Surface mount devices (SMDs), such as flip chip packages and ball grid array (BGA) packages are gaining in popularity in microelectronics industry because they provide high density inputs/outputs, better electrical and thermal performance. However, these solder bump interconnections in SMDs are sandwiched between the silicon die and the substrate, which makes them challenging to be inspected. Current non-destructive solder bump inspection techniques like electrical testing, X-ray and acoustic microscopy have some application gaps. New solder bump inspection technique is urgently needed to fill these gaps. Previous work has shown the potential of using a non-contact, non-destructive laser ultrasonics and laser interferometer based inspection system for assessing solder bump qualities. The system uses a pulsed Nd:YAG laser to induce ultrasound in the chip packages and a laser interferometer to measure the transient out-of-plane displacement on the package surface. The quality of the solder bumps can be evaluated by analyzing the out-of-plane displacement. However, there are still some gaps that need to be addressed before the system is ready on the shelf. This dissertation focuses on addressing some of these existing issues. The research work consists of the following: 1) a control interface was developed to integrate all the different modules to achieve automation. 2) a new signal-processing method for analyzing the transient out-of-plane displacement signals without requiring a known-good reference chip was developed. 3) the application scope of the system was expanded to inspect the second level solder bumps in BGA packages. Two types of process-induced defects including poor-wetting and solder bump voids were investigated. Meanwhile, solder bump fatigue caused by cyclic mechanical bending and thermal cycle was also studied using this system. 4) a finite element analysis was performed to study the thermo-mechanical reliability of solder bumps in PBGA package under cyclic thermal loads. The successful completion of the research objectives has led to a laser ultrasound solder bump inspection system prototype with more user-friendliness, higher throughputs, better repeatability and more flexibility, which accelerate the commercialization the system.

One often wishes to understand the probability distribution of stochastic data from experiment or computer simulations. However, where no model is given, practitioners must resort to parametric or non-parametric methods in order to gain information about the underlying distribution. Others have used initially a nonparametric estimator in order to understand the underlying shape of a set of data, and then later returned with a parametric method to locate the peaks. However they are interested in estimating spectra, which may have multiple peaks, where in this work we are interested in approximating the peak position of a single-peak probability distribution. One method of analyzing a distribution of data is by fitting a curve to, or smoothing them. Polynomial regression and least-squares fit are examples of smoothing methods. Initial understanding of the underlying distribution can be obscured depending on the degree of smoothing. Problems such as under and oversmoothing must be addressed in order to determine the shape of the underlying distribution. Furthermore, smoothing of skewed data can give a biased estimation of the peak position. We propose two new approaches for statistical mode estimation based on the assumption that the underlying distribution has only one peak. The first method imposes the global constraint of unimodality locally, by requiring negative curvature over some domain. The second method performs a search that assumes a position of the distribution's peak and requires positive slope to the left, and negative slope to the right. Each approach entails a constrained least-squares fit to the raw cumulative probability distribution. We compare the relative efficiencies [12] of finding the peak location of these two estimators for artificially generated data from known families of distributions Weibull, beta, and gamma. Within each family a parameter controls the skewness or kurtosis, quantifying the shapes of the distributions for comparison. We also compare our methods with other estimators such as the kernel-density estimator, adaptive histogram, and polynomial regression. By comparing the effectiveness of the estimators, we can determine which estimator best locates the peak position. We find that our estimators do not perform better than other known estimators. We also find that our estimators are biased. Overall, an adaptation of kernel estimation proved to be the most efficient. The results for the work done in this thesis will be submitted, in a different form, for publication by D.A. Rabson and J.K. Looper.

<p>The different mapping techniques that are used increases the details on surfaces without increasing the number of polygons. Image Based Sculpting tools in the program Modo and Z-Brush is used to create folds and wrinkles from photographs of actual fabrics instead of trying to create these shapes by modeling them. This method makes it easier to achieve photorealistic renderings and produce as realistic fabric dynamics as possible when they are applied on objects.</p>

Microelectronic interconnections require advanced joining techniques. Direct metal bonding methods, which include thercomsonic and thermocompression bonding, offer remarkable advantages over soldering and adhesives joining. These processes are reviewed in this paper. The progress made in this area is outlined. Some work concerned with the bonding modeling is also presented. This model is based on the joint interface mechanics resulting from compression. Both bump and substrate deformation are taken into account. The improved understanding of the relationship between the deformation and bonding formation may provide more accurate joint evaluation criterion. / Singapore-MIT Alliance (SMA)

