A hybrid polynomial dimensional decomposition for uncertainty quantification of high-dimensional complex systems

This paper presents a novel hybrid polynomial dimensional decomposition (PDD) method for stochastic computing in high-dimensional complex systems. When a stochastic response does not possess a strongly additive or a strongly multiplicative structure alone, then the existing additive and multiplicative PDD methods may not provide a sufficiently accurate probabilistic solution of such a system. To circumvent this problem, a new hybrid PDD method was developed that is based on a linear combination of an additive and a multiplicative PDD approximation, a broad range of orthonormal polynomial bases for Fourier-polynomial expansions of component functions, and a dimension-reduction or sampling technique for estimating the expansion coefficients. Two numerical problems involving mathematical functions or uncertain dynamic systems were solved to study how and when a hybrid PDD is more accurate and efficient than the additive or the multiplicative PDD. The results show that the univariate hybrid PDD method is slightly more expensive than the univariate additive or multiplicative PDD approximations, but it yields significantly more accurate stochastic solutions than the latter two methods. Therefore, the univariate truncation of the hybrid PDD is ideally suited to solving stochastic problems that may otherwise mandate expensive bivariate or higher-variate additive or multiplicative PDD approximations. Finally, a coupled acoustic-structural analysis of a pickup truck subjected to 46 random variables was performed, demonstrating the ability of the new method to solve large-scale engineering problems.     

Novel computational methods for high‐dimensional stochastic sensitivity analysis

This paper presents three new computational methods for calculating design sensitivities of statistical moments and reliability of high‐dimensional complex systems subject to random input. The first method represents a novel integration of the polynomial dimensional decomposition (PDD) of a multivariate stochastic response function and score functions. Applied to the statistical moments, the method provides mean‐square convergent analytical expressions of design sensitivities of the first two moments of a stochastic response. The second and third methods, relevant to probability distribution or reliability analysis, exploit two distinct combinations built on PDD: the PDD‐saddlepoint approximation (SPA) or PDD‐SPA method, entailing SPA and score functions; and the PDD‐Monte Carlo simulation (MCS) or PDD‐MCS method, utilizing the embedded MCS of the PDD approximation and score functions. For all three methods developed, the statistical moments or failure probabilities and their design sensitivities are both determined concurrently from a single stochastic analysis or simulation. Numerical examples, including a 100‐dimensional mathematical problem, indicate that the new methods developed provide not only theoretically convergent or accurate design sensitivities, but also computationally efficient solutions. A practical example involving robust design optimization of a three‐hole bracket illustrates the usefulness of the proposed methods. Copyright © 2014 John Wiley & Sons, Ltd.     

A dimensional decomposition method for stochastic fracture mechanics

This paper presents a new dimensional decomposition method for obtaining probabilistic characteristics of crack-driving forces and reliability analysis of general cracked structures subject to random loads, material properties, and crack geometry. The method involves a novel function decomposition permitting lower-variate approximations of a crack-driving force or a performance function, Lagrange interpolations for representing lower-variate component functions, and Monte Carlo simulation. The effort required by the proposed method can be viewed as performing deterministic fracture analyses at selected input defined by sample points. Compared with commonly-used first- and second-order reliability methods, no derivatives of fracture response are required by the new method developed. Results of three numerical examples involving both linear-elastic and nonlinear fracture mechanics of cracked structures indicate that the decomposition method provides accurate and computationally efficient estimates of probability density of the J-integral and probability of fracture initiation for various cases including material gradation characteristics and magnitudes of applied stresses and loads.     

Development of stochastic isogeometric analysis (SIGA) method for uncertainty in shape

In this paper, a new method is proposed that extend the classical deterministic isogeometric analysis (IGA) into a probabilistic analytical framework in order to evaluate the uncertainty in shape and aim to investigate a possible extension of IGA in the field of computational stochastic mechanics. Stochastic IGA (SIGA) method for uncertainty in shape is developed by employing the geometric characteristics of the non-uniform rational basis spline and the probability characteristics of polynomial chaos expansions (PCE). The proposed method can accurately and freely evaluate problems of uncertainty in shape caused by deformation of the structural model. Additionally, we use the intrusive formulation approach to incorporate PCE into the IGA framework, and the C++ programming language to implement this analysis procedure. To verify the validity and applicability of the proposed method, two numerical examples are presented. The validity and accuracy of the results are assessed by comparing them to the results obtained by Monte Carlo simulation based on the IGA algorithm.     

