Dissertations / Theses on the topic 'Design-under-test'

Composite materials hold great potential for the replacement of traditional materials in machines utilized on a daily basis. One such example is within an engine block assembly where massive components inherently reduce the efficiency of the system they constitute. By replacing metal elements such as connecting rods, cylinder caps, or a crank shaft with composite alternatives, a significant increase in performance may be achieved with respect to mechanical strength, thermal stability, and durability, while also reducing mass. Exploration of this technology applied to a connecting rod geometry was investigated through a combination of process development, manufacturing, numerical analysis and testing. Process development explored composite material options based on experimental characterization, fabrication, and machining methods. Finite element analysis provided insight into model and data accuracy, as well as a basis for study on a unidirectional composite I-beam geometry. Destructive testing of the composite connecting rods provided data for a strength to weight ratio comparison with the original steel component. The composite connecting rods exhibited weight savings of 15%-17% that of the steel component. The rod made of woven composite material exhibited an elastic modulus of 68.1 Msi in its linear behavior before failure, thereby exhibiting a higher stiffness than the steel rod tested. Although the failure strengths were 25% below the required design load, the calculated strength to weight ratios showed favor for the composite alternatives.

Crack propagation in integrated circuits is a major failure cause of electronic components. In the automotive branch, a highly competitive market, failure is not allowed and reliability can make the difference to gain market shares. For this reason, many efforts and researches were conducted in this field during the last decades. In automotive applications, electronic parts have to challenge heavy life conditions, such as vibrations, wide temperature variations, humidity, oil aggressions, etc. In particular, cyclic thermal variations bring incompatible thermal strains along materials interfaces, and leads to interfacial crack propagation (delamination) between copper leadframes used as conductor and polymeric molding compound used to encapsulate devices. Delamination causes then the cracking of the encapsulant and finally the failure of the electronic component. Predicting the interfacial delamination in electronic packages is key to enhance reliability. As detailed in this work, delamination strength of interfaces involving molding compounds depends on many influences. Its characterization required numerous kinds of test, involving heavy costs and long testing periods. In this work, a low cost test setup is proposed to characterize the molding compound/copper interface regarding the different influences already mentioned. The delamination toughness of the copper/polymeric molding compound is characterized with a classic molding compound of the market with respect to the mode mixity (ratio of loading mode II/mode I), temperature, and load rate. Moisture effect is also investigated. The critical energy release rate for the investigated interface is estimated using the classical numerical-experimental correlation used in fracture mechanics. For the investigated material pair, the interfacial toughness exhibits viscoelastic properties and an own time-temperature superposition principle. Obtained interfacial parameters are defined in Abaqus. The latter is widely benchmarked for problems involving delamination and viscoelastic materials. In the case of cracking between two dissimilar materials, the analytical solution of the stress and strain fields at delamination front depends on a characterisctic reference length. The choice of this refence length, conditioning the the mode mixity estimation, is also addressed. The reference length is experimentally characterized for the different investigated temperature
La propagation des fissures dans les circuits intégrés est l’une des causes principales de défaillance de ces circuits. Sur le marché automobile fortement concurrentiel, les défaillances ne sont pas permises, et la fiabilité demeure un critère pouvant faire la différence pour gagner des parts de marché. Pour cette raison, de nombreux efforts et recherches ont vu le jour durant les dernières décennies. Dans le cas des applications liées à l’automobile, les composants électroniques sont confrontés à de lourdes contraintes, telles que vibrations, importants changements de température, projections d’eau ou de dérivés pétroliers, etc. Les variations de température génèrent en particulier des déformations thermiques souvent incompatibles au niveau des interfaces entre matériaux, ce qui provoque dans les circuits intégrés le décollement des conducteurs cuivrés de l’enrobage polymérique utilisé pour la protection du composant (délaminage). Le délaminage aboutit par la suite à la fissuration propre de l’encapsulant puis à la défaillance du composant. L’amélioration de la fiabilité des composants requiert par conséquent de pouvoir prédire le délaminage aux interfaces cuivre/matériau d’encapsulation. Tel que démontré dans ces travaux, la ténacité interfaciale dépend de plusieurs influences. Une caractérisation complète requiert de nombreux tests, mettant en jeu différents types d’appareillages et d’éprouvettes, ce qui génère de nombreux coûts et un temps d’exécution important. Une méthode de test à faible coût est proposée dans ce document pour la caractérisation des interfaces cuivre/encapsulant dans les différentes conditions évoquées. Cette méthode a été mise en pratique avec succès pour caractériser un matériau d’encapsulation classiquement utilisé dans l’industrie automobile. L’influence de la mixité des modes de rupture, de la température, de la vitesse de chargement ainsi que de l’humidité ont été étudiées. Les techniques habituelles en mécanique de la rupture de corrélation résultats expérimentaux-simulations ont été utilisées pour déterminer le taux de restitution d’énergie critique dans les différents cas. La ténacité de l’interface cuivre/matériau d’encapsulation s’est révélée suivre un comportement viscoélastique, incluant des propriétés propres de superposition temps-température. Les paramètres interfaciaux obtenus ont finalement été définis dans l’outil de simulation MEF Abaqus. Ce dernier a d’ailleurs été l’objet de nombreuses vérifications en ce qui concerne la simulation des problèmes de délaminage des matériaux viscoélastiques. Dans le cas de la fissuration entre deux matériaux hétérogènes, la description analytique des champs de contrainte et de déformation en pointe de fissure fait appel à une longueur de référence. Le choix de cette longueur de référence est abordé dans ce document. Cette dernière est déterminée expérimentalement pour les différentes températures testées

