Mechanical Engineering

Permanent URI for this communityhttp://hdl.handle.net/1903/2263

Browse

Now showing 1 - 20 of 1223

2021 Annual Report - Center for Engineering Concepts Development
(2021-10-01) Anand, Davinder; Hazelwood, Dylan
CECD is twenty-one years old and continues to be a platform for experimenting with new ideas in engineering research and education with special attention to the impact of engineering on society. I’m pleased to report that we continue to be supported by ARL, NSWC-IHEODTD, the State of Maryland, and the Neilom Foundation. One hundred guests helped us celebrate our twenty years of activities highlighting innovative activities of contemporary interest that benefit the economic welfare of the State of Maryland and the Nation. This report provides a brief overview of those accomplishments as well as ongoing activities that bring great credit to our faculty and students that comprise CECD.
3D-Printed Microinjection Needle Arrays via a Hybrid DLP-Direct Laser Writing Strategy
(Wiley, 2023-02-05) Sarker, Sunandita; Colton, Adira; Wen, Ziteng; Xu, Xin; Erdi, Metecan; Jones, Anthony; Kofinas, Peter; Tubaldi, Eleonora; Walczak, Piotr; Janowski, Miroslaw; Liang, Yajie; Sochol, Ryan D.
Microinjection protocols are ubiquitous throughout biomedical fields, with hollow microneedle arrays (MNAs) offering distinctive benefits in both research and clinical settings. Unfortunately, manufacturing-associated barriers remain a critical impediment to emerging applications that demand high-density arrays of hollow, high-aspect-ratio microneedles. To address such challenges, here, a hybrid additive manufacturing approach that combines digital light processing (DLP) 3D printing with “ex situ direct laser writing (esDLW)” is presented to enable new classes of MNAs for fluidic microinjections. Experimental results for esDLW-based 3D printing of arrays of high-aspect-ratio microneedles—with 30 µm inner diameters, 50 µm outer diameters, and 550 µm heights, and arrayed with 100 µm needle-to-needle spacing—directly onto DLP-printed capillaries reveal uncompromised fluidic integrity at the MNA-capillary interface during microfluidic cyclic burst-pressure testing for input pressures in excess of 250 kPa (n = 100 cycles). Ex vivo experiments perform using excised mouse brains reveal that the MNAs not only physically withstand penetration into and retraction from brain tissue but also yield effective and distributed microinjection of surrogate fluids and nanoparticle suspensions directly into the brains. In combination, the results suggest that the presented strategy for fabricating high-aspect-ratio, high-density, hollow MNAs could hold unique promise for biomedical microinjection applications.
A 1D Reduced-Order Model (ROM) for a Novel Latent Thermal Energy Storage System
(MDPI, 2022-07-14) Kailkhura, Gargi; Mandel, Raphael Kahat; Shooshtari, Amir; Ohadi, Michael
Phase change material (PCM)-based thermal energy storage (TES) systems are widely used for repeated intermittent heating and cooling applications. However, such systems typically face some challenges due to the low thermal conductivity and expensive encapsulation process of PCMs. The present study overcomes these challenges by proposing a lightweight, low-cost, and low thermal resistance TES system that realizes a fluid-to-PCM additively manufactured metal-polymer composite heat exchanger (HX), based on our previously developed cross-media approach. A robust and simplified, analytical-based, 1D reduced-order model (ROM) was developed to compute the TES system performance, saving computational time compared to modeling the entire TES system using PCM-related transient CFD modeling. The TES model was reduced to a segment-level model comprising a single PCM-wire cylindrical domain based on the tube-bank geometry formed by the metal fin-wires. A detailed study on the geometric behavior of the cylindrical domain and the effect of overlapped areas, where the overlapped areas represent a deviation from 1D assumption on the TES performance, was conducted. An optimum geometric range of wire-spacings and size was identified. The 1D ROM assumes 1D radial conduction inside the PCM and analytically computes latent energy stored in the single PCM-wire cylindrical domain using thermal resistance and energy conservation principles. The latent energy is then time-integrated for the entire TES, making the 1D ROM computationally efficient. The 1D ROM neglects sensible thermal capacity and is thus applicable for the low Stefan number applications in the present study. The performance parameters of the 1D ROM were then validated with a 2D axisymmetric model, typically used in the literature, using commercially available CFD tools. For validation, a parametric study of a wide range of non-dimensionalized parameters, depending on applications ranging from pulsed-power cooling to peak-load shifting for building cooling application, is included in this paper. The 1D ROM appears to correlate well with the 2D axisymmetric model to within 10%, except at some extreme ranges of a few of the non-dimensional parameters, which lead to the condition of axial conduction inside the PCM, deviating from the 1D ROM.