Stochastic sensitivity analysis by dimensional decomposition and score functions

This article presents a new class of computational methods, known as dimensional decomposition methods, for calculating stochastic sensitivities of mechanical systems with respect to probability distribution parameters. These methods involve a hierarchical decomposition of a multivariate response function in terms of variables with increasing dimensions and score functions associated with probability distribution of a random input. The proposed decomposition facilitates univariate and bivariate approximations of stochastic sensitivity measures, lower-dimensional numerical integrations or Lagrange interpolations, and Monte Carlo simulation. Both the probabilistic response and its sensitivities can be estimated from a single stochastic analysis, without requiring performance function gradients. Numerical results indicate that the decomposition methods developed provide accurate and computationally efficient estimates of sensitivities of statistical moments or reliability, including stochastic design of mechanical systems. Future effort includes extending these decomposition methods to account for the performance function parameters in sensitivity analysis.     

Probabilistic fracture mechanics by Galerkin meshless methods – part II: reliability analysis

 This is the second in a series of two papers generated from a study on probabilistic meshless analysis of cracks. In this paper, a stochastic meshless method is presented for probabilistic fracture-mechanics analysis of linear-elastic cracked structures. The method involves an element-free Galerkin method for calculating fracture response characteristics; statistical models of uncertainties in load, material properties, and crack geometry; and the first-order reliability method for predicting probabilistic fracture response and reliability of cracked structures. The sensitivity of fracture parameters with respect to crack size, required for probabilistic analysis, is calculated using a virtual crack extension technique described in the companion paper [1]. Numerical examples based on mode-I and mixed-mode problems are presented to illustrate the proposed method. The results show that the predicted probability of fracture initiation based on the proposed formulation of the sensitivity of fracture parameter is accurate in comparison with the Monte Carlo simulation results. Since all gradients are calculated analytically, reliability analysis of cracks can be performed efficiently using meshless methods. Received 20 February 2001 / Accepted 19 December 2001     

Probabilistic fracture mechanics analysis of linear-elastic cracked structures by spline fictitious boundary element method

The inherent uncertainties in crack geometry, material properties and loadings have large influence on fracture response characteristics of cracked structures. This paper presents the probabilistic fracture mechanics analysis of linear-elastic cracked structures subjected to mixed-mode loading conditions using the spline fictitious boundary element method (SFBEM). The response surface method (RSM) is used to predict the fracture probability of the cracked structure. To determine the unknown coefficients of the response surface function, the SFBEM based on the Erdogan fundamental solutions for infinite cracked plates is adopted to perform deterministic analyses of stress intensity factors (SIFs) corresponding to different test points with given parameters. Numerical examples based on mode-I and mixed-mode crack problems are presented to illustrate the present method. The results show that the predicted failure probability obtained by the present approach is accurate in comparison with the Monte Carlo simulation (MCS) results. Since a much lesser number of numerical tests are required in RSM as compared with that needed in MCS, and since the SFBEM based on the Erdogan fundamental solutions has been used to conduct the numerical tests, reliability analysis of cracked structures can be performed efficiently using the present method.     

A Probabilistic Method to Predict Fatigue Crack Initiation

A probabilistic method to predict macrocrack initiation due to fatigue damage is presented in this paper. Acoustic non-linearity is used to quantify pre-macrocrack initiation damage. This data is then used in a probabilistic analysis of fatigue damage. The probabilistic fatigue damage analysis consists of a suitably chosen damage evolution equation to model accumulated damage coupled with a procedure to calculate the probability of macrocrack initiation. The probability of macrocrack initiation is evaluated using the Monte Carlo Method with Importance Sampling. Numerical results for the probabilistic assessment of fatigue damage for a sample problem are presented and compared with experimental results.     