The purpose of this research was to compare the equating performance of various equating procedures for the multidimensional tests. To examine the various equating procedures, simulated data sets were used that were generated based on a multidimensional item response theory (MIRT) framework. Various equating procedures were examined, including both unidimensional and the multidimensional equating procedures based on an IRT framework in addition to traditional equating procedures. Specifically, the performance of the following six equating procedures under the random groups design was compared: (1) unidimensional IRT observed score equating, (2) unidimensional IRT true score equating, (3) full MIRT observed score equating, (4) unidimensionalized MIRT observed score equating, (5) unidimensionalized MIRT true score equating, and (6) equipercentile equating. A total of four factors (test length, sample size, form difficulty differences, and correlations between dimensions) were expected to impact equating performance, and their impacts were investigated by creating two conditions per each factor: long vs. short test, large vs. small sample size, some vs. no form differences, and high vs. low correlation between dimensions. This simulation study over 50 replications yielded several patterns of equating performance of the six procedures across the simulation conditions. The following six findings are notable: (1) the full MIRT procedure provided more accurate equating results (i.e., less degree of error) than other equating procedures especially when the correlation between dimensions was low; (2) the equipercentile procedure was more likely than the IRT methods to yield a larger amount of random error and overall error across all the conditions; (3) equating for multidimensional tests was more accurate when form differences were small, sample size was large, and test length was long; (4) even when multidimensional tests were used (i.e., the unidimensionality assumptions were violated), still the unidimensional IRT procedures were found to yield quite accurate equating results; and (5) whether an equating procedure is an observed or a true score procedure did not seem to yield any differences in equating results. Building upon these findings, some theoretical and practical implications are discussed, and future research directions are suggested to strengthen the generalizability of the current findings. Given that only a handful of studies have been conducted in the MIRT literature, such research is expected to examine the various specific conditions where these findings are likely to be hold, thereby leading to practical guidelines that can be used in various operational testing situations.

The base tGSt method has been proposed by Carballo, Holzer and Murray [5] as a means of determining the strength of standing seam roof systems under gravity loading. The objective of this thesis is to evaluate the accuracy of the base test method. To do this, eleven sets of tests were performed at Virginia Polytechnic Institute & State University (VPl&SU). Each test set consisted of a single span base test from which a failure load was predicted for the corresponding three span confirming test. Results of two test sets recorded in Reference [5] were also used to evaluate the method. A secondary objective of this thesis is to comment on the effects that system components (purlin orientation, clip type, bracing configuration, panel type, insulation and purlin type) have on the strength of the system. Results from proprietary tests conducted at VPI & SU were used in conjunction with the results from this research to accomplish the secondary objective.
Master of Science