A Deep Adversarial Approach Based on Multi-Sensor Fusion for Semi-Supervised Remaining Useful Life Prognostics
(MDPI, 2019-12-27) Verstraete, David; Droguett, Enrique; Modarres, Mohammad
Multi-sensor systems are proliferating in the asset management industry. Industry 4.0, combined with the Internet of Things (IoT), has ushered in the requirements of prognostics and health management systems to predict the system’s reliability and assess maintenance decisions. State of the art systems now generate big machinery data and require multi-sensor fusion for integrated remaining useful life prognostic capabilities. When dealing with these data sets, traditional prediction methods are not equipped to handle the multiple sensor signals in unison. To address this challenge, this paper proposes a new, deep, adversarial approach to any remaining useful life prediction in which a novel, non-Markovian, variational, inference-based model, incorporating an adversarial methodology, is derived. To evaluate the proposed approach, two public multi-sensor data sets are used for the remaining useful life prediction applications: (1) CMAPSS turbofan engine dataset, and (2) FEMTO Pronostia rolling element bearing data set. The proposed approach obtains favorable results when against similar deep learning models.
A Thermodynamic Entropy Approach to Reliability Assessment with Applications to Corrosion Fatigue
(MDPI, 2015-10-16) Imanian, Anahita; Modarres, Mohammad
This paper outlines a science-based explanation of damage and reliability of critical components and structures within the second law of thermodynamics. The approach relies on the fundamentals of irreversible thermodynamics, specifically the concept of entropy generation as an index of degradation and damage in materials. All damage mechanisms share a common feature, namely energy dissipation. Dissipation, a fundamental measure for irreversibility in a thermodynamic treatment of non-equilibrium processes, is quantified by entropy generation. An entropic-based damage approach to reliability and integrity characterization is presented and supported by experimental validation. Using this theorem, which relates entropy generation to dissipative phenomena, the corrosion fatigue entropy generation function is derived, evaluated, and employed for structural integrity and reliability assessment of aluminum 7075-T651 specimens.
A Unique Failure Mechanism in the Nexus 6P Lithium-Ion Battery
(MDPI, 2018-04-04) Saxena, Saurabh; Xing, Yinjiao; Pecht, Michael
Nexus 6P smartphones have been beset by battery drain issues, which have been causing premature shutdown of the phone even when the charge indicator displays a significant remaining runtime. To investigate the premature battery drain issue, two Nexus 6P smartphones (one new and one used) were disassembled and their batteries were evaluated using computerized tomography (CT) scan analysis, electrical performance (capacity, resistance, and impedance) tests, and cycle life capacity fade tests. The “used” smartphone battery delivered only 20% of the rated capacity when tested in a first capacity cycle and then 15% of the rated capacity in a second cycle. The new smartphone battery exceeded the rated capacity when first taken out of the box, but exhibited an accelerated capacity fade under C/2 rate cycling and decreased to 10% of its initial capacity in just 50 cycles. The CT scan results revealed the presence of contaminant materials inside the used battery, raising questions about the quality of the manufacturing process.
