International Journal of Mechanics and Materials in Design最新文献

This research extensively used different progressive machine learning (ML) techniques to predict the compressive strength (CS) and tensile strain (TSt) of engineered cementitious composites (ECC) with 14 input variables and six algorithms. Specifically, random forest (RF), support vector machine, extreme gradient boosting (XGBoost), light gradient boosting machine, categorical gradient boosting (CatBoost), and natural gradient boosting techniques were used in the present study, to understand mechanical properties of ECC meanwhile these properties are crucial for design codes and developing new reliable models for mixtures. The discrepancy between the ML technique and specific ECC expected outputs is novel in this study and will aid researchers in better understanding of ECC features. To estimate the CS and TSt of the ECC, 2535 and 1469 input data points, respectively, were incorporated based on the material ratio, W/B, and different properties of the fibers. In addition, hyperparameter optimization techniques have also been used in ML to improve over fitting and make the model more accurate and robust. Moreover, an error analysis was highlighted between the actual and predicted CS and TSt of the ECC with each ML technique. Also, the significance and influence of the variable inputs that affect the CS and TSt were explained using the Shapley additive explanation (SHAP) approach. Among all approaches, CatBoost and XGBoost predicted the CS and TSt of ECC with greater accuracy than other techniques in terms of the coefficient of determination (R²), mean square error, mean absolute error, root mean square error, and symmetric mean absolute percentage error. The training and testing R² values of CatBoost and XGBoost for predicting the CS and TSt of ECC were 0.96, 0.89, 0.89, and 0.76, respectively. SHAP analysis revealed that W/B and fiber elongation were the most significant input variables for the CS and TSt of the ECC.

Graphical Abstract

An approach to estimating the dynamic characteristic of one-dimensional orthorhombic quasicrystal sector plates with various boundary conditions is presented. These investigated structures have an arbitrary angle span, and the constituent materials of these models possess a certain porosity distribution. The Green–Naghdi generalized thermoelastic equations have been applied to the sector plates in a thermal environment to obtain an explicit expression, from which the heat fluxes equation can be converted into a state equation. The state-space method is conducted in basic equations of phason and phonon fields to capture governing equations of the structures along the r-direction. Fourier series expansions are utilized to control the span of the sector plates, subject to simply supported boundary conditions along the circumferential direction. Meanwhile, the remaining boundary conditions are treated by differential quadrature technique. The global propagator matrix is proposed to overcome the numerical instabilities caused by high-order frequency and large discrete points. Numerical examples show that each order frequency follows an increasing pattern from small to large with the increase of fixed boundary conditions. The changes in angular span can cause variations in stiffness, leading to alterations in structural stability.

In view of the fact that composite material structures will be affected by multi-source uncertainty factors during use, a new method for predicting the uncertainty of the residual strength of porous laminates in hot and humid environments is proposed based on the PCE method. The influence of different hot and humid environments on the uncertainty of residual strength of porous composite laminates was studied through experimental and numerical methods. The elastic mechanical parameters of composite materials are used as random variables, and the expression of a generalized polynomial chaos expansion model for porous laminates in hot and humid environments is derived; The Hermite polynomial is introduced and the pseudo spectral projection method is used to solve the polynomial coefficients. The probability density and moment function of composite laminates under temperature and humidity changes are calculated. The results show that the output results based on polynomial chaos expansion method are generally consistent with the experimental values, and it can effectively and accurately predict the dispersion of residual strength of composite plates with holes under actual working conditions; Comparing the output results of the polynomial chaos expansion method with the Monte Carlo method, the deviations of the average and standard deviation of the two methods are controlled within 1 and 2.1%, respectively. However the calculation time of the PCE method is only 6% of that of the Monte Carlo method. The proposed polynomial chaos algorithm has the advantages of high efficiency and fast calculation speed in solving the uncertain response problem of composite materials.

Abstract

In this paper, we delve into the evolving landscape of vibration-based structural damage detection (SDD) methodologies, emphasizing the pivotal role civil structures play in society's wellbeing and progress. While the significance of monitoring the resilience, durability, and overall health of these structures remains paramount, the methodology employed is continually evolving. Our focus encompasses not just the transformation brought by the advent of artificial intelligence but also the nuanced challenges and future directions that emerge from this integration. We shed light on the inherent nonlinearities civil engineering structures face, the limitations of current validation metrics, and the conundrums introduced by inverse analysis. Highlighting machine learning's (ML) transformative role, we discuss how techniques such as artificial neural networks and support vector machine's have expanded the SDD's scope. Deep learning's (DL) contributions, especially the innovative capabilities of convolutional neural network in raw data feature extraction, are elaborated upon, juxtaposed with the potential pitfalls, like data overfitting. We propose future avenues for the field, such as blending undamaged real-world data with simulated damage scenarios and a tilt towards unsupervised algorithms. By synthesizing these insights, our review offers an updated perspective on the amalgamation of traditional SDD techniques with ML and DL, underlining their potential in fostering more robust civil infrastructures.