Smoothing techniques are designed to improve the accuracy of equating functions. The main purpose of this dissertation was to propose a new statistic (CS) and compare it to existing model selection strategies in selecting smoothing parameters for polynomial loglinear presmoothing (C) and cubic spline postsmoothing (S) for mixed-format tests under a random groups design. For polynomial loglinear presmoothing, CS was compared to seven existing model selection strategies in selecting the C parameters: likelihood ratio chi-square test (G2), Pearson chi-square test (PC), likelihood ratio chi-square difference test (G2diff), Pearson chi-square difference test (PCdiff), Akaike Information Criterion (AIC), Bayesian Information Criterion (BIC), and Consistent Akaike Information Criterion (CAIC). For cubic spline postsmoothing, CS was compared to the ± 1 standard error of equating (± 1 SEE) rule. In this dissertation, both the pseudo-test data, Biology long and short, and Environmental Science long and short, and the simulated data were used to evaluate the performance of the CS statistic and the existing model selection strategies. For both types of data, sample sizes of 500, 1000, 2000, and 3000 were investigated. In addition, No Equating Needed conditions and Equating Needed conditions were investigated for the simulated data. For polynomial loglinear presmoothing, mean absolute difference (MAD), average squared bias (ASB), average squared error (ASE), and mean squared errors (MSE) were computed to evaluate the performance of all model selection strategies based on three sets of criteria: cumulative relative frequency distribution (CRFD), relative frequency distribution (RFD), and the equipercentile equating relationship. For cubic spline postsmoothing, the evaluation of different model selection procedures was only based on the MAD, ASB, ASE, and MSE of equipercentile equating. The main findings based on the pseudo-test data and simulated data were as follows: (1) As sample sizes increased, the average C values increased and the average S values decreased for all model selection strategies. (2) For polynomial loglinear presmoothing, compared to the results without smoothing, all model selection strategies always introduced bias of RFD and significantly reduced the standard errors and mean squared errors of RFD; only AIC reduced the MSE of CRFD and MSE of equipercentile equating across all sample sizes and all test forms; the best CS procedure tended to yield an equivalent or smaller MSE of equipercentile equating than the AIC and G2diff statistics. (3) For cubic spline postsmoothing, both the ± 1 SEE rule and the CS procedure tended to perform reasonably well in reducing the ASE and MSE of equipercentile equating. (4) Among all existing model selection strategies, the ±1 SEE rule in postsmoothing tended to perform better than any of the seven existing model selection strategies in presmoothing in terms of the reduction of random error and total error; (5) pseudo-test data and the simulated data tended to yield similar results. The limitations of the study and possible future research are discussed in the dissertation.

Depuis le développement des composants d'électronique de puissance qui ont permis la fabrication d'onduleurs fiables et efficaces, les entraînements à vitesse variable utilisant des moteurs asynchrones sont devenus de plus en plus populaires. La technique MLI s'est avérée être une méthode très efficace de contrôle de la vitesse de rotation. Cependant, les impulsions de tension, avec des pentes très raides (de l'ordre de quelques kV/µs), ont apporté de nouveaux risques pour le système d'isolation électrique des moteurs. La richesse harmonique de la tension MLI entraînera une surtension significative due à une différence d'impédance entre le câble et le moteur. En effet, la tension observée par certaines parties du système d'isolation peut dépasser la tension d'apparition des décharges partielles (ang. PDIV); ce qui amorcera une activité de décharges partielles localisée. Le système d'isolation des machines basse tension (appelé type I) est basé presque entièrement sur des matériaux polymères qui ne sont pas conçus pour supporter des décharges partielles tout au long de leur vie. En raison de l'utilisation de variateurs de fréquence, l'isolation primaire du fil émaillée est en danger par rapport aux machines alimentées par réseau. En conséquence, c'est souvent le point le plus faible du système d'isolation qui conduira à la panne prématurée d'une machine. Le but de cette thèse est d'étudier et d'analyser le processus de vieillissement du fil émaillé exposé aux différents facteurs et de proposer une méthode permettant de prédire les durées de vie dans des conditions fixées. Cette étude introduit une prédiction basée sur la méthode des plans d'expériences et la distribution statistique de Weibull. Grâce au modèle obtenu avec des tests de vieillissement courts multicontraintes (température, tension, fréquence) il est possible de prédire les résultats de tests significativement plus longs. De plus, la méthodologie proposée permet de prédire la dispersion des essais longs en se basant sur la dispersion des résultats à court terme. Les prédictions sont comparées avec les données expérimentales afin de prouver la précision du modèle
Since the development of power electronic components, which allowed the manufacturing of reliable and efficient inverters, variable speed drives using inductive motors have become more and more popular. The PWM technique has proven to be a very effective method of rotational speed control. However, the fast changing voltage pulses, with very steep slopes (in the order of a few kV/µs), has brought new hazards for the electrical insulation system of such motors. Very high frequency harmonic components of PWM voltage will result in significant overvoltage due to an impedance mismatch between the cable and the motor. As an effect, the voltage seen by some parts of the insulation system may exceed the Partial Discharge Inception Voltage (PDIV) stating localized partial discharges activity. The insulation system in low-voltage machines (called type I) is based almost entirely on polymer materials, which are not able to support partial discharge activity throughout their lives. Due to the use of frequency inverters especially the primary insulation of the magnet wire is endangered in comparison with system-powered machines. As a result this is often the weakest link of the insulation system leading to a premature breakdown of the machine. The aim of this thesis is to investigate and analyze the aging process of the enameled wire exposed to different factors and to propose a method allowing to predict their lifespans in given conditions. This study introduces a prediction based on the Design of Experiments method and the statistical Weibull distribution. Thanks to the model obtained with short multi-stress (temperature, voltage, frequency) aging tests, it is possible to predict the results of significantly longer ones. Moreover, the adapted methodology is proposed that allows to predict the scatter of the long tests basing on the short-time results dispersion. The predictions are compared with the experimental data in order to prove the model accuracy