Flip-chip underfilling is a technology by which silica-filled epoxy resin is used to fill the micro-cavity between a silicon chip and a substrate, by dispensing the liquid encapsulant at elevated temperatures along the periphery of one or two sides of the chip and then allowing capillary action to draw the material into the gap. Since the chip, underfill material, and substrate solidify together as one unit, thermal stresses on solder joints during the temperature cycling (which are caused by a mismatch in the coefficients of thermal expansion between the silicon chip and the organic substrate) can be redistributed and transferred away from the fragile bump zone to a more strain-tolerant region. Modeling of the flow behaviour of a fluid in the underfill process is the key to this technology. One of the most important drawbacks in the existing models is inadequate treatment of non-Newtonian fluids in the underfill process in the development of both analytical models and numerical models. Another important drawback is the neglect of the presence of solder bumps in the existing analytical models. This thesis describes a study in which a proper viscosity constitutive equation, power-law model, is employed for describing the non-Newtonian fluid behaviour in flip-chip package. Based on this constitutive equation, two analytical models with closed-form solutions for predicting the fluid filling time and fluid flow front position with respect to time were derived. One model is for a setting with two parallel plates as an approximate to flip-chip package, while the other model is for a setting with two parallel plates within which an array of solder bumps are present. Furthermore, a numerical model using a general-purpose finite element package ANSYS was developed to predict the fluid flow map in two dimensions. The superiority of these models to the existing models (primarily those developed at Cornell University in 1997) is confirmed based on the results of the experiments conducted in this study. This thesis also presents a finding of the notion of critical clearance in the design of a flip-chip package through a careful simulation study using the models developed. The flip-chip package design should make the clearance between solder bumps larger than the critical clearance.

Flip-chip underfilling is a technology by which silica-filled epoxy resin is used to fill the micro-cavity between a silicon chip and a substrate, by dispensing the liquid encapsulant at elevated temperatures along the periphery of one or two sides of the chip and then allowing capillary action to draw the material into the gap. Since the chip, underfill material, and substrate solidify together as one unit, thermal stresses on solder joints during the temperature cycling (which are caused by a mismatch in the coefficients of thermal expansion between the silicon chip and the organic substrate) can be redistributed and transferred away from the fragile bump zone to a more strain-tolerant region. Modeling of the flow behaviour of a fluid in the underfill process is the key to this technology. One of the most important drawbacks in the existing models is inadequate treatment of non-Newtonian fluids in the underfill process in the development of both analytical models and numerical models. Another important drawback is the neglect of the presence of solder bumps in the existing analytical models. This thesis describes a study in which a proper viscosity constitutive equation, power-law model, is employed for describing the non-Newtonian fluid behaviour in flip-chip package. Based on this constitutive equation, two analytical models with closed-form solutions for predicting the fluid filling time and fluid flow front position with respect to time were derived. One model is for a setting with two parallel plates as an approximate to flip-chip package, while the other model is for a setting with two parallel plates within which an array of solder bumps are present. Furthermore, a numerical model using a general-purpose finite element package ANSYS was developed to predict the fluid flow map in two dimensions. The superiority of these models to the existing models (primarily those developed at Cornell University in 1997) is confirmed based on the results of the experiments conducted in this study. This thesis also presents a finding of the notion of critical clearance in the design of a flip-chip package through a careful simulation study using the models developed. The flip-chip package design should make the clearance between solder bumps larger than the critical clearance.

This thesis presents further experimentation and modeling for bump-type gas foil bearings used in oil-free turbomachinery. The effect of shaft temperature on the measured structural force response of foil bearings is of importance for reliable high temperature applications. During actual operation with shaft rotation, the bearing structural parameters are coupled to the effects of a hydrodynamic gas film layer, thus determining the overall bearing load performance. A 38.17 mm inner diameter foil bearing, Generation II, is mounted on an affixed non-rotating hollow shaft with an outer diameter of 38.125 mm. A cartridge heater inserted into the shaft provides a controllable heat source. The clearance between the shaft and the foil bearing increases with increasing shaft temperatures (up to 188°C). A static load (ranging from 0 N to 133 N) is applied to the bearing housing, while measuring the resulting bearing displacement, which represents the compliant structure deflection. Static load versus displacement tests render the bearing static structural stiffness. As the shaft temperature increases, the static test results indicate that the bearing structural stiffness decreases by as much as 70% depending on the bearing orientation. A dynamic load test setup includes a rigid shaft support structure and a suspended electromagnetic shaker. Dynamic load (from 13 N to 31 N) test results show that the test foil bearing stiffness increases by as much as 50% with amplitude of dynamic load above a lightly loaded region, nearly doubles with frequency up to 200 Hz, and decreases by a third as shaft temperature increases. A stick slip phenomenon increases the bearing stiffness at higher frequencies for all the amplitudes of dynamic load tested. The test derived equivalent viscous damping is inversely proportional to amplitude of dynamic load, excitation frequency, and shaft temperature. Further, the estimated bearing dry friction coefficient decreases from 0.52 to 0.36 with amplitude of dynamic load and stays nearly constant as shaft temperature increases. Test results identify static and dynamic bearing parameters for increasing shaft temperature. These experimental results provide a benchmark for predictions from analytical models in current development and are essential to establish sound design practices of the compliant bearing structure.