Decomposition methods for structural reliability analysis

A new class of computational methods, referred to as decomposition methods, has been developed for predicting failure probability of structural and mechanical systems subject to random loads, material properties, and geometry. The methods involve a novel function decomposition that facilitates univariate and bivariate approximations of a general multivariate function, response surface generation of univariate and bivariate functions, and Monte Carlo simulation. Due to a small number of original function evaluations, the proposed methods are very effective, particularly when a response evaluation entails costly finite-element, mesh-free, or other numerical analysis. Seven numerical examples involving elementary mathematical functions and solid-mechanics problems illustrate the methods developed. Results indicate that the proposed methods provide accurate and computationally efficient estimates of probability of failure.     

Prediction of reliability analysis of composite tubular structure under hygro-thermo-mechanical loading

The present paper focuses on reliability prediction of composite structure under hygro-thermo-mechanical loading, conditioned by Tsai-Wu failure criterion, where the Monte–Carlo method is used to estimate the failure probability(Pf). This model was developed in two steps: first, the development of a deterministic model, based on an analytical and numerical approach, and then, a probabilistic computation. Using the hoop stress for each ply, a sensitivity analysis was performed for random design variables, such as materials properties, geometry, manufacturing, and loading, on composite cylindrical structure reliability. The probabilistic results show the very high increase of failure probability when all parameters are considered.     

Stochastic meshless analysis of elastic–plastic cracked structures

This paper presents a stochastic mesh-free method for probabilistic fracture-mechanics analysis of nonlinear cracked structures. The method involves enriched element-free Galerkin formulation for calculating the J-integral; statistical models of uncertainties in load, material properties, and crack geometry; and the first-order reliability method (FORM) for predicting probabilistic fracture response and reliability of cracked structures. The sensitivity of fracture parameters with respect to crack size, required for probabilistic analysis, is calculated using a virtual crack extension technique. Numerical examples based on mode-I fracture problems have been presented to illustrate the proposed method. The results from sensitivity analysis indicate that the maximum difference between sensitivity of the J-integral calculated using the proposed method and reference solutions obtained by the finite-difference method is about three percent. The results from reliability analysis show that the probability of fracture initiation using the proposed sensitivity and meshless-based FORM are very accurate when compared with either the finite-element-based Monte Carlo simulation or finite-element-based FORM. Since all gradients are calculated analytically, the reliability analysis of cracks can be performed efficiently using meshless methods. The authors would like to acknowledge the financial support of the U.S. National Science Foundation (NSF) under Award No. CMS-9900196. The NSF program director was Dr. Ken Chong.     

Optimization-oriented exponential-polynomial-closure approach for analyzing nonlinear stochastic oscillators

A novel method named optimization-oriented exponential-polynomial-closure approach is proposed in this article. The main idea of this attempt is to extend the original exponential-polynomial-closure solution procedure methodologically by minimizing the resulted residual error square of the governing equation, which is achieved after an exponential polynomial is adopted as the approximate solution. The objective function for computing the parameters in the approximate solutions of nonlinear random oscillators is then formulated. The probabilistic solutions of the oscillators obtained by the presented approach are verified by the exact solutions in some special cases or by Monte Carlo simulation. Numerical examples indicate that the solutions attained by the presented approach match with the exact or Monte Carlo simulation solutions. The advantage of the presented solution procedure is that it can provide a much more accurate solution than the Equivalent Linearization approach and it is much more efficient than Monte Carlo simulation as demonstrated by the numerical examples.     

Uncertainty quantification of high‐dimensional complex systems by multiplicative polynomial dimensional decompositions

