Stable Voids Research Articles

CFD-DEM study on heat transfer characteristics and microstructure of the blast furnace raceway with ellipsoidal particles

Coupled computational fluid dynamics and discrete element method (CFD-DEM) simulations are carried out to investigate the heat transfer characteristics and microstructure of the blast furnace (BF) raceway with different particle shapes. Three kinds of particles are considered including prolate and oblate ellipsoidal particles and spheres. The coupled CFD-DEM model is first validated based on experimental data and the capability of the combined sphere method (CSM) to investigate the heat transfer is also demonstrated. Then, the validated model is used to systematically explore formation and evolution of the BF raceway, particle behavior, force structure, heat transfer and temperature distribution with the effect of particle shapes on these issues discussed. Numerical results show that the raceway size formed by prolate ellipsoidal particles is the smallest among the three cases whereas the coordination number of which is the highest. Uniform contact force networks and typical orientation inclined angles of ellipsoidal particles influence the stable packing structure and homogeneous void distribution therein. The large surface area and homogeneous void structure of ellipsoidal particles can enhance the convective heat transfer. All the aforementioned factors collectively promote ellipsoidal particles to possess higher heat transfer efficiency than spheres. These findings could be helpful for the design and optimization of the BF.

Void structure stability in wet granular matter and its application to crab burrows and cometary pits

Cohesive granular matter can support stable void structures, which can universally be found in various scenes from everyday lives to space. To quantitatively characterize the stability and strength of a void structure in cohesive granular matter, we perform a simple tunnel-compression experiment with wet granular matter. In the experiment, a horizontal tunnel in a wet granular layer is vertically compressed with a slow compression rate. The experimental result suggests that the tunnel deformation can be classified into the following three types: (i) shrink, (ii) shrink with collapse, and (iii) subsidence by collapse. Using the experimental result, we estimate the stable limit of various void structures in a cohesive granular layer from crab burrows on a sandy beach to the pits observed on cometary surfaces.

On the mobility of defect clusters and their effect on microstructure evolution in fcc Ni under irradiation

µs-scale molecular dynamics studies of vacancy and interstitial clusters in fcc Ni revealed three- and one- dimensional (3-D and 1-D) modes of the cluster motion. The 1-D mobility of interstitial defects is known to enhance swelling rate. The theoretical analysis performed here suggests two novel mechanism by which the 3-D mobile clusters affect microstructure evolution under irradiation. First, the mobility of vacancy clusters hinders nucleation of stable voids due to recombination with interstitial-type loops and edge dislocations. Second, the capture efficiency of dislocations is higher for 3-D mobile vacancy and interstitial clusters than single defects, with the combined effect depending on relative fractions of clustered vacancies and interstitials produced in cascades. The observed differences in radiation damage of most fcc and bcc metals is attributed to the difference in cascade-produced vacancy defects: immobile SFTs and loops in fcc metals and mobile clusters in bcc metals. In this context, fcc Ni is similar to bcc metals due to its high stacking fault energy.

Propagation of electrostatic surface wave along the dust void boundary

An analytical model has been developed using fluid theory to study the propagation of high-frequency electrostatic surface wave along the dust void boundary. Fluid equations, which govern the formation of stable dust void, are suitably coupled with Maxwell’s fourth equation (for electrostatic limit) to obtain the linear dispersion relation under suitable boundary conditions. The study contains two situations where the excitation of the surface wave takes place along the void boundary (interface) due to the propagation of ion acoustic (IA) and dust ion acoustic (DIA) waves as well as the ion acoustic and dust acoustic waves in the void and dusty plasma region respectively. The linear dispersion relations in both the cases worked out to be complex in nature for wave frequency much higher than collisional frequencies between various plasma species. Special wave modes, namely the symmetric and asymmetric modes are identified in the first case. However, in the second case, surface wave modes identified have the real component of wave frequency varies linearly within small wave vector range and thereafter saturates to a constant value for large wave vector range. Dependence of surface wave frequency on plasma parameters is also investigated and significant results are highlighted.

