Sensors and Instrumentation, Aircraft/Aerospace, Energy Harvesting & Dynamic Environments Testing, Volume 7

In general, existing methods to develop an effective input for multiple-input/multiple-output (MIMO) control do not offer flexibility to account for limitations in experimental test setups or tailor the control to specific test objectives. The work presented in this paper introduces a method to leverage global optimization approaches to define a MIMO control input to match a data set representing field data. This contrasts with traditional MIMO input estimation methods which rely on direct inverse methods. Efficacy of the iterative optimization method depends on the objective function and optimization method used as well as the definition of the format of the input cross-power spectral density (CPSD) matrix for the optimization routine. Various objective functions are explored in this work through sampling as well as implementation within the iterative optimization process and their impact on the resulting output CPSD. Performance of iterative optimization is assessed against the traditional, direct pseudoinverse method of obtaining the input CPSD as well as the buzz method and weighted least squares (LS). Constraints can be used within the optimization process to control the magnitude and other aspects of the input CPSD, which allows for shaker limitations to be accounted for, among other considerations. Iterative optimization can provide the best input CPSD possible for a test setup while accounting for any shortcomings the setup may have, including force and voltage constraints, which is not possible with traditional methods.

As part of the National Aeronautics and Space Administration New Aviation Horizons initiative to demonstrate and validate future high-impact concepts and technologies, the X-57 Maxwell airplane – the first all-electric X-plane – was conceived to advance research in the area of electric propulsion to show the feasibility of minimizing fuel use, reducing emissions, and lowering noise during flight. Through several configuration modifications to the X-57 airplane, validation of electrical-powered flight with increasing efficiency between each modification when compared to the baseline original airplane is anticipated. In the case of the X-57 Modification II airplane, a ground vibration test was needed to identify the airplane structural modes and use them to update and validate the finite element model. To determine the airworthiness of the airplane, the updated finite element model will be utilized to investigate both classical and whirl flutters. The X-57 Modification II ground vibration test was performed by the National Aeronautics and Space Administration Armstrong Flight Research Center Flight Loads Laboratory. This paper will highlight the testing performed to acquire the modal data as well asthe results.

The purpose of mechanical environment testing is to prove that designs can withstand the loads imparted on them under operating conditions. This is dependent not only on the test article construction but also on the loads imparted through its boundary conditions. Current practices develop environment test specifications from field responses using a single degree of freedom input control with no consideration for the mild to severe deviations from the field motion caused by the laboratory boundary condition. Test specifications are considered conservative with the assumption that most of the steps taken to generate them (e.g., straight-line envelopes and adding 3 dB) result in appropriately conservative specifications. However, without an accurate quantifiable measure of conservatism, designs can be easily mis-tested yielding unnecessarily high costs. Previous work showed a modal model for components excited through base-mounted fixtures to generate specifications with much lower uncertainty and with guaranteed quantifiable conservatism. The method focused on reproducing in-service modal energy in the test configuration by controlling the 6 degree-of-freedom input motion. That work generated test specifications with enough conservatism to account for unit-to-unit variability in the damping of the test article. This paper focuses on generating conservative specifications while considering resonant frequency shifts as a parameter for unit-to-unit variability.

Response reconstruction tests are typically performed on newly designed structures to ensure they can survive their intended environment. Recent works have shown that using multiple small shakers in a multiple-input multiple-output (MIMO) setup is better at reproducing a random vibration environment everywhere on a structure than a traditional single-axis test. The authors’ prior work sought to reproduce the environment everywhere on a part while controlling only to a subset of accelerometers on a component called the transmission simulator (TS) and found that the results were quite sensitive to how well the impedance matched between the test and the environment of interest. This paper studies this further, varying the degree to which the impedance matches between test and the in-service environment, to see the effect that this has on the fidelity of the response reconstruction at both the controlled and uncontrolled locations. To obtain additional insights, the comparison is performed utilizing results from both simulated and experimental MIMO tests. The results show in a systematic way how the response reconstruction becomes more accurate at the uncontrolled locations as the impedance of the TS improves. Furthermore, the results show that one can simulate the MIMO test to predict much of what is seen in real experiments, providing a rudimentary metric for quantifying the impedance match of TS prior to performing the MIMO tests.

