Browse Topic: Statistical analysis

Items (546)
This paper deals with the uncertainty estimation of identified frequency and damping trends of whirl flutter modes, obtained by applying system identification methods on experimental data. In particular, two different identification approaches are considered, namely the free-decay analysis by using Matrix Pencil algorithm and the Data-Driven Stochastic Subspace Identification method (SSI), applied to system response to stochastic input. The two approaches lead to as many uncertainty estimation methodologies, both leveraging the bootstrapping statistical process. A full validation procedure is then set up to assess the accuracy of such methods in correctly quantifying the uncertainty of the estimated statistics. To do so, a wing-rotor state-space linear numerical model is used to simulate system response to both dwell and stochastic inputs. The state space numerical system aims to replicate the ATTILA wing-rotor wind-tunnel model, which falls in the framework of Clean Sky 2 European program to investigate the possible occurrence of whirl-flutter instability in tiltrotor configurations. Hence, one of the proposed methodologies is applied to ATTILA experimental data for both modal identification and uncertainty estimation, and the processed flutter trends are reported in a few conditions of particular interest due to the progressive approaching of whirl-flutter condition.
Velo, AlessandroFonte, FedericoFavale, MarcoSoal, KeithBöswald, MarcVolkmar, RobinSchwochow, Jan
This study investigates Reynolds number effects on rotor wake vortex development using a hyperbaric rotor facility capable of pressurizing air up to 100 bar. Background-oriented schlieren (BOS) and hot-wire anemometry (HWA) were applied to characterize vortex trajectories, core growth, and circumferential velocity distribution. BOS measurements revealed consistent blade-to-blade trajectory deviations and vortex pairing across all operating conditions, despite that the investigated three-bladed rotor was milled from a single piece of aluminum, ensuring precise manufacturing and a highly symmetric geometry. A statistical scheme was developed to analyze the radial structure of fluctuating tip vortices, which traverse the pointwise fiber-film sensor in a fixed position. With increasing vortex Reynolds number, the tip vortices are more compact with a reduction in core growth. The circulation in the vortices grows with the vortex radial coordinate, and converges at a radial position basically independent of the vortex core size. Observed asymmetries in young vortices at low Reynolds numbers indicate enhanced roll-up dynamics. The results demonstrate the facility’s ability to isolate Reynolds number effects in rotor wake dynamics.
Bartzsch, Hauke T.Wolf, C. ChristianGalli, EricaRaffel, MarkusBraune, MarcLöhr, Markus
Bench-level tribological experiments were utilized to evaluate material, coating, and lubricant formulation effects on the loss-of-lubricant survivability of tapered roller end and cone rib contacts. Cone rib and roller end contacts were simulated using a single rotating roller and rotating flat disk. The applied load and rotational speeds of the roller and disk were controlled to simulate representative rotorcraft gearbox bearing operating conditions. The contacts were lubricated for an initial period before the lubricant supply was shut off, and the supply tube was then removed. Tests continued to run, without additional oil, until the measured friction force reached a predetermined cutoff value. Weibull-based statistical analysis was used to compare the loss-of-lubrication runtimes.
Hager Jr., CarlCarl, MatthewMurtiff, Cole
Launch, recovery, and deck handling operational performance on smaller ship platforms like Corvettes, Frigates and Destroyers are qualified as the most challenging tasks in the UAS ship-deployment of a VTOL Uncrewed Air System (UAS). One of the main hurdles is the random nature of seaway-created deck motions coupled with ship structure disturbed air wake patterns. The MoD has supported a range of work aimed at bringing Quiescent Period Prediction (QPP) technology to fruition. QPP firstly requires Wave Profiling RADAR to measure the sea wave system out to approximately 2km in the region around a vessel. Secondly these measurements are employed in a wave propagation model to predict the actual wave forces acting on a vessel. Using the wave predictions as inputs to a vessel model makes possible to predict the actual (deterministic as opposed to statistical) motions of a vessel. Wave systems naturally alternate groups of large waves with smaller waves, this property, combined with the predictive ability, allows to identify the quietest (most quiescent) periods in which to conduct wave limited naval operations. Naval mission planners in the Royal Navy, and elsewhere in the World, appreciate the need to maintain rapid, but safe, deck tempo. The fundamental concept is to measure remote sea surface profiles to predict the future wave forces acting upon a vessel. The objective is to expand ship operating deck limits to approximately Sea State 6+. The deck definitions generally empirically measured by using standard rating scales, are replaced by instrumented devices reporting the status of the deck prior to touch-down. In this paper, a thorough discussion describing the QPP deck measuring devices designed to replace piloted cueing is provided. Theory, previous simulation studies and current at-sea testing along with data results, are also discussed. To conclude, the interface of the deck measuring device into the next version of the UK UAS system, is provided. The results of the RADAR trial indicated that the RADAR data was reliable, with the RADAR images matching the physical map. The two-dimensional surface plot showed both the RADAR blocking fence along with an additional target. An additional observation concerning the operation over the deck whilst the ship is experiencing a quiescent ship motion period. The coupled secondary effect documents minimized air wake confusion. This is owing to fewer ship structure excursions into and out of the air flow. To better define deck airflow around the ship the integration of a Doppler LIDAR instrumented federate is proposed. This is meant to predict the future vessel air wake and look for quiescent periods in this paralleling the vessel motion QPP technique.
Ferrier, BernardChristmas, JacquelineBelmont, MichaelWatson, RN, Commander Brad
In the context of Rotorcraft Pilot Couplings, the biomechanics of the pilot body play a fundamental role in determining the stability of the pilot-vehicle closed loop system. The response of the pilot body is, in turn, inherently stochastic, being a function of pilot biometrics and muscular activation. Coupling the statistical distribution of pilot biomechanical behavior determined in specialized experimental campaign with linear models of the helicopter heave dynamics, an uncertainty propagation procedure is developed, with the aim of estimating the statistical distribution of the stability margins of the closed loop pilot-vehicle system. Results obtained varying the collective lever characteristics, as well as the helicopter model parameters, align well with results obtained previously in deterministic settings. However, the new scheme allows to define quantitative robustness indices.
