Search

For efficient reliability analysis of buildings and structures, robust load models are required in stochastic dynamics, which can be estimated in particular from environmental processes, such as earthquakes or wind loads. To determine the response behaviour of a dynamic system under such loads, the power spectral density (PSD) function is a widely used tool for identifying the frequency components and corresponding amplitudes of environmental processes. Since the real data records required for this purpose are often subject to aleatory and epistemic uncertainties, and the PSD estimation process itself can induce further uncertainties, a rigorous quantification of these is essential, as otherwise a highly inaccurate load model could be generated which may yield in misleading simulation results. A system behaviour that is actually catastrophic can thus be shifted into an acceptable range, classifying the system as safe even though it is exposed to a high risk of damage or collapse. To address these issues, alternative loading models are proposed using probabilistic and non-deterministic models, that are able to efficiently account for these uncertainties and to model the loadings accordingly. Various methods are used in the generation of these load models, which are selected in particular according to the characteristic of the data and the number of available records. In case multiple data records are available, reliable statistical information can be extracted from a set of similar PSD functions that differ, for instance, only slightly in shape and peak frequency. Based on these statistics, a PSD function model is derived utilising subjective probabilities to capture the epistemic uncertainties and represent this information effectively. The spectral densities are characterised as random variables instead of employing discrete values, and thus the PSD function itself represents a non-stationary random process comprising a range of possible valid PSD functions for a given data set. If only a limited amount of data records is available, it is not possible to derive such reliable statistical information. Therefore, an interval-based approach is proposed that determines only an upper and lower bound and does not rely on any distribution within these bounds. A set of discrete-valued PSD functions is transformed into an interval-valued PSD function by optimising the weights of pre-derived basis functions from a Radial Basis Function Network such that they compose an upper and lower bound that encompasses the data set. Therefore, a range of possible values and system responses are identified rather than discrete values, which are able to quantify the epistemic uncertainties. When generating such a load model using real data records, the problem can arise that the individual records exhibit a high spectral variance in the frequency domain and therefore differ too much from each other, although they appear to be similar in the time domain. A load model derived from these data may not cover the entire spectral range and is therefore not representative. The data are therefore grouped according to their similarity using the Bhattacharyya distance and k-means algorithm, which may generate two or more load models from the entire data set. These can be applied separately to the structure under investigation, leading to more accurate simulation results. This approach can also be used to estimate the spectral similarity of individual data sets in the frequency domain, which is particularly relevant for the load models mentioned above. If the uncertainties are modelled directly in the time signal, it can be a challenging task to transform them efficiently into the frequency domain. Such a signal may consist only of reliable bounds in which the actual signal lies. A method is presented that can automatically propagate this interval uncertainty through the discrete Fourier transform, obtaining the exact bounds on the Fourier amplitude and an estimate of the PSD function. The method allows such an interval signal to be propagated without making assumptions about the dependence and distribution of the error over the time steps. These novel representations of load models are able to quantify epistemic uncertainties inherent in real data records and induced due to the PSD estimation process. The strengths and advantages of these approaches in practice are demonstrated by means of several numerical examples concentrated in the field of stochastic dynamics., Für eine effiziente Zuverlässigkeitsanalyse von Gebäuden und Strukturen sind robuste Belastungsmodelle in der stochastischen Dynamik erforderlich, die insbesondere aus Umweltprozessen wie Erdbeben oder Windlasten geschätzt werden können. Um das Antwortverhalten eines dynamischen