The bump at the end of the railway bridge is a result of differential movement between the bridge deck and the approach embankment. The movement can have the form of a bump or a dip. Either defect in the track geometry can cause significant problems in track performance. The current state of practice was evaluated by conducting a literature review and an industry survey. According to the survey, approximately half of all railway bridges are affected by the bump/dip. The total annual cost for repairing these bridge transition problems is estimated at $26 million. This does not take into account the considerable cost resulting from speed reductions that railroads must place on trains at these locations. In addition to the increased maintenance costs, the bump/dip leads to higher impact loads, uncomfortable rides and possible safety hazards. The track response due to the bump at the end of the bridge was evaluated by creating a 4-D finite element model of the train, track structure and track substructure. The motion of the train model across a bridge/approach transition, with and without a bump/dip, was then simulated using LS-DYNA. It was found that a track modulus differential alone (no bump/dip) at a bridge/approach location leads to impact forces as well as increased ballast and subgrade pressures on the approach. This instigates the formation of a bump or dip in the track. The track response is increased when a bump/dip is present in the track profile. A parametric study looking at the influence of train direction, train speed, bump/dip size, approach embankment soil modulus, approach tie material, bridge tie material, bridge deck type, ballast thickness and approach tie length on the magnitude of impact forces, track deflection, ballast and subgrade pressures was also performed with the model. Finally, a design solution to minimize the bump at the end of the bridge is proposed. The solution involves installing varying length steel bars into a soft subgrade approach embankment. The solution addresses both the settlement and track modulus differential between the bridge and the embankment. A full-scale field test of this prototype solution is underway.

Title from PDF of title page. / Document formatted into pages; contains 93 pages. / Thesis (M.S.)--University of South Florida, 2003. / Includes bibliographical references. / Text (Electronic thesis) in PDF format. / ABSTRACT: One often wishes to understand the probability distribution of stochastic data from experiment or computer simulations. However, where no model is given, practitioners must resort to parametric or non-parametric methods in order to gain information about the underlying distribution. Others have used initially a nonparametric estimator in order to understand the underlying shape of a set of data, and then later returned with a parametric method to locate the peaks. However they are interested in estimating spectra, which may have multiple peaks, where in this work we are interested in approximating the peak position of a single-peak probability distribution. One method of analyzing a distribution of data is by fitting a curve to, or smoothing them. Polynomial regression and least-squares fit are examples of smoothing methods. Initial understanding of the underlying distribution can be obscured depending on the degree of smoothing. / ABSTRACT: Problems such as under and oversmoothing must be addressed in order to determine the shape of the underlying distribution.Furthermore, smoothing of skewed data can give a biased estimation of the peak position. We propose two new approaches for statistical mode estimation based on the assumption that the underlying distribution has only one peak. The first method imposes the global constraint of unimodality locally, by requiring negative curvature over some domain. The second method performs a search that assumes a position of the distribution's peak and requires positive slope to the left, and negative slope to the right. / ABSTRACT: Each approach entails a constrained least-squares fit to the raw cumulative probability distribution.We compare the relative efficiencies [12] of finding the peak location of these two estimators for artificially generated data from known families of distributions Weibull, beta, and gamma. Within each family a parameter controls the skewness or kurtosis, quantifying the shapes of the distributions for comparison. We also compare our methods with other estimators such as the kernel-density estimator, adaptive histogram, and polynomial regression. By comparing the effectiveness of the estimators, we can determine which estimator best locates the peak position. We find that our estimators do not perform better than other known estimators. We also find that our estimators are biased. / ABSTRACT: Overall, an adaptation of kernel estimation proved to be the most efficient.The results for the work done in this thesis will be submitted, in a different form, for publication by D.A. Rabson and J.K. Looper. / System requirements: World Wide Web browser and PDF reader. / Mode of access: World Wide Web.

No description available.