This research was conducted to advance the knowledge and understanding of the fire performance of light gauge steel frame wall systems through thermal property tests, full-scale fire tests of magnesium oxide board lined walls, 3-D uncoupled and coupled thermal-structural finite element analyses and design of walls with both unstiffened and web-stiffened channel stud sections. It has provided experimental and numerical data and improved finite element strategies and design methods to undertake structural fire design of light gauge steel frame wall systems.

碩士
國立中央大學
土木工程學系
103
Taiwan has very high density of population. Because the wind turbines make noise, so it is not suitable for the construction of offshore wind turbines in the western land. It is necessary to develop the offshore wind turbines at the offshore environment. The offshore wind turbines usually subjected to cyclic lateral loadings induced by wind and sea wave. In this research, designs the equipment and plans a series of experiments on a model aluminum pile to discussion the behavior of the pile under lateral load. In this research, use Zhunan sand to make the soil of the experiment. Three static load tests were conducted. Two of the test was fixed-end and another test was free-end. Two cyclic load tests were conducted and were fixed end. The number of cycles was one thousand and ten thousand. The mending moment measured by the strain gauges on the pile. Use the regression analysis to calculate the function of the pile deflection and soil resistance. Collecting the pile deflection and soil resistance deduce the p-y curves. Calculated the characteristic parameters of coefficient of subgrade reaction with static load and cyclic load. The load-displacement curves of the pile head were discussed the permanent displacement and the pile secant stiffness under cyclic lateral loads.

碩士
國立成功大學
統計學系碩博士班
97
For highly reliable products, traditional accelerated life tests (ALTs) may be difficult to estimate the life time of products since the products are not likely to fail in a reasonable test time. For this reason, accelerated degradation tests (ADTs) take measurements of degradation along experiment and monitor it over time. These measurements provide valuable information for inferring distribution of life time more precisely. Hence, accelerated degradation tests are increasingly used to assess the life time information of highly reliable products today. In addition, for some products, destructive test is needed for obtaining the degradation measurements. For example, the test of a adhesive bond needs to break the product to obtain the strength of the bond. This kind of test which only one meaningful measurement can be taken on each unit is called an accelerated destructive degradation test (ADDT). The lognormal and Weibull distribution are often used to describe the distribution of product characteristics in life and degradation tests. When the distribution is misspecified, the life time quantile, often of interest to the practitioner, may differ significantly between these two distributions. In this study, under a specific ADDT, we give the information of bias and inefficiency under distribution misspecification. And we provide robust plans that will give more unbiased and efficient estimate of life time quantile. Furthermore, if the scale parameters of two distributions satisfy certain specific relationship by previous experience, we use the adjusted profile likelihood of working model to obtain robust test plans. After the implement of the robust plan, the estimate of the life time quantile by using adjusted profile likelihood will be more precise under distribution misspecification than those estimations without using adjusted profile likelihood.

