Flexible viscoelastic foams with relative density in the range of 5-50% are widely used to control interaction forces, notably in cushions for impact protection, shoes, mechanical grippers, etc. The design of new foams to improve performance is challenging, due to the lack of efficient models to connect cellular architecture, viscoelasticity of the base resin, and non-linear responses that stem from large deformation of open pores. This paper presents highly efficient solutions for the response of linearly viscoelastic, kinematically non-linear and shear-deformable struts that are loaded at an oblique angle to their axis. The models can be used to predict structure-property relationships for strut-based metamaterials, including non-linear stress-strain response and accompanying hysteresis. To illustrate this, the entire design space of a simple unit cell is mapped in terms of a broad range of strut angles and thicknesses; the unit cell is inspired by features in stochastic foams that dominate their response. An enormous range of stress-strain responses can be obtained from this single foam-like cell, including purely softening, purely stiffening, and non-monotonic stiffness with adjustable tangent moduli. The model is also used to illustrate connections between viscoelastic material parameters, loading rate, and hysteresis for two distinct classifications of non-linear stress-strain response.

The transient thermal response of a 15-cell, 48 V, lithium-ion battery pack for an unmanned ground vehicle (UGV) was simulated using ANSYS fluent. Heat generation rates and specific heat capacity of a single cell were experimentally measured and used as input to the thermal model. A heat generation load was applied to each battery, and natural convection film boundary conditions were applied to the exterior of the enclosure. The buoyancy-driven natural convection inside the enclosure was modeled along with the radiation heat transfer between internal components. The maximum temperature of the batteries reached 65.6 °C after 630 s of usage at a simulated peak power draw of 3600 W or roughly 85 A. This exceeds the manufacturer's maximum recommended operating temperature of 60 °C. We present a redesign of the pack that incorporates a passive thermal management system consisting of a composite expanded graphite (EG) matrix infiltrated with a phase-changing paraffin wax. The redesigned battery pack was similarly modeled, showing a decrease in the maximum temperature to 50.3 °C after 630 s at the same power draw. The proposed passive thermal management system kept the batteries within their recommended operating temperature range.

The friction-excited dynamics of a two-sided interface are examined experimentally, and complementary modeling and simulation efforts are used to understand the experimental observations. The experiment uses a thin fiber pulled from between two clamping pads. The experiment shows that stick–slip behaviors of the two clamping pads are periodic but out of phase, a behavior dubbed “walking stick–slip”, because the pullout specimen pulls out of the fixture in discrete, alternating steps. The friction behavior is understood using two dynamic models: (i) a three degree-of-freedom translational model, and (ii) a model augmented with rotational degrees-of-freedom. The out-of-phase walking arises from the component rotations. “Walking” stick–slip (to our knowledge) is a previously unreported phenomenon which may have implications in a variety of other situations.

We present an analysis of the fatigue failure of an 18 tooth star-ratchet gear (SRG). The subject gear was implemented in the freewheel assembly of a mountain bicycle. After 6 years of service, the gear failed unexpectedly during a typical off-road ride. The unique geometry of SRGs precluded a simple comparison to existing gear lifetimes. Scanning Electron Microscopy (SEM) analysis of the failed gears showed crack initiation at the root of the gear teeth, followed by fatigue crack propagation and eventual chip-out. A biomechanical analysis of pedaling forces, coupled with explicit power data obtained from instrumented rides over the same trails, in conjunction with a Finite Element Analysis (FEA) of the gear, were used to determine stress amplitudes for fatigue calculations. Energy dispersive spectroscopy (EDS) determined the alloy composition of the gear and thus set the strength and fatigue properties of the gear. Basquin's law, Goodman's mean stress correction, and Miner's rule were used to estimate the lifetime, in bike rides, of the gear. Our analysis led to an estimate of 2288 rides, while failure was reported after roughly 312 rides. Given the uncertainties in fatigue life estimation and service use, we find this estimate acceptable. Published by Elsevier Ltd.

