Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “attack modeling”

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 361 records · Page 20
All sourcesNASA NTRSDOE OSTI
NASA NTRS1989

Techniques for extreme attitude suspension of a wind tunnel model in a magnetic suspension and balance system

Although small scale magnetic suspension and balance systems (MSBSs) for wind tunnel use have been in existence for many years, they have not found general application in the production testing of flight vehicles. One reason for this is thought to lie in the relatively limited range of attitudes over which a wind tunnel model may be suspended. Modifications to a small MSBS to permit the suspension and control of axisymmetric models over angles of attack from less than zero to over ninety degrees are reported. Previous work has shown that existing arrangement of ten electromagnets was unable to generate one of the force components needed for control at extreme attitudes. Examination of possible solutions resulted in a simple alteration to rectify this deficiency. To generate the feedback signals to control the suspended model, an optical position sensing system using collimated laser beams and photodiode arrays was installed and tested. An analytical basis was developed for distributing the demands for force and moment needed for model stabilization amonge the electromagnets over the full attitude range. This was implemented by an MSBS control program able to continually adjust the distribution for the instantaneous incidence in accordance with prescheduled data. Results presented demonstrate rotations of models from zero to ninety degrees at rates up to ninety degrees per second, with pitching rates rising to several hundred degrees per second in response to step-change demands. A study of a design for a large MSBS suggests that such a system could be given the capability to control a model in six degrees of freedom over an unlimited angle of attack range.

Parker, David Huw↗
NASA NTRS1993

Unsteady aerodynamic models for maneuvering aircraft

Forced oscillation tests over a large angle-of-attack range for an F-18 model are conducted in the NASA Langley 12-foot low-speed tunnel. The resulting dynamic longitudinal data are analyzed with an unsteady aerodynamic modeling method based on Fourier functional analysis and the indicial formulation. The method is extensively examined and improved to automate the calculation of model coefficients, and to evaluate more accurately the indicial integral. The results indicate that the general model equation obtained from harmonic test data in a range of reduced frequency is capable of accurately modeling the nonlinear responses with large hysteresis effect, except in the region where a delayed flow reattachment occurs at low angles of attack in down strokes. The indicial formulation is used to calculate the response to harmonic motion, harmonic ramp motion, constant-rate pitching motion and smaller-amplitude harmonic motion. The results show that more accurate results can be obtained when the motion starts from a low angle of attack where hysteresis effect is not important.

Hu, Chien-Chung↗
DOE OSTI2024

Semantic Stealth: Crafting Covert Adversarial Patches for Sentiment Classifiers Using Large Language Models

Deep learning models have been shown to be vulnerable to adversarial attacks, in which perturbations to their inputs cause the model to produce incorrect predictions. As opposed to adversarial attacks in computer vision, where small changes introduced to pixel values can drastically alter a model's output while remaining imperceptible to humans, text-based attacks are difficult to conceal due to the discrete nature of tokens. Consequently, unconstrained gradient-based attacks often produce adversarial examples that lack semantic meaning, rendering them detectable through visual inspection or perplexity filters. In contrast to methods that rely on gradient-based optimization in the embedding space, we propose an approach that leverages a Large Language Model's ability to generate grammatically correct and semantically meaningful text to craft adversarial patches that seamlessly blend in with the original input text. These patches can be used to alter the behavior of a target model, such as a text classifier. Since our approach does not rely on gradient backpropagation, it only requires access to the target model's confidence scores, making it a grey-box attack. We demonstrate the feasibility of our approach using open-source LLMs, including Intel's Neural Chat, Llama2, and Mistral-Instruct, to generate adversarial patches capable of altering the predictions of a distilBERT model fine-tuned on the IMDB reviews dataset for sentiment classification.

Roa Carvajal, Maria↗
NASA NTRS1975

Dynamic-stability tests on an aircraft escape module at Mach numbers from 0.40 to 2.16

Wind-tunnel measurements of the aerodynamic damping and oscillatory stability of a model of a proposed escape module for a military aircraft have been made using a small-amplitude forced-oscillation technique in pitch and yaw at Mach numbers from 0.40 to 2.16 and in roll at Mach numbers from 0.40 to 1.20. The results in pitch indicate regions in the angle-of-attack range where the model exhibits large and rapid changes in both damping and stability with angle of attack, probably caused by vortex flow over the fins. There was no pronounced effect of change in angle of attack on damping in yaw. Except for the highest Mach number, negative damping in roll was produced at high negative angles of attack.

Davenport, E. E.↗
NASA NTRS1984

Design parameters for flow energizers

One of the most significant uses of flow energizers, which are small highly swept strakes mounted immediately above a lifting surface, is in flow control over regions where a lifting surface is joined to another body, such as a fuselage or nacelle. In the presently reported systematic wind tunnel test study of flow energizers, 14 different geometric configurations using a 75-deg sweep flow energizer were tested on a light twin-engine general aviation aircraft model. It is found that cambered flow energizers perform better than their flat counterparts. All but two of the energizer installations developed lower L/D at cruise angles of attack, lower maximum lift coefficients, and lower stall angles of attack than the baseline model.