Zanoni, AndreaMasarati, PierangeloColombo, FrancescaZilletti, MicheleMarchesoli, DavideTalamo, CarmenCassoni, Gianni
A framework for statistical comparison between analytical and experimental structural loads has been developed and applied to approximately 100 counters within the UH-60A Airloads test program. This framework relies on established structural load variability methods with novel applications to analytical structural load development maneuver time transient analysis. The analytical results are from Rotorcraft Comprehensive Analysis System (RCAS) spanwise structural loads developed with hub load and spanwise aerodynamic loads prescribed. RCAS consistently under predicted the Coefficient of Variation (COV) associated with spanwise Normal bending when compared to flight data. This resulted in significant scale factors required to achieve a μ+2σ reliability for structural load development. RCAS results for Edgewise bending scale factors proved slightly better than Normal bending in addition to more even over / under prediction of COV when compared to flight data.
Viall, WesleyShotorban, BabakFahimi, Farbod
This study presents a statistical approach for detecting and estimating damage to multicopter propellers through a comprehensive probabilistic model. The methodology is derived from model-based analysis and applied within the time series statistical techniques. This research accounts for uncertainties in the estimation process and offers confidence intervals for assessing the extent of damage to the propellers. The framework employs functionally pooled (FP) models characterized by parameters that depend on damage sizes, proper statistical estimation, and decision-making schemes. The validation and assessment are assessed via a hexacopter flying in circles with a constant velocity and altitude under turbulence. The damage size ranges from healthy to 10 mm. The method achieves fast damage detection and precise magnitude estimation based on a segment of a single measured signal obtained from aircraft sensors during flight.
Huang, ShinanKopsaftopoulos, FotisVining, CassandraZhou, PeiyuanZhu, Jingxi
When the target value of functional geometrical specification is too tight, its cascade of tolerances is at the feasibility limit of production. In this case, the geometrical Tolerancing method loses its benefits and generates an excessive level of non-Conformity which induces additional costs that are not acceptable. The aim of this paper is first to introduce the background concerning chain of dimension method and tolerances capabilities based on test specimen results. Secondly, demonstrate ability to apply statistical calculation. Thirdly extend conventional chain of dimension in one dimension to multi-holes system installation. And, then analyze potential effect by stress evaluation. And confirm the demonstration of improvement on Tolerancing installation calculations, by onboarding all stakeholder (design, manufacturing, stress) early in design phase (interfaces maturation) and by analyzing more in detail installations constraints. This method should be applied first on "non-critical" junction, because it needs to be further matured and so it is not yet mature enough for primary structure and associated quality checks. In conclusion, as a result, it is possible to increase tolerance specification of parts and manage risks of non-assembly. In conclusion, tolerances for holes localization could be approximatively multiplied by two compared to basic calculation method.
Gatti, Jean-LoupDayan, DavidAnthonioz, HugoFruitet, Pierre
Rotor blade optimization presents a multifaceted challenge as traditional design methodologies rely on computationally exhaustive high-fidelity computational fluid dynamics (CFD). Conversely, low-fidelity techniques such as potential flow based codes are inaccurate, especially in the regions of flow separation. This paper proposes leveraging artificial neural networks (ANNs) to predict the performance polar of a given airfoil geometry, and to facilitate the inverse design of airfoil, a modified form of ANNs (known as Tandem Neural Networks (T-NNs)) is implemented. The airfoil inverse design is a multi-point optimization problem (at multiple angles of attack) and therefore, the T-NNs are trained on the vectors of performance polar instead of individual angles of attack. The paper also delves into a comprehensive analysis of data wrangling, airfoil parametrization and design of experiments to cover a wide range of rotorcraft airfoils. A novel way of including practical design constraints for airfoil geometry is also included. Finally, this work demonstrates the application of the proposed methodology for airfoil inverse design, statistical analysis for generating a family of airfoils and optimization of HART-II rotor using T-NNs and Genetic Algorithm (GA).
Anand, ApurvaBaeder, James DMarepally, Koushik
ABSTRACT
Geyer, WilliamGordon,  BarbaraMattei,  ChristopherRobinson,  Dwight
A robust framework for fault detection and identification of rotor degradation in multicopters while effectively rejecting the effects of gusts is introduced. The rotor fault detection and identification methods employed in this study are based on excitation-response signals of the aircraft under ambient turbulence to distinguish between an aircraft response to gusts and rotor faults. A concise overview of the development of statistical time series model for healthy aircraft using the aircraft attitudes as the output and controller commands as the input is presented. This model is utilized to extract quality features for training a simple neural network to perform effective online rotor fault detection and identification in a hexacopter exceptional speed of making a decision and accuracy of fault classification. It is shown that using a statistical time series model assisted neural network employed for online monitoring is capable of rejecting gusts, sensitive to even 20% rotor degradation and achieves fault detection and identification in less than 2 s after the fault with an accuracy over 99%.
Dutta, AirinGandhi, FarhanKopsaftopoulos, FotisMcKay, Michael
This work introduces the use of "global" stochastic models to detect and identify rotor failures in multicopters under different operating conditions, turbulence, and uncertainty. The identification of an extended class of time-series models known as Vector-dependent Functionally Pooled AutoRegressive models, which are characterized by parameters that depend on both forward velocity and gross weight, using scalar or vector aircraft response signals under white noise excitation has been described. A concise overview of the residual based statistical decision making schemes for fault detection and identification of rotor failures is provided. The scalar and vector statistical models, along with residual variance and residual uncorrelatedness methods were validated and their effectiveness was assessed by a proof-of-concept application to aircraft flight for healthy and faulty states under severe turbulence and intermediate operating conditions. The results of this study demonstrate the effectiveness of all the proposed residual-based time series methods in terms of prompt rotor fault detection, although the methods based on Vector AutoRegressive models exhibit improved performance compared to their scalar counterparts with respect to their performance in identifying rotor failures in the post-failure controller compensated state.