Systems unter solchen Belastungen zu bestimmen, ist die Funktion der Leistungsspektraldichte (PSD) ein weit verbreitetes Werkzeug zur Identifizierung der Frequenzkomponenten und der entsprechenden Amplituden von Umweltprozessen. Da die zu diesem Zweck benötigten realen Datensätze häufig mit aleatorischen und epistemischen Unsicherheiten behaftet sind und der PSD-Schätzprozess selbst weitere Unsicherheiten induzieren kann, ist eine strenge Quantifizierung dieser Unsicherheiten unerlässlich, da andernfalls ein sehr ungenaues Belastungsmodell erzeugt werden könnte, das zu fehlerhaften Simulationsergebnissen führen kann. Ein eigentlich katastrophales Systemverhalten kann so in einen akzeptablen Bereich verschoben werden, so dass das System als sicher eingestuft wird, obwohl es einem hohen Risiko der Beschädigung oder des Zusammenbruchs ausgesetzt ist. Um diese Probleme anzugehen, werden alternative Belastungsmodelle vorgeschlagen, die probabilistische und nicht-deterministische Modelle verwenden, welche in der Lage sind, diese Unsicherheiten effizient zu berücksichtigen und die Belastungen entsprechend zu modellieren. Bei der Erstellung dieser Lastmodelle werden verschiedene Methoden verwendet, die insbesondere nach dem Charakter der Daten und der Anzahl der verfügbaren Datensätze ausgewählt werden. Wenn mehrere Datensätze verfügbar sind, können zuverlässige statistische Informationen aus einer Reihe ähnlicher PSD-Funktionen extrahiert werden, die sich z.B. nur geringfügig in Form und Spitzenfrequenz unterscheiden. Auf der Grundlage dieser Statistiken wird ein Modell der PSD-Funktion abgeleitet, das subjektive Wahrscheinlichkeiten verwendet, um die epistemischen Unsicherheiten zu erfassen und diese Informationen effektiv darzustellen. Die spektralen Leistungsdichten werden als Zufallsvariablen charakterisiert, anstatt diskrete Werte zu verwenden, somit stellt die PSD-Funktion selbst einen nicht-stationären Zufallsprozess dar, der einen Bereich möglicher gültiger PSD-Funktionen für einen gegebenen Datensatz umfasst. Wenn nur eine begrenzte Anzahl von Datensätzen zur Verfügung steht, ist es nicht möglich, solche zuverlässigen statistischen Informationen abzuleiten. Daher wird ein intervallbasierter Ansatz vorgeschlagen, der nur eine obere und untere Grenze bestimmt und sich nicht auf eine Verteilung innerhalb dieser Grenzen stützt. Ein Satz von diskret wertigen PSD-Funktionen wird in eine intervallwertige PSD-Funktion umgewandelt, indem die Gewichte von vorab abgeleiteten Basisfunktionen aus einem Radialbasisfunktionsnetz so optimiert werden, dass sie eine obere und untere Grenze bilden, die den Datensatz umfassen. Damit wird ein Bereich möglicher Werte und Systemreaktionen anstelle diskreter Werte ermittelt, welche in der Lage sind, epistemische Unsicherheiten zu erfassen. Bei der Erstellung eines solchen Lastmodells aus realen Datensätzen kann das Problem auftreten, dass die einzelnen Datensätze eine hohe spektrale Varianz im Frequenzbereich aufweisen und sich daher zu stark voneinander unterscheiden, obwohl sie im Zeitbereich ähnlich erscheinen. Ein aus diesen Daten abgeleitetes Lastmodell deckt möglicherweise nicht den gesamten Spektralbereich ab und ist daher nicht repräsentativ. Die Daten werden daher mit Hilfe der Bhattacharyya-Distanz und des k-means-Algorithmus nach ihrer Ähnlichkeit gruppiert, wodurch zwei oder mehr Belastungsmodelle aus dem gesamten Datensatz erzeugt werden können. Diese können separat auf die zu untersuchende Struktur angewandt werden, was zu genaueren Simulationsergebnissen führt. Dieser Ansatz kann auch zur Schätzung der spektralen Ähnlichkeit einzelner Datensätze im Frequenzbereich verwendet werden, was für die oben genannten Lastmodelle besonders relevant ist. Wenn die Unsicherheiten direkt im Zeitsignal modelliert werden, kann es eine schwierige Aufgabe sein, sie effizient in den Frequenzbereich zu transformieren. Ein solches Signal kann möglicherweise nur aus zuverlässigen Grenzen bestehen, in denen das tatsächliche Signal liegt. Es wird eine Methode vorgestellt, mit der diese Intervallunsicherheit automatisch durch die diskrete Fourier Transformation propagiert werden kann, um die exakten Grenzen der Fourier-Amplitude und der Schätzung der PSD-Funktion zu erhalten. Die Methode ermöglicht es, ein solches Intervallsignal zu propagieren, ohne Annahmen über die Abhängigkeit und Verteilung des Fehlers über die Zeitschritte zu treffen. Diese neuartigen Darstellungen von Lastmodellen sind in der Lage, epistemische Unsicherheiten zu quantifizieren, die in realen Datensätzen enthalten sind und durch den PSD-Schätzprozess induziert werden. Die Stärken und Vorteile dieser Ansätze in der Praxis werden anhand mehrerer numerischer Beispiele aus dem Bereich der stochastischen Dynamik demonstriert.