Access Scheduling and Controller Design in Networked Control Systems
(2005-10-05) Zhang, Lei; Hristu-Varsakelis, Dimitrios; Mechanical Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)
A Networked Control System (NCS) is a control system in which the sensors and actuators are connected to a feedback controller via a shared communication medium. In an NCS, the shared medium can only provide a limited number of simultaneous connections for the sensors and actuators to communicate with the controller. As a consequence, the design of an NCS involves not only the specification of a feedback controller but also that of a communication policy that schedules access to the shared communication medium. Up to now, this task has posed a significant challenge, due in large part to the modeling complexity of existing NCS architectures, under which the control and communication design problems are tightly intertwined. This thesis proposes an alternative NCS architecture, whereby the plant and controller choose to ``ignore'' the actuators and sensors that are not actively communicating. This new architecture leads to simpler NCS models in which the design of feedback controller and communication polices can be effectively decoupled. In that setting, we propose a set of medium access scheduling strategies and accompanying controller design methods that address a broad range of stabilization, estimation, and optimization problems for a general class of NCSs. The performance of the proposed methods is illustrated through a set of simulations and hardware experiments.
Acoustic Black Hole with Functionally Graded Perforated Rings
(2024) Petrover, Kayla; Baz, Amr; Mechanical Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)
This thesis investigates a novel class of acoustic black hole waveguides (ABH) that harnesses the functionality of an array of optimally designed Functionally Graded Perforated Rings (FGPR). Through this approach, the developed ABH exhibits inherent energy dissipation characteristics derived from the flow through the perforations, which enhances its acoustic absorption behavior, resulting in rapid attenuation of the propagating waves as it traverses the length of the waveguide. Furthermore, the proposed ABH structure facilitates the incorporation of additional porous absorbing layers sandwiching the rings to further enhance its absorption characteristics. Consequently, the operational mechanism of this new class of ABH waveguides diverges significantly from that of the conventional ABH waveguide, which generates the black hole effect by employing sequential solid-flat rings of decreasing inner radius to create the necessary virtual power law taper. Instead, the new class of ABH generates the black hole effect through reactive means rather than the effective dissipative means of the conventional ABH. Therefore, this thesis develops a transfer matrix modeling (TMM) approach and a finite element method (FEM) approach to model the absorption and reflection characteristics of the novel class of ABH, aiming to predict its behavior and, more importantly, demonstrate its merits as effective means for controlling sound propagation. The interior-point method for optimization was employed to select optimal geometric design parameters for the FGPR inside the proposed ABH. Accordingly, the ABH with FGPR is manufacturable, unlike the conventional, and its acoustic properties are tuned to minimize the reflection of incoming acoustic waves across the frequency range 0-5 kHz. This optimization process is then repeated for the ABH with FGPR sandwiched by absorbing layers. From the pool of optimal designs generated, those that offer manufacturing advantages are chosen for further testing and evaluation. Numerical simulations are conducted to showcase the advantages and behavior of the proposed ABH configurations. The predictions of the TMM and FEM are compared and validated against experimental results which are collected with the ACUPRO impedance tube. Furthermore, comparisons between the ABH with FGPR and conventional ABH are made to elucidate and distinguish their respective behaviors and underlying principles of operation.
ACOUSTIC EMISSION-BASED STRUCTURAL HEALTH MANAGEMENT AND PROGNOSTICS SUBJECT TO SMALL FATIGUE CRACKS
(2014) Keshtgar, Azadeh; Modarres, Mohammad; Mechanical Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)
One of the major concerns in structural health management (SHM) is the early detection of growing crack. Using this, future consequential damage due to crack propagation can be reduced or eliminated by scheduling maintenance which can prevent costly downtime. Early crack detection can also be used to predict the remaining useful life of a system. Acoustic Emission (AE) is a non-destructive testing (NDT) method with potential applications for locating and monitoring fatigue cracks during SHM and prognosis. The research presented in this dissertation focuses on the structural health monitoring using AE. In this research a correlation between AE signal characteristics and crack growth behavior is established, and a probabilistic model of fatigue crack length distribution based on certain AE signal features is developed. In order to establish the AE signal feature versus the fatigue crack growth model and study the consistency and accuracy of the model, several standard fatigue experiments have been performed using standard test specimens subjected to cyclic loading with different amplitude and frequencies. Bayesian analysis inference is used to estimate the parameters of the model and associated model error. The results indicate that the modified AE crack growth model could be used to predict the crack growth rate distribution at different test conditions. In the second phase of this research, an AE signal analysis approach was proposed in order to detect the time of crack initiation and assess small crack lengths, which happen during the early stages of damage accumulation. Experimental investigation from uniform cyclic loading tests indicated that initiation of crack could be identified through the statistical analysis of AE signals. A probabilistic AE-based model was developed and the uncertainties of the model were assessed. In addition, a probabilistic model validation approach was implemented to validate the results. The developed models were properly validated and the results were accurate. It was shown that the updated model can be used for detection of crack initiation as well as prediction of small crack growth in early stages of propagation. It was found that the novel AE monitoring technique facilitates early detection of fatigue crack, allows for the original life predictions to be updated and helps to extend the service life of the structure. Finally, a quantification framework was proposed to evaluate probability of failure of structural integrity using the observed initial crack length. The outcome of this research can be used to assess the reliability of structural health by estimating the probability density function of the length of a detected crack and quantifying the probability of failure at a specified number of cycles. The proposed method has applications in on-line monitoring and evaluation of structural health and shows promise for use in fatigue life assessment.