2429 Optogenetic stimulation of corticotropin-releasing hormone expressing neurons in Barrington’s nucleus recapitulates the social stress voiding phenotype in mice

OBJECTIVES/SPECIFIC AIMS: Voiding postponement is common cause of LUT dysfunction in which children void infrequently with large volumes. This condition is modeled in mice that are subjected to social stress who show decreased voiding frequency and increased voided volumes along with increases in corticotropin-releasing hormone (CRH) expression in Barrington’s nucleus (BN) (i.e., the pontine micturition center). Optogenetics is a technique to selectively stimulate cells or neurons of interest via light activated channel receptors [i.e., channel-2 rhodopsin (ChR2)]. Here we examined the effects of optogenetic manipulation of CRH BN neurons on the in vivo voiding phenotype and urodynamics in awake mice. We hypothesized that stimulating these neurons at higher frequencies (10–50 Hz) would lead to CRH-dependent alterations in voiding phenotype (i.e., larger voided volumes and longer intermicturition intervals. METHODS/STUDY POPULATION: Double transgenic mice expressing ChR2 in CRH cells were generated using the Cre-lox recombinase system and had fiberoptic probes implanted into BN at 8 weeks of age. The mice also underwent simultaneous catheter placement into the bladder for in vivo cystometry. In vivo cystometry before and during optogenetic stimulation at various frequencies was performed 5–7 days postoperatively. Saline was perfused at 10 µL/minute and baseline stable voiding cycles were established. Bladder capacity, threshold pressure, voiding pressure, and voided volume were recorded at baseline and at each optogenetic setting. In some mice, the protocol was repeated in the presence of CRH antagonist (NBI 30775). RESULTS/ANTICIPATED RESULTS: Fiberoptic stimulation (470 nm at 25 and 50 Hz) produced a significant rise in the intermicturition interval, bladder capacities and increased void volumes. This effect was especially pronounced in females in whom bladder capacity and intermicturition interval more than doubled at 50 Hz stimulation. Fluoroscopic images confirmed complete bladder emptying with each void. The increased bladder capacity at higher frequencies (25 and 50 Hz) was CRH-dependent as injection of a CRH antagonist (NBI 30775) blocked the optogenetic effect. Control non-double mice showed no effects from optogenetic stimulation. DISCUSSION/SIGNIFICANCE OF IMPACT: Our results suggest that optogenetic stimulation of CRH-expressing neurons in BN at high frequency (25 and 50 Hz) inhibits micturition and recapitulates the voiding phenotype seen in socially stressed mice (large, infrequent voids). Lower frequencies of optogenetic stimulation (2 and 10 Hz) had no effects on cystometry parameters or voiding phenotype. In addition, females had a greater response to optogenetic stimulation compared with males with larger bladder capacities and longer intermicturition intervals. The changes in voiding phenotype seen were CRH dependent as blockage of the CRH receptor prevented changes in cystometry parameters with optogenetic stimulation. Further elucidation of these and other neural subpopulations in BN are warranted to understand micturition and how it may be manipulated in disease states such as voiding postponement and acute urinary retention.

Acute hypoxia alters reflex micturition behavior in urethane‐anesthetized carotid sinus intact and denervated adult female Sprague‐Dawley rat