To analyse the reliability and durability of large complex structures such as high-rise buildings, most realistically, it is advisable to utilize site-specific load characteristics. Such load characteristics can be made available as data records, e.g. representing measured wind or earthquake loads. Due to various circumstances such as measurement errors, equipment failures, or sensor limitations, the data records underlie uncertainties. Since these uncertainties affect the results of the simulation of complex structures, they must be mitigated as much as possible. In this work, the Procrustes analysis, finding similarity transformations between two sets of points in n-dimensional space is used and is extended to uncertainties so that data records can be analysed regarding the uncertainty. To find the best matching of two sets of points the Kabsch algorithm is used. In this manner, a basis is created to simulate and assess the reliability of high-rise buildings under load due to wind and earthquakes.

International audience; This paper deals with the transient response of a non-linear dynamical system with random uncertainties. The non-parametric probabilistic model of random uncertainties recently published and extended to nonlinear dynamical system analysis is used in order to model random uncertainties related to the linear part of the finite element model. The non-linearities are due to restoring forces whose parameters are uncertain and are modeled by the parametric approach. Jayne's maximum entropy principle with the constraints defined by the available information allows the probabilistic model of such random variables to be constructed. Therefore, a non-parametric-parametric formulation is developed in order to model all the sources of uncertainties in such a non-linear dynamical system. Finally, a numerical application for earthquake engineering analysis is proposed concerning a reactor cooling system under seismic loads.

International audience; This paper presents a novel probabilistic model of random uncertainties for dynamical system in the medium-frequency (MF) range. This approach combines a nonparametric probabilistic model of random uncertainties for the reduced matrix models in structural dynamics with a reduced matrix model adapted to the MF range. The theory is presented and the random energy matrix relating to a given MF band, its random trace, and its random eigenvalues are studied. A numerical example is presented allowing convergence properties and stability of random responses With respect to the bandwidth to be analyzed.

International audience; The paper is devoted to the identification of stochastic loads applied to a non-linear dynamical system for which experimental dynamical responses are available. The identification of the stochastic load is performed using a simplified computational non-linear dynamical model containing both model uncertainties and data uncertainties. Uncertainties are taken into account in the context of the probability theory. The stochastic load which has to be identified is modelled by a stationary non-Gaussian stochastic process for which the matrix-valued spectral density function is uncertain and is then modelled by a matrix-valued random function. The parameters to be identified are the mean value of the random matrix-valued spectral density function and its dispersion parameter. The identification problem is formulated as two optimization problems using the computational stochastic model and experimental responses. A validation of the theory proposed is presented in the context of tubes bundles in Pressurized Water Reactors.

International audience; In this paper, we present an energy density field approach for low- and medium-frequency vibroacoustic analysis of complex structures using a statistical computational model. The theory presented is adapted to a structure consisting of an automotive vehicle coupled with its internal cavity. The objective of this paper is to take advantage of the statistical properties of the frequency response functions observed from previous experimental and numerical experiences. In this approach, the matrix-valued frequency response function is expressed in terms of a dimensionless matrix which is estimated using the vibroacoustic computational model and the proposed energy method. Using these dimensionless matrix-valued frequency response function a simplified vibroacoustic model is proposed.

International audience; The focus of this investigation is on the formulation and validation of a novel approach for the inclusion of uncertainty in the modeling of the boundary conditions of linear structures and of the coupling between linear substructures. First, a mean structural dynamic model that includes boundary condition/coupling flexibility is obtained using classical substructuring concepts. The application of the nonparametric stochastic modeling approach to this mean model is next described and thus permits the consideration of both model and parameter uncertainty. Finally, a dedicated identification procedure is proposed to estimate the two parameters of this stochastic model, i.e. the mean boundary condition/coupling flexibility and the overall level of uncertainty.

International audience; A general methodology is presented for the consideration of both parameter and model uncertainty in the determination of the response of geometrically nonlinear structural dynamic systems. The approach is rooted in the availability of reduced order models of these nonlinear Systems with a deterministic basis extracted from a reference model (the mean model). Uncertainty, both from parameters and model, is introduced by randomizing the coefficients of the reduced order model in a manner that guarantees the physical appropriateness of every realization of the reduced order model, i.e. while maintaining the fundamental properties of symmetry and positive definiteness of every such reduced order model. This randomization is achieved not by postulating a specific joint statistical distribution of the reduced order model coefficients but rather by deriving this distribution through the principle of maximization of the entropy constrained to satisfy the necessary symmetry and positive definiteness properties. Several desirable features of this approach are that the uncertainty can be characterized by a single measure of dispersion, affects all coefficients of the reduced order model, and is computationally easily achieved. The reduced order modeling strategy and this stochastic modeling of its coefficients are presented in details and several applications to a beam undergoing large displacement are presented. These applications demonstrate the appropriateness and computational efficiency of the method to the broad class of uncertain geometrically nonlinear dynamic systems.