Performance and reliability of silicon based microelectromechanical systems (MEMS) and photovoltaic (PV) devices are often strongly affected by defects and residual stresses. As a result, both industry and academia need new tools that can rapidly locate and quantify defects. The grey-field photoelastic technique has been shown to improve detection and residual stress quantification of defects in wafer bonded MEMS structures. In the PV industry, the tool has proven capable of detecting high bulk stress in wafers, damage due to wafer cutting, and stresses associated with trapped cracks. In this paper, we describe the development of an infrared photoelastic tool that captures full-field residual stress images at camera framing rates. This solid state tool, which adapts technology developed for visible light residual stress inspection, reduces the data collection time for residual stress maps from several minutes to a fraction of a second. We demonstrate the tool on canonical samples including a beam in bending and disk in compression to verify stress results and identify the lower limits of detection. We then demonstrate the tool on several industrial applications.

Designing a surface treatment process that generates high-strength bonding at low annealing temperatures with high overall process yields has been hampered by the development and growth of gas bubbles trapped at the bonded interface. This paper compares three methods for detecting and quantifying gas bubble defects in wafer-bonded silicon subjected to different surface treatments: (i) scanning acoustic microscopy (SAM), (ii) IR transmission (IRT) imaging, and (iii) IR photoelasticity [using the grey-field polariscope (IR-GFP)]. The SAM and IR-GFP are shown to detect the same defects, which are not possible to visualize with IRT, even with a modified optical train. In addition, the magnitude of the residual stress fields associated with each defect was quantified using IR-GFP shear stress imaging. Residual stresses are found to vary by more than a factor of 2, depending on the wafer pretreatment. These results also indicate that residual stress levels from interfacial gas bubbles are similar to those reported for particles trapped at the bond interface.

Characterizing the mechanical properties of materials and biological systems at the nanoscale requires accurate measurement of forces on the order of μN and less. Due to the scale of the measurements and size of the instrumentation, calibration of nanoscale devices presents a new challenge in metrology. In order to ensure accuracy of results, traceable calibrations must be performed on nanoscale instrumentation. Our group recently developed a novel MEMS-based high resolution load cell with force resolution on the order of μN. This paper reports on a simple method for traceably calibrating our device using dead weights that could be generalized to other MEMS-based load cells. In this article, fabrication of a MEMS load cell is detailed and we compare our calibrated force–displacement curves to a non-linear theoretical prediction, revealing errors as great as 29%.

We present ultra low-cycle fatigue experiments of axisymmetric nano-crystalline gold films of nano-thickness deformed by a spherical indenter using a recently developed freestanding membrane test. Freestanding membranes of gold were centrally deflected using a spherical indenter attached to a MEMS load cell. Fabrication of the films for these experiments yielded films 100 nm thick and 500 μm in diameter. We observed that the plastically deformed thin films recover completely in time at room temperature. One particular film was loaded 4 consecutive times, with the fourth loading leading to its fracture after full recovery from the three prior loadings. Observation of this phenomenon directed us to term this behavior “Ultra Low cycle fatigue”.

In this paper, we present experimental measurements of the strain energy release rate for stiction-failed polysilicon microcantilevers using a newly developed single cantilever beam peel test. Our experiments show that dry-contacting microcantilevers adhere exclusively as tip-stuck, "arc-shaped" stiction failures, while adhesion under "wet" conditions generate exclusively "s-shaped" stiction failures. Microcantilevers were "peeled" from the substrate under displacement control using a piezoelectric actuator attached to one end of an array of microcantilever beams. The crack length was monitored using interferometric imaging, and related to the applied displacement using established equations from linear elastic fracture mechanics. The pull-off forces associated with "arc-shaped" stiction failures were an average value of 89.7 nN, for 1000 mum long beams, and an average value of 123 nN for 1500 mum long beams. Adhesion energies for s-shaped failures were measured as 13.7 mJ/m2 for IPA released beams and 15.4 mJ/m2 for deionized water released beams. These values are in good agreement with previous measurements. The proposed experimental method enables application of a simple fracture mechanics model using a standard specimen geometry. These experiments, using both wet and dry adhesion failure conditions, show that the quality of the adhesive failure depends upon the magnitude of the forces pulling the microcantilever into contact with the underlying substrate