ACTIVE ACOUSTIC METAMATERIAL
(2015) Althamer, Saeed; Baz, Amr; Mechanical Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)
A class of active acoustic metamaterial (AAMM) is presented. The proposed AAMM consists of an acoustic transmission line connected in parallel to an array of Helmholtz resonators that are provided with actively controlled boundaries. In this manner, the AAMM is in effect an assembly of periodic cells, each of which consists of a Helmholtz resonator connected in parallel to two sections of the transmission line. The two sections meet the Helmholtz resonator at its neck. The local control action at each Helmholtz resonator of a unit cell is generated by using a Proportional and Derivative (PD) as well as Fractional Derivative (FD) controllers. The controllers that rely in their operation on the measurement of the flow resulting from the deflection of the resonator boundary and the flow rates inside the two transmission line sections before and after the resonator. Such a single local control action is shown to be capable of controlling the local effective density and elasticity of each unit cell. Lumped-parameter models are developed to model the dynamics and control characteristics of the AAMM under different gains for the PD controller and exponents of the FD controller. The models are exercised to demonstrate the ability of the FD controller in generating metamaterials with double negative effective density and elasticity over broad frequency ranges as compared to conventional Proportional and Derivative (PD) controllers. With such capabilities, the development of AAM with FD control action may provide viable means for generating desirable spatial distributions of density and elasticity over broad frequency band using a small number of control actuators.
ACTIVE CONTROL OF NON-RECIPROCAL ACOUSTIC METAMATERIAL WITH A DYNAMIC CONTROLLER
(2019) Raval, Suraj; Baz, Amr; Mechanical Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)
Reciprocity is one of the fundamental properties in the field of wave propagation. In acoustics, this property helps in various practical applications. But, breaking this reciprocity also has useful applications. As a result of which, many researchers have tried to break the reciprocity in acoustics, which is comparatively difficult, unlike in fields such as electro-magnetics. Majority of these proposed methods to break the reciprocity are hard-wired systems, which work for a very limited frequency range. Thus, we have introduced a non-reciprocal metamaterial having boundary control with the help of piezoelectric sensors and actuators. A theoretical model is introduced to induce the nonreciprocal behavior, and it is backed up by providing experimental evidence. Our setup consists of a cylindrical cell made up of acrylic, filled with water, having four piezo sensors/actuators, two on each end. The idea is to excite the piezo cell through an actuator on one side, collect the resulting signal from the piezo sensor on the other side, and perform appropriate mathematical operations on this signal to produce a feedback/control signal via a specially designed dynamic control action. This control signal affects the propagation of pressure waves through the water medium inside the cell as it introduces a virtual gyroscopic effect of controlled magnitude and direction. Thus, this is how non-reciprocity is introduced and controlled into the metamaterial cell. The obtained theoretical and experimental results demonstrate the effectiveness of the dynamic controller in breaking the acoustic reciprocity. Extension of this work to multi-cell metamaterial configuration is natural extension to be pursued.