Dutta, AirinMcKay, MichaelKopsaftopoulos, FotisGandhi, Farhan
Sikorsky has developed a specification outlining the use of three casting technologies: simulation, additive manufacturing of the mold and low pressure casting. This specification has been used in the past on new development projects with positive results, reducing lead times and number of pours to produce a useable part. When the S-92 program needed to develop a second source for a casting, they worked with Magellan Aerospace to implement the specification. The project proceeded on time with all castings able to be used. Some elements of the specification were modified to work with a legacy part design, including the use of statistical process controls to reduce variability in crucible pouring.
Woodworth, HeatherFeatheringham, Andrew
The ongoing electrification and data-intelligence trends in logistics industries enable efficient powertrain design and operation. In this work, the commercial package delivery vehicle powertrain design space is revisited with a specific combination of optimization and control techniques that promise accurate results with relatively fast computational time. The specific application that is explored here is a Class 6 pickup and delivery truck. A statistical learning approach is used to refine the search for the most optimal designs. Five hybrid powertrain architectures, namely, two-speed e-axle, three-speed and four-speed automatic transmission (AT) with electric motor (EM), direct-drive, and dual-motor options are explored, and a set of Pareto-optimal designs are found for a specific driving mission that represents the variations in a hypothetical operational scenario. The modeling and optimization processes are performed on the MATLAB™-Simulink platform. A cross-architecture performance and cost comparison is performed, which shows that two-speed e-axle is the optimal architecture for the selected application.
Anil, Vijay SankarZhao, TongZhao, MingjieVillani, ManfrediAhmed, QadeerRizzoni, Giorgio
Research on Tracking Algorithm for Forward Target-Vehicle Using Millimeter-Wave Radar2020-01-07024/14/2020
In order to solve such problems that the millimeter-wave radar is of large computation, poor robustness and low precision of the target tracking algorithm, this paper presents an algorithmic framework for millimeter-wave radar tracking of target-vehicles. The target measurement information outside the millimeter- wave radar detection range is eliminated by the data plausibility judgment method based on the millimeter-wave radar detection parameters. Target clustering is made using Manhattan distance, to eliminate clutter interference and cluster multiple target measurements into one. The data association is made by use of nearest neighbor to determine the correspondence between information received measured by the radar and the real target. The vehicle is the key detection target of the vehicle millimeter-wave radar during road driving. These target-vehicles generally have no vertical movement or small moving speed in the vertical direction, so only the movement of the target-vehicle in the XY plane needs to be considered. Since the target-vehicle motion state has the characteristics of small mobility, a constant acceleration model is established based on the millimeter-wave radar motion coordinate system to describe the motion state of the front target-vehicle. The motion state are tracked and optimized by the algorithm of improved adaptive extended Kalman filter (IAEKF), because it is difficult to determine the statistical property of its measurement noise. A differential position system is formed by installing a base station on the ground and RT3000s on the ego-vehicle and target-vehicle, respectively. Differential Position System is formed by installing Base Station on the ground and high-precision inertial navigator RT3000s and RT-XLANs on the ego-vehicle and target-vehicle, respectively. By use of the differential position system, with effective communication, the relative distance and speed information between both vehicles can be obtained in real time to verify the accuracy of the millimeter-wave radar target tracking algorithm. Results show the proposed algorithm is feasible and of high estimation accuracy.
Song, ShipingWu, JianYang, YuHe, RuiChen, XuesongLi, Xin
The Right Stuff for Aging Electronics/Intermittence/No Fault Found2019-01-18899/16/2019
For those in the avionics repair and maintenance business, the acronyms NFF (No Fault Found), NTF (No Trouble Found), and CND (Cannot Duplicate) are, unfortunately, all too familiar terms. After several decades of frustration with this illusive phenomenon, it continues to consume an enormous amount of test and diagnostic effort and is the source of considerable cost and discomfort within the multi-level avionics repair model. There are undoubtedly many causes of NFF and all of them should be addressed. The question is: Where do you start and which solution will be the most beneficial? Our particular efforts have focused on the literal or statistical analysis of NFF, recognizing that if the system’s MTBF (Mean Time Between Failure) has decreased, or if the device's NFF rate has increased with age and deterioration, a physical fault is most likely present. However, if it isn’t found during conventional testing then it probably only fails intermittently. Similarly, having an intermittent failure mode, it in all probability cannot be detected or diagnosed at testing time because of known and demonstrated limitations in the conventional measurement equipment used to perform the tests. In this paper we will outline the problem of intermittence and its testing difficulties. More importantly, we will describe the unique equipment and process which has produced overwhelming success in Intermittence / NFF resolution and MTBF extension. Our team-developed overhaul system called IFDIS 2.0 (Intermittent Fault Detection and Isolation System 2.0) incorporates all the necessary testing procedures and technological capabilities that are proving to be critical to the resolution of the chronic intermittent / NFF problem.
Knudsen, Hector I.
Uncertainty of the Ice Particles Median Mass Diameters Retrieved from the HAIC-HIWC Dataset: A Study of the Influence of the Mass Retrieval Method2019-01-19836/10/2019
In response to the ice crystal icing hazard identified twenty years ago, aviation industry, regulation authorities, and research centers joined forces into the HAIC-HIWC international collaboration launched in 2012. Two flight campaigns were conducted in the high ice water content areas of tropical mesoscale convective systems in order to characterize this environment conducive to ice crystal icing. Statistics on cloud microphysical properties, such as Ice Water Content (IWC) or Mass Median Diameter (MMD), derived from the dataset of in situ measurements are now being used to support icing certification rulemaking and anti-icing systems design (engine and air data probe) activities. This technical paper focuses on methodological aspects of the derivation of MMD. MMD are estimated from PSD and IWC using a multistep process in which the mass retrieval method is a critical step. Complementary to previous studies reporting on MMD values calculated from the HAIC-HIWC dataset, this paper deals with the uncertainty in MMD by comparing two different approaches for the retrieval of the mass-size (m-D) relationship. The analysis encompasses the data collected in the high IWC areas (IWC > 1g.m-3) sampled during the two HAIC-HIWC field campaigns. MMD series are computed using three different mass-size relationships and statistical values are compared. Overall, MMD values are in good agreement, at least for two methods although they imply quite different assumptions. On the variability in MMD values at a given temperature level, results show that MMD may vary significantly from one flight to the other, even though MMD series produced with the different mass retrieval methods follow a similar pattern. A strong temperature dependence is observed regardless the assumption on the m-D relationship, making MMD to increase by more than a factor of 2 as temperature increases from -50°C to -10°C. Finally, the influence on calculated MMD of two different definitions for particle size (Deq and Dmax) is demonstrated. Generally, MMD computed with Dmax are a few percent larger as compared to MMD calculated from Deq definition, supporting the conclusions from previous studies.