Acute hypoxia is well known to increase respiratory drive by stimulating peripheral chemoreceptors located in the carotid bodies. Ongoing work in our laboratory has been examining the effects acute hypoxia on lower urinary tract (LUT) function in urethane‐anesthetized adult female rats. While these experiments have revealed that exposure to acute hypoxia is capable of modifying LUT function, the contribution of peripheral chemoreceptor‐mediated effects is unknown. Moreover, hypoxia‐induced modulation of LUT behavior is present both during hypoxia and transiently after the hypoxic stimulus is removed, such that hypoxia increases void frequency while the return to normoxia increases the threshold bladder pressure for voiding and prolongs the inter‐contraction interval (ICI). The present study was designed to evaluate whether the changes in LUT function noted are mediated by mechanisms that are dependent on carotid body hypoxia chemoreception. To this end, we conducted a series of experiments using continuous flow bladder cystometry with simultaneous recording of external urethral sphincter (EUS) EMG activity in urethane‐anesthetized spontaneously breathing adult female Sprague‐Dawley rats (n=10). Micturition events were elicited by saline infusion into the bladder via suprapubic catheter with the saline flow rate adjusted to induce a voiding frequency of approximately one void every 4–5 minutes. We also simultaneously recorded tongue, external oblique, and diaphragm muscle EMG activity via bipolar stainless steel EMG electrodes. Each rat was allowed to settle into a stable baseline voiding rhythm which was then recorded for approximately 1‐hr under normoxic conditions, after which an hypoxic challenge (12% O2, 15 min) was performed. In one group of rats, the carotid sinus nerves (CSN) were transected bilaterally prior to beginning the cytometry experiment while in a second group of rats the CSN were left intact. Success of the CSN denervation was confirmed by the absence of an increase in diaphragm EMG burst amplitude and frequency during the hypoxic challenge. Consistent with our prior observations, CSN intact rats responded to hypoxia with an increased voiding frequency, particularly at the onset of hypoxia; however, peak bladder pressure during contraction was variable with no consistent modulation by hypoxia. Upon return to normoxia, the initial voiding event was delayed, and subsequent events displayed increased ICI with elevated bladder pressure threshold. In contrast to CSN intact rats, CSN denervated rats exhibited a marked decrease in peak bladder contraction pressure during hypoxic exposure as well as a smaller increase in bladder pressure threshold after return to normoxia. No notable differences in hypoxia‐induced changes in ICI between CSN intact or transected rats was observed. These results suggest that hypoxia‐mediated carotid body afferent input is not required for producing a modulation of reflex micturition in response to hypoxia, but certain aspects of the hypoxic response may be impacted by carotid body dependent mechanisms.Support or Funding InformationNIH NS096514This abstract is from the Experimental Biology 2018 Meeting. There is no full text article associated with this abstract published in The FASEB Journal.

Pillar Fracturing a Sandstone Reservoir Shows Benefit Over Conventional Fracturing

This article, written by Special Publications Editor Adam Wilson, contains highlights of paper SPE 186199, “Unlocking Depleted and Low-Modulus Telisa Sandstone Reservoir With Pillar-Fracturing Technique: Well-Performance-Improvement Comparison With Conventional Fracturing,” by M. Azhari, SPE, N.F. Prakoso, and D. Ningrum, Medco E&P Indonesia, and L. Soetikno and A. Makmun, Schlumberger, prepared for the 2017 SPE/IATMI Asia Pacific Oil and Gas Conference and Exhibition, Bali, Indonesia, 17–19 October. The paper has not been peer reviewed. The Kaji Semoga Field in South Sumatra consists of three main reservoirs—Telisa sandstone (TLS), Baturaja limestone, and Talangakar sandstone. The successful development of TLS with hydraulic fracturing led to further efforts to maximize oil recovery. After a study with suitable samples and cases, pillar fracturing was considered. This method is similar to conventional fracturing techniques where fluid and proppant are used to create conductive paths in the reservoir layer; however, pillar fracturing relies on open flow channels. Pillar Fracturing A major step change in the development of the Kaji Semoga Field was to use the hydraulic-fracturing technique referred to as pillar fracturing. The pillar-fracturing technique creates stable voids within the proppant pack that serve as infinite-conductivity channels for fluid flow rather than the intragranular flow of conventional fracturing techniques (Fig. 1). Hydrocarbon will flow preferentially through the channels rather than through the proppant pack. In addition, pillar fracturing allows for better fracture cleanup, longer fracture effective half-lengths, and lower pressure drops along the fracture. Consequently, production after pillar fracturing is greater than that following conventional fracturing. Another advantage of the pillar-fracturing technique is a low screenout rate compared with conventional fracturing. The prevention of screen out is related to the different bridging characteristics of conventional- and pillar-fracturing slurries. The stable voids within the proppant pack are created through a combination of specific pumping schedules, specific equipment, and fracturing-fluid design. The voids are achieved by alternate pumping of gelled fluid in two types of pulses, proppant-laden pulses and proppant-free pulses. Fibers are added continuously during pumping to mitigate the dispersion of the proppant-laden pulses as they are conveyed throughout the surface lines and the wellbore and within the fracture. A dispersed pulse can create narrow fracture widths and a reduction in the number and quality of the channels, which leads to a reduction of fracture conductivity. The fiber also enhances the proppant-carrying capacity and prevents proppant settling within the fracture. The pulse stage is then followed by a continuous proppant tail-in stage near the end of the treatment before the flush stage. This tail-in stage ensures good connectivity between the wellbore and the channels created during the treatment. Specially engineered blending equipment was needed to ensure stable and consistent pulse delivery.