In this paper, we extend our earlier single blade shedding studies by examining the dynamics of multiple blade shedding in a fan disc of an aviation gas turbine engine experimentally using a scaled-down test rig with improved instrumentation and numerically using nonlinear finite element simulations. The newly improved scaled-down rig is designed using dimensional analysis to maintain its dynamic equivalency with a fan disc in a medium size engine. The improved instrumentation includes additional strain gauges, accelerometer, temperature and speed sensors for improved characterisation of the shedding dynamics. High speed photography was also used to capture the time history of the multiply released blades. The shedding experiments were compared with high resolution finite element simulations of a fully bladed fan disc of a realistic gas turbine engine. We took account of blade airfoil, strain rate effects, and multiple contacts between the blades and the containment ring in our finite element simulations. The results of the current investigations reveal that (i) the released multiple blades interact with the trailing blades causing maximum damage to the trailing blades, (ii) large strains develop in the containment ring due to the multiple blade shedding and (iii) the predicted transient response of the finite element simulations of multiple blade interactions are in agreement with the findings of the scaled-down experiments, confirming the validity of our scaled-down test rig as a possible alternative or a compliment to full engine shedding tests.

The smoothed finite element methods (SFEM) have demonstrated their ability to generate more flexible models, offering increased reliability compared to traditional FEM in certain straightforward and idealized situations. To explore the potential of SFEM in complex engineering problems, this paper, for the first time, combining with multiple point constraints to develop a simple and general procedure to study various analysis types of multi-component structures, via (1) the global matrix is constructed by eliminating independent degrees of freedom; (2) the local matrix generated by the SFEM is divided into four kinds of sub-domains, and any entry of the local matrix is assembled to the global matrix depending on the type of sub-domain. By implementing this approach without augmenting the number of equations, the current method excels not only in the analysis of multi-component structures but also outperforms ABAQUS and NASTRAN in terms of effectiveness and efficiency. This superiority has been convincingly demonstrated through several numerical examples, providing strong validation for the proposed method.

The gossamer space structures can be stowed effortlessly because of a lack of out-of-plane stiffness, but structural strength is needed on partial or complete out-gassing to maintain their deployed state. This study demonstrates a novel approach to producing a self-maintaining shape ability of an inflatable cylindrical boom using heat-actuated SMA wires when the inflation gas is vented out from the assembly after complete deployment. Kapton-based and Kapton-SMA-based booms are analyzed numerically for bending stiffness under inflation and no-inflation pressure, followed by experimental validation. At this end, a customized heat test chamber is developed to conduct the required experiments. Furthermore, a parametric study is also performed to find the effect of materials and design parameters on the boom’s stiffness. Before all, the non-linear behavior of double-layered laminated Kapton is found by curve fitting of stretch test data with the optimized different material model parameters to find the best-fitted material model under the hyperelastic materials category. The study helps to find the membrane behavior and rigidization of the inflatable boom in a reversible manner.

The primary focus of traditional topological optimization in continuum structures is addressing stress, compliance, and other relevant factors associated with single-phase materials. However, the optimal design of structural buckling performance has gained increasing attention due to its significant economic loss and safety risk. Furthermore, the versatility, lightweight nature, and adjustability of composite multiple-phase materials offer significant potential for application in various fields. Therefore, this paper presents a novel methodology for optimizing multi-phase materials’ design by concurrently incorporating structural buckling criteria and compliance design. Linear buckling analysis is utilized to determine the critical buckling load of the structure, and a buckling constraint is incorporated into the topology optimization model to regulate its buckling performance. A refined material interpolation model scheme is introduced to enhance the algorithm’s robustness and eliminate pseudo-eigenmode in buckling analysis. The numerical results demonstrate that the final topology optimization design exhibits distinct and discernible boundaries for the topological configurations of multiple-phase materials. Moreover, it is possible to effectively regulate the buckling property while minimizing any compromise on stiffness.

Harvesting electrical energy from mechanical vibrations through piezoelectric-based resonant devices is a suitable form of generating alternative electrical sources for several applications, most dedicated to powering small electronic devices. This technique has attracted considerable attention over the past decades, mainly due to piezoelectric materials’ high electrical charge density. However, the amount of harvestable energy is usually small and sensitive to variabilities in design, manufacturing, operation, and environmental conditions. Hence, it is essential to account for predictable and potentially relevant uncertainties during the design of energy harvesting devices. This work presents strategies for the robust design of resonant piezoelectric energy harvesters, considering the presence of uncertainties in design, manufacturing, and mounting conditions, such as the bonding of the piezoelectric materials and the clamping of the resonant device. The work proposes and discusses strategies for finite element modeling, accounting for adhesive bonding of piezoelectric materials and imperfect clamping; harvestable power output mean value and dispersion estimation with Polynomial Chaos Expansion; and robust optimization using multiobjective optimization techniques. Relevant general conclusions concerning harvesting devices include but are not limited to, devices with shorter resonating beams and larger tip masses tend to present performances that are nominally better but also less robust. Additionally, reducing the effective electrical resistance may improve robustness without significantly losing the mean value performance. Also, through an assessment of the most relevant design variables and uncertain parameters, some aspects that should receive special attention when designing, manufacturing, and mounting these devices are discussed, such as the bonding of piezoelectric patches and the clamping of cantilever beams due to their essential effect on the robustness of the device. It is also shown that including well-selected design variables may mitigate the impact of uncertainties and, thus, improve the robustness of the device.

Based on stress transfer relationship of fiber reinforced composite layer, damping layer and stiffeners, this study presents a novel dynamic analytical model in order to evaluate the dynamic characteristics of a composite damping plate with randomly oriented carbon nanotube reinforced stiffeners. Both an energy method and complex modulus theory are used to derive the vibration equations. Experiments and numerical simulations are adopted for confirming the correctness of the analytical findings. Furthermore, the model is utilized to investigate the impact of structural variables on the dynamic properties, including the modal loss factor and first-order natural frequency.