Coutris, PierreSchwarzenboeck, AlfonsLeroy, DelphineGrandin, AliceDezitter, FabienStrapp, J. Walter
Vibroacoustic Model’s Likelihood Computation Based on a Statistical Reduction of Random FRF Matrices2019-01-15936/5/2019
Improvement of vibroacoustic models prediction capabilities requires an adapted indicator to compare experimental measurements with the results of the computational model. When dealing with highly uncertain objects such as series production cars, a probabilistic approach is mandatory to be able to describe the dispersion of experimental results. Moreover, a probabilistic non-parametric model also account for modeling uncertainties and simplifications that are part of any engineering process. The proposed approach deals with Frequency Response Functions since FRFs are the common way to handle vibroacoustic models. When considering multiple input and output points configuration, FRFs are frequency dependent complex matrices. Since the probabilistic modeling is available in current vibroacoustic software, collections of random realizations of the FRF matrix can be computed from the existing FE model. The model’s likelihood naturally appears as the probability of a measured quantity to be part of its model. It is a single number that can advantageously be used as an indicator of the model‘s relevance regarding measurements. A novel complex FRF matrix statistical reduction is proposed, allowing the model’s likelihood computation. This reduction relies on the separation of statistically independent components such that the probability of the whole is the product of the probability of the components. The reduction is performed by a two stage Independent Component Analysis, first along the frequencies and second on frequency independent complex matrices. For each of the components, the joint probability density function of the complex coefficient is constructed from the various realization of the considered FRF matrix. The projection of any experimental or computed matrix on the components basis provides the complex coefficients which probabilities are known. The product of the component probabilities is the model’s likelihood. The proposed approach is applied to a mid-size vehicle body.
Gagliardini, LaurentSoize, ChristianReyes, Justin
This article introduces a probabilistic model-based programming language called AURA-Sim that enables the quantification, propagation, and analysis of arbitrary random distributions through nonlinear systems. Probabilistic programming languages are a relatively recent innovation that are intended to automate much of the low-level programming required to implement common statistical computations. Model-based programming languages are designed to simplify the specification of complex dynamical systems, having many separate components that interact over time, and are widely used for developing complex systems in numerous disciplines. This articles explains the harmonious combination of these two programming concepts into a unified programming language that enables system designers to directly solve many of the most important problems of uncertainty management for dynamical systems. The resulting language is developed as a set of C++ libraries and exposed to the user in the model-based language of SIMULINK and MATLAB. AURA-Sim allows system designers to model essentially arbitrary random processes and to propagate them through a wide variety of nonlinear dynamical systems. This capability is not currently available in any model-based programming language. The AURA-Sim library is based on generalized polynomial chaos (gPC) theory which is reviewed in the following. Traditionally, uncertainty quantification, propagation, and analysis has been conducted using Monte Carlo simulation; however, Monte Carlo simulations often incur a high computational cost, are time consuming, and slow to converge. Even after dedicating the time and computational resources to perform exhaustive Monte Carlo analysis, comprehensive coverage of the uncertainty space is not assured and reasoning over the simulation results requires additional cost. The AURA-Sim approach offers the potential to provide comprehensive coverage with a single simulation run, drastically reducing the required cost. Reasoning and calculating inferential statistics from the results does not required large data sets because the simulation signals are represented as random quantities.
Cooper, JaredKlyde, DavidDeVore, Dr.Reed, Adam
Existing Structural Health Monitoring (SHM) techniques generally depend on deterministic parameters in order to detect, localize, and quantify damage. This limits the applicability of such systems in real-life situations, where stochastic, time-varying structural response, as well as complex damage types immersed in operational/environmental uncertainties are almost always encountered. Thus, there lies a need for the proposal of statistical quantities and methods for assessing structural health. That is, a holistic probabilistic SHM framework involving damage detection, localization, and quantification, is due if such systems are to become standard on VTOL platforms. In this work, a novel probabilistic approach for active-sensing acousto-ultrasound SHM targeting damage detection and quantification is proposed based on stochastic non-parametric time series representations. Statistical signal processing techniques are used to formulate statistical hypothesis tests, based on which a decision can be made to whether a component is healthy or damaged within pre-defined confidence bounds. The methods presented herein can also be used for damage quantification. The proposed framework is first applied to a notched Aluminum coupon with different damage sizes within an active-sensing, local "hot-spot" monitoring framework. After that, experimental data collected over a stiffened Aluminum panel, representing a sub-scale fuselage component, is analyzed using the probabilistic framework for validation of the proposed methods on more real-life structures. Results show the advantage of the proposed techniques in citing confidence to the decision-making process when compared with state-of-the-art damage indicators. In addition, insights into damage localization within a probabilistic framework are also presented, which may be used as a preliminary step to damage localization
Amer, AhmadKopsaftopoulos, Fotis
Guided-wave-based acousto-ultrasound structural health monitoring (SHM) methods have attracted the interest of the SHM community as guided waves can travel long distances without significant dissipation and are capable of detecting small damage sizes of several types. However, when subject to changing environmental and operational conditions (EOC), guided-wave-based methods may give false indications of damage as they exhibit increased sensitivity to varying EOC. In order to improve the reliability and enable the large-scale applicability of these methods, and to build a robust SHM system, it is necessary to quantify the uncertainty in guided wave propagation due to changing EOC. In this paper, a rigorous investigation on the uncertainty involved in the propagation of Lamb waves due to the variation in temperature and material properties of nominally-identical structures has been performed both numerically and experimentally. A high fidelity finite element model is established to study the effect of small temperature perturbation on the S0 and A0 modes of Lamb waves and the associated uncertainty is quantified. Then experiments are performed under ambient laboratory temperature variations during an eleven day period. The experimental results have indicated that temperature variations as small as 0.5°C may result variations in the amplitude of Lamb waves and affect the damage index. Then uncertainty due to the variation in material properties has been considered by taking into account the statistical Gamma distributed dependency between Young's modulus and Poisson ratio jointly and the associated variation in the damage index is also investigated.