Self-monitoring of blood glucose has become an important and critical tool for effective management of diabetes. A closed-loop, real-time sensing and dispensing system, however, has been elusive. We describe here, a novel MEMS-based sensor intended to measure, in realtime, the concentration of glucose in a solution. The device utilizes a bi-material membrane with a functional hydrogel layer that swells reversibly, and in proportion to, the presence of glucose. The hydrogel is composed of 2-hydroxyethyl methacrylate (HEMA) functionalized with 3-acrylamidophenylboronic acid (AAPBA) groups as the glucose-sensitive moiety. Phenylboronic acid (PBA) derivatives interact with glucose by forming a charged complex, causing the hydrogel to swell by solvent intake [1].

Semiconductor wafer bonding has been identified as an enabling technology for a wide variety of semiconductor device processing applications such as wafer level encapsulation, three-dimensional structures and interconnects, and silicon-on-insulator substrates. In many of these applications accurate measurement and control of local residual stresses is critical for acceptable device yields and quality control. In this paper, synchrotron X-ray topography (XRT) and the infrared gray field polariscope (IR-GFP) are employed as full-field tools for the detection and measurement of residual stresses in wafer-bonded silicon. Both tools are used to inspect samples with varying levels of residual stresses from both wafer nanotopography and patterned interfacial features, resulting in excellent qualitative correlation between the tools. While the XRT offers higher spatial resolution and greater sensitivity to strain, the IR-GFP provides dramatically faster imaging rates, simple operating procedures, and instrument affordability. Based on these comparisons, the two techniques were shown to be complimentary tools for semiconductor processing control, the XRT being ideally suited as a laboratory research and development tool, while the IR-GFP is applicable for rapid process or quality control.

Microcantilever beams are frequently utilized as sensor platforms in microelectromechanical system devices. These highly compliant surface-micromachined structures generally fail by adhering to the underlying substrate during processing or subsequent operation. Such failures, which are commonly known as ldquostictionrdquo failures, can be prevented or repaired in a number of ways, including low adhesion coatings, rinsing with low surface energy agents, and active approaches such as laser irradiation. Gupta [ J. Microelectromech. Syst. vol. 13, pp. 696-700, 2004] recently demonstrated that stress waves could be used to repair stiction-failed structures. This paper extends the work of Gupta by developing a fracture mechanics theory of the repair process and compares that theory with corresponding experiments. We show that: 1) incremental crack growth is associated with each laser pulse, the extent of which is directly related to the laser fluence; 2) repeated pulsing fully repairs all of the microcantilevers; and 3) a fracture mechanics model accurately predicts the observed experimental results. [1664].

This paper identifies dynamic excitation parameters that promote decohesion of stiction-failed microcantilevers. The dynamic response of "s-shaped" adhered beams subjected to harmonic loading is described using modal analysis; this model is then used to predict the onset of debonding in the context of a critical interface energy. These theoretical results are used to rationalize preliminary experiments, which illustrate that dynamic excitation may be used to affect partial or complete repair of stiction-failed microcantilevers. The theoretical results provide fundamental insight regarding regimes where resonant effects trigger debonding and can serve as a potential mechanism for stiction repair. The models illustrate that driving a structure at resonance is usually beneficial with regards to debonding. However, this is not universally true; there is no benefit to driving a device at frequencies with unfavorable mode-shapes. Thus, these results provide a reasonable physical and mathematical explanation for the preliminary experimental results, while providing a roadmap for identifying parameters in future tests.