International audience; This paper deals with the experimental identification and the validation of a non-parametric probabilistic approach allowing model uncertainties and data uncertainties to be taken into account in the numerical model developed to predict low- and medium-frequency dynamics of structures. The analysis is performed for a composite sandwich panel representing a complex dynamical system which is sufficiently simple to be completely described and which exhibits, not only data uncertainties, but above all model uncertainties. The dynamical identification is experimentally performed for eight panels. The experimental frequency response functions are used to identify the non-parametric probabilistic approach of model uncertainties. The prediction of the low- and medium-frequency dynamical responses obtained with the stochastic system is compared with the experimental measurements.

International audience; This paper deals with uncertainties in vibroacoustics of a bounded cavity whose wall is constituted of a rigid wall and of a deformable part constituted of a Composite Sandwich Panel (CSP). Such a CSP has two thin carbon-resin skins and one high stiffness closed-cell foam core. The objectives of this paper is (1) to study the robustness of acoustic response with respect to the dispersion of the CSP induced by the manufacturing process, (2) to develop a predictive mean mechanical model of the vibroacoustic system and (3) to use a nonparametric probabilistic approach for data and model uncertainties of the CSP in order to analyze the robustness of the mean vibroacoustics model in the LF and MF bands to predict internal acoustic level.

International audience; The aim of this paper is to propose probabilistic methodologies for the dynamic analysis of the mistuning of rotating bladed-disks with cyclic symmetry in the low frequency range. A recent nonparametric probabilistic model of random uncertainties is used for modeling the random uncertainties in the context of the blade mistuning. An inverse probabilistic approach, based on the identification of the dispersion parameters controlling the nonparametric probability model with respect to the blade tolerances is constructed in order to define the blade geometric tolerances yielding a given probability level of the dynamic amplification of the forced response. A numerical application is presented in details.

International audience; The dynamical analysis of complex mechanical systems is in general very sensitive to random uncertainties. In order to treat the latter in a rational way and to increase the robustness of the dynamical predictions, the random uncertainties can be represented by probabilistic models. The structural complexity of the dynamical systems arising in these fields results in large finite element models with significant random uncertainties. Parametric probabilistic models capture the uncertainty in the parameters of the numerical model of the structure, which are often directly related to physical parameters in the actual structure, e.g. Young's modulus. Model uncertainties would have to be modeled separately. On the other hand, the proposed nonparametric model of random uncertainties represents a global probabilistic approach which, in addition, takes directly into account model uncertainty, such as that related to the choice of a particular type of finite element. The uncertain parameters of the structure are not modeled directly by random variables (r.v.'s); instead, the probability model is directly introduced from the generalized matrices of a mean reduced matrix model of the structure by using the maximum entropy principle (Soize 2001). In this formulation the global scatter of each random matrix is controlled by one real positive scalar called dispersion parameter. An example problem from aerospace engineering, specifically the FE model of the scientific satellite INTEGRAL of the European Space Agency (ESA) (Alenia 1998) is used to elucidate the two approaches. First the analysis based on the parametric formulation is carried out; the associated results are then used to calibrate the dispersion parameters and to construct the reduced matrices of the non-parametric model.

International audience; This paper deals with data uncertainties and model uncertainties issues in computational mechanics. If data uncertainties can be modeled by parametric probabilistic methods, for a given mean model, a nonparametric probabilistic approach can be used for modeling model uncertainties. The first part is devoted to random matrix theory for which we summarize previous published results and for which two new ensembles of random matrices useful for the nonparametric models are introduced. In a second part, the nonparametric probabilistic approach of random uncertainties is presented for linear dynamical systems and for nonlinear dynamical systems constituted of a linear part with additional localized nonlinearities. In a third part, a new method is proposed for estimating the parameters of the nonparametric approach from experiments. Finally, examples with experimental comparisons are given.

International audience; The random character of blade mistuning is a motivation to construct probability models of random uncertainties. Recently, a new approach known as a nonparametric model of random uncertainties, based on the entropy optimization principle, was introduced for modeling random uncertainties in linear and nonlinear elastodynamics. This paper presents an extension of this nonparametric model for vibration analysis of structures with cyclic geometry. In particular this probability model allows the blade eigenfrequencies uncertainties and the blade-modal-shape uncertainties to be modeled.