Active Control of Sound Transmission Into Three-Dimensional Enclosures
(2004-05-12) Al-Bassyiouni, Moustafa; Balachandran, Balakumar; Mechanical Engineering
The aim of this dissertation work is to understand active control of sound fields inside a three-dimensional rectangular enclosure into which noise is transmitted through a flexible boundary. To this end, analytical and numerical studies have been conducted. In the modeling efforts, a spherical wave excitation, which is generated by a noise source located in the near field of the flexible panel, is considered. Piezoelectric patches, which are bonded symmetrically to the top and bottom surfaces of the panel, are used as actuators. Microphones located inside and outside the enclosure serve as pressure sensors. The efforts account for panel interactions with both the external sound field and the enclosed sound field, and this feature makes it appealing for model-based active control schemes. The feasibility of implementing two zero spillover schemes for active structural-acoustic control systems has been studied through analysis and experiments. These schemes have been developed to ensure that spillover does not occur outside the control bandwidth. The numerical results are found to be in good agreement with the corresponding experimental observations; attenuations ranging up to 18.1 dB are experimentally obtained for narrowband disturbances and an attenuation of 8.3 dB is obtained for broadband excitation in the frequency range of 40 Hz £ f £ 230 Hz. The following contributions have resulted from this work: i) an analytical model capable of predicting the external pressure fields due to both the noise source and structural?acoustic interactions and that accounts for the general case of spherical wave propagation, ii) development of zero spillover, active structural-acoustic control schemes for controlling three?dimensional sound fields, and iii) a new relaxed zero spillover control scheme to ensure that the controlled response is bounded over the entire frequency range.
AN ACTIVE NON-INTRUSIVE SYSTEM IDENTIFICATION APPROACH FOR CARDIOVASCULAR HEALTH MONITORING
(2014) Fazeli, Nima; Hahn, Jin-Oh; Mechanical Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)
In this study a novel active non-intrusive system identification paradigm is developed for the purpose of cardiovascular health monitoring. The proposed approach seeks to utilize a collocated actuator sensor unit devised from the common blood pressure cuff to simultaneously 1) produce rich transmural blood pressure waves that propagate through the cardiovascular system and 2) to make measurements of these rich peripheral transmural blood pressures utilizing the pressure oscillations produced within the cuffs bladder in order to reproduce the central aortic blood pressure accurately. To achieve this end a mathematical model of the cardiovascular system is developed to model the wave propagation dynamics of the external (excitation applied by the cuff) and internal (excitation produced by the heart) blood pressure waveforms through the cardiovascular system. Next a system identification protocol is developed in which rich transmural blood pressures are recorded and used to identify the parameters characterizing the model. The peripheral blood pressures are used in tandem with the characterized model to reconstruct the central aortic blood pressure waveform. The results of this study indicate the developed protocol can reliably and accurately reproduced the central aortic blood pressure and that it can outperform its intrusive passive counterpart (the Individualized Transfer Function methodology). The root-mean-square error in waveform reproduction, pulse pressure error and systolic pressure errors were evaluated to be 3.31 mmHg, 1.36 mmHg and 0.06 mmHg respectively for the active nonintrusive methodology while for the passive intrusive counterpart the same errors were evaluated to be 4.12 mmHg, 1.59 mmHg and 2.67 mmHg indicating the superiority of the proposed approach.
ACTIVE NON-RECIPROCAL ACOUSTIC METAMATERIALS
(2022) Zhou, Han; Baz, Amr AM; Mechanical Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)
This dissertation presents different configurations of active Acoustic MetaMaterials (AMM) which are proposed in order to control the flow of vibration and acoustic wave propagations in various applications. Distinct among these configurations is a 1-dimensional (1D) periodic array which consists of an assembly of active acoustic unit cells which are provided with programmable piezoelectric elements. By tuning the structural properties of these cells, the 1D array can impede the wave propagation over specific frequency ranges. In order to achieve non-reciprocal acoustic wave transmission of the AMMs, three different methodologies are introduced including active control of the piezoelectric elements using virtual gyroscopic control actions, eigenstructure shaping controller, and finally spatial-temporal modulation algorithm.Theoretical models are developed to investigate the fundamentals and the underlying physical phenomena associated with all the considered three AMM configurations. Experimental prototypes of all these AMM configurations are built and tested to demonstrate their effectiveness in controlling the propagation of vibration and noise through these materials. Furthermore, the experimental results are used to validate the developed theoretical models. The developed theoretical and experimental approaches are envisioned to be valuable tools in the design of arrays of AMM for various applications which are only limited by our imagination.