Annular flow stability within small-sized channels

An analytical study based on a variational thermodynamic principle is presented to evaluate the influence of surface tension on the stability of annular flow within small-sized channels. The model introduces phenomenological assumptions in the interfacial structure of the flow regime and theoretically draws the equilibrium transition line from an annular regime to the initiation of the partial wetting condition on the inner surface. By including surface tension, this model expands previous theories and identifies the stable flow configuration in terms of void fraction and interfacial extension. The significant influence of a higher surface tension and smaller diameter (i.e. lower Weber number) are responsible for a lower stable void fraction and higher slip ratio. A complete screening of the main influential parameters is conducted to explore the descriptive ability of the model. This analysis aims at contributing to the understanding of the stability of two-phase flow regimes and can be extended to the transition between other neighbouring regimes, including wall friction as well as liquid entrainment phenomena.

Nanoscale Mechanics of the Solid Electrolyte Interphase on Lithiated-Silicon Electrodes

The ∼300% volume changes of lithiated silicon electrodes (LixSi) during electrochemical cycling lead to cracking of the solid electrolyte interface (SEI). Here, we report how strain is transferred from LixSi to two primary inorganic SEI components: LiF and Li2O. Our first principle calculations show that LiF, effectively bonded on LixSi at x > 1, enables the entire interface structure to deform plastically by forming delocalized stable voids. In contrast, Li2O tightly bonded to LixSi is stiffer, and deforms rigidly across all x. Our results explain the significantly improved ductility of SEI with higher LiF versus Li2O content observed experimentally.

Active Adoption of Void Formation in Metal-Oxide for All Transparent Super-Performing Photodetectors.

Could ‘defect-considered’ void formation in metal-oxide be actively used? Is it possible to realize stable void formation in a metal-oxide layer, beyond unexpected observations, for functional utilization? Herein we demonstrate the effective tailoring of void formation of NiO for ultra-sensitive UV photodetection. NiO was formed onto pre-sputtered ZnO for a large size and spontaneously formed abrupt p-NiO/n-ZnO heterojunction device. To form voids at an interface, rapid thermal process was performed, resulting in highly visible light transparency (85–95%). This heterojunction provides extremely low saturation current (<0.1 nA) with an extraordinary rectifying ratio value of over 3000 and works well without any additional metal electrodes. Under UV illumination, we can observe the fast photoresponse time (10 ms) along with the highest possible responsivity (1.8 A W−1) and excellent detectivity (2 × 1013 Jones) due to the existence of an intrinsic-void layer at the interface. We consider this as the first report on metal-oxide-based void formation (Kirkendall effect) for effective photoelectric device applications. We propose that the active adoption of ‘defect-considered’ Kirkendall-voids will open up a new era for metal-oxide based photoelectric devices.

Replication Experiments in Microgravity Liquid Phase Sintering

Although considerable experience exists with sintering on Earth, the behavior under reduced gravity conditions is poorly understood. This study analyzes replica microgravity liquid phase sintering data for seven tungsten alloys (35 to 88 wt pct tungsten) sintered for three hold times (1, 180, or 600 minutes) at 1773 K (1500 °C) using 0.002 pct of standard gravity. Equivalent sintering is performed on Earth using the same heating cycles. Microgravity sintering results in a lower density and more shape distortion. For Earth-based sintering, minimized distortion is associated with low liquid contents to avoid solid settling and slumping. Distortion in microgravity sintering involves viscous spreading of the component at points of contact with the containment crucible. Distortion in microgravity is minimized by short hold times; long hold times allow progressive component reshaping toward a spherical shape. Microgravity sintering also exhibits pore coalescence into large, stable voids that cause component swelling. The microgravity sintering results show good replication in terms of mass change and sintered density. Distortion is scattered but statistically similar between the replica microgravity runs. However, subtle factors, not typically of concern on Earth, emerge to influence microgravity sintering, such that ground experiments do not provide a basis to predict microgravity behavior.