The central theme of this paper is multiplicative polynomial dimensional decomposition (PDD) methods for solving high‐dimensional stochastic problems. When a stochastic response is dominantly of multiplicative nature, the standard PDD approximation, predicated on additive function decomposition, may not provide sufficiently accurate probabilistic solutions of a complex system. To circumvent this problem, two multiplicative versions of PDD, referred to as factorized PDD and logarithmic PDD, were developed. Both versions involve a hierarchical, multiplicative decomposition of a multivariate function, a broad range of orthonormal polynomial bases for Fourier‐polynomial expansions of component functions, and a dimension‐reduction or sampling technique for estimating the expansion coefficients. Three numerical problems involving mathematical functions or uncertain dynamic systems were solved to corroborate how and when a multiplicative PDD is more efficient or accurate than the additive PDD. The results show that indeed, both the factorized and logarithmic PDD approximations can effectively exploit the hidden multiplicative structure of a stochastic response when it exists. Since a multiplicative PDD recycles the same component functions of the additive PDD, no additional cost is incurred. Finally, the random eigensolutions of a sport utility vehicle comprising 40 random variables were evaluated, demonstrating the ability of the new methods to solve industrial‐scale problems. Copyright © 2013 John Wiley & Sons, Ltd.     

Thermoelectromechanically induced stochastic post buckling response of piezoelectric functionally graded beam

This paper deals with the stochastic post buckling response the functionally graded material (FGMs) beam with surface bonded piezoelectric layers subjected to thermoelectromechanical loadings. A C⁰ nonlinear finite element method using higher order shear deformation theory with von-Karman nonlinearity is used for basic formulation. The random system parameter such as material properties of FGM and piezoelectric layers and thermoelectromechanical loadings are modeled as uncorrelated random input variables. The first and second order perturbation method and Monte Carlo sampling (MCS) are proposed to examine the mean, coefficient of variation, probability distribution function and probability of failure of critical post buckling load. Typical numerical results are presented for volume fraction indexes, slenderness ratios, boundary conditions, piezoelectric layers and thermoelectromechanical loadings with random system properties. The present outlined approach is validated with the results available in the literature and by employing MCS.     

Light-transmittance predictions under multiple-light-scattering conditions. I. Direct problem: hybrid-method approximation

Our aim is to present a method of predicting light transmittances through dense three-dimensional layered media. A hybrid method is introduced as a combination of the four-flux method with coefficients predicted from a Monte Carlo statistical model to take into account the actual three-dimensional geometry of the problem under study. We present the principles of the hybrid method, some exemplifying results of numerical simulations, and their comparison with results obtained from Bouguer-Lambert-Beer law and from Monte Carlo simulations.     

Modeling of fatigue crack propagation using dual boundary element method and Gaussian Monte Carlo method

This paper studies the modeling of fatigue crack propagation on a multiple crack site of a finite plate using deterministic and probabilistic methods. Stress intensity factor has been calculated by the combined deterministic approach of the dual boundary element method (DBEM) and the probabilistic approach of the Gaussian Monte Carlo method. The Gaussian Monte Carlo method has been incorporated to simulate the random process of the fatigue crack propagation. A finite plate of aluminum alloy 2024-T3 with a thickness of 1.6 mm and 14 holes is analyzed and the fatigue life of the plate is predicted by following a linear elastic law of fracture mechanics. The results of fatigue life predicted by DBEM-Monte Carlo method are in good agreement with experimental ones. The same approach is also applied to two other engineering applications of a gear tooth and a bracket.     

Global sensitivity analysis by polynomial dimensional decomposition

This paper presents a polynomial dimensional decomposition (PDD) method for global sensitivity analysis of stochastic systems subject to independent random input following arbitrary probability distributions. The method involves Fourier-polynomial expansions of lower-variate component functions of a stochastic response by measure-consistent orthonormal polynomial bases, analytical formulae for calculating the global sensitivity indices in terms of the expansion coefficients, and dimension-reduction integration for estimating the expansion coefficients. Due to identical dimensional structures of PDD and analysis-of-variance decomposition, the proposed method facilitates simple and direct calculation of the global sensitivity indices. Numerical results of the global sensitivity indices computed for smooth systems reveal significantly higher convergence rates of the PDD approximation than those from existing methods, including polynomial chaos expansion, random balance design, state-dependent parameter, improved Sobol's method, and sampling-based methods. However, for non-smooth functions, the convergence properties of the PDD solution deteriorate to a great extent, warranting further improvements. The computational complexity of the PDD method is polynomial, as opposed to exponential, thereby alleviating the curse of dimensionality to some extent.