ACTIVE NONRECIPROCAL ACOUSTIC METAMATERIALS
(2017) Baqai, Bilal; Baz, Amr M; Mechanical Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)
In this thesis, the emphasis is placed on the development of a class of active acoustic diodes and metamaterials in an attempt to control the flow and distribution of acoustic energy in acoustic cavities and systems. Such development departs radically from the currently available approaches where the non-reciprocities are generated by hard-wired designs, favoring one transmission direction which is dictated by the arrangement of the hardware and hence it cannot be reversed, or without the presentation of rigorous control theory analysis. The proposed active nonreciprocal acoustic metamaterial (ANAM) cell consists of only one-dimensional acoustic cavity provided with active flexible boundaries. These boundaries are made from piezoelectric bimorphs with the inner layers which interact directly with cavity acting as sensors for monitoring the pressures of the propagating acoustic waves. The outer layers of the bimorphs provide the necessary control actions by direct application of the appropriate control voltage on each layer or by proper connection of nonlinearly activated shunted networks of electrical components such as the switching resistor networks. The control of the switching is carried out using the robust Sliding Mode Control (SMC) strategy. In this strategy, a lumped-parameter model of the ANAM cell is developed to control the strength of the nonreciprocal characteristics of the cell by proper selection of the slope of the switching surfaces. Appropriate optimization strategies are developed to enable a rational selection of the characteristics of the switching surfaces. Numerical examples are presented to demonstrate the effectiveness of the proposed ANAM in tuning and programming the directivity, flow, and distribution of acoustic energy propagating though the metamaterial. Experimental demonstration of the proposed ANAM is presented and includes a comprehensive investigation of the effect of the parameters of the SMC on the system performance. Such investigations are carried out in an attempt to validate the capabilities of ANAM in controlling the non-reciprocity in magnitude and direction. The presented theoretical and experimental techniques provide invaluable tools for designing and predicting the performance of this class of ANAM.
Adaptive Gradient Assisted Robust Optimization with Applications to LNG Plant Enhancement
(2012) Mortazavi, Amir Hossein; Azarm, Shapour; Radermacher, Reinhard K; Mechanical Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)
About 8% of the natural gas feed to a Liquefied Natural Gas (LNG) plant is consumed for liquefaction. A significant challenge in optimizing engineering systems, including LNG plants, is the issue of uncertainty. To exemplify, each natural gas field has a different gas composition, which imposes an important uncertainty in LNG plant design. One class of optimization techniques that can handle uncertainty is robust optimization. A robust optimum is one that is both optimum and relatively insensitive to the uncertainty. For instance, a mobile LNG plant should be both energy efficient and its performance be insensitive to the natural gas composition. In this dissertation to enhance the energy efficiency of the LNG plants, first, several new options are investigated. These options involve both liquefaction cycle enhancements and driver cycle (i.e., power plant) enhancements. Two new liquefaction cycle enhancement options are proposed and studied. For enhancing the diver cycle performance, ten novel driver cycle configurations for propane pre-cooled mixed refrigerant cycles are proposed, explored and compared with five different conventional driver cycle options. Also, two novel robust optimization techniques applicable to black-box engineering problems are developed. The first method is called gradient assisted robust optimization (GARO) that has a built-in numerical verification scheme. The other method is called quasi-concave gradient assisted robust optimization (QC-GARO). QC-GARO has a built-in robustness verification that is tailored for problems with quasi-concave functions with respect to uncertain variables. The performance of GARO and QC-GARO methods is evaluated by using seventeen numerical and engineering test problems and comparing their results against three previous methods from the literature. Based on the results it was found that, compared to the previous considered methods, GARO was the only one that could solve all test problems but with a higher computational effort compared to QC-GARO. QC-GARO's computational cost was in the same order of magnitude as the fastest previous method from the literature though it was not able to solve all the test problems. Lastly the GARO robust optimization method is used to devise a refrigerant for LNG plants that is relatively insensitive to the uncertainty from natural gas mixture composition.