Ward, D. T.↗
DOE OSTI2023

Making Corgis Important for Honeycomb Classification: Adversarial Attacks on Concept-based Explainability Tools

Methods for model explainability have become increasingly critical for testing the fairness and soundness of deep learning. Concept-based interpretability techniques, which use a small set of human-interpretable concept exemplars in order to measure the influence of a concept on a model's internal representation of input, are an important thread in this line of research. In this work we show that these explainability methods can suffer the same vulnerability to adversarial attacks as the models they are meant to analyze. We demonstrate this phenomenon on two well-known concept-based interpretability methods: TCAV and faceted feature visualization. We show that by leveraging the geometry of the problem and carefully perturbing the examples of the concept that is being investigated, we can radically change the output of the interpretability method. The attacks that we propose can either induce positive interpretations (polka dots are an important concept for a model when classifying zebras) or negative interpretations (stripes are not an important factor in identifying images of a zebra). Our work highlights the fact that in safety-critical applications, there is need for security around not only the machine learning pipeline but also the model interpretation process.

Brown, Davis R.↗
DOE OSTI2023

Virtual sensor supervised learning for cyber-attack neutralization

An industrial asset may have monitoring nodes that generate current monitoring node values. A dynamic, resilient estimator may split a temporal monitoring node space into normal and one or more abnormal subspaces associated with different kinds of attack vectors. According to some embodiments, a neutralization model is constructed and trained for each attack vector using supervised learning and the associated abnormal subspace. In other embodiments, a single model is created using out-of-range values for abnormal monitoring nodes. Responsive to an indication of a particular abnormal monitoring node or nodes, the system may automatically invoke the appropriate neutralization model to determine estimated values of the particular abnormal monitoring node or nodes (e.g., by selecting the correct model or using out-of-range values). The series of current monitoring node values from the abnormal monitoring node or nodes may then be replaced with the estimated values.

97 MATHEMATICS AND COMPUTING↗
NASA NTRS1973

Effect of nozzle lateral spacing, engine interfairing shape, and angle of attack on the performance of a twin-jet afterbody model with cone plug nozzles

Twin-jet afterbody models were investigated by using two balances to measure separately the thrust minus total axial force and the afterbody drag at Mach numbers from 0 to 1.3. Angle of attack was varied from minus 2 deg to 8.5 deg. Translating shroud cone plug nozzles were tested at dry-power and maximum-afterburning-power settings with a high-pressure air system used to provide jet total-pressure ratios up to 9.0. Two nozzle lateral spacings were studied by using afterbodies with several interfairing shapes. The close- and wide-spaced afterbodies had identical cross-sectional area distributions when similar interfairings were installed on each. The results show that the highest overall performance was obtained with the close-spaced afterbody and basic interfairings. Increasing angle of attack decreased performance for all configurations and conditions investigated.

Berrier, B. L.↗
NASA NTRS1989

High-alpha aerodynamic characteristics of crescent and elliptic wings

Static longitudinal and lateral-directional forces and moments were measured for elliptic- and crescent-wing models at high angles of attack in the NASA Langley 14-by-22-ft Subsonic Tunnel. The forces and moments were obtained for an angle-of-attack range including stall and post-stall conditions at a Reynolds number based on the average wing chord of about 1.8 million. Flow-visualization photographs using a mixture of oil and titanium-dioxide were also taken for several incidence angles. The force and moment data and the flow-visualization results indicated that the crescent wing model with its highly swept tips produced much better high-angle-of-attack aerodynamic characteristics than the elliptic model. Leading-edge separation-induced vortex flow over the highly swept tips of the crescent wing is thought to produce this improved behavior at high angles of attack. The unique planform design could result in safer and more efficient low-speed airplanes.

Van Dam, C. P.↗
NASA NTRS2014

Hybrid Wing Body Model Identification Using Forced-Oscillation Water Tunnel Data

Static and dynamic testing of the NASA 0.7 percent scale Hybrid Wing Body (HWB) configuration was conducted in the Rolling Hills Research Corporation water tunnel to investigate aerodynamic behavior over a large range of angle-of-attack and to develop models that can predict aircraft response in nonlinear unsteady flight regimes. This paper reports primarily on the longitudinal axis results. Flow visualization tests were also performed. These tests provide additional static data and new dynamic data that complement tests conducted at NASA Langley 14- by 22-Foot Subsonic Tunnel. HWB was developed to support the NASA Environmentally Responsible Aviation Project goals of lower noise, emissions, and fuel burn. This study also supports the NASA Aviation Safety Program efforts to model and control advanced transport configurations in loss-of-control conditions.