International audience; This paper presents a novel probabilistic model of random uncertainties for complex dynamical system in the medium frequency (MF) range. This approach combines a nonparametric probabilistic model of random uncertainties for the reduced matrix models in structural dynamics with a reduced matrix model method in the MF range. The theory is presented, the random energy matrix relative to a given MF band is studied and a simple numerical example is analyzed.

International audience; This paper deals with the transient response of a nonlinear dynamical system with random uncertainties and subjected to earthquake. The nonparametric probabilistic model of random uncertainties recently published and extended to nonlinear dynamical system analysis is used in order to model random uncertainties related to the linear part of the finite element model. The nonlinearities are due to restoring forces whose parameters are uncertain and are modeled by the parametric approach. Jayne's maximum entropy principle with the constraints defined by the available information allow the probabilistic model of such random variables to be constructed. Therefore, a nonparametric-parametric formulation is developed in order to model all the sources of uncertainties in such a nonlinear dynamical system. Finally, a numerical application for earthquake engineering analysis is proposed and concerned a reactor coolant system under seismic loads.

International audience; Recently, a new approach, called a non-parametric model of random uncertainties, has been introduced for modelling random uncertainties in linear and non-linear elastodynamics in the low-frequency range. This non-parametric approach differs from the parametric methods for random uncertainties modelling and has been developed in introducing a new ensemble of random matrices constituted of symmetric positive-definite real random matrices. This ensemble differs from the Gaussian orthogonal ensemble (GOE) and from the other known ensembles of the random matrix theory. The present paper has three main objectives. The first one is to study the statistics of the random eigenvalues of random matrices belonging to this new ensemble and to compare with the GOE. The second one is to compare this new ensemble of random matrices with the GOE in the context of the non-parametric approach of random uncertainties in structural dynamics for the low-frequency range. The last objective is to give a new validation for the non-parametric model of random uncertainties in structural dynamics in comparing, in the low-frequency range, the dynamical response of a simple system having random uncertainties modelled by the parametric and the non-parametric methods. These three objectives will allow us to conclude about the validity of the different theories.

International audience; This paper presents a new approach, called a nonparametric approach, for constructing a model of random uncertainties in dynamic substructuring in order to predict the matrix-valued frequency response functions of complex structures. Such an approach allows nonhomogeneous uncertainties to be modeled with the nonparametric approach. The Craig-Bampton dynamic substructuring method is used. For each substructure, a nonparametric model of random uncertainties is introduced. This nonparametric model does not require identifying uncertain parameters in the reduced matrix model of each substructure as is usually done for the parametric approach. This nonparametric model of random uncertainties is based on the use of a probability model for symmetric positive-definite real random matrices using the entropy optimization principle. The theory and a numerical example are presented in the context of the finite-element method. The numerical results obtained show the efficiency of the model proposed.

International audience; This paper is devoted to numerical models for prediction of transient dynamical response induced by shocks upon structures with inhomogeneous random uncertainties. The usual numerical methods for analyzing such structures in the low- and medium-frequency ranges employ reduced matrix models based on the use of the elastic modes. The contribution of the higher modes is very sensitive to the modeling and data errors. Here, a recent nonparametric probabilistic method is applied to construct the random matrix model allowing modeling and data errors to be taken into account. The paper presents an extension of the nonparametric method for the case of inhomogeneous uncertainties in the structures. A dynamic substructuring method is used and every substructure gets its own uncertainty level. Experiments have been developed in order to validate the nonparametric probabilistic approach of random uncertainties for such dynamical systems.

International audience; Random matrix theory was intensively studied in the context of nuclear physics. For physical applications, the most important ensemble is the Gaussian Orthogonal Ensemble (GOE) whose elements are real symmetric random matrices with statistically independent entries and are invariant under orthogonal linear transformations. Recently, a new approach, called a nonparametric model of random uncertainties, has been introduced by the author for modeling random uncertainties in vibration analysis. This approach has been developed in introducing a new ensemble of random matrices constituted of symmetric positive-definite real random matrices, called the "positive-definite" ensemble, which differs from the GOE. The first objective of this paper is to compare the GOE with the "positive-definite" ensemble of random matrices in the context of the nonparametric approach of random uncertainties in dynamic systems for the low-frequency range. The second objective of this paper is to give a new validation for the nonparametric model of random uncertainties in dynamic systems in comparing, in the low-frequency range, the dynamical response of a simple system having random uncertainties modeled by the parametric and the nonparametric methods. It is proved that the "positive-definite" ensemble of random matrices, which has been introduced in the context of the development of this nonparametric approach, is well adapted to the low-frequency vibration analysis, while the use of the Gaussian orthogonal ensemble (GOE) is not.