Recent advances in the microelectronics industry have led to increasing demand for on-line quality control throughout the entire fabrication process. The complexity of modern devices requires an increasing number of processing steps that amplifies the need for reliable inspection methods. It has become particularly important to develop techniques that quantify the residual stresses generated during each processing step. Localized stresses arising from defects at bonded interfaces can cause downstream processing failures. We conducted a broad range of experiments using controlled defects in order to better understand stresses that arise from bond defects. These defects consist of silicon dioxide ‘mesas’ that are patterned onto a silicon wafer that is subsequently bonded to a ‘smooth’ wafer. The defect dimensions were chosen to represent particles commonly found in a cleanroom. Each bonded pair was inspected using an infrared grey-field polariscope (IR-GFP) to obtain quantitative measures of the residual stress fields surrounding the ’mesas’. High-resolution x-ray topography (XRT) measurements were taken on the same samples to validate the measurements made using the IR-GFP with a standardized tool. While the XRT is capable of much higher spatial and stress resolution than the IR-GFP, both methods agreed qualitatively. These experiments show that, as expected, the debond radius increases with mesa height and decreases with increasing bond energy. The maximum residual stress around trapped particles was found to vary between 1.60 and 3.43 MPa for wafers treated using the standard RCA cleaning method.

We have developed a new scientific tool for the mechanical characterization of axisymmetric freestanding nano-thickness films. The tool consists of a calibrated MEMS-based load cell with a spherical loading tip, and a positioning system capable of sub-nm accuracy. The loading tip is positioned over the center of a free-standing film using a two-axis positioning stage. A Michelson Interferometer is positioned below the film to record the displacement field of the membrane, in situ, as the load cell deflects the membrane from above. As the MEMS load cell is lowered into the axisymmetric freestanding thin film, concentric fringes emanate from the center of the thin film sample. This tool allows us to record the applied load and membrane displacement simultaneously. To date, load cells with stiffnesses as low as 2nN/μm have been fabricated and calibrated. We have validated the operation of the apparatus using gold films 1080μm in diameter and 170nm thick.