Ahmed, ShabbirKopsaftopoulos, Fotis
This work introduces the use of statistical time series methods to detect rotor failures in multicopters. A concise overview of the development of various time series models using scalar or vector signals, statistics, and fault detection methods is provided. The fault detection methods employed in this study are based on parametric time series representations and response-only signals of the aircraft state, as the external excitation is non-observable. The comparative assessment of the effectiveness of scalar and vector statistical models and several residual-based fault detection methods are presented in the presence of external disturbances, such as various levels of turbulence and uncertainty, and for different rotor failure scenarios. The results of this study demonstrate the effectiveness of all the proposed residual-based time series methods in terms of prompt rotor fault detection, although the methods based on Vector AutoRegressive (VAR) models exhibit improved performance compared to their scalar counterparts with respect to their robustness and effectiveness for different turbulence levels and ability to distinguish between healthy and fault compensated condition after rotor failure.
Dutta, AirinMcKay, MichaelKopsaftopoulos, FotisGandhi, Farhan
Pitching airfoil measurements are known to exhibit significant scatter at near- and post-stall angles of attack. Applying data-driven algorithms revealed the presence of bimodal distribution within the data scatter, suggesting that the statistical mean and standard deviation often used to represent cycle-to-cycle variations are incorrect. Considering the historical significance of dynamic stall measurements, a thorough assessment was undertaken to ascertain that the observed furcation in the data is not a result of facility or post processing error. Once confirmed, cluster-averages, associated variances, and group probability were identified as the best alternative to represent groups in the data. Several existing clustering techniques were tested, however, their shortcomings led to the development of two new data-driven algorithms. A uniqueness that is common to both of the new algorithms is that the clustering process is based on the flow phenomena that contribute the most energy to the overall flow variations. By operating in the optimal basis that maximizes the variance in the measurements, separation of clusters became efficient. When applied to several test cases, the clusters revealed the causes for such grouping, such as the variations in the separation location, occurrence of LE/TE stall, presence/absence of a dynamic stall vortex (or vortices), reattachment angle, etc. In all the cases, the physical processes and their effects were obscured by the phase-average curves. Further analyses to study the effects of Mach number, reduced frequency, mean angle and amplitude of oscillation revealed trends in the group probability. Aerodynamic damping, peak values of pitching moment and lift were substantially different between the clusters, as well as with the phase-average. Considering future semi-empirical models that need to account for cycle switching from one group to another, Markov process (and chain) was studied.
Ramasamy, ManikandanHarms, TannerSanayei, ArmaunWilson, JacobMartin, PrestonNikoueeyan, PouryaNaughton, Jonathan
Naturalistic Driving Behavior Analysis under Typical Normal Cut-In Scenarios2019-01-01244/2/2019
Cut-in scenarios are common and of potential risk in China but Advanced Driver Assistant System (ADAS) doesn’t work well under such scenarios. In order to improve the acceptance of ADAS, its reactions to Cut-in scenarios should meet driver’s driving habits and expectancy. Brake is considered as an express of risk and brake tendency in normal Cut-in situations needs more investigation. Under critical Cut-in scenarios, driver tends to brake hard to eliminate collision risk when cutting in vehicle right crossing lane. However, under less critical Cut-in scenarios, namely normal Cut-in scenarios, driver brakes in some cases and takes no brake maneuver in others. The time when driver initiated to brake was defined as key time. If driver had no brake maneuver, the time when cutting-in vehicle right crossed lane was defined as key time. This paper focuses on driver’s brake tendency at key time under normal Cut-in situations. Environment factors (for example, traffic condition and road type), cutting-in vehicle type and motion factors were considered as influence factors. To comprehensively take those factors into account, cluster analysis was adopted to extract typical Cut-in scenarios from naturalistic driving database. Seven typical scenarios as well as driver’s behavior data were obtained. It is found that there exists certain linear relationship between Time headway (THW, equals to relative distance divided by subject vehicle velocity) and relative velocity. The linear correlation of brake cases is good, but linearity of no brake cases is weak. In each typical Cut-in scenario, driver's brake behavior was demonstrated by scatter plot on THW-relative velocity plane. This plane was divided into several zones where driver tends to brake or not to brake. A special situation is that cutting-in vehicle surpasses subject vehicle and merges in a brutal way and under such situation driver will brake hard.
Ma, XuehanMa, ZhixiongZhu, XichanCao, JianyongYu, Feng
Behavior of Electric Scooter Operators in Naturalistic Environments2019-01-10074/2/2019
The use of electric scooters (e-scooters), which are more generally categorized as motorized scooters, has undergone explosive growth owing to “scooter share” programs in which an e-scooter is rented for a limited period of time. The near-spontaneous ubiquity of e-scooters has prompted government and scooter share companies to address issues partly motivated by concerns related to the inclusion of a large population of e-scooters into vehicular traffic. These issues are influenced by the decisions and behaviors of the scooter operators, who, despite being licensed to drive passenger vehicles, potentially have limited experience operating an e-scooter in the presence of traffic. E-scooters are in a relative unique position where they are small enough to negotiate pedestrian traffic, yet fast enough to travel on roadways. This enables an e-scooter operator to change when and where he rides, e.g., from traveling on a sidewalk to riding in a clear traffic lane in order to avoid a group of pedestrians standing at an intersection. Such changes may catch nearby motorists off-guard, thereby increasing the risk of a collision with the e-scooter. The present observational study assessed e-scooter rider behavior in west Los Angeles, a region with a robust presence of rental e-scooters. The large population, preponderance of e-scooters, and high traffic volumes provide an exemplary area to observe not just how drivers and e-scooter riders adapt to one-another’s presence, but also the increased risk of an interaction between e-scooters with other vehicles and pedestrians. Operator behavior of rented e-scooters is quantified and reviewed according to current regulations, public concerns regarding e-scooters, and behaviors present that may affect an individual’s ability to safely operate an e-scooter in the presence of traffic, including both vehicular and pedestrian.