碩士
中原大學
電子工程研究所
105
With the increasing number of embedded memory cores in modern electronic system designs, the cost of memory testing becomes significant. Built-in-self-test (BIST) is an effective approach for memory testing. However, memory BIST for three-dimensional integrated circuits (3D ICs) has not been well studied. Different from 2D SOCs, the testing of 3D ICs consists of both pre-bond testing and post-bond testing. Therefore, extra memory BIST controllers may be required for each layer to reduce the total test application time. In this thesis, we propose a two-stage approach: the first stage performs memory grouping under distance constraints and the second stage performs test scheduling under power constraints. Compared to the previous work, our approach can improve both BIST area cost and total test time simultaneously.

碩士
世新大學
法律學系
92
The study investigates the design patent infringement cases in USA, and to make clear the matter of law and the matter of fact in the design patent infringement lawsuits, as well as to solve the issues of the law and the fact in the suits by the tests, the viewpoint of the tests, the legal rule in these cases. In the investigating process, it has been observed the legal rule and the rationales developed from the design patent infringement cases, have to sustain the test and the forging of the successive cases for period of time, they could be stated as the leading cases. Therefore, this analysis and organization the procedure, the test and the counterclaim to be applied to the design patent infringement cases have been done case by case, which offer the nature, the rationales, the viewpoint and the content of the test, and then confirm the test and the procedure of the design patent infringement. The results indicate that by learning from the experiences of the foreign cases and to fill up the deficiency of the foreign experience according to our patent law system, a generalized procedure and the test for determining the design patent infringement that are suitable in Taiwan. It is expected that the results not only could apply to the practice of the design patent infringement suits to the court in Taiwan, but also that could provide to the governing patent organizations in Taiwan for reference. Hopefully, the results could attract some scholars, experts, the patent attorneys, the judgers and the ministers of the governing patent organizations to do further research.

碩士
國立高雄第一科技大學
機械與自動化工程系碩士班
105
The aim of this paper is to present an integrated procedure for the optimization of dimensions of an on-road bicycle frame under horizontal, vertical and pedaling fatigue test and stiffness simulation. The procedure is composed of uniform design of experiments, explicit dynamics finite element analysis, Kriging interpolation, compromise programming method. The experimental design is used to plan a set of experiments with multiple factors of bike frame size by uniform design. Then, the bicycle frame in each experiment is analyzed by ANSYS/Workbench to obtain the maximum stress and deformation value. Then, Kriging interpolation is applied to construct the surrogate model of permanent deformation, maximum stress and bicycle mass based on the input and output data of experiment simulations. In order to get minimize the mass, maximum stress and permanent deformation of bicycle frame at the same time. First, to compose the each target be a single objective function by compromise programming method with weighting factors. Then, the bicycle frame stiffness simulation and mass is used to find the best weighting factors. Finally, generalized reduced gradient algorithm combine GRG algorithm method applied to find the optimal solution of dimensions of bicycle frame under the goal of minimize the mass, maximum stress and permanent deformation. From result, after performing the optimization procedure presented in this paper, the improvement rate of the horizontal test of maximum von Mises stress is 2.12%, the vertical test of maximum von Mises stress is 6.73%, the pedal test of maximum von Mises stress is 1.97%, the stiffness of maximum deformation is 1.28%, the mass of bicycle is 3.81%. Generally, successful achieve multi object design of optimization and the lightweight and high strength design of the bicycle frame.

The installation of acoustic liners can represent an effective strategy for noise reduction in aero engines. The performance of an acoustic liner as a sound absorbing device is expressed by the acoustic impedance which mainly depends on its geometry and on the peculiar flow conditions acting on its surface. Several parameters are involved in the estimation of the impedance through a non-trivial dependence; this is even more complex when such devices are installed within a high-speed and high-temperature flow field such the one characteristic of the core nozzle of an aeroengine. Within such a scenario the experimental investigation is fundamental in order to assess the liner performance accurately. The most reliable way to determine the acoustic impedance is through a grazing flow rig which allows to experimentally investigate the acoustic liner performances under different flow conditions by using a controlled acoustic signal as excitation. At the Department of Industrial Engineering in Florence, a novel grazing flow rig for testing liners at flow conditions representative of the Low-Pressure Turbine (LPT) exit, i.e. at high speed and high temperature, has been designed and realized. The present work describes the design of the test rig and the preliminary experimental outcomes achieved for assessing the rig performances. The design of all the main components has been addressed following a comprehensive approach in order to meet the acoustic and aerodynamic requirements and by considering the thermo-structural effects as well. The design process has been supported by the use of Finite Element Analyses, the main results and the design solutions achieved are shown. As the rig is supposed to work under high temperature conditions, the sensors to be used for collecting the acoustic data shall withstand the hot flow as well. Hence, an experimental analysis has been carried out in order to compare the main market solutions available. As a result, the most suitable measurement device was identified. Once the design and the construction of the rig was finished, a preliminary set of tests was performed to prove the eligibility of the rig for the designed purpose. The uniformity of the flow field into the test section has been verified through a preliminary measurement of the flow profile. From the acoustic point of view, the reflection coefficient was measured at the duct ends and the results demonstrated very low reflections proving the effectiveness of the designed anechoic terminations. In addition, a first test on an acoustic liner sample was performed, i.e. the transmission loss was measured at cold flow condition. The overall results demonstrate that the rig is properly designed for the desired purposes.