The development of wafer bonded silicon and semiconductor technology is critical to many microelectronic and optoelectronic industries. Bonding processes are inherent to many silicon-on-insulator (SOI) substrates, optoelectronic integration, wafer level packaging, and a host of sensor and actuator applications. The interfacial quality of wafer bonded silicon structures, as measured by adhesion energy and defect population, is critical to achieving high yield processes. Several defects that result from wafer surface preparation (nanotopography, surface patterning) and bonding environment (trapped particles, trapped gasses) may generate significant residual stresses that adversely affect device performance. This paper will discuss how residual stresses are generated by interfacial defects and demonstrate the application of an infrared grey-field polariscope (IR-GFP) as a non-destructive inspection technology for locating, imaging and quantifying these defects. We will also present the development of a new model to quantify the residual stresses from trapped particles. 1. RESIDUAL STRESSES FROM WAFER BONDING DEFECTS Even as bonding is being conducted in specialized cleanroom environments with dedicated equipment, very small particles and trapped gasses are nearly impossible to eliminate. Furthermore, it has been shown that surface morphology has a significant influence on the quality of bonded silicon wafers. Elimination of nanometer scale waviness has proven to be a difficult process that has yet to be accomplished in mass production settings. These defects may result in high residual stresses at a bonded interface, which can alter device performance and durability. The most common method for locating defects in fusion bonded wafers is infrared transmission (IRT) exposure. However, IRT imaging is incapable of distinguishing between trapped gas or particle defects and suffers from spatial limitations due to optical diffraction. More importantly, IRT imaging provides no information on the residual stress state associated with each defect. Finally, IRT imaging is incapable of detecting the variations in bonding due to inherent surface nanotopography as the gap between the interfaces may be only on the order of 1-10 nanometers. Until now, X-ray topography is the only tool that has been shown to detect the influence of wafer surface nanotopography. 1.1. TRAPPED PARTICLES Particle contamination along a bonded interface prevents uniform contact between layers, resulting in poor bonding and locally high stresses. Particulate contamination at the interface arises from any number of sources including; air borne organic particles, spacer wear, or contaminants in cleaning chemicals. The effects of particle contamination are dramatic: trapped contaminant particles measuring just 250nm in diameter, are known to generate debond diameters greater than 2.5 mm in fusion bonded silicon [1]. Furthermore, dense dislocation networks are found to spread well beyond the debond radius during routine annealing of particle-contaminated fusion-bonded structures [2]. The yield strength of defect free silicon is roughly 230 MPa at 800°C [3], while the critical resolved shear stress of silicon at 1050°C is just 3.2 MPa [4]. These strengths are used as an order of magnitude approximation for stresses present during the annealing process, which cause dislocation motion. In addition to the scattering phenomenon introduced by the presence of dislocations, residual stresses may cause local variations in electron and hole mobility of the silicon substrates [5]. It is apparent that trapped particles can potentially lead to mechanical reliability problems while also affecting electrical properties. In the processing phase, trapped particles are especially troublesome for fusion bonded structures. Relatively weak van der Waals forces are utilized to bring the wafers into initial contact, so that large particles may completely prevent successful bonding. High strength room temperature bonding as well as high temperature annealing processes can pull substrates into more intimate contact and nearly complete bonding in the vicinity of fairly large particles. Though a smaller debond region will surround these trapped particles, the resulting structure will have higher residual stresses due to geometrically necessary bending around the particle and indentation forces at the contact points. 1.2. TRAPPED GAS BUBBLES Gasses may be trapped by all bonding processes when irregular bond fronts entrain gas in the wake of the bond front. These entrained gas pockets result in local debonds supported by gas pressure. Most commercial bonders have attempted to minimize this problem by bonding under vacuum conditions and by including fixturing that initiates a single bond front from a central contact point. Even with these additional measures it has not been possible to eliminate all gas bubbles. While particle contamination results in highly localized contact stresses, the stress field associated with gas bubbles is quite different. In fusion and anodic bonding, trapped pressures may result in stresses of the same magnitude as the trapped particles depending on the byproducts of the local chemical reaction. Anthony presents a detailed mechanism for developing debonded regions from multiple bonding fronts in anodically bonded structures [6]. As multiple bond fronts approach each other, one or both wafers may begin to bow due to a dimensional mismatch between the bonding fronts (as seen in Figure 2.7). The unbonded regions are not necessarily supported by trapped gas, but the area is likely to be highly stressed. A good analogy for this type of defect is the wrinkling of wallpaper when application goes awry. If two ends of the paper are inadvertently stuck to the wall and, adhesion progresses from these points, the central portion where the bond fronts meet will not lie flat on the wall. Harendt et al. [7] also report voids in fusion bonded structures resulting from bond fronts simultaneously initiating at different locations. 1.3. SURFACE NANOTOPOGRAPHY Several investigators have shown that surface morphology has a significant influence on the ability to bond silicon wafers. Further, these investigators have shown that surface morphology can have a large effect on the residual stress state of the bonded pair. Silicon wafers display important surface morphology on several length scales. The length scale of interest to the current study, first investigated by Mazsara et al. [8] and Yamada et al. [9] (often called surface waviness or nanotopography), is on the order of 1 mm, with peak-to-valley heights of only 20 nm (as determined by optical profilometry on their samples). Both researchers employed x-ray topography (XRT) to determine local strain contrast. They note that the periodicity of the strain pattern resulting from bonding forces should be comparable to the original wafer surface roughness. Local regions that do not bond as a result of trapped gasses or particles at the interface appear very smooth in these images because no local strain variations are present where the surfaces do not contact. Surface waviness is initially generated by the wire sawing operation, which is used to separate multiple wafers from a single silicon crystal ingot. The wire sawing technique has gained almost complete industrial acceptance over internal diameter (ID) sawing because of higher throughput and less wasted raw material [10]. Subsequent flattening processes, employing either lapping or grinding, are used to reduce topography variations. Both lapping [11] and soft pad grinding [12] appear to eliminate surface waviness at the micron level, but elastic springback will result in some degree of persistent underlying waviness. The magnitude of residual stresses in bonded wafers will depend on the relative coordination of the surface roughnesses of the mating wafers during bonding. The maximum residual stress state will occur where two peaks are aligned. Maszara et al. [8] estimated the local stress that results from elastic flattening of the rough surface using the theory of elastic spheres. They calculated that the stress normal to the surface, averaged over the asperity base for a wavelength of 1mm, with a height of 10nm is approximately σmax ≅ 10 MPa. 2. EXPERIMENTAL MEASUREMENT OF RESIDUAL STRESSES An infrared grey field polariscope (IR-GFP) has recently been developed for qualitatively inspecting bonded silicon devices [13]. Previous research has shown that the IR-GFP can qualitatively locate trapped particles and gas bubbles that are undetectable with traditional infrared transmission imaging [14,15]. This improved resolution is attributed to the larger spatial extent of the residual stress field compared with the debond region associated with a given defect. Experimental and theoretical quantification of residual stresses from these defects has yet to be investigated. Furthermore, residual stresses due to surface nanotopography have only previously been measured qualitatively using XRT measurements. Here, we present some quantified residual stress measurements from each type of defect and in the next section outline the initial efforts to model the residual stresses from bonding defects. In the following experiments, fusion bonded structures were generated by bonding two double-side polished silicon wafers. 2.1. TRAPPED PARTICLE EXPERIMENTS As shown in Figure 1 and Table 1, the bow-tie patterns that arise from imaging defects with the infrared polariscope are used to qualitatively locate defects can also quantify the residual stresses. Table 1 presents the average shear stress measured through the sample thickness, which are then utilized, with models outlined in Section 3, to quantify the maximum residual stress. In Figure 1a, a large particle has generated a debond in the upper right corner as seen using IRT imaging. However, several other defects are located by imaging the re