Todd, JayKrauss, DavidZimmermann, JacquelineDunning, Amber
A Trajectory-Based Method for Scenario Analysis and Test Effort Reduction for Highly Automated Vehicle2019-01-01394/2/2019
Unlike the test of passive safety of traditional vehicles, highly automated vehicles (HAV) need more capabilities to be tested. Besides, there are more parameter combinations for the scenarios that need to be tested for each capability, resulting in a high time-consuming and costs for the autonomous vehicle tests. This paper proposes a method for scenario analysis and test effort reduction. Firstly, the trajectories of the vehicle under test (VUT) in the scenario are analyzed, and the trajectories which lead to the test mission failure are obtained. Based on the above trajectories, the threshold that lead to the test mission failure, or a combination of thresholds are analyzed. The above thresholds or a combination of thresholds values are defined as Scenario Character Parameter (SCP). The process of the analysis of the SCPs are related to the abilities of the HAV, but does not depend on the specific algorithm of the HAV. Therefore, through the above analysis of trajectories and SCPs, the ability of the scenario to measure the performance of HAVs can be quantized. After completing the analysis of scenarios that are used in HAVs evaluation, the SCPs corresponding to each scenario are obtained. The SCPs have the relationships such as overlapping or inclusive. Then, a set of scenarios with minimum number but still cover all SCPs can be searched. Use this set of scenarios to replace the original combination of test scenarios, the number of scenarios that need to be tested can be reduced. The method proposed in this paper reduces the amount of tests and costs for HAVs, which will be a promote to the development of the HAV technology.
Qi, YunlongLuo, YugongLi, KeqiangKong, WeiWang, Yongsheng
Failure Prediction for Robot Reducers by Combining Two Machine Learning Methods2019-01-05084/2/2019
There are many production robots used at car manufacturing plants, and each of them is fitted with several reducers. A breakdown of one of these reducers may cause a huge loss due to the stoppage of all production lines. Therefore, condition-based maintenance is currently being used to predict failures by predetermined thresholds for average and standard deviations. However, this method can cause many false alarms or some false negatives. There are some ways of suppressing false alarms, such as detecting a change in the probability density function. However, when false alarms are suppressed using the probability density function in the operational range, some false negatives may occur, leading to a breakdown of a reducer and huge loss. A false negative is caused by overlooking an anomaly with slight changes and it is difficult to detect using only the probability density function. Therefore, we developed the Difference Signum Method (DSM) to detect an anomaly with slight changes by focusing on such changes. Although DSM reduces false negatives, it can cause many false alarms. This paper proposes a new failure prediction method using ensemble learning of the probability density function and DSM in order to reduce both false positives and false negatives. Using this new failure prediction method, the number of alerts is now fewer than four times/week, a substantial reduction from nine times/week with the previous method. The number of false negatives reached the target value of zero times/year from two times/year using the probability density function. Therefore, the performance of this new failure prediction method makes it applicable to actual production lines.
Tanaka, YasuhiroTakagi, Toru
Estimation of the Relative Roles of Belt-Wearing Rate, Crash Speed Change, and Several Occupant Variables in Frontal Impacts for Two Levels of Injury2019-01-12194/2/2019
Driver injury probabilities in real-world frontal crashes were statistically modeled to estimate the relative roles of five variables of topical interest. One variable pertained to behavior (belt-wearing rate), one pertained to crash circumstances (speed change), and three pertained to occupant demographics (sex, age, and body mass index). The attendant analysis was composed of two parts: (1) baseline statistical modeling to help recover the past, and (2) sensitivity analyses to help consider the future. In Part 1, risk functions were generated from statistical analysis of real-world data pertaining to 1998-2014 model-year light passenger cars/trucks in 11-1 o’clock, full-engagement frontal crashes documented in the National Automotive Sampling System (NASS, 1997-2014). The selected data yielded a weighted estimate of 1,269,178 crash-involved drivers. Those data were parsed for four subpopulations: two levels of belt use (properly-belted vs. unbelted) and two levels of driver injury (moderate-to-maximum, MAIS2+ vs. serious-to-maximum, MAIS3+). For each subpopulation, a baseline statistical model was generated via logistic regression, cast as a function of the studied variables. Each risk function was assessed for statistical significance (p-value for each term) and statistical associativity (Goodman-Kruskal Gamma). The four resulting risk functions had some statistical insignificance and fair fidelity, with Gammas ranging from 0.54 to 0.73. However, the risk functions demonstrated excellent fidelity for estimating aggregate injury rates (function-estimated vs. directly-estimated). They were accordingly applied in Part 2. In Part 2, sensitivity studies were conducted by (a) perturbing the studied variables in the NASS dataset to generate thousands of hypothetical NASS files, (b) applying the risk functions to estimate attendant net injury rates, and (c) relating the net injury rates to the variations. Specifically, net injury rates and mean statistics were generated for 15,552 hypothetical NASS datasets involving both belted and unbelted drivers. Those data were then normalized by the means of the baseline NASS file. Finally, power functions were developed to relate the resulting dimensionless net injury-rate data to the five dimensionless predictor variables. Those functions demonstrated excellent fidelity (R2≥0.95), and their exponents helped quantify the relative role of the five studied variables. Belt-wearing rate and speed change were determined to be the most influential, followed by age, body mass index, and sex. These findings might help guide engineers and regulators.