In recent years, the Boundary Condition Challenge has gained acceptance in the structural dynamics community. In this challenge problem, an example dynamic system known as the Box and Removable Component, or BARC, is subjected to a single point shock load. The BARC consists of a Removable Component mounted to a box-shaped fixture. The challenge problem specifies a shock load applied to the Box fixture. Here, an additional environment for the challenge problem is proposed. This new environment will be stationary random vibration due to multiple exciters on the Box fixture. In this work, the response of the BARC to this environment will be explored with mod/sim. The goal is to provide the structural dynamics community with all the pieces necessary to examine the various facets of the challenge problem in the context of random vibration and enable researchers to more easily explore problems in random vibration. A data set including input and output degrees of freedom, model modes, model frequency response functions, and input and output time histories and power spectral densities will be created and placed on the challenge problem shared site for others to download and use.

Unlike traditional vibration testing that involves driving a single axis, Multi-Input/Multi-Output (MIMO) testing has become increasingly popular due to its ability to more accurately replicate field responses and failure modes. Quantifying the mismatch between test response and field response is critical to understanding whether the field environment was adequately replicated by the vibration test. Ideally, a vibration test would replicate the field response in terms of deflection shape and magnitude and therefore also the stresses in the test article. However, a clear and concise process or metric to quantify the difference with respect to stress between the test and field environment does not exist.

This paper first considers how the Cross Power Spectral Density (CPSD) metrics are affected by part to part variability between the field and the test. Basic properties of an analytical beam model, such as damping and stiffness, are incrementally varied and the effect on the metrics is observed. A more complex model is used to study the correlation between the CPSD metrics and failure mechanisms such as stress and fatigue. A synthetic field environment is generated so that the field stresses and fatigues are known. Many imperfect MIMO tests are constructed as samples for comparison, and several CPSD metric methods are computed for each MIMO case. The calculated CPSD metrics are correlated to the stress and fatigue differences, and the metrics that best correlate to the failure criteria are identified.

Many different structures are tested in laboratory environments to replicate the operational or field environment. Structures subjected to vibration environments are typically affixed to an electrodynamic shaker table via a test fixture. The fixture with the shaker table system provides the boundary condition and dynamic impedance to the structure under test. Because the shaker system coupled with the test fixture defines the impedance to the structure under test, it is important to be able to model the shaker table system. This is a difficult task due to the complicated interface between the shaker table, oil film, bearings, and seismic mass on which the table rests. This paper uses optimization analysis with a modal projection error objective function to develop a representative shaker table model. This technique uses data to update and provide confidence in the realization of the shaker table model.

Expansion techniques are powerful tools that can take a limited measurement set and provide information on responses at unmeasured locations. Expansion techniques are used in dynamic environments specifications, full field stress measurements, model calibration, and other calculations that require response at locations not measured. However, the process of modal expansion techniques such as SEREP (System Equivalent Reduction Expansion Process) has error with the projection of the measurement set of degrees of freedom to the expanded degrees of freedom. Empirical evidence has been used in the past to qualitatively determine the error. In recent years, the modal projection error was developed to quantify the error through a projection between different domains. The modal projection error is used in this paper to demonstrate the use of the metric in quantifying the error of the expansion process and to quantify which modes of the expansion process are the most important.

In experimental modal analysis as well as in sound testing, a fully automated impact measurement is usually very important. Fully automatic in this context means an excitation point search and an adjustment of the excitation force are controlled directly in the hammer. No manual configuration steps are necessary for the user. The excitation force depends on several boundary conditions like the used impact tip, the acceleration and the mass of the hammer but also on the material properties of the excited structure. Therefore, the impact energy of the modal hammer is currently iteratively changed by the user until the target force is reached.

Existing semi-automatic impulse hammers do not support fully automatic, reproducible and highly precise single hit excitation with freely adjustable force amplitude. Also other useful functionalities are currently missing in semi-automatic impulse hammers. The internal calculation of pulse describes parameters such as the pulse height, pulse width and frequency range of the excitation or if the single hit is reached for autonomous quality assurance. Single hits are possible with position changes of the structure or the hammer in impact mode. In addition to the use of a tripod, the hammer can also be held by a person and the hammer can be used manually guided for impact operation.