An analysis of hydrated proton diffusion in ab initio molecular dynamics

A detailed understanding of the inherently multiscale proton transport process raises a number of scientifically challenging questions. For example, there remain many (partially addressed) questions on the molecular mechanism for long-range proton migration and the potential for the formation of long-lived traps giving rise to burst-and-rest proton dynamics. Using results from a sizeable collection of ab initio molecular dynamics (AIMD) simulations (totaling ∼2.7 ns) with various density functional approximations (Becke-Lee-Yang-Parr (BLYP), BLYP-D3, Hamprecht-Cohen-Tozer-Handy, B3LYP) and temperatures (300-330 K), equilibrium and dynamical properties of one excess proton and 128 water molecules are studied. Two features in particular (concerted hops and weak hydrogen-bond donors) are investigated to identify modes in the system that are strongly correlated with the onset of periods of burst-and-rest dynamics. The question of concerted hops seeks to identify those time scales over which long-range proton transport can be classified as a series of sequential water hopping events or as a near-simultaneous concerted process along compressed water wires. The coupling of the observed burst-and-rest dynamics with motions of a fourth neighboring water molecule (a weak hydrogen-bond donor) solvating the protonated water molecule is also investigated. The presence (absence) of hydrogen bonds involving this fourth water molecule before and after successful proton hopping events is found to be strongly correlated with periods of burst (rest) dynamics (and consistent with pre-solvation concepts). By analyzing several realizations of the AIMD trajectories on the 100-ps time scale, convergence of statistics can be assessed. For instance, it was observed that the probability for a fourth water molecule to approach the hydronium, if not already proximal at the beginning of the lifetime of the hydronium, is very low, indicative of the formation of stable void regions. Furthermore, the correlations of the neighboring water atoms are identified as the fourth water approaches the hydronium. Finally, the temperature effects on structural and dynamical properties are studied.

Effects of operating parameters and fluid properties on the efficiency of a new vacuum evaporation method

A new process for vacuum evaporation was developed where evaporation takes place near the inner surface of a vortex as produced by a rotor submerged in the liquid. Contrary to the state of the art the new process does not need a vacuum vessel but the rotating liquid creates a geometrically stable low pressure void surrounded by a vortex stabilized by the equilibrium between centrifugal forces and the pressure difference. First tests with water and sugar solutions at concentrations similar to wine must showed evaporation rates in the upper range of thin-film evaporators. A test series was conducted to study the effect of the variation of process parameters. The heating power and thus the fluid temperature has the most important influence on the vaporisation rate. A second test series using sucrose solution of different concentration comes to the conclusion that this method is suitable for aqueous solutions but the vapour production rate drops significantly with increased sugar content using the current rotor design. The simplicity of the construction and the process handling make this new method a promising development for the wine production.

Two-nanometer voids in single-layer hexagonal boron nitride: formation via the "can-opener" effect and annihilation by self-healing.

The exposure of hexagonal boron nitride single layers to low energy ions leads to the formation of vacancy defects that are mobile at elevated temperatures. For the case of h-BN on rhodium, a superhoneycomb surface with 3 nm lattice constant (nanomesh), a concerted self-assembly of these defects is observed, where the "can-opener" effect leads to the cut-out of 2 nm "lids" and stable voids in the h-BN layer. These clean-cut voids repel each other, which enables the formation of arrays with a nearest neighbor distance down to about 8 nm. The density of voids depends on the Ar ion dose, and can reach 10(12) cm(-2). If the structures are annealed above 1000 K, the voids disappear and pristine h-BN nanomesh with larger holes is recovered. The results are obtained by scanning tunneling microscopy and density functional theory calculations.

Stability of Porous Platinum Nanoparticles: Combined In Situ TEM and Theoretical Study.

Porous platinum nanoparticles provide a route for the development of catalysts that use less platinum without sacrificing catalytic performance. Here, we examine porous platinum nanoparticles using a combination of in situ transmission electron microscopy and calculations based on a first-principles-parametrized thermodynamic model. Our experimental observations show that the initially irregular morphologies of the as-sythesized porous nanoparticles undergo changes at high temperatures to morphologies having faceted external surfaces with voids present in the interior of the particles. The increasing size of stable voids with increasing temperature, as predicted by the theoretical calculations, shows excellent agreement with the experimental findings. The results indicate that hollow-structured nanoparticles with an appropriate void-to-total-volume ratio can be stable at high temperatures.