ADAPTIVE SAMPLING METHODS FOR TESTING AUTONOMOUS SYSTEMS
(2018) Mullins, Galen Edward; Gupta, Satyandra K; Mechanical Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)
In this dissertation, I propose a software-in-the-loop testing architecture that uses adaptive sampling to generate test suites for intelligent systems based upon identifying transitions in high-level mission criteria. Simulation-based testing depends on the ability to intelligently create test-cases that reveal the greatest information about the performance of the system in the fewest number of runs. To this end, I focus on the discovery and analysis of performance boundaries. Locations in the testing space where a small change in the test configuration leads to large changes in the vehicle's behavior. These boundaries can be used to characterize the regions of stable performance and identify the critical factors that affect autonomous decision making software. By creating meta-models which predict the locations of these boundaries we can efficiently query the system and find informative test scenarios. These algorithms form the backbone of the Range Adversarial Planning Tool (RAPT): a software system used at naval testing facilities to identify the environmental triggers that will cause faults in the safety behavior of unmanned underwater vehicles (UUVs). This system was used to develop UUV field tests which were validated on a hardware platform at the Keyport Naval Testing Facility. The development of test cases from simulation to deployment in the field required new analytical tools. Tools that were capable of handling uncertainty in the vehicle's performance, and the ability to handle large datasets with high-dimensional outputs. This approach has also been applied to the generation of self-righting plans for unmanned ground vehicles (UGVs) using topological transition graphs. In order to create these graphs, I had to develop a set of manifold sampling and clustering algorithms which could identify paths through stable regions of the configuration space. Finally, I introduce an imitation learning approach for generating surrogate models of the target system's control policy. These surrogate agents can be used in place of the true autonomy to enable faster than real-time simulations. These novel tools for experimental design and behavioral modeling provide a new way of analyzing the performance of robotic and intelligent systems, and provide a designer with actionable feedback.
Adaptive Superposition of Finite Element Meshes in Linear and Nonlinear Dynamic Analysis
(2005-12-05) Yue, Zhihua; Robbins, Donald; Mechanical Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)
The numerical analysis of transient phenomena in solids, for instance, wave propagation and structural dynamics, is a very important and active area of study in engineering. Despite the current evolutionary state of modern computer hardware, practical analysis of large scale, nonlinear transient problems requires the use of adaptive methods where computational resources are locally allocated according to the interpolation requirements of the solution form. Adaptive analysis of transient problems involves obtaining solutions at many different time steps, each of which requires a sequence of adaptive meshes. Therefore, the execution speed of the adaptive algorithm is of paramount importance. In addition, transient problems require that the solution must be passed from one adaptive mesh to the next adaptive mesh with a bare minimum of solution-transfer error since this form of error compromises the initial conditions used for the next time step. A new adaptive finite element procedure (s-adaptive) is developed in this study for modeling transient phenomena in both linear elastic solids and nonlinear elastic solids caused by progressive damage. The adaptive procedure automatically updates the time step size and the spatial mesh discretization in transient analysis, achieving the accuracy and the efficiency requirements simultaneously. The novel feature of the s-adaptive procedure is the original use of finite element mesh superposition to produce spatial refinement in transient problems. The use of mesh superposition enables the s-adaptive procedure to completely avoid the need for cumbersome multipoint constraint algorithms and mesh generators, which makes the s-adaptive procedure extremely fast. Moreover, the use of mesh superposition enables the s-adaptive procedure to minimize the solution-transfer error. In a series of different solid mechanics problem types including 2-D and 3-D linear elastic quasi-static problems, 2-D material nonlinear quasi-static problems, and 2-D transient problems for linear elastic and material nonlinear materials, the s-adaptive solution is compared to a solution obtained using a non-adaptive, uniform refined mesh. These comparisons clearly demonstrate that the s-adaptive method is capable of generating a solution with the same accuracy level as a non-adaptive, uniform refined mesh; however, the s-adaptive solution uses far fewer DOF and consequently executes much faster.
