The other objective of the SNPP mission is risk reduction for the follow-on JPSS missions. This paper summarizes the on-orbit geolocation calibration and validation (Cal/Val) activities for both VIIRS sensors onboard SNPP and NOAA-20 in the past 10 years. These activities include nominal geolocation Cal/Val activities, risk reduction activities, and improvements for the on-orbit VIIRS sensor operations. After these activities, sub-pixel geolocation accuracy is achieved. Nadir equivalent geolocation uncertainty is generally within 75 m (1-σ), or 20% imagery band pixels, in either the along-scan or along-track direction for both SNPP and NOAA-20 VIIRS sensors. The worst 16-day measured geolocation errors (radial, 3-σ) are 280 m and 267 m, respectively, in the latest SNPP and NOAA-20 VIIRS data collections, which are better than the required accuracy of 375 m (radial, 3-σ). The risk reduction activities also improved VIIRS builds for JPSS-3 and JPSS-4 satellites, and provide lessons learned for other VIIRS-like sensor builds.

VIIRS↗

A practical discrete-adjoint method for high-fidelity compressible turbulence simulations

Methods and computing hardware advances have enabled accurate predictions of complex compressible turbulence phenomena, such as the generation of jet noise that motivates the present effort. However, limited understanding of underlying physical mechanisms restricts the utility of such predictions since they do not, by themselves, indicate a route to design improvements. Gradient-based optimization using adjoints can circumvent the flow complexity to guide designs, though this is predicated on the availability of a sufficiently accurate solution of the forward and adjoint systems. These are challenging to obtain, since both the chaotic character of the turbulence and the typical use of discretizations near their resolution limits in order to efficiently represent its smaller scales will amplify any approximation errors made in the adjoint formulation. Formulating a practical exact adjoint that avoids such errors is especially challenging if it is to be compatible with state-of-the-art simulation methods used for the turbulent flow itself. Automatic differentiation (AD) can provide code to calculate a nominally exact adjoint, but existing general-purpose AD codes are inefficient to the point of being prohibitive for large-scale turbulence simulations. Here, we analyze the compressible flow equations as discretized using the same high-order workhorse methods used for many high-fidelity compressible turbulence simulations, and formulate a practical space–time discrete-adjoint method without changing the basic discretization. A key step is the definition of a particular discrete analog of the continuous norm that defines our cost functional; our selection leads directly to an efficient Runge–Kutta-like scheme, though it would be just first-order accurate if used outside the adjoint formulation for time integration, with finite-difference spatial operators for the adjoint system. Its computational cost only modestly exceeds that of the flow equations. We confirm that its accuracy is limited by computing precision, and we demonstrate it on the aeroacoustic control of a mixing layer with a challengingly broad range of turbulence scales. For comparison, the error from a corresponding discretization of the continuous-adjoint equations is quantified to potentially explain its limited success in past efforts to control jet noise. The differences are illuminating: the continuous-adjoint is shown to suffer from exponential error growth in (reverse) time even for the best-resolved largest turbulence scales. Implications for jet noise reduction and turbulence control in general are discussed.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