International audience; A new approach is presented for analyzing random uncertainties in dynamical systems. This approach consists of modeling random uncertainties by a nonparametric model allowing transient responses of mechanical systems submitted to impulsive loads to be predicted in the context of linear structural dynamics. The information used dues not require the description of the local parameters of the mechanical model. The probability model is deduced from the use of the entropy optimization principle. whose available information is constituted of the algebraic properties related to the generalized mass, damping, and stiffness matrices which have to he positive-definite symmetric matrices, and the knowledge of these matrices for the mean reduced matrix model. An explicit construction and representation of the probability model have been obtained and are very well suited to algebraic calculus and to Monte Carlo numerical simulation in order to compute the transient responses of structures submitted to impulsive loads. The fundamental properties related to the convergence of the stochastic solution with respect to the dimension of the random reduced matrix model are analyzed. Finally, an example is presented.

International audience; A new nonparametric probabilistic approach is presented for modeling random uncertainties in transient linear elastrodynamics. The information used does not require a description of the local parameters of the mechanical model. The probability model is constructed in the generalized coordinates associated with the elastic eigenmodes. The available information is constituted of the algebraic properties of the generalized mass. damping and stiffness matrices which have to be positive-definite symmetric matrices. and the knowledge of these matrices for the mean reduced matrix model. The convergence of the stochastic solution with respect to the dimension of the random reduced matrix model is analysed.

International audience; Random uncertainties in finite element models in linear structural dynamics are usually modeled by using parametric models. This means that: (1) the uncertain local parameters occurring in the global mass, damping and stiffness matrices of the finite element model have to he identified; (2) appropriate probabilistic models of these uncertain parameters have to be constructed; and (3) functions mapping the domains of uncertain parameters into the global mass, damping and stiffness matrices have to be constructed. In the low-frequency range, a reduced matrix model can then be constructed using the generalized coordinates associated with the structural modes corresponding to the lowest eigenfrequencies. In this paper we propose an approach for constructing a random uncertainties model of the generalized mass, damping and stiffness matrices. This nonparametric model does not require identifying the uncertain local parameters and consequently, obviates construction of functions that map the domains of uncertain local parameters into the generalized mass, damping and stiffness matrices. This nonparametric model of random uncertainties is based on direct construction of a probabilistic model of the generalized mass, damping and stiffness matrices, which uses only the available information constituted of the mean value of the generalized mass, damping and stiffness matrices. This paper describes the explicit construction of the theory of such a nonparametric model.

International audience; This paper concerns uncertainties in structural dynamics for composite sandwich panels constituted of two thin carbon-resin skins and one high stiffness closed-cell foam core. Each skin is constituted of two unidirectional plies [60/-60]. Such light composite sandwich panels, manufactured with a same process, generally present a significant dispersion for their Frequency Response Functions (FRF) in the Low-Frequency (LF) and Medium-Frequency (MF) ranges. The objectives of this paper are (1) to study the dispersion due to the process by using experiments (2) to develop a predictive mean mechanical model based on the use of the laminated composite thin plate theory in dynamics and (3) to use a nonparametric probabilistic approach for data and model uncertainties to improve the predictability of the mean model in the MF dynamics.

This paper deals with the characterization of the blade manufacturing geometric tolerances in order to get a given level of amplification in the forced response of a mistuned bladed-disk. It is devoted to an industrial application in order to validate the theory previously developed [1] and in order to show that this theory is suited to any industrial bladed-disks. It should be noted that the development of an adapted methodology for solving the inverse problem, in order to characterize the manufacturing tolerances, is an important challenge for industries in this area. Let us recall that this theory is based on the use of a nonparametric probabilistic model of random uncertainties in the blade [2]. The dispersion parameters controlling the nonparametric model are estimated as a function of the geometric tolerances. Such an identification is carried out in a computational context by using the numerical Monte Carlo simulation and by using the reduced model method presented in [3]. The industrial application is devoted to the mistuning analysis of a 22 blades wide chord fan stage. Centrifugal stiffening due to rotational effects is also included. The results obtained validate the efficiency and the reliability of the method on three dimensional bladed disks.