The properties of natural soil are inherently variable and influence design decisions in geotechnical engineering. Apart from the inherent variability of the soil, the variability may arise due to measurement of soil properties in the field or laboratory tests and model errors. These wide ranges of variability in soil are expressed in terms of mean, variance and autocorrelation function using probability/reliability based models. The most common term used in reliability based design is the reliability index, which is a probabilistic measure of assurance of performance of structure. The main objective of the reliability based design is to quantify probability of failure/reliability of a geotechnical system considering variability in the design parameters and associated safety. In foundation design, reliability based design is useful compared to deterministic factor of safety approach. Several design codes of practice recommend the use of limit state design concept based on probabilistic models, and suggest that, development of reliability based design methodologies for practical use are of immense value. The objective of the present study is to propose reliability based design methodologies for pile foundations under static and seismic loads. The work presented in this dissertation is subdivided into two parts, namely design of pile foundations under static vertical and lateral loading; and design of piles under seismic loading, embedded in non-liquefiable and liquefiable soil. The significance of consideration of variability in soil parameters in the design of pile foundation is highlighted. A brief review of literature is presented in Chapter 2 on current pile design methods under vertical, lateral and seismic loads. It also identifies the scope of the work. Chapter 3 discusses the methods of analysis which are subsequently used for the present study. Chapter 4 presents the reliability based design methodology for vertically and laterally loaded piles based on cone penetration test data for cohesive soil. CPT data from Konaseema area in India is used for analysis. Ultimate limit sate and serviceability limit state are considered for reliability based design using CPT data and load displacement curves. Chapter 5 presents the load resistance factor design (LRFD) of vertically and laterally loaded piles based on load test data. Reliability based code calibrated partial factors are determined considering bias in failure criteria, model bias and variability in load and resistance. Chapter 6 illustrates a comprehensive study on the effect of soil spatial variability on response of vertically and laterally loaded pile foundations in undrained clay. Two-dimensional finite difference program, FLAC2D (Itasca 2005) is used to model the soil and pile. The response of pile foundations due to the effect of variance and spatial correlation of undrained shear strength is studied using Monte Carlo simulation. The influence of spatial variability on the propagation and formation of failure near the pile foundation is also examined. Chapter 7 describes reliability based design methodology of piles in non-liquefiable soil. The seismic load on pile foundation is determined from code specified elastic design response spectrum using pseudo-static approach. Variability in seismic load and soil undrained shear strength are incorporated. The effects of soil relative densities, pile diameters, earthquake predominant frequencies and peak acceleration values on the two plausible failure mechanisms; bending and buckling are examined in Chapter 8. The two-dimensional finite difference analysis is used for dynamic analysis. A probabilistic approach is proposed to identify governing failure modes of piles in liquefiable soil in Chapter 9. The variability in the soil parameters namely SPT-N value, friction angle, shear modulus, bulk modulus, permeability and shear strain at 50% of modulus ratio is considered. Monte Carlo simulation is used to determine the probability of failure. A well documented case of the failed pile of Showa Bridge in 1964 Niigata earthquake is considered as case example. Based on the studies reported in this dissertation, it can be concluded that the reliability based design of pile foundations considering variability and spatial correlation of soil enables a rational choice of design loads. The variability in the seismic design load and soil shear strength can quantify the risk involved for pile design in a rational basis. The identification of depth of liquefiable soil layer is found to be most important to identify failure mechanisms of piles in liquefiable soil. Considerations of soil type, earthquake intensity, predominant frequency of earthquake, pile material, variability of soil are also significant.