Papers by Keyword: Uncertainty Analysis

Abstract: This study analysed the buckling behaviour of thin-cylindrical shells under axial compression, addressing the persistent disparity between theoretical predictions and numerical simulations. The research investigated the influence of key parameters height, Young's modulus, and thickness on the critical buckling load. A Finite Element Analysis (FEA), specifically a Geometrically and Materially Non-Linear Analysis (GMNA), was performed using the software ABAQUS to model the shells. To bridge the gap between simulation and theory, a mathematical model for uncertainty analysis was developed in MATLAB, employing the Monte-Carlo Simulation (MCS) and referencing Rankine's theory. This study introduces a novel analytical framework that integrates Finite Element Analysis (FEA) and uncertainty analysis to resolve discrepancies in buckling predictions for thin cylindrical shells. The model's accuracy was validated with a maximum error of less than 13% compared to existing studies, and the uncertainty analysis demonstrated a robust standard deviation of 0.249 (less than 1%). The findings revealed that thickness is the most influential parameter; a 10% increase in thickness led to a 10.86% increase in the buckling load. Young's modulus had a moderate impact, with a 10% increase causing a 0.28% rise in the buckling load, while height was the least influential, with a 10% increase leading to only a 0.1% increase. This research provides valuable insights into the complexities of predicting critical buckling loads, highlighting the distinct impact of geometric and material properties on the structural behaviour of cylindrical shells.

Abstract: The uncertainty of mechanical system performance is strongly influenced by the properties of system components such as mass, stiffness-damping coefficient, and friction coefficient. Based on computational simulations, the system performance under uncertainty conditions can be estimated. However, the nonlinear dynamic behavior of friction is difficult to simulate in numerical simulations, this research is therefore employed a smooth stick-slip friction force model instead of the Coulomb friction force model. Monte Carlo simulation (MCS) combined with multibody dynamic (MBD) simulation is proposed to evaluate the uncertainty characteristics of the system components and stick-slip friction force between two contacting bodies. Numerical simulations applied the proposed method were performed to consider the effects of uncertainty of friction coefficient on the machining accuracy of a three axes CNC (Computer Numerical Control) machine tool.

Abstract: The presented paper aims at thorough uncertainty analysis of combined experimental and computational approach for determination of equivalent thermal conductivity of highly perforated bricks. The thermal conductivity is determined using experimental guarded hot plate method. However, due to large dimensions of analyzed specimens, the measurement is influenced by experimental errors due to abnormal heat leakage. Therefore, numerical modeling of heat transfer is performed in order to quantify this leakage and the resulting value of thermal conductivity is subsequently corrected. The uncertainty analysis is essential part of presented approach leading to its justification. It involves statistical errors, measurement errors as well as results of merged sensitivity analysis. Within the frame of this paper, highly perforated clay brick without cavity fillings was investigated. The results indicated that accuracy of presented approach is very sufficient and combined standard uncertainty is lower than 10 %.

Abstract: SOCRAT-BN code is developed for the analysis of design and beyond design basis accidents at sodium cooled fast reactors. To simulate the behavior of the coolant in the reactor core heat transfer and friction in rod bundle geometry are required to consider. The article describes the validation of the code SOCRAT-BN on the experiment with fuel rod imitators in the triangular geometry with wire-wound taking into account experiment and some code model uncertainties.

Abstract: In view of robustness of objective function and constraints in robust design, the method of maximum variation analysis is adopted to improve the robust design. In this method, firstly, we analyses the effect of uncertain factors in design variables and design parameters on the objective function and constraints, then calculate maximum variations of objective function and constraints. A two-level optimum mathematical model is constructed by adding the maximum variations to the original constraints. Different solving methods are used to solve the model to study the influence to robustness. As a demonstration, we apply our robust optimization method to an engineering example, the design of a machine tool spindle. The results show that, compared with other methods, this method of HPSO(hybrid particle swarm optimization) algorithm is superior on solving efficiency and solving results, and the constraint robustness and the objective robustness completely satisfy the requirement, revealing that excellent solving method can improve robustness.

Abstract: Steam pressure and resonant frequency of microwave cavity are important measure parameters of microwave wetness measurement system. The measuring precision of pressure and frequency directly affect the accuracy of wetness measurement. This paper takes different pressure and wetness of steam as example to analyze the parameter uncertainty of measurement relationship and deduces standard uncertainty of wetness measurement. The system comprehensive uncertainty is identified, when the parameters measurement, cavity heat expansion, sampling error, sedimentary water film and so on are considered. The result shows that the system uncertainty is less than 0.004%. The system uncertainty introduced by pressure measurement is small and can be neglected, but the system uncertainty from frequency measurement has a great effect. The precisely measurement of resonance frequency is key to ensure the accuracy of the system.

Abstract: The impact of purity of raw material on the accuracy of gas standard is studied. The R142b with purity of 99.6% and N₂ with purity of 99.996% was used to prepare R142b in N₂ gas standard based on gravimetric method. Results showed that if the impurity in R142b raw material was not dectected, it would cause 0.20-0.23% deviation relative to the reference value; if the uncertainty for impurity in R142b reach to 100%, the uncertainty in each step of dilution was amplified for almost 66.9%-144%, seriously reducing the reliability of the gas standard. Conversely, the neglect of 4.85×10^-5 mol/mol Ar in dilutent gas N₂ would neither significantly affect the mole fraction nor its uncertainty value. These results clarified the importance of purity analysis on standard gas mixture preparation.

Abstract: Great accuracy is required in the settlement process of coal trade. In order to improve the accuracy of coal measurement, a set of online belt weigher calibration device based on the method of material superimposed has been developed. The mechanical structure of the whole device is well designed realizing material circulation conveying. At the same time the software and hardware of calibration system are fully designed.Finally the experimental results indicate that the uncertainty of the calibration device for belt weigher is within 2% and the minimum theoretical ratio of belt material superimposed is about 42% which attains the error requirement of grade 0.5.

Abstract: Cyclic characteristic of offshore environmental load may cause the accumulation of foundation settlement and excess pore pressure in the subsoil. This may lead to conditions that jeopardize the structural stability and operability of the wind turbine. This paper discusses the analysis steps of the responses of non-cohesive subsoil under cyclic loads for a gravity base foundation system.

Abstract: Monte Carlo simulation provides a probabilistic method to evaluate the physical behavior of earth dam. Therefore, the behavior could be got in a more realistic manner. Based on the theory, an innovative software program code is developed by combining the Monte Carlo and finite difference methods to predict the performance of earth dams after impounding. In order to assess the efficiency of the method, the case study of earth dam, located at Southeast of China, has been studied in detail. The performance of this dam is predicted and compared with the field monitoring by using the monitoring data. The results shows the robustness of the proposed method.