Sedimentation Behavior of Four Dredged Slurries in China

Sedimentation model tests were carried out on four clays in China for investigating the sedimentation behavior of slurries. The sedimentary process can be divided into fluid state and consolidation state. The consolidation state of slurries can be divided into Non-Terzahi soil and Terzaghi soil. The void ratio responsible for the intersect point of fluid state and consolidation state is determined as 8.6 times the void ratio at liquid limit based on the difference of settling type which can be divide into Consolidation settling and Zone settling with the initial void ratio increase. According to the relationship between the stable void ratio (at the end of consolidation) and the initial void ratio, the suggested value of void ratio at the transition point from Non-Terzaghi soil to Terzaghi soil is 2.8 times its value at liquid limit.

A direct observation of nanometer-size void dynamics in an ultra-thin water film

Here we report the direct dynamic observation of electron beam induced nucleation of nanometer-diameter voids in a thin water film (<12 nm) on a hydrophilic substrate using an in situ TEM platform tailored towards imaging of liquids. The growth dynamics of these nano-voids in a thin water film reveal that there is a critical diameter below which voids are unstable and close up. However, the coalescence of several of such small and unstable voids allows for pathway to form larger and stable voids. A simple model based on the minimization of free energy explains our experimental observations.

Binder Based on Polyelectrolyte for High Capacity Density Lithium/Sulfur Battery

Lithium/sulfur (Li/S) battery is unique in that it is indeed a liquid electrochemical system. In discharge, elemental sulfur is first reduced into high order polysulfide (PS, Li2Sx, x = 8), which dissolves into liquid electrolyte and leaves voids in the cathode, whereas in the last stage, the dissolved low order PS (Li2Sx, x = 4) is reduced into insoluble Li2S2 and Li2S, which fill the voids up. In order to facilitate reversible cycling, the sulfur cathode is required to be capable of retaining the stable void structure. For this purpose, we introduce a poly(acrylamide-co-diallyldimethylammonium chloride) (AMAC) as the binder of high sulfur loading cathodes. AMAC is a cationic polyelectrolyte, which is substantially insoluble in organic solvent while being highly soluble in water. This feature enables the cathode to retain the stable void structure formed repeatedly during discharge and charge cycling of the sulfur cathode. However, in the process of aqueous slurry coating, AMAC causes severe pitting corrosion of Al current collector. In this work, we study AMAC binder by comparing it with the conventional poly(ethylene oxide) (PEO) binder, discussing the mechanism of Al pitting corrosion, and finally solving the problem of Al corrosion.

Quantitative microscopic measurement of void distribution in shear bands in Zr66.7Cu33.3metallic glass

We employ an electron phase retrieval technique in the transmission electron microscope to reconstruct the projected thickness maps of metallic glass specimens and measure the void distribution at a microscopic level. We examine an as-spun melt-spun ${\text{Zr}}_{66.7}{\text{Cu}}_{33.3}$ glass and the shear bands formed in this glass from inhomogeneous deformation in tension and compression. Both as-spun and deformed glasses show no variation in projected thickness indicative of voids down to the limit of this medium-resolution technique ($0.32\phantom{\rule{0.28em}{0ex}}\text{nm}$). This demonstrates that the free volume generated in deformation does not condense into stable voids larger than $0.32\phantom{\rule{0.28em}{0ex}}\phantom{\rule{4pt}{0ex}}\text{nm}$ in radius, but is distributed diffusely in shear bands.

Analysis of current-driven motion of morphologically stable voids in metallic thin films: Steady and time-periodic states

We report results of self-consistent numerical simulations of current-induced migration of morphologically stable voids in metallic thin films accounting rigorously for current crowding, surface curvature, and surface diffusional anisotropy effects. In a previous study, we demonstrated that as the morphological stability limit is approached, the migration speed of a morphologically stable steady void deviates substantially from being inversely proportional to the void size. We also derived a scaling relationship for the void migration speed, rescaled properly with a shape factor, as a function of the void size as described by Cho et al. [Appl. Phys. Lett. 85, 2214 (2004)]. In this study, we calculate accurately shape factors for stable steady void morphologies, as well as for stable time-periodic void morphologies with surface waves propagating on the voids. We predict the effects of surface diffusional anisotropy strength on the migration of stable steady voids, as well as the effects of void size on void migration speed for both steady and time-periodic states. The results validate fully our scaling theory for the current-driven migration of morphologically stable voids and establish its universality. This theory provides an enabling tool for better design of interconnects in integrated circuits toward optimal reliability under conditions that render void migration an important source of metallic thin-film failure.