Laituri, TonyHenry, ScottLi, Guosong
Risk Assessment of Fuel Property Variability Using Quasi-Random Sampling/Design of Experiments Methodologies2019-01-13873/19/2019
Increases in on-board heat generation in modern military aircraft have led to a reliance on thermal management techniques using fuel as a primary heat sink. However, recent studies have found that fuel properties, such as specific heat, can vary greatly between batches, affecting the amount of heat delivered to the fuel. With modern aircraft systems utilizing the majority of available heat sink capacity, an improved understanding of the effects of fuel property variability on overall system response is important. One way to determine whether property variability inside a thermal system causes failure is to perform uncertainty analyses on fuel thermophysical properties and compare results to a risk assessment metric. A sensitivity analysis can be performed on any properties that cause inherent system variability to determine which properties contribute the most significant impact. For the current study, a quasi-random sampling based uncertainty analysis was combined with a surrogate model based sensitivity analysis. Combining sample based and surrogate-based methodologies provided statistical information from the sampling based approach and sensitivity information from the design of experiments from one test series. Using the two methods simultaneously combined the advantages of both methods, while reducing the number of trials required for a statistically significant sample. The methodologies were applied to fuel property variability by sampling deviating thermophysical properties and analyzing the system impact for a sample mission profile. The fuel property variabilities examined were fuel density, specific heat, viscosity, and thermal conductivity. The analysis utilized the architecture’s feed tank temperature as the failure metric and the probability of system failure was determined using a confidence interval. The determined point of failure was analyzed using a sensitivity analysis to determine dominant fuel properties. The sample sizes were compared using computation time and sample size effect on statistical variation to determine the optimal setting for risk analysis.
McCarthy, KevinJackson, Galen R.
A Method for System Identification in the Presence of Unknown Harmonic Excitations Based on Operational Modal Analysis2019-01-50071/23/2019
Operational modal analysis techniques classically have been developed based on the assumption that the input to the system is a stationary white noise. While, in many practical cases, the systems are excited by combination of white noise and colored noises (harmonic excitations). Consequently, in conditions where non-white noises are present, the existing OMA methods cannot completely distinguish between the system poles and the induced poles due to colored noises. In order to overcome this weakness of OMA methods, some researches have been conducted in the field. In this paper, a new method is proposed for identifying the modal parameters of the system under the unknown colored noises, based on the Power Spectral Density Transmissibility (PSDT) function. In this work, the proposed methodology is established upon applying the auxiliary force, which can re-excite the system under operational conditions. In order to identify the modal parameters through the PSDT function, an appropriate parametric identification method such as the Poly-reference Least Squares Complex Frequency-domain method (PLSCF), or Poly-Max method, is utilized. Thus, modal parameters of the system poles are identified using a Stabilization Diagram (SD) by overestimating the system model order. To illustrate the efficiency of the proposed methodology, a four DOF vibrational system is considered as a case study through a computer simulation, and the obtained results are compared and discussed for verification.
Khodaygan, S.
Characterization of GDI PM during Vehicle Start-Stop Operation2019-01-00501/15/2019
As the fuel economy regulations increase in stringency, many manufacturers are implementing start-stop operation to enhance vehicle fuel economy. During start-stop operation, the engine shuts off when the vehicle is stationary for more than a few seconds. When the brake is released by the driver, the engine restarts. Depending on traffic conditions, start-stop operation can result in fuel savings from a few percent to close to 10%. Gasoline direct injection (GDI) engines are also increasingly available on light-duty vehicles. While GDI engines offer fuel economy advantages over port fuel injected (PFI) engines, they also tend to have higher PM emissions, particularly during start-up transients. Thus, there is interest in evaluating the effect of start-stop operation on PM emissions. In this study, a 2.5L GDI vehicle was operated over the FTP75 drive cycle. Runs containing cold starts (FTP-75 cycle Phases 1 & 2) and multiple runs containing hot starts (FTP-75 cycle Phases 3 & 4) were performed each day. Note that the FTP-75 Phases 3 & 4 are identical to Phases 1 & 2 except that the engine is warmed up. Three fuels were evaluated: an 87 AKI gasoline (E0), a 21% splash blend of ethanol and the 87 AKI gasoline (E21), and a 12% splash blend of iso-butanol and the 87 AKI gasoline (iBu12). PM mass, transient particle number concentration and size distribution, and soot mass concentration were evaluated for both start-stop operation and no start-stop operation on each fuel. Three Phase 1 & 2 cycles and as many as 27 Phase 3 & 4 cycles were performed for each fuel-mode combination. Composite FTP mass emissions for E0 and iBu12 showed increased total PM emissions with start-stop operation, but E21 showed no difference. Statistical analysis of the effects of start-stop on PM number and soot emissions showed different trends for different fuels. For example, when E0 is used with start-stop operation, the particle number decreased but the soot mass tended to increase. The results of this study have implications for hybrid vehicle operation as well because the internal combustion engine in hybrid vehicles must stop and re-start during normal operation.
Storey, John M.Moses-DeBusk, MelanieHuff, SheanThomas, JohnEibl, MaryLi, Faustine
Evaluation of Atomization Timing and Optimal Water Content for an Emulsified Fuel Droplet2018-32-005910/30/2018
The emulsified fuel means that it is mixed fuel with water and stabilized by surfactant. The difference of boiling points between fuel and water occur the secondary atomization during heating process. The water content strongly influence on the timing of secondary atomization(1). However, the water content is determined empirically. It means that it is the doubtful of compatibility fuel and a combustor. Then the emulsified fuel is needed the engineering evaluation (not empirically) to take advantage of sure secondary atomization. This research focuses on the timing and behavior of secondary atomization with an emulsified fuel droplet and the proposal of engineering evaluation. Moreover, we propose novel test method without the suspending wire to avoid heat transfer from itself. Namely, the novel point is heating process by floating in the high temperature silicone oil. This method can reveal the atomization behavior of a fuel droplet similar to the spray combustion. The measured data are waiting time of atomization and direct photos during heating process. The waiting time is fitted by Weibull plots which is a statistical treatment of reliability engineering. The inclination of Weibull plots means the timing of secondary atomization. This is the engineering evaluation on this research. The experimental results show the optimal water content of n-Hexadecane is 23%. If the fuel droplets have different timing of the secondary atomization, it is hard to control the combustion, for example, ignition delay, rate of heat release and so on. The typical experimental results show the inclination of Weibull plots converge to a point. This is mean that even various sizes of fuel droplets occur secondary atomization in the similar timing during heating process. The optimal water content of Bio diesel fuel is wide range under 30%. We propose the statistical evaluation to determine the optimal water content for practical emulsified fuel use.