Murphy, Patrick C.↗
NASA NTRS1982

Determination of airplane aerodynamic parameters from flight data at high angles of attack

The problem of determining airplane aerodynamic model equations and estimating the associated parameters from flight data taken at high angles of attack is addressed. Two representations of the aerodynamic function based on the polynomial and spline representations are given. Then the technique of building an adequate model using a stepwise regression is presented with examples demonstrating the construction of the model and various approaches to model verification.

Klein, V.↗
NASA NTRS1981

Estimation of dynamic stability parameters from drop model flight tests

A recent NASA application of a remotely-piloted drop model to studies of the high angle-of-attack and spinning characteristics of a fighter configuration has provided an opportunity to evaluate and develop parameter estimation methods for the complex aerodynamic environment associated with high angles of attack. The paper discusses the overall drop model operation including descriptions of the model, instrumentation, launch and recovery operations, piloting concept, and parameter identification methods used. Static and dynamic stability derivatives were obtained for an angle-of-attack range from -20 deg to 53 deg. The results of the study indicated that the variations of the estimates with angle of attack were consistent for most of the static derivatives, and the effects of configuration modifications to the model (such as nose strakes) were apparent in the static derivative estimates. The dynamic derivatives exhibited greater uncertainty levels than the static derivatives, possibly due to nonlinear aerodynamics, model response characteristics, or additional derivatives.

Chambers, J. R.↗
NASA NTRS1983

Pressure distributions obtained on a 0.04-scale and 0.02-scale model of the Space Shuttle Orbiter's forward fuselage in the Langley 20-inch Mach 6 air tunnel

Results from pressure distribution tests on 0.04-scale and 0.02-scale models of the forward fuselage of the Space Shuttle Orbier are presented without analysis. The tests were completed in the Langley 20-Inch Mach 6 Tunnel. The 0.04-scale model was tested at angles of attack from 0 to 35 and angles of sideslip from 0 to -4. The 0.02-scale model was tested at angles of attack from -10 to 45 and angles of sideslip from 0 to -4. The tests were conducted in support of the development of the Shuttle Entry Air Data System (SEADS). In addition to modeling the 20 SEADS pressure orifices, the wind-tunnel to models were also instrumented with orifices to match Development Flight Instrumentation (DFI) port locations currently existing on the Space Shuttle Orbiter Columbia (OV-102). This DFI simulation had provided a means for comparisons between reentry flight pressure data and wind-tunnel data.

Bradley, P. F.↗
DOE OSTI2022

Adaptive Hierarchical Cyber Attack Detection and Localization in Active Distribution Systems

Development of a cyber security strategy for the active distribution systems is challenging due to the inclusion of distributed renewable energy generations. Here this paper proposes an adaptive hierarchical cyber attack detection and localization framework for distributed active distribution systems via analyzing electrical waveforms. Cyber attack detection is based on a sequential deep learning model, via which even minor cyber attacks can be identified. The two-stage cyber attack localization algorithm first estimates the cyber attack sub-region, and then localize the specified cyber attack within the estimated subregion. We propose a modified spectral clustering-based network partitioning method for the hierarchical cyber attack ‘coarse’ localization. Next, to further narrow down the cyber attack location, a normalized impact score based on waveform statistical metrics is proposed to obtain a ‘fine’ cyber attack location by characterizing different waveform properties. Finally, compared with classical and state-of-art methods, a comprehensive quantitative evaluation with two case studies shows promising estimation results of the proposed framework.

42 ENGINEERING↗
NASA NTRS1981

Noise of the SR-3 propeller model at 2 deg and 4 deg angle of attack

The noise effect of operating supersonic tip speed propellers at angle of attack with respect to the incoming flow was determined. Increases in the maximum blade passage noise were observed for the propeller operating at angle of attack. The noise increase was not symmetrical with one wall of the wind tunnel having significantly more noise increase than the other wall. This was apparently the result of the rotational direction of the propeller. The lack of symmetry of the noise at angle of attack to the use of oppositely rotating propellers on opposite sides of an airplane fuselage as a way of minimizing the noise due to operation at angle of attack.

Dittmar, J. H.↗
NASA NTRS1991

An experimental study of the effects of aft blowing on a 3.0 caliber tangent ogive body at high angles of attack

An experimental study of the effects of aft blowing on the forebody vortex asymmetry over a 3.0 caliber tangent ogive body at high angles of attack was conducted. The tip of the ogive body was equipped with a single blowing nozzle whose position could be adjusted. The tests were conducted in a subsonic wind tunnel at laminar flow conditions. The effects of model roll, angle of attack, blowing coefficient, and blowing nozzle axial position were independently studied. Surface pressure measurements and flow visualization results were obtained. Aft blowing was observed to alleviate the degree of vortex asymmetry at all angles of attack. The blowing was found to be more effective at the higher angles of attack. However, proportional control of the degree of vortex asymmetry was not observed, because the initial flowfield was highly asymmetric.

Gittner, Nathan M.↗