Based on these requirements, a new smart impulse hammer was developed. This new and unique automatic impulse hammer is presented as part of this work.

To characterize the dynamics of an aircraft, ground vibration testing is necessary. Typically, this testing is used to obtain the aircraft’s modal characteristics, which are then used to correlate and update a finite element model, which in turn is used in an aeroelastic analysis. However, building a dynamics model may not be feasible in some situations (such as when a non-OEM third party modifies an older aircraft for a new purpose) due to a lack of detailed engineering data or due to the cost compared to the scope of the overall project. The ability to use properly scaled free-free experimental mode shapes as inputs to the aeroelastic analysis could eliminate the time and cost associated with building and correlating a finite element model.

This paper investigates the use of experimental ground vibration test data from an “iron bird” demonstration test article in aeroelastic analysis. Issues associated with the use of experimental mode shapes directly in a flutter analysis are discussed, including obtaining rigid body modes and generating mass-normalized mode shapes.

In the field of acoustics, much work has been done to extensively research the localization of perturbations in a multiple scattering medium. One of the most popular methods for perturbation localization is time-reversal, which is the process of using sensors to receive an acoustic wave and send it back toward the source in reversed time order. It has been shown that perturbations can also be localized without a priori knowledge of the scattering medium by subtracting baseline from perturbation acoustic pressure field measurements and applying time-windowing to localize the perturbation. However, the localization of perturbations using pressure field subtraction and time-windowing breaks down in environments where the acoustic path to the perturbations is not well-known due to a lack of information about either the distribution of the scatterers or geometry of the environment. In this study, both pressure field subtraction and time-reversal are used to localize hidden perturbations in a complex reverberant environment by utilizing a sparse sensor grid placed throughout the medium. This research aims to utilize distributed acoustic sensors in areas where optical camera systems may not be able to be utilized by analyzing the acoustic propagation of the time-reversed signal on the sparse sensor array (LA-UR-20-29698).

The new generation of smartphones, equipped with various sensors such as a three-axis accelerometer, have shown potential as an intelligent, low-cost, crowd-based infrastructure monitoring platform over the past few years. This paper reports the results of an experimental study on using smartphones to identify different types of pavement deteriorations using artificial neural network (ANN). In an experimental study conducted in Blacksburg, VA, 92 responses, i.e., acceleration versus time responses in z direction, were recorded using smartphone accelerometers located in a moving vehicle. These responses were collected from different types of pavement deteriorations including speedbump, pothole, alligator cracking, and intact pavement. Then, ten different features were selected using signal-processing-based statistical techniques in both time-domain and frequency-domain to distinguish between different pavement deterioration types. ANN was then used for classification. The training techniques were Patternnet, Learning Vector Quantization 1 (LVQ1), and LVQ2 algorithms and their combination, i.e., being first trained using one of these techniques and being again trained using another technique. For model evaluation, repeated hold-out, leave-one-out cross-validation, and accuracy were used, and the average errors were reported for model comparison. According to the results, Patternnet and Patternnet+ LVQ2 provided the most accurate results with 93.48% and about 90% accuracies, respectively, while LVQ1 and LVQ1+LVQ2 did not reveal acceptable results.

The Box Assembly with Removable Component (BARC) structure has been recently introduced as a challenge problem for the study of the effects of boundary conditions on vibrational testing and modal analysis. Current efforts in studying shaker input excitations on the BARC structure have focused on either varying the degrees of freedom of the test or varying the input signal. The effects that the bolted joints introduced into the BARC’s dynamical response have not been fully investigated. This study presents an investigation on the influences of test fixture connections on the dynamical responses of BARC systems for the purpose of establishing a standard fixture connection for general testing and test replication. This investigation is done by varying the distance between bolted connections to compare the effects of stiffness of fixture contribution on the dynamic response. In addition to modal analysis, experimental random vibrations are carried out to determine the dominant frequencies in the system. The experimental measurements are compared to finite element simulations in order to determine the possible variability and the reason behind that. The finite element simulations show that the wider connection geometry leads to a softening effect in natural frequencies of the system. On the other hand, the experimental measurements indicate that the effects of gap between the fixture connections have a negligible impact on the system’s dominant frequencies. This study shows the importance of accurately considering the same experimental setup as the computational modeling and vice versa.