ADAPTIVITY IN WALL-MODELED LARGE EDDY SIMULATION
(2022) Kahraman, Ali Berk; Larsson, Johan; Mechanical Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)
In turbulence-resolving simulations, smaller eddies account for most of the computational cost. This is especially true for a wall-bounded turbulent flow, where a wall-resolved large eddy simulation might use more than 99% of the computing power to resolve the inner 10% of the boundary layer in realistic flows.The solution is to use an approximate model in the inner 10% of the boundary layer where the turbulence is expected to exhibit universal behavior, a technique generally called wall-modeled large eddy simulation. Wall-modeled large-eddy simulation introduces a modeling interface (or exchange location) separating the wall-modeled layer from the rest of the domain. The current state-of-the-art is to rely on user expertise when choosing where to place this modeling interface, whether this choice is tied to the grid or not. This dissertation presents three post-processing algorithms that determine the exchange location systematically. Two algorithms are physics-based, derived based on known attributes of the turbulence in attached boundary layers. These algorithms are assessed on a range of flows, including flat plate boundary layers, the NASA wall-mounted hump, and different shock/boundary-layer interactions. These algorithms in general agree with what an experienced user would suggest, with thinner wall-modeled layers in nonequilibrium flow regions and thicker wall-modeled layers where the boundary layer is closer to equilibrium, but are completely ignorant to the cost of the simulation they are suggesting. The third algorithm is based on the sensitivity of the wall-model with the predicted wall shear stress and a model of the subsequent computational cost, finding the exchangelocation that minimizes a combination of the two. This algorithm is tested both a priori and a posteriori using an equilibrium wall model for the flow over a wall-mounted hump, a boundary layer in an adverse pressure gradient, and a shock/boundary-layer interaction. This third algorithm also produces exchange locations that mostly agree with what an experienced user would suggest, with thinner layers where the wall-model sensitivity is high and thicker layers where this sensitivity is low. This suggests that the algorithm should be useful in simulations of realistic and highly complex geometries.
Additive Manufacturing of Microfluidic Technologies via In Situ Direct Laser Writing
(2021) Alsharhan, Abdullah; Sochol, Ryan; Mechanical Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)
Innovations in microfluidic technologies hold great promise for a wide range of chemical, biomedical, and soft robotic applications. Unfortunately, key drawbacks associated with soft lithography-based microfabrication processes hinder such progress. To address these challenges, we advance a novel submicron-scale additive manufacturing (AM) strategy, termed “in situ direct laser writing (isDLW)”. IsDLW is an approach that benefits from the architectural versatility and length scales inherent to two-photon polymerization (2PP), while simultaneously supporting the micro-to-macro interfaces required for its effective utilization in microfluidic applications. In this dissertation, we explore isDLW strategies that enable passive and active 3D microfluidic technologies capable of enhancing “on-chip” autonomy and sophistication. Initially, we use poly(dimethylsiloxane) (PDMS)-based isDLW to fabricate microfluidic diodes that enable unidirectional rectification of fluid flow. We introduce a novel cyclic olefin polymer (COP)-based isDLW strategy to address several limitations related to structural adhesion and compatibility of PDMS microchannels. We use this COP-based approach to print microfluidic transistors comprising flexible and free-floating components that enable both “normally open” (NO) and “normally closed” (NC) functionalities—i.e., source-to-drain fluid flow (QSD) through the transistor is either permitted (NC) or obstructed (NO) when a gate input (PG) is applied. As an exemplar, we employ COP-based isDLW to print an integrated microfluidic circuit (IMC) comprised of soft microgrippers downstream of NC microfluidic transistors with distinct PG thresholds. All of these microfluidic circuit elements are printed within microchannels ≤ 40 μm in height, representing the smallest such components (to our knowledge). Theoretical and experimental results illustrate on the operational efficacy of these components as well as characterize their performance at different input conditions, while IMC experimental results demonstrate sequential actuation of the microrobotic components to realize target gripper operations with a single PG input. Furthermore, to investigate the utility of this strategy for static microfluidic technologies, we fabricate: (i) interwoven bioinspired microvessels (inner diameters < 10 μm) capable of effective isolation of distinct microfluidic flow streams, and (ii) deterministic lateral displacement (DLD) microstructures that enable continuous sorting of submicron particles (860 nm). In combination, these results suggest that the developed AM strategies offer a promising pathway for advancing state-of-the-art microfluidic technologies for various biological and soft robotic applications.