Nanoscale surface features at the interface of bonded silicon wafers are critical to achieving high yield wafer bonding processes. Interfacial features such as surface patterning, trapped particles and gasses, and wafer surface nanotopography can generate residual stresses that may affect subsequent device lifetimes, processing and performance. This paper presents a numerical model that captures a more complete description of bonding physics and mechanics than extensions of closed form thin-plate solutions. The model is verified by experimentally measuring residual stresses arising from patterned nanoscale surface features using an infrared grey- field polariscope. The proposed numerical model captures the functional form of the observed residual stress fields arising from interfacial defects and is in good agreement with experiments. These simulations indicate that high bond energies may produce debonds around small defects that are not possible to detect using infrared transmission inspection techniques, yet the associated residual stresses are increasingly large in magnitude.

We present results from a new procedure developed to quantify the pull-off force and strain energy release rates associated with stiction-failure in microdevices. The method is analogous to a standard, macro-scale peel test, but carried out using micro-scale devices. Adhesion is initiated by lowering an array of microcantilevers that protrude from a substrate into contact with a separate substrate. Displacement is controlled by a piezo-actuator with sub-nm resolution while alignment is controlled using linear and tilt stages. An interferometric microscope is used to align the array and the substrate and to record deflection profiles and adhesion lengths during peel-off. This geometry is accurately modeled using linear elastic fracture mechanics, creating a robust, reliable, standard method for measuring adhesion energies in stiction-failed microdevices.

We present experimental measurements of the mechanical properties of nanoscale-thick materials using a recently developed free-standing membrane test. Freestanding membranes of gold and a nano-porous silicate were centrally deflected using a spherical indenter attached to a microfabricated fixed-fixed beam device. Experiments of this type allow us to extract mechanical properties of the thin films without the complicating effects of the substrate, enabling accurate study of length scale effects. These experiments differ from previous work by testing freestanding films and by including the finite contact size of the indenter tip.