For most dynamical systems, there is much uncertainty for boundary conditions between experiments, simulations, and service environments. The Box Assembly with Removable Component attempts to investigate these uncertainties to provide a higher degree of certainty in replicating service environment fixtures. This study investigates the previous nonuniformities in the literature between numerical simulations and experimental testing for the dynamical system and fixture interface. The common experimental setup is to attach the system directly to the fixture, allowing for heavy contact between the system and the fixture that corrupts the results. In this work, the experimental results are carried out by adding a spacer between these contact points to provide tests that better match numerical simulations and show the mismatch for past studies in the literature.

Nixus is the world’s first fly-by-wire (FBW) sailplane, with a custom 92-foot span, 53.3 aspect ratio wing. With the second-largest-aspect-ratio wing ever built for a crewed airplane, Nixus faces unique challenges for aeroelastic design, requiring detailed study and special considerations for safe operation. The use of an FBW system for the wing control surfaces allows the exploration of new strategies for automatic flap positioning, tailored aileron deflections, load alleviation, and, in the future, aeroelasticity control. This paper briefly describes some aspects of this sailplane’s fabrication and covers in detail the ground vibration test (GVT) and use of the GVT results in the flight envelope expansion campaign. The preparation and execution of the GVT are presented, including aspects related to the involvement of Cal Poly’s Aerospace Engineering undergraduate students in this process as part of an extracurricular activity. Results of the GVT, comparison with finite element analysis (FEA), and flutter predictions from the test-verified FEA model are provided. Additionally, the execution of flutter flight tests that used the FBW system to excite the wing is presented.

NASA is developing an expendable heavy-lift launch vehicle capability, the Space Launch System, to support lunar and deep space exploration. To support this capability, an updated ground infrastructure is required including modifying an existing Mobile Launcher system. The Mobile Launcher is a very large heavy beam/truss steel structure designed to support the Space Launch System during its buildup and integration in the Vehicle Assembly Building, transportation between the Vehicle Assembly Building and Launch Pad 39B by the Crawler Transporter and provides the launch platform at the launch pad. As part of the Verification and Validation of the Mobile Launcher, and Crawler Transporter, two rollouts of the Mobile Launcher transported by the Crawler Transporter, Integrated System Verification, and Validation (ISVV)-005 and ISVV-010, have been performed to demonstrate the Crawler Transporter’s ability to transport the Mobile Launcher. ISVV-005 occurred in September 2018 and ISVV-010 occurred in late June 2019. ISVV-005 and ISVV-010 also provided the opportunity to gather data that can be used to identify the Mobile Launcher on the Crawler Transporter rollout modal characteristics and refine the estimates of the Artemis I integrated launch vehicle rollout forcing functions. While the rollout environment has historically produced relatively small launch vehicle structural loads for the Saturn/Apollo and Space Shuttle programs in comparison to launch and ascent loads, these relatively small structural loads are inputs to structural fatigue analyses. The same holds true for the Space Launch System. Because the rollout forces acting on the Mobile Launcher and the Crawler Transporter are not directly measurable, Operational Modal Analysis techniques, instead of traditional Experimental Modal Analysis techniques, provide an empirical means to identify the Mobile Launcher on the Crawler Transporter rollout modal characteristics. The ISVV-010 rollout modal characteristics provide important supplemental modal information, which, along with the Mobile Launcher modal test that was performed in June 2019 immediately prior to ISVV-010 rollout, combine to reduce uncertainty in the test-correlated Mobile Launcher on the Crawler Transporter finite element model. A well-test-correlated finite element model will play a key role in the Building Block Approach the Space Launch System program has implemented as part of its certification process for the Artemis I flight and in refining the Artemis I rollout forcing functions. At the time of the ISVV-005 rollout in September 2018, the Mobile Launcher was still undergoing construction, and therefore its modal characteristics are not directly comparable to those of the Mobile Launcher during the June 2019 Mobile Launcher modal test and subsequent ISVV-010 rollout. Hence, the ISVV-005 rollout modal characteristics will not be looked at in this paper. This paper will briefly describe the Mobile Launcher and Crawler Transporter physical characteristics, ISVV-010 rollout data collection, the challenges in implementing Operational Modal Analysis techniques due in part to the Crawler Transporter harmonics, and how these challenges were overcome to obtain the ISVV-010 Mobile Launcher on the Crawler Transporter rollout modal characteristics.