Aoki, JunichiTanaka, Junya
Feature-Based Response Classification in Nonlinear Structural Design Simulations10-02-03-00127/24/2018
An applied system design analysis approach for automated processing and classification of simulated structural responses is presented. Deterministic and nonlinear dynamics are studied under ideal loading and low noise conditions to determine fundamental system properties, how they vary and possibly interact. Using powerful computer resources, large amounts of simulated raw data can be produced in a short period of time. Efficient tools for data processing and interpretation are then needed, but existing ones often require much manual preparation and direct human judgement. Thus, there is a need to develop techniques that help to treat more virtual prototype variants and efficiently extract useful information from them. For this, time signals are evaluated by methods commonly used within structural dynamics and statistical learning. A multi-level multi-frequency stimulus function is constructed and simulated response signals are combined into frequency domain functions. These are associated with qualitative system features, such as being periodic or aperiodic, linear or nonlinear and further into subcategories of nonlinear systems, such as fundamental, sub or super harmonic and even or odd order types. Appropriate classes are then determined from selected feature metrics and rules-of-thumb criteria. To automate the classification of large data sets, a support vector machine is trained on categorised responses to determine whether a single feature, or combinations of features, applies or not. The trained classifier can then efficiently process new sets of data and pick out cases that are associated with possible vibrational problems, which subsequently can be further analysed and understood. This article describes elements of the analysis, discuss the effectiveness of evaluated feature metrics, reports practical considerations and results from two separate training study examples.
Andersson, NiclasRinaldo, Raoul
ABSTRACT An analytical approach to identify manufacturing span reduction opportunities is presented. Span and variation reduction are key elements in operations management that lead directly to improved inventory turns. Generic solutions (such as implementing one type of corrective action across all gates of a production line) is a conventional approach that can lead to poor long term decisions. Statistical measures, such as coefficient of variation, standard deviation, and average span, are calculated and compared against target values to identify span and variation reduction opportunities. The focus is on gated areas of the production flow, where improved targets need to be achieved. This approach identifies Lean, Variation Reduction and Design Improvement opportunities and was used to implement corrective actions that resulted in 29.3% reduction in the average span for a production Bell main rotor blade in 2017.
Tiwari, ChandrashekharSkinner, Gregg
ABSTRACT This article describes an approach to learning gearbox operating conditions, defined by torque, rotational speed, and power, from acceleration data. Learning operating conditions paves the way to learning gearbox state-of-health because health indicators have to be normalized with respect to operating conditions to avoid false alarms. Moreover, because operational data is vastly larger than data associated with faults, representation learning is easier (and often only possible) from the operational data. The article compares two different solutions, one based on a multi-layer perceptron and the other on a recurrent network using the first four statistical moments as input features. The decision process, including heuristics and domain knowledge, used for selection of the network topology is described in detail. Models were found most effective in estimating the mechanical power transmitted through the gearbox and provided improvements over the second moment (RMS) alone.
Nenadic, NenadHood, AdrianThurston, Michael
ABSTRACT Major challenges of high quality requirements are associated with internal statistical variability, i.e., too severe gap between "as-designed" vs. "as-built" composite micro-structures. Therefore, the objective of this work is development of general physics-based methodology to correlate structural performance of RCS with inevitable variability of internal designs and demonstrate it on examples of representative composite sub-elements. Generated results can be used to provide guidance for simplifications and relaxation of existing quality requirements with obvious cost and availability improvement. Demonstration of implementation is shown on quantification of damage initiation in laminated structures under conditions of tension and bending.
Gurvich, MarkPhan, NamLong, QuanLaBarre, BobGu, WenjiongShishkin, SergeyRahman, Anisur
Fuel Effects on PM Emissions from Different Vehicle/Engine Configurations: A Literature Review2018-01-03494/3/2018
Particulate matter (PM) emitted from gasoline combustion continues to be a subject of research and regulatory interest. This is particularly true as new technology gasoline direct injection (GDI) engines can produce significantly higher levels of PM compared to older technology port fuel injection (PFI) engines. The goal of this study was to conduct a comprehensive literature search and subsequent statistical analysis related to the effects of gasoline properties, such as aromatics, octane indices, and fuel volatility, on PM (mass and number) emissions from PFI and GDI vehicles/engines. The statistical analyses showed a range of positive and negative correlations between different fuel properties and PM mass, total particle number (PN) and solid particle number (SPN) for different engine types (GDI, PFI, and for subdivisions of these engine types), numbers of engine cylinders and driving cycles. For GDI vehicles, total aromatic content, T70, T90 (the temperature when 70% and 90% of a fuel by volume boils away during a distillation test), and distillation end point (EP) [(the highest temperature achieved during a distillation test)] were positively correlated with PM mass emissions, PN emissions, or both. Anti-Knock index (AKI), research octane number (RON), and motor octane number (MON), and T10 (the temperature when 10% of a fuel by volume boils away during a distillation test) were negatively correlated with PM mass emissions, PN emissions, or both. For PFI vehicles for the Federal Test Procedure (FTP), LA92 and US06 cycles, T50, T70, T90, AKI and MON showed more mixed results, with both positive and negative correlations, while distillation EP and RON showed a negative correlation with PM mass emissions. Many of these analyses also showed statistically significant interactions, which indicates that the magnitude and direction of the regression coefficient (slope) estimated between the fuel property and PM emissions component varied as of function of at least one of the categorical variables (i.e., vehicle engine technology or model year, number of cylinders, and/or drive cycle). The presence of such statistical interactions demonstrates the underlying complexity in the data set. The details related to the interactions can provide valuable information to researchers for interpreting data sets that include combinations of different vehicle technologies. The information can also be used in the design of test programs, where a better understanding of how the effects of different fuel properties can vary as a function of different vehicle technologies and drive cycles can aid in study planning.
Karavalakis, GeorgeDurbin, Thomas D.Yang, JiachengVentura, LucianaXu, Karen
Items per page:
1 – 50 of 546