This paper presents a brief overview of vibration-based structural damage detection studies that are based on machine learning (ML) in civil engineering structures. The review includes both parametric and nonparametric applications of ML accompanied with analytical and/or experimental studies. While the ML tools help the system learn from the data fed into, the computer enhances the task with the learned information without any programming on how to process the relevant data. As such, the performance level of ML-based damage identification methodologies depends on the feature extraction and classification steps, especially on the classifier choices for which the characteristic nature of the acceleration signals is recorded in a feasible way. Yet, there are several issues to be discussed about the existing ML procedures for both parametric and nonparametric applications, which are presented in this paper.

The effects of freeplay and multi-segmented nonlinearities in the pitch degree of freedom on the dynamical responses of a two-degree-of-freedom piezoaeroelastic energy harvesting system are investigated. The nonlinear governing equations of the considered piezoaeroelastic energy harvesting system are derived along with the use of the unsteady representation based on the Duhammel formulation to model the aerodynamic loads. The nonlinear piezoaeroelastic response is carried out in the presence of freeplay and multi-segmented nonlinearities before and after the linear onset of flutter. Such nonlinearities can be introduced to piezoaeroelastic energy harvesters for performance enhancement through the possible existence of sudden jumps and chaotic responses due to the grazing bifurcation. It is shown that the existence of discontinuous effects results in the possibility of harvesting energy at lower speeds than the linear onset speed of instability. Additionally, the increase of the strength of the multi-segmented nonlinearities leads to the presence of aperiodic responses with the presence of several bifurcations limiting the system’s dynamics at low pitch angles.

The use of virtual instruments in a laboratory classroom setting for basic vibration analyses is presented. The objective of the laboratory experience is to experimentally measure the natural frequencies and damping ratio of a cantilever beam. The acceleration amplitudes due to an impact loading are measured with an accelerometer and processed with a National Instruments myDAQ data acquisition system. The myDAQ system comes with a standard set of virtual instruments including a dynamic signal analyzer and an oscilloscope, both of which are used in this laboratory. These virtual instruments are used to interface and measure actual sensor data. The dynamic signal analyzer is used in real time to analyze the accelerometer data and identify the natural frequencies of the system. It has many settings that allow the user to change the frequency range and sampling rate along with various windowing options. In this manner, the students are exposed to how each of these parameters, as well as the impact type and location, influences the signal captured and the resulting natural frequencies observed. The oscilloscope is used to capture the accelerometer signal for a relatively short period of time to calculate the damping ratio of the beam. Several periods of oscillations are captured to then find the logarithmic decrement and the corresponding damping ratio. The use of virtual instruments has been found to be a cost-effective means to expose students to basic vibration analysis without the expense and complexities of multiple pieces of hardware for multiple lab stations. The system also provides an effective tool for many other laboratory exercises with the use of the other virtual instruments such as the digital multimeter, function generator, and power supplies.

The dynamic behavior of piezoelectric energy harvesters has been widely studied in the last decade. Different deterministic modeling techniques and simplifications have been adopted to describe their electromechanical coupling effect in order to increase the accuracy on the output power estimation. Although it is a common practice to use deterministic models to predict the input-output (I/O) behavior of piezoelectric harvesters, perfect predictions are not expected since these devices are not exempt of uncertainties. The accuracy of the output estimation is affected mainly by the uncertainties on its electromechanical properties, requiring in many cases a parameter identification based on experimental measurements. In this context, two main questions arise: (1) how to properly perform the electromechanical properties identification and (2) how to select the most adequate prediction model. The interest of this work is to answer both questions employing a Bayesian inferential scheme. In particular, a model class selection is established employing predictive models with different grades of nonlinearities, while the updated model parameters are identified using a transitional Markov chain Monte Carlo over the device’s frequency response. Different recommendations to achieve the mentioned tasks are offered based on the number of experiments, the output type (voltage, electrical power, and displacement), and the method